Converging roles of ion channels, calcium, metabolic stress, and activity pattern of Substantia nigra dopaminergic neurons in health and Parkinson's disease

Functional anatomy of the dopaminergic midbrain system

Dopamine‐releasing catecholaminergic neurons within the midbrain are part of the basal ganglia network, and are crucial for a variety of fundamental brain functions, such as voluntary movement, goal‐directed behavior and habit formation, and also motivation, emotion, cognition, reward, memory, associative learning, and decision making (Bjorklund and Dunnett 2007; Gerfen and Surmeier 2011; D'Ardenne et al. 2012; Pignatelli and Bonci 2015). Consequently, their loss or their dysfunction is associated with respective pathophysiologies, like schizophrenia, substance abuse, or Parkinson's disease (PD) (Wise 2009; Brisch et al. 2014; Kalia and Lang 2015).

The cell bodies of tyrosine hydroxylase‐positive dopaminergic (DA) neurons within the midbrain are mainly arranged within three nuclei in close proximity: the ventral tegmental area (VTA, or A10 group), the retrorubral field (RRF, or A8 group), and the Substantia nigra (SN, or A9 group) (Dahlstrom and Fuxe 1964; Bjorklund and Dunnett 2007). Simplified, according to their widespread axonal projections, DA midbrain neurons are subdivided into the mesocorticolimbic system (VTA and RRF DA neurons), and the mesostriatal system (SN DA neurons). However, in‐depth cell‐specific analyses of DA projections and inputs, biophysical properties, neuronal wiring, gene expression, and physiological functions have revealed a much more complex picture, defining a variety of different subpopulations of midbrain DA neurons (Bromberg‐Martin et al. 2010; Liss and Roeper 2010; Bourdy and Barrot 2012; Lammel et al. 2012; Ungless and Grace 2012; Watabe‐Uchida et al. 2012; Jhou et al. 2013; Roeper 2013; Volman et al. 2013; Poulin et al. 2014; Beier et al. 2015; Hoglinger et al. 2015; Lerner et al. 2015).

Electrical activity patterns of SN DA neurons in vivo and in vitro

The number of SN DA neurons ranges from about 10.000 (bilaterally) in adult mice to about 400.000 in adult humans (Damier et al. 1999a; Brichta and Greengard 2014), and they are predominantly found within the SN pars compacta. They receive direct excitatory inputs from the subthalamic nucleus, the pedunculopontine nucleus, as well as from somatosensory and motor cortices, and they receive inhibitory inputs particularly from the (dorsolateral) striatum (Grillner and Mercuri 2002; Henny et al. 2012; Watabe‐Uchida et al. 2012; Lerner et al. 2015).

In vivo, within the intact basal ganglia network, SN DA neurons predominantly display two types of firing patterns: tonic irregular single‐spike activity in the frequency range of 1–10 Hz (compare Fig. 2) (Hage and Khaliq 2015), or phasic so‐called burst activity at higher frequencies (~ 13–20 Hz in anesthetized animals, or up to 80 Hz in awake animals or humans), as originally described by the pioneering work of Grace and Bunney (Grace and Bunney 1984a,b) and Johnson (Johnson et al. 1992), reviewed, e.g. in (Johnson and Wu 2004; Bean 2007a; Deister et al. 2009; Lee and Tepper 2009; Paladini and Roeper 2014; Dragicevic et al. 2015). Burst activity of SN DA neurons causes a phasic supralinear increase in released dopamine, predominantly in response to novel, unexpected or salient events (Schultz et al. 2015), and it encodes the so‐called ‘Go’ signals that facilitate movement initiation and motor learning (Jin and Costa 2010; Costa 2011; Sommer et al. 2014). As illustrated in Figs 1 upper and 2, in vitro, even in complete synaptic isolation, SN DA neurons still display an intrinsically generated, very regular activity with relatively low frequencies (~ 0.5–4 Hz), the so‐called pacemaker activity (Grace and Onn 1989; Bean 2007a; Chan et al. 2007; Lammel et al. 2008; Margolis et al. 2010).

SN DA neurons are particularly vulnerable to degenerative triggers

Although seemingly more homogeneous than the VTA/RRF DA neurons, recent work identified distinct subpopulations of DA neurons also within the SN: the medial and lateral SN DA neurons, with axonal projections to the dorsomedial and dorsolateral striatum, respectively (Bjorklund and Dunnett 2007; Schiemann et al. 2012; Lerner et al. 2015), as well as the ventral tier and dorsal tier SN DA neurons (Damier et al. 1999a; Watabe‐Uchida et al. 2012; Lehericy et al. 2014). However, dorsal tier SN DA neurons are more ‘VTA/RRF DA‐type’ like, in view of their electrophysiological properties, their expression of the cytosolic calcium‐binding protein calbindin‐D28K, the absence of DCC (deleted in colorectal cancer, the receptor for the axon guidance molecule netrin), and their lower susceptibility to degenerative triggers and to cell death in PD (Yamada et al. 1990; Damier et al. 1999a; Neuhoff et al. 2002; Lammel et al. 2008; Margolis et al. 2010; Brichta and Greengard 2014). In contrast, ventral tier SN DA neurons build the ‘classical’ calbindin‐D28K‐negative and DCC‐positive mesostriatal SN DA population (located particularly in the ventrolateral SN) that is crucial for control and modulation of voluntary movements and for habit learning/formation. These SN DA neurons are particularly vulnerable to degeneration in PD, and their loss causes the major motor symptoms of PD patients: bradykinesia, rigor, resting tremor, and postural instability (Fearnley and Lees 1991; Agid et al. 1993; Damier et al. 1999b; Braak et al. 2004; Osborne et al. 2005; Kordower et al. 2013; Brichta and Greengard 2014; Dopeso‐Reyes et al. 2014; Kalia and Lang 2015). It is noteworthy that other, non‐DA neurons are also progressively degenerating in PD, particularly noradrenergic neurons within the locus coeruleus (LC), but also neurons, e.g. within the pedunculopontine nucleus, or the dorsal motor nucleus (Jellinger 1991; Zarow et al. 2003; Del Tredici and Braak 2013; Goedert et al. 2013; Greene 2014; Sanchez‐Padilla et al. 2014).

In this review, we focus on the ‘classical’ mesostriatal (or nigrostriatal) SN DA neurons, and on their particularly high vulnerability to degeneration in PD. The cause for the differential vulnerability of the different types of SN DA and VTA DA neurons to degenerative triggers is still not clear; however, it is present not only in PD and its animal models, but it also occurs in the non‐PD brain during aging (Damier et al. 1999b; Hindle 2010; Collier et al. 2011; Reeve et al. 2014; Rodriguez et al. 2015). Indeed, aging is the most prominent risk factor for PD, and PD is discussed to reflect a form of premature or accelerated aging (Surmeier 2007; Hindle 2010; Collier et al. 2011).

Mitochondrial (dys‐)function and metabolic stress in SN DA neurons

SN DA neurons appear to be intrinsically particularly vulnerable to metabolic stress and mitochondrial dysfunction. Thus, it is not too surprising that inhibitors of the complex I of the mitochondrial electron transfer chain (ETC), like MPTP, induce selective SN DA neuron degeneration and parkinsonism in humans and animals, and are widely used to generate PD‐animal models (Langston and Ballard 1983; Schapira 2010; Blesa and Przedborski 2014). The dopamine metabolism itself already seems to cause intrinsically high levels of metabolic stress and free reactive oxygen and nitrogen species (ROS/RNS) in SN DA neurons (Berman and Hastings 1999; Sulzer 2007; Surmeier et al. 2011). ROS/RNS can damage proteins, lipids, and DNA of mitochondria, leading to their dysfunction, while mitochondrial dysfunction in turn provides a source of metabolic stress and ROS/RNS – propagating a self‐accelerating cycle of increasing metabolic stress in DA neurons (Henchcliffe and Beal 2008; Schapira 2008; Vila et al. 2008). Mitochondria build a dynamic network (Youle and van der Bliek 2012), where important metabolic pathways take place, like the tricarboxylic acid (TCA) cycle and the oxidative phosphorylation within the ETC, and thus ATP production, as well as control of metabolic stress and cell death, e.g. via apoptosis (Perier et al. 2012). Consequently, cells possess defense responses against mitochondrial dysfunction. This is particularly important for the highly vulnerable SN DA neurons, which are subjected to continued metabolic stress. In this context, a double‐edged role of neuromelanin is discussed. It is build by catecholamine metabolites, and accumulates particularly in highly vulnerable SN DA (and LC) neurons with age. On one hand, neuromelanin seems to protect SN DA neurons from (iron‐catalyzed) dopamine oxidation, ROS production, metabolic stress, and from degeneration, while its neuronal release seems to have opposite effects on SN DA neurons, as it can propagate microglia activation, neuroinflammation, and neurodegeneration (Hirsch et al. 1988; Zucca et al. 2015). Additional defense mechanisms in SN DA neurons against mitochondrial dysfunction and metabolic stress are mild mitochondrial uncoupling by uncoupling proteins (UCPs), and a variety of other antioxidative mechanisms, mediated e.g. by PARK‐gene functions (Liss et al. 2005; Guzman et al. 2010; Sanchez‐Padilla et al. 2014). The PARK7 gene product DJ‐1 (Bonifati et al. 2003) for instance suppresses mitochondrial metabolic stress and ROS levels particularly in SN DA neurons (Guzman et al. 2010). However, despite their defense mechanisms and their DNA‐repair machinery (Alexeyev et al. 2013), mitochondrial DNA damage accumulates in SN DA neurons with age, likely contributing to their already high metabolic stress levels, and to their progressive degeneration (Bender et al. 2006; Kraytsberg et al. 2006; Reeve et al. 2008).

Lysosomal (dys‐)function and metabolic stress in SN DA neurons

As a further protection mechanism against mitochondrial dysfunction and metabolic stress, defective mitochondria are disposed of by mitophagy, an autophagic‐lysosomal elimination (Pilsl and Winklhofer 2012; Ryan et al. 2015). This dynamic process seems to be particularly important for SN DA neurons, as a variety of PARK‐gene products are involved in this pathway, like the mitochondrial kinase PINK1/PARK6 (Valente et al. 2004; Hao et al. 2010), the mitochondrial protease HtrA2/Omi/PARK13 (Plun‐Favreau et al. 2007), the E3‐ubiquitin ligase parkin/PARK2, (Kitada et al. 1998), and the F‐box only protein 7/PARK15 (Di Fonzo et al. 2009). Loss of function mutations in these PARK‐genes trigger SN DA degeneration and cause autosomal recessive familial forms of PD (Gasser et al. 2011; van Brug et al. 2015). A variety of further PARK‐genes are associated with the autophagy–lysosome pathway (Bourdenx et al. 2014; Gan‐Or et al. 2015). For instance, the lysosomal p‐type ATPase13A2/PARK13 (Ramirez et al. 2006; Kruger et al. 2011), or alpha‐synuclein (PARK1, PARK4), which is partly degraded by lysosomes (Dehay et al. 2013; Hunn et al. 2015), the leucine‐rich repeat kinase LRRK2/PARK8 (Funayama et al. 2002; Khan et al. 2005), the LIM homeodomain transcription factor LMX1b, crucial not only for SN DA development and maintenance but also for autophagy–lysosome pathway function (Laguna et al. 2015), and the lysosomal glucocerebrosidase (GBA) (Siebert et al. 2014; Schapira 2015).

Heterozygous mutations in the GBA gene occur in about 10% of ‘sporadic’ PD cases and have just recently emerged as a major risk factor for PD (Mazzulli et al. 2011; Schapira and Gegg 2013). Reduced GBA function triggers accumulation of misfolded alpha‐synuclein and PD (Asselta et al. 2014; Schapira 2015). Mutant GBA‐protein is trapped in the endoplasmic reticulum (ER), causing not only elevated metabolic stress but also elevated Ca²⁺ release from the ER (Westbroek et al. 2011; Tan et al. 2014; Kilpatrick et al. 2015). Furthermore, dopaminergic neurons derived from pluripotent stem cells of GBA‐associated PD patients not only display lysosomal and autophagic dysfunction but also elevated intracellular Ca²⁺ levels and impaired calcium homeostasis (Schondorf et al. 2014). These findings point to another factor that contributes to metabolic stress, as well as to mitochondrial and lysosomal (dys)function in SN DA neurons: their calcium homeostasis.

Calcium homeostasis and metabolic stress in SN DA neurons

Intracellular Ca²⁺ levels, Ca²⁺ microdomains, Ca²⁺ buffering, and the mode of Ca²⁺ entry are tightly controlled in SN DA neurons, as Ca²⁺ modulates a variety of cellular functions, like excitability, neurotransmitter release, ATP‐production, apoptosis, and general regulation of enzymes and gene expression (Dolmetsch 2003; Brini et al. 2014; Burgoyne and Haynes 2014). In view of SN DA neuron vulnerability to metabolic stress, Ca²⁺ regulates mitochondrial motility (Wang and Schwarz 2009), stimulates the mitochondrial nitric oxide synthase, leading to elevated levels of ROS/RNS (Traaseth et al. 2004; Murphy 2009; Sanchez‐Padilla et al. 2014), and it stimulates enzymes of the TCA cycle and the ETC, and thus ATP production, thereby further boosting ROS generation (Gleichmann and Mattson 2011). Ca²⁺ also stimulates calpain, a protease that is involved in alpha‐synuclein processing, SN DA degeneration, and PD (Crocker et al. 2003; Dufty et al. 2007; Diepenbroek et al. 2014; Games et al. 2014). Furthermore, when mitochondrial Ca²⁺ levels rise above a certain threshold, opening of the mitochondrial permeability transition pore, and thus neurodegeneration is triggered (Martin et al. 2014). In view of differential vulnerability of DA neurons to degenerative triggers, physiological mitochondrial matrix Ca²⁺ levels appear to be higher in highly vulnerable SN DA neurons compared to more resistant VTA DA neurons (J. Surmeier, personal communication), possibly contributing to their differential vulnerability to metabolic stress and PD‐trigger factors.

As illustrated in Fig. 1, Ca²⁺ can enter SN DA neurons from the extracellular space via channels and transporters located in their plasma membrane. More precisely, via voltage‐gated calcium channels (VGCCs) and/or glutamate receptors like NMDA‐R or AMPA‐R (Bezprozvanny and Mattson 2008), or by store operated calcium channels, like STIM/ORAI and transient receptor potential channel complexes (Dietrich et al. 2014; Majewski and Kuznicki 2015). Ca²⁺ is also released from mitochondria (e.g. via the mitochondrial sodium–calcium exchanger) and the ER (via RyR and IP3 receptors) (Brini 2003; Malli et al. 2003). Spatial and temporal dynamics of cytosolic Ca²⁺ levels in SN DA neurons are further controlled by its extrusion, e.g. via plasma membrane Ca²⁺ ATPases, or sodium–calcium exchangers (NCX), and by intracellular Ca²⁺‐buffering (Schwaller 2010; Hajnoczky et al. 2014). Because of low levels of Ca²⁺‐binding proteins in SN DA neurons, like calbindin, Ca²⁺ is mainly buffered by uptake into the ER, via the sarco‐ER Ca²⁺ ATPases (SERCAs), and by mitochondria (Verkhratsky 2005; Cali et al. 2012; Hoppins and Nunnari 2012). Mitochondria take up Ca²⁺ directly via a mitochondrial Ca²⁺ uniporter mCU, or a mitochondrial Ca²⁺/H⁺ exchanger Letm1 (Baughman et al. 2011; De Stefani et al. 2011; Waldeck‐Weiermair et al. 2011), or indirectly via the ER‐mitochondria pathway (Rizzuto et al. 2009; Drago et al. 2011; Bakowski et al. 2012). This pathway is involved in defining SN DA vulnerability and in PD (Cali et al. 2014; Guardia‐Laguarta et al. 2015), and survival of cultured SN DA neurons is stimulated by modulating Ca²⁺ release from the ER (Guerreiro et al. 2015), while MPTP can induce mitochondrial as well as ER stress in DA neurons (Selvaraj et al. 2012).

Besides the ER‐mitochondria compartment, lysosomes also buffer cytosolic Ca²⁺ – via uptake by a yet unidentified Ca²⁺/H⁺ exchanger (Gomez‐Suaga et al. 2012; Osellame and Duchen 2014). The Ca²⁺ concentration is assumed to be about 5000‐fold higher in lysosomes than in the cytoplasm (Christensen et al. 2002; Xu and Ren 2015). In view of novel neuroprotective PD‐therapies, ambroxol, a well‐established drug for treating conditions with productive cough or viscous mucus (Weiser 2008), has recently emerged as a promising new tool. Ambroxol, which can pass the blood–brain barrier, reduces neuronal metabolic stress and was shown to improve GBA‐activity and lysosomal function in human fibroblasts derived from PD‐patients (Bendikov‐Bar et al. 2013; Albin and Dauer 2014; McNeill et al. 2014; Siebert et al. 2014). Ambroxol normalizes PD‐associated GBA dysfunction (and associated calcium homeostasis) seemingly by increasing its activity and its expression (Maegawa et al. 2009; McNeill et al. 2014; Ambrosi et al. 2015; Schapira 2015). Furthermore, recent findings indicate that Ambroxol can accumulate in lysosomes, and directly affect their H⁺ and Ca²⁺ homeostasis, and it stimulates exocytosis from secretory lysosomes via pH‐dependent Ca²⁺ release (Fois et al. 2015).

In conclusion, all these findings illustrate a complex homeostatic interplay of calcium homeostasis, lysosomal and mitochondrial function and metabolic stress, as well as PARK‐gene products in SN DA neurons, whose physiological action is already metabolically challenging (see below). Accordingly, dysfunction of this complex interplay triggers degeneration, particularly of metabolically compromised SN DA neurons, and it constitutes a converging pathological PD‐pathway. Consequently, its pharmacological modulation offers novel strategies for neuroprotective PD therapies – as currently under investigation for Ambroxol.

The electrical activity of SN DA neurons causes particular metabolic stress and triggers dopamine release

There is one central factor that critically modulates the complex interplay of calcium, mitochondrial and lysosomal function particularly in SN DA neurons that we have not yet considered: their electrical activity. Neuronal activity per se implies high energy demand and metabolic stress, mainly because of stimulation of the Na⁺/K⁺ ATPase, that is necessary to maintain the asymmetric ion distribution after action potentials (intracellular high K⁺, low Na⁺) and that is consuming about 50% and more ATP in active neurons (Laughlin et al. 1998; Howarth et al. 2012; Johar et al. 2014). Again, SN DA neurons seem to be particularly dependent on proper Na⁺/K⁺ ATPase activity, as mutations in its ATP1A3 subunit causes a form of rapid‐onset dystonia parkinsonism, RDP (Heinzen et al. 2014). This might be attributed to the fact that the metabolic cost of SN DA neuron activity is particularly high, as specific voltage‐gated calcium channels (VGCC, see below) are active during their physiological action, causing an activity‐related, oscillatory increase in intracellular Ca²⁺ levels (compare Fig. 1, upper insert). These oscillatory Ca²⁺ changes cause related oscillatory changes of mitochondrial membrane potentials, ROS‐levels, and of plasma membrane Ca²⁺ ATPase activity (Fig. 1) (Guzman et al. 2010; Surmeier et al. 2011; Pissadaki and Bolam 2013). It is important to emphasize that more resistant VTA DA neurons display an electrical activity that is similar to that of highly vulnerable SN DA neurons; however, it is not associated with oscillatory elevated Ca²⁺ and metabolic stress levels (Guzman et al. 2010; Khaliq and Bean 2010). Highly vulnerable locus coeruleus neurons however display similar activity‐dependent Ca²⁺ oscillations and related metabolic stress levels – that are interestingly sensitive to orexinergic peptides (orexins) (Sanchez‐Padilla et al. 2014).

Taken together, the specific mode of electrical activity of SN DA neurons generates periodically elevated Ca²⁺ and metabolic stress levels that likely render them particularly vulnerable to additional metabolic stressors and PD‐triggers. Not only because of the general metabolic cost of neuronal activity, but predominantly as a result of their Ca²⁺ channel activity. In conclusion, the elevated activity‐related Ca²⁺ and metabolic stress levels in SN DA neurons provide a molecular rationale for their particularly high vulnerability to mitochondrial, lysosomal, and PARK‐gene dysfunction, and thus for the differential vulnerability of SN DA and VTA DA neurons to degeneration in PD. Moreover, given its high metabolic costs, the oscillatory Ca²⁺ signal in SN DA neurons must be crucial for their specific physiological functions. Indeed, as SN DA neurons release dopamine in a calcium‐ and activity‐dependent manner (Rice et al. 2011; Brimblecombe et al. 2015) – from presynaptic axonal sites into the striatum (De‐Miguel and Nicholls 2015), as well as from somatodendritic sites into the midbrain (Ford et al. 2010; Rice and Patel 2015) – action potential frequency and its pattern, as well as intracellular presynaptic and somatodendritic Ca²⁺ levels in fact mediate eventually all physiological functions of SN DA neurons.

GIRK channel‐coupled D2‐autoreceptors mediate calcium‐ and dopamine‐dependent activity control of SN DA neurons

The activity of SN DA neurons in vitro and in vivo is inhibited by dopamine itself in a negative feedback loop via dopamine autoreceptors of the D2‐subtype (D2‐AR) (De Mei et al. 2009; Beaulieu et al. 2015). D2‐ARs belong to the seven‐transmembrane G‐protein‐coupled receptor family, and are located at the presynapses as well as on the somatodendritic compartment of SN DA neurons. Upon dopamine binding, the G‐protein beta‐gamma subunit gets released from the D2 autoreceptor, and directly opens G‐protein‐coupled inwardly rectifying potassium channels (GIRK2) (Luscher and Slesinger 2010). The GIRK2 activation in turn leads to hyperpolarization of the SN DA neuron membrane potential and thereby inhibits their electrical activity (Bozzi and Borrelli 2006; De Mei et al. 2009; Ford 2014; Dragicevic et al. 2015). In essence, in a negative feedback loop and via D2‐ARs, dopamine itself modulates and adapts its own activity‐dependent, presynaptic, and somatodendritic release from SN DA neurons, and consequently its physiological functions (De‐Miguel and Nicholls 2015; Rice and Patel 2015). It is noteworthy, that this D2‐AR/GIRK2 feedback regulation is particularly pronounced in highly vulnerable mature SN DA neurons, compared to more resistant VTA DA neurons (Lammel et al. 2008; Brichta and Greengard 2014; Dragicevic et al. 2015). Moreover, within the population of SN DA neurons, D2‐AR/GIRK2 activity control displays a homeostatic plasticity (Fig. 1). D2‐AR responses of SN DA neurons are sensitized in a dopamine‐, calcium‐, and age‐dependent manner: their dopamine autoinhibition – in response to phasically elevated in vivo dopamine levels (e.g. caused by cocaine or L‐DOPA), during maturation, and/or because of elevated free intracellular Ca²⁺ levels – is strengthened by functional coupling to the neuronal calcium sensor NCS‐1 (Borgkvist et al. 2014; Dragicevic et al. 2014; Ng et al. 2016; Poetschke et al. 2015). As illustrated in Fig. 1, NCS‐1 binds to D2‐receptors at their third intracellular loop in a calcium‐dependent fashion, and thereby prevents G‐protein‐coupled receptor kinase (GRK2) mediated phosphorylation, and thus beta‐arrestin mediated internalization and D2‐receptor desensitization (Kabbani et al. 2012; Beaulieu et al. 2015). We showed that this calcium‐dependent D2‐AR sensitization preferentially depends on VGCCs of the Cav1.3 channel subtype (that are discussed below) (Dragicevic et al. 2014).

In view of SN DA vulnerability in PD, sensitized D2‐AR responses appear less pronounced in PARK7 KO (DJ‐1) and in PARK8 transgenic (LRRK2‐R1441C) mice, and their mature SN DA neurons display juvenile‐like desensitizing D2‐AR responses (Goldberg et al. 2005; Tong et al. 2009; Tong and Shen 2012). Moreover, remaining human SN DA neurons from PD patients display a transcriptional dysregulation of the NCS‐1/D2‐AR/GIRK2 signaling network (Dragicevic et al. 2014). These findings indicate that dopamine‐ and calcium‐mediated sensitization of D2‐AR responses also adapts SN DA neuron activity in vivo and in PD. In context of SN DA degeneration, sensitized D2‐AR/GIRK2 responses will reduce elevated activity, and thus oscillatory calcium load, and related metabolic stress in SN DA neurons, and thus could protect them from overexcitability and excitotoxic cell death (compare Fig. 2). SN DA neurons are particularly prone to excitotoxicity, as it is triggered by elevated activity, perturbed calcium homeostasis, high levels of ROS/RNS, and metabolic stress (Prentice et al. 2015).

In conclusion, flexible dopamine‐dependent D2‐AR/GIRK2 responses in SN DA neurons allow modulation and homeostatic adaptation of their activity, their oscillatory calcium signaling and their dopamine release to physiological needs. Beyond that, D2‐AR/GIRK2 sensitization, e.g. as a result of pathophysiologically elevated SN DA activity and/or calcium overload, will counteract excitotoxic cell death, and thus represents another defense mechanism against metabolic stress. As summarized next, in this view, two different classes of ion channels have emerged to be particularly relevant for SN DA neuron activity, related calcium and metabolic homeostasis, as well as for PD‐pathophysiology: voltage‐gated calcium channels (VGCCs), and metabolically gated ATP‐sensitive potassium (K‐ATP) channels.

Bidirectional roles of voltage‐gated Ca²⁺ channels in SN DA neurons

Bidirectional roles of metabolically gated K‐ATP channels in SN DA neurons