Phencyclidine Inhibits the Activity of Thalamic Reticular Gamma-Aminobutyric Acidergic Neurons in Rat Brain

Author's Accepted Manuscript

Phencyclidine inhibits the activity of thalamic reticular GABAergic neurons in rat brain Eva Troyano-Rodriguez PhD, Laia Lladó-Pelfort PhD, Noemí Santana PhD, Vicent Teruel-Martí PhD, Pau Celada PhD, Francesc Artigas PhD

www.elsevier.com/locate/bps

PII: DOI: Reference:

S0006-3223(14)00386-2 http://dx.doi.org/10.1016/j.biopsych.2014.05.019 BPS12227

To appear in:

Biological Psychiatry

Cite this article as: Eva Troyano-Rodriguez PhD, Laia Lladó-Pelfort PhD, Noemí Santana PhD, Vicent Teruel-Martí PhD, Pau Celada PhD, Francesc Artigas PhD, Phencyclidine inhibits the activity of thalamic reticular GABAergic neurons in rat brain, Biological Psychiatry, http://dx.doi.org/10.1016/j.biopsych.2014.05.019 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Phencyclidine inhibits the activity of thalamic reticular GABAergic neurons in rat brain Troyano-Rodriguez E1,2,3, Lladó-Pelfort L1,2,3, Santana N2,1,3 Teruel-Martí V4, Celada P3,1,2#, Artigas F1,2,3# 1

Department of Neurochemistry and Neuropharmacology, Institut d’Investigacions Biomèdiques de Barcelona (IIBB-CSIC) 2

Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM) 3

4

Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS)

Dept. Anatomia i Embriologia Humana, Facultat de Medicina, Universitat de València

Running title. Phencyclidine and reticular thalamus #

The last two authors contributed equally to this study

Corresponding

author:

Francesc

Artigas,

PhD

Dept.

of

Neurochemistry

and

Neuropharmacology, IIBB-CSIC (IDIBAPS), Rosselló, 161, 6th floor, 08036 Barcelona, Spain. Phone: +3493-363 8315; Fax: +3493-363 8301; e-mail: [email protected]

Word account: Abstract: 249 words; full manuscript: 3691 words 5 Figures, 1 Supplemental material

1

Abstract Background: The neurobiological basis of action of non-competitive NMDA receptor (NMDA-R) antagonists is poorly understood. Electrophysiological studies indicate that phencyclidine (PCP) markedly disrupts neuronal activity -with an overall excitatory effectand reduces the power of low frequency oscillations (LFO, <4 Hz) in thalamocortical networks. Since the reticular nucleus of the thalamus (RtN) provides tonic feed-forward inhibition to the rest of thalamic nuclei, we examined the effect of PCP on RtN activity, under the working hypothesis that NMDA-R blockade in RtN would disinhibit thalamocortical networks. Methods: Assessment of drug effects (PCP followed by clozapine) on the activity of RtN (single unit and local field potential recordings) and prefrontal cortex -PFC(electrocorticogram) in anesthetized rats. Results: PCP (.25-.5 mg/kg IV) reduced the discharge rate of 19/21 RtN neurons to 37% of baseline (p < .000001) and the power of LFO in RtN and PFC to ~20% of baseline (p < .001). PCP also reduced the coherence between PFC and RtN in the LFO range. A low clozapine dose (1 mg/kg IV) significantly countered the effect of PCP on LFO in PFC, but not in RtN, and further reduced the discharge rate of RtN neurons. However, clozapine administration partly antagonized the fall in coherence and phase-locking values produced by PCP. Conclusions: PCP activates thalamocortical circuits in a bottom-up manner by reducing the activity of RtN neurons, which tonically inhibit thalamic relay neurons. However, clozapine reversal of PCP effects is not driven by restoring RtN activity and may involve a cortical action.

Keywords: antipsychotic drugs; clozapine; NMDA receptor antagonists; psychotic symptoms; thalamocortical networks; schizophrenia

2

Introduction The non-competitive NMDA receptor (NMDA-R) antagonists phencyclidine (PCP) and ketamine are extensively used as pharmacological models of schizophrenia due to their ability to evoke and aggravate schizophrenia symptoms in healthy individuals and schizophrenic patients, respectively (1, 2). PCP, ketamine and dizocilpine (MK-801) also evoke behavioral alterations in experimental animals, which are partly or totally antagonized by antipsychotic drugs (3-5). The brain networks and cellular elements involved in these actions are not fully understood, yet neuroimaging studies indicate that these agents increase the activity of prefrontal cortex (PFC) and cingulate areas (6-10). Work in experimental animals has enabled to identify the actions of NMDA-R antagonists in various brain areas, including the PFC, the thalamus and the hippocampal formation (5, 11). Previous studies showed that PCP markedly altered the activity of pyramidal neurons in the medial PFC (mPFC) (12) and of thalamic relay neurons of the centromedial (CM) and mediodorsal (MD) thalamic nuclei (13) in anesthetized rats. PCP had a mixed action in PFC and CM/MD neuronal activity, yet with an overall excitatory effect, resulting from the higher percentage of neurons excited (45% in PFC, 57% in CM/MD) and the large increase in neuronal discharge (286% in PFC, 424% in CM/MD). Concurrently, PCP reduced low frequency oscillations (LFO, <4 Hz) in mPFC and CM/MD and increased c-fos expression in thalamocortical areas (12, 13). PCP effects were blocked by antipsychotic drugs (12-14), which supports the association of these alterations to psychotic symptoms. The primary cellular/regional site of action for these PCP effects is unclear. Given the reciprocal connectivity of the PFC and CM/MD (15, 16), PCP blockade of NMDA-R inputs onto PFC GABA interneurons, as observed for MK-801 (17), would disinhibit projection pyramidal neurons and increase corticothalamic inputs (top-down process). PCP might also block NMDA-R inputs onto GABAergic neurons in basal ganglia structures and/or the reticular nucleus of the thalamus (RtN), which provide tonic feed-forward inhibition to thalamic relay neurons (bottom-up process). We conducted the present study to examine whether PCP inhibits the activity of RtN neurons in vivo, which -according to the working

3

hypothesis- would reduce GABA inputs onto CM/MD nuclei and increase thalamocortical activity, as observed (12, 13).

Methods Animals and treatments Male albino Wistar rats weighting 250-320 g (Iffa Credo, Lyon, France) were used in this study. Animal care, treatments and recording procedures were essentially as described (12, 13) (see also Supplemental material-Methods). Animal procedures followed European Union regulations (OJ of EC L358/1, December 18, 1986). Rats were deeply anaesthetised (chloral hydrate, 400 mg/kg intraperitoneal -IP). Constant level of anesthesia was obtained by infusing 50-70 mg/kg/h chloral hydrate IP. The femoral vein was cannulated for intravenous (IV) drug administration. Animals were implanted with a 2-mm screw fixed in the skull skimming the cortical brain surface (duramater was accurately removed) for electrocorticogram (ECoG) measurements in PFC (AP+3.2, L-.5 and DV brain surface). RtN neurons (Fig. 1) were recorded with glass micropipettes filled with 2 mol/L sodium chloride containing 2% pontamine sky blue (Avocado Research Chemicals Ltd, Lancaster, United Kingdom). Descents were carried out in the ipsilateral RtN (AP-1.4 to -3.14, L-1.8 to -3.8 and DV-5 to -6.5 mm below brain surface) according to Paxinos and Watson (18). Drugs PCP hydrochloride (.25-.5 mg/kg IV) was from Sigma/RBI (Natick, MA) and clozapine (CLZ; 1 mg/kg IV) was from Tocris (Bristol, UK). Doses were chosen from previous experiments (12, 13). Electrophysiological experiments. Electrophysiological procedures were essentially as described in (13). For single-unit, local field potential (LFP) and ECoG recordings, the original signal was amplified and filtered between 30 Hz and 1 KHz, .1-100 Hz and .1-200 Hz, respectively. Sampling rate was 25 kHz 4

for single unit recordings and 2.5 kHz for local field potential and ECoG recordings. Data storage was made with Spike2 software (CED, Cambridge, UK). RtN neurons were identified by their electrophysiological characteristics (19-24). RtN neurons exhibited firing rates higher than thalamic relay cells and great variability in burst characteristics (see Results). Likewise, histological verification of the recording site was carried out. Neurons outside the RtN were excluded from the study. Histological procedures Tissue treatment for the localization of pontamine has been described elsewhere (13) (see also Supplemental material-Methods). Briefly, tissue sections adjacent to the pontamine dot were thaw-mounted onto individual slides. The two best consecutive sections containing an intense pontamine dot were processed, one for neutral red staining and the other one for in situ hybridization of parvalbumin (PV) to label RtN cells (25). Low-magnification images of the two sections were then obtained and merged with Adobe Photoshop to verify electrode localization (Fig. 1). Parvalbumin (PV)-positive cells were identified by an oligodeoxyribonucleotide probe complementary to bases 205-249 (GenBank accession number X63070.1; Isogen Bioscience BV (De Meern, The Netherlands)). Labelling of the probes and in situ hybridization procedures were carried out as described previously (26, 27) (see Supplemental materialMethods). Data analysis Electrophysiological data analysis was performed off-line (12, 13). A detailed description is given in Supplemental material-Methods. Briefly, each treatment condition (basal, PCP and CLZ/saline) was recorded for 5 minutes, of which the last 2-min periods were used for data analysis. Bursts analysis of RtN neurons was made as previously described in (24). Burst strength was analysed using inter-spike intervals (ISI) and defined as a burst index (BI): ISI<10 ms divided by ISI<200 ms.

5

For LFP and ECoG recordings, signals were downsampled 16 times and raw data were imported to the Matlab development environment (The MathWorks, Natick, MA, USA) for off-line analysis, using built-in and self-developed routines. LFP and ECoG analyses were performed in the same time segments than spike analysis. To compare common oscillatory patterns between signals, wavelet coherence was firstly used (28) as a measure of power and phase difference stationarity. On the other hand, phase-locking values were defined specifically (29) as a measure of the stationarity of the phase differences in a temporal window and therefore of the phase-depending synchrony (see Supplemental material-Methods). Due to the specificity of the latter measure, this was calculated considering only the dominant oscillation. Likewise, we analysed the phase-locking between the dominant 1 Hz oscillation (.5-2 Hz) in PFC and discharge of RtN neurons by using the first spike of any burst event. The difference between each treatment condition was evaluated with paired Student’s t-test and one/two-way ANOVA followed by Newman-Keuls as post hoc tests, as appropriate. Data are mean ± SEM and significance has been set at p < .05.

Results Firing patterns of RtN neurons The effects of PCP were examined in a total of 21 RtN neurons (one per rat), localized in different sectors of the RtN, with a mean DV coordinate of 5.87 ± .09 mm (n = 21) (Fig. 1). The spontaneous discharge rate of recorded neurons was 16.4 ± 1.2 spikes/s (n = 21), nearly 10-fold that of CM/MD neurons (1.7 ± .2 spikes/s) (13). Unlike thalamic relay neurons, RtN neurons fired in a variety of combinations of tonic and burst firing patterns. Some neurons fired exclusively in tonic mode while others fired only in burst mode. The average burst index was .37 ± .04 (n = 21), within the range of reported values (24). Figure 1C-1E shows representative examples of firing patterns of RtN neurons. Effect of PCP on the discharge of RtN neurons. 6

Overall, PCP administration (.25-.5 mg/kg IV) markedly reduced the discharge rate of RtN neurons, from 16.4 ± 1.2 to 6.7 ± .9 spikes/s (p < .000001, paired Student’s t–test, n=21). Nineteen neurons were inhibited by PCP, from 16.9 ± 1.2 to 6.2 ± .9 spikes/s (37 ± 5 % of baseline; p < .000001), and two neurons did not respond (from 11.6 ± .7 to 11.8 ± .5 spikes/s) despite PCP simultaneously reduced the power of LFO in PFC (see below). The effect of PCP persisted for at least 10 min (from 13.8 ± 1.5 to 4.4 ± 1.2 and 3.4 ± .9 spikes/s in basal conditions, 5 and 10 minutes after PCP administration, respectively; one-way ANOVA: F2,10 = 37.72, p < .001; n=6). Figure 2A shows a representative example of the effect of PCP on the discharge of a RtN neuron. PCP did not alter the firing pattern of inhibited neurons. Burst index was .36 ± .04 and .39 ± .08 in basal conditions and after PCP, respectively (n.s., paired Student’s t-test, n = 19).

Effect of PCP on LFO in RtN and PFC In parallel with the effects on the discharge of RtN neurons, PCP significantly reduced the power of LFO (.2-4Hz) in RtN and PFC to ca. 20% of baseline (from 1.06 ± .14 to .20 ± .02 µV 2 in RtN; from 1.06 ± .21 to .22 ± .05 µV2 in PFC; t-tests p < .001; n = 13 both). This effect also persisted for ≥10 min after PCP administration (one-way ANOVAs; RtN: F2,8 = 15.14, p < .002; PFC: F2,8 = 6.2, p < .05; n = 5 both) (Fig. 2). PCP administration significantly decreased coherence between PFC and RtN in the .24 Hz range, from .6 ± .01 to .4 ± .02 (p < .0001; paired Student’s t-test; n = 13). When restricting the analysis to the dominant 1 Hz oscillation (.5-2 Hz), phase-locking values decreased from .57 ± .05 to .28 ± .04 (t-test; p < .002; n = 13). We also found a significant phase-locking between the 1 Hz oscillation in PFC and the initiation of bursting activity in RtN which was lost after PCP treatment (basal = .31 ± .05, PCP = .19 ± .03, t-test p < .05; n = 8). Effect of clozapine on PCP-induced alteration 7

The administration of a low dose of clozapine (1 mg/kg IV, 5 min after PCP) significantly countered the effect of PCP on LFO in PFC (.92 ± .18 to .16 ± .04 and .53 ± .1 µV2; basal, PCP and PCP + CLZ values, respectively; one-way repeated measures ANOVA: F2,14 = 17.23, p < .001; n = 8) (Fig. 3). However, CLZ had a completely different effect on RtN activity, not significantly antagonizing PCP effect on LFO (.95 ± .13 to .22 ± .03 and .35 ± .06; basal, PCP and PCP + CLZ values, respectively; one-way ANOVA: F2,14 = 34.81, p < .0001; n = 8). Moreover, CLZ administration further reduced the discharge rate of RtN neurons (from 17.9 ± 1.5 to 6.8 ± .8 and 1.0 ± .3 spikes/s in basal, PCP and PCP + CLZ conditions, respectively; one way ANOVA: F2,22 = 96.50, p < .001; n = 12) (Fig. 3). Likewise, CLZ markedly reduced the discharge rate of RtN neurons when administered alone to previously untreated rats (from 18.8 ± 2.0 to 7.1 ± 2.8 spikes/s; p < .04; paired Student’s t-test; n = 5) (Supplemental material-Fig. S1). The CLZ-induced reduction in discharge rate was accompanied by an increased burst index (from .33 ± .05 to .39 ± .11 and .71 ± .10 in basal, PCP and PCP + CLZ conditions respectively; F2,22 = 4.65, p < .03; n=12, with significant post-hoc differences between CLZ vs basal and PCP). Interestingly, CLZ administration also reversed -yet partially- the effect of PCP on coherence values (from .59 ± .02 to .43 ± .03 and .5 ± .03; basal, PCP and PCP + CLZ values, respectively; one-way ANOVA: F2,14 = 16.48, p < .001; n = 8) and phase-locking values between PFC and RtN oscillations (from .67 ± .05 to .28 ± .06 and .51 ± .06; basal, PCP and PCP + CLZ values, respectively; one-way ANOVA: F2,14 = 15.28, p < .001; n = 8) (Fig. 4). CLZ also significantly antagonized the PCP-induced reduction of phase-locking between the 1 Hz oscillation in PFC and the initiation of bursting activity in RtN (PCP = .19 ± .02, PCP + CLZ = .52 ± .07; t-test p < .003, n = 6). Similarly, the administration of CLZ alone evoked a significant increase of this phase-locking value (basal = .33 ± .07, CLZ = .50 ± .11, t-test p < .05; n = 5).

Discussion 8

The present study shows that acute PCP administration markedly reduces the discharge rate of GABAergic neurons in the RtN and the power of LFO in the PFC (as reported) (12) and RtN. Together with previous studies (12, 13), the present observations indicate that PCP activates thalamocortical networks by markedly decreasing the activity of RtN neurons that feed-forward inhibit thalamic relay neurons (Fig. 5). The inability of CLZ to reverse the PCP effect on RtN discharge suggests that the normalization of thalamocortical activity by CLZ (12, 13) is not driven by an action on RtN neurons and may involve changes in mPFC, as suggested earlier (12). In contrast to the mPFC and CM/MD thalamic nuclei, where PCP preferentially activated excitatory neurons, here we show that PCP inhibited the majority of the recorded GABAergic RtN neurons (19/21, two neurons did not respond). These observations provide novel in vivo evidence supporting that non-competitive NMDA-R antagonists block NMDA-R in GABAergic elements, thus disinhibiting the rest of thalamic nuclei (10, 30, 31). Effect of PCP on RtN firing rate Using the same experimental conditions in the same setting, we showed that PCP markedly activated mPFC pyramidal neurons (12) and CM/MD thalamic neurons (13) and, at the same dose, it dramatically reduced the discharge of RtN GABAergic neurons (present data). This is perhaps the most remarkable observation of the present study. Moreover, mPFC pyramidal and CM/MD thalamic neurons were excited or inhibited by PCP (likely reflecting indirect and direct effects of NMDA-R blockade, respectively), whereas PCP inhibited most RtN neurons (19/21, a 90%). This, together with the large fall in RtN neuronal discharge (from 16.4 to 6.8 spikes/s), supports the view that the inhibition of RtN cells by PCP is a primary effect driving the activation of thalamocortical networks (12, 13). Effect of PCP on LFO In parallel to decreasing RtN neuron discharge, PCP markedly reduced the power of LFO in PFC and RtN. These actions may underlie psychotomimetic effects (32), since i) they are shared by PCP and serotonergic hallucinogens, and ii) they are reversed by antipsychotic drug treatment (12, 33, 34). Previous data indicate that NMDA-R blockade by the competitive antagonist APV leads to delta bursting in brain slices containing the RtN (35) 9

whereas in the present experimental conditions (in vivo, chloral hydrate anesthesia), PCP reduced LFO in RtN. While both studies support the involvement of RtN in NMDA-R blockade, the distinct effect likely derives from the different experimental conditions used. Interestingly, PCP reduced the coherence between PFC and RtN, in agreement with the disrupted thalamocortical activity. The great effect of PCP on coherence and phase-locking when considering the dominant 1 Hz oscillation suggests a major role of this frequency in the functional connectivity between the PFC and RtN. Cellular/regional site of action of PCP Previous in vivo studies reported an opposite action of MK-801 on the discharge rate of putative pyramidal neurons (increase) and GABAergic interneurons (decrease) in rat mPFC (17), suggesting that the blockade of NMDA-R in GABAergic interneurons disinhibits pyramidal neurons. In agreement with these electrophysiological observations, PCP increased c-fos expression in vGluT1-positive (pyramidal) but not in GAD65/67-positive GABAergic interneurons (12). Given the reciprocal connectivity between the PFC and CM/MD (15, 16) and the top-down control exerted by the PFC on most cortical and subcortical brain areas (36), PCP might alter thalamocortical activity by the same mechanism. While this possibility has not been examined so far (and cannot be ruled out), the present study shows that PCP activates thalamocortical circuits through a bottom-up mechanism, by reducing the activity of RtN GABAergic neurons, which provide feed-forward inhibition to the rest of thalamic nuclei (37, 38) (Fig. 5). Both ways of action (top-down and bottom-up) are not mutually exclusive and may simultaneously occur, as previously discussed (13). Hence, PCP produced a generalized c-fos activation of thalamic (CM, MD, paracentral, centrolateral, reuniens, rhomboid, etc.) and cortical areas (PFC, somatosensory and retrosplenial cortices, etc.), with the conspicuous exception of RtN, which led us to consider this GABAergic structure as a primary site of action for PCP (12, 13). This regional pattern has also been observed with MK-801 (39). The thalamus provides most vGluT2-positive inputs to PFC and somatosensory cortex (40) and PCP markedly increased c-fos expression in layers IV and VI of somatosensory cortex, indicating the existence of an increased thalamocortical (layers IV and VI) and 10

corticothalamic (layer VI) functional connectivity (for review see (41)). Also, PCP increased cfos expression in a narrow band of cells between layers III and V in PFC (12), which receive thalamic inputs (15) (rat PFC lacks layer IV). Interestingly, unlike in mPFC (12), GABAergic neurons in somatosensory and retrosplenial cortices expressed c-fos in response to PCP treatment, which agrees with the dual projection of thalamocortical fibers to cortical glutamatergic and GABAergic neurons (42) and the preferential thalamocortical inputs on fast-spiking cortical interneurons (43). Hence, the contribution of top-down and bottom-up mechanisms may vary in different cortical areas. The reasons for the preferential action of non-competitive NMDA-R antagonists on GABAergic elements are not fully understood and may include differences in the role of AMPA and NMDA-R in the generation of excitatory postsynaptic potentials in pyramidal and GABAergic neurons and a different subunit composition of NMDA-R in excitatory and inhibitory neurons (44). In line with the present observations, the competitive NMDA-R antagonist APV hyperpolarized RtN (but not PFC) cells in vitro (35). Likewise, functional differences may also account, given the greater discharge rate of RtN neurons (and of fastspiking cortical interneurons) compared to thalamic and cortical excitatory neurons. The higher discharge rate of some local-circuit and projection GABAergic neurons may lead to a higher accessibility of non-competitive NMDA receptor antagonists to their binding site inside the NMDA channel during more prolonged depolarized periods. Effects of clozapine Unlike in mPFC and CM/MD nuclei (12, 13), the administration of CLZ did not reverse the effect of PCP on RtN neuron discharge. On the contrary, CLZ further reduced RtN neuronal discharge. A tentative explanation for this CLZ effect is the following one. Corticothalamic inputs account for 70% of RtN excitatory inputs (38) and CLZ reduced the PCP-induced pyramidal discharge, which should decrease cortical excitatory inputs onto RtN. Further, a direct inhibitory action of CLZ on RtN neurons cannot be excluded given the expression of α1A/D-adrenergic receptors by RtN neurons (45) and the moderate affinity of CLZ for this receptor subtype (~40 nM (46)). Likewise, RtN neurons also express 5-HT1A receptors (47), where CLZ behaves as a functional agonist (48, 49). The possibility that CLZ exerts direct 11

actions on RtN neurons is supported by pilot experiments showing that CLZ markedly reduced the discharge of RtN neurons when administered alone (Supplemental material-Fig. S1). Further studies are required to assess the relative weight of these two factors in the effect of CLZ on RtN activity. Interestingly, CLZ (alone and after PCP) increased the phase-locking between the 1 Hz oscillation in PFC and the initiation of burst events in RtN above baseline. Given the marked reduction of RtN neuronal discharge produced by CLZ, the increase in phase-locking may be related to the removal of the α1-adrenoceptor excitatory tone, thus leaving RtN neurons more dependent on cortical excitatory inputs. Functional implications The present study indicates that GABAergic neurons of the RtN are a primary target for the non-competitive NMDA-R antagonist PCP. Given the functional connectivity between the RtN and the rest of thalamic nuclei, the dramatic reduction in RtN activity produced by PCP likely results in disinhibition of thalamocortical activity, as observed (12, 13) (Fig. 5). Interestingly, a very recent PET scan study shows that systemic ketamine administration increases metabolic activity in PFC and reduces it in the RtN of mice (50), in full agreement with present and past observations on PCP. The RtN, -together with the substantia nigra reticulata- is a key structure in the function of basal ganglia circuits, since both provide tonic, feed-forward inhibition to thalamic relay neurons projecting to cortical and limbic structures. Hence, a reduced number of GABAergic neurons in these nuclei concentrate large amounts of information via descending cortico-striatal pathways and expand it further via thalamocortical pathways. In particular, the RtN serves a nexus to facilitate the interaction between functionally-related cortical and thalamic areas, following a precise anatomical connectivity (51-54). As an example, cognitive and emotional information pathways converge in primate RtN (55). Thus, the marked inhibition of RtN activity by PCP likely has a deep impact on sensory, cognitive and emotional signals, in agreement with its widespread effects on mood, cognition and perception. 12

The dramatic effect of PCP on RtN neuronal discharge agrees with the marked sensitivity to competitive NMDA-R blockade with APV (35) and with more recent observations indicating that NMDA and mGlu2 receptors tonically excite RtN neurons (56). Notwithstanding a possible effect of PCP on cortical GABAergic interneurons to modulate corticothalamic pathways, the marked effects of PCP on RtN activity and the functional role of RtN in the control of multiple cortical areas. One of the most characteristic effects of noncompetitive NMDA-R antagonist is motor hyperactivity, which is partly or totally antagonized by antipsychotic drugs (3-5). In line of the present observations, the restoration of the GABAergic transmission in the anterior thalamic nucleus by local application of the GABAA agonist muscimol prevented the motor hyperactivity induced by the systemic administration of MK-801 (57). In summary, we suggest that NMDA blockade in RtN plays a major role in the psychotomimetic actions of non-competitive NMDA-R antagonists, which leads to disinhibition of thalamocortical circuits involved in motor, sensory and cognitive functions.

Acknowledgements Supported by the Innovative Medicines Initiative Joint Undertaking (IMI) under Grant Agreement N° 115008 (NEWMEDS). IMI is a public/private partnership between the European Union and the European Federation of Pharmaceutical Industries and Associations. Support from the grant PI12/00156 (PN de I+D+I 2008-2011, ISCIIISubdirección General de Evaluación y Fomento de la Investigación. Co-financed by the ERDF (European Union -“A way to build Europe”)) is also acknowledged. Also supported by the Instituto de Salud Carlos III, Centro de Investigación Biomédica en Red de Salud Mental, CIBERSAM. The Researcher Stabilization Program of the Health Department of the Generalitat de Catalunya supports P.C. We thank Mercedes Núñez and Noemí Jurado for skilful technical assistance.

13

Financial disclosure FA declares having received lecture and consultation fees from Lundbeck on antidepressant drugs. He is also PI of grants from Lundbeck and nLife Therapeutics on antidepressant mechanisms. The other authors report no biomedical financial interests or potential conflicts of interest.

14

Figure legends

Figure 1. Histological localization and firing pattern characteristics of recorded RtN neurons. A1-A3: Histological verification of a representative RtN neuron. A1. Neutral red staining of the thalamus showing the pontamine sky blue dot (arrow). A2. In situ hybridization immunohistochemistry of parvalbumin mRNA to label RtN GABAergic neurons in the tissue section adjacent to A1. A3. Merge of the images in A1 and A2 together with the enlargement of the area containing the pontamine dot. B. Scheme showing the localization of pontamine (blue dots) in the 21 experiments, according to the atlas of Paxinos and Watson (18). The small letters A, C, D and E in panel B identify the examples given in the respective panels of the figure. C-E: Firing pattern characteristics of three RtN neurons. C1E1: Examples of interspike Interval (ISI) histograms of three neurons. Note the bimodal distribution in D1. C2-E2: Recordings of the same neurons as in C1-E1 (5 s): C2: Neuron mostly firing in tonic mode. D2: Neuron firing with a burst followed by a long tail of tonic spikes (see bimodal distribution of ISI in D1). E2: Neuron firing mostly in burst mode with occasional single spikes. Black bars under the recordings (300 ms) are expanded below (C3, D3 and E3). The squares (30 ms) in C3-E3 are shown as insets in the respective ISI histograms.

Figure 2. Effects of PCP on firing rate and oscillatory activity. A. Integrated firing rate histogram showing the inhibitory action of PCP (.25 mg/kg IV; arrow) on the discharge of a RtN neuron. Note the marked inhibition of the firing activity, which persisted for at least 15 min. B, C. Color-coded spectrograms (red, high power; blue, low power) showing the concurrent reduction of low frequency oscillations (.2-4 Hz) in the RtN (B) and PFC (C). Note the marked effect on the dominant 1 Hz oscillation (time scale is the same in panels A, B and C). D. Raw LFP (RtN) and ECoG (PFC) recordings corresponding to 10 s periods in baseline conditions and after PCP administration (indicated by the bars below C). E-G. Bar graphs showing the effect of PCP on RtN neuron discharge (E), oscillatory activity of RtN (F), and oscillatory activity in PFC (G). *p < .05 vs. basal, n = 5-6. 15

Figure 3. Effects of PCP and CLZ administration on RtN neuronal discharge and low frequency oscillations. A-C are three different experiments showing the simultaneous reduction of RtN neuronal discharge (integrated firing rate histograms; upper panels) and low frequency oscillations in PFC (color-coded spectrograms, lower panels) after PCP administration, together with the opposite effect of CLZ on both variables: CLZ further reduced RtN neuronal discharge but countered the fall of low frequency oscillations in PFC. Blue bars below the integrated firing rate histograms indicate the 2-min periods corresponding to spectrograms in the lower panels. D-F: Bar graphs showing the average effects of PCP and CLZ on RtN discharge (D), low frequency oscillations in the RtN (E) and low frequency oscillations in PFC (F). Note that CLZ significantly countered PCP effect on low frequency oscillations in PFC but not in RtN (for simplicity, RtN spectrograms are not shown). *p < .001 vs. basal, and p < .001 vs. PCP alone (n = 8-12).

Figure 4. Effect of PCP and CLZ on coherence and phase-locking between RtN and PFC in the low frequency oscillation range. A: Color-coded coherogram (red: high coherence; blue: low coherence) between RtN and PFC in basal conditions and after the administration of PCP and CLZ showing a loss of coherence after PCP administration and a partial reversal after CLZ. Note that the coherence is more marked around the 1 Hz oscillation. B. Phase diagrams in basal (B1), PCP (B2) and PCP + CLZ (B3) conditions. Note the loss of coherence after PCP. CLZ partially countered the loss of coherence. C and D are bargraphs showing respectively, coherence and phase locking values (arbitrary units; 1=fully coherent) in the different experimental conditions. *p < .05 vs. basal p < .03 vs. PCP alone (n=8).

16

Figure 5. Activation of thalamocortical circuits by the action of PCP on the RtN. The blockade of NMDA-R in the RtN (and perhaps other basal ganglia structures such as substantia nigra reticulata -SNR- or ventral pallidum -VP-) would disinhibit thalamic relay neurons leading to increased excitatory thalamocortical inputs in various cortical areas, as observed (12, 13). PCP might also activate cortical and thalamic areas by potential action on cortical GABA interneurons, as observed with MK-801 (see discussion). Redrawn from (13).

17

References 1. Javitt DC and Zukin SR (1991): Recent advances in the phencyclidine model of schizophrenia. Am J Psychiatry 148: 1301-1308 2. Krystal JH, D'Souza DC, Mathalon D, Perry E, Belger A and Hoffman R (2003): NMDA receptor antagonist effects, cortical glutamatergic function, and schizophrenia: toward a paradigm shift in medication development. Psychopharmacology (Berl) 169: 215-233 3. Carlsson M and Carlsson A (1989): Dramatic synergism between MK-801 and clonidine with respect to locomotor stimulatory effect in monoamine-depleted mice. J Neural Transm 77: 65-71 4. Geyer MA, Krebs-Thomson K, Braff DL and Swerdlow NR (2001): Pharmacological studies of prepulse inhibition models of sensorimotor gating deficits in schizophrenia: a decade in review. Psychopharmacology (Berl) 156: 117-154 5. Bubenikova-Valesova V, Horacek J, Vrajova M and Hoschl C (2008): Models of schizophrenia in humans and animals based on inhibition of NMDA receptors. Neurosci Biobehav Rev 32: 1014-1023 6. Breier A, Malhotra AK, Pinals DA, Weisenfeld NI and Pickar D (1997): Association of ketamine-induced psychosis with focal activation of the prefrontal cortex in healthy volunteers. Am J Psychiatry 154: 805811 7. Vollenweider FX, Leenders KL, Scharfetter C, Antonini A, Maguire P, Missimer J, et al. (1997): Metabolic hyperfrontality and psychopathology in the ketamine model of psychosis using positron emission tomography (PET) and [18F]fluorodeoxyglucose (FDG). Eur Neuropsychopharmacol 7: 9-24 8. Holcomb HH, Lahti AC, Medoff DR, Cullen T and Tamminga CA (2005): Effects of noncompetitive NMDA receptor blockade on anterior cingulate cerebral blood flow in volunteers with schizophrenia. Neuropsychopharmacology 30: 2275-2282 9. Rowland LM, Bustillo JR, Mullins PG, Jung RE, Lenroot R, Landgraf E, et al. (2005): Effects of ketamine on anterior cingulate glutamate metabolism in healthy humans: a 4-T proton MRS study. Am J Psychiatry 162: 394-396 10. Vollenweider FX and Kometer M (2010): The neurobiology of psychedelic drugs: implications for the treatment of mood disorders. Nat Rev Neurosci 11: 642-651 11. Celada P, Llado-Pelfort L, Santana N, Kargieman L, Troyano-Rodriguez E, Riga MS, et al. (2013): Disruption of thalamocortical activity in schizophrenia models: relevance to antipsychotic drug action. Int J Neuropsychopharmacol 16: 2145-2163 12. Kargieman L, Santana N, Mengod G, Celada P and Artigas F (2007): Antipsychotic drugs reverse the disruption in prefrontal cortex function produced by NMDA receptor blockade with phencyclidine. Proc Natl Acad Sci U S A 104: 14843-14848

18

13. Santana N, Troyano-Rodriguez E, Mengod G, Celada P and Artigas F (2011): Activation of thalamocortical networks by the N-methyl-D-aspartate receptor antagonist phencyclidine: reversal by clozapine. Biol Psychiatry 69: 918-927 14. Kargieman L, Riga MS, Artigas F and Celada P (2012): Clozapine Reverses Phencyclidine-Induced Desynchronization of Prefrontal Cortex through a 5-HT(1A) Receptor-Dependent Mechanism. Neuropsychopharmacology 37: 723-733 15. Kuroda M, Yokofujita J and Murakami K (1998): An ultrastructural study of the neural circuit between the prefrontal cortex and the mediodorsal nucleus of the thalamus. Prog Neurobiol 54: 417458 16. Groenewegen HJ and Uylings HB (2000): The prefrontal cortex and the integration of sensory, limbic and autonomic information. Prog Brain Res 126: 3-28 17. Homayoun H and Moghaddam B (2007): NMDA receptor hypofunction produces opposite effects on prefrontal cortex interneurons and pyramidal neurons. J Neurosci 27: 11496-11500 18. Paxinos G, Watson C (2005): The Rat Brain in stereotaxic coordinates , 5th ed. San Diego: Elsevier Academic Press. 19. Contreras D, Curro Dossi R and Steriade M (1993): Electrophysiological properties of cat reticular thalamic neurones in vivo. J Physiol 470: 273-294 20. Domich L, Oakson G and Steriade M (1986): Thalamic burst patterns in the naturally sleeping cat: a comparison between cortically projecting and reticularis neurones. J Physiol 379: 429-449 21. Mukhametov LM, Rizzolatti G and Tradardi V (1970): Spontaneous activity of neurones of nucleus reticularis thalami in freely moving cats. J Physiol 210: 651-667 22. Mulle C, Madariaga A and Deschenes M (1986): Morphology and electrophysiological properties of reticularis thalami neurons in cat: in vivo study of a thalamic pacemaker. J Neurosci 6: 2134-2145 23. Steriade M, Domich L and Oakson G (1986): Reticularis thalami neurons revisited: activity changes during shifts in states of vigilance. J Neurosci 6: 68-81 24. Ushimaru M, Ueta Y and Kawaguchi Y (2012): Differentiated participation of thalamocortical subnetworks in slow/spindle waves and desynchronization. J Neurosci 32: 1730-1746 25. Celio MR (1990): Calbindin D-28k and parvalbumin in the rat nervous system. Neuroscience 35: 375475 26. Schmitz GG, Walter T, Seibl R and Kessler C (1991): Nonradioactive labeling of oligonucleotides in vitro with the hapten digoxigenin by tailing with terminal transferase. Anal Biochem 192: 222-231 27. Serrats J, Artigas F, Mengod G and Cortes R (2003): GABAB receptor mRNA in the raphe nuclei: coexpression with serotonin transporter and glutamic acid decarboxylase. J Neurochem 84: 743-752

19

28. Hurtado JM, Rubchinsky LL and Sigvardt KA (2004): Statistical method for detection of phase-locking episodes in neural oscillations. J Neurophysiol 91: 1883-1898 29. Lachaux JP, Rodriguez E, Martinerie J and Varela FJ (1999): Measuring phase synchrony in brain signals. Hum Brain Mapp 8: 194-208 30. Tsai G and Coyle JT (2002): Glutamatergic mechanisms in schizophrenia. Annu Rev Pharmacol Toxicol 42: 165-179 31. Lisman JE, Coyle JT, Green RW, Javitt DC, Benes FM, Heckers S, et al. (2008): Circuit-based framework for understanding neurotransmitter and risk gene interactions in schizophrenia. Trends Neurosci 31: 234-242 32. Celada P, Puig MV and Artigas F (2013): Serotonin modulation of cortical neurons and networks. Front Integr Neurosci 7: 25 33. Celada P, Puig MV, Diaz-Mataix L and Artigas F (2008): The hallucinogen DOI reduces low-frequency oscillations in rat prefrontal cortex: reversal by antipsychotic drugs. Biol Psychiatry 64: 392-400 34. Riga MS, Soria G, Tudela R, Artigas F and Celada P (2014): The natural hallucinogen 5-MeO-DMT, component of Ayahuasca, disrupts cortical function in rats: reversal by antipsychotic drugs. Int J Neuropsychopharmacol: 1-14 35. Zhang Y, Llinas RR and Lisman JE (2009): Inhibition of NMDARs in the Nucleus Reticularis of the Thalamus Produces Delta Frequency Bursting. Front Neural Circuits 3: 20 36. Miller EK and Cohen JD (2001): An integrative theory of prefrontal cortex function. Annu Rev Neurosci 24: 167-202 37. Steriade M (2001): The GABAergic reticular nucleus: a preferential target of corticothalamic projections. Proc Natl Acad Sci U S A 98: 3625-3627 38. Jones EG (2002): Thalamic circuitry and thalamocortical synchrony. Philos Trans R Soc Lond B Biol Sci 357: 1659-1673 39. Vaisanen J, Ihalainen J, Tanila H and Castren E (2004): Effects of NMDA-receptor antagonist treatment on c-fos expression in rat brain areas implicated in schizophrenia. Cell Mol Neurobiol 24: 769780 40. Hur EE and Zaborszky L (2005): Vglut2 afferents to the medial prefrontal and primary somatosensory cortices: a combined retrograde tracing in situ hybridization study [corrected]. J Comp Neurol 483: 351373 41. Shipp S (2007): Structure and function of the cerebral cortex. Curr Biol 17: R443-449 42. Sun QQ, Huguenard JR and Prince DA (2006): Barrel cortex microcircuits: thalamocortical feedforward inhibition in spiny stellate cells is mediated by a small number of fast-spiking interneurons. J Neurosci 26: 1219-1230 20

43. Hull C, Isaacson JS and Scanziani M (2009): Postsynaptic mechanisms govern the differential excitation of cortical neurons by thalamic inputs. J Neurosci 29: 9127-9136 44. Zhang Y, Buonanno A, Vertes RP, Hoover WB and Lisman JE (2012): NR2C in the thalamic reticular nucleus; effects of the NR2C knockout. PLoS One 7: e41908 45. Pieribone VA, Nicholas AP, Dagerlind A and Hokfelt T (1994): Distribution of alpha 1 adrenoceptors in rat brain revealed by in situ hybridization experiments utilizing subtype-specific probes. J Neurosci 14: 4252-4268 46. Sleight AJ, Koek W and Bigg DC (1993): Binding of antipsychotic drugs at alpha 1A- and alpha 1Badrenoceptors: risperidone is selective for the alpha 1B-adrenoceptors. Eur J Pharmacol 238: 407-410 47. Pompeiano M, Palacios JM and Mengod G (1992): Distribution and cellular localization of mRNA coding for 5-HT1A receptor in the rat brain: correlation with receptor binding. J Neurosci 12: 440-453 48. Diaz-Mataix L, Scorza MC, Bortolozzi A, Toth M, Celada P and Artigas F (2005): Involvement of 5HT1A receptors in prefrontal cortex in the modulation of dopaminergic activity: role in atypical antipsychotic action. J Neurosci 25: 10831-10843 49. Cosi C, Waget A, Rollet K, Tesori V and Newman-Tancredi A (2005): Clozapine, ziprasidone and aripiprazole but not haloperidol protect against kainic acid-induced lesion of the striatum in mice, in vivo: role of 5-HT1A receptor activation. Brain Res 1043: 32-41 50. Dawson N, Morris BJ and Pratt JA (2013): Subanaesthetic ketamine treatment alters prefrontal cortex connectivity with thalamus and ascending subcortical systems. Schizophr Bull 39: 366-377 51. Guillery RW, Feig SL and Lozsadi DA (1998): Paying attention to the thalamic reticular nucleus. Trends Neurosci 21: 28-32 52. Pinault D (2004): The thalamic reticular nucleus: structure, function and concept. Brain Res Brain Res Rev 46: 1-31 53. Fuentealba P and Steriade M (2005): The reticular nucleus revisited: intrinsic and network properties of a thalamic pacemaker. Prog Neurobiol 75: 125-141 54. Lam YW and Sherman SM (2011): Functional organization of the thalamic input to the thalamic reticular nucleus. J Neurosci 31: 6791-6799 55. Zikopoulos B and Barbas H (2012): Pathways for emotions and attention converge on the thalamic reticular nucleus in primates. J Neurosci 32: 5338-5350 56. Crabtree JW, Lodge D, Bashir ZI and Isaac JT (2013): GABAA , NMDA and mGlu2 receptors tonically regulate inhibition and excitation in the thalamic reticular nucleus. Eur J Neurosci 37: 850-859

21

57. Lopez Hill X and Scorza MC (2012): Role of the anterior thalamic nucleus in the motor hyperactivity induced by systemic MK-801 administration in rats. Neuropharmacology 62: 2440-2446

22

Fig 1

23

Fig 2

24

Fig 3

25

Fig 4

26

Fig 5

27