Learning and acquisition of new behaviors rely on the physiological context in which new information is processed. Neuromodulators play a permissive role by regulating the behavioral state and by facilitating the computations in neurons organized in functional ensembles across large, integrated brain areas.
One such area is the striatum, located in the basal ganglia and involved in reward-oriented learning and action selection (i.e., the promotion of preferred actions over undesired or competing actions). The activity of striatal neurons is dependent on glutamatergic afferent transmission (mainly from cortex and thalamus) timed with the dopaminergic input that promotes synaptic integration (arising in the ventral tegmental area and the substantia nigra pars compacta). Importantly, the activity of thalamic and dopaminergic neurons are modulated by cholinergic terminals originating in the midbrain (pedunculopontine nucleus, PPN and laterodorsal tegmental nucleus, LDT), suggesting that cholinergic neurons are capable of modulating the coincident thalamic and dopaminergic inputs at the level of striatal microcircuits.
We investigate the structural and functional organization in these circuits, the neuronal dynamics underlying the encoding of behaviorally-relevant (sensory) information (i.e. learning) into patterns of activity that will determine an outcome (i.e. action), and how different levels of arousal determine the optimization of neuronal computations and consequently the behavioral outcome.
Brainstem cholinergic projections to the striatum (see Dautan et al., 2014)
Identification of the functional connectome
The striatum, including its dorsal and ventral (nucleus accumbens) regions, receives a variety of synaptic inputs originating in areas associated primarily with motor and limbic functions. Such functional dissociation can be traced back to the cholinergic midbrain, where the PPN is associated with movement and attention, and the LDT is associated with limbic responses. Given the molecular heterogeneity of these pathways, novel tracing tools are necessary to make a fine dissection of their connectivity. We use newly developed viral vectors with anterograde, retrograde or transsynaptic capabilities to target specific neurons based on their phenotype or their connectivity. These pathways are revealed by histofluorescence and immunohistochemistry, and provide information on the specific connectivity of neuromodulatory neurons; furthermore, these data set the structural basis for our functional studies.
Electrophysiological mapping of neuromodulatory effects in subcortical circuits
Neuromodulatory neurons project widely to many brain areas through long and sometimes diffuse axonal collaterals. We use high-density electrophysiological recordings to record over large dorsal/ventral striatal areas or from different neuronal circuits functionally related (e.g. thalamus). We use optogenetic manipulations of specific subsets of neurons based on their phenotype or their connectivity. These experiments are sometimes combined with juxtacellular recordings to reveal the identity of certain neurons. We analyze the discharge properties and temporal correlations between neurons and correlate it with the population activity (local field potentials), and evaluate linear interactions in the oscillatory activity between pairs of synaptically connected structures. We can thus characterize the spatial and temporal organization of functional neuronal ensembles that are susceptible of neuromodulation.
Optogenetic stimulation of cholinergic axons from the PPN modulates the activity of dopamine neurons in the VTA (see Dautan et al., 2016)
Behavioral electrophysiology
The effective transfer of information depends on neuronal computations occurring during narrow windows of coincident activity between structures that are synaptically connected. This coincident activity is strongly dependent on the behavioral context, where neuromodulators play a determinant role. A full understanding of how neuromodulatory neurons operate in different physiological contexts requires high-resolution electrophysiological recordings during specific behavioral paradigms.
