Summary

The neocortex and basal ganglia operate together through cortico-basal ganglia-thalamocortical circuits to execute movements and goal directed behaviors. The neocortex communicates with the basal ganglia through parallel sets of glutamatergic corticostriatal projections that synapse on striatal medium-sized spiny neurons. These excitatory signals are modified at the medium spiny neuron by dopamine, released from midbrain projections and by acetylcholine released from striatal interneurons. Combined, these neurotransmitters provoke state changes in medium spiny neurons and determine whether signalling proceeds through direct or indirect striatal pathways that ultimately determine basal ganglia output. Neurodevelopmental and neurodegenerative disorders that affect motor function are associated with abnormal activation of the corticostriatal pathway. The abnormal release of dopamine, acetylcholine and glutamate have been implicated in numerous childhood diseases including juvenile Parkinson's disease, L- dopa-responsive dystonia, drug dependence, Huntington's disease, Tourette syndrome and obsessive compulsive disorder . Our laboratory has examined how alterations in dopamine, acetylcholine and glutamate might produce bradykinesia in parkinsonism, hyperkinesias in drug dependence and progressive motor degeneration in Huntington's disease.

The effect of dopamine on the excitability of striatal cells is complex and depends on the interplay between dynamic intrinsic cellular and synaptic properties. Dopamine is thought to play a crucial role in reinforcement and learning because it signals error in the prediction of future reward and switches attentional and behavioral selections to unexpected, behaviorally important stimuli. This switching response prepares the organism for an appropriate reaction to biological significant events. Through direct corticostriatal modulation, dopamine reinforces cortically-initiated activation of a particular basal ganglia-thalamocortical circuit by facilitating conduction through the circuit's direct pathway, which has a net excitatory effect on the thalamus, and suppresses conduction through the indirect pathway, which has a net inhibitory effect on the thalamus.

In addition to its known direct effects at the medium spiny neuron, our laboratory has demonstrated that dopamine modulates corticostriatal neurotransmission through presynaptic D2 receptors located on corticostriatal terminals. These presynaptic D2 receptors selectively inhibit a subset of corticostriatal terminals with a low probability of release, while terminals with a high probability of release remain unperturbed. The effects of dopamine on corticostriatal release are frequency dependent and increase with the frequency of cortical stimulation. Dopamine inhibition of glutamate release is minimal with low-frequency stimulation (1 Hz) but increases with higher rates of stimulation (20 Hz), with dopamine inhibition remaining selective for terminals with a lower probability of release. This dependence on frequency is also evident post-synaptically, with a reduction in excitatory postsynaptic currents, best seen at higher rates of cortical stimulation. Thus, dopamine acts as a low-pass filter, but the filtering is applied selectively to terminals with a low probability of release. In this way, dopamine released by behavioral stimuli can directly regulate striatal neurotransmission by selecting sets of corticostriatal projections and can process bursts of cortical information while rejecting others.

Striatal filtering is altered in diseases that affect dopamine release. In dopamine-deficient states, the reduction in dopamine availability prevents normal striatal filtering by sensitizing D2 dopamine receptor responses. Using optical recordings of presynaptic release, our laboratory determined the effect of dopamine deficiency and replenishment at single cortical synaptic terminals in mouse models for acute and chronic dopamine depletion. Using reserpine-treated and dopamine-deficient transgenic mice, our investigations demonstrated that dopamine depletion produces sensitized presynaptic D2 receptor responses and altered dopamine-mediated responses in subsets of corticostriatal terminals. In control mice, dopamine was found to inhibit ~85% of cortical terminals though D2 receptor actions that depressed exocytosis from terminals with a low probability of release. In contrast, for both reserpine-treated and dopamine-deficient mice, D2 receptor stimulation depressed release from both fast and slow-releasing terminals. Sensitized D2 receptor responses more broadly inhibited corticostriatal release and promoted further inhibition at the slow-releasing terminals. We also discovered that dopamine is not required for the formation of normal corticostriatal cytoarchitecture, glutamate density or basal striatal glutamate concentrations, suggesting that dopamine is essential for movement, but is not required for the development of neural circuits that control those behaviors .

Similar to dopamine-depleted states, dopamine excess, as modelled by repeated psychostimulant use, also produces abnormal corticostriatal filtering. Theoretical models of the synaptic changes underlying the neuronal basis for addiction require two processes 1) a long-term synaptic change that is selectively produced by chronic addictive drugs, and 2) a synaptic basis for renormalization by re-administering that drug after withdrawal, i.e., a "paradoxical" or "allostatic" renormalization. These hypothesized synaptic changes, however, have never been identified. We introduced an approach that establishes the effects of repeated methamphetamine at single cortical synaptic terminals. Our results showed precisely how repeated methamphetamine when administered in a paradigm that results in behavioral changes abolishes normal striatal filtering by inducing a very long-term depression of corticostriatal release. This chronic presynaptic depression is the longest-lasting synaptic change due to drug exposure ever identified, lasting at least 4 months in adult mice following withdrawal from repeated drugs. A second phenomenon, which we call paradoxical presynaptic potentiation, showed that readministration of the addictive drug, only in mice with previous repeated drug administration, renormalizes their synapses: it even has an opposite effect in animals that are drug naive. We found that chronic presynaptic depression is maintained through altered cholinergic receptor responses while paradoxical presynaptic potentiation is induced through a dopamine D1 receptor mechanism seen only in mice with previous psychostimulant experience. Surprisingly, these long-term synaptic effects are not due to long-term changes in dopamine release but rather occur secondary to modulation of corticostriatal activity; corticostriatal modulation is initiated by dopamine but does not require further dopamine modulation once synaptic plasticity occurs.

We have identified the candidates for the long-term synaptic changes that have long been thought to underlie addiction. These newly identified synaptic mechanisms might underlie drug addiction and other forms of habit learning, and may show how such habits are unlearned at a synaptic basis. While none of these synaptic phenomena have been identified in any previous study, they are consistent with and serve to explain many observations in habit learning and drug dependence.

Huntington disease (HD) is a fatal autosomal dominant disorder produced by a polyglutamine expansion within the gene product, huntingtin. HD is characterized by progressive neurodegeneration within the cortex and striatum that produces dementia, progressive motor disturbances and early death. Although the progression of symptoms in HD has been classically attributed to cellular neurodegeneration, mutant huntingtin may promote abnormal synaptic transmission. Since dementia and signs of cortical degeneration precede that of basal ganglia injury, altered cell-cell interactions with cortical neuronal dysfunction producing an abnormal release of glutamate from cortical projections to the striatum may ultimately play a significant role in the development of the movement disturbance.

We have taken advantage of new optical approaches to identify developmental changes in corticostriatal release in the YAC128 HD transgenic mouse. The YAC128 mouse model was created with a yeast artificial chromosome (YAC) containing the entire human HD gene with 128 CAG repeats, including promoter regions. The YAC128 mouse might recapitulate HD pathophysiology since it expresses a full length form of huntingtin protein and induces selective neurodegeneration in the lateral striatum that is highly correlated to the motor deficit. To document changes occurring in corticostriatal release over time, we evaluated three groups of YAC128 transgenic mice, defined according to the development of motor symptoms: a young group (1 month) showing no evidence of abnormal behaviors, a middle aged group (6 months) corresponding to the onset of functional motor impairment, and an older group (12 months) displaying locomotor hypokinesis. We found that neurotransmitter release from corticostriatal terminals is enhanced in young YAC128 transgenic mice but diminishes in the aging animal. These age-dependent changes in corticostriatal activity are produced by selective alterations in subpopulations of individual cortical terminals that may target D2 receptor-expressing striatal neurons. These mechanisms might provide a synaptic basis contributing to excitotoxic injury in Huntington's disease.

Our experiments demonstrated that young YAC128 mice exhibit accelerated exocytosis from corticostriatal terminals at 1 month of life. This age-specific potentiation in release is manifest in all cortical terminals and is consistent with the hypothesis that enhanced synaptic glutamate promotes motor hyper-kinesis in young YAC128 mice. However, while this temporal increase in glutamate release might prime the cell for programmed cell death it may not be the primary cause of injury to striatal medium spiny neurons.

Our data also demonstrates that corticostriatal release progressively declines in aging YAC128 mice. A progressive deterioration in corticostriatal function in YAC128 mice was indicated by a decline in active corticostriatal terminals, accompanied by a decreased rate of exocytosis of the recycling synaptic vesicle pool or by a smaller pool of recycling synaptic vesicles. The ensuing disconnection between the cortex and striatum and the consequential loss of regulated glutamate release and trophic factors such as BDNF from cortical projections might cause a progressive loss in dendritic spines of medium spiny neuron that might otherwise be neuroprotective. This reduction in glutamate release from cortical projections would diminish striatal excitation and might account for the rapid decline in locomotor activity found in older HD mice. This late-onset reduction in glutamate release may also provoke excitotoxic changes in postsynaptic neurons that attempt to compensate by decreasing the number of dendritic spines. Such effects could work synergistically with enhanced postsynaptic NMDA receptor responses, mitochondrial dysfunction and postsynaptic membrane depolarization to further facilitate degeneration of medium spiny neurons.