Calcium and Parkinson’s disease

Abstract

Parkinson’s disease (PD) is the second most common neurodegenerative disease in the world. Its causes are poorly understood and there is no proven therapeutic strategy for slowing disease progression. The core motor symptoms of PD are caused by the death of dopaminergic neurons in the substantia nigra pars compacta (SNc). In these neurons, Ca²⁺entry through plasma membrane Cav1 channels drives a sustained feed-forward stimulation of mitochondrial oxidative phosphorylation. Although this design helps prevent bioenergetic failure when activity needs to be sustained, it leads to basal mitochondrial oxidant stress. Over decades, this basal oxidant stress could compromise mitochondrial function and increase mitophagy, resulting in increased vulnerability to other proteostatic stressors, like elevated alpha synuclein expression. Because this feedforward mechanism is no longer demanded by our lifestyle, it could be dispensed with. Indeed, use of dihydropyridines – negative allosteric modulators of Cav1 Ca²⁺ channels – comes with little or no effect on brain function but is associated with decreased risk and progression of PD. An ongoing, NIH sponsored, Phase 3 clinical trial in North America is testing the ability of one member of the dihydropyridine class (isradipine) to slow PD progression in early stage patients. The review summarizes the rationale for the trial and outlines some unanswered questions.

What causes PD?

PD poses a major health care challenge worldwide, disabling millions. The cardinal symptoms of the disease are motoric: slowness of movement, rigidity and resting tremor [1,2]. Although there are palliative treatments, there is no proven strategy for slowing the progression of the disease. This hole in our therapeutic arsenal stems from an incomplete understanding of what causes the disease.

Most of what we know about PD etiology comes from two very different lines of investigation. The first line of study is built upon clinical observation and pathological analysis of human PD brains. Using this approach, the core motor symptoms have been traced to the loss of dopaminergic neurons in the substantia nigra pars compacta (SNc), a small region in the mesencephalon. These dopaminergic neurons innervate the basal ganglia, a subcortical structure involved in the control of movement and action selection. In early stage patients, boosting the production of dopamine (DA) by systemic administration of a brain penetrant DA precursor (levodopa) alleviates symptoms, establishing a clear causal connection between the loss of dopaminergic neurons and motoric symptoms[2].

More detailed study of the brains of PD patients has revealed two additional features of the disease that are important to understanding its causes. First, there is neurodegeneration in other parts of the brain aside from the SNc. Although reliable quantitative estimates of neuronal loss have not been systematically compiled, in advanced PD patients there is compelling evidence of neuronal loss in a handful of other structures, like the locus ceruleus (LC), pedunculopontine nucleus (PPN), intralaminar nuclei of the thalamus (ILT) and lateral hypothalamus (LH), in advanced PD patients [3]. Importantly, these other neurons do not use DA as a neurotransmitter, arguing that DA is not uniquely linked to degeneration. Second, there is evidence of proteostatic failure in several of the neuron types affected in PD. The principal piece of evidence for this conclusion is the observed accumulation of intracellular protein aggregates in the brains of PD patients. These were first described by Lewy [4] and called Lewy bodies when they appear in the somatic region of neurons or Lewy neurites when in an axon or dendrite. A major component of this Lewy pathology (LP) is phosphorylated alpha synuclein (αSYN). What is less clear is the relationship between LP and neurodegeneration. Although commonly observed in neurons affected by PD, the question of whether these aggregates represent a cause, as opposed to a marker, of cellular degeneration is not established.

The second major line of study that has informed us about PD mechanisms is built upon genetics [5]. About one in ten PD cases is familial; that is, the appearance of the disease can be traced to a genetic mutation. Both recessive and dominant mutations associated with PD have been identified [6]. The best characterized, recessive, loss-of-function mutations have been clearly linked to mitochondrial maintenance. PINK1 and PARKIN mutations in non-neuronal cells lead to deficits in mitochondrial turnover or quality control[7]; whether this is also the case in neurons is controversial. Parkin also has been implicated in mitochondrial biogenesis through its negative regulation of the transcriptional repressor PARIS [8]. Mitochondrial biogenesis is promoted by the transcriptional co-activator PGC-1a and its target gene, NRF-1. In the absence of parkin, paris accumulates and leads to the progressive loss of dopaminergic neurons in the SNc. Mutations in DJ-1 also fall into this category; DJ-1 helps regulate cellular and mitochondrial oxidant defenses. All three of these genetic mutations are linked to mitochondrial function and increase the probability of an early onset form of PD [9,10]. Of note, LP is not an obligatory component of the pathological profile in many of the PD patients with these mutations.

Autosomal dominant mutations associated with PD have not yet been linked to any particular cellular pathway. Mutations or copy duplications/triplications of SNCA are associated with normal age-of-onset (~60 years old) PD. As SNCA codes for αSYN – a major component of LP – the proposition that elevated or misfolded αSYN drives PD pathogenesis has gained widespread acceptance. But precisely how this might happen is still unclear. One hypothesis is that oligomeric fibrils of αSYN behave like a prion, catalyzing the creation of new fibrils that spread from neuron to neuron leaving a trail of LP [11,12,13]. Although intriguing, it is not obvious how a process like this could create a relatively stereotyped pattern of pathology in PD patients without additional factors coming into play. Mutations in a number of other genes, including LRRK2, also are associated with PD [6,14,15] Considerable progress has been made toward understanding the function of the proteins encoded by these genes. While functions including autophagy and vesicular trafficking and exocytosis appear to be affected a coherent thesis regarding pathogenesis has not yet emerged. A persistent limitation in this line of study is that none of the genetic mutations associated with PD recapitulates the disease in mice, suggesting that additional factors contribute to the disease phenotype.

Although identification of specific genetic defects in familial cases of the disease offers important clues about pathogenesis, the vast majority of PD cases are idiopathic and lack a discernible genetic cause. Environmental toxins have been implicated in a few cases, but they cannot explain the vast majority. The factor most closely linked with PD onset is aging. The average age at the time of idiopathic PD diagnosis is around 60 years; PD risk increases steadily beyond that point [16,17]. Unfortunately, little data exists on the effects of aging in the brain that would lead to a better understanding of why this disease appears late in life and how it could be stopped.

What does the pattern of pathology in PD tell us?

Another clue about pathogenesis in PD is the regionally selective nature of the pathology. As mentioned above, the selective loss of SNc dopaminergic neurons is responsible for the core motor symptoms of the disease. What makes them vulnerable? There are three features of these neurons that may be mechanistically linked to their selective vulnerability: 1) the reliance upon DA as a neurotransmitter; 2) a highly branched and long axonal arbor; 3) a distinctive pacemaker phenotype that relies heavily upon Ca²⁺ entry through Cav1 channels and leads to basal mitochondrial oxidant stress.

Regarding the first two, DA has long been thought be a risk factor for SNc DA neurons [18,19,20]. The combination of elevated DA, cytosolic Ca²⁺ and αSYN could be a particularly toxic combination, especially in axon terminals [21,22,23]. Consistent with this idea, DA axon terminals are lost early in the development of PD, preceding the loss of DA cell bodies [24]. Thus, cellular stresses associated with DA metabolism could put these cells at risk. However, three facts are difficult to reconcile with the DA hypothesis. One is that levodopa therapy does not seem to accelerate disease progression [25]; if DA was the toxin driving the disease, this would not be true. Another is that many non-DA neurons are affected in PD, either degenerating or manifesting LP. Third, DA neurons in the ventral tegmental area (VTA) are relatively protected in PD. Thus, while DA is associated with vulnerability of SNc neurons, there must be other factors in PD [3].

The long, highly branched axon of SNc DA neurons is another potential risk factor. SNc DA neurons in the rodent have axons that branch profusely in the striatum and possess as many as 200,000 vesicular release sites [26]. It has been proposed that the bioenergetic demand of sustaining electrical excitability in a long unmyelinated axon like this puts SNc DA neurons at risk [27,28]. The aggregate energy consumption of a long axon is undoubtedly substantial. The question is whether this creates a bioenergetic stress, possibly related to active axonal transport functions. Recent work has provided some direct support for this hypothesis [29], by showing that in vitro mitochondrial oxidant stress is higher in SNc DA neuron axons than in the less branched axons of VTA DA neurons. Moreover, reducing the size of the arbor decreases this stress. It is also possible that this long axon poses a proteostatic burden on the somatic compartment [30]. Another possibility is that a highly branched axon is an efficient conduit for retrogradely transmitted pathogens, such as misfolded αSYN, simply because of the large number of synaptic uptake sites [31]. However, there are neurons with long, highly branched axons with as many release sites as those of SNc DA neurons, which do not succumb in PD [32]– arguing that a long axon alone is not a primary cause of neurodegeneration or LP in PD.

The third major hypothesis of selective vulnerability of SNc DA neurons is based upon physiological studies. SNc DA neurons are autonomous pacemakers, generating action potentials (spikes) at a relatively slow rate (2–10 Hz) in the absence of synaptic input [33,34,35,36,37,38,39]. The spikes in these neurons are broad, which maximizes Ca²⁺ entry and promotes slow rhythmic activity [40]. The autonomous activity is accompanied by slow oscillations in intracellular Ca²⁺ concentration that are caused by the opening of plasma membrane Cav1 Ca²⁺ channels and release of Ca²⁺ from endoplasmic reticulum (ER) and by the absence of strong buffering of intracellular Ca²⁺ [36,38,41,42,43,44,45,46]. Channels with a pore-forming Cav1.2 subunit and channels with a Cav1.3 subunit contribute to this oscillation in SNc DA neurons. The Cav1.3 channels are particularly interesting because they open at relatively hyperpolarized membrane potentials [47,48,49]. In fact, because of the depolarized operating range of SNc DA neurons, the Cav1.3 channels never close fully during the pacemaking cycle [36,39]; this and the expression of C-terminal splice variants of the Cav1.3 subunit, which minimize Ca²⁺-dependent inactivation, result in a sustained Ca²⁺ influx [50,51,52] Recently it has also been shown that low threshold L-type Ca²⁺ channels, most likely Cav1.3 channels, contribute strongly to retrograde-propagation of spikes in dendrites, further increasing Ca²⁺ entry and modulating the synaptic gain upon stimulation and consequently the initiation of burst firing [53]. Underscoring the potential importance of this Ca²⁺ oscillation, there is a clear inverse correlation between the expression of the Ca²⁺ binding protein calbindin and risk of degeneration in PD patients [54,55]; this correlation is also found in transgenic mice [56]. Hence, loss of localized intracellular control of Ca²⁺ oscillations may contribute to neuronal vulnerability to other stresses.

Initially, pacemaking activity was thought to be dependent upon Cav1 Ca²⁺ channels because dihydropyridine (DHP) inhibitors slowed or stopped pacemaking [38,44,57,58]. However, the DHP concentrations used in these experiments were orders of magnitude greater than the IC₅₀ of recombinant channels [59,60], at these concentrations both Na⁺ and K⁺ channels were also antagonized [38,44,45,57,58,61]. Using Ca²⁺ imaging of dendrites to monitor DHP antagonism of L-type channels, it was shown that nanomolar or low micromolar DHP concentrations eliminated dendritic calcium oscillations (reflecting near complete antagonism of L-type channels) without altering pacemaking rate or regularity [45]. In the absence of Cav1 channels, pacemaking was sustained by a combination of voltage-dependent Na⁺ and K⁺ channels in conjunction with a voltage-independent leak channel [39,45]; hyperpolarization and cyclic nucleotide-activated cation (HCN) channels, which are prominent in SNc dopaminergic neurons [37], also contributed to the sustained pacemaking activity. These channels appear to reside in the soma, proximal dendrites and axon initial segment.

The slow cytosolic Ca²⁺ oscillations in SNc DA neurons sub-serve three complementary functions. First, even though they are not required for pacemaking, they help maintain the slow tonic spiking in these neurons [39,44,45,62]. Second, Ca²⁺ positively modulates the expression and activity of enzymes involved in DA synthesis, ensuring that there is a match between the supply and demand of the neurotransmitter [63,64,65]. Third, they promote Ca²⁺ entry into mitochondria at specialized junctions with the ER; mitochondrial Ca²⁺ entry stimulates oxidative phosphorylation (OXPHOS) and the production of Adenosine Triphosphate (ATP) [43,66,67]. In principle, this feed-forward control of OXPHOS helps to ensure that bioenergetic needs are met [68,69,70,71] and that intracellular ATP levels do not fall into a range that would trigger protective activation of K-ATP channels, which would cause cessation of on-going activity [72]. But this bioenergetic control system comes at a cost. First, stimulating OXPHOS in the absence of strong ATP demand (that is, most of the time) leads to mitochondrial hyperpolarization, retrograde electron flux through the electron transport chain and increased production of reactive oxygen species (ROS) and reactive nitrogen species (RNS) [29,66,73]. Although antioxidant defenses of mitochondria are substantial [74], they are imperfect [75]. ROS can damage proteins, lipids and DNA, particularly in mitochondria. Sustained mitochondrial oxidant stress could be a major factor underlying a decline in mitochondrial function in at-risk neurons in humans with age [76]. The progressive loss of mitochondrial function with time – as a consequence of accumulated oxidant damage – has been proposed to be a primary driver of aging-related decline in neuronal viability [77,78]. In either case, the ongoing mitochondrial oxidant stress in SNc DA neurons could lead to an increased rate of turnover of these organelles by mitophagy [74,79]. Aging-related decline in the ability to recycle effete mitochondria could lead to premature senescence and death of the cells. Morover an enhanced need to eliminate damaged mitochondria could compromise other proteostatic functions, including the degradation of misfolded proteins, like αSYN. If so, this might suggest that accumulation of αSYN is a consequence of mitochondrial damage in SNc cells. Another downside of a Ca²⁺-mediated feed-forward system of bioenergetics supply is that it results a sustained elevation in cytosolic Ca²⁺. The bioenergetic expense of sequestering Ca²⁺ is substantial because pumping divalent Ca²⁺ against a steep electrochemical gradient (10⁴ difference in concentration across the plasma membrane). requires active transport. These gradients are maintained by ATP-dependent pumps, whose sustained engagement creates a contuous bioenergetic burden for the cells [80,81]. Lastly, Ca²⁺ promotes αSYN aggregation [82,83,84,85], because the Ca²⁺-dependent protease calpain cleaves off the carboxyl-terminus of αSYN, promoting aggregation [86]. This is further enhanced by oxidative stress [87] or DA [88]. In animal models, inhibiting calpain markedly reduces the neuronal toxicity of αSYN over-expression or mutation [89]. Interestingly, Ca²⁺ oscillations and incubation with αSYN fibrils creates a cytosolic oxidant stress that may combine with mitochondrial oxidant stress in SNc DA neurons [90]. To make matters worse, the Ca²⁺-dependent protein phosphatase calcineurin promotes αSYN toxicity [22]. Thus, Ca²⁺ entry in SNc DA neurons is Janus faced; there is clearly an upside to it, but there also is a dark side.

A common, at-risk neuronal phenotype in PD?

Do other at-risk neurons in PD exhibit a similar physiological phenotype? Although the data are sparse, the tentative answer appears to be yes. Neurons in the LC, dorsal motor nucleus of the vagus (DMV) and PPN also exhibit pacemaking activity and cytosolic Ca²⁺ oscillations [43,91,92,93,94]. Although not autonomous pacemakers, neurons in Raphe nuclei (RN) and in the nucleus basalis of Meynert (NBM) spike continuously during the waking state [95,96]. LC and DMV have also been demonstrated to express Cav1.3 channels and to exhibit elevated mitochondria oxidative stress [43,94]. In contrast, VTA DA neurons, which are relatively spared in PD, are also autonomous pacemakers but have been shown to have modest Cav1 channel currents [66,97,98], robust cytosolic Ca²⁺ buffering [56,99] and low mitochondrial oxidant stress. Additional studies are needed to determine whether there are other physiological phenotypes that contribute to vulnerability in PD.

The extent to which the axonal phenotype of SNc DA neurons is shared by other at-risk neurons is not clear at this point because the experiments to answer this question are difficult. However, it appears that neurons in the DMV, gigantocellular reticular nucleus, dorsal and median RN, LC, LH, PPN, basal forebrain and ILT all have large, diffuse axonal projections to varying degrees, distinguishing them from the vast majority of sensory or motor neurons in the brain, which typically have spatially focused, modestly branched axons that conform to topographic maps [100,101,102,103,104,105,106,107]

Does the unusual use of Ca²⁺ by at-risk neurons cause PD?

The proposition that Ca²⁺ entry through Cav1 channels and Ca²⁺-dependent feed-forward control of mitochondrial metabolism sets up SNc DA neurons (and other at-risk neurons) to be knocked over by other stressors linked to PD (i.e. aging, genetic mutations) is plausible but untested. Unfortunately, there are no genetic animal models of PD with predictive validity. That said, inhibiting Cav1 channels diminishes the damage induced by 6-hydroxydopamine and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) when given at doses that produce partial loss of SNc DA neurons [38,108,109,110]. Inhibiting Cav1 channels also protects SNc DA neurons from low doses of the environmental toxin rotenone [38]. The rodent chemical models of PD have generally proven disappointing, as they lack LP and manifest a pattern of pathology that does not resemble that in PD, even though there is cell loss in SNc.. A major obstacle in the field continues to be a lack of genetic models that recapitulate the phenotype seen in PD. Lacking such models, it has been difficult to obtain clear mechanistic insight into the factors regulating cell loss, and to test potential therapeutic interventions designed to modify the damage-inducing pathways. Recently, the use of stereotaxically injected, αSYN fibril fragments has gained popularity as a model of PD [13,111,112,113,114]. Fibrils injection into the striatum has been reported to produce degeneration in SNc DA neurons following retrograde transport. However, to date, there has been no attempt to determine whether inhibition of Cav1 channels blunts the toxicity of αSYN fibrils in vivo. What has recently been shown is that over-expression of αSYN increases oxidant stress in SNc DA neurons (see [90])and, in so doing, down-regulates the function of K⁺ (Kv4) channels that slow the spiking rate [115]. Thus, the convergence of LP and mitochondrial stress could force SNc DA neurons into a ‘death spiral’, where growing oxidant stress elicits an increase in firing rate that further augments bioenergetic demand

Despite of the limitations of the animal models, there are several observations from humans that are consistent with the proposition that Cav1 channels and Ca²⁺-stimulated mitochondrial oxidant stress are drivers of neurodegeneration and LP in PD. First, as mentioned above, the expression of the Ca²⁺ buffering protein calbindin is inversely correlated with risk of degeneration in the mesencephalon [54,116]. Second, in the human mesecephalon, there are clear signs of sustained mitochondrial oxidant stress. In the SNc from PD patients, mtDNA deletions (like those produce by ROS) are common [117,118,119,120] and Complex I activity is diminished [121,122].

But the most compelling evidence for this theory comes from epidemiological studies of patients being treated for hypertension. DHPs are FDA approved, negative allosteric modulators of Cav1 Ca²⁺ channels [123]. They are not channel blockers, in that they do not plug the channel pore. DHPs instead interact with a site buried in the lipid bilayer and impede channel activation. This interaction is dependent upon membrane potential; during membrane depolarization, DHPs bind efficiently, whereas with membrane hyperpolarization, they hardly bind at all. This is important because it makes DHPs effective inhibitors of Cav1 channels only in cells that are relatively depolarized. This explains why, in spite of widespread expression of Cav1 channels in the brain, use of brain penetrant DHPs have few, if any, short-term cognitive, sensory or motor effects in humans. For decades, DHPs have been used in humans to treat hypertension. They lower blood pressure primarily by inhibiting Cav1 channels in vascular smooth muscle, causing them to relax and dilate blood vessels. More importantly for PD, use of DHPs to treat hypertension has consistently been linked to a substantial reduction in the risk of being diagnosed with PD (~30% reduction in risk)[124,125,126,127,128]. This is similar to the risk reduction linked to smoking tobaccco [129,130,131]. This reduction in risk was not related to simply lowering blood pressure as treatment with beta-adrenergic blockers or acetylcholinesterase inhibitors was not associated with reduced risk, in spite of the fact both are typically more effective treatments for hypertension. Treatment with DHPs also is associated with slowing of disease progression even if initiated after diagnosis [132].

Although consistent with the hypothesis, these epidemiological studies are retrospective, and a definitive test requires a prospective, randomized, placebo-controlled study. Phase II tolerability trial in early stage PD patients was conducted and found that the DHP isradipine is well tolerated at doses below 10 mg/day [133,134]. Isradipine was chosen for this study because of its excellent pharmacokinetics and nearly equal affinity for Cav1.2 and Cav1.3 channels [48,135,136]. In 2013, NIH funded a 5 year, randomized, placebo-controlled, double blind, Phase III multicenter study to determine whether a 3 year regimen of isradipine treatment could slow the progression of PD in recently diagnosed patients. The study has enrolled over 300 patients and is slated for completion in 2019.

Conclusions and Challenges

The causes of PD in humans remain unresolved, although growing understanding of the common characteristics of neurons affected by the disease has led to hypothesis-driven studies to determine the significance of these in the pathogenesis. One such theory is that the pattern of pathology in PD is dictated by an unusual neuronal phenotype. This phenotype includes a long highly branched axon and an unusual pacemaking phenotype that engages Cav1 Ca²⁺ channels in a feed-forward control of mitochondrial OXPHOS, leading to sustained mitochondrial oxidant stress. This oxidant stress, over time, could lead to progressive loss of mitochondrial function that results in bioenergetic deficits. Oxidant stress also could increase proteostatic dysfunction by damaging cellular proteins and increasing mitophagic flux, broadly compromising the ability to deal with protein dysfunction. In addition, the elevation in cytosolic Ca²⁺ concentration accompanying this feedforward mechanism could promote aggregation of αSYN, a major component of LP. Like Humpty Dumpty who by sitting precariously on a high wall becomes susceptible to falling and breaking into pieces, this phenotype could poise SNc dopaminergic neurons, and other neurons with a similar phenotype, up for a ‘fall’ when pushed by other stressors that normally could be tolerated, including genetic mutations, environmental toxins or reactive neurotransmitter (i.e., DA).

This ‘Humpty Dumpty’ hypothesis is supported by observations in humans, most compellingly epidemiological data linking use of DHP inhibitors of Cav1 channels to decreased risk of developing PD and slowed progression. The ongoing NIH sponsored Phase III clinical trial with the DHP isradipine will be the first prospective test of this theory of pathogenesis. One limitation in this trial is the outcome measure: the unified Parkinson’s disease rating scale (UPDRS). The UPDRS is comprehensive but noisy, as it is somewhat subjective, so with 56 centers involved in this trial, with several clinicians at each site rating patients, there will be unavoidable experimental variability that could obscure a modest effect. However there are not any validated, objective biomarkers that would provide a more uniform assessment of disease progression. Another limitation is that the trial is restricted to patients who are symptomatic. Unfortunately, symptomatic presentation in PD appears relatively late in the disease process when a significant fraction of their SNc dopaminergic neurons are already gone. As we have learned from many clinical trials of disease modifying interventions, efficacy is often optimized when treatment is initiated at very early stages. This is not possible in PD, so the hope is that relief of the mitochondrial stresses will extend survival of the remaining neurons. If isradipine treatment were to significantly slow disease progression, this could have a profoundly positive impact on the quality and duration of life of these patients, in addition to the obvious benefits in terms of limiting health care costs.

There are many unanswered questions and challenges related to the general question of selective vulnerability and pathogenesis in PD. Answering these questions could lead to new neuroprotective strategies. From our perspective, these include:

• Why is age the number one risk factor, and what factors that change with aging contribute to neurodegeneration in PD?

• Are neurons at risk in PD preferentially sensitive to the genetic mutations that have been linked to PD? For the recessive mutations affecting mitochondrial function the connection seems clear. But it is not as evident for the dominant mutations in SNCA, LRRK2 and GBA, for example.

• Are the long, branched axons of SNc DA neurons responsible for increased vulnerability? Recent work in vitro has shown that SNc DA axons have greater engagement of Cav1 channels [23] and elevated mitochondrial oxidant stress than VTA DA axons [29], but it isn’t clear how they are related and whether DA is a factor.

• To what extent do other neurons at-risk in PD have the same anatomical and physiological phenotype as SNc DA neurons? This has been addressed for LC and DMV neurons but the vast majority of at-risk neurons have not been studied in a way that would inform theories of pathogenesis.

• Does the at-risk phenotype promote αSYN aggregation or propagation of αSYN pathology? Does LP preferentially kill some types of neuron and, if so, how? The two major hallmarks of PD are neurodegeneration and LP. Yet, it is unclear whether the latter is a marker or a mediator of the former. As outlined above, it is possible that Ca²⁺ handling in at-risk neurons promotes LP, but this needs to be tested rigorously

Highlights.

• Parkinson’s disease (PD) is the second most common neurodegenerative disease

• Core PD motor symptoms are caused by the death of pacemaking dopaminergic neurons

• These neurons use Cav1 Ca²⁺ channels to support pacemaking and bioenergetics

• Resulting high cytosolic Ca²⁺ and mitochondrial stress could drive degeneration

• Inhibitors of Cav1 channels may decrease PD progression