Of mice and memory.

The recent paper by Tang et al. (1999) describing the genetic enhancement of learning and memory in mice sparked a media frenzy that even a reclusive scientist could not avoid. For a brief period we were saturated by stories from newspapers, television, and radio reporting on these “smart” mice and the “IQ gene” and, inevitably, the possibility of enhancing human intelligence. For those of us in the field of neuroscience, the hypothetical applications of this paper have encountered more skepticism, but the results of the project make a tangible contribution to our current understanding of learning and memory. The inspiration for the study of Tang et al. comes from the predominant theory on the cellular basis for learning, called Hebb’s rule, which states that synaptic transmission between two neurons will be strengthened if the two cells are simultaneously active and one cell repeatedly excites or causes the other cell to fire (Hebb 1949). This form of synaptic plasticity called long-term potentiation (LTP) clearly exists, as well as the reverse of LTP called long-term depression (LTD), a use-dependent decrease in synaptic strength. However, the relationship of these processes to memory is still in dispute. One implication of Hebb’s rule suggests that improved coincidence detection between neurons will result in improved learning. Tang et al. set out to test this hypothesis. They did this by altering the characteristics of the protein responsible for coincidence detection, the N-methyl-D-aspartate (NMDA) receptor. This receptor will only activate when it binds glutamate released by activity in the presynaptic cell and when it senses postsynaptic activity in the form of membrane depolarization. When the NMDA receptor detects these two events within a short time window, it allows calcium into the cell and this calcium is critical for inducing LTP. The NMDA receptor is composed of an obligatory NR1 subunit and a selection of NR2 subunits (A–D). The NR2 subunit confers different channel properties on the receptor complex. One intriguing subunit difference is the longer excitatory postsynaptic potential (EPSP) observed in NR2B-containing receptors in comparison to NR2A-containing receptors. This longer EPSP suggests that NR2B-containing receptors have a longer time window for detecting preand postsynaptic activity resulting in increased coincidence detection and enhanced LTP when activity is less synchronized. The developmental regulation of NR2B supports the relationship between this subunit and increased LTP because the expression of NR2B is highest when the animal is young and decreases in adulthood, correlating well with the higher amount of LTP