Chemical Senses Vol. 30 No. suppl 1 © Oxford University
Press 2005; all rights reserved
Neural Basis of Odor-source Searching Behavior in Insect Brain Systems Evaluated with a Mobile Robot
Department of Mechano-Informatics, Graduate School of Information Science and Technology, University of Tokyo, 7-3-1 Hongo, Bonkyo-Ku, Tokyo 113-8656, Japan
Correspondence to be sent to: Ryohei Kanzaki, email: kanzaki{at}i.u-tokyo.ac.jp
Key words: behavior, model, neuron, network, olfaction, pheromone, simulation
 |
Introduction |
|---|
 Top Introduction Strategies of the odor-source...Neural mechanisms of odor-source...Neural networkLong-lasting excitationReciprocal inhibitionEvaluation by an insect-size...AcknowledgementsReferences |
|---|
More than 3 million species of insects live around the world in a variety of environments, and display a diversity of sophisticated behaviors adapted to these environments. Our research is aimed at understanding how the brain systems of insects process constantly changing environmental information and generate adaptive behaviors. Specifically, we are investigating how odor information is processed and modified by other sensory modalities and experience (i.e. learning and memory). It is well known that males of many moth species can detect their species-specific pheromones at low concentrations and orient successfully toward the odor source (e.g. females) even though the odor-source is far away. This may depend not only on high sensitivity to olfactory information by insect olfactory receptors, but also on superior behavioral strategies or algorithms based on processing by neural networks in the insect brain (Arbas et al., 1993
;
Kanzaki, 1998). Insects have become an
excellent model for understanding adaptive control in biological systems which has
inspired research and development of control and communication in engineered systems. It is our long-term goal to understand the behavioral and neural basis of behavior of insects. To this end we have investigated the algorithms used to search for and locate a pheromone source and its underlying control mechanisms in the brain of the male silk moth, Bombyx mori. To evaluate the behavioral model we have implemented it in an insect-size mobile robot as a controller for the robot behavior.
|
Strategies of the odor-source searching |
|---|
|
TopIntroductionStrategies of the odor-source...Neural mechanisms of odor-source...Neural networkLong-lasting excitationReciprocal inhibitionEvaluation by an insect-size...AcknowledgementsReferences |
|---|
Male B. mori exhibit a characteristic zigzagging pattern as they walk upwind to pheromones released by females of the same species (Kanzaki et al., 1992
;
Kanzaki, 1998). Upwind walking towards
the pheromone source is controlled by a largely internally generated program of steering
which is triggered by the detection of an intermittent distribution of pheromone by the
antennae. Once initiated by a single puff of pheromone, this program consists of brief
bout of straight-line walking, zigzag turns and subsequent looping behavior (turns of
>360°). Upon stimulation male moths exhibit straight-line walking in the direction
of the antenna to which the stimulation was applied. Upon the loss of pheromone
stimulation, males exhibit zigzagging walking with the time interval between turns
increasing significantly after each turn, followed by looping behavior. This
pre-programmed sequence of movements is ‘reset’ and ‘restarted’
from the beginning in response to pulsed pheromonal stimulation.
It is known that odor is typically distributed by wind and therefore exists not as a
continuous concentration gradient but as a patchy intermittent plume (Murlis et al., 1992). As a result, the
male moths often show a variety of walking patterns depending on the distribution of the
odor filaments in the air. The pheromone-triggered pre-programmed zigzag turns and reset
mechanism together with the intermittent structure of the pheromone plume are the
underlying behavioral basis for the odor-searching behavior in B. mori
males.
|
Neural mechanisms of odor-source searching behavior |
|---|
|
TopIntroductionStrategies of the odor-source...Neural mechanisms of odor-source...Neural networkLong-lasting excitationReciprocal inhibitionEvaluation by an insect-size...AcknowledgementsReferences |
|---|
In B. mori males specific subsets of descending interneurons (DNs) which link the brain and the thoracic motor system show a characteristic state-dependent activity resembling an electronic ‘flip-flop circuit’ which has two distinct firing frequencies: high and low (Olberg, 1983
). Switching back and forth between the two states occurs upon pheromonal
stimulation. The flip-flop is the basic element of ‘memory’ in electronic
circuits. We have characterized two types of DNs (Group IIA,D and Group IIC) which
respond with a flip-flopping activity pattern and a brief excitation pattern,
respectively, in response to pulsed pheromonal stimulation (Kanzaki et al., 1994;
Mishima and Kanzaki, 1999). Both
groups of DNs have dendritic arborizations in particular areas of neuropile in the
protocerebrum of the brain: the lateral accessory lobe (LAL) and the ventral
protocerebrum (VPC). Our results indicate that the pheromone-mediated pre-programmed orientation behavior is controlled by these two types of DNs; i.e. straight-line walking is controlled by the brief excitation by the Group IIC DNs and the subsequent zigzagging turns and looping are controlled by the flip-flopping activity of Groups IIA and IID DNs.
|
Neural network |
|---|
|
TopIntroductionStrategies of the odor-source...Neural mechanisms of odor-source...Neural networkLong-lasting excitationReciprocal inhibitionEvaluation by an insect-size...AcknowledgementsReferences |
|---|
To test our hypothesis for the neural control of pheromone-mediated orientation in B. mori males we built a behavioral model for odor-source searching behavior. It seems obvious from the results of our studies of the DNs involved in this behavior that long-lasting excitation and reciprocal inhibition are important for generating the flip-flop responses, and these ideas were incorporated into our model.
|
Long-lasting excitation |
|---|
|
TopIntroductionStrategies of the odor-source...Neural mechanisms of odor-source...Neural networkLong-lasting excitationReciprocal inhibitionEvaluation by an insect-size...AcknowledgementsReferences |
|---|
We have characterized serotonin immunoreactive protocerebral bilateral neurons (PBNs) which link both LALs and VPCs (M. Iwano and R. Kanzaki, unpublished observations). Moreover, pressure injection of the serotonin into the LAL drove an enhancement of the pheromone response in some LAL neurons (E.S. Hill and R. Kanzaki, unpublished observations). Therefore, we also incorporated the demonstrated neuromodulatory actions of serotonin in generating long-lasting excitation into our model.
Dynamics of the concentration of neuromodulator is described below.
si(t) is damped slowly and the time constant
ph determines this long-lasting response.
|
|
where i represents the side, i.e. right or left LAL–VPC. When the
antennae receive a pheromone stimulus, an input signal goes into both LAL–VPC
regions. We assumed that a short time lag occurred according to which antenna receives
the stimulus. i represents the side that received a stimulus,
represents the
opposite side and the time lag was described as:
|
|
where tstim is the moment at which a pheromone stimulus reaches the antenna.
|
Reciprocal inhibition |
|---|
|
TopIntroductionStrategies of the odor-source...Neural mechanisms of odor-source...Neural networkLong-lasting excitationReciprocal inhibitionEvaluation by an insect-size...AcknowledgementsReferences |
|---|
We have characterized
-aminobutyric acid (GABA) immunoreactive protocerebral
bilateral neurons (PBNs) which link the left and right LALs (M. Iwano and R. Kanzaki,
unpublished observations). GABA is an inhibitory neurotransmitter in the insect brain. It
is consistent with what is known from other rhythmically active neural circuits
(Calabrese et al., 1989) that
alternating activity patterns could be generated by these GABA immunoreactive PBNs which
may make reciprocal connections (Mishima and
Kanzaki, 1999). It is also known from other reciprocally active systems that
fatigue of the cell plays an important role for alternating the activity state. The
membrane potential of the LAL–VPC region of our model is shown below.
|
|
where
|
|
Ui(t) is the average membrane potential of the
region, Xi(t) is the activity ratio of the region
determined by equation (4) and U is time constant of the
membrane potential. The output function sigm depends on the membrane potential
Ui(t) and the threshold level
hi(t) that has the fatigue effect described
below.
|
|
The time constant h(Ui) is a
variable dependent on the membrane potential Ui(t),
so that as the membrane potential Ui(t) becomes
higher, the cell fatigues more quickly. Since
h(Ui) has a low value under
conditions where the Ui(t) is less than the static
level of the threshold hi0, the threshold
hi(t) recovers to the initial static level
hi0 quickly when the region becomes inactive.
|
Evaluation by an insect-size robot |
|---|
|
TopIntroductionStrategies of the odor-source...Neural mechanisms of odor-source...Neural networkLong-lasting excitationReciprocal inhibitionEvaluation by an insect-size...AcknowledgementsReferences |
|---|
In order to evaluate the behavioral model under circumstances in which it controls a real body interacting with a real environment, we built an insect-size mobile robot and used our model as the control system for the robot. To make the pheromone field around the robot equivalent to the pheromone field around the moth, our robot was built to the approximate dimensions of the real insect (length: 31 mm, width: 18 mm, height: 30 mm). Antennae excised from live B. mori males were used as pheromone sensors, which were attached in front of the robot with an inter-antennal spacing similar to the moth. The electroantennogram, the depolarization of the antennal nerve which appears between the tip and the base of the antenna upon pheromone detection, was used as the odor input signal to the behavioral control system. Sensor values recorded in the pheromone field were transmitted to a host computer every 20 ms using a wire. The host computer calculated the next action according to the behavioral model by solving simultaneous differential equations. System parameters used in this model were determined in the simulation. The robot received action commands from the host computer and then drove the motors. Power was supplied to the robot by thin wires.
The experiments were performed in a wind tunnel so that the behavior of the robot
could be compared directly with the behavior of our insects in response to the same sort
of pheromone plume. Upon detection of pheromone the robot responded by executing
moth-like behavioral sequences, and by repeating these sequences of the behavior, the
robot could reach the pheromone source. Thus, our results indicate that even a
pre-programmed sequence of the behavior, without the influence of memory and learning,
can support a complex task such as oriention towards and location of the odor source by
simply repeating the set and reset of the program according to the distribution of odor
filaments in the environment. We now know that pheromone tracking in B. mori
males is modified by sensory modalities other than olfaction, and by experience
(Gatellier et al., 2004;
Seki et al., 2005). It is
the ultimate goal of our ongoing research to understand how these basic neural systems
are modified by other sensory modalities and experience for generating adaptive
behavior.
For many decades, invertebrate neuroethology has provided insight into how nervous systems organize and generate behavior, due in part to invertebrates being uniquely suited for multidisciplinary studies at different levels of organization using a variety of methodological approaches. For the coming decades, multidisciplinary contributions of biology (analysis) and engineering (synthesis) will be important for expanding our understanding of neuroethology.
|
Acknowledgements |
|---|
|
TopIntroductionStrategies of the odor-source...Neural mechanisms of odor-source...Neural networkLong-lasting excitationReciprocal inhibitionEvaluation by an insect-size...AcknowledgementsReferences |
|---|
This research was supported by a grant from the BRAIN.
|
References |
|---|
|
TopIntroductionStrategies of the odor-source...Neural mechanisms of odor-source...Neural networkLong-lasting excitationReciprocal inhibitionEvaluation by an insect-size...AcknowledgementsReferences |
|---|
Arbas, E.A., Willis, M.A. and Kanzaki, R. (1993) Organization of goal-oriented locomotion: pheromone-modulated flight behavior of moths. In Beer, R.D., Ritzmann, R.E. and McKenna, T. (eds)., Biological Neural Networks in Invertebrate Neuroethology and Robotics. Academic Press, New York, pp. 159–198.
Calabrese, R.L. and Arbas, E.A. (1989) Central and peripheral oscillators generating heartbeat in the leech, Hirndo medicinalis. In Jacklet, J.W. (ed.), Neuronal and Cellular Oscillators. Marcel Dekker, NewYork, pp. 237–267.
Gatellier, L., Nagao, T. and Kanzaki, R. (2004) Serotonin enhances the sensitivity of the male silkmoth to pheromone. J. Exp. Biol., 207, 2487–2496.
Kanzaki, R. (1998) Coordination of wing motion and walking suggests common control of zigzag motor program in a male silkworm moth. J. Comp. Physiol. A, 182, 267–276.[CrossRef]
Kanzaki, R., Sugi, N. and Shibuya, T. (1992) Self-generated zigzag turning of Bombyx mori males during pheromone-mediated upwind walking. Zool. Sci., 9, 515–527.
Kanzaki, R., Ikeda, A. and Shibuya, T. (1994) Morphological and physiological properties of pheromone-triggered flipflopping descending interneurons of the male silkworm moth, Bombyx mori. J. Comp. Physiol. A, 175, 1–14.
Mishima, T. and Kanzaki, R. (1999) Physiological and morphological characterization of olfactory descending interneurons of the male silkworm moth, Bombyx mori. J. Comp. Physiol. A, 184, 143–160.[CrossRef]
Murlis, J., Elkinton, J.S. and Cardé, R.T. (1992) Odor plumes and how insects use them. Annu. Rev. Entomol., 37, 505–532.[CrossRef][Web of Science]
Olberg, R.M. (1983) Phermone-triggered flip-flopping interneurons in the ventral nerve cord of the silkworm moth, Bombyx mori. J. Comp. Physiol., 152, 2973–307.
Seki, Y., Aonuma, H. and Kanzaki, R. (2005) Pheromone processing center in the protocerebrum of Bombyx mori revealed by NO-induced anti-cGMP immunocytochemistry. J. Comp. Neurol. (in press).
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||