in Table 1. Any experimental outcome could include enhanced, unchanged, or diminished LTP, and enhanced, unchanged or diminished memory, giving us a nine-compartment matrix. A quick glance at Table 1 reveals that of the nine possible combinations, eight are represented in the literature.3 What this means is that LTP and memory have been dissociated from each other in almost every conceivable fashion. LTP can be decreased and memory enhanced. Hippocampus-dependent memory deficits can occur with no discernable effect on LTP.

Does this mean that LTP is not involved in hippocampus-dependent memory formation? Certainly not. There are ways to rationalize each of the outcomes and maintain the hypothesis that LTP is involved in memory, as I will discuss later. Most of the explanations are variations of either (1) compensation by multiple, redundant memory systems or (2) measurement of the wrong type of LTP. However, what Table 1 does make clear, and the first point of the exercise, is that hippocampal LTP does not equal hippocampal memory. There is no one-to-one correspondence between LTP and memory.

One must bring to bear subtlety of thinking when considering the role of LTP in memory. There will be no direct quantitative or even qualitative relationship between LTP measured experimentally and memory measured experimentally—that is already abundantly clear from the available literature. This probably seems like a statement of the obvious to anyone who is an active researcher in the area. However,

3The ninth position, LTP unchanged and memory enhanced, can also be filled in by stepping slightly outside the bounds of the table. There are examples in the literature of two mouse strains that have equivalent LTP in vitro but that differ in their spatial learning capacity. As this is not a genetic manipulation in the same sense as a knockout or transgenic mouse line, I did not include this example in the chart. If anyone knows of other interesting mouse lines that dissociate LTP and memory, please let me know.

I am constantly amazed at the extent to which the general scientific audience is surprised when results from an LTP experiment do not "match up" with results from a behavior experiment. I think that those of us who work on LTP and behavior perhaps have not done a very good job of conveying this point in our publications.

In terms of testing the hypothesis of a role for LTP in memory (versus an equivalency), the most damning observations probably are those examples where LTP is completely lost and there is no effect on hippocampus-dependent memory formation. A specific example of this finding is the GluRA knockout mouse (see reference 7 and Figure 1), but there are several other similar observations in the literature. If LTP is lost and there is no effect on memory, that seems like a pretty hard spot to wiggle out of. Nevertheless, there are important caveats to interpreting these observations. For example, overtraining the animals could lead to memory formation by an alternate route that is not normally utilized. Or perhaps the particular learning paradigm is not measuring the relevant type of memory. (This is likely the case in the GluRA knockout animal; see reference 8.) Conversely, perhaps the physiology paradigm is not measuring the relevant form of LTP, or LTP in the relevant brain region. These are all issues that must be taken into consideration when interpreting these results.

The point here is not to flip-flop around like a politician and evade the question but rather to make two points. First, the discussion illustrates several specific issues that must be considered when interpreting experimental results comparing LTP with behavior, such as the multitude of experimental variables for both LTP and behavior. The second point is that, as I mentioned at the outset of the section, we have to bring some subtlety of thinking to bear on the question of the role of LTP in memory. The available data already make it clear that there will be no one-to-one correspondence of LTP to memory. We must begin to

TABLE 1 LTP and Memory in Genetic Mouse Models

0 0

Post a comment