A wide variety of behaviors are generated by rhythmic pattern-generating circuits. These include ongoing and stereotyped movements such as breathing, chewing, walking, running, flying, and swimming. Central pattern generators (CPG) are small discrete neural circuits and this together with the repetitive nature of the behavior they produce make them very good candidates for studying behavioral mechanisms at all levels of analysis.

    One of the first and most important tasks when one is about to study rhythmic behavior is the identification of the neurons and synapses that form the neural network that generates the rhythmic output. This has proved to be a very difficult and usually impossible task in vertebrate preparations. Historically it was proven to be more rewarding in invertebrates.

    The locust, and in particular locust flight behavior, were used to show for the first time the ability of the central nervous system to generate a fictive motor output. Using the locust preparation it was possible to show that rhythmic motor patterns can be generated in the absence of patterned sensory inputs. It was again locust flight that was used to demonstrate the importance of sensory control in producing an adaptive motor pattern. Indeed one important characteristic of many rhythmic behaviors is the need of constant modulatory control to produce a virtually endless repertoire of variation on a single motor pattern.

    David Rand (PhD student) and Daniel Knebel (undergraduate) are currently studying the rhythmic output of a central pattern generator networks situated in the stomatogastric nervous system of the locust. The frontal ganglion (~100 cells, see image) constitute the major source of innervation to the locust front gut. We are investigating the characteristic neural patterns that can be recorded from nerves leaving the ganglion and from the network's neurons in two very distinct behavioral contexts; feeding and molting (the periodical shedding of the insect's cuticle during metamorphosis). We are interested in the interactions between this neural circuit and other neural centers as well as endocrine factors specific to the different behavioral states.

The figure shows a section through the locust frontal ganglion. The lower panel shows an example of the rhythmic motor pattern that can be recorded from the indicated nerves in a totally isolated ganglion in-vitro. Ayali and Zilberstein (2002)

    Animal movements result from a dynamic interplay among neural commands, muscle and body mechanics and the environment. It is increasingly evident that a comprehensive understanding of animal locomotion must address the interactions of all these components.

    A collaborative interdisciplinary study (with Dr. Einat Fuchs, a post doc fellow at the laboratory of Prof.  Philip Holmes, Princeton University) takes advantage of the cockroach preparation and a combined theoretical and experimental approach to study the functional organization of central pattern generating circuits and inter-limb coordination during locomotion. Einat Fuchs and Tzachy David (a MSc student) monitor the rhythmic motor output of distinct cockroach leg efferents as they control for and manipulate inputs to the rhythm-generating networks. Time series recorded under different experimental conditions are compared and analyzed, utilizing advanced mathematical tools based on coupled oscillator models. The results are used to further improve dynamical models and computer simulations of six-legged locomotion, as well as (potentially) insect-inspired robots. See image on the right.

    In a different, yet similarly collaborative and interdisciplinary, study (with Dr. Gal Ribak and Prof. Daniel Weihs, Faculty of Aerospace Engineering, Autonomous Systems Program, Technion, Haifa, Israel) we are investigating locust flight kinematics and neural control, aiming at the possible use of insects as controllable miniature Unmanned Aerial Vehicles.

Bottom right: Studying locust optomotor responses. A computer screen on the right of the tethered insect is showing a pattern of moving vertical stripes that are seen through a mirror placed to the left of the insect. The locust is steering to the left by dragging the hindlegs. This behavior is typical steering response.     

    We also contribute to a recently initiated effort aiming at building a robot, composed of Assur tensegrity structures, that mimics caterpillar locomotion (led by Dr. Offer Shai, Faculty of Engineering, Tel-Aviv).

    Locust phase polymorphism is defined as the ability of some grasshopper species to show within the species, forms or morphs that differ in their morphology as well as their biology, dependent on the population density. Density dependent changes in locusts have been described in many different research areas; from morphology and Anatomy, biochemistry and physiology, to ecology and behavior. The polymorphic characteristics are quantitative, there are innumerable intermediates between the extreme phases and the change is reversible.

    The behavior of individual locusts in the presence of others is a major phase characteristic. The behavior change is the first noticeable change when previously isolated locusts are crowded. This change facilitates the appearance of the various morphological and physiological phase changes that follow it. Yet, the neurophysiological basis of the behavioral phase characteristics has received very little attention.

    We have recently demonstrated, for the first time, neural correlates of locust behavioral density-dependent phase polymorphism. We have studied phase related differences in identified flight related interneurons as well as in DUM neurons (insect octopaminergic neurons) activity. Dr. Moshe Gershon and Yevgenia Rozenblum (MSc student) currently further analyze behavioral phase characteristics and the neurobiological basis of locust phase polymorphism. Ronit Kornfein (MSc student) is working to establish the molecular basis of locust phase polymorphism and its behavioral application. This line of research will fill a long lasting gap in the understanding of locust phases and will provide insights into environmental effects on neural plasticity in general.

The photograph is of a crowded (left, orange-black) and an isolated reared (right, green) Vth instar locusts. This photo (by Amir Ayali) was used as the cover for Fuchs et al. 2003 (see publication list).

Desert locust (Schistocerca gregaria) in the gregarious phase. The figure shows a swarm of desert locust in Morocco during the 2004 outbreak. Locust density dependent polymorphism is an extreme example of the effects of environmental factors on the animal's behavior. In high population density, locusts actively aggregate, forming large hopper bands or adult swarms. In marked contrast, isolated animals move away from fellow locusts and from crowded groups. Photo is curtsey of Philip Dalton, John Downer Productions.

    In the developing nervous system of an animal the growth pattern of single neurons designated to constitute future neural circuits is a dominant factor influencing the nature of the developing networks. The branching pattern of the neurons defines the basic hardware framework of the nervous system. It is thus instrumental in the future output of neural circuits for behavior. In collaboration with Prof. Eshel Ben-Jacob from exact sciences (TAU School of Physics) we have cultured neurons from the locust frontal ganglion and investigated mechanisms of neuronal network self organization. We further investigated correlation and interactions between the neuron's and networks' growth pattern and electrical activity by culturing the neurons on multi electrode arrays (MEA; silicon chips in which an array of electrodes is embedded), and recording the neurons electrical activity as networks evolve.

    Sarit Anava , Ya'ara Saad (PhD. students) and Mai Anabusi (MSc student) are further investigating various aspects of the development of cultured neuronal network. In collaboration with Dr. Yael Hanien of the TAU school of electrical engineering, Sarit Anava has looked at various aspects related to the role of mechanical tension in neuron and network development. As part of this collaboration neurons are cultured on quartz substrates decorated by islands of carbo-nano-tubes.

The figure shows an example of a locust frontal ganglion neuron in primary culture after 2 days. Click image to open a 90 h timelapse video of network development.

Left: A photograph of a locust and a MEA chip over a tilling map describing the electrical activity of a cultured neuronal network in the time frequency domain. The photo was used as the cover for Ayali et al. 2004 (see publications list).

Right: A processed SEM image of a locust neuron over carbo-nano-tube islands (Hanien, Ayali et al. 2008)

Click image below to follow link