Do dentate gyrus granule cells tag time-specific experiences?
1. ARL NSMA
2. Evelyn F. McKnight Brain Inst., Univ. Arizona, Tucson, AZ
3. Neurosci., NTNU, Trondheim, Norway
CA1 and CA3 networks can transmit information simultaneously about location and the sensory-behavioral components of unique episodes at a given location. Cells in these regions undergo two different forms of pattern separation, i.e., rate remapping and global remapping (Leutgeb et al., 2005). Contrary to previous speculations about the role of the dentate gyrus in pattern separation, however, the population of active granule cells appears to be more consistent over multiple environments than predicted by chance (Leutgeb et al., 2007). A possible explanation is that only a small fraction of granule cells is competent to become incorporated into new cell assemblies within a given time window (e.g., Aimone et al., 2006). Thus, if a rat experiences several different environments separated in time by several days or weeks, different granule cells might be allocated to represent different environments. More cells will be activated if the environments are revisited.
Granule cell activity was estimated using the activity dependent immediate early gene Arc. Arc+ cells were counted in the dorsal blade of dentate gyrus in 5 rats from each of 4 groups that were pretrained and/or tested in various combinations of 4 unique environments (A,B,C,D): Group 1: Train A, B, C, D-Test A, B, C, D; Group 2: Train A, B, C, D-Test D, D, D, D; Group 3: Train D, D, D, D-Test A, B, C, D; Group 4: Train D, D, D, D-Test D, D, D, D. Each group received four 5-day training blocks over 72 days, spaced at intervals of 30, 21, 14, and 7 days. On the test day, the animals experienced either the same environment in 4 sessions or all four environments (one session each).
Contrary to the prediction that granule cells represent the time at which an experience first occurs, there was no difference in the proportion of Arc+ cells between Group 1 (remote memories) and Group 3 (recent memories). Groups 1 and 3 expressed about twice the proportion of Arc+ cells as Groups 2 and 4, however, which was less than predicted by the random selection with replacement hypothesis. The simplest explanation for the current data is that the probability for activation among the granule cell population is not uniform, with some cells thus showing a higher than expected involvement in representing multiple experiences. Animals that were trained repeatedly in the same environment over several weeks (Groups 3 and 4) had slightly higher proportion (0.5%) of Arc positive granule cells on the test day than their corresponding test-day Groups (1 and 2 respectively). This may have resulted from increased survival of a small population of newborn granule cells as a result of repeated activation (e.g., Kempermann et al., 1997).
Grant/Other Support: AG023309; McKnight Brain Research Foundation; state of Arizona and ADHS; NS020331; Centre of Excellence grant from the Norwegion Research Council
Keyword: IEG Arc; hippocampus; spatial experience; behavior