These pages ouline the results of a four-year collaboration between Sidarta Ribeiro and Claudio Mello, of Rockefeller's Neuroethology Lab, and Guillermo Cecchi and myself. Newer members of this interest group are Tim Gardner and Pabel Delgado.
When a canary song is played back to a canary, there is a blush of gene expression in some nuclei of the canary's brain. One of the interesting genes that gets expressed is called ZENK. ZENK is to some level a novelty detector: expression is only induced by novel stimuli. This is important because, as much as one might try to control the environment of an alert, behaving animal, there always is (and has to be) an environment; and the brain is never, ever quiet. But if the animal is fully accustomed to its environment, then ZENK allows us to get a huge signal: a factor of 12 induction between a conspecific song and silence.
One of the interesting places where ZENK is induced is in a nucleus called the caudiomedial neostriatum, or NCM. Observe at left a micrograph of a 20 micron thick slice of NCM, which has been immunoreacted with an antibody for ZENK. The field of view is about 2mm wide. (Click on the photo to view an enlarged version). This photo was reconstructed from a much higher magnification scan at 3 pixels per micron.
We generate high resolution maps, by scanning with a computer controlled stage. Then we go ahead and perform image recognition to discriminate labelled nuclei from every other piece of histological debris lying around. After we do so, we classify the cells into three different groups:
We do this because it is important that we distinguish between a small number of cells expressing a lot of zenk, and a lot of cells expressing a little ZENK. Finally we construct "average density" maps for each color channel. On the left you can see a map constructed in this fashion for the "control" experiment, namely an animal that sat for hour and a half listening to nothing; on the right is the result of playing back a conspecific song. These maps have been modified from their original version: they have been reduced to fit your browser. Please click on the images to see a "normal" resolution map. It should be clear that there is a distinct and most robust activation of gene expression at every possible level. It should also be clear that there's no likely way we can analyse this pattern and relate it to the complexities of the original song.
See our Java animated version of the figure panels!!!