Enhanced methamphetamine sensitisation in a rat model of the brain-derived neurotrophic factor Val66Met variant: Sex differences and dopamine receptor gene expression

Brain-derived neurotrophic factor (BDNF) and the Val66Met polymorphism may play a role in the development of psychosis and schizophrenia. The aim of this study was to investigate long-term effects of methamphetamine (Meth) on psychosis-like behaviour and dopamine receptor and dopamine transporter gene expression in a novel rat model of the BDNF Val66Met polymorphism. At the end of a 7-day subchronic Meth treatment, female rats with the Met/Met genotype selectively showed locomotor hyperactivity sensitisation to the acute effect of Meth. Male rats showed tolerance to Meth irrespective of Val66Met genotype. Two weeks later, female Met/Met rats showed increased locomotor activity following both saline treatment or a low dose of Meth, a hyperactivity which was not observed in other genotypes or in males. Baseline PPI did not differ between the groups but the disruption of PPI by acute treatment with apomorphine was absent in Meth-pretreated Met/Met rats. Female Met/Met rats selectively showed down-regulation of dopamine D2 receptor gene expression in striatum. Behavioural effects of MK-801 or its locomotor sensitisation by prior Meth pretreatment were not influenced by genotype. These data suggest a selective vulnerability of female Met/Met rats to short-term and long-term effects of Meth, which could model increased vulnerability to psychosis development associated with the BDNF Val66Met polymorphism.


Introduction
Schizophrenia is a devastating mental illness that disrupts thoughts, perceptions and behaviour and is categorised by a range of positive, negative and cognitive symptoms (Saha et al., 2007;American Psychiatric Association, 2013). A range of genetic and environmental factors have been implicated in the development of schizophrenia, including drug abuse (Gururajan et al., 2012;Saha et al., 2022). Methamphetamine (Meth) is one of the most widely used drugs in the world (Chomchai and Chomchai, 2015;Degenhardt et al., 2016). It increases monoamine release, including dopamine (Chu et al., 2008;Fleckenstein et al., 2009;Hedges et al., 2018;Hogarth et al., 2022), in the central nervous system to induce feelings of euphoria and alertness . Longer-term use of Meth may lead to a psychosis similar to paranoid schizophrenia (Gururajan et al., 2012;McKetin, 2018, Wearne andCornish, 2018). However, not all Meth users develop psychosis, with an estimate of 40 percent experiencing psychotic symptoms (Glasner--Edwards and Mooney, 2014;McKetin et al., 2019). This could be due to genetic factors, in addition to drug dose and duration of use (Medhus et al., 2013;McKetin, 2018;Wearne and Cornish, 2018;McKetin et al., 2019).
Brain-derived neurotrophic factor (BDNF) is a protein that plays a central role in neuronal development, survival and differentiation (Notaras and van den Buuse, 2019). In adulthood it is involved in neuronal plasticity, and learning and memory, with deficient levels of BDNF implicated in a range of psychological and pathophysiological conditions (Notaras et al., 2015, Notaras and. Post-mortem studies have shown significantly reduced BDNF protein levels in the prefrontal cortex (Weickert et al., 2003) and hippocampus (Thompson Ray et al., 2011) of individuals with schizophrenia. The Val66Met single nucleotide polymorphism (SNP) of the BDNF gene (rs6265) has been studied extensively in the context of psychiatric disorder susceptibility, given it results in diminished activity-dependent BDNF secretion and aberrant trophic function, and has been suggested as a putative locus of risk for psychosis and affective disorders (Notaras et al., 2015, Notaras and. Genetic association studies have furthermore suggested that Val66Met is associated with vulnerability for Meth use disorder (Iamjan et al., 2015;He et al., 2020).
Dopaminergic dysregulation has been shown to be a key factor in psychosis (Vanderschuren et al., 2000;van den Buuse, 2010;Javitt et al., 2012;Hogarth et al., 2022) and BDNF may impact dopaminergic signalling through eliciting long-term neuronal adaptations and changes in dopamine receptor gene expression (Martin-Iverson et al., 1994;Guillin et al., 2001). Glutamate hypofunction may also play a significant role in psychosis (Belsham, 2001;van den Buuse, 2010;Javitt et al., 2012) with N-methyl-D-aspartate (NMDA) receptor hypoactivity associated with dopamine pathway hyperactivity in a bi-directional manner (McCutcheon et al., 2020). It has previously been shown that depletion of BDNF in rats reduced Meth-induced dopamine release, while not affecting baseline dopamine release (Narita et al., 2003), suggesting BDNF has a stimulatory effect on the release of dopamine. Indeed, infusion of BDNF into the substantia nigra increased dopamine turnover in the striatum (Martin-Iverson et al., 1994). We previously observed that BDNF heterozygous mutant mice showed endogenous sensitisation to the acute action of Meth on locomotor activity and decreased additional sensitisation effects of chronic Meth treatment (Manning et al., 2016). In a proteomic analysis in Val66Met mice, Meth pretreatment induced markedly different long-term changes in protein expression, depending on the genotype (Greening et al., 2021). Deficient activity-dependent BDNF release in carriers of the Met-allele has been linked to enhanced dopamine signalling compared to the Val carriers and has been proposed as a possible mechanism in psychosis development (Notaras et al., 2015;Corrone et al., 2023).
Current understanding of the role BDNF, and more specifically the Val66Met genetic variant, plays in Meth-induced psychosis and schizophrenia remains limited. In previous studies in a humanized transgenic Val66Met mouse model, we found no significant difference in long-term sensitisation of locomotor hyperactivity induced by an acute Meth challenge between Val/Val, Val/Met and Met/Met genotypes (Greening et al., 2021;Corrone et al., 2023) despite extensive genotype differences in Meth-induced changes in the mesocorticolimbic proteome including dopamine signalling markers (Dat, Comt, and Th) (Greening et al., 2021). In the present study we compared the effect of chronic Meth on psychosis-like behaviour in a novel rat model of the Val66Met genetic variant, Val68Met rats (Mercado et al., 2021;Jaehne et al., 2022). In addition to investigating the long-term effects of chronic Meth, behavioural testing was done during treatment as well (Wearne et al., 2015). Moreover, we investigated locomotor hyperactivity as well as disruption of prepulse inhibition (PPI) following acute administration of both dopaminergic agonists and glutamatergic antagonists (van den Buuse, 2010;Schonfeld et al., 2022). Finally, to elucidate mechanisms involved in genotype-dependent behavioural changes, we assessed gene expression of all five dopamine receptors (D1, D2, D3, D4 and D5 (Beaulieu et al., 2015)) and the dopamine transporter (DAT) in the striatum during adulthood.

Animals
Male and female outbred rs6265 (Val68Met) rats with the Met/Met genotype (Mercado et al., 2021) were originally obtained from Dr Caryl Sortwell (Michigan State University, MI, USA). Briefly, these rats carry the Valine (Val) to Methionine (Met) polymorphism (Val68Met) in the rat Bdnf gene (GenBank: NM_001270630; Ensembl: ENSR-NOG00000047466) (Mercado et al., 2021). Met/Met founder rats were mated with Sprague-Dawley controls to generate heterozygous Val/Met rats which were shipped to the La Trobe University Animal Research and Training Facility (LARTF) where they were used to produce offspring of all genotypes (Val/Val, Val/Met, Met/Met) for experimentation.
A total of 129 offspring of all three genotypes and both sexes was used in this study (Table 1). Rats were housed 2-4 per individuallyventilated cage (Tecniplast, Buguggiate, Italy) with ad libitum access to rodent chow and tap water, and were kept in a temperature (21 • C ± 2 • C) and humidity (55% ± 15%) controlled environment on a 12 h light cycle (lights on 0700 h; light inside the cage 12-22 lux). All behavioural testing was conducted between 8am and 4pm. All procedures were compliant with guidelines of the Australian Code of Practice for the Care and Use of Animals for Scientific Purposes set by the National Health and Medical Research Council of Australia and were approved by the La Trobe University Animal Experimentation Ethics Committee.

Drugs
All drugs used were dissolved and diluted in 0.9% saline solutions. Meth was purchased from the National Measurement Institute (Australia) and given either at 1 or 5 mg/kg via intraperitoneal (i.p.) injection or at 0.2 and 0.5 mg/kg via subcutaneous (s.c.) injection (Wearne et al. 2016a(Wearne et al. , 2017. Apomorphine and MK-801 were purchased from Sigma-Aldrich (Australia) and given at 0.3 mg/kg and 0.05 mg/kg s.c., respectively (Adams et al., 2008;Gururajan et al., 2015).

Meth treatment and behavioural testing
Starting at 6 weeks of age, the rats were pretreated for one week with daily injection of either i.p. saline or Meth. On days 1 and 7 of pretreatment, an i.p. dose of 1 mg/kg was given, while on days 2-6 a 5 mg/ kg dose was administered (Wearne et al., 2016a(Wearne et al., ,b, 2017. On days 1 and 7 rats also underwent locomotor hyperactivity testing. Following a two-week withdrawal period, rats were tested for changes in PPI following injection of saline, apomorphine (0.3 mg/kg) or MK-801 (0.05 mg/kg) s.c. in a pseudo-randomised order. Following completion of these sessions, rats were given either saline, MK-801 (0.05 mg/kg) or Meth (0.2 mg/kg and 0.5 mg/kg) s.c. and were tested for changes in locomotor hyperactivity after each dose. Behavioural tests were done with 3-4 days intervals, to allow for washout of acute drug treatments ( Supplementary Fig. 1). Following testing of an initial cohort of rats, response to 0.5 mg/kg acute Meth was seen to be particularly high in females, therefore, testing the effect of a lower acute dose (0.2 mg/kg) was added to later cohorts to ensure a 'ceiling' effect did not occur. As a result, in the combined data set, some rats did not undergo testing at this lower dose. Three to four days was allowed between all PPI and locomotor hyperactivity testing sessions to allow for drug washout. The age of the rats during treatment and testing coincided with late adolescence (6 weeks) and early adulthood (9-12 weeks) in humans respectively (Hill et al., 2012). Although body weight gain was slightly slower during the 7-day treatment period, there were no differences between the genotypes in body weight over the course of the experiments (Supplementary Fig. 1).

Locomotor hyperactivity
Drug-induced locomotor hyperactivity may be used as a proxy Table 1 Number of rats per group. measure for psychosis-like behaviour, as both psychosis and hyperlocomotion are associated with increased dopamine transmission (van den Buuse, 2010; Pratt et al., 2012;Hogarth et al., 2022). To examine locomotor hyperactivity, rats were placed in a 43 × 43 cm automated photocell chamber (Med Associates Inc., USA), where they could move freely. Based on patterns of behaviour in pilot experiments, locomotor hyperactivity testing included a 15-min habituation period followed by 60 min of observation post drug administration.

Prepulse inhibition
PPI is a measure of sensorimotor gating, which is a pre-attentional process that filters out irrelevant stimuli from the environment and prevents an overload of unnecessary information (van den Buuse, 2010; Swerdlow et al., 2016). Overall, reduced PPI is thought to reflect sensorimotor dysfunction (Shoji and Miyakawa, 2018), similar to information processing deficits involved in a range of disorders, including schizophrenia (van den Buuse, 2010; Swerdlow et al., 2016). To measure PPI, rats were placed in a plexiglass cylinder, 9 cm in diameter, which was fitted with a movement sensor mounted underneath to detect any spontaneous movements and which was placed in a sound-attenuating startle chamber (SR-Lab, San Diego, USA). Startle stimuli were 40ms white noise bursts, at 115 dB. Prepulse stimuli were 2, 4, 8 or 16 dB over the 65 dB background noise, and were delivered at 30 or 100ms prior to the startle stimulus. The PPI session comprised 104 trials with variable inter-trial intervals (8-27 s). Of the 104 trials, 32 were pulse-alone (or startle) trials (four blocks of eight 115 dB); 8 were no-stimulus trials to check for non-specific movements, and the remaining 64 were prepulse-pulse trials. %PPI was calculated by [(pulse-alone trials startle response amplitudeprepulse-pulse trials startle amplitude)/(pulse-alone trials startle amplitude)] × 100% . Increasing levels of prepulse intensity resulted in increased %PPI ( Supplementary Fig. 2). However, because there were no statistical interactions of sex or genotype with prepulse intensity (not shown), and to simplify data presentation, only results for the average PPI of all four prepulse intensities will be presented here. Because PPI at the 30ms inter-stimulus interval (ISI) was much lower and more variable than at the 100ms ISI, only the 100ms data are presented here. Raw data for the 30ms ISI are included in Supplementary Fig. 3. Saline and apomorphine were injected immediately prior to placing rats in PPI chambers, while MK-801 was given 15 min prior (Gururajan et al., 2015;van den Buuse et al., 2017).

Quantitative reverse transcriptase-PCR
Several days following the final behavioural tests, i.e. 6 weeks after the end of chronic dosing, the rats were deeply anaesthetised with CO 2 and decapitated for collection of brain samples. Brains were dissected on a cold plate and tissue samples were rapidly frozen on dry ice and stored at − 80 • C. A random selection of dissected striatum (n = 6/group) were prepared for gene expression analysis. Behavioural data for this subgroup are shown in Supplementary Figs. 4-6.
Total RNA was isolated from dissected tissue using Trizol (Ambion, Austin, TX, USA) with single-stranded complementary DNA synthesized using the QuantiTect Reverse transcription kit (Qiagen, Frankfurt, Germany). Quantitative PCR was performed with SYBR Green reagent (Qiagen) using the Rotor-Gene 6000 real-time PCR system (Corbett Life Science, Frankfurt, Germany). Forward (F) and reverse (R) primers used were. Two housekeeping genes were used (Cypa and Gapdh), and based on the finding of similar results against both genes, only Cypa is used here for data presentation. Relative mRNA levels were quantified using the 2 − ΔΔCT method (normalised to Cypa) with saline-pretreated Val/Val rats of each sex used as controls. Each PCR was performed in technical triplicates.

Data analysis
Data were analysed using the SPSS statistical package (version 27) and Graph Pad Prism (version 9). The data were first checked for multivariate outliers using z-scores, and scores falling outside z = ± 3.29 were removed. The data were then checked for normality violations using skewness and kurtosis z-scores and scores outside z = ± 1.96 considered to be violating normality. The sphericity assumption of repeated-measures was examined using Mauchly's test, and upon violation, the Greenhouse-Geisser degrees of freedom adjustment was applied to the critical p-values.
Between-group differences were analysed using a factorial Analysis of Variance (ANOVA), with repeated measures where appropriate. Between-subject factors were genotype (Val/Val, Val/Met, Met/Met), pretreatment (saline, Meth) and sex (male, female). Because of capacity limitations of the equipment, PCR gene expression analysis was done separately in samples from male and female rats. Therefore, statistical analysis could not include direct male/female comparison but, instead, focused on genotype and METH pretreatment effects in males and females separately. Where appropriate, post hoc analysis using Bonferroni's multiple comparison was used to investigate significant main effects and interactions. The level of significance was set at p < 0.05 for all statistical analyses. As a measure of effect size, partial eta squared (ηp 2 ) values were used, with cut-offs being ≥0.01 small, ≥0.06 medium, and ≥0.14 large.
There was, however, a significant effect of pretreatment (F(1,116) = 10.58, P = 0.001, ηp 2 = 0.084), with Meth-pretreated rats displaying higher baseline activity, although there were no significant interactions of pretreatment with either sex or genotype.
As some rats did not undergo testing with the lower dose of Meth, analysis of locomotor activity over 60 min following acute s.c. 0.5 mg/ kg ( Fig. 2A  that sensitisation also affected the time course of the effect of the acute Meth challenge in a sex-specific manner. There was, however, no interaction of genotype with acute Meth, pretreatment, or with pretreatment and acute Meth, suggesting there is no involvement of Val68Met genotype in Meth-induced hyperactivity or its sensitisation, when tested with the acute 0.5 mg/kg dose ( Fig. 2A and B).
Analysis of the effect of acute s.c. injection of 0.2 mg/kg Meth ANOVA of data from female rats split by genotype showed that in Met/ Met females, locomotor activity after both saline injection and following the 0.2 mg/kg Meth challenge was significantly higher in Methpretreated than saline-pretreated rats (52% and 68% higher, respectively, main effect of acute Meth challenge, F(1,10) = 39.40, P < 0.001, ηp 2 = 0.798; main effect of Meth pretreatment, F(1,10) = 6.90, P = 0.025, ηp 2 = 0.408; acute Meth × pretreatment interaction, F(1,10) = 3.73, P > 0.05, ηp 2 = 0.272). In contrast, in Val/Val and Val/Met females (Fig. 2D) and all of the male genotype groups (Fig. 2C) there was no significant effect of Meth pretreatment or interaction with the effect of the acute 0.2 mg/kg Meth challenge.

Post-treatment MK-801-induced locomotor hyperactivity
There was a significant main effect of MK-801 (F(1,115)
There were no significant interactions between MK-801 and sex (F (1,117) = 0.55, P > 0.05, ηp 2 = 0.005), or pretreatment (F(1,117) = 1.64, P > 0.05, ηp 2 = 0.014), and also no significant interaction with genotype (F(2,117) = 0.74, P > 0.05, ηp 2 = 0.001). These data suggest Fig. 3. Total 1-h locomotor activity following acute saline, 0.2 mg/kg or 0.5 mg/kg of Meth, or 0.05 mg/kg MK-801 in saline and Meth-pretreated male (A) and female (B) Val68Met rats. Male rats pretreated with Meth showed significantly greater locomotor distance moved following an acute challenge with 0.5 mg/kg Meth (* P < 0.05) than saline-pretreated male rats independent of genotype. Female Met/Met rats pretreated with Meth showed significantly greater locomotor distance moved following acute saline or 0.2 mg/kg Meth (# P < 0.05, see down arrows). Independent of genotype, female rats pretreated with Meth showed significantly greater distance moved following acute challenge treatment with 0.5 mg/kg Meth and MK-801 than female rats pretreated with saline (* P < 0.05). VV = Val/Val; VM = Val/Met; MM = Met/Met. Data are presented as mean ± SEM. that all groups respond similarly to the effects of 0.05 mg/kg with no effect of Val68Met genotype or Meth pretreatment ( Fig. 4A and B).
Analysis of baseline startle amplitude following saline control treatment ( Fig. 4C and D)  suggest no involvement of Val68Met genotype or Meth pretreatment in the drug-induced changes in startle amplitudes seen in these rats (Fig. 4).

Gene expression
PCR analysis of gene expression of D1, D2, D3 and D4 receptors and the dopamine transporter in the striatum of male rats showed no significant main effects of either genotype or Meth pretreatment, and no significant interactions between these factors (Fig. 5). There was a significant main effect of genotype for D5 (F(2,27) = 3.59, P = 0.041, ηp 2 = 0.21), however Bonferroni post hoc testing showed no significant differences between any on the genotypes although there was a trend for D5 receptor gene expression to be lower in male Val/Met rats compared to other genotypes (VM vs VV, p = 0.076, VM vs MM, p = 0.11). Similarly in females, there was no effect of Meth pretreatment on the expression of any of the genes, and no effect of or interaction with genotype except for D2 receptors, where there was a main effect of genotype (F(2,30) = 4.50, P = 0.019, ηp 2 = 0.23). Bonferroni post hoc testing showed that female Met/Met rats had significantly lower D2 receptor gene expression than Val/Val rats (P = 0.022; Fig. 5).

Discussion
We investigated the long-term effect of a sub-chronic Meth treatment regimen on behaviour and dopamine receptor and dopamine transporter gene expression in a rat model of the BDNF Val66Met polymorphism. Previous studies have suggested involvement of BDNF in the development of psychosis but the mechanisms responsible remain unclear. The main findings of this study were that female rats with the Met/Met genotype selectively showed sensitisation to the acute effect of Meth at the end of the one-week Meth pretreatment period. This effect was not seen in female Val/Val or Val/Met, while male rats showed tolerance irrespective of Val66Met genotype. Two weeks after the last Meth injection, female Met/Met rats showed increased locomotor activity following both an injection of saline or of a low dose of Meth, an effect again not seen in other genotypes or in males. PPI was disrupted by acute treatment with either apomorphine or MK-801 but these effects were not altered by Meth pretreatment in Val/Val or Val/Met rats. In contrast, while in saline-pretreated Met/Met rats apomorphine disrupted PPI, Meth-pretreated Met/Met rats did not show PPI disruption following apomorphine treatment. Female Met/Met rats were also the only group to show down-regulation of dopamine D2 receptor gene expression in striatum. Behavioural effects of MK-801 or its locomotor sensitisation by prior Meth pretreatment were not influenced by genotype. Combined, these effects suggest a subtle, but significant selective vulnerability of female Met/Met rats to the short-term and long-term effects of Meth, a genotype effect which could model increased vulnerability to psychosis development.
The acute effects of Meth on dopamine release in the brain include an action on vesicular release and dopamine uptake (Chu et al., 2008;Fleckenstein et al., 2009;Hedges et al., 2018;Hogarth et al., 2022), resulting in intracellular dopamine accumulation and eventually dopamine efflux through the dopamine transporter. Chronically, as shown by post-mortem and imaging studies, Meth use may induce reduced levels of dopamine, tyrosine hydroxylase and the dopamine transporter in nucleus accumbens, caudate and putamen, although levels of DOPA decarboxylase and the vesicular monoamine transporter were found to be normal (Wilson et al., 1996;Vanderschuren et al., 2000;Volkow et al., 2001). In animal models, chronic Meth treatment induces long-term behavioural sensitisation to its acute effects (Vanderschuren et al., 2000;Manning and van den Buuse, 2013;Hogarth et al., 2022), an effect also found in the present study. BDNF interacts with the effects of Meth at several levels, including acute dopamine release (Narita et al., 2003). Specifically, injection of a BDNF antibody into the nucleus accumbens suppressed dopamine release and behavioural activation induced by acute Meth treatment (Narita et al., 2003). It could then be expected that Val66Met, with its associated reduction of activity-dependent BDNF release, would similarly lead to reduced acute Meth-induced locomotor hyperactivity and reduced chronic Meth-induced sensitisation. This was not observed in the present study or in our previous mouse studies (Greening et al., 2021;Corrone et al., 2023). Instead, at the end of the one-week treatment period female rats with the Met/Met genotype showed sensitisation to the acute effect of Meth, which was not observed in female Val/Val or Val/Met rats or in male rats. The explanation for this selective effect in female Met/Met rats remains unclear, but it could represent enhanced vulnerability of the Met/Met genotype to psychosis induced by repeated Meth administration, at least in females. Several studies have shown sex differences in the effects of Meth in the brain (Daiwile et al., 2022). In addition, in mice we recently showed that the metabolism of Meth to amphetamine was greater in females than in males (Jaehne et al., 2022b). BDNF interacts with female sex hormones at several levels (Scharfman et al., 2006;Spencer et al., 2010;Wu et al., 2013) and it is therefore possible that high levels of estrogen and other female sex hormones create a homeostatic environment which exposes the effect of reduced BDNF release in Met/Met rats to induce behavioural changes, in this case enhanced Meth-induced locomotor hyperactivity. Of note, the 7-day Meth treatment regimen produced tolerance to the effect of Meth in male rats, which may be associated with their lower levels of estrogen and other hormones.
The selective locomotor hyperactivity sensitisation, seen in female Met/Met rats at the end of the treatment regimen, was not observed in our previous studies which only focused on long-term effects of chronic Meth (Greening et al., 2021;Corrone et al., 2023). Therefore it is possible that similar effects occurred in those mouse studies at the end of the three-week treatment with Meth and future studies will have to focus on those possible early effects to ascertain if the changes seen in this rat model generalise to other Val66Met models, specifically mice. Two weeks after the last Meth injection, female Met/Met rats no longer showed selective enhanced Meth-induced locomotor hyperactivity to the high dose of Meth. Instead, following both saline and a low dose of Meth, Meth-pretreated Met/Met rats showed greater locomotor activity than saline-pretreated Met/Met rats. This indicates that the initial hypersensitivity of this group to acute Meth developed into a subtle increase in the response to relatively mild environmental stimulation, in this case induced by acute saline and acute low-dose Meth. When administered the higher dose of Meth or MK-801, these subtle differences between the female Met/Met rats and the other genotypes were masked, i.e. a 'ceiling' effect. In previous studies in Val66Met mice (Greening et al., 2021;Corrone et al., 2023), we did not include a lower dose of an acute Meth challenge and therefore this subtle long-term effect of Meth sensitisation may have been overlooked.
Because it is disrupted in schizophrenia and other psychotic illnesses, PPI is a widely used measure of sensorimotor gating in animal models (van den Buuse, 2007;Powell et al., 2009;van den Buuse, 2010). However, baseline PPI can be decreased by a wide range of drugs and neurotransmitter mechanisms and we have therefore argued that animal model studies should include investigation of changes in PPI induced by relevant selective drugs (van den Buuse, 2010). It is possible that baseline PPI appears not to be altered by genetic or developmental factors but the effect of dopaminergic and/or glutamatergic is diminished or enhanced, unmasking underlying changes in the activity of brain pathways involved in PPI regulation (van den Buuse, 2010). Here, and in previous studies, we measured baseline PPI following saline injection, as well as disruption of PPI following acute treatment with the dopamine receptor agonist, apomorphine, and the NMDA receptor antagonist, MK-801, representing dopaminergic and glutamate NMDA receptor regulatory mechanisms. While baseline PPI was not altered by chronic Meth pretreatment, apomorphine-induced disruption of PPI was absent in female Met/Met rats, again showing this group as vulnerable to the effects of chronic Meth. It may have been expected that, similar to Meth-induced locomotor hyperactivity sensitisation, the effect of apomorphine on PPI would be enhanced in this group while, instead, the opposite was found. However, regulatory pathways involved in locomotor hyperactivity and PPI disruption are overlapping at best  and it is conceivable that the direction of effects caused by chronic Meth differs for the two behaviours. The lack of effect of apomorphine shows that acute dopaminergic regulation of PPI is disrupted in the female Meth-pretreated Met/Met group, an effect similar to the result we observed in a developmental model of schizophrenia, poly (I:C) induced maternal immune activation . In that model, PPI disruption following acute apomorphine and Meth treatment was absent in poly (I:C) offspring, similar to the present findings in Meth-pretreated Met/Met rats.
Meth interacts with NMDA receptor expression (Yamamoto et al., 2006;Simões et al., 2008) and, conversely, pretreatment with MK-801 has been shown to inhibit the long-term effects of Meth (Muraki et al., 1992;Ohmori et al., 1994, Kim andJang, 1997). Confirming a close interaction of Meth-induced dopamine release and NMDA receptors, here we show that Meth pretreatment leads to sensitisation of the effect of acute MK-801 on locomotor activity, albeit only in females. However, this interaction was independent of Val68Met genotype, suggesting the changes in drug-induced locomotor hyperactivity in Met/Met rats were specific to dopaminergic regulation of behaviour.
Analysis of the expression of dopamine receptors and the dopamine transporter in the striatum revealed a significantly lower expression of D2 receptors in female Met/Met rats, independent of Meth pretreatment. The relevance of these changes for the observed effects of Val66Met on behaviour remains to be clarified. One possibility could be that the reduced D2 receptor expression represents an adaptive response to higher dopamine release in the Met/Met female brain, which could also be the underlying mechanism of this group's higher vulnerability to the effects of chronic Meth. Micro-dialysis and voltammetry studies have shown enhanced baseline and stimulated dopamine release in female BDNF heterozygous mice (Birbeck et al., 2014), which could similarly be present in Met/Met rats. In neural transplantation studies, Met/Met rats showed enhanced graft-derived neurite outgrowth compared to wild-type rats (Mercado et al., 2021), consistent with dopaminergic hyperactivity. Clearly, this explanation remains speculative and analysis of dopamine release and receptor gene expression in other brain regions such as prefrontal cortex nuclei and the nucleus accumbens needs to be included in future studies for a more broad understanding of the neurochemical changes underlying Meth effects. Nevertheless, selectivity of the reduction of D2 expression for female Met/Met rats does parallel the selectivity of the effect of chronic Meth on Meth-induced locomotor hyperactivity for this sub-group. Previously, we used high-resolution quantitative mass spectrometry-based proteomics to biochemically map the long-term effects of Meth within the brain of Val66Met mice, including 4808 proteins across the mesocorticolimbic circuitry (Greening et al., 2021). Of note, Meth altered levels of several dopamine signalling markers, such as the dopamine transporter, catechol-O-methyltransferase (Comt) and tyrosine hydroxylase (TH) in Met/Met mice compared to Val/Val mice. Similar detailed studies are yet to be done in the Val68Met rat model used in the present study.
Our results show a subtle, but significant interaction of Val66Met with the effects of chronic Meth on psychosis endophenotypes. Results were female-specific, and depended on the time after chronic Meth treatment, the acute drug challenge used to induce psychosis-like behaviour, as well as on the dose of that drug challenge. Clinical studies on the role of BDNF Val66Met in psychosis have given mixed results (Notaras et al., 2015). For example, while some studies suggested genetic linkage of the Val66Met polymorphism with age of onset and symptoms of schizophrenia (Suchanek et al., 2013), other such studies produced negative results (Xu et al., 2007). Studies on an effect of Val66Met on Meth use disorder and the occurrence of Meth psychosis have similarly shown variable results (Haerian, 2013;Iamjan et al., 2015;He et al., 2020). Factors such as ethnicity, inclusion of both the Met/Met and Val/Met genotypes to compare to Val/Val controls rather than simply 'Met carriers', lack of focus on specific symptom clusters, but also limitations such as small cohort sizes, may all play a role in a dearth of clear understanding of how Val66Met may be involved in psychosis (for detailed discussion, see Notaras et al. (2015)). The use of animal models allows for tight control of experimental variables, while limiting the possible influence of environmental and genetic factors (van den Buuse, 2010; Gururajan et al., 2012;Tsai, 2018). This allows detection of subtle effects, as seen in the present study. Indeed, the relatively small magnitude of the effect of the Met/Met genotype is in line with its reported effect on BDNF release: there was only a 29% reduction of regulated BDNF release from hippocampal-cortical neurons obtained from Met/Met mice compared to Val/Val mice with no changes in constitutive release or resting levels of BDNF (Chen et al., 2006), unlike what can be expected in models of homozygous or heterozygous knockouts which are unrealistic representations of functional genetic variants generally observed in human populations.
The present findings sex and genotype specificity are in line with previous findings in female Val66Met mice and rats. For example, female Met/Met rats showed reduced reinstatement of lever pressing following extinction in an operant model of alcohol self-administration . Female hBDNF Met/Met mice exhibited reduced impulsivity compared to hBDNF Val/Val mice of the same sex as shown by a lower number of premature responses in the 5-choice serial reaction time task (Hogan et al., 2021). While female Val/Val mice showed a greater propensity toward stable operant ethanol self-administration relative to male mice of the same genotype, no sex difference was found in Met/Met mice (Hogan et al., 2021). Following adolescent corticosteroid treatment, female Met/Met mice showed reduced fear extinction in a fear conditioning paradigm and interneuron dysfunction in the amygdala as shown by reduced somatostatin and calretinin cell density, effects not seen in males (Raju et al., 2022). Thus, our results are consistent with the notion of major sex differences in the involvement of BDNF in behaviour and brain function (Scharfman et al., 2006;Spencer et al., 2010;Wu et al., 2013).
This study has some limitations. In the PPI protocol, we used a longer pre-injection time for MK-801 than for apomorphine and saline, based on the slow development of effect of MK-801 compared to APO. The protocol did not include a separate control group where saline was injected at the same longer pre-injection time point as was used for MK-801. Therefore it cannot be excluded that the differences in pre-injection times somehow influenced the outcome of PPI testing. Future experiment should include a separate saline control group with the same preinjection time as used here for MK-801. Another limitation is, that we used the rats from the behavioural experiments to assess dopamine receptor gene expression, rather than a separate, non-behavioural cohort. It cannot be excluded that the drugs used for behavioural testing (Meth, MK-801, apomorphine) had a long-term effect of dopamine gene expression. This should be confirmed in future experiments in a nonbehavioural cohort which receives the same Meth pretreatment as used here for the behavioural studies. Finally, while using the same Val68Met rat model it has been shown that cultures from embryonic Met/Met hippocampus display markedly reduced activity-dependent BDNF release (Mercado et al., 2021), in line with previous studies in mice (Egan et al., 2003;Chen et al., 2006), it is unclear whether there are sex differences in this release deficit. Therefore, the mechanism by which this Val66Met-induced differences in BDNF release may mediate sex-specific behavioural changes like we found in the present study, remains unclear.
In conclusion, these data suggest a selective vulnerability of female Met/Met rats to short-term and long-term effects of Meth, which could model increased vulnerability to psychosis development associated with the BDNF Val66Met polymorphism.

Declaration of competing interest
The authors have no conflict of interest to declare.

Data availability
Data will be made available on request.