Citation: Yiu et al. (2014). Neurons are recruited to a memory trace based on relative neuronal excitability immediately before training. Neuron 83(3): 722-735
This is one of those nights where this challenge got hard. Given the long NYC commute to and from Queens College, it’s always easy to read a paper, but typing up a review was pretty difficult as it was one of those hardcore grant writing days.
But, with a little over an hour to go in the day, I give you Yiu et al. (2014)
So, the last couple of posts have been examining memory traces and how the coactive cells involved in learning change their connectivity to retain information for later recall. But how do these cells become incorporated into the memory trace anyway? How do they, and not their neighbors, win the sweepstakes of cognition and get the chance to participate in an engram?
According to Yiu et al., it is their excitability. That is, those which are particularly excited at the time when sensory information hits will retain the memory.
What led the authors here were the observations that cells with increased CREB function seem to be preferentially active in memory processing and that CREB likely increases neuronal excitability. So, they manipulated lateral amygdala (LA) neurons in several ways. A) They mutated KCNQ2, a potassium channel subunit responsible for controlling neuronal excitability (in fact, it was shown that many members of a Czech family with a rare form of epilepsy had mutations in this subunit), in a small subpopulation of neurons. B) They injected CREB into a subpopulation. C) Expressed Kir2.1 in a subpopulation to decrease excitability. Then the mice were fear conditioned. They found that cells expressing CREB or the KCNQ2 mutant were much more likely than other cells to be c-fos positive upon memory retrieval, while cells with Kir2.1 were less likely. Behaviorally, animals who received excitation in their LA demonstrated enhanced fear memory.
Interestingly, the authors note that no matter the manipulation, the proportion of LA neurons allocated to the memory trace was constant across manipulations. Thus, participation in a memory trace is competitive, with the deciding factor of participation apparently being excitation at the time of trace formation. What is absolutely amazing about this paper is that it is the only paper I have ever seen in which the authors called a particular group of cells “losers.”
“…these winning neurons may also actively inhibit ‘‘loser’’ neurons.”
So, to summarize, neurons must be more electrically excited than their neighbors to be involved in a fear memory trace, which may result from their quicker activating of the local inhibitory circuits to shut down other cells.
What I’m not clear on is what exactly CREB is doing. I couldn’t tell if yes, enhancing CREB expression will excite the cells and allows them to be preferentially involved in a trace, or if perhaps excitation by other means can do so and enhanced CREB is doing something else. For instance, excitation preps the cell for trace incorporation in a number of ways including inducing CREB, which in and of itself isn’t the cause of preferential activation, but allows the cell to more quickly undergo LTP.
Of course, that is all exhausted speculation. Will certainly revisit this paper at a more alert time.