Realistic neural circuit simulation of the moth antennal lobe that recognizes relative pheromonal concentration
-
1
The University of Tokyo, Graduate School of Information Science and Technology, Japan
-
2
The University of Tokyo, Research Center for Advanced Science and Technology, Japan
-
3
The University of Tokyo, Graduate School of Engineering, Japan
-
4
Johns Hopkins University, Department of Neurology, United States
Olfaction is the most important sensory information for a male silkmoth because of their female-searching behavior. A male moth finds a female moth via sensing pheromones which were emitted by a conspecific female. Odorants such as a pheromone are intermittently distributed in the air. When a male moth tries to find a female using olfactory information, the recognition of the pheromonal concentration in high spatio-temporal resolution is inevitable for efficient odor detection. Therefore, the concentration discrimination of intermittent pheromone plumes is the critical function for moth brain. However, the conventional research topics [1, 2] for the antennal lobe which had been studied by experiments or simulations were the discrimination of odorant species rather than concentrations. Hence, we are trying to build the realistic antennal lobe simulation of silkmoth based on experimental studies to understand dynamics and functions of this first olfactory center in the natural condition.
We experimentally found that relative pheromonal concentration discriminations are processed in the antennal lobe of silkmoth, especially in macroglomerular complex (MGC). The antennal lobe, the first olfactory processing center in the insect brain, has two types of neurons. One is a projection neuron (PN), which innervates the MGC and higher olfactory centers in a protocerebrum. The other is a local interneuron (LN), most of which are inhibitory and GABAergic. We measured dynamic changes of PN’s response when an antenna intermittently and consistently stimulated by pheromone [3]. Responses of PNs which were measured by calcium imaging and electrophysiology revealed following facts (Fig. 1A).
1) When PNs are stimulated by high concentration pheromone, action potentials are inhibited by LNs after PNs generate a small number of spikes.
2) Repetitive stimulations by pheromone make inhibition weak.
3) After the many repetitions of same concentration stimuli, PNs show a concentration-independent and constant response at any concentration.
That does not mean the antennal lobe simply desensitize the stimuli but the sensing range is tuned to stimuli because the PNs’ responses are enhanced or weakened when higher and lower concentration stimuli are applied just after repetitive stimuli.
We preliminary build an antennal lobe network model by the single compartment H-H type model to verify two following things. First, we expected to verify abstract network connection properties obtained from our experiments. Second, we expected to verify the cause of delay between PNs’ first spike after strong stimulation and strong GABAergic inhibition to PNs’ response when antenna was stimulated by high concentration pheromone. The cause of delay is predicted the time costs of the action potential propagation in the LNs. Therefore, to reproduce this delay as a physical phenomenon, it is suitable to make realistic 3D models of a neural circuit each cell is presented as a multi-compartment model. However, building a multi-compartment neural circuit model from scratch require very hard efforts such as parameter estimations, setting network connection properties. There are a lot of parameters for active and passive membrane properties and synaptic connection properties. Hence, before we make multi-compartment model of antennal lobe, we built single compartment model to verify network connection properties and reproduce dynamic changes of PN’s response. Fig.1C shows the schematic diagram of single compartment antennal lobe network consisting of PNs and spiking LNs. Each arrow indicates GABAergic inhibitory synapse. There are reciprocal GABAergic inhibition synapses among LNs, and the connection probability from LNs to LNs is 0.5. And from LNs to PNs, there are also have GABAergic inhibition and the connection probability is 0.5. As a result, responses of PNs in our single compartment network model could reproduced dynamic changes like our experiments (Fig. 1B). That means, the dynamic properties 1)-3) in response of PNs can be explained by only two types of synaptic connections, among LNs and from LNs to PNs.
However, as mentioned above, it is difficult to estimate realistic time delays due to spike propagations inside of neurons. Therefore, we are in progress of building multi-compartment model of antennal lobe. To make the multi-compartment model, we collected morphology data of confocal laser scanning microscope (CLSM) images of neuronss in the antennal lobe. The data were brought from our database system (BoND [4]). By these kinds of data, we extracted the suitable morphology for simulation by KNEWRiTE [5] and construct the standard brain model [5] of the antennal lobe. Using CLSM images which have a LN and a PN in one image, we could estimate synaptic connection area, and used the Peter’s rule to select synapse position. Presently, we are developing the multi-compartment model simulation of antennal lobe with K computer (RIKEN AICS, Japan). Afterward, we are planning to estimate membrane potential parameters to make realistic model.
In summary, we built the antennal lobe network model by the single compartment model to verify network connection properties and the cause of delay. As a result of the simulation, we were able to reproduce dynamic changes of responses of PNs. And, we were able to find out the LNs’ inhibition to PNs with delay can cause the antennal lobe to modify its sensitivity in response with the stimuli from our preliminary single compartment model simulation. Furthermore, we will understand more specifically about neuronal dynamics and characteristics of the antennal lobe’s microcircuit which can detect relative concentration difference in detailed 3D multi-compartment model simulation.
Acknowledgements
This research used computational resources of the K computer and other computers of the HPCI system provided by the AICS and ( the names of the HPCI System Providers) through the HPCI System Research Project (Project ID: hp140151 hp150074). And also, this research was supported by funding from Neuroinformatics Japan Center, RIKEN BSI to INCF Japan Node IVB-PF Committee.
References
1. Wehr M, Laurent G (1996) Odour encoding by temporal sequences of firing in oscillating neural assemblies. Nature 384: 162–166.
2. Assisi C, Stopfer M, Bazhenov M (2012) Excitatory local interneurons enhance tuning of sensory information. PLoS Comput Biol. 8:e1002563.
3. Fujiwara T, Kazawa T, Sakurai T, Fukushima R, Uchino K, et al. (2014) Odorant concentration differentiator for intermittent olfactory signals. J.Neurosci,34(50):16681-16593
4. Kazawa T, Ikeno H, Kanzaki R (2008) Development and application of a neuroinformatics environment for neuroscience and neuroethology. Neural Networks 21(8):1047-1055
5. Ikeno H, Kazawa T, Namiki S, Miyamoto D, Haupt SS, Nishikawa I, Kanzaki R (2012) Development of a scheme and tools to construct a standard moth brain for neural network simulations. Comput Intell Neurosci. doi:10.1155/2012/795291
Keywords:
Olfaction,
antennal lobe,
multi-compartment model,
Insects,
simulation
Conference:
Neuroinformatics 2015, Cairns, Australia, 20 Aug - 22 Aug, 2015.
Presentation Type:
Poster, to be considered for oral presentation
Topic:
Computational neuroscience
Citation:
Park
H,
Kazawa
T,
Miyamoto
D,
Goto
A,
Tabuchi
M and
Kanzaki
R
(2015). Realistic neural circuit simulation of the moth antennal lobe that recognizes relative pheromonal concentration.
Front. Neurosci.
Conference Abstract:
Neuroinformatics 2015.
doi: 10.3389/conf.fnins.2015.91.00044
Copyright:
The abstracts in this collection have not been subject to any Frontiers peer review or checks, and are not endorsed by Frontiers.
They are made available through the Frontiers publishing platform as a service to conference organizers and presenters.
The copyright in the individual abstracts is owned by the author of each abstract or his/her employer unless otherwise stated.
Each abstract, as well as the collection of abstracts, are published under a Creative Commons CC-BY 4.0 (attribution) licence (https://creativecommons.org/licenses/by/4.0/) and may thus be reproduced, translated, adapted and be the subject of derivative works provided the authors and Frontiers are attributed.
For Frontiers’ terms and conditions please see https://www.frontiersin.org/legal/terms-and-conditions.
Received:
31 May 2015;
Published Online:
05 Aug 2015.
*
Correspondence:
Mr. Heewon Park, The University of Tokyo, Graduate School of Information Science and Technology, Tokyo, 113-8656, Japan, park@brain.imi.i.u-tokyo.ac.jp