Handling prevents and reverses cognitive deficits induced by sub-chronic phencyclidine in a model for schizophrenia in rats

Treatments for schizophrenia are not effective in ameliorating cognitive deficits. Therefore, novel therapies are needed to treat cognitive impairments associated with schizophrenia (CIAS), which are modelled in rats through administration of sub-chronic phencyclidine (scPCP). We have previously shown that enrichment via voluntary exercise prevents and reverses impairments in novel object recognition (NOR) in this model. The present study aimed to investigate if handling could prevent delay-induced NOR deficits and prevent and reverse scPCP-induced NOR deficits. Two cohorts of adult female Lister Hooded rats were used. In experiment one, handling (five minutes/day, five days/week for two weeks), took place before scPCP administration (2 mg/kg, i.p. twice-daily for seven days). NOR tests were conducted at two, four, and seven weeks post-handling with a one-minute inter-trial interval (ITI) and at five weeks post-dosing with a six-hour ITI. In experiment two, rats were handled after scPCP administration and tested immediately in the one-minute ITI NOR task and again at two weeks post-handling. In both handling regimens, the scPCP control groups failed to discriminate novelty, conversely the scPCP handled groups significantly discriminated in this task. In the 6 h ITI test, vehicle control and scPCP control failed to discriminate novelty; however, the vehicle handled and scPCP handled groups did significantly discriminate. Handling rats prevented and reversed scPCP-induced deficits and prevented delay-induced NOR deficits. These findings add to evidence that environmental enrichment is a viable treatment for cognitive deficits in rodent tests and models of relevance to schizophrenia, with potential to translate into effective treatments for CIAS.


Introduction
Schizophrenia is a complex psychiatric disorder. While several approved antipsychotics are available to treat positive symptoms, there is an unmet clinical need for effective treatments of cognitive or negative symptoms [1]. As a result, patients with schizophrenia have a significantly impaired quality of life [2] which comes at a considerable cost, not only to patients and carers but also to the economy, with schizophrenia rated as one of the top 25 leading causes of disability worldwide [3]. Cognitive Impairment Associated with Schizophrenia (CIAS), including dysfunction in seven domains of cognition, is therefore a critical target for the development of novel therapeutic strategies. Visual recognition memory is also a target, with evidence of deficits in this domain being widely described both in animal models [4] and in patients [5,6]. Recognition memory is mediated through the hippocampus [7,8] and the prefrontal cortex [9,10] in primates [5] and rats [11,12]. Dysfunction in these areas is often measured via tests of object recognition whereby a rat's preference for novelty can be used to determine whether memory for a previously encountered object has been retained. The Novel Object Recognition (NOR) task is used in the present study to quantify deficits in visual recognition memory both before and after sub-chronic phencyclidine (scPCP) administration, with manipulation of the length of the inter-trial interval (ITI) enabling the testing of short-term visual discrimination (one minute ITI) [13] as well as determining whether handling impacts on the rat's threshold of natural forgetting, as measured using a 6 h ITI. This longer ITI was chosen based on previous work from our lab, including data from McLean et al. [14] whereby socially housed rats could distinguish novelty up to a 4h ITI, but not at a 6h ITI.
We have developed and thoroughly validated a sub-chronic dosing regimen of phencyclidine, an NMDA receptor antagonist, in female rats and have demonstrated reliable cognitive and social behaviour deficits and pathology of relevance to schizophrenia, for certain types of patients, those characterised by cognitive deficits as described by Tamminga and colleagues (see [15,16,4] for reviews on our model). The model is robust and easy to work with as shown by its establishment in other laboratories (e.g. see [17]). It is therefore a very useful model for research into identifying efficacy of new therapeutic strategies for alleviating these deficits in patients (e.g. [18]). We have used this model in combination with the NOR to test the hypothesis that environmental enrichment in the form of handling can reverse CIAS.
Environmental enrichment is very important for rodent welfare and improving husbandry [19]. Enrichment can be categorised into social (e. g. contact with cagemates, exposure to conspecifics and handling by experimenters) and physical (e.g. bedding, toys and access to running wheels) [20,21]. A lack of appropriate enrichment can lead to altered behaviours, for example increased anxiety and poor learning and memory as a result of a lack of nesting materials for mice [22]. There is evidence that good environmental enrichment practices can have a positive effect on laboratory animals, both physically and behaviourally [23], including changes to exploration levels during a number of behavioural tasks [21]. Our previous work has shown that enrichment using voluntary aerobic exercise can improve function in multiple domains including visual recognition memory and executive function in scPCP-treated female rats [18]. In humans, others have found that high levels of physical fitness, as measured by walking endurance (distance travelled in two minutes of walking as fast as possible) has a beneficial positive correlation with all aspects of cognition in healthy young adults (mean age 28.8 [24]) and on age-related differences in executive function and memory (aged 22-36 [25]). This indicates that non-pharmacological interventions have important potential for treating CIAS, most likely as add-on to pharmacotherapy.
Enrichment via the introduction of toys and running wheels into the housing cage has previously been found to improve performance in object recognition tasks in the CA1 knockout mouse model (NMDA receptor knockout in the CA1 region of the hippocampus) [26], while combined access to toys, running wheels and an increased number of cagemates (N=14) in a larger enriched cage has been shown to increase dendritic branching and synaptogenesis in the cortex and hippocampus [27]. A 45% reduction in apoptotic cell death has also been observed in the hippocampus of young male Wistar rats who were weaned into enriched cages at three weeks old [28].
Handling by experimenters is listed as a form of environmental enrichment in rats [19], and it is common practice in many laboratories and research institutes to regularly handle experimental animals. Previous work has looked at the effects of handling on behaviour or neurochemistry in rats at multiple life stages. One week of daily handling (five minutes/day), along with cages enriched with toys, platforms and nesting materials, has been found to reduce levels of anxiety as measured by time spent interacting with stimuli or not moving (freezing) during exposure to cat odour compared to non-handled rats [29]. A lack of neonatal handling (within the first three weeks of life) has been shown to impair hippocampal-dependant memory in adulthood [30,31], as well as higher glucocorticoid levels in response to a 20min restraint stress [32]. A lack of handling in adolescent rats may also impact behaviour, for example when the elevated plus maze was conducted in adult male Sprague-Dawley rats and then repeated after 48 h, the handled group had a significantly lower escape latency (time taken to move from an open arm to a closed arm) during the repeated task than the unhandled group, indicating learning and memory may be worsened by a lack of handling [33]. Decreased anxiety was also observed in male handled albino, brown-hooded and black-hooded rats in adulthood (at both 90d and 180d) [34]. Handling can be linked to social play in rats, as evidenced by ultrasonic vocalisations (USVs). Playful handling by experimenters has been shown to increase the amount of 50 kHz USVs, which are associated with positive affect in both male and female rats and are used as play signals in Long Evans [35], Wistar [36] and Lister Hooded rats [37]. Social isolation has been found to induce schizophrenia-relevant behaviour in mice [38] and rats [39], while patients with schizophrenia are 2.3x more likely to self-report being lonely compared to the general population [40]. Improving the cognitive deficits seen in scPCP rats through handling may provide a basis for encouraging increased social contact in schizophrenia patients or those considered clinically ultra-high risk (criteria for which is described in [41]) with an aim to prevent or reverse CIAS in the clinic.
While it is clear that a lack of environmental enrichment is associated with cognitive deficits in a number of studies [29][30][31]33,34], the present study aims to determine whether a minimal level of handling by experimenters contributes to the cognitive deficits observed in the scPCP rat model. Our hypotheses are therefore two-fold; we predict that a lack of handling will produce deficits in visual recognition memory in scPCP treated rats, and that handling the rats for 5 min/day for 10 days will be sufficient to prevent (experiment one) and reverse (experiment two) these deficits.

Animals
Subjects were 80 (40 in experiment one, and 40 in experiment two) adult female Lister Hooded rats (Charles River Laboratories, UK) aged 9-12 weeks old and weighing 223±10g at the beginning of testing. We use female rats in this paradigm as we have found males to be less sensitive to the deficit induced by PCP [42], and females show more robust performance following increasing inter-trial intervals compared with male rats [43]. Furthermore, we have extensively pharmacologically validated the performance of female rats in a variety of cognitive tests in our laboratory (See review [15]). Group sizes were determined based on Cohen's D power analysis, further details of which can be found in the Supplementary Materials document. Rats were housed in groups of five, separated according to treatment status (scPCP or Vehicle, and handled or non-handled) and kept at a constant temperature of 19-23ᵒC and a relative humidity of 45-65% in individually ventilated two-tier cages (38cm x 59cm x 24cm, GR1800 Double Decker Cage, Tecniplast, UK). Cages were enriched with sizzle nesting materials and cardboard tubes as our standard practice. They were provided food (Special Diet Service, Ltd., Essex, UK) and water ad libitum. Studies were carried out under Project License no. P763B36B8 in accordance with the Animals Scientific Procedures Act (UK, 1986) and approved by the local University AWERB (Animal Welfare Ethical Review Body).

Drug administration
In each experiment, rats were randomly housed upon arrival into home cages numbered 1-8. The cages were then assigned to four treatment groups; cages 1-2 as vehicle control (VehC), caged 3-4 as vehicle handled (VehH), cages 5-6 as scPCP control (PCPC) and cages 7-8 as scPCP handled (PCPH). All underwent an i.p dosing regimen consisting of twice-daily (9am and 4pm) administration of 2mg/kg phencyclidine hydrochloride dissolved in 0.9% saline (Sigma-Aldrich, Gillingham, Dorset, UK) (PCPC and PCPH groups) or 0.9% saline (VehC and VehH groups) for 7 days as described previously by us (used in [44]; and also used by [45] and others, see [4] for review). This was followed by a seven day wash-out period to ensure direct drug effects or drug withdrawal effects were no longer present.

Apparatus
Five identical opaque Plexiglas open topped boxes (52cm L; 52cm W; 31cm H) were used as the testing arena, with the floor divided into a grid. Objects were chosen to be easily distinguishable from each other while being the same general size and shape, and to ensure there was no natural preference for either object over the other [46], with different pairs of objects used for each NOR test. Objects were positioned six cm away from the walls of the box, identified using the grid markings.

Behavioural testing
A timeline of procedures undertaken for experiment one (preventing the deficit) are shown in Fig. 1a while experiment two (reversing the deficit), is detailed in Fig. 1b.

Handling
VehH and PCPH groups received daily handling sessions for 10 days over two weeks (Monday to Friday). Each cage was handled for 30 min, allowing for an initial period of 5 min introducing the experimenter's hand to the cage and interacting with all five rats, followed by approximately five minutes of handling per rat. The period of 10 days, and schedule of five minutes per rat were chosen based on Cloutier et al., 2018 [47] and Costa et al., 2012 [33] respectively. Two experimenters, blinded to treatment, alternated the cages so that the rats became familiar with both people. Rats first became accustomed to the experimenter through gentle touches and being picked up and moved within their home cage. Subsequent days involved the rats being picked up by the body and placed in the arms of the experimenter where they were permitted to roam freely. Finally, from day six to day 10 of the handling regimen, the rats were briefly grasped around the shoulders with hindquarters supported, similar to the technique shown in [48] as used for intraperitoneal injection, although the experimenters were careful to maintain a light hold whereby the rat was able to move freely, and no true restraint was used at any point. This technique was used in addition to previously described handling methods for the final five days of handling. In experiment one (preventing the deficit), handling occurred before drug administration while in experiment two (reversing the deficit), handling began immediately after the drug washout period (see Fig. 1 for timelines). Control rats remained in their home cage and were not handled by experimenters. Cages were changed once weekly by technical staff, involving minimal handling of the rats.

Novel object recognition
Rats were habituated to the NOR testing apparatus by being placed in the boxes in home cage groups and allowed to explore once for 20 min.
During testing, rats were placed individually into the test arena to explore two identical objects (objects A and B) for three minutes (acquisition phase) before being moved into an open Plexiglass box (24 cm W, 44 cm L, 19 cm H) for a 1 min Inter-Trial Interval (ITI) for short term visual discrimination memory testing [42], or to their home cage for a 6 hour ITI (to measure natural forgetting [14]). Rats were then returned to the testing arena for three minutes (requisition phase) where they explored an object identical to one used in the acquisition phase (the familiar object) and a new (novel) object. The side on which the novel object was placed (left or right) was randomised and counterbalanced. After each phase, 70% ethanol cleaning solution was used in an attempt to limit the influence of olfactory cues on the objects or box. The apparatus is further described in Supplementary Figure 1.
Performance was video-recorded for blind-scoring by an experimenter using an online stopwatch [49]. Discrimination Indices (DIs) were calculated using the following equation (N=novel object exploration, F=familiar object exploration, T=total time exploring both novel and familiar objects): The DI value represents the difference in exploration between the novel and familiar objects in the retention phase, with a positive number indicating a preference for the novel object, a larger absolute value implying a greater preference for one object over the other, and a DI of zero therefore indicating no difference in exploration between either object.

Locomotor activity
Locomotor activity was measured by the total number of lines showing all procedures undertaken from arrival through to collection of tissue. Rats were handled for 10 days in a two week period (Monday to Friday), beginning one week after arrival. After handling, they were dosed twice daily for seven days followed by a seven day washout period. At the end of the washout period, the two week post-handling NOR was conducted, followed by a two week gap before the four week post-handling NOR. At five weeks post-handling, rats were tested using a 6 h ITI, then a final NOR was conducted at seven weeks post-handling. Rats were euthanised immediately following the final behavioural test. b. Timeline of experiment two, showing all procedures undertaken from arrival through to collection of tissue. Rats were dosed twice daily for seven days, beginning seven days after arrival. This was followed by a seven day washout period before handling began, consisting of 10 days of handling during a two week period (Monday to Friday). Immediately after this period, the first NOR was conducted followed by another NOR two weeks later. Rats were euthanised at six weeks post-handling.
crossed by the rats while exploring the NOR boxes in both the acquisition and retention phases.

Statistical analysis
Statistical analysis was performed using IBM SPSS Statistics for Windows, version 25.0 (IBM Corp., Armonk, N.Y., USA). Four-way ANOVA was used to assess differences between and within groups across time-points, and three-way repeated measures ANOVAs were used to compare exploration times between and within groups where time was a significant factor. Post hoc 2-way repeated-measures ANOVA were run where interactions of item*treatment*handling were found. Bonferroni correction for multiple comparisons was applied where appropriate. Where time had a significant main effect on total object exploration, pairwise comparisons were undertaken to identify which time points differed. Mixed-model ANOVA was used to compare the exploration of objects between treatment groups. Where time was not found to be a significant factor, data was averaged across time-points.
Both DIs and total exploration times were compared using 2-way ANOVA and, where significant interactions were found, Bonferonnicorrected post hoc independent Student's t-tests were used in order to find differences between each treatment group. The 6 h ITI NOR was analysed using a 2-way ANOVA with exploratory post hoc paired Student's t-tests to compare exploration times within groups, and independent t-tests to compare DIs across groups. All DIs were compared to an expected mean of zero using one-way t-tests, with a significant value indicating a preference for the novel object. Line crossing data was analysed using three-way between subjects ANOVA with post hoc t-tests comparing lines crossed between time points.

Handling rats prior to scPCP administration prevented a deficit in NOR performance
There was a significant main effect of time when comparing exploration of objects in the acquisition (F(2,72) = 65.067, p<0.001) and retention (F(2,72) = 58.115, p<0.001) phases of the task. Exploration was significantly reduced across groups at the four-week and sevenweek time points when compared to the two-week time point (details of these differences can be found in Supplementary Table 1). 3-way ANOVA were therefore conducted for exploration times at each timepoint, and a significant item*treatment*handling interaction was found in the acquisition phase at seven weeks post-handling (F(1,36) = 6.503, p<0.05), as well as in the retention phase at two weeks (F(1,36) = 10.204, p<0.01) and four weeks post-handling (F(1,36) = 5.798, p<0.05). Post-hoc mixed two-way ANOVA with Bonferroni correction (α = 0.01) were then used to identify group differences in item preference. Significant differences between novel and familiar object exploration were found at two weeks post-handling between VehC and PCPC There was no effect of time on DI, and therefore DI values were averaged across all three time-points.

Handling enhances performance in the 6 h ITI NOR test in both the VehH and PCPH groups
A main effect of item was found on exploration times in the 6 h ITI NOR (F(1,35) = 23.791, p<0.001), but no significant item*treatment or item*handling interactions were identified. Exploratory post-hoc testing found no significant differences in exploration of identical objects A and B in the acquisition phase of the test, but both handled groups showed a difference in exploration of the familiar and novel objects in the retention phase (VehH t(9) = 2.841, p<0.05; PCPH t(9) = 3.835, p<0.01).
There were no significant effect of treatment or handling in the DIs of any treatment groups at this time point, nor was there a significant treatment*handling interaction. However, DIs for both handled groups were significantly different from zero (VehH t(9) = 2.774, p<0.05; PCPH t(9) = 4.618, p<0.01). Table 2 shows the total number of lines crossed as well as total exploration by the rat during this NOR task.

Handling can reverse scPCP induced deficits in NOR performance
There was a significant main effect of time when comparing exploration of objects in the acquisition (F(1,35) = 90.478, p<0.001) and retention (F(1,35) = 27.195, p<0.001) phases of the task. Exploration was significantly reduced across groups at the two week time point when Table 1 Total line crossings and exploration times (mean±SEM) for NOR tests. No significant differences were found between total line crossings for each treatment group at each time point (analysed with mixed model ANOVA). Differences were found in total exploration times at four and seven weeks post-handling (*p<0.01 compared to VehC). There was no effect of time on DI, and therefore DI values were averaged across both time-points. 2-way ANOVA found a significant effect of treatment (Veh or PCP; F(1,36) = 8.235, p<0.01) but not of handling, and no treatment*handling interaction was found. DIs in all but the PCPC group were found to be significantly different to zero (VehC t(9) = 6.750, p<0.0001; VehH t(9) = 5.925, p<0.001; PCPC t(9) = 3.520, p<0.01). Table 3 shows the total number of lines crossed by the rat and total object exploration during each NOR task. There was a main effect of time on lines crossed (F(1,36 = 6.033, p<0.05). 2-way ANOVA with post-hoc independent Student's t-tests Bonferroni correction (α = 0.01) found significant differences in total exploration times at two weeks (F (3,36) = 6.366, p<0.01) post-handling, with post-hoc independent Student's t-tests Bonferroni correction (α = 0.01) finding significant differences from vehicle control in both scPCP control (t(18) = 3.276, p<0.01) and scPCP handling (t(18) = 3.873, p<0.01) groups. Exploration times were also significantly different between vehicle handled and scPCP handled rats (t(18) = 2.892, p<0.01).

Discussion
The aims of the present study were to establish whether a lack of handling is necessary to produce cognitive deficits in the scPCP model of schizophrenia, and to determine if our handling regimen would be able to prevent and reverse scPCP induced cognitive deficits in rats, using the NOR test as a measure of visual discrimination memory. NOR deficits have been consistently demonstrated in our rodent scPCP model for schizophrenia [4], and visual discrimination memory tasks in schizophrenia patients, such as the recurring figures task [50], have also identified deficits in this domain, therefore an intervention which is able to prevent or reverse the onset of cognitive deficits in scPCP rats may hold promise as a treatment for CIAS when translated into patient trials. The relationship between handling and scPCP-induced cognitive deficits may also provide insight into the mechanisms underlying CIAS, and the importance of enrichment in patients.
In experiment one (preventing the deficit), we observed a significant difference in performance on the NOR task between scPCP control and each of the other treatment groups at two weeks post-handling, and between scPCP control and scPCP handled at four weeks post-handling (see Fig. 2). DI values were not impacted by time-point, and analysis of averaged DIs across time-points showed that scPCP control rats performed significantly poorer in the task than both vehicle groups as well as the scPCP handled group (Fig. 3). Overall NOR performance appears to be less robust at four weeks and seven weeks post-handling, which is likely due to the significant reduction in overall object exploration at these time-points. The lack of a significant time*item*handling*treatment interaction indicates that the effect of handling on NOR performance may not be affected by time, and this conclusion is also supported by the lack of a time-point effect on DI values. In future studies, it will be advantageous to test separate cohorts at each time-point in order to eliminate the effects of repeated testing, as any conclusions regarding the effect of time-point on exploration of the familiar and novel objects in the current study should be taken with caution.
Total exploration of the objects reduced significantly between the first and second NORs but not between the second and third NORs, indicating that rats habituated to the task after the initial test but that the effect of multiple NOR trials on exploration may level off after two tests (Supplementary Table 1). The reduction in exploration time at four weeks and seven weeks post-handling may have contributed to the lack of statistical significance in the analysis shown in Fig. 2. As DIs are not impacted by this reduction in total exploration, these values may therefore be a more robust measure than exploration times when comparing NOR performance in this study. There were no differences in the number of line crossings between treatment groups within timepoints, however scPCP handled rats showed significantly higher total exploration than vehicle control at weeks 4 and 7 post-handling (Table 1). This unexpected finding warrants further investigation in order to determine why handling appears to have reduced the rate of adaptation to the NOR apparatus in the scPCP treated rats.
Exploratory post-hoc testing found no significant differences in exploration times during the retention phase of the 6 h ITI NOR task (Fig. 4) which is to be expected, as prior work has shown six hours to be the above the threshold for natural forgetting in rats [14]. However, DI values were found to be significantly different from zero in the scPCP handled and vehicle handled groups (Fig. 5). We have therefore observed a preference for novelty in the 6 hr ITI task in the handled groups, but not in the non-handled control groups. These data indicate that handling may extend the threshold of natural forgetting in rats, however the effect was not strong enough to produce significant differences between groups. The 6 h ITI NOR was the third task to be conducted in this cohort, and the effect of repeated testing on overall object exploration may have contributed to the lack of significance shown in Fig. 4, however analysis of DI values indicates that any effect of Table 2 Mean total line crossings±SEM for NOR test at 5 weeks post-handling with a 6 h ITI. No significant differences were found between total line crossings or object exploration for each treatment group. Data were analysed with between subjects ANOVA.  Table 3 Total line crossings and exploration times (mean±SEM) for NOR tests. No significant differences were found between total line crossings for each treatment group at each time point (analysed with mixed models ANOVA). Differences were found in total exploration times at two weeks post-handling (*p<0.01 compared to VehC; #p<0.01 compared to VehH). handling at this task is minor. This data, alongside Mitsadali et al., 2020 [18], is the first to show that enrichment prior to scPCP administration is able to prevent the deficit in visual recognition memory from occurring. Evidence that prevention of cognitive deficits in the scPCP model is possible after enrichment intervention indicates that early intervention therapies in at-risk populations have the potential to prevent the onset of cognitive symptoms, which is much more beneficial on an individual [51] and societal [52] level than treating symptoms as they appear.
In experiment two (reversing the deficit), scPCP control rats performed significantly differently from vehicle control and vehicle handled rats at zero weeks post-handling (Fig. 6), while no differences were found at two weeks post-handling. Exploration was different between the two time-points, but no effect of time was found for DI values and therefore DI values were averaged across time-points. Comparing DIs between groups showed an effect of treatment but not of handling, however the scPCP control group was the only group that did not show a significant preference for the novel object (Fig. 7). These data indicate Fig. 2. The effects of handling on the exploration time(s) of the familiar and novel object during the retention phase of the NOR task at two weeks, four weeks, and seven weeks posthandling. Veh = vehicle; PCP = sub-chronic phencyclidine; C = control (non-handled) and H = handled. The data are expressed as the mean±SEM (N = 10/group). 4-way ANOVA found an effect of time (^^^p<0.0001), and 2-way ANOVA with Bonferroni correction (α=0.01) compared group performance within each timepoint; *p<0.05; **p<0.01; ***p<0.001. Fig. 3. The effects of handling on the average DIs of NOR test performance at two, four and seven weeks post-handling. The data are expressed as the mean ±SEM (N = 10), and are analysed with a 2-way ANOVA and post-hoc Student's t-tests for comparison between groups, and one-sample t-tests comparing each group to an expected mean of zero. ***p<0.001; ****p<0.0001 between group comparison; ####p<0.0001 comparison to zero.  that our handling regimen may be able to reverse scPCP induced cognitive deficits, however this reversal is less pronounced than the prevention protocol described in experiment one.
Total exploration of the objects was reduced significantly between the two NOR tasks (Supplementary Table 2). There were no differences between the number of lines crossed between treatment groups at each time-point, but total exploration at two weeks post-handling was significantly higher in both scPCP groups when compared to vehicle control, and scPCP handled rats also explored significantly more than vehicle handled rats (Table 3). Vehicle exploration was particularly low in this cohort, which is likely to have driven these differences.
Additional studies should be carried out to determine which, if any, other cognitive domains are affected by this enrichment. For example, whether performance in the Attentional Set Shifting Task, a measure of executive function, or in the T-maze spontaneous alternation task, a spatial working memory task, which have previously been found to be impaired in scPCP rodent models ( [53] and [54] respectively), could also be reversed or prevented using this handling regimen. Using environmental enrichment as a complementary treatment along with pharmacological intervention will also be important to understand whether this combination is able to increase the effects and longevity of treatment. It will also be necessary to determine whether the effects seen here are consistent across strains, and whether they are also present in other rodent models of schizophrenia. It has been found that isolation rearing has the ability to impair NOR task performance in rats [14], but it has yet to be determined whether environmental enrichment via handling by the experimenter is able to reverse these deficits. Additionally, developmental models of schizophrenia in rats such as the maternal immune activation (mIA) model have demonstrated that post pubertal environmental enrichment is able to rescue deficits in spatial memory in male Sprague-Dawley rats [55], and additional research into whether these improvements would be found after a handling regimen is still to be undertaken. Further investigation into the effect of handling on other rodent models of CIAS would also allow us to determine whether the mechanisms underlying the effect of handling by the experimenter are shared between models, and to therefore understand whether environmental enrichment has the potential to be explored as an adjunct therapy for patients.
We acknowledge that the lack of ex-vivo mechanistic data is a limitation in this study and, therefore, our future work will focus on identifying the mechanism by which enrichment via handling can prevent and reverse cognitive deficits in the rat scPCP model, in particular investigating the role of neurogenesis. Environmental enrichment has been shown to restore neurogenesis in aged rats [56], and exercise has also been shown to increase levels of brain derived neurotrophic factor (BDNF) in the hippocampus [57,58]. A link has been made between environmental enrichment, NOR performance, and BDNF upregulation in male Wistar rats [59]. This study found that environmental enrichment with a running wheel ('physical enrichment') and without a wheel ('cognitive enrichment') were both able to improve performance on a novel object recognition task, however they only saw upregulation of BDNF in the hippocampus of physically enriched rats. Further investigation is needed into the role of BDNF and neurogenesis in mediating the improvements in cognition that are seen as a result of handling enrichment, in order to provide additional insight into the mechanisms by which enrichment is able to prevent and reverse cognitive deficits in our model and, more widely, how this intervention could be translated into a treatment for CIAS in the future.
A necessary next step will be to understand the relevance of the findings in the current study as they translate to the clinic by establishing which aspects of handling are driving the changes seen here. For example, it will be important to determine whether the lack of scPCP induced cognitive deficits in our handled rats are caused directly by physical contact, and if this could be further manipulated by changing the type and intensity of handling, or introducing techniques such as tickling. Physical manipulation via tickling has been found to improve affect in rats [60], while repeated handling in the current study may have improved mood or reduced anxiety during dosing and behavioural testing simply via increased familiarity with the experimenters. There is evidence that gentle stroking stimuli can activate oxytocin synthesising neurons [61], while oxytocin has been able to improve some CIAS domains in schizophrenia patients (see [62] for review). This link should be further investigated, as an intervention which improves cognitive deficits via increased oxytocin levels would prove to be a promising avenue for translation into use in the clinic. Our results, in conjunction with our previous work into the ability of voluntary exercise to prevent and reverse scPCP-induced NOR deficits [18] demonstrates the significant effect that enrichment can have on cognition and, therefore, how a lack of enrichment can facilitate the development of cognitive impairments in a rodent model of schizophrenia. The impact of our handling regimen on NOR performance was more pronounced in experiment one, indicating that the intervention is more efficacious when used to prevent a deficit than to reverse it. This study provides promising insights into Fig. 6. The effects of handling on the exploration time(s) of the familiar and novel objects during the retention phase of the NOR task at zero and two weeks post-handling. Veh = vehicle; PCP = sub-chronic phencyclidine; C = control (non-handled) and H = handled. The data are expressed as the mean±SEM (N = 10/group). 4-way ANOVA found an effect of time (^^^p<0.0001), and 2-way ANOVA with Bonferroni correction (α=0.01) compared group performance within each timepoint; *p<0.05. Fig. 7. The effects of handling on the average DIs of NOR test performance at zero and two weeks post-handling. The data are expressed as the mean±SEM (N = 10), and are analysed with one-sample t-tests comparing each group to an expected mean of zero. ##p<0.01; ##p<0.001; ####p<0.0001 comparison to zero. possible future non-pharmacological interventions for the prevention of CIAS in ultra-high risk individuals, which may be accessible to a wider range of people, and have the potential to be used in addition to pharmacological treatment. Cognitive impairments are associated with worse clinical and functional outcomes in patients [63], while the use of Cognitive Remediation Therapy (CRT) [64,65] has been successful in improving functionality in patients [66] via their focus on cognitive processes. The success of cognitive therapies in schizophrenia highlights the importance of addressing CIAS, and indicate that translating the effects of handling in scPCP rats into the clinic may hold promise for treatment of CIAS and for improving patient outcomes. Additionally, in the context of the scPCP model and wider pre-clinical cognitive research, these findings indicate that care should be taken to minimise handling by experimenters in order to ensure that cognitive deficits are maintained at appropriate levels.
In summary, enriching rats through a handling regimen prevented, and reversed, scPCP-induced deficits in visual discrimination memory, as measured by NOR task performance, with more robust evidence being found for preventing, rather than reversing, this deficit. This effect was found in short term (1 min ITI) tests, with some indication of an effect in a longer term (6 h ITI) task. Furthermore, we have shown the protective effect of enrichment in rats against cognitive impairments in the NOR task, something which, if able to be translated to CIAS in patients and atrisk individuals, holds some promise for effective early intervention in schizophrenia, as well as preventing further decline after diagnosis. In a preclinical context, these findings highlight the importance of controlling for handling by experimenters when conducting preclinical research in the scPCP model. Minimising differences in standard handling practices between research groups will be a necessary step in obtaining consistent and high quality data and will contribute to the ongoing effort to bridge the gap in translation between preclinical and clinical testing of novel treatments for CIAS.

Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability
Data will be made available on request.

Supplementary materials
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.physbeh.2023.114117.