Group II metabotropic glutamate receptor (mGlu2 and mGlu3) roles in thalamic processing

As the thalamus underpins almost all aspects of behaviour, it is important to understand how the thalamus operates. Group II metabotropic glutamate (mGlu2/mGlu3) receptor activation reduces inhibition in thalamic nuclei originating from the surrounding thalamic reticular nucleus (TRN). Whilst an mGlu2 component to this effect has been reported, in this study, we demonstrate that it is likely, largely mediated via mGlu3.


| INTRODUCTION
With over 30 distinct nuclear groups, the thalamus co-ordinates the transfer of information to facilitate many aspects of behaviour, from sensation and movement, to cognition and attention (Sherman & Guillery, 2001). It composes of highly organised circuits in conjunction with the neocortex to process and filter incoming and outgoing information. The somatosensory thalamic nucleus of rodents, the ventrobasal thalamus (VB), is a central tool in the study of structurefunction relationships of these thalamocortical circuits due to the highly conserved somatotopic representation of each individual facial vibrissae as a single barreloid (see review: Diamond et al., 2008).
Identifying the structure and function of somatosensory VB microcircuits enables identification of basic principles of thalamic function, which is essential for understanding how the rest of the thalamus operates.
The VB receives excitatory inputs exclusively from the principal sensory trigeminal nucleus via the lemniscal pathway (see review: Diamond et al., 2008). Common to all thalamic nuclei, VB thalamocortical afferents project to layer IV of the cortex and also receive reciprocal modulatory corticothalamic inputs from layer VI, which modulate how driver inputs are transmitted (see review: Jones, 2009). Both thalamocortical and corticothalamic afferents also innervate the associated thalamic reticular nucleus (TRN), which serves to provide both feedback and feedforward inhibition to thalamic nuclei upon thalamocortical and corticothalamic innervation, respectively (Jones, 2009) (Figure 1). It is important to understand how this inhibition is controlled, as its maladaptation has been implicated in several neurophysiological disease states, including schizophrenia (Ferrarelli & Tononi, 2017;Steullet et al., 2018;Young & Wimmer, 2017 (Copeland et al., 2012(Copeland et al., , 2017Cox & Sherman, 1999;Salt & Eaton, 1995a, 1995bSalt & Turner, 1998;Turner & Salt, 2003), with ultrastructural studies indicating the presence of Group II mGlu receptors on TRN terminals and surrounding glial processes (Liu et al., 1998;Mineff & Valtschanoff, 1999;Tamaru et al., 2001).
The similarity in sequence homology and subsequent pharmacology of the mGlu 2 and mGlu 3 receptors (Conn & Pin, 1997) has made it difficult to discern the relative contributions of each subtype to the overall Group II mGlu receptor effect on inhibition from the TRN to the VB. However, with careful use and application of selective pharmacological compounds in a series of electrophysiological experiments, we have previously been able to demonstrate mGlu 2 receptor-mediated disinhibition of sensory-evoked responses in the VB (Copeland et al., 2012(Copeland et al., , 2017) and, as presented in this study, we are now able to demonstrate contribution of mGlu 3 receptors to this same effect. from Jackson Laboratories, USA, on a C57BL6/J background) and What is already known • The thalamic reticular nucleus provides feedback inhibition to the ventrobasal thalamus upon somatosensory stimulation.

What does this study add
• mGlu 3 receptors likely majority mediate the Group II mGlu effect on sensory-evoked inhibition.

What is the clinical significance
• This mechanism likely enables important sensory information to be discerned from background activity.
• In higher order circuits, this mechanism may contribute to cognitive and attentional processes.
wild-type C57BL6/J mice (18-23 days old; n = 6; Harlan, USA) were deeply anaesthetised with 4.0% isoflurane and decapitated into a container of crushed ice. Wild types (À/À) were bred from heterozygous (+/À) bred pairs. Power calculations were performed to determine animal numbers required based on an estimated signal-to-noise ratio of 2 [mean: SD].

| Slice preparation and recording
Mouse brains were quickly removed and placed in an oxygenated, ice cold beaker of slicing solution that contained (in mM) 110 NaCl; 10 MgCl 2 ; 2 KCl; 26 NaHCO 3 ; 1.25 NaH 2 PO 4 ; 0.5 CaCl 2 ; 10 HEPES; and 15 glucose (pH adjusted to 7.45 with NaOH, osmolarity was 308 to 312 mOsm). After cooling in slicing solution for 2 to 3 min, the whole brains were blocked (portions of anterior and posterior tissue removed) using a razor blade and then glued to the microslicer (DTK Zero 1, DSK) tray using cyanoacrylate. The tray containing the blocked and mounted brain was filled with oxygenated, ice-cold slicing solution, and serial, coronal sections containing the TRN and VB were cut at a thickness of 300-400 μm. Slices were then placed in a larger recovery chamber containing oxygenated slicing solution at room temperature (18 C to 20 C). The recovery chamber was in a large water bath at room temperature. Slices for IPSP and mini-inhibitory postsynaptic current (mIPSC) recordings were processed as follows.

IPSP recordings
After 1 h in the recovery chamber, medium was replaced by a continuously oxygenated Krebs medium containing (in mM) 124 NaCl,  acid (LY341495) stocks were made in 100% DMSO at 1000Â the desired working concentration. Compound was diluted into the recording solution containing CNQX, APV and TTX immediately before application to the brain slice. All solutions applied to the brain slices contained 0.1% to 0.2% DMSO. DMSO content was matched between solutions for each experimental protocol and in the vehicle controls. Compound treatment periods were from 10 to 12 min in duration.

mIPSC recordings
After a 10-min period in the recovery chamber, 500 μl of 0.5-M CaCl 2 solution was slowly added (500-ml volume) to increase the calcium concentration to 1 mM. The water bath was then turned on, and the temperature was monitored inside the recovery chamber. The recovery chamber temperature was allowed to reach 33 C to 34 C for a period of approximately 30 min, after which the water bath was turned off and the recovery chamber was allowed to slowly return to room temperature (18 C to 20 C). Slices were used for recording after at least 1 h of recovery time.
Slices were placed in a superfusion chamber mounted on a Nikon Eclipse FN-1 microscope. Neurons within the VB area of F I G U R E 1 Thalamic circuitry underlying responses to vibrissal deflection. Branching collaterals from excitatory thalamocortical and corticothalamic axons (black), which originate from functionally linked topographical areas in the thalamus/ cortex, innervate the thalamic reticular nucleus (TRN), and the TRN sends a reciprocal inhibitory projection (grey) back to the thalamic area from which it receives its thalamocortical innervation the thalamus were visualised using IR/DIC water immersion optics.
Compound-containing solutions were applied to the slice via whole chamber superfusion. Glass recording electrodes were filled with (in mM)
Throughout the experiments, EEG and ECG were monitored. Additional urethane anaesthetic was administered i.p. as required and the experiment was terminated with an overdose of the same anaesthetic.
A tracheostomy tube was used to facilitate spontaneous breathing.
Body temperature was monitored and maintained using a heat probe and mat. A stereotaxic frame was used to fix the head in a flat skull position.

| Recording and iontophoresis
Seven-barrel recording and iontophoretic glass pipettes were advanced into the VB. Extracellular recordings were made from single VB neurons responsive to somatosensory input through the central barrel (filled with 4-M sodium chloride [NaCl]). Iontophoretic drug applications were performed unblinded using the outer barrels (Salt, 1987). On each occasion, one of the outer barrels was filled with

| Stimulation protocol
Neurons were identified as VB neurons on the basis of stereotaxic location (Paxinos & Watson, 1986) and responses to vibrissa deflection. Vibrissa deflection was performed using fine air-jets directed through 23 gauge needles mounted on micromanipulators positioned and orientated close to the vibrissa to ensure deflection of a single vibrissa was achieved. Air-jets were electronically gated with solenoid valves that produced a rising air pulse at the vibrissa 8 ms after switching. Response latencies were calculated from the start of the gating pulse. Using such an approach, it is possible to use air-jets to evoke an excitatory response from stimulation of a single vibrissa, as described previously (Salt, 1987

| In vivo electrophysiology protocols
Throughout the study, extracellular single neuron action potentials were gated, timed and counted using a window discriminator, a CED1401 interface and Spike2 software (Cambridge Electronic Design, Cambridge, UK), which recorded the output from the iontophoresis unit and also triggered the sensory stimuli sequence. Data were analysed by plotting post-stimulus time histograms (PSTHs) from these recordings by counting the spikes evoked by sensory stimulation. We used conventional criteria to divide neuronal responses into burst and tonic activity (Lu et al., 1992). Any action potentials with interspike intervals of ≤4 ms were considered to be part of a burst. All other spikes were regarded as tonic. We com-

| RESULTS
The interpretation of the results in this study clearly rely upon careful and appropriate use of the electrophysiological preparations in conjunction with the Group II mGlu receptor selective pharmacological compounds.
The rodent ventrobasal thalamus (VB) comprises one major cell type, thalamocortical relay neurons, and is otherwise largely devoid of intrinsic inhibitory interneurons, receiving its inhibitory input almost exclusively from the adjacent thalamic reticular nucleus (TRN) (Diamond et al., 2008). This VB circuitry is homologous between rat and mouse species (Diamond et al., 2008). In this study, in vitro electrophysiology was performed with mouse brain slices to take advantage of the genetic manipulations possible in this species to generate mGlu 3 knockout animals (Linden et al., 2005). Rats were used for the in vivo components to take advantage of their larger vibrissal pad size (Haidarliu et al., 2010), which increases feasibility of individual vibrissa deflections.
The Group II orthosteric agonist LY354740 (Monn et al., 1997;Schoepp et al., 2003) has been widely employed to investigate Group II mGlu receptor function in both in vitro and in vivo neurophysiological assays in both animal and human CNS studies (Copeland et al., 2012(Copeland et al., , 2015(Copeland et al., , 2017Moldrich et al., 2003;Nordquist et al., 2008;Schoepp et al., 2003). The Group II orthosteric antagonist LY341495 possesses high selectivity with nanomolar potencies for the Group II mGlu receptors. Whilst LY341495 also has submicromolar potencies at other mGlu receptor subtypes (Kingston et al., 1998), its use in this study follows the parameters previously demonstrated to produce selective antagonism for the Group II mGlu receptors only (Kingston et al., 1998). The mGlu 2 selective positive allosteric modulator LY487379, which possesses no intrinsic agonist activity at mGlu 2 receptors, acts to enhance responses to submaximal mGlu 2 receptor agonism without activity at other receptors or ion channels (Johnson et al., 2003). The use of LY487379 in rodent in vitro and in vivo neurophysiological studies is The Group II mGlu receptor agonist effect on evoked and spontaneous presynaptic quantal release events is nullified in mGlu 3 receptor knockout (Grm3 -/-) mice. (ai) Traces of the effects of the Group II mGlu receptor agonist LY354740 (30 nM) alone and in conjunction with increasing concentrations of the Group II mGlu receptor antagonist LY341495 (10, 30 and 300 nM) on IPSP amplitude in wild-type mouse brain slices. (aii) Traces of the effects of the Group II mGlu receptor agonist LY354740 (30 and 300 nM) on IPSP amplitude in mGlu 3 receptor knockout mouse brain slices. (bi) Overall effects on IPSP amplitude in wild-type mouse brain slices of the same compound application combinations as described in (ai). * indicates significance at the P < 0.05 threshold of responses under drug conditions in comparison with baseline. (bii) Overall effects on IPSP amplitude in mGlu 3 knockout mouse brain slices of the same compound application combinations as described in (aii). (ci) Traces from individual neurons illustrating the effects of the Group II mGlu receptor agonist LY354740 (30 nM) alone or in conjunction with the Group II mGlu receptor antagonist LY341495 (100 nM) on the number of spontaneous mIPSC events in the ventrobasal thalamus (VB) in wild-type mouse brain slices. (cii) Traces from individual neurons illustrating the effects of the Group II mGlu receptor agonist LY354740 (30 nM) on the number of spontaneous mIPSC events in the VB in mGlu 3 receptor knockout mouse brain slices. (di) Effects of the same compound application combinations on the cumulative fraction of the calculated interevent intervals of the spontaneous mIPSCs in the VB in wild-type mouse brain slices as described in (ci). (dii) Effects of the same compound application combinations on the cumulative fraction of the calculated interevent intervals of the spontaneous mIPSCs in the VB in mGlu 3 receptor knockout mouse brain slices as described in (cii) well documented (Cie slik et al., 2020;Copeland et al., 2012Copeland et al., , 2015Copeland et al., , 2017Galici et al., 2005;Mango et al., 2019). LY395756 is a mixed compound in that it possesses opposite activity at each of the Group II mGlu receptor subtypes. It is an agonist at mGlu 2 receptors (EC 50 0.40 μM) and an antagonist of mGlu 3 receptors (IC 50 2.94 μM; see compound 13, Dominguez et al., 2005). Whilst the mGlu 2 partial agonist activity of LY395756 has been debated to account for some of its

| Synaptic transmission at the TRN-VB synapse can be modulated by mGlu 3 receptor activation
It has been previously demonstrated that IPSPs evoked in vitro in the VB upon TRN stimulation in rat brain slices can be suppressed by Group II mGlu receptor activation (Turner & Salt, 2003) and that there is an mGlu 2 component to this overall Group II mGlu effect (Copeland et al., 2012(Copeland et al., , 2017. Therefore, we wanted to determine if there was a complementary mGlu 3 component to this same effect. Whilst the reversible Group II agonist LY354740 effect on evoked IPSP amplitude in the VB was reproducible in wild-type mouse brain slices (Copeland et al., 2017;Turner & Salt, 2003) where the circuitry connecting the VB and TRN is preserved (LY354740 30 nM reduced baseline evoked IPSP amplitude by 29% ± 2%; n = 6 cell recordings from 6 slices; Figure 2ai,bi), this was completely ablated in the mGlu 3 receptor knockout (Grm3 À/À ) mouse brain slices (LY354740 30 and 300 nM reduced baseline evoked IPSP amplitude by 2% ± 1% and 1% ± 1%, respectively; n = 6 cell recordings from 6 slices; Figure 2aii,bii). Interestingly, upon application of the highest concentration of the Group II antagonist LY341495 in combination with the Group II agonist LY354740 in the wild-type preparation, an increase in evoked IPSP amplitude when compared with baseline was observed (LY354740 30 nM and LY341495 100 nM increased evoked IPSP amplitude by 13% ± 2%; Figure 2ai,bi). Application of vehicle controls had no effect in comparison with baseline responses (data not shown).
One component that contributes to IPSP amplitude suppression is direct inhibition of GABAergic vesicle fusion with the presynaptic TRN membrane (Turner & Salt, 2003). In vitro recording of mIPSCs agonist LY354740 effect on mIPSC frequency was reproducible (Copeland et al., 2017) and reversible in wild-type mouse brain slices ( Figure 2ci,di), again, this was not evident in the mGlu 3 receptor knockout mouse brain slice preparations (Figure 2cii,dii). The frequency of mIPSCs between wild-type and mGlu 3 knockout preparations were comparable (4.1 ± 1.9 Hz vs. 3.8 ± 1.8 Hz, respectively; each n = 6 cell recordings from 6 slices; Wilcoxon matched pairs test P > 0.05). Application of vehicle controls had no effect in comparison with baseline responses (data not shown).
Taken together, these in vitro data indicate that there is a considerable mGlu 3 receptor-mediated component to the overall Group II mGlu receptor effect on TRN-induced GABAergic inhibition evoked in the VB. We therefore sought to confirm this effect in vivo.

| mGlu 3 activation contributes to the gating of neuronal responses to somatosensory stimulation
Ventrobasal thalamus (VB) neuronal responses to short-and longduration vibrissal stimulation can be potentiated by Group II mGlu receptor activation leading to somatosensory disinhibition, which comprises an mGlu 2 component (Copeland et al., 2012(Copeland et al., , 2017Salt & Eaton, 1995a, 1995bSalt & Turner, 1998). We used iontophoretic application of the mixed compound LY395756 (mGlu 2 agonist, mGlu 3 antagonist) onto VB neurons in vivo to confirm functional contribution of an mGlu 3 component to this overall Group II effect, as suggested by the in vitro experiments. Application of the mixed compound LY395756 alone inhibited total VB neuronal responses to longduration vibrissal stimulation (83% ± 3% of baseline; n = 11 cell recordings from 9 rats; Figure 3), an effect similar to that observed upon application of the Group II mGlu antagonist LY341495 (81% ± 5% of baseline; n = 6 cell recordings from 3 rats Figure 4b). As the mixed compound is an mGlu 3 antagonist, this has revealed an mGlu 3 component to the overall Group II mGlu effect on sensory disinhibition. Upon co-application of the mixed compound LY395756 with the mGlu 2 positive allosteric modulator LY487379, the reverse was seen, with a potentiation of VB neuronal responses to long-duration vibrissal stimulation (146% ± 10% of baseline; n = 6 cell recordings from 5 rats): an effect similar to that observed upon application of the Group II mGlu agonist LY354740 (177% ± 24% of baseline; n = 6 cell recordings from 6 rats; Figure 4b). As the mixed compound is an mGlu 2 agonist, this confirms the mGlu 2 component to the overall Group II mGlu effect on sensory disinhibition that has been previously described (Copeland et al., 2012(Copeland et al., , 2017. Furthermore, this action of the mGlu 2 positive allosteric modulator confirms that the Harlan Wistar rats used in this study express functional mGlu 2 receptors (Ceolin et al., 2011). Application of vehicle controls had no effect in comparison with baseline responses (data not shown).
To further investigate the agonist/antagonist activity of the mixed compound LY395756, we performed further subanalysis of its effects on burst firing of VB thalamic neurons, which exhibit two distinct response patterns, tonic and burst-mode responses (Copeland et al., 2015;Ramcharan et al., 2000;Rivadulla et al., 2003). Tonic responses are associated with a linear transmission of information and occur when thalamic neurones have been depolarised from resting potential following the inactivation of a voltage-and time-dependent calcium current (I T ), whilst burst-mode firing occurs following hyperpolarisation of thalamic neurones where I T is de-inactivated (Jahnsen & Llinás, 1984;Llinás & Jahnsen, 1982). As such, during tonic firing, synaptic transmission through the thalamus is faithfully relayed, whereas during burst firing, transmission through the thalamus is less reliable with impulses occurring at low and irregular rates punctuated by high-frequency bursts. Iontophoretic application of the mixed compound LY395756 alone increased burst firing (baseline burst firing: 65% ± 7%; +LY395756 burst firing 75% ± 6%; n = 11 cell recordings from 9 rats; Figure 4a), an effect similar to that observed upon application of the Group II mGlu antagonist LY341495 (baseline burst firing: 49% ± 6%; +LY341495 burst firing 65% ± 4%; n = 6 cell recordings from 3 rats; Figure 4b). Iontophoretic application of the mixed compound LY395756 in combination with the mGlu 2 positive allosteric modulator LY487379 decreased burst firing (baseline burst firing: 66% ± 6%; +LY395756 and LY487379 burst firing 53% ± 6%; n = 6 cell recordings from 5 rats; Figure 5a), an effect similar to that observed upon application of the Group II mGlu agonist LY354740 (baseline burst firing: 74% ± 5%; +LY354740 burst firing 62% ± 4%; n = 6 cell recordings from 6 rats; Figure 5b). Application of vehicle controls had no effect in comparison with baseline responses (data not shown).

| DISCUSSION AND CONCLUSIONS
Thalamic nuclei can be classed as either first-order or higher order nuclei based upon the source of their driver inputs of information: first-order nuclei receive driver inputs from the periphery (e.g. auditory, visual and somatosensory), whereas higher order nuclei receive driver inputs from cortical layer V (Jones, 2009) thalamic circuitries, such as those of the higher order thalamic nuclei that serve to support cognitive processes (see Jones, 2009, for review). As the rodent VB composes only of excitatory VB neurons (i.e. no interneurons) (Barbaresi et al., 1986;Harris & Hendrickson, 1987;Ohara & Lieberman, 1993;Ralston, 1983), which when coupled with our in-depth understanding of its simple circuitry (Sherman & Guillery, 2001) (Figure 1), makes it an ideal candidate with which to understand the basic principles of thalamic function.
There is substantial evidence from in vitro and in vivo electrophysiological studies that upon driver afferent stimulation, the Group II mGlu receptors reduce inhibition in thalamic nuclei, likely via a reduction in GABAergic transmission from the TRN (Alexander & Godwin, 2005;Copeland et al., 2012Copeland et al., , 2015Copeland et al., , 2017Salt & Eaton, 1995a;Salt & Turner, 1998;Turner & Salt, 2003). In the VB, it has been previously demonstrated that this mechanism comprises an mGlu 2 component (Copeland et al., 2012(Copeland et al., , 2017 and, in this study, we are able to provide further evidence to support this and also additional evidence for co-contribution from mGlu 3 receptor activation to this effect. In fact, these data suggest that the overall Group II mGlu receptor effect is mainly mediated via mGlu 3 . The mixed compound LY395756 is a more than sevenfold more potent mGlu 2 receptor agonist (EC 50 0.40 μM) than it is an mGlu 3 receptor antagonist (IC 50 2.94 μM; Dominguez et al., 2005), yet the overriding effect of the mixed compound when applied alone in this study was that of antagonism.
Indeed, the mGlu 2 positive allosteric modulator used to reveal the mGlu 2 component is in itself very effective. As application of submaximal glutamate (1 μM) to produce a $3% of maximal glutamate response can be potentiated upon co-application of LY487479 up to $75% of maximal glutamate response, a 2,500% increase (Johnson et al., 2003), meaning that even low levels of mGlu 2 activation can be revealed. Therefore, the mGlu 2 effect revealed by the potent mGlu 2 positive allosteric modulator in this study and our previous work (Copeland et al., 2012(Copeland et al., , 2017 Ohishi et al., 1993aOhishi et al., , 1993bTamaru et al., 2001). Taken together, the electrophysiological and ultrastructural studies suggest that mGlu 3 receptors localised on TRN axon terminals, likely the majority, mediate the reduction in inhibition in thalamic nuclei from the TRN, with any Group II mGlu receptor subtypes present on surrounding glial processes likely contributing a smaller modulatory role to the same effect.
The location of Group II mGlu receptors on TRN terminals and surrounding astrocytic processes indicates that they play a pivotal role in modulating inhibition in the thalamus. As TRN terminals are not directly targeted by synaptically released glutamate, these receptors may be activated by endogenous 'glutamate spillover' (Kullmann, 2000) from synapses formed between excitatory driver afferents and thalamocortical neurons upon normal sensory processing (Copeland et al., 2012) or under conditions of intense synaptic activation (Alexander & Godwin, 2006). This is further evidenced by the in vitro IPSP experiments conducted in this study, as application of the highest concentration of the Group II mGlu receptor antagonist LY341495 was able to increase IPSP amplitude above that seen under baseline conditions. As the mGlu 3 knockout data indicate that reductions in IPSP amplitude and mIPSC frequency are majority mediated via mGlu 3 , taken together, these data suggest that it is basal mGlu 3 receptor activation that is likely occurring upon driver afferent stimulation via 'glutamate spillover' to reduce inhibition in the VB from the TRN. Ultrastructural studies support such a mechanism as they have evidenced close association of sensory afferent terminals with TRN afferent GABAergic terminals, upon which mGlu 3 receptors are heavily localised (Lourenço Neto et al., 2000;Ohishi et al., 1993aOhishi et al., , 1993bTamaru et al., 2001), on the soma and proximal dendrites of neurons in the rodent VB (Ohara & Lieberman, 1993;Ralston, 1983).
This study contributes to the growing evidence demonstrating that GABAergic transmission from the TRN can be reduced through activation of Group II mGlu receptors, resulting in a reduction in sensory-evoked inhibition (Copeland et al., 2012(Copeland et al., , 2017. Such a mechanism could play an important role in discerning relevant information from background activity to enhance sensory perception. In higher order nuclei, this mechanism could be of importance in attentional and cognitive processes, and there is evidence to suggest that it is the mGlu 3 receptor subtype that plays a role in reducing inhibition from the TRN in such nuclei (Copeland et al., 2015).
As increased inhibition in the higher order mediodorsal thalamic nucleus has been associated with onset of cognitive deficits and impairments in working memory, such as is seen in schizophrenia (DeNicola et al., 2020;Parnaudeau et al., 2013;Peräkylä et al., 2017).
Targeting of the Group II mGlu receptors and, specifically the mGlu 3 receptor subtype, may be of therapeutic importance for this disease state. As antagonism of mGlu 3 receptors increases evoked inhibition from the TRN, as evidenced in this study by a decrease in total firing but an increase in burst firing upon application of the mixed compound LY395756 alone. Thus, activation of mGlu 3 receptors to reduce inhibition may be an appropriate target-specific treatment for schizophrenia. This mechanism of mGlu 3 -mediated reduction in inhibition is further supported by the IPSP and mIPSC data presented in this study. It has been previously demonstrated that activation of mGlu 3 receptors in the mediodorsal thalamus is able to reduce inhibition from the TRN (Copeland et al., 2015). Thus the present study suggests that mGlu 3 receptor activation will decrease the unreliable and irregular synaptic transmission associated with burst firing and increase the proportion of tonic firing where synaptic transmission through the thalamus is faithfully relayed. Indeed, the mGlu 3 receptor has been implicated in the aetiological, pathophysiological and pharmacotherapeutic aspects of schizophrenia, with polymorphisms in the mGlu 3 receptor gene and protein, but not the mGlu 2 receptor, detected in patients with schizophrenia (see review: Stansley & Conn, 2018). The design of future novel therapies targeted to treat deficits in cognitive function may therefore achieve greater success if selectivity and higher efficacy for mGlu 3 receptors were achieved.

| Conclusions
The Group II mGlu receptor effect in reducing evoked inhibition from the TRN to the VB is likely to mainly u3mediated via mGlu 3 receptors.
This mechanism may be of importance when identifying important sensory information in an environment with background activity and may be an overarching principle applicable to higher order thalamic nuclei function in the control of cognitive and attentional processes.
As the mGlu 3 receptor subtype appears to be activated by endogenous 'glutamate spillover' upon afferent stimulation, it may be advantageous to develop mGlu 3 positive allosteric modulators, as opposed to direct agonists, to alleviate instances of increased thalamic inhibition, as is believed to occur in schizophrenia. Indeed, there is emerging evidence to suggest that mGlu 3 positive allosteric modulators could act to enhance performance in cognitive tasks (Walker & Conn, 2015).

| Nomenclature of targets and ligands
Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Harding et al., 2018), and are permanently archived in the Concise Guide to PHARMACOLOGY 2019/20 (Alexander et al., 2019).