Limits...
The role of inhibition in a computational model of an auditory cortical neuron during the encoding of temporal information.

Bendor D - PLoS Comput. Biol. (2015)

Bottom Line: Using a computational neuronal model, we find that stimulus-locked responses are generated when sound-evoked excitation is combined with strong, delayed inhibition.In contrast to this, a non-synchronized rate representation is generated when the net excitation evoked by the sound is weak, which occurs when excitation is coincident and balanced with inhibition.Together these data suggest that feedforward inhibition provides a parsimonious explanation of the neural coding dichotomy observed in auditory cortex.

View Article: PubMed Central - PubMed

Affiliation: Institute of Behavioural Neuroscience, Department of Experimental Psychology, University College London, London, United Kingdom.

ABSTRACT
In auditory cortex, temporal information within a sound is represented by two complementary neural codes: a temporal representation based on stimulus-locked firing and a rate representation, where discharge rate co-varies with the timing between acoustic events but lacks a stimulus-synchronized response. Using a computational neuronal model, we find that stimulus-locked responses are generated when sound-evoked excitation is combined with strong, delayed inhibition. In contrast to this, a non-synchronized rate representation is generated when the net excitation evoked by the sound is weak, which occurs when excitation is coincident and balanced with inhibition. Using single-unit recordings from awake marmosets (Callithrix jacchus), we validate several model predictions, including differences in the temporal fidelity, discharge rates and temporal dynamics of stimulus-evoked responses between neurons with rate and temporal representations. Together these data suggest that feedforward inhibition provides a parsimonious explanation of the neural coding dichotomy observed in auditory cortex.

No MeSH data available.


Impact of spontaneous rate on computational model.a. Relationship between spontaneous rate of simulated neuron and the amplitude of the Gaussian noise added to the excitatory and inhibitory conductances. The arrow indicates the amplitude of noise used for the simulated neurons in analyses conducted in Figs. 2–8. The gray dashed lines indicate spontaneous rates of 0 spk/s (bottom) and 40 spk/s (top). b-c. Classification of neuron-type [non-sync (o), sync (x), mixed (+), atypical (square)] for two different spontaneous rates: a low spontaneous rate (b) of 0 spk/s (noise input of 3x10-8) and a high spontaneous rate (c) of ~40 spk/s (noise input of 6x10-8). Responses outside the allowable range (pure tone response between 1–50 spk/s) are indicated in cyan.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4400160&req=5

pcbi.1004197.g009: Impact of spontaneous rate on computational model.a. Relationship between spontaneous rate of simulated neuron and the amplitude of the Gaussian noise added to the excitatory and inhibitory conductances. The arrow indicates the amplitude of noise used for the simulated neurons in analyses conducted in Figs. 2–8. The gray dashed lines indicate spontaneous rates of 0 spk/s (bottom) and 40 spk/s (top). b-c. Classification of neuron-type [non-sync (o), sync (x), mixed (+), atypical (square)] for two different spontaneous rates: a low spontaneous rate (b) of 0 spk/s (noise input of 3x10-8) and a high spontaneous rate (c) of ~40 spk/s (noise input of 6x10-8). Responses outside the allowable range (pure tone response between 1–50 spk/s) are indicated in cyan.

Mentions: Our computational model operated with a fixed spontaneous rate (~4 spk/s), close to the median spontaneous rate encountered in our real neuronal population (3.8 spk/s). To generate a spontaneous rate, we added Gaussian noise to the excitatory and inhibitory conductances of the neuron. If the amplitude of this Gaussian noise was increased, the spontaneous rate increased monotonically (Fig. 9a). We next examined how sensitive our computational model was to changes in spontaneous rate. We examined spontaneous rates covering the entire range observed in our real neuronal population (0–40 spk/s) and found that the model parameters for generating synchronized and non-synchronized neurons were similar, albeit with a slight shift in the threshold I/E ratio for observing synchronized responses (Fig. 9b,c). We next examined how robust each neuronal representation (sync, non-sync, mixed) was across varying spontaneous rates (0–40 spk/s), and observed that a large fraction of synchronized (67%) and non-synchronized (52%) neurons did not change their neural coding regime across the entire range of spontaneous rates tested (Fig. 10, S7 Fig). In contrast to this, only 15% of mixed response neurons showed a similar invariance (Fig. 10).


The role of inhibition in a computational model of an auditory cortical neuron during the encoding of temporal information.

Bendor D - PLoS Comput. Biol. (2015)

Impact of spontaneous rate on computational model.a. Relationship between spontaneous rate of simulated neuron and the amplitude of the Gaussian noise added to the excitatory and inhibitory conductances. The arrow indicates the amplitude of noise used for the simulated neurons in analyses conducted in Figs. 2–8. The gray dashed lines indicate spontaneous rates of 0 spk/s (bottom) and 40 spk/s (top). b-c. Classification of neuron-type [non-sync (o), sync (x), mixed (+), atypical (square)] for two different spontaneous rates: a low spontaneous rate (b) of 0 spk/s (noise input of 3x10-8) and a high spontaneous rate (c) of ~40 spk/s (noise input of 6x10-8). Responses outside the allowable range (pure tone response between 1–50 spk/s) are indicated in cyan.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4400160&req=5

pcbi.1004197.g009: Impact of spontaneous rate on computational model.a. Relationship between spontaneous rate of simulated neuron and the amplitude of the Gaussian noise added to the excitatory and inhibitory conductances. The arrow indicates the amplitude of noise used for the simulated neurons in analyses conducted in Figs. 2–8. The gray dashed lines indicate spontaneous rates of 0 spk/s (bottom) and 40 spk/s (top). b-c. Classification of neuron-type [non-sync (o), sync (x), mixed (+), atypical (square)] for two different spontaneous rates: a low spontaneous rate (b) of 0 spk/s (noise input of 3x10-8) and a high spontaneous rate (c) of ~40 spk/s (noise input of 6x10-8). Responses outside the allowable range (pure tone response between 1–50 spk/s) are indicated in cyan.
Mentions: Our computational model operated with a fixed spontaneous rate (~4 spk/s), close to the median spontaneous rate encountered in our real neuronal population (3.8 spk/s). To generate a spontaneous rate, we added Gaussian noise to the excitatory and inhibitory conductances of the neuron. If the amplitude of this Gaussian noise was increased, the spontaneous rate increased monotonically (Fig. 9a). We next examined how sensitive our computational model was to changes in spontaneous rate. We examined spontaneous rates covering the entire range observed in our real neuronal population (0–40 spk/s) and found that the model parameters for generating synchronized and non-synchronized neurons were similar, albeit with a slight shift in the threshold I/E ratio for observing synchronized responses (Fig. 9b,c). We next examined how robust each neuronal representation (sync, non-sync, mixed) was across varying spontaneous rates (0–40 spk/s), and observed that a large fraction of synchronized (67%) and non-synchronized (52%) neurons did not change their neural coding regime across the entire range of spontaneous rates tested (Fig. 10, S7 Fig). In contrast to this, only 15% of mixed response neurons showed a similar invariance (Fig. 10).

Bottom Line: Using a computational neuronal model, we find that stimulus-locked responses are generated when sound-evoked excitation is combined with strong, delayed inhibition.In contrast to this, a non-synchronized rate representation is generated when the net excitation evoked by the sound is weak, which occurs when excitation is coincident and balanced with inhibition.Together these data suggest that feedforward inhibition provides a parsimonious explanation of the neural coding dichotomy observed in auditory cortex.

View Article: PubMed Central - PubMed

Affiliation: Institute of Behavioural Neuroscience, Department of Experimental Psychology, University College London, London, United Kingdom.

ABSTRACT
In auditory cortex, temporal information within a sound is represented by two complementary neural codes: a temporal representation based on stimulus-locked firing and a rate representation, where discharge rate co-varies with the timing between acoustic events but lacks a stimulus-synchronized response. Using a computational neuronal model, we find that stimulus-locked responses are generated when sound-evoked excitation is combined with strong, delayed inhibition. In contrast to this, a non-synchronized rate representation is generated when the net excitation evoked by the sound is weak, which occurs when excitation is coincident and balanced with inhibition. Using single-unit recordings from awake marmosets (Callithrix jacchus), we validate several model predictions, including differences in the temporal fidelity, discharge rates and temporal dynamics of stimulus-evoked responses between neurons with rate and temporal representations. Together these data suggest that feedforward inhibition provides a parsimonious explanation of the neural coding dichotomy observed in auditory cortex.

No MeSH data available.