Discussion
The present study provides strong evidence that RGS4 inhibits NOP receptor responses in vitro. Moreover, this study demonstrates that pharmacological inhibition of RGS4 potentiates the ability of a NOP agonist to attenuate LID and its neurochemical correlates in vivo, without worsening its sedative properties. These data point to a functional RGS4-NOP receptor interaction, and add to previous reports that RGS4 fine-tunes opioid receptor signaling and behaviors (Dripps et al. , 2017; Han et al. , 2010; Stratinaki et al. , 2013; Xie et al. , 2005). In fact, RGS4 negatively regulated reward and physical dependence induced by the MOP receptor preferential agonist morphine but did not affect morphine-induced analgesia or tolerance (Han et al. , 2010). RGS4 also differentially modulated DOP receptor-mediated behavioral outcomes in mice (Dripps et al. , 2017; Stratinaki et al. , 2013). Indeed, genetic deletion of RGS4 as well as acute pharmacological inhibition of RGS4 with CCG-203769, increased SNC80-induced antinociception and antihyperalgesia, but did not affect the pro-convulsant action of this DOP agonist.
The interaction between RGS4 and NOP receptor was originally evaluated in COS-7 cells transfected with a dual-expression plasmid containing RGS4 and each of the opioid receptor subtypes (Xie et al. , 2005). In that study a single concentration of N/OFQ was tested, which was poorly modulated by RGS4. In fact, RGS4 increased by 28% N/OFQ GTPase activity but left unaffected N/OFQ inhibition of forskolin-induced cAMP levels. In the present study in HEK293T cells, RGS4 did not affect N/OFQ efficacy but significantly reduced its potency. Moreover, it reduced both the efficacy and potency of the small molecule potent and selective agonist AT-403. In NOP-transfected HEK293T cells, N/OFQ and AT-403 inhibited the D1-stimulated cAMP production, a Gi/o protein-mediated intracellular function (Feng et al. , 2017) with similar potencies (9.43 and 9.92, respectively).These values are consistent with those reported for N/OFQ and AT-403 in the [35S]GTPγS and calcium mobilization assays (Ferrari et al. , 2017) confirming that AT-403 is a potent full agonist at the NOP receptor. The rightward shift of N/OFQ and AT-403 curves and the reduction of AT-403 efficacy in the presence of RGS4 suggests that RGS4 negatively couples to Gi/o to inhibit NOP receptor signaling. However, this effect is shared by another RGS protein, RGS19, functionally very close to RGS4, which was reported to negatively modulate NOP receptor signaling in vitro (Xie et al. , 2005). Interestingly, RGS19 reduced the efficacy of AT-403, but showed only a trend for reduction of N/OFQ efficacy. Thus, RGS4 and RGS19 seem to modulate more efficiently the efficacy of AT-403 than of N/OFQ. This could be due to the different modes of interaction and activation of the NOP receptor by the endogenous ligand N/OFQ versus a small-molecule nonpeptidic ligand like AT-403.
The occurrence of a RGS4-NOP interaction was confirmed in native systems, i.e. striatal primary neurons and striatal slices, using an RGS4 inhibitor to block the endogenous activity of RGS4. Primary striatal neurons express both RGS4 and NOP (Buzas et al. , 1998; Runne et al. , 2008) and the two functionally interact to modulate D1 receptor evoked responses since CCG-203920 increased N/OFQ potency in inhibiting D1-stimulated cAMP levels. This interaction occurs in DARRP32-positive, likely GABAergic, neurons indicating this interaction is not an artifact of protein overexpression in HEK273 cells but is constitutively active and physiologically relevant. This interaction occurs not only in developing tissues in rats but also in adult mice. We previously demonstrated that N/OFQ and AT-403 inhibited the increase of ERK-positive striatal neurons (likely MSNs) induced by D1 receptor agonist SKF-38393 in striatal slices (Arcuri et al. , 2018; Marti et al. , 2012), a biochemical predictor of antidyskinetic activity. In this model, we now show that CCG-203920 potentiates the effect of a submaximal dose of AT-403 without affecting the D1 response alone. Consistently, CCG-203920 potentiated the antidyskinetic effect of AT-403, significantly extending the delay in AIM onset induced by AT-403. Since this effect was not accompanied by the worsening of the positive effect of AT-403 on rotarod performance, it is likely due a true potentiation of its antidyskinetic properties. Since we show that the same dose of CCG-203920 reversed raclopride-induced akinesia in mice through selective interaction with RGS4, we are confident that also the antidyskinetic effect of CCG-20920 is mediated by RGS4 targeting. To confirm that the effect of CCG-203920 is truly mediated by interference with the molecular pathways underlying LID, CCG-203920 potentiated the inhibition of ERK phosphorylation induced by AT-403. LID is characterized by aberrant enhancement of Gα and Gβγ signaling pathways downstream of the D1 receptor, such as the cAMP/PKA and MAPK cascades (Santiniet al. , 2007). Specifically, the increased activity along the canonical and non-canonical D1 pathways leads to the phosphorylation of several downstream effectors in striatal direct pathway MSNs, such as the GluR1 subunit of glutamate AMPA receptor and ERK (Pavon et al. , 2006; Santini et al. , 2007) . We previously reported that a dose of AT-403 as high as 0.1 mg kg-1 normalized pERK levels and blunted LID (Arcuri et al. , 2018). We now report that a 3-fold lower dose, ineffective alone, normalized pERK levels when combined with RGS4 inhibitor CCG-203920, confirming that RGS4 blockade potentiates the ability of a NOP receptor agonist to modulate MAPK pathway changes underlying LID. As far as pGluR1 levels are concerned, AT-403 alone fully inhibited the rise of pGluR1 associated with dyskinesia, which might have precluded further inhibition by CCG-203920. Overall, these data indicate that RGS4 blockade improves the antidyskinetic effect induced by a NOP receptor agonist and the underlying signaling pathways, without amplifying its sedative effects. This suggests that as for δ opioid receptor agonists (Dripps et al. , 2017), RGS4 blockade might differentially impact NOP behaviors, and perhaps widen the therapeutic window between sedation and pharmacodynamic effects of potent NOP agonists like AT-403.
Interestingly, previous studies indicated the involvement of RGS4 in the pathogenesis of LID, showing that RGS4 blockade induced therapeutic antidyskinetic effects (Ko et al. , 2014; Shen et al. , 2015). Specifically, chronic treatment with antisense oligonucleotides targeting RGS4 reduced AIM development during L-Dopa priming in a rat model of LID (Ko et al. , 2014). Pharmacological blockade of RGS4 is expected to affect GPCRs other than the NOP receptor, inducing off-target effects in brain or neuronal populations not involved in LID. To confirm that an RGS4 inhibitor would selectively target and correct a pathological condition, we show that striatal RGS4 level are reduced after DA depletion and rapidly upregulated after L-Dopa administration and dyskinesia onset. This further supports the rationale for therapeutic application of RGS4 inhibitors in LID therapy. Our data confirm previous evidence of reduction of RGS4 expression in DA-depleted animals (Geurts et al. , 2003; Ko et al. , 2014). More specifically, they nicely complement an ex-vivo study in 6-OHDA hemilesioned dyskinetic rats (Koet al. , 2014), where RGS4 expression was found to be reduced in the lesioned striatum after DA depletion and increased following chronic L-Dopa treatment, the increase being more marked at 1 hour (i.e. ON L-Dopa) than at 24 hours (i.e. OFF L-Dopa) after L-Dopa administration. Surprisingly, however, our study also revealed a reduction in the unlesioned striatum, suggesting a powerful influence of the cortico-basal ganglia-thalamo-cortical loop and/or cross-striatal dopaminergic projections over RGS4 levels.