A Clue to Neuron Vulnerability

Three recent papers published in the Journal of Neuroscience address the mechanisms that underlie the activity of dopamine neurons in the Substantia Nigra.  The interest in this work stems from the hypothesis that the persistent influx of calcium resulting from the combination of constant pacemaker and burst firing, is a major contributor to vulnerability of these cells to the toxic consequences of elevated intracellular calcium.

A substantial portion of the inward current that underlies single action potentials in dopamine neurons of the substantia nigra is carried by calcium (Bean, 2007).  The fact that these neurons are constantly firing in one of two modes, pacemaker or bursting, suggests that significant energy is required to maintain a low level of calcium in the cytoplasm.  By understanding the mechanisms that underlie the firing characteristics of these neurons, it may be possible to devise ways to maintain the normal physiological properties of these cells and at the same time reduce the energy burden resulting from constant influx of calcium.

The combination of ion conductances that underlie the spontaneous pacemaker activity of dopamine neurons has been studied extensively and remains the subject of controversy and interest.  Putzier et al (2009) used the dynamic clamp method to examine the role of the dihydropyridine sensitive channel, Ca_v1.3, in pacemaker firing.  Addition of the Ca_v1.3 conductance to neurons with the dynamic clamp after the endogenous conductance was blocked with a dihydropyridine antagonist resulted in normal pacemaker activity.  There were several significant conclusions of this work.  First, calcium entry mediated by Ca_v1.3 was not required.  Second, it was the voltage dependence of activation that was critical for pacemaker activity rather than the kinetics of activation or ion selectivity.  Third, pacemaker activity was restored with a conductance applied at the soma.  Finally, the activation of NMDA receptors caused a transient increase in firing that was not maintained.  Part of the interpretation of these results may be limited by the use of relatively high concentrations of the dihydropyridine antagonist that may affect other channels that contribute to pacemaker activity (Guzmann et al., 2009).  In spite of that potential limitation, there is agreement that calcium entry is not required to maintain tonic activity.  The fact that these two studies reach this conclusion is a significant advance.

Bursts of action potentials in dopamine neurons are thought to be critical for the phasic release of dopamine in projection areas.  The combination of excitatory synaptic drive and intrinsic membrane properties are necessary to produce this activity (Blythe et al., 2009; Deister et al., 2009).  Different approaches were used in the two studies to reach common conclusions.  One conclusion was that activation of NMDA receptors on the cell soma rather than dendrites was sufficient to activate bursts action potentials.  Each of these studies indicated that bursts of action potentials were dependent on a combination of intrinsic properties and NMDA receptor activation.  The action potential was initiated near the axon hillock on a proximal dendrite or soma, the block of sodium channels near the site of action potential initiation decreased burst firing, bursts of action potentials did not require calcium entry through Ca_v1.3 channels (Blythe et al., 2009).  It was the voltage dependence of the NMDA conductance that was necessary for the generation of bursts and intrinsic (but unidentified) conductances contributed to the NMDA current to drive the high frequency firing (Deister et al., 2009). 

Taken together this group of papers significantly advances the understanding of the firing rate and pattern of dopamine neurons in the substantia nigra.  By understanding these basic properties it will be possible to determine how dopamine neurons in the substantia nigra differ from those in the more medial Ventral Tegmental Area.  The differences between the two groups of neurons will be important clues to the factors that put the neurons in the substantia nigra at risk.  Given that the primary risk factor for Parkinson’s disease is age, it seems critical to extend the work that is presented in these studies to much older animals. 

Dr. Williams has written an informative and insightful assessment of some of the recent work bearing on calcium and the selective vulnerability of substantia nigra pars compacta (SNc) dopaminergic neurons – a topic of great importance to the Parkinson’s disease community. As pointed out in the post, the studies of Deister et al. (Deister et al., 2009) and Blythe et al. (Blythe et al., 2009) provide complementary insights into the mechanisms underlying burst firing that could be particularly valuable in assessing the role glutamatergic synaptic inputs in triggering excitotoxic damage to SNc dopaminergic neurons. The interaction between NMDA receptors – identified by both studies as central to burst firing – and metabotropic glutamate receptors that regulate the release of calcium from intracellular stores promises to be a rich topic of investigation in the near future  (Morikawa et al., 2003; Harnett et al., 2009).

The role of calcium and Cav1.3 calcium channels in governing pacemaking – the other major spiking mode of SNc dopaminergic neurons – is dealt with parenthetically by Dr. Williams’ post, but merits more extended comment in my view. The recent paper by Putzier et al. (Putzier et al., 2009) makes the assertion that calcium influx is not necessary for pacemaking, as do earlier papers by our group (Chan et al., 2007; Guzman et al., 2009). However, there is a crucial difference between their conclusions and ours. Putzier et al. claim that Cav1.3 calcium channels are necessary for pacemaking. We do not. Their conclusion is based upon 1) the ability of a high concentration (10 µM) of the dihydropyridine nimodipine to stop pacemaking and 2) the ability of dynamic clamp injection of current mimicking that expected to flow through Cav1.3 channels to restore pacemaking in the presence of nimodipine. There is no question about the quality of the data. But for the author’s interpretation to be correct, two assumptions must hold. First, the concentration of nimodipine used (10 µM) must selectively antagonize Cav1.3 channels. Our previous work shows that this isn’t the case and that at this concentration, which is roughly 500 times the IC50 for the high affinity state of these channels (Sinnegger-Brauns et al., 2009), voltage-dependent Nav1 sodium channels are also antagonized (Guzman et al., 2009). These sodium channels help support pacemaking by providing depolarizing inward current in the same voltage range as Cav1.3 calcium channels. Our studies also show that much lower concentrations of another dihydropyridine (isradipine, a more potent Cav1.3 antagonist) eliminate any measureable calcium influx (directly assessed with calcium imaging) without altering pacemaking frequency. Hence, the pharmacological manipulation in the Putzier et al. experiments antagonized at least two channels that are important to the pacemaking process, not one. The second key assumption made by Putzier et al. is that the current provided by the dynamic clamp in their experiments is unique to Cav1.3 channels. But, clearly this isn’t the case either. Voltage-dependent Nav1 sodium channels provide a depolarizing inward current with similar kinetics in the same voltage range. Because our data argue that the ability of high concentrations of dihydropyridine to stop pacemaking is due to the collateral antagonism of sodium channels (not to antagonism of Cav1.3 calcium channels), the restoration of pacemaking by adding back a similar conductance is completely expected. 

There are two other observations that argue against the necessity of Cav1.3 channels in pacemaking. First, in mice lacking functional Cav1.3 channels, pacemaking in SNc dopaminergic neurons is normal in rate and regularity (Chan et al., 2007). Second, even in the presence of high concentrations of dihydropyridine, pacemaking can be restarted either by waiting for transcriptional compensation or by antagonizing Kv1 potassium channels (Chan et al., 2007; Guzman et al., 2009). There is one other point that needs to be made about the conclusions advanced by Putzier et al. Although Cav1.3 calcium channels are not necessary for pacemaking, when they are engaged their ability to allow the influx of calcium has important electrophysiological consequences. There is a substantial literature showing that in SNc dopaminergic neurons small conductance calcium-activated potassium channels (SK) channels are activated by calcium influx through L-type channels and that this activation regulates pacemaking pattern and rate (Ping et al., 1996; Wilson and Callaway, 2000; Wolfart et al., 2001). Our modeling results are consistent with this experimental work, showing that antagonism of Cav1.3 calcium channels has little effect on pacemaking in large measure because the antagonism is accompanied by a reduction in calcium influx and down-regulation of SK potassium channel opening, serving to balance the transmembrane current ledger.

Taken together, these studies argue that while Cav1.3 channels normally help to support pacemaking in SNc dopaminergic neurons, they are not necessary and that the functional deficits accompanying their antagonism with dihydropyridines or other more selective drugs should be minimal. This conclusion not only is consistent with the absence of significant motor or cognitive side-effects with dihydropyridine use by humans, it is consistent with the tolerability of isradipine in early stage Parkinson’s disease patients, where any drug-induced impairment of SNc dopaminergic neurons should be immediately apparent.

We were pleased that Dr. Williams noted the significance of Putzier et al. (2009), which showed that Cav1.3 channels support nigral dopamine neuron pacemaker activity and amplify bursts by virtue of their voltage dependent gating rather than their selectivity for calcium. However, the subsequent opinion by Dr. Surmeier suggested that Putzier et al. (2009) misinterpreted dynamic clamp results. Here we address Dr. Surmeier's concerns.

1. Dr. Surmeier begins by stating, "Putzier et al. (Putzier et al., 2009) makes the assertion that calcium influx is not necessary for pacemaking ...", a conclusion that he later states is incompatible with the known role of calcium-activated potassium channels. However, our conclusion was that Cav1.3 channel calcium selectivity is not functionally significant, and we specifically stated that a role for calcium influx through other calcium channels that are known to activate calcium-activated potassium channels (Wolfart, 2002) is compatible with our results.

2. The use of 10 μM nimodipine to block native Cav1.3 channels was criticized because this dose was asserted to be higher than needed. However, this point rests on problematic evidence and assumptions. First, Dr. Surmeier cites equilibrium binding data from depolarized membranes measured under nonphysiological conditions that are known to overestimate the functional effects of dihydropyridines by orders of magnitude and to not account for differences in sensitivity among Cav1.2 and Cav1.3 channels (Koschak et al., 2001). In contrast, the dose used in our study is supported by papers that have directly measured inhibition of Cav1.3 channel function in intact cells and brain slices (Koschak et al., 2001; Xu and Lipscombe, 2001; Safa et al., 2001; Puopolo et al., 2007). Second, Guzman et al. (2009) rely on distal dendritic calcium responses to ascertain effectiveness of their dihydropyridine treatment, but this assay is an indirect nonlinear assay of Cav1.3 channel function. Finally, Putzier et al. demonstrates that Cav1.3 channels in the soma can support pacemaking, thus lowering the significance of the effect in distal dendrites. Based on published direct measurements (Koschak et al., 2001; Xu and Lipscombe, 2001; Safa et al., 2001; Puopolo et al., 2007), the dihydropyridine dose used in Putzier et al. is necessary, while lower doses will not completely inhibit somatic Cav1.3 channels.

3. Surmeier used modeling and results with 20 μM isradipine to argue that dihydropyridines inhibit pacemaking by blocking leak channels and voltage-gated sodium channels rather than Cav1.3 channels (Guzman et al., 2009). However, the participation of these non-Cav1.3 channels without a larger contribution of dihydropyridine-sensitive Cav1.3 channels is questionable. First, the isradipine inhibition of sodium channels reported by Surmeier (Guzman et al., 2009) is modest compared to the known isradipine inhibition of Cav1.3 channels (Koschak et al., 2001). Second, the contribution of sodium channels to subthreshold currents in dopamine neurons is smaller than calcium channels (Puopolo et al., 2007). Third, Mercuri et al. (1994) showed that much lower dihydropyridine doses, which do not affect sodium channels, are sufficient to slow pacemaker activity in dopamine neurons. Fourth, the inhibition of sodium channels would change the upstroke and peak of the action potential, but these effects are not apparent in nimodipine experiments (Putzier et al., 2009 and unpublished results). Finally, the dynamic clamp showed that leak channels cannot support pacemaker activity in dopamine neurons (Putzier et al., 2009). This illustrates a major strength of the dynamic clamp: a conductance is altered in a defined manner and native channels are allowed to govern the cell's response. In contrast, modeling relies on assumptions about the excitability of dopamine neurons. Taken together, the above experimental data show that Cav1.3 channels are the major target of dihydropyridines.

4. Dr. Surmeier states that functional deficits accompanying Cav1.3 antagonism with dihydropyridines or other more selective drugs should be minimal. However, three papers cited by Dr. Williams (Putzier et al., 2009; Deister et al., 2009; Blythe et al., 2009) each independently demonstrate that dihydropyridines reduce the intensity of NMDA receptor-evoked bursts, an effect that may be significant for cognition and motor function. It is also not clear that the doses used to partially inhibit the more sensitive Cav1.2 channels in vascular smooth muscle and the heart (Koschak et al., 2001;, Xu and Lipscombe, 2001) are sufficient to prevent Parkinson's disease-induced death of dopamine neurons. If higher doses are needed to protect dopamine neurons, then disastrous effects may be induced in the cardiovascular system. Thus, further work with Cav1.3 inhibitors will be required to assess their impact on dopamine neurons in vivo.

5. Dr. Surmeier implies that the conclusions in Putzier et al. (2009) do not apply in knockout mice or after remodeling of dopamine neuron excitability after prolonged Cav1.3 channel inhibition. On this point we agree: the results in Putzier et al. (2009) only apply to the normal substantia nigra dopamine neuron.

6. The authors of Putzier et al. (2009) hold no patents regarding calcium channels in dopamine neurons and so arrived at their conclusions in the absence of conflict of interest.

Based on the above responses, the conclusions in Putzier et al. (2009) remain valid. Of course, we hope that recent publications will stimulate new experiments to further clarify the role of Cav1.3 channels in dopamine neurons.