The mechanism of local cross-talk between spine heads.

Karel Svoboda's paper in Nature studying LTP cross-talk using 2-photon uncaging of glutamate is finally out. He had hinted at this in his presidential address at SfN in San Diego 2 months ago (see entry for 15 November). Using the technique developed by Masanori Matsuzaki and Haruo Kasai in 2001 & 2004, LTP at small spine heads is reproduced by Hebbian pairing of post-synaptic depolarisation with repetitive quantal uncaging at visually identified spine heads. This photochemical tetanus "mimics" the widely used electrical methods, with the clear advantage that you know which spine is stimulated, and for how long/much. One difference between Kasai and Svoboda's approach is that Kasai used 3D photochemical mapping (Nature Neuroscience 2001) mapping of all near-by spines, whereas Svoboda tests only one nearby spine (see Figure above comparing the new and old data). No matter. The extra wrinkle Karel adds is that they go on to probe near, non-stimulated spines with a sub-LTP photochemical tetanus (LTP: 0 mV, 20 mW, 4 ms, 30 flashes at 0.5 Hz; sub-LTP used 1 ms flashes). It turned out that if you probe locally (within 6 microns) and quickly (less than 6 mins), the sub-LTP tetanus produces LTP with morphological enlargement. In a very elegant experiment, this surprising finding is shown to be "physiological" in that the same thing can be seen if electrical stimulation is used for the first LTP, followed by sub-LTP at a nearby spine (their Figure 3-boy that's a hard expt!).

This is interesting, but what is the mechanism of this spread? Obviously a rise in intracellular Ca is required for LTP. But it is not Ca that carries the message to the nearby spine. We "know" this already, as Svoboda's dogma is that the spine head is a previleged compartment for Ca, trapping >99% of Ca in its small volume. It is not protein synthesis, as the spread occurs in the presence of anisomycin. It is not ras, even though this does spread out of the spine into the dendrite (see figure from 15 Nov entry), as Karel said this during his talk at SfN. So what is the mechanism? Answer: we don't know (a Nature article with no mechanism). I would note in passing that all the experiments, except one, were performed at room temperature. Svoboda and Sabatini (who wrote the News & Views for this paper) have given Kasai a really hard time for doing this. I also wonder what Jeff Magee thinks about this, as all these studies are of juvenile neurons. Jeff always uses mature (> 2 months) brain slices for his work. Obviously there are many details left to discover in the next few years!

Local dynamic synaptic learning rules in pyramidal neuron dendrites. Nature (2007) 450:1195-1200.

Structural basis of long-term potentiation in single spine heads. Nature (2004) 429:761-765.

Dendritic spine morphology is critical for AMPA receptor expression in hippocampal CA1 pyramidal neurons. Nature Neuroscience (2001) 4:1086-1092.