linking back to

My lab:
After just barely making a grant deadline yesterday, I'm back just in time for the learning and memory session of this high-caliber symposium where people from literally all over the world have congregated to honor Randolf Menzel on his 70th birthday.

Ken Cheng started off this day with his talk on "Universal laws of behavior tested in the honeybee". He began by showing comparative data on generalization from pigeons and bees. Interestingly, he framed the problem of generalization as the question of delineating the class of stimuli which can be grouped according to their common consequences, i.e. which stimuli belong to the same 'consequential region'. In his experiments on Honeybees, Ken found that this consequential region forms an exponential around the stimulus for which generalization is tested. The second 'universal law' was the temporal weighting rule, meaning that the weight being given to a piece of evidence is inversely related to the square of the time that has passed since the last evidence was obtained, i.e. recent evidence weighs more than earlier evidence. He showed that bees seem to modulate their preference for recent evidence with circadian time. His conclusion of the two examples is that optimality analysis may be a good way to study universal laws, but, maybe not so surprising, that a considerable degree of abstraction is required.

Next up was Joachim Erber talking about operant conditioning of tactile cues. He told us that bees, when they scan an object with their antennae, they make contact up to 300 times per minute with contacts usually shorter than 10ms. Interestingly, antennal positions remain changed after 10 minutes of scanning an object such that the antennae are positioned where the object used to be. This happens independently of any sucrose reward, but it is comparatively slow. Using two objects, one dorsal and one ventral of the bee, one can reward antennal scanning of one of these plates, but not the other and get fast motor learning. Recording from the fast flagellum motor muscle, which controls some of the scanning movements of the antenna, they managed to operantly condition the contractions of this muscle by rewarding firing frequencies 1SD over baseline. The third experiment he described was to condition the animals to extend their proboscis after antennal scanning of an object was rewarded, ie., the sequence was to first present the object and then reward the animal with a drop of sucrose for scanning the object with its antennae.

Dorothea Eisenhard followed with her talk on "Formation of conflicting memories in honeybees". She works just across the hallway from me and works on extinction and reconsolidation using classical conditioning of the proboscis extension reflex. She showed that a single extinction trial does not lead to extinction, whereas both two and five extinction trials do. She also showed that bees show spontaneous recovery over night after 5 extinction trials. The extinction memory, just like the regular acquisition memory, consolidates dependent on protein synthesis.

Uli Müller
was up next, talking about "Molecular biology of learning and memory: from memory phases to signaling cascades". His talk was mainly about the work he did in his time here in the lab of Randolf Menzel, so not really any new data, but some important basic information for the students in the auditorium. For instance, Protein Kinase C (PKC) is only cleaved into the continually active form PKM after multiple, but not single conditioning trials. Of course, he also talked about cAMP and PKA, CaMKII, Calpain and all these now classic processes.

Last in this session was Brian Smith talking about "Distributed plasticity in the honeybee brain". He presented results on octopamine receptor knockdown on classical olfactory learning, adding to the evidence that octopamine signaling mediates the unconditioned stimulus, (US, sucrose) in this type of learning. He also studied conditioned stimulus (CS, odor) processing using optical imaging of neural activity in the antennal lobes. From these data he concluded that conditioning apparently stretches the perceptual space represented in the activity of the neurons in the antennal lobe. Further data suggest that octopamine generally excites these neurons. Brian also looked at the effect of leaving the US away when the odor CS is presented and found latent inhibition. The magnitude of this effect is proportional to the concentration of the odor and context-dependent. This effect can also be seen when mixtures of the pre-exposed odor with a novel odor both in imaging and in conditioning experiments. Interestingly, quantitative trait analysis of this effect yielded data pointing towards the receptor of the precursor to octopamine, tyramine. Corroborating this pointer are results showing that blocking tyramine receptors causes overgeneralization of inhibition from the pre-exposed to the novel odor, i.e., even the novel odor is learned more slowly if tyramine receptors are blocked.

The next session on brain anatomy and physiology was kicked off by Jürgen Rybak talking about the virtual 3D bee brain. He explained to us the basic principles behind transforming image data from microscopic scans of honeybee brains into a standard 3D atlas in the computer. One of the nice things about this standard is that one can image the structure of any individual neuron and then register it with the standard brain to quantitatively analyze the morphology of these neurons. Using this technique with multiple neurons allows the scientist to study the projection areas of multiple neurons and deduce potential areas of overlap/connectivity, even if these neurons were stained in different animals. In the second part of his talk, Jürgen talked about an autosegmentation procedure, which is based on knowledge about the individual variability in brain morphology. Statistics about the most common shapes instruct a dynamic model which is then used to autosegment newly acquired brain data to morph them into the standard atlas. Currently, the group is working on an ontology organizing all the single-neuron data in the standard brain and connect them to other databases such as genomic and proteomic databases.

Second speaker for this session was Alison Mercer from New Zealand talking about "Dopamine signalling in the bee". She started out by telling us that dopamine levels in the brains of young workers increased when the queen was removed from the hive. She presented data suggesting that Queen mandibular pheromone (QMP) is mediating this manipulation of the young worker dopamine levels. But not only dopamine levels are manipulated by QMP, but also the expression of one of the dopamine receptors, AmDOP1. Because dopamine is involved in aversive learning in insects, they tested classical conditioning of the sting extension reflex. It turned out that QMP blocks aversive learning in this paradigm, but only in the young animals that also show the sensitivity to QMP in their dopamine levels. Homovanillyl alcohol (HVA) seems to be the key component mediating the QMP effects. Her data suggest that HVA acts as an AmDOP1 receptor antagonist, but as a AmDOP3 agonist. Given the action of these receptors on cAMP levels, cAMP level analysis in the mushroom-bodies suggests QMP acting on cAMP levels via homeostatic mechanisms. Using behavioral tests, Alison showed that young workers are attracted to QMP, whereas older workers (foragers) avoid QMP. Interestingly, expression levels of the dopamine receptors in the brain did not have any relation to attraction/avoidance of QMP, but expression levels in the antenna: bees that showed strong attraction to QMP had higher levels of AmDOP3 and AmOA1 receptors, compared to bees that were not attracted to QMP. Further data indicated that QMP acts to enhance the attraction of young workers to their queen.

Final speaker this morning (and actually a good 45 minutes behind schedule so not 'morning' anymore) was Monique Gauthier talking about "Modification of olfactory learning and memory induced by siRNA targeting nicotinic acetylcholine subunits in the honeybee". The talk started with an overview of muscarinic and nicotinic binding sites in the honeybee brain and some basic information on the pharmacological properties of these cholinergic receptors. Conditioning experiments showed that nicotinic acetylcholine receptors are involved in olfactory learning. The receptors in the alpha lobe of the mushroom-bodies are involved in retrieval and the ones in the calyces are involved in the acquisition of classical olfactory conditioning (single trial learning). Monique went on to show data on receptor subunit distribution and the behavioral effects of RNAi-mediated knockdown of single receptor units in the entire brain. Knockdown of the alpha 7 subunit before retrieval had no direct effect on memory, whereas systemic knockdown during training significantly reduced memory levels. Knocking down alpha 8 systemically during acquisition had no effect, but knockdown after training did have an effect on retrieval. Restricting the knockdown of these subunits in the calyces and alpha lobes of the mushroom-bodies, respectively, showed that the different subunits differentially mediate the retrieval/acquisition effects in these two regions of the mushroom-bodies.
Posted on Saturday 12 June 2010 - 12:50:00 comment: 0

Render time: 0.1290 sec, 0.0038 of that for queries.