The Harvey-Svoboda paper does indeed usefully advance our understanding of the vital topic of the synapse-specificity of LTP. Although the paper does not mention it, other studies have shown that LTP occurs in an all-or-none manner, so that the "threshold-crosstalk" described by these authors is really equivalent to the "LTP-crosstalk" previously described by others (notably, Engert-Bonhoeffer).

However, your summary is slightly wrong on one point: the Harvey-Svoboda evidence that depletion of calcium stores does not affect crosstalk does not rule out that calcium itself might be the "intracellular diffusible factor". This is still a very real possibility. The main argument Harvey-Svoboda advance against calcium is the evidence (in the supplementary material) that in the conditions of their experiments the spread of calcium from synapse to synapse is "only" 1 percent, a number that is not significantly different from zero. However, that number is clearly even less significantly different from 1 percent, a level that could (if repeated 30 times, as in their protocol) combine...  Read more

The Harvey-Svoboda paper does indeed usefully advance our understanding of the vital topic of the synapse-specificity of LTP. Although the paper does not mention it, other studies have shown that LTP occurs in an all-or-none manner, so that the "threshold-crosstalk" described by these authors is really equivalent to the "LTP-crosstalk" previously described by others (notably, Engert-Bonhoeffer).

However, your summary is slightly wrong on one point: the Harvey-Svoboda evidence that depletion of calcium stores does not affect crosstalk does not rule out that calcium itself might be the "intracellular diffusible factor". This is still a very real possibility. The main argument Harvey-Svoboda advance against calcium is the evidence (in the supplementary material) that in the conditions of their experiments the spread of calcium from synapse to synapse is "only" 1 percent, a number that is not significantly different from zero. However, that number is clearly even less significantly different from 1 percent, a level that could (if repeated 30 times, as in their protocol) combine with sub-threshold calcium signals to trigger synapse-inspecific LTP.

The main, and probably insuperable, difficulty, that synapses encounter in generating completely specific LTP is that they must remain well-coupled electrically to the parent dendrite. While this may seem a minor technicality, it may turn out to be the single most important problem the brain faces, and responsible for much of its baffling circuitry.

View all comments by Paul Adams