Multiple Hits Explain Selective Loss of Dopamine Neurons in PD
One of the enduring mysteries of Parkinson disease has been the exquisite specificity of neuronal death. The movement disorder appears when a cluster of dopaminergic neurons in the substantia nigra (SN) degenerate, but at the same time their dopaminergic neighbors in the ventral tegmental area (VTA) stay alive and well. Explanations for SN neuron death have invoked dopamine, calcium, and the α-synuclein protein in turn as the pathologic players, but none of these factors on its own explains specificity. A new study from David Sulzer and colleagues at Columbia University Medical Center in New York proposes a solution, though not a simple one. Their data, appearing in the April 30 issue of Neuron, suggest that SN neurons brew up a perfect storm of dopamine plus calcium plus synuclein that leads to their selective destruction. Knocking out any one leg of the toxic tripod prevents cell death, which opens up a trio of pathways for potential new therapies.
The study centers on a novel technique, developed by first author Eugene Mosharov, to measure cytosolic dopamine concentrations in living cells (see ARF related news story). When the investigators turned their technique on SN neurons in ventral mesencephalon cultures from mice, they found that the cells have undetectable cytosolic dopamine (less than 100 nM). However, adding L-DOPA to the cells in high concentrations boosted cytosolic dopamine to measurable levels, and caused neurotoxicity. Further studies indicated that cell death was tied to cytosolic dopamine levels: Neurotoxicity depended on the conversion of L-DOPA to dopamine, and reducing cytosolic dopamine levels by overexpression of the vesicular dopamine transporter VMAT2 protected the cells from death. When the experiments were done in cells from α-synuclein knockout mice, cytosolic dopamine levels were still elevated, but toxicity was reduced. That suggested that α-synuclein was downstream of dopamine in the neurotoxic cascade. Finally, the investigators showed that blocking dopamine metabolism by inhibiting monoamine oxidase also protected the cells from death, suggesting it is not dopamine per se, but a metabolite that is the toxin. All together, the data support a model where high cytosolic dopamine leads to generation of oxidized metabolites that interact with α-synuclein to cause neurotoxicity (e.g., see Burke et al., 2008 and Martinez-Vicente et al., 2008).
Next, things became really interesting when Mosharov and colleagues compared SN cells to VTA dopaminergic neurons. They found that after treatment with L-DOPA, the SN cells developed two to three times higher levels of cytosolic dopamine than VTA cells. Along with lower levels of cytosolic dopamine, the VTA cells were more resistant to the toxic effects of L-DOPA. There was no apparent difference in uptake, storage, or degradation of dopamine between the two cell types, suggesting that elevated synthesis was to blame for the higher cytosolic dopamine levels and neurotoxic response to L-DOPA in the SN.
But why? Previous work from the lab of James Surmeier at Northwestern University in Chicago, Illinois, showed that SN neurons have an unusual physiology involving a calcium-dependent pacemaking activity that results in higher overall calcium fluxes than in VTA neurons (see ARF related news story). The calcium flows through dihydropyridine-sensitive L-type CaV1.3 channels, and the Chan paper suggests that calcium-related toxicity may explain the age-related degeneration of the cells. To test if calcium was regulating dopamine synthesis or toxicity, Mosharov and colleagues treated SN cells with inhibitors of the CaV1.2/1.3 channels and found that the L-DOPA-induced elevation in cytosolic dopamine was diminished to where it resembled that in VTA cells. The SN cells also became resistant to the neurotoxic effects of L-DOPA. The channel blockers had no effect on dopamine uptake or breakdown, leaving the likely target of regulation to be the calcium-sensitive amino acid decarboxylase enzyme, which catalyses the conversion of L-DOPA to dopamine.
“The results suggest you need a convergence of high cytosolic dopamine, synuclein, and high calcium to explain the selective sensitivity of SN neurons,” Sulzer told ARF. That makes the picture complicated, he said, but it also opens up multiple opportunities for therapies. Based on Surmeier’s previous work, clinical trials have already started on L-type calcium channel blocker isradipine for PD. The drug is attractive because of its long track record in humans, although a compound more selective for the CaV1.3 might be even better, Sulzer said. Strategies to lower α-synuclein are another possibility. A third prospect is to target cytosolic dopamine, and Sulzer and Mosharov said they are working on overexpression of the VMAT2 transporter as one approach. By loading dopamine into vesicles, VMAT2 lowers cytosolic dopamine and protects cells, as they show in the current paper. Finally, inhibitors of monoamine oxidase including selegiline and rasagiline are used to treat PD, and are being investigated for neuroprotective actions. “People are finding they work, but they don’t know why," Sulzer said. “Our paper may explain that, by suggesting that they block the metabolism of dopamine to a form that interacts with α-synuclein to make a toxic species.”
In an accompanying commentary, Surmeier ties the higher calcium flux in SN cells to the idea that mitochondria are also important in the pathogenesis of PD. He speculates that perhaps SN cells that constantly handle a higher calcium flux end up taxing their mitochondria because these organelles buffer calcium, and over the long term they have to supply the energy to maintain calcium levels. “Because aging diminishes the mitochondrial capacity to generate ATP through oxidative phosphorylation, neurons with a calcium-reliant pacemaking phenotype should be operating nearer and nearer their metabolic capacity, diminishing their ability to withstand episodic challenges,” Surmeier writes.
A related paper by Kim Tieu of the University of Rochester, New York, and Serge Przedborski of Columbia University solves another mystery by providing a missing link in a common model of PD, namely MPTP toxicity. MPTP is a contaminant in heroin that was found to cause an acute parkinsonism in some drug users, and it has become widely used in animal models to mimic the SN degeneration of PD. Researchers know that MPTP is taken up by astrocytes and converted to the mitochondrial toxin MPP+, which is selectively taken up by neurons that express the dopamine transporter. However, it was not clear how MPP+ got out of astrocytes. Tieu and Przedborski present evidence that the organic cation transporter-3 (Oct3) is expressed in astrocytes adjacent to midbrain dopaminergic neurons and is necessary for death of dopaminergic neurons in mice treated with MPTP. The bidirectional transporter may also protect cells under some conditions, the data suggest, by taking up excess extracellular dopamine or other toxins. The paper appears in the PNAS early edition this week.—Pat McCaffrey
Ann Romney Center for Neurologic Diseases, BWH and HMS
This paper by Mosharov et al. takes a very contemporary look at an old question: Why are the dopaminergic neurons of the substantia nigra pars compacta (SNc) more vulnerable in Parkinson disease than those in the neighboring ventral tegmental area (VTA)? By considering the role of α-synuclein and the unique calcium-dependent pacemaking activity of SNc neurons recently shown by Jim Surmeier’s group, David Sulzer and colleagues propose a novel and provocative model for the molecular basis of the unique susceptibility of the SNc in disease.
By demonstrating not only an enhanced vulnerability of nigral neurons to L-DOPA-induced toxicity, but also providing a pharmacologic basis for their observations, they put forth an interesting hypothesis regarding the selective accumulation of neuromelanin in this brain region, as well.
One caveat that must be openly considered, and is likewise raised in the thoughtful commentary by Jim Surmeier that accompanies the Mosharov report, is that the endogenous levels of cytoplasmic dopamine cannot be measured by currently available technologies. Thus, the authors are only able to follow the kinetics of L-DOPA and dopamine metabolism in cultures treated with micromolar concentrations of L-DOPA, thus raising the cytoplasmic dopamine levels into detectable range. As the authors are first to point out, the levels of L-DOPA provided to their neuronal cultures are about three orders of magnitude above those that occur following therapeutic administration of L-DOPA to Parkinson’s patients and likely seven orders of magnitude above those seen in normal human brain. One wonders whether the differences in aromatic amino acid decarboxylase (AADC) activities suggested by these data at 100 μM L-DOPA would hold true in the presence of only 5 nM L-DOPA. It seems possible that at potentially saturating concentrations of L-DOPA one observes differences that would not exist at the extremely lower physiological or even therapeutic levels of substrate. This potential confound is also consistent with the classical view that tyrosine hydroxylase and not AADC is the rate-limiting step in dopamine synthesis, in vivo. With the argument that differential AADC activities could play a determining role in cytoplasmic dopamine levels, one then places AADC at the bottleneck. Even with this caveat in mind, however, it must be acknowledged that the methods applied here define the state-of-the-art for measuring dopamine and the technical hurdles already surpassed by this group are not trivial.
One last note regarding the authors’ hypothesis that calcium may regulate the activity of AADC; in Figure 6 of the report the authors nicely show the differential response of SNc and VTA neurons to inhibitors of the L-type calcium channels. While not specifically emphasized by the authors, I found it quite interesting that the calcium chelator BAPTA-AM, or use of cadmium, had significant inhibitory effects on L-DOPA metabolism in both culture systems. I would interpret these data as very consistent with their model of a potent, calcium-dependent effect on dopamine synthesis and/or metabolism. Further work will be required to specifically evaluate the influence of calcium, whether it be calcium that specifically passes across L-type calcium channels, or not. These data, along with the incrimination of α-synuclein in the mechanism L-DOPA- or dopamine-dependent toxicity in this culture system, are likely to influence the future experiments of many of us who study the etiology of Parkinson disease.
National Institute on Aging
One of the difficulties with the longstanding idea that dopamine is a contributor to neuronal cell loss in Parkinson disease (PD) and related disorders is that it appears neither to be necessary nor sufficient to cause cell death. Dopamine can’t be necessary for cell death as there are non-dopaminergic neurons that are lost throughout the disease process. And dopamine on its own is unlikely to be sufficient to cause neurodegeneration, as there appear to be dopamine neurons that are relatively spared in PD.
This paper by Mosharov et al. highlights the idea that simple rules are unlikely to explain complex diseases. The authors address the sufficiency argument by showing that dopamine contributes to toxicity only in some contexts. Calcium is also highlighted and is a pretty good candidate for a modifier, especially given the relatively poor calcium buffering capacity of dopaminergic nigral neurons that is probably intrinsically related to the physiological role they have to serve. Again, it seems unlikely that calcium alone is either necessary or sufficient to explain the complex patterns of relative vulnerability in PD, but one could imagine that both neurotransmitter identity added to a specific physiological role of this group of neurons might impact their long-term survival.
The most interesting candidate identified here, albeit briefly in one figure, is α-synuclein. We know from other evidence that α-synuclein can be a causal factor in PD and that the difference between no disease and an aggressive, progressive diffuse Lewy body disorder is determined by α-synuclein expression over a twofold range. The data here that α-synuclein influences nigral neuron survival without shifting intracellular dopamine levels strongly implicate this protein as having a detrimental effect on cells downstream of other toxic stressors. This is consistent with a number of previous observations, including the relative resistance of α-synuclein null mice to different stressors and the proposal that there might be a toxic pairing of α-synuclein and dopamine.
But why α-synuclein would have this property of cascading damage is not at all obvious. Why even would neurons make a protein at relatively high levels if that protein is a perennial bad apple? It is interesting that smaller, simpler organisms such as fruit flies do apparently fine without any synuclein genes. Perhaps α-synuclein has some important benefit to the function of a neuron in the context of either longer-lived species (we know that aging is another contributory factor in PD) or specific physiological needs (which might relate in some way to the calcium-mediated effects seen here, although that link seems immediately tenuous). I think this question might be approached either by understanding what the normal of function of α-synuclein is and by identifying why exactly it is toxic to cells under some conditions. The Mosharov paper is an important start to this process.
This is a great paper. The idea of cytosolic dopamine being toxic has been unsubstantiated dogma in the field for years. There have been numerous indirect indications of free dopamine contributing to the vulnerability of the substantia nigra neurons, but we had to take leaps of faith when discussing this. The lingering comment was always "if we only had a way of measuring cytoplasmic dopamine in nigral neurons." Well, the Sulzer laboratory has utilized an elegant combination of techniques to achieve this.
More importantly, they used the technique to help unify several of the hot topics in the field, namely, calcium regulation and α-synuclein expression. Application of this technique in other transgenic models related to Parkinson disease should help advance the field even further.
From a patient standpoint, these findings do suggest that there could be new treatments on the horizon. The better we can manage dopamine inside and outside the dopamine neuron, the better we can manage the disease progression and therapeutic treatment.