Introduction
The contribution of striatal cholinergic interneurons (ChIs) to striatal functions and parkinsonism has long been recognized (Calabresi et al. , 2000; Pisani et al. , 2007). In the last decade, the role of striatal ChIs in levodopa (L-DOPA) induced dyskinesia (LID) has also emerged (Bordiaet al. , 2019; Moehle et al. , 2019). Dyskinesia is a major side-effect of L-DOPA pharmacotherapy of Parkinson’s disease (PD), and is clinically characterized by abnormal involuntary movements (AIMs) of dystonic and choreic nature (Bastide et al. , 2015). LID has a complex and partially unknown etiology, but the final neurobiological event is the sensitization of D1 expressing striato-nigral GABAergic medium-sized spiny neurons (MSNs) to L-DOPA (Bastide et al. , 2015). Striatal ChIs profusely innervate both striato-pallidal and striato-nigral MSNs and influence their activity either directly or indirectly, via modulation of dopamine, serotonin, glutamate release from striatal afferents, and GABA release from striatal interneurons. It was originally found that ablation or functional inhibition of ChIs attenuate LID (Ding et al. , 2011; Won et al. , 2014), although a more recent optogenetic study revealed that ChIs dually modulate LID, facilitating it at low firing rates and attenuating it at higher ones (Bordia et al. , 2016). Pharmacological studies showed that both muscarinic (Bordia et al. , 2016; Chambers et al. , 2019; Ding et al. , 2011; Shen et al. , 2015) and nicotinic (Quik et al. , 2015) receptors are involved in LID. However, while it has been convincingly demonstrated that nicotinic receptor blockade (with receptor antagonists) or desensitization (with receptor agonists) prevents LID (Quik et al. , 2015), the role of muscarinic receptors is far from clear (Bordia et al. , 2019). This might be due to the different neuronal distribution and functional roles of muscarinic receptor subtypes. In fact, muscarinic receptors are GPCR belonging to two major families which couple to Gi/o (M2 and M4 subtypes) or Gq/11 (M1, M3 and M5 subtypes), and are unevenly distributed on the different populations of striatal neurons and nerve terminals. In particular, striato-pallidal MSNs predominantly express M1 receptors, while striato-nigral MSNs express both M1 and M4 receptors (Bernard et al. , 1992; Hersch et al. , 1994). The M4 receptor is considered the main autoreceptor type at striatal ChIs (Zhang et al. , 2002), although also M2 autoreceptors contribute to negative auto feed-back (Bonsi et al. , 2008). Another major problem in dissecting out the role of endogenous ACh and muscarinic receptor subtypes in LID is the poor selectivity of muscarinic antagonists available. Thus, the unselective muscarinic antagonists dicyclomine and atropine prevented LID in 6-hydroxydopamine (6-OHDA) hemilesioned mice (Bordia et al. , 2016; Ding et al. , 2011) whereas atropine also prevented LID inhibition induced by optogenetic elevation of ChIs firing rates (Bordia et al. , 2016). M4 selective positive allosteric modulators (PAMs) reduced LID in murine and non-human primate models of LID, pointing out a crucial role for M4 receptors in this movement disorder (Shen et al. , 2015). In partial agreement, the putative M4 “preferential” antagonist tropicamide (Lazareno et al. , 1993; Lazareno et al. , 1990) worsened AIMs induced by L-DOPA but attenuated those induced by a D1 receptor agonist in 6-OHDA hemilesioned rats (Chambers et al. , 2019), suggesting the involvement of different sets of M4 receptors. However, the M4 “selectivity” of tropicamide, if any (Croy et al. , 2016), is negligible which calls for a re-evaluation of its effects with relatively more selective compounds. Finally, the role of ACh in LID might extend beyond striatal ChIs. In fact, cholinergic inputs from the pedunculopontine tegmental nucleus (PPN) modulate basal ganglia function (Xiao et al. , 2016), and D1 signaling at striato-nigral MSNs terminals (Moehle et al. , 2017). Although a preliminary report showed that injection of tropicamide in PPN did not modulate LID expression (Chambers et al. , 2019), whether M4 receptor stimulation or blockade in SNr shape LID remains to be established.
In the present study, in vivo dual probe microdialysis was used to clarify the role of striatal and nigral M1 and M4 receptors in the modulation of LID and the underlying striato-nigral MSNs activation. A reverse microdialysis approach was adopted to deliver the M1 and/or M4 preferential antagonists telenzepine, PD-102807, tropicamide as well as the selective M4 PAM VU0152100 in the dorsolateral striatum or SNr, concurrently with systemic L-DOPA administration. This approach had the two-fold advantage of i) targeting muscarinic receptors in specific brain areas, ii) minimizing the poor selectivity of muscarinic receptor antagonists by setting pharmacological selective concentrations in the perfusion Ringer (see Methods). Dyskinetic behavior was monitored concurrently with GABA release in SNr, a readout of striato-nigral MSNs activation (Marti et al. , 2012; Mela et al. , 2012; Mela et al. , 2007), and glutamate release in striatum, a possible index of cortico-basal ganglia-thalamo-cortical loop activation (Bastide et al. , 2015). The feasibility of this approach was previously demonstrated showing that reverse dialysis of the D1 antagonist SCH23390, but not the D2 antagonist raclopride, in striatum prevented LID and its neurochemical correlates (Mela et al. , 2012).