Drug-induced change in transmitter identity is a shared mechanism generating cognitive deficits

Cognitive deficits are a long-lasting consequence of drug use, yet the convergent mechanism by which classes of drugs with different pharmacological properties cause similar deficits is unclear. We find that both phencyclidine and methamphetamine, despite differing in their targets in the brain, cause the same glutamatergic neurons in the medial prefrontal cortex to gain a GABAergic phenotype and decrease their expression of the vesicular glutamate transporter. Suppressing the drug-induced gain of GABA with RNA-interference prevents the appearance of memory deficits. Stimulation of dopaminergic neurons in the ventral tegmental area is necessary and sufficient to produce this gain of GABA. Drug-induced prefrontal hyperactivity drives this change in transmitter identity. Returning prefrontal activity to baseline, chemogenetically or with clozapine, reverses the change in transmitter phenotype and rescues the associated memory deficits. The results reveal a shared and reversible mechanism that regulates the appearance of cognitive deficits upon exposure to different drugs.

Brain impairments are often characterized by constellations of symptoms and behavioral alterations, some of which are shared across disorders.Cognitive deficits are found in mood and neuropsychiatric disorders such as drug misuse, schizophrenia and depression, raising the possibility that shared mechanisms could produce the same impairments in response to different stimuli.While studies have focused on the actions of single drugs, less attention has been given to investigating the mechanisms of action that different drugs of abuse have in common.We investigated the effect of sub-chronic treatment with phencyclidine (PCP) or methamphetamine (METH), two drugs belonging to different classes of chemicals.PCP affects glutamatergic transmission by acting as an NMDA antagonist 1 , while METH affects signaling by dopamine and other monoamines 2 .While differing in their molecular targets in the brain and in some behavioral effects [3][4][5][6] , PCP and METH have been extensively studied for their ability to cause longlasting cognitive deficits and mimic symptoms of schizophrenia [7][8][9] .However, the process by which they generate similar behavioral impairments has been unknown.Understanding shared neuronal mechanisms underlying drug-induced cognitive deficits could foster development of effective treatments and be beneficial for a spectrum of disorders 10,11 .
When neuronal activity is altered for a sustained period, neurons can change the neurotransmitter they express, often switching from an excitatory to an inhibitory transmitter or vice versa and causing changes in behavior 12,13 .Using a combination of genetic labeling strategies, RNA interference, chemogenetics and optogenetics, we investigated whether changes in cortical neuron transmitter phenotype are a shared mechanism involved in the generation of both PCP-and METH-induced cognitive deficits.

Phencyclidine induces a change in the transmitter phenotype of neurons in the medial prefrontal cortex that is required for the appearance of cognitive deficits
To determine whether changes in neurotransmitter phenotype are involved in generating cognitive deficits, we tested the effect of a 10-day treatment with PCP (10 mg/kg/day), which induces long-lasting cognitive impairments and recapitulates deficits observed in schizophrenia 8,14 .We focused on the medial prefrontal cortex (mPFC), which is a major hub for cognitive control 15,16 , and examined the transmitter phenotype of glutamatergic neurons expressing vesicular glutamate transporter 1 (VGLUT1) 16 because they represent the largest neuronal population in the mPFC.To identify these neurons following changes in transmitter profile, we labeled them permanently with a nuclear mCherry reporter using VGLUT1 CRE ::mCherry mice (Fig. 1a).
These results suggested that PCP induces the synthesis and expression of GABA in PL glutamatergic neurons not previously expressing this transmitter.
To learn whether the observed co-expression of mCherry and GABA is caused by unspecific mCherry expression, we used fluorescent in situ hybridization (FISH) to determine the extent to which mCherry labelling recapitulates VGLUT1 expression in the PL of VGLUT1 CRE ::mCherry mice (Supplementary Fig. 3a,b).In both saline-and PCP-treated mice, 96% of neurons expressed both VGLUT1 and mCherry, while 3% expressed VGLUT1 but not mCherry and only ~1% expressed mCherry + but not VGLUT1 (Supplementary Fig. 3c).To find out whether these VGLUT1 -/mCherry + neurons co-expressed GABAergic markers, we quantified mCherry and VGLUT1 colocalization with transcripts for the GABA synthetic enzyme (GAD1).0.08% of mCherry + neurons in saline-and 0.05% in PCP-treated mice were VGLUT1 -and GAD1 + (Supplementary Fig. 3d).Instead, GAD1 co-localization was most frequently detected in neurons labelled with both VGLUT1 and mCherry (0.8% of total glutamatergic neurons in saline-treated controls, and 2.2% in PCP-treated mice) (Supplementary Fig. 3e).Thus, in both PCP-and saline-treated VGLUT1 CRE ::mCherry mice, unspecific expression of mCherry contributes £0.08% to the observed coexpression of GABA and mCherry in PL neurons (Fig. 1b,c).
We next examined the expression levels of the GABA vesicular transporter (VGAT) and VGLUT1 in mCherry + neurons that gained GABA after PCP treatment or that co-expressed GABA in drug-naïve conditions, and compared those levels with the expression levels of VGAT and VGLUT1 in neurons expressing only GABA or only glutamate.We used FISH to detect transcripts for mCherry, GAD1, and for either VGAT or VGLUT1.To reveal changes in expression levels, we selectively decreased the amplification for VGAT and VGLUT1 to obtain punctate staining (Fig. 1e,f).In the PL of PCP-treated mice, neurons expressing both mCherry and GAD1 (mCherry + /GAD1 + ) expressed VGAT at the level of GABAergic neurons (labeled with GAD1 and not mCherry) (Fig. 1g).At the same time, the expression level of VGLUT1 in mCherry + /GAD1 + neurons decreased by ~55% compared to that of glutamatergic cells expressing mCherry and not GAD1 in PCP-treated mice (Fig. 1h).Neurons co-expressing mCherry and GAD1 in drugnaïve conditions also showed high expression levels of VGAT and low VGLUT1 (46% less VGLUT1 than in purely glutamatergic neurons expressing only mCherry), as evident from analyses of these cells in salinetreated controls (Supplementary Fig. 4a-c).The expression level of VGAT and VGLUT1 in mCherry + /GAD1 + neurons did not differ between PCP-and saline-treated mice (Supplementary Fig. 4d,e), but the number of mCherry + /GAD1 + neurons was higher in the PL of PCP-treated mice than in controls (6.6±0.5 neurons/mm 2 , saline; 10.6±0.9 neurons/mm 2 , PCP), mirroring the PCP-induced increase in the number of mCherry + /GAD67 + neurons (Fig. 1c) and suggesting that PCP induced gain of GAD1 in mCherry + neurons that were not expressing it earlier.Thus, both the glutamatergic neurons that gained GABA after PCP-exposure, as well as those expressing GABA in drug-naïve conditions, express high levels of VGAT and lowered levels of VGLUT1.
We then asked whether PCP-treatment affects the transmitter phenotype of PL GABAergic neurons.No difference was observed in the number of neurons expressing GABA and not mCherry between PCP-treated animals and controls (8594±340 vs 8837±271) (Supplementary Fig. 1e).However, PCP and other NMDA receptor antagonists have been shown to reduce the expression of GAD67 and parvalbumin in prefrontal cortex parvalbumin-positive (PV + ) interneurons 17,18 .Because variability in the number of GABAergic neurons scored could have prevented the detection of loss of GABA from a small number of PV + neurons, we quantified the number of PV + neurons expressing GAD67 after PCP-treatment by permanently labeling them with a PV CRE ::TdTomato mouse line.PCP caused 223±45 TdTomato + neurons (6% of the TdTomato + population) to stop expressing GAD67 (Supplementary Fig. 5), confirming that PCP treatment reduces the expression of GABAergic markers in a subpopulation of PV + neurons 18 .
To investigate whether glutamatergic neurons that have gained GABA contribute to cognitive deficits, we selectively suppressed GABA expression in PL glutamatergic neurons by injecting a Credependent adeno-associated virus (AAV) expressing shRNA for GAD1 (AAV-DIO-shGAD1-GFP or AAV-DIO-shScr-GFP as control) in the PL of VGLUT1 CRE mice before exposure to PCP (Fig. 1i,j).shGAD1 suppressed GABA expression in neurons expressing the virus (Supplementary Fig. 6a-c), and reduced the number of PL neurons co-expressing GAD1 and VGLUT1 transcripts in both PCP-and saline-treated mice to half of that in saline-ShScr controls (Fig. 1k).Having efficiently suppressed PCP-induced gain of GABA, we examined the impact of shGAD1 on PCP-induced behavior.While shGAD1 did not affect PCP-induced hyperlocomotion on the first day of treatment, it prevented appearance of locomotor sensitization after a 10-day PCP-exposure (Fig. 1l and Supplementary Fig. 6d), indicating that PCP-induced gain of GABA is required for sensitization to the acute locomotor effect of the drug.We next focused on deficits in recognition and working memory, since these behaviors are affected by repeated exposure to PCP 14,19 and are regulated by the PL 20,21 .shGAD1 prevented PCP-induced impairments of recognition memory in the novel object recognition test (NORT) (Fig. 1m) and deficits in spatial working memory in the spontaneous alternation task (SAT) (Fig. 1n).Neither PCP-treatment nor shGAD1 changed exploratory behaviors in the NORT and in the SAT (Supplementary Fig. 6e,f).shGAD1 did not affect the behavioral performances of saline-treated controls (Fig. 1l-n), suggesting that glutamatergic neurons gaining GABA upon PCPexposure, but not those co-expressing GABA before drug-exposure, mediated drug-induced locomotor sensitization and memory impairments.The number of PL VGLUT1 + /GAD1 + neurons was positively correlated with locomotor sensitization, and negatively correlated with object recognition and working memory performance (Fig. 1o-q).Overall, these data indicate that PCP-induced gain of GABA in PL glutamatergic neurons is necessary for the appearance of these behavioral alterations.

Methamphetamine changes the transmitter phenotype of the same prelimbic neurons affected by phencyclidine
Because METH causes memory deficits similar to those induced by PCP [21][22][23] , we asked whether METH-treatment would also affect the transmitter identity of PL glutamatergic neurons.Resembling the effect of PCP, 10 days of METH-treatment (1 mg/kg/day) increased the number of mCherry + PL neurons co-expressing GABA and GAD67 by 1.7-and 1.9-fold (Fig. 2a-d and Supplementary Fig. 7a), without changing the number of mCherry + and GABA + /mCherry -cells (Supplementary Fig. 7b) or inducing apoptosis or neurogenesis (Supplementary Fig. 2).As previously observed for PCP-and saline-treated animals, unspecific expression of mCherry could not account for the observed co-expression of GABA and mCherry in the PL of mice that received METH (Supplementary Fig. 3).We next used FISH to examine the expression of the VGAT and VGLUT1 in mCherry + neurons that co-expressed or gained GABA. Similar to the effects of PCP-treatment, mCherry + /GAD1 + neurons in the PL of METH-treated mice expressed a level of VGAT equal to that of neurons expressing GAD1 but not mCherry (Fig. 2e).The expression level of VGLUT1 decreased by ~76% compared to that of glutamatergic cells expressing mCherry and not GAD1 (Fig. 2f).VGAT and VGLUT1 expression levels in mCherry + /GAD1 + neurons did not differ from those measured in PCP-and saline-treated mice (Supplementary Fig. 4d,e), and the number of mCherry + /GAD1 + neurons in the PL of mice that received METH resembled the number observed in PCP-treated mice (10.6±0.9 neurons/mm 2 , PCP; 10.2±0.5 neurons/mm 2 , METH), indicating that both drugs affect PL glutamatergic neurons similarly.
Glutamatergic neurons that co-express GABA or gain it after treatment with either drug were most prevalent in layer 2/3 and layer 5 of the PL (Supplementary Fig. 8a).These PL layers innervate the nucleus accumbens (NAc) 24 , which modulates behaviors that are affected by repeated intake of PCP or METH [25][26][27] .
To determine whether neurons that change transmitter identity project to the NAc, we injected fluoro-gold (FG) into the NAc of VGLUT1 CRE ::mCherry mice, treated them with PCP, METH or saline, and screened the PL for mCherry + /GABA + neurons expressing the retrograde tracer (Supplementary Fig. 8b-d).In both PCP-and METH-treated mice, ~0.9% of FG + neurons were mCherry + /GABA + .Such cells were less frequent in controls (~0.3% of the total number of FG + neurons) (Supplementary Fig. 8e), indicating that neurons changing transmitter identity with drug-treatment project to the NAc.
Since both PCP and METH affect the transmitter phenotype of PL glutamatergic neurons that have the NAc as a shared downstream target, we asked whether both drugs change the transmitter identity of the same cells.If PCP and METH changed the transmitter identity of different cells, administering the two drugs one after the other should induce gain of GABA in neurons that have not gained it after treatment with the first drug.To determine whether this was the case, we genetically labeled neurons expressing GABAergic markers during the interval between the delivery of PCP and METH, using VGAT FLP ::CreER T ::TdTomato cON/fON mice (see Methods) in which neurons expressing VGAT at the time of tamoxifen administration are permanently labeled with TdTomato (Fig. 3a,b, Supplementary Fig. 9).We first injected tamoxifen in saline-treated controls and determined that TdTomato labels neurons co-expressing VGLUT1 and GAD1 in drug-naïve conditions with 77% efficiency and 79% specificity (Supplementary Fig. 9d,e).
We next used VGAT FLP ::CreER T ::TdTomato cON/fON mice to determine if PCP and METH cause the same neurons to change transmitter phenotype.Mice were treated first with PCP followed by tamoxifen administration, and then treated with either saline, PCP or METH (Fig. 3e).Across treatment groups the total number VGLUT1 + /GAD1 + neurons was unchanged (1169±46, PCP+saline; 1273±69, PCP+PCP; 1177±45, PCP+METH) (Fig. 3f), indicating that consecutive administration of drugs does not cause additional glutamatergic neurons to gain GAD1.Furthermore, we did not detect differences in the number of VGLUT1 + /GAD1 + /TdTomato + neurons (861±73, PCP+saline; 831±43, PCP+PCP; 832±38, PCP+METH) nor in the number of VGLUT1 + /TdTomato + neurons (180±44, PCP+saline; 108±41, PCP+PCP; 237±103, PCP+METH) (Fig. 3f).A decrease in the first population and an increase in the second population would have indicated loss of GAD1 from some VGLUT1 + neurons and gain of GAD1 in another population of VGLUT1 + neurons.These results indicated that consecutive administration of PCP and METH does not cause gain of GAD1 by additional neurons, nor induces neurons that gained GAD1 upon PCP-treatment to revert to their original transmitter phenotype.Thus, PCP and METH change the transmitter identity of a largely overlapping population of PL neurons.

Signaling by dopaminergic neurons is necessary and sufficient to change the transmitter identity of prelimbic neurons
Demonstration that both PCP and METH have the same effect on the transmitter phenotype of the same PL glutamatergic neurons prompted investigation of the underlying mechanism of drug action.PCP, METH, and other addictive substances increase phasic firing of dopaminergic neurons in the ventral tegmental area (VTA) 28,29 and increase the levels of extracellular dopamine (DA) in the striatum and prefrontal cortex 30,31 .Could signaling by dopaminergic neurons in the VTA be a common mediator of the PCP-and METH-induced change in transmitter identity?To address this question, we tested whether suppressing the activity of VTA dopaminergic neurons during treatment with PCP or METH affects the number of PL neurons that change transmitter phenotype.We expressed the PSAML-GlyR chemogenetic receptor in the VTA of DAT CRE mice (Supplementary Fig. 10a-d).Administration of the PSEM 308 ligand before drug-injection suppressed the acute PCP-and METH-induced increase in c-fos + dopaminergic neurons (Supplementary Fig. 10e,f).Combining VTA suppression with drug administration for the entire duration of treatment, by co-administration of PSEM 308 and PCP or METH, prevented the increase in the number of PL VGLUT1 + /GAD1 + neurons (Fig. 4a-d).These results show that drug-induced increase in activity of VTA dopaminergic neurons is required for PL neurons to change their transmitter phenotype upon treatment with PCP or METH.
It remained unclear, however, whether stimulation of dopaminergic neurons in the VTA is by itself sufficient to induce PL neurons to change transmitter identity, or whether other effects of PCP or METH are involved.Phasic firing of VTA neurons can be mimicked by optogenetic stimulation of VTA dopaminergic neurons 32 .To learn whether repeated optogenetic stimulation of VTA dopaminergic neurons is sufficient to induce PL glutamatergic neurons to change transmitter identity, we expressed ChR2-YFP (or YFP as control) in VTA DAT CRE neurons and implanted an optic fiber above the VTA (Fig. 4e,f).Administration of 80 sets of laser stimulation (each consisting of 30 bursts of five 8.5 ± 1 mW pulses of 4 ms duration at 20 Hz) over the course of 1h increased the number of VTA dopaminergic neurons expressing c-fos by 7.3-fold in mice expressing ChR2 (Supplementary Fig. 11a-d).We then exposed mice to 1h of VTA stimulation per day for 10 days and analyzed the transmitter phenotype of PL glutamatergic neurons.Remarkably, the number of VGLUT1 + /GAD1 + was 1.7-fold higher in ChR2-expressing mice compared to controls (Fig. 4g,h, Supplementary Fig. 11h-j), demonstrating that phasic firing of dopaminergic neurons in the VTA changes the transmitter phenotype of PL glutamatergic neurons.These findings establish signaling by dopaminergic neurons in the VTA as a common mediator for PCP-and METH-induced gain of GABA in PL glutamatergic neurons and suggest that exposure to other addictive substances that activate the VTA could produce similar effects.

Drug-induced prelimbic hyperactivity mediates the change in transmitter phenotype and linked cognitive deficits
We next asked how the effects of three different stimuli, PCP, METH and stimulation of dopaminergic neurons in the VTA, converge to change the transmitter identity of glutamatergic neurons in the PL.Increased neuronal activity can cause neurons to change the transmitter they express 12,33,34 .Could PCP, METH and optogenetic VTA stimulation induce alterations in PL activity that mediate the change in PL neuron transmitter phenotype?PCP and METH increased c-fos expression in PL glutamatergic neurons by 3.8-and 3.5-fold after a single injection and by 2.6-and 3.7-fold throughout a 10-day treatment (Supplementary Fig. 12a-d,f,g).This PL hyperactivity was still present 2 days after the end of drug-treatment (Supplementary Fig. 12i,j).Similarly, expression of c-fos in PL glutamatergic neurons was increased by 2.7-fold in mice expressing ChR2 after 1h of phasic stimulation of VTA dopaminergic neurons (Supplementary Fig. 11e-g).To determine whether this increase in activity promoted the change in transmitter phenotype, we tested whether suppression of PL hyperactivity during treatment with PCP or METH would prevent glutamatergic neurons from gaining GABA.Glutamatergic cells in the PL receive perisomatic inhibition from local PV + interneurons, which do not show changes in c-fos expression after administration of PCP or METH (Supplementary Fig. 12e,h,k).We hypothesized that chemogenetic activation of PV + neurons would suppress drug-induced hyperactivity of glutamatergic cells 35,36 .To test this idea, we expressed the chemogenetic receptor PSAML-5HT3HC in mPFC PV + neurons using AAV-flex-PSAML 141F,Y115F 5HT3HC-GFP and administered the PSEM 308 ligand immediately before acute injection of either PCP or METH (Supplementary Fig. 13a-e).While GFP + neurons infected with the virus showed high c-fos expression after PSEM 308 treatment, consistent with their expected activation (Supplementary Fig. 13f), the PCP-or METHinduced increase in c-fos expression in PL glutamatergic neurons was suppressed (Supplementary Fig. 13g-i).
We then combined chemogenetic activation of PL PV + interneurons with either PCP-or METHadministration for the duration of drug-treatment (Fig. 5a,b).The number of VGLUT1 + /GAD1 + neurons in the PL of PCP-and METH-treated mice that received PSEM 308 was half of that of mice that did not (586±18 and 611±23 vs 1262±66 and 1222±12) and was indistinguishable from that of saline-treated controls (Fig. 5c,d).Thus, suppression of drug-induced PL hyperactivity is sufficient to prevent glutamatergic neurons from changing their transmitter identity, indicating that hyperactivity is required for the change in transmitter phenotype.

Normalizing prelimbic neuron activity after drug-exposure reverses the change in transmitter identity and the associated behavioral alterations
PL glutamatergic neurons that change their transmitter upon exposure to PCP or METH retain the GABAergic phenotype for at least 11 days of drug washout (Fig. 5d, and PCP+saline group in Fig. 3f).As PCP-and METH-induced memory deficits are also long-lasting 14,38 , we asked whether the persistence of behavioral deficits is linked to retention of the GABAergic phenotype and whether both are reversible.
Clozapine, an antipsychotic drug that acts as a serotonin and dopamine receptor antagonist, reverses PCPinduced deficits in the NORT 14 , leading us to investigate whether it also reverses the change in transmitter phenotype.VGLUT1 CRE ::mCherry mice that received PCP displayed 1.94 fold more mCherry + /GABA + PL neurons than controls 17 days after the end of PCP-treatment, indicating that neurons had maintained the acquired GABAergic phenotype (Fig. 6a-c).In mice that received clozapine treatment after PCP, the number of mCherry + /GABA + neurons was reduced compared to that of mice treated with PCP alone (559±55 vs 1124±94) and was not different from that of saline-treated controls (Fig. 6a-c).Clozapine did not affect the number of mCherry + /GABA + neurons in saline-treated mice, suggesting that this drug selectively reverses the PCP-induced change in glutamatergic neuron transmitter identity.We found that clozapine rescued PCP-induced memory deficits in the NORT and SAT, without affecting the behavioral performance of controls (Fig. 6d-g and Supplementary Fig. 14a-d).
We next investigated the mechanisms underlying clozapine-induced reversal of the change in transmitter identity.Because clozapine suppresses the acute PCP-induced increase in PL c-fos expression 39,40 , we asked whether reversal of the gain of GABA depends on suppression of neuronal activity.After drug-treatment, the number of c-fos + glutamatergic neurons was 2.1-and 2.8-fold higher in PCP-and METH-treated mice compared to controls (Supplementary Fig. 12i-k).Administration of clozapine after PCP-treatment returned c-fos expression to baseline (Supplementary Fig. 14e-h), suggesting that PL hyperactivity during drug-washout is necessary to maintain the newly acquired transmitter phenotype.If this were the case, suppressing PL hyperactivity after the transmitter change has occurred could be expected to reverse the change.To test this hypothesis, we chemogenetically activated PV + neurons for 10 days to normalize c-fos expression in the PL of PCP-or METH-treated mice after the change in transmitter phenotype had taken place (Supplementary Fig. 15a-d).More than 3 weeks after the end of drug-treatment, glutamatergic neurons in the PL of both PCP and METH-treated mice still displayed the drug-induced GABAergic phenotype (Fig. 7a-c).Normalizing PL activity decreased the number of VGLUT1 + /GAD1 + neurons in the PL of PCP and METH-treated mice to the level of controls (588±27 and 575±9 vs 1225±38 and 1071±42) (Fig. 7a-c).Thus, PL neuronal activity maintains the change in transmitter identity once it has been induced.Chemogenetically activating PL PV + interneurons after the change in transmitter phenotype had occurred also rescued memory deficits in the NORT and SAT and suppressed locomotor sensitization to both PCP and METH (Fig. 7d-i and Supplementary Fig. 15e-k).Overall, these data show that suppressing PL hyperactivity following drug-exposure reverses the change in transmitter phenotype and the associated behavioral alterations.

Discussion
We show that gain of a GABAergic phenotype by PL glutamatergic neurons is a shared mechanism involved in generating both PCP-and METH-induced cognitive deficits.Both drugs cause the same PL neurons to acquire a new transmitter phenotype characterized by expression of GABA, GAD67, and VGAT, combined with lower levels of VGLUT1.Other PL neurons show the same transmitter phenotype in drug naïve conditions, as suggested by earlier studies 41,42 .This change in transmitter phenotype regulates the appearance of locomotor sensitization and memory deficits in the NORT and SAT, consistent with the involvement of the PL and the NAc, which receives input from PL neurons that change their transmitter identity 20,21,[25][26][27]43 . Give the role of GABAergic long-range projections in modulating brain oscillation and synchronization 44 , gain of GABA by PL neurons projecting to the NAc may contribute to the reduction of NAc firing rates and disruption of cortex-accumbens synchronization after PCP-treatment 45 .
Signaling by dopaminergic neurons in the VTA is necessary and sufficient to change the transmitter identity of PL neurons.Chemogenetic suppression of dopaminergic hyperactivity in the VTA during PCPor METH-treatment prevents the change in transmitter phenotype.Furthermore, optogenetic stimulation of phasic firing of VTA dopaminergic neurons, in the absence of drug administration, is sufficient to induce PL hyperactivity and produce the change in transmitter identity.These results are likely achieved through monosynaptic and polysynaptic connections from the VTA to the PL 46,47 .Furthermore, VTA dopaminergic neurons can release other transmitters than dopamine 48 , which may also contribute to the change in transmitter phenotype of PL neurons.Our results are consistent with evidence that many addictive substances, including PCP and METH, promote phasic firing of dopaminergic neurons in the VTA 49 and that inducing phasic firing of these neurons enhances DA release in the NAc 50 mimicking the effect of drug intake.
Neuronal hyperactivity in the PL mediates the change in transmitter identity, as expected for activity-dependent neurotransmitter switching 12,33,34 , and is necessary to maintain the newly acquired transmitter phenotype after the end of drug treatment.Midbrain cholinergic neurons that change transmitter identity in response to sustained exercise spontaneously revert to expression of their original transmitter within a week of cessation of the stimulus 12 .In contrast, PL glutamatergic neurons maintain their GABAergic phenotype for more than 3 weeks after the end of drug-treatment and the linked cognitive deficits are long-lasting 14,38 .c-fos expression in the PL increases after acute treatment with PCP or METH 39, 51 , as well as after one hour of phasic stimulation of VTA dopaminergic neurons, and remains elevated for at least two weeks 52,53 .Suppressing drug-induced hyperactivity during or after drug-treatment, respectively, prevents or rescues the change in transmitter phenotype and the coupled behavioral alterations, indicating that PL hyperactivity is necessary to produce and maintain these changes.Because exposure to either PCP or METH decreases expression of PV and GAD67 18,54 , impaired function of PFC PV + interneurons may contribute to PL glutamatergic neuron hyperactivity and maintenance of the newly acquired transmitter phenotype.Chronic treatment with clozapine also reduces c-fos expression in the PL of PCP-treated mice and reverses PCP-induced changes in transmitter phenotype and behavior.This effect may be mediated by increased inhibitory input to PL glutamatergic neurons 55 .
Overall, we identify a novel, shared and reversible cellular mechanism involved in the ability of different drugs to generate cognitive deficits.Since multiple drugs of abuse acutely promote mPFC hyperactivity 56,57 , and stimulation of VTA dopaminergic neurons is sufficient to change the transmitter identity of PL neurons, drugs other than METH and PCP may result in the appearance of cognitive deficits by promoting this change in the transmitter phenotype of PL glutamatergic neurons.
Homozygous DAT CRE mice were bred with wild-type animals to obtain DAT CRE heterozygous experimental mice.

Drugs and pharmacological treatments
Phencyclidine hydrochloride (PCP; Sigma-Aldrich, P3029) was dissolved in sterile saline and administered subcutaneously (s.c.) at a dose of 10 mg/kg/day for 10 consecutive days 14 .Methamphetamine hydrochloride (METH; Sigma-Aldrich, M8750) was dissolved in saline and administered s.c. at a dose of 1 mg/kg/day for 10 consecutive days (modified from 21 ).To investigate whether drug-treatment affects the transmitter identity of PL neurons, mice were sacrificed either 2 days (Fig .1b,c,g,h, Supplementary Fig. 1c-e, Fig. 2c,e,f, Supplementary Fig. 3, Supplementary Fig. 4, Supplementary Fig. 5, Supplementary Fig. 7, Supplementary Fig. 8e, Fig. 4d), 11 days (Fig. 1k, Fig. 5d), 17 days (Fig. 6c) or 23 days (Fig. 7c) after the end of PCP or METH treatment.To investigate the effect of PCP or METH on PL neuronal activity, mice were sacrificed 2h after a single injection of the drug (Supplementary Fig. 12c-e), 2h after the last of 10 daily injections (Supplementary Fig. 12f-h) or 2 days after the end of 10 days of drug treatment (Supplementary Fig. 12i-k).
Tamoxifen (Sigma-Aldrich, T5648) was dissolved in corn oil (Sigma-Aldrich, C8267)/ethanol 9:1 and administered intraperitoneally (i.p.) at a dose of 75 mg/kg/day.To test whether PCP and METH affect the neurotransmitter phenotype of the same PL neurons, VGAT FLP ::CreER T ::TdTomato cON/fON mice received a 10-day treatment with either saline or PCP (10 mg/kg/day).Beginning on day 9 of saline/PCP administration, mice received an injection of tamoxifen each day at 2 pm for 7 consecutive days.Mice were then left untreated for 7 days to enable tamoxifen washout before exposing them to additional treatments 58 .After the end of the washout period, mice received PCP, METH, or saline for 10 additional days and were sacrificed 2 days after the last injection.
PSEM 308 hydrochloride (Fisher Scientific, 64-252-5) was dissolved in DMSO to obtain a stock solution of 5 mg/ml.Before use, the stock solution was further diluted in saline to a final concentration of 1 mg/ml.To investigate the effect of acute chemogenetic manipulation on neuronal activity, mice were injected i.p. with PSEM 308 (5 mg/kg) in DMSO/saline (1:4) or DMSO/saline alone (vehicle) 10 min prior to injecting METH, PCP or saline, and were sacrificed 2h after drug-injection.To investigate the effect of repeated chemogenetic manipulation mice were injected i.p. twice a day for 10 days with either PSEM 308 (5 mg/kg at 10-11 am and 2.5 mg/kg at either 1-2 pm or 4-5 pm) or vehicle.For experiments in which PSEM 308 was administered together with PCP or METH, the first PSEM 308 injection was administered 10 min prior to injecting METH, PCP or saline, and the second PSEM 308 injection occurred 3 h later.For experiments in which the pharmacogenetic treatment occurred after the end of drug treatment, the two PSEM 308 daily injections were administered 6 h apart.
Clozapine (Sigma-Aldrich, C6305) was dissolved in DMSO to obtain a stock solution of 10 mg/ml.The stock solution was further diluted in sterile saline before use to a final concentration of 0.5 mg/ml.DMSO dissolved 1:19 in saline (vehicle) was used as a control solution.Three days after the end of PCP treatment, mice began receiving daily i.p. injections of clozapine (5 mg/kg/day) or DMSO/saline (1:19) alone (vehicle) for 2 additional weeks 14 .

TUNEL assay
TUNEL assay was used to detect in situ apoptosis.The assay was performed using the In Situ Cell Death Detection (TUNEL) Kit with TMR Red (Roche, 12156792910) as previously described 12 .Briefly, 30 µm thick PL frozen sections were post-fixed with 1% PFA for 20 min at 22-24 °C and rinsed with PBS (three times, 5 min each).Sections were then permeabilized in 0.1% sodium citrate and 1% Triton X-100 for 1 h at 22-24 °C.After rinsing in PBS (three times, 5 min each), sections were incubated with TUNEL reaction solution according to the vendor's instructions.Incubation was performed in a humidified chamber for 3 h at 37 °C in the dark.Sections were rinsed and mounted with Fluoromount containing DRAQ-5 (1:1000).As positive controls, sections were treated with DNase I (10 U/mL, New England Biolabs, M0303S) for 1 h at 37 °C, rinsed in PBS (three times, 5 min each), and incubated with the TUNEL mixture.

Fluorescent in situ hybridization (FISH)
FISH was performed using the RNAscope Multiplex Fluorescent v2 Kit (Advanced Cell Diagnostics, 323100) according to the manufacturer's instructions, with a few adjustments.Briefly, the 30-μm fixed brain sections were mounted on Superfrost Plus slides and air-dried in a 50 °C oven for 30 min.Sections were rehydrated in PBS for 2 min and incubated for 5 min in 1X target retrieval solution at 95 °C.After one rinse in distilled water (2 min), sections were dehydrated with 100% ethanol for 5 seconds, air-dried, and incubated for 10 min in 5% hydrogen peroxide at RT.Sections were then incubated in a HybEZ humidified oven at 40 °C with protease III for 30 min, and later with the probe solution for 2 h.After incubation with the probes, slides were incubated o/n in a solution of SSC5X at 22 °C.The following day, sections were incubated with the following solutions in a HybEZ humidified oven at 40 °C with three rinsing steps in between each: amplification Amp-1, 30 min; Amp-2, 30 min; Amp-3, 15 min.For each probe used, sections were incubated in a HybEZ humidified oven at 40 °C with the following solutions: HRP-C1, -C2, or -C3 (depending on the probe) for 15 min, Opal dye of choice for 30 min, and HRP-blocker for 15 min.Opal 520 (Akoya Biosciences, FP1487001KT), Opal 570 (Akoya Biosciences, FP1488001KT) and/or Opal 690 (Akoya Biosciences, FP1497001KT) dyes were used for fluorescent labelling.
Depending on experimental needs, we obtained different levels of signal amplification by changing the dilution of the Opal dyes.To achieve a fully amplified signal, the Opal dyes were used at a dilution of 1:1500.This dilution was used to identify the cell-boundaries of neurons co-expressing mCherry and GAD1 (used to detect mCherry and GAD1 transcripts (Fig. 1e-h, Fig. 2e,f, Supplementary Fig. 4) and to quantify the number of VGLUT1 + / GAD1 + PL neurons (Fig. 1k, Supplementary Fig. 3 7c,d).Because VGLUT1 expression is decreased but not completely lost in neurons that have changed their transmitter identity, obtaining a fully amplified VGLUT1 signal allows detection of these cells.

Imaging
Images were acquired with a Leica SP5 confocal microscope with a 25x/0.95water-immersion objective and a z resolution of 1 μm, or with Leica Stellaris 5 with a 20x/0.75CS2 dry objective and a z resolution of 1 μm for immunohistochemistry and 0.7 μm for RNAscope.Low magnification images were acquired with Leica Stellaris 5 with a 10x/0.40CS2 dry objective with a z resolution of 2.5 μm.

Cell counting
All counts were performed by investigators double-blinded to the origin of each image.Either Image-J/Fiji or Imaris9 (Fig. 2c, Supplementary Fig. 7, Fig. 6c, Supplementary Fig. 12, Supplementary Fig. 13d-i, Supplementary Fig. 14g,h, Supplementary Fig. 15c,d) was used for cell counting.When using Image-J/Fiji, cell counts were performed by examining all sections within the confocal stacks without maximal projection.
When using Imaris9, cell counts were performed semiautomatically using the Spot detection function and the Colocalize Spot plug-in, and later corrected manually.Independently of the software used for analysis, only neurons showing colocalization in at least 3 consecutive z-planes were included in the co-expression group.
PL sections were analyzed from Bregma +2.8 mm to Bregma +1.54 mm, according to the Paxinos Mouse Brain Atlas.PL boundaries were determined based on PL cytoarchitecture, as previously described 59 .Pilot experiments determined that counting 1 in 6 sections was sufficient to estimate the number glutamatergic and GABAergic cells in the PL.Consequently, to determine the number of PL mCherry + /GABA + or GAD67 + neurons (Fig. 1c, Fig. 2c, Fig. 6c), TdTomato + /PV + neurons (Supplementary Fig. 5), as well as VGLUT1 + /GAD1 + neurons (Fig. 5d), 7-to-8 sections were counted for each mouse brain.The total number of co-expressing neurons was calculated by multiplying the number of counted cells by 6.We later determined that by counting 1 in 12 sections and multiplying the number of counted cells by 12, instead of 6, the final result was not different from that obtained by counting 1 section every 6.We therefore adopted this strategy for the rest of the counts (Fig. 1k, Fig. 3d,f, Fig. 4 d,h, Fig. 7c).
To check for equal sampling across experimental groups, we counted the total number of PL mCherry+, TdTomato + or VGLUT1 + neurons and determined that their number was constant across treatment groups (Supplementary Fig. 1d, Supplementary Fig. 5e, Supplementary Fig. 7b, Supplementary Fig. 11j).For TUNEL assay, as well as quantification of DCX + , Ki67 + , and c-fos + cells, we scored 1 in 9 PL sections, consequently counting 4-to-6 sections for each mouse brain.To quantify c-fos, GFP, and TH expression in the VTA, we collected 1 in 6 sections from Bregma -2.92 mm to Bregma -3.88 mm, according to the Paxinos Mouse Brain Atlas, thus quantifying 4-5 sections per brain.

Quantification of VGAT and VGLUT1 mRNA expression
To quantify the expression level of VGLUT1 and VGAT mRNA, we used multiplex RNAscope against mCherry, GAD1, and either VGLUT1 or VGAT in sections from VGLUT1 CRE ::mCherry mice.The RNA signals for mCherry and GAD1 were fully amplified to allow clear detection of mCherry + and/or GAD1 + cell boundaries.To facilitate quantification of mRNA expression levels, we obtained unsaturated, puncta-like RNA-signals for VGLUT1 and VGAT.
After staining, 4-to-7 optical sections (1 µm z step) of each physical section were examined and regions of interest (ROIs) were drawn around the boundaries of mCherry + , GAD1 + and mCherry + /GAD1 + cells using the optical section in which the cross-sectional area of the cell was the largest.
VGLUT1 or VGAT expression was quantified using Image-J/Fiji as percent of the ROI occupied by VGLUT1 or VGAT RNA fluorescent signal.For each mouse, we analyzed the ROIs of 25 cells co-expressing mCherry and GAD1, 25 cells expressing only mCherry, and 25 cells expressing only GAD1.These cells were found in sections at different levels of the PL rostrocaudal axis and were distributed across all layers of the PL.

Stereotaxic injections
4-5-week-old mice were deeply anesthetized using 3-4% vaporized Isoflurane and head-fixed on a stereotaxic apparatus (David Kopf Instruments Model 1900) for all stereotaxic surgeries.Anesthesia was maintained throughout the procedure at a level that prevented reflex response to a tail/toe pinch, using a continuous flow of 1-2% vaporized Isoflurane.Eye drops (Puralube Vet Ointment, Fisher Scientific, 2024927) were placed in each eye to prevent them from drying out, and vitals were checked every 10 min.
An incision was made to expose the bregma and lambda point of the skull.A 1 mm drill was used to perforate the skull at the desired coordinates.Stereotaxic coordinates for the injection sites were determined using the Paxinos Brain atlas and adjusted experimentally.Using a syringe pump (PHD Ultra™, Harvard apparatus, no.70-3007) installed with a microliter syringe (Hampton, 1482452A) and capillary glass pipettes with filament (Warner Instruments, G150TF-4), we infused the brain with 500 nl (for injections in the PL) or 1 µl (for injections in the VTA) of AAV solution for each injection site at a rate of 100 nl/min.
After injection of the PL, the pipette was left in place for 8 min to allow diffusion of the virus.When the surgery was completed, the scalp was disinfected with betadine and sutured with tissue adhesive glue (Vetbond tissue adhesive, 1469SB).For post-op pain treatment, mice received an injection of Buprenorphine SR (0.5 mg/kg) or Ethiqa XR (3.25 mg/kg).
When targeting the VTA, we injected the AAV solutions bilaterally (from bregma: AP, -3.2 mm; ML, ±0.5 mm; DV, -4.0 mm from the dura).After injection of the VTA, the pipette was left in place for 16 min to allow diffusion of the virus.Mice that later received optogenetic VTA stimulation were implanted during the same surgical procedure with a fiber optic cannula with Ceramic Ferrule (RWD Life Sciences, R-FOC-F200C-39NA).We used the same coordinates used for AAV injection with the following modifications: the fiber was implanted at a 10° angle, and the DV coordinate was reduced to -3.9.To secure the implant to the skull, the skull was covered with a layer of OptiBond XTR Primer (OptiBond XTR Bottle Primer -5 ml Bottle.Self-Etching) followed by OptiBond XTR Bottle Universal Adhesive (OptiBond XTR Bottle Universal Adhesive 5 ml Bottle.Self-Etching, Light-Cure).Finally, a thick layer (up to 0.5 cm thick) of Nano-optimized Flowable Composite (Tetric EvoFlow A2 Syringe -Nano-optimized Flowable Composite 1 -2 Gram) was used to create a scaffold and secure the optic fiber to the skull.Polymerization of OptiBond XTR Primer, Adhesive, and Flowable Composite was achieved with dental LED light (Fencia Premium Silver LED Light, 5W).

Optogenetic stimulation
The first stimulation session occurred 4-5 weeks after surgery.After being moved to the room where stimulation was performed, mice were acclimatized to the room for 1 h.Stimulation occurred in the home cage while the mouse was allowed to move.The ceramic ferrule protruding from the animal's head was coupled to a DPSS blue light laser (473 nm Blue DPSS Laser with Fiber Coupled, BL473T3-050FC, with ADR-700A Power Supply; Shanghai Laser & Optics Century) via custom-made patch cords.Patch cords were assembled by epoxying optical fibers (200 μm, 0.39 numerical aperture, Thorlabs, FT200EMT) to Fiber Connectors FC/PC with Ceramic Ferrule (Thorlabs, 30140E1) and polishing the optic fiber with a fiber polishing kit (Thorlabs) to achieve a minimum of 85% transmission.The laser power was measured before each mouse/stimulation session using a Compact Power and Energy Meter Console, Digital 4" LCD (Thorlabs, PM 100D).Before each mouse/stimulation session, we attached an unused fiber optic cannula to the optic cable and measured the laser power at the tip of the fiber optic cannula using the Compact Power and Energy Meter Console, Digital 4" LCD and adjusted the laser power to 8.5 ± 1 mW at the tip of the fiber optic cannula.Each session of optogenetic stimulation lasted 1h, during which 80 sets of laser stimulation, each in turn consisting of 30 bursts of 5 pulses of 4 ms at 20 Hz, were delivered (modified from 32 ).For c-fos quantification experiments, mice received a single session of optogenetic stimulation and were sacrificed 1 h after the end of the session.For the 10-days stimulation protocol, mice received a daily session of 1 h stimulation between 11 am and 5pm and were sacrificed 2 days after the last stimulation.

Drug-induced locomotor sensitization
Locomotor activity was measured as total distance traveled in the home cage during the 90 min immediately after a single injection of PCP, METH or saline.Mouse movements were recorded using a camera suspended 2 m above the home cage and were automatically scored using AnyMaze 5.For experiments in which mice received either clozapine treatment or chemogenetic manipulation of PL activity after the end of PCP-, METH-or saline-treatment, locomotor quantification was performed at the end of the experimental timeline, two days after the spontaneous alternation task (SAT).In this case, baseline locomotor activity was first recorded in the home cage for 90 min, after which all mice received an acute injection of PCP 10 mg/kg (PCP challenge) or METH 1 mg/kg (METH challenge).

Novel object recognition test (NORT)
The novel object recognition test was used to assess recognition memory performance and performed as described 61

Spontaneous alternation task (SAT)
The spontaneous alternation task was performed as described 19,63 and used to measure immediate working memory performance.The test was conducted in a T-shaped, dark gray plastic maze, composed of 3 arms 30 cm long, 9 cm wide and 20 cm tall.Mice naïve to the maze were placed at the end of one of the 3 arms, and left free to explore the maze for 8 min.Exploration was recorded with a camera suspended 1 m above the apparatus, and the series of arm entries was scored using BORIS software by investigators blinded to the mouse's previous treatment history.The mouse was considered to have entered an arm of the maze only when both the forelimbs and hindlimbs were completely within the arm.Alternation was defined as successive entries into the three arms of the maze on overlapping triplet sets.Alternation percent was calculated as ratio of actual alternation to total possible alternations (defined as the total number of arm entries minus two), multiplied by 100.Total arm entries were also scored to compare the degree of exploration across treatment groups.

Supplementary Files
This is a list of supplementary les associated with this preprint.Click to download.113023SupplementaryFiguresandlegendsCLEAN.pdf Fig.4c,d,g,h, Fig.5c,d, Fig.7c,d).Because VGLUT1 expression is decreased but not completely lost in

2 (
Stoelting, WoodDale, IL, USA).In experiments aimed at preventing drug-induced changes in neurotransmitter phenotype by combining PCP, METH, or saline treatment with chemogenetic activation of PV+ interneurons, or by suppressing the gain of GABA with shGAD1 interference, locomotor activity was measured on the first day of treatment (i.e.immediately after the first drug/saline injection, DAY1) and on the last day of treatment (i.e.immediately after the tenth drug/saline injection, DAY10).
with the modifications outlined below.The test was conducted in an open field box made of dark gray plastic 40 x 40 x 21 cm, dimly illuminated(30-40 lux).In the 3 days before the test, mice were acclimatized to the open field box for 5 min/day.24 h after the last acclimatization session, mice were placed back in the open field with two identical objects and allowed to explore the 2 objects for 10 minutes (familiarization phase).24 h after the familiarization phase, memory retention was tested by placing the mouse back in the open field where one of the familiar objects had been replaced by a novel object (test phase).The mouse was left to explore the 2 objects for 12 minutes and video recordings were collected with a camera (Sony HDR-CX405) suspended 1 m above the apparatus.Because mice (including salinetreated controls) often failed to reach an accepted exploration criterion 61 (20 s of total exploration in the 10 min of the test), we extended the time available for exploration from 10 to 12 min, and mice that did not reach the criterion in less than 12 min were excluded.Mice were also excluded if they moved or overturned one of the two objects.The time spent exploring each of the two objects (novel and familiar) was manually scored using BORIS software (version v. 5.0.1 62 ) by investigators blinded to the mouse's previous treatment-history.The mouse was considered to be exploring the object when it had its nose directed toward the object at a distance of <1 cm.Chewing or climbing on the object was not considered an exploratory behavior.Exploratory behavior was scored until the mouse reached the criterion of 20 s spent exploring the two objects.We calculated a recognition index (RI) as the percent of time spent exploring the novel object relative to the total time spent exploring both objects (novel and familiar) [RI=time exploring novel object/(time exploring novel object + time exploring familiar object)].Total exploration time and time to reach the criterion were also recorded.
Statistical analyses of the data were performed using Prism 9 software.The Excel Real Statistics package was used for non-parametric Aligned Rank Transform (ART) ANOVA.The Shapiro-Wilk test was used to assess whether the data were normally distributed.All statistical tests were two-tailed.Details about the number of animals and statistical tests used for each experiment are reported in the figure legends and in the supplementary figure legends.Measurements were always taken from distinct samples, except for measurements of locomotor activity on first and last day of treatment, which were on the same mice.Means and SEMs are reported for all experiments.

Fig. 5 |
Fig. 5 | Suppression of drug-induced PL hyperactivity during drug delivery prevents the change in transmitter identity and prevents drug-induced behavioral alterations.a Experimental protocol.b Expression of PSAML-5HT3HC-GFP in the PL.Scale bar, 500 µm.c, VGLUT1 and GAD1 expression in the PL following saline/drug treatment combined with chemogenetic activation of PV + neurons.(Blue rectangles) neurons co-expressing VGLUT1 and GAD1 illustrated at high magnification in the insets.Scale bar, 50 µm.d Quantification of (c) (n=4-5 mice).e-h Chemogenetic activation of PV + neurons during drug treatment prevents the deficits in the NORT and SAT induced by both PCP and METH (n=4-9mice).i-lThe number of PL VGLUT1+/GAD1+ neurons is negatively correlated with performance in the NORT and SAT (n=2-4 mice).Statistical significance (*P<0.05,**P<0.01,***P<0.001,****P<0.0001) was assessed using two-way ANOVA followed by Tukey's test (d-h) and linear regression and Pearson's correlation analysis (i-l).Data are presented as mean ± SEM.

998Fig. 7 |
Fig. 7 | Suppression of drug-induced PL hyperactivity after the end of drug-treatment reverses the