Olfaction: Processing, Learning and Cognition

Séminaire FSER organisé par Pierre-Marie Lledo (Institut Pasteur), Catherine Dulac (Harvard University) et Ivan Rodriguez (Université de Genève) du 27 août au 1er septembre 2007

L’étude de l’olfaction a connu une démarche contraire à celle appliquée à l’étude des sens (analyse d’une fonction du niveau intégré vers l’étage cellulaire voire moléculaire).

Participants

Ron L. Davis, Mario De Bono, Barry Dickson, Catherine Dulac, Rainer Friedrich, Martin Giurfa, Leslie M. Kay, Pierre-Marie Lledo, Liqun Luo, Harold McGee, Aurélie Mouret, Venky Murthy, Ivan Rodriguez, Hitoshi Sakano, Noam Sobel, Nao Uchida, Donald Wilson, Rachel Wilson

Résumé

Jusqu’à présent, la démarche physiologique classique appliquée à l’étude des sens appréhendait l’analyse d’une fonction du niveau intégré vers l’étage cellulaire voire moléculaire. L’étude de l’olfaction a connu une démarche contraire. La révolution génétique et l’apport considérable des méthodes d’exploration non-invasives ont introduit une démarche physiologique inverse en permettant, à partir de l’identification d’un gène, et donc d’une protéine, d’inférer des règles de construction et de fonctionnement des circuits nerveux du système olfactif. En retour, l’exposition des organismes aux molécules odorantes se traduit par des modifications des niveaux macroscopiques (circuits, cellules…) aux niveaux moléculaires (gènes, enzymes…). De ce va-et-vient entre les différents niveaux d’organisation émerge une approche nouvelle de l’étude de l’olfaction.

Cette nouvelle physiologie a permis de révolutionner la façon dont nous pensions les mécanismes moléculaires et cellulaires de la réception du signal olfactif, du traitement central de l’information sensorielle ou encore la mise en place de réponses comportementales spécifiques. En conséquence, la physiologie de l’odorat subit actuellement une mutation spectaculaire. L’objectif de cette rencontre était de cerner l’ampleur de ce renouveau. Elle a permis de rassembler des spécialistes utilisant divers modèles expérimentaux (vers, insectes, poissons ou mammifères), afin d’évaluer le degré de congruence de ces nouveaux concepts.

Ce colloque a permis également de promouvoir des orientations pluridisciplinaires dans ce domaine de recherche. Il apparaît que les démarches expérimentales retenues par chacun des intervenants ont imposé un choix critique : celui du niveau d’organisation qui servira de référence. Or, cette rencontre nous a permis de constater qu’indépendamment de l’échelle considérée par ces différents laboratoires —gène, cellule, assemblées cellulaires, système fonctionnel, organisme— tout système biologique ad hoc, est un système complexe et ouvert, dont l’équilibre ne peut être maintenu que de façon dynamique. La recherche des règles sous-jacentes au fonctionnement de ces systèmes dynamiques représente un enjeu particulièrement stimulant pour les recherches futures.

Reste maintenant à définir une ligne de conduite autour des stratégies futures qui permettront d’aborder les particularités de l’olfaction. Une synthèse devra être opérée pour réunir autour de ce même sujet, les Sciences de la Vie, les Sciences de l’Ingénierie et les Sciences Humaines. Atteindre cet objectif nécessitera d’apporter un certain nombre de réponses aux questions que cette rencontre nous a permis de poser :

Quelle est la place d’une approche phénoménologique appliquée à l’étude de l’odorat? Doit-elle être nécessairement associée à une approche systémique permettant de dégager les interactions entre les composantes biologiques impliquées ?
L’approche quantitative, aux niveaux organisationnel et fonctionnel, nécessite le dialogue constant entre le macroscopique et le microscopique pour des systèmes non-linéaires. Dans ce contexte, qu’en est-il des propriétés émergentes des systèmes biologiques.
Viser à la compréhension du fonctionnement des systèmes sensoriels implique la mise en œuvre, dans un cadre conceptuel rigoureusement défini, de multiples outils expérimentaux spécifiquement destinés à une démarche résolument transdisciplinaire. Quelles sont donc les passerelles nécessaires à développer dans un proche avenir ?

Compte-rendu

In one session on olfactory system plasticity, Ron Davis reviewed his studies of olfactory memory trace formation after olfactory classical conditioning. Using functional optical imaging to detect calcium influx and synaptic transmission, he discussed studies that suggest that multiple memory traces are formed in the olfactory nervous system after learning. One is a short-term trace lasting only a few minutes. It forms in the projection neurons of the antennal lobe. A second is a middle-term trace that forms in dorsal paired medial neurons. This memory trace forms by 30 min after training and persists for at least 2 hr. The last two traces appear to represent long-term memory. They form in the alpha/beta and gamma neurons of the mushroom body neurons respectively, but not until many hours after training (~9) and they persist for as long as 48 hrs. The combined studies suggest that behavioral memory is guided from multiple memory traces that persist for different durations after olfactory learning. Martin Giurfa extended this discussion by analyzing levels of complexity in associative olfactory learning in honeybees. Using a combination of conditioning assays and calcium imaging of antennal lobe activity, he showed that odor maps in the antennal lobe directly reflect perceptual evaluation of odors and vice versa. Studies on olfactory learning suggest that experience-dependent changes in neural activity at the level of the antennal lobe depend on the nature of the association and its complexity. In simple learning forms, changes in odor maps were quantitative while in higher-order forms, they were qualitative and implied the recruitment of new glomeruli. In both cases, however, changes allowed decorrelating odorant representations that had to be distinguished. Comparative studies between aversive and appetitive olfactory learning showed that both learning forms are insulated from each other due to the involvement of different aminergic systems, with octopamine mediating appetitive reinforcement and dopamine mediating aversive reinforcement. It is concluded that olfactory learning has to be studied both at the levels of the olfactory and the reinforcement circuits.

Then, Pierre-Marie Lledo talked about another form of plasticity of the adult brain. To ensure that the mature nervous system’s control of behavior is flexible in the face of a varying environment, he showed that morphological and physiological changes are possible at many levels, including that of the entire cell. In the olfactory bulb of the adult brain, new neurons are generated throughout life and form an integral part of normal functional circuitry. P-M Lledo demonstrated that this process is not fixed, but highly modulated, revealing a plastic mechanism by which the brain’s performance can be optimized for a given environment. The functional benefits of this whole-cell plasticity, was discussed in the context of olfactory discrimination.

Nao Uchida talked about the rapid information processing in the olfactory system. First, he discussed about behavioral experiments about the speed and accuracy in the olfactory discrimination. Next, he presented electrophysiological recording of the piriform cortex neurons during the performance in an odor discrimination task. He found that the neurons in the piriform cortex responded rapidly with a transient burst of activity tightly locked to inhalation onsets.

Don Wilson presented the degree of cortical plasticity and odor perception, in rodents. One role of the piriform cortex is to read complex patterns of olfactory bulb output, generated by odor-specific patterns of receptor and mitral cell activity. Cortical output is then believed to help generate behavioral responses to (perception of) those odors. Data presented at the conference demonstrated that in contrast to odor-specific topography in responses of the olfactory bulb, neurons responsive to a specific odor were widely distributed within the piriform cortex, and even neighboring neuron pairs had only a 50% probability of both responding to the same individual odor. Through analysis of cortical responses to complex mixtures, evidence was presented that cortical neuron ensembles performed both pattern separation and pattern completion as the mixtures were morphed by subtracting or replacing individual components. This ensemble activity predicted behavioral judgments of mixture similarity and distinctiveness. For example, removal of a single component from a 10 component mix produced minimal change in ensemble activity (pattern completion) and animals could not discriminate the odors behaviorally. In contrast, introduction of a novel component into a mixture produced a dramatic shift in ensemble activity (pattern separation) and this change was easily discriminated behaviorally. Finally, the processes of pattern separation and completion appear to be anatomically distributed, with pattern separation dominant in the anterior piriform cortex and pattern completion more pronounced in the posterior piriform cortex.

Leslie Kay talked about the mammalian olfactory bulb that receives dense synaptic and neuromodulatory input from many central olfactory and limbic areas. Central input to the olfactory bulb granule cell layer desynchronizes fast oscillatory activity in waking mammals, and when it is temporarily or permanently abolished results in high amplitude narrow band gamma (40-100 Hz) oscillations. She showed that in the intact system, rats can modulate the amount of gamma power dependent on the demands of the discrimination in a 2 alternative choice task, with very high power in the gamma frequency band accompanying discrimination of odors that have highly overlapping input patterns. However, performance of the same discrimination in a Go/No-Go task erases the difference in difficulty between fine and coarse odor discrimination and produces strong beta band oscillations (15-30 Hz) in the olfactory bulb and hippocampus. The gamma oscillations represent local processing, as they are restricted to the olfactory bulb, while the beta oscillations represent distributed processing, being coherent in many areas of the olfactory and hippocampal systems. This suggests that the Go/No-Go task benefits from involvement of many cortical areas, while the 2 alternative choice task isolates processing to the olfactory bulb and makes learning more difficult.

Noam Sobel presented a statistical method to reduce dimensionality in both odor percepts and physico-chemical descriptors for a large set of molecules. Although it is agreed that physico-chemical features of molecules determine their perceived odor, the rules governing this relationship remain unknown. A significant obstacle to such understanding is the high dimensionality of features describing both percepts and molecules. Noam Sobel and colleagues applied a statistical method to reduce dimensionality in both odor percepts and physico-chemical descriptors for a large set of molecules. They found that the primary axis of perception was odor pleasantness, and critically, that the primary axis of physico-chemical properties reflected the primary axis of olfactory perception. This allowed us to predict the pleasantness of novel molecules by their physico-chemical properties alone. Olfactory perception is strongly shaped by experience and learning. However, our findings suggest that olfactory pleasantness is also partially innate, corresponding to a natural axis of maximal discriminability amongst biologically relevant molecules.

Finally Harold McGee concluded by a keynote lecture where he introduced us into the fascinating word of molecular gastronomy. Contemporary avant-garde cooking is exploring new ways of preparing food and new ways of influencing the diner’s experience. This approach to food is moving away from simple nourishment and satisfaction of expectations, and toward innovation, surprise, and expression: something closer to theatre or art, as the recent Kessel art exhibition featuring the work of a Catalan chef indicates. Avant-garde chefs are using industrial and laboratory ingredients, techniques, and machines to make their foods, and they are following developments in sensory science to understand and control the diner’s experience.