All living organisms obtain information about their environment, process it and respond with behavioral output. Often, but not exclusively, information processing is performed by networks of neurons in the brain. However, studying complex brains is extremely challenging both conceptually and technically, and tends to impose a coarse grained approach. Fortunately, some model systems exhibit non-trivial behavioral patterns that are governed by neuronal circuits of limited size. A handful of such identifiable neurons are ideal for a detailed scrutiny on multiple levels: from circuitry to individual molecules within single cells. They can thus facilitate the understanding of basic principles of neural information processing and the engineering of neural circuits. In other words, a model system combining small neural circuits and a high experimental resolution is a promising candidate for the title "the phage of neuroscience" (or the hydrogen atom, if you will) and it could be reasonably argued that C. elegans is precisely that.
The nematode C. elegans possesses 302 neurons, the anatomy and connectivity of which are known with great precision. It was the first multi-cellular organism to have its genome fully sequenced and many of its genes have been cloned and characterized. Furthermore, a slew of existing genetic, genomic and biophysical techniques make it possible to perturb and assay this model organism with unparalleled experimental resolution. Specifically, it is possible to study concurrently the behavior of the entire animal, sub-circuits in its nervous system, individual neuron physiology and the molecules involved in intra-neuronal signaling pathways.
Even on the seemingly elementary level of worm behavior not all is clear. For instance, the precise fashion by which the worm navigates its environment is not fully understood. More detailed questions regarding the manner by which nematodes sense their environment are only partially answered and a comprehensive picture of how their nervous system goes about processing the information provided by the sensory apparatus simply does not exist.
In broad strokes, the goal of our research is to contribute to a collective effort to truly understand small networks of neurons. Asymptotically, I identify this sought-after understanding with the ability to answer all of the questions about the system that might come to mind, a small subset of which were presented here. These questions span multiple scales and various degrees of abstraction, from the molecular components of the system to the engineering and ingenuity of computational devices made of biochemical materials. This challenge must therefore be met with a parallel effort of both forward (e.g., revealing the molecular components and their interactions in detail) and reverse (e.g., systems analysis of higher function and statistical properties) approaches. Similar to the case of bacterial chemotaxis, such detailed scrutiny is likely to spawn theoretical insights. It is my hope that these, in turn, will transcend the boundaries of small neural circuits and contribute to the thinking about more complex biological neural networks. At the same time, I believe that explaining the design and properties of information processing in the framework of small neural circuits is a worthy challenge in its own right.
A good starting point for investigating the sensing of the environment and related navigational behavior is to focus on the responses of the worm to a specific stimulus. In our case the stimulus of choice is temperature, and the resulting behavior is known as "thermotaxis". We combine molecular biology techniques, quantitative analysis, and optics and engineering methods in order to identify and characterize the components of the thermotaxis neural circuit, i.e. the "hardware" underlying this sensory dependent navigation paradigms. We view this behavior as a model with which we can shed light on fundamental questions such as the relative contributions of the intra- and inter-cellular scales to the function of the neural circuit, the neuronal origin of stochastic behavioral patterns and the nature and limitations of information flow between neurons.
How does a coherent and predictable neuronal activity arise from the underlying molecular dynamics? It is known that chemical reactions in cells are stochastic in nature, yet sensory transduction often requires predictable responses. It is also known that homeostasis mechanisms can preserve the overall performance of neurons despite significant changes in the expression levels of channels, receptors and other gene products. However, for the purpose of long term adaptation, it is essential that significant changes in neuron performance do take place in a controlled and robust manner. Furthermore, the manner by which noisy and largely varying pathways might conspire to produce a consistent and adaptive behavior of an entire neural circuit, even if the responses of individual neurons are significantly altered, is even less understood. In part, the poor understanding of these problems arises from the difficulty of studying them in vivo. We aim to make use of optical methods for the purpose of analyzing fluctuations, correlations and the effects of external conditions on gene expression patterns. Questions we find interesting in this context cencern the coordinated regulation of gene expression that enables the homeostatic function of neurons in a noisy setting and limitations of information flow inside neurons.
We will use a similar combination of experimental approaches for the investigation of the recently reported sleep like behavior of worms. C. elegans molts its cuticle at the end of each of four developmental stages. Raizen et al. recently described the quiescent behavioral state that precedes each molt, as well as its similarities to sleep. Since the evolutionary origin and the very purpose of sleep are largely unknown our hope is that the combination of a powerful model system and our experimental tools will assist in unveiling at least part of the mystery.