Optimization of Efficient Neuron Models With Realistic Firing Dynamics. The Case of the Cerebellar Granule Cell

Roles for cerebellum and subsumption architecture in central pattern generation

Within vertebrates, central pattern generators drive rhythmical behaviours, such as locomotion and ventilation. Their pattern generation is also influenced by sensory input and various forms of neuromodulation. These capabilities arose early in vertebrate evolution, preceding the evolution of the cerebellum in jawed vertebrates. This later evolution of the cerebellum is suggestive of subsumption architecture that adds functionality to a pre-existing network. From a central-pattern-generator perspective, what additional functionality might the cerebellum provide? The suggestion is that the adaptive filter capabilities of the cerebellum may be able to use error learning to appropriately repurpose pattern output. Examples may include head and eye stabilization during locomotion, song learning, and context-dependent alternation between learnt motor-control sequences.

Propagation of Oscillations in the Indirect Pathway of the Basal Ganglia

Oscillatory signals propagate in the basal ganglia from prototypic neurons in the external globus pallidus (GPe) to their target neurons in the substantia nigra pars reticulata (SNr), internal pallidal segment, and subthalamic nucleus. Neurons in the GPe fire spontaneously, so oscillatory input signals can be encoded as changes in timing of action potentials within an ongoing spike train. When GPe neurons were driven by an oscillatory current in male and female mice, these spike-timing changes produced spike-oscillation coherence over a range of frequencies extending at least to 100 Hz. Using the known kinetics of the GPe→SNr synapse, we calculated the postsynaptic currents that would be generated in SNr neurons from the recorded GPe spike trains. The ongoing synaptic barrage from spontaneous firing, frequency-dependent short-term depression, and stochastic fluctuations at the synapse embed the input oscillation into a noisy sequence of synaptic currents in the SNr. The oscillatory component of the resulting synaptic current must compete with the noisy spontaneous synaptic barrage for control of postsynaptic SNr neurons, which have their own frequency-dependent sensitivities. Despite this, SNr neurons subjected to synaptic conductance changes generated from recorded GPe neuron firing patterns also became coherent with oscillations over a broad range of frequencies. The presynaptic, synaptic, and postsynaptic frequency sensitivities were all dependent on the firing rates of presynaptic and postsynaptic neurons. Firing rate changes, often assumed to be the propagating signal in these circuits, do not encode most oscillation frequencies, but instead determine which signal frequencies propagate effectively and which are suppressed.SIGNIFICANCE STATEMENT Oscillations are present in all the basal ganglia nuclei, include a range of frequencies, and change over the course of learning and behavior. Exaggerated oscillations are a hallmark of basal ganglia pathologies, and each has a specific frequency range. Because of its position as a hub in the basal ganglia circuitry, the globus pallidus is a candidate origin for oscillations propagating between nuclei. We imposed low-amplitude oscillations on individual globus pallidus neurons at specific frequencies and measured the coherence between the oscillation and firing as a function of frequency. We then used these responses to measure the effectiveness of oscillatory propagation to other basal ganglia nuclei. Propagation was effective for oscillation frequencies as high as 100 Hz.

On the Use of a Multimodal Optimizer for Fitting Neuron Models. Application to the Cerebellar Granule Cell

This article extends a recent methodological workflow for creating realistic and computationally efficient neuron models whilst capturing essential aspects of single-neuron dynamics. We overcome the intrinsic limitations of the extant optimization methods by proposing an alternative optimization component based on multimodal algorithms. This approach can natively explore a diverse population of neuron model configurations. In contrast to methods that focus on a single global optimum, the multimodal method allows directly obtaining a set of promising solutions for a single but complex multi-feature objective function. The final sparse population of candidate solutions has to be analyzed and evaluated according to the biological plausibility and their objective to the target features by the expert. In order to illustrate the value of this approach, we base our proposal on the optimization of cerebellar granule cell (GrC) models that replicate the essential properties of the biological cell. Our results show the emerging variability of plausible sets of values that this type of neuron can adopt underlying complex spiking characteristics. Also, the set of selected cerebellar GrC models captured spiking dynamics closer to the reference model than the single model obtained with off-the-shelf parameter optimization algorithms used in our previous article. The method hereby proposed represents a valuable strategy for adjusting a varied population of realistic and simplified neuron models. It can be applied to other kinds of neuron models and biological contexts.