Saturday, February 17, 2007

New Leutgeb/Moser - place cell paper   posted by amnestic @ 2/17/2007 07:11:00 AM

The Leutgebs and the Mosers have brought us another interesting datapoint regarding how the hippocampus segregates or lumps the representation of spatial environmens. They recorded from the CA3 and dentate gyrus subregions of the hippocampus while they moved rats between a series of 'morphed' environments, moving gradually from a circular to a square arena. The idea is that CA3 tends to lump representations while DG tends to split them apart. No time to chat right now, but here's an excerpt from the commentary by Andre Fenton:
The authors confirm that CA3 place cells respond to small deviations in the spatial environment by lumping. In other words, the same neuronal discharge patterns were observed in CA3 regardless of whether the rat was in a morphed or unmorphed circular or square box. Larger deviations from either environment caused rate remapping in CA3. The dentate gyrus was quite different. Single dentate granule cells had more firing fields than did individual CA3 cells. Granule cells responded to small morph deviations in the rat's environment by changing both firing rates and firing fields unpredictably. Thus, the dentate gyrus proves a consummate information splitter and the CA3 more of a lumper (see the figure). Small changes in spatial input information caused large changes in dentate gyrus output to CA3 but virtually no changes in CA3 output to CA1.

I had a dream last night that I was at a seminar where Stefan Leutgeb was explaining these findings and Edvard Moser was in the audience and began correcting Leutgeb and they got into an argument but started speaking in French. Miraculously in my dream I could understand French and was able to impress the student next to me by explaining the argument. I don't remember what they were arguing about though. Too-da-loo.

Labels: ,