Biological Physics of Olfaction:
from Genes to Networks

Link to postdoc openings

The group was established in October 2002 and is expected to function for 4 to 5 years.  It is part of the Department of Biological Physics at the Max Planck Institute for Physics of Complex Systems. We seek to understand in detail a variety of aspects of olfaction - the sense of smell. Our approach is theoretical and computational, but involves close collaboration with relevant biology groups. We also explore the applicability of insights gained from olfaction to other parts of the brain, and aim to contrast olfactory mechanisms to those encountered in vision and hearing.

Within the last 10 years, remarkable progress has been achieved in the understanding of molecular and genetic aspects of olfaction. The cloning of olfactory receptors (the largest gene superfamily in vertebrates) has permitted elegant new experiments that give precise information on how the detection and the first stages of processing of odorant information are organized.

A good understanding of many functional aspects of the olfactory system has yet to emerge, however. The character of information processing in the olfactory bulb, and in particular the role played there by oscillations and by temporal coding, is currently the subject of intense scrutiny and debate. The mechanisms of storage of short- and long-term olfactory memories have only begun to be elucidated. The processes of development of neural connectivity in the olfactory system differ strongly from their counter-parts in the visual system, and their study can be expected to yield basic insights with wide applicability to other parts of the brain. The mechanism through which olfactory sensory neurons achieve singular (i.e., monogenic and monoallelic) expression of odorant receptors remains ununderstood. In general, olfaction is exceptionally well suited to the study of the fundamental question of how genetic information gives rise to the physical structure of a neural system.

Of particular interest to our group at present are aspects of olfaction that are directly related to the property of glomerular convergence. Each of the millions of sensory neurons in the nasal olfactory epithelium chooses, apparently in a stochastic way, to express one - and only one - odorant receptor from the repertoire of about 1000 odorant-receptor genes (in mouse). Such exclusive monogenic choice is unusual in the expression of large gene families. On the phenotype level, this results in a mosaic-like distribution of sensory neurons of 1000 distinct types throughout the area of the olfactory epithelium. The axons of these sensory neurons link the epithelium to the next stage of the olfactory system, the olfactory bulb. Even though the axons of neurons expressing a particular receptor start from a spread-out spatial configuration that is intermingled with similar configurations of hundreds of other types, at the point of termination in the olfactory bulb the axons are observed to be almost perfectly sorted by type and converge together in a highly compact structure - the glomerulus. Thus there are 1000 glomeruli (per hemisphere) in the bulb, each collecting sensory input resulting from odorants binding to only one type of olfactory receptor. The mechanisms through which this sorting and convergence of axons is achieved are at present unclear. It has been shown, however, that the glomerular convergence of axons is present even in animals in which the transduction of odorant-to-receptor binding to electrical signal is disrupted. Glomerular convergence is thus directly determined by genetic information. The resulting pattern of neural connectivity provides the substrate for combinatorial coding of odor quality and quantity, as well as for further information processing.

Glomerular convergence in the mouse olfactory system. Olfactory sensory neurons expressing a particular odorant receptor (transcribed from a tagged transgene) are labeled in dark blue. Their axons are seen to project from the somas scattered in the olfactory epithelium (left half of figure) into a single highly compact glomerulus (at far right) in the olfactory bulb. We model the mechanisms underlying this convergence. [Figure adapted from Extern linkVassalli et al.(2002).] Combinatorial coding of odor information in the antennal lobe of honeybees. (a) Schematic view of the antennal lobe with 38 distinct glomeruli labeled. (b) Representation of glomerular activity (as measured by in vivo calcium imaging) in response to presentation of the odorant 1-octanol. (c) Presentation of clove oil leads to a different, but partially overlapping glomerular activation pattern. We study how information encoded through such combinatorial codes is processed. [Figure from Extern linkGalizia et al. (1999); for movies see Extern linkatlas at FU-Berlin.]

Contact for this page: Martin Zapotocky.