Genetic deletion of Rhes or pharmacological blockade of mTORC1 prevent striato-nigral neurons activation in levodopa-induced dyskinesia

Highlights

• Rhes deletion and mTORC1 blockade attenuate levodopa-induced dyskinesia in mice.

• Rapamycin but not Rhes deletion fully preserves rotarod performance on levodopa.

• Rhes deletion and rapamycin prevent the levodopa-induced rise of nigral amino acids.

• Rapamycin but not Rhes deletion prevents the rise of striatal glutamate.

Abstract

Ras homolog enriched in striatum (Rhes) is a small GTP-binding protein that modulates signal transduction at dopamine receptors, and also activates mammalian target of rapamycin complex 1 (mTORC1). Rhes binding to mTORC1 is hypothesized to play a role in motor disorders such as levodopa-induced dyskinesia. Here, we investigate the behavioral and in vivo neurocircuitry changes associated with genetic deletion of Rhes or inhibition of mTORC1 signaling in the mouse model of levodopa-induced dyskinesia. 6-Hydroxydopamine-hemilesioned Rhes knockout mice and wild-type littermates were chronically treated with levodopa. In parallel, 6-hydroxydopamine-hemilesioned naïve mice were chronically treated with levodopa or levodopa plus rapamycin, to block mTORC1 pathway activation. Dyskinetic movements were monitored during levodopa treatment along with motor activity on the rotarod. Finally, dyskinetic mice underwent microdialysis probe implantation in the dopamine-depleted striatum and ipsilateral substantia nigra reticulata, and GABA and glutamate levels were monitored upon acute challenge with levodopa. Both Rhes knockouts and rapamycin-treated mice developed less dyskinesia than controls, although only rapamycin-treated mice fully preserved rotarod performance on levodopa. Levodopa elevated nigral GABA and glutamate in controls but not in Rhes knockouts or rapamycin-treated mice. Levodopa also stimulated striatal glutamate in controls and Rhes knockouts but not in rapamycin-treated mice. We conclude that both genetic deletion of Rhes and pharmacological blockade of mTORC1 significantly attenuate dyskinesia development by reducing the sensitization of striato-nigral medium-sized spiny neurons to levodopa. However, mTORC1 blockade seems to provide a more favorable behavioral outcome and a wider effect on neurochemical correlates of dyskinesia.

Introduction

Ras homolog enriched in striatum (Rhes) is a 266 amino acid protein isolated in the rat brain through a subtractive hybridization procedure (Falk et al., 1999, Usui et al., 1994). Rhes belongs to the superfamily of Ras proteins, a group of small GTP-binding proteins that exert pleiotropic effects on cell function (Harrison, 2012). Moreover, Rhes has been shown to influence dopamine (DA) and adenosine-related behaviors and signal transduction pathways (Errico et al., 2008, Ghiglieri et al., 2015, Sciamanna et al., 2015, Vitucci et al., 2015). Rhes is highly expressed in striatum, and to a lesser extent in other areas of the rodent and human brain, including hippocampus (Ghiglieri et al., 2015, Spano et al., 2004). In human brain, the ortholog gene, RASD2, has been localized also in the V layer of prefrontal cortex, with a peculiar pattern of expression (Vitucci et al., 2015). Rhes mRNA expression is essentially postnatal and positively regulated by thyroid hormone (Vargiu et al., 2001). Likewise, DA is a positive regulator of Rhes expression in adult striatum (Harrison and LaHoste, 2006, Harrison et al., 2008). Rhes found in striatal medium-sized spiny neurons (MSNs) and cholinergic interneurons is involved in sensorimotor gating functions and modulates striatal DA receptor-dependent signaling and behaviors (Errico et al., 2008, Ghiglieri et al., 2015, Sciamanna et al., 2015, Vitucci et al., 2015). However, Rhes has received much attention for its pathogenic role in Huntington's disease (Mealer et al., 2013, Subramaniam et al., 2009, Subramaniam and Snyder, 2011). Consistently, Rhes acts as a specific striatal E3 ligase (Subramaniam et al., 2010) able to bind and SUMOylate mutant huntingtin, thereby inhibiting its aggregation and increasing its cytotoxicity (Subramaniam et al., 2009). Additionally, it has been postulated that binding to mutant huntingtin prevents Rhes from activating mammalian target of rapamycin complex 1 (mTORC1), thus removing its trophic effects on striatal cells (Lee et al., 2015, Subramaniam and Snyder, 2011). On the other hand, striatal Rhes-induced mTORC1 pathway activation has been proven to be crucial for another motor disorder, i.e. levodopa (L-DOPA) induced dyskinesia (LID) (Subramaniam et al., 2012). LID represents a group of dystonic and choreic involuntary movements emerging in the majority of Parkinson's disease (PD) patients chronically administered with L-DOPA (Bastide et al., 2015, Fabbrini et al., 2007). LID can be very disabling, particularly in the advanced stages of PD, and the yet incomplete knowledge of the pathophysiological mechanisms underlying LID has hindered the identification of a drug capable of preventing the sensitization of striatal MSNs to L-DOPA, the cellular mechanism associated with LID development (Bastide et al., 2015, Santini et al., 2010).

Nonetheless, a pathogenic role in LID onset for aberrant striatal D1 receptor (D1R) signaling in MSNs has been well established (Bastide et al., 2015, Feyder et al., 2011). Indeed, D1R coupling to Gαolf proteins and adenylyl cyclase is enhanced in the dyskinetic brain, which results in an increased activity of cAMP-related protein kinase A (PKA) and phosphorylation (activation) of downstream effectors such as Dopamine- and cAMP-regulated phosphoprotein M_r. 32,000 (DARPP-32) (Nishi et al., 1997). DARRP-32 phosphorylation at Thr34 residue promotes extracellular-regulated kinase 1 and 2 (ERK) signaling (Pavon et al., 2006) which, in turn, activates mTORC1, ultimately responsible for LID onset in mice and rats (Decressac and Bjorklund, 2013, Santini et al., 2008). ERK activates mTORC1 by maintaining Ras homolog enriched in brain (Rheb) in the active state. Until the discovery that Rhes is capable to bind and activate mTORC1 (Subramaniam et al., 2012, Subramaniam and Snyder, 2011), Rheb was believed to be the only striatal mTORC1 activator. Interestingly, in line with Rhes ability to bind and activate mTORC1 signaling, which is known to be instrumental to LID development (Santini et al., 2009), Rhes knockout (Rhes^−/−) mice displayed significantly less dyskinesia than wild-type littermates during L-DOPA treatment (Subramaniam et al., 2012).

In the present study, we used an in vivo microdialysis approach to investigate the role of Rhes and mTORC1 in modulating the neurocircuitry underlying dyskinesia associated with striatal dopamine denervation and chronic L-DOPA administration. Specifically, we investigated whether L-DOPA is able to elevate GABA release in substantia nigra reticulata (SNr), an in vivo neurochemical marker of striato-nigral MSNs activation (Bido et al., 2011, Mela et al., 2012, Mela et al., 2007, Paolone et al., 2015), and glutamate release in striatum, a neurochemical LID correlate (Dupre et al., 2011, Paolone et al., 2015), in 6-OHDA-hemilesioned, L-DOPA primed Rhes^−/− mice and rapamycin-treated mice.