Dentate dependent-CA3 Network Pattern Separation can occur in the Absence of Neurogenesis

PIATTI, V. C.; HON, O.; LEUTGEB, J. K.; SCHLENNER, A. ; AN, Y.; LEUTGEB, S.; EWELL, L.A.; CAMERON H.A.

San Diego, CA

Congreso; Neuroscience 2018; 2018

https://www.sfn.org/Meetings/Neuroscience-2018/General-Information/Itinerary-Planner-and-Mobile-App

In the dentate gyrus (DG) neurons are born and integratedinto the DG network throughout the lifespan, which implies that a proportion ofdentate granule cells are immature. The increased excitability of immaturecells has been proposed to sparsify dentate network activity such that adultdentate neurogenesis is essential for supporting network discrimination ofnovel or overlapping network patterns. To determine how immature granule cellsin the DG influence network processes of pattern separation in the downstreamCA3, we recorded CA3 single unit activity in rats with and withoutneurogenesis. We used a task where dentate network activity was shown to benecessary for CA3 principal neurons to discriminate between novel sensory cues(McHugh et al., 2007). In this paradigm, rats were allowed to explore a novelenvironment over four 10-minute sessions. The walls of the environment wereswitched in each session from black (A) to white (B) in an A-B-B-A order. Theexperiment was repeated across multiple days. CA3 pyramidal cells were recordedin the transgenic GFAP-TK rat strain in which adult neurogenesis can becompletely ablated by daily administration of the drug Valganciclovir. Byperforming doublecortin and bromodeoxyuridine immunohistochemistry, weconfirmed that neurogenesis was ablated (99 ± 0.5%) in GFAP-TK⁺ rats (n = 6)6-8 weeks after continuous drug administration compared to GFAP-TK⁻ controlanimals (n = 8). In animals with confirmed ablation of dentate neurogenesis incomparison to controls we did not observe significant differences in the degreeof CA3 network discrimination for novel sensory cues, a dentate-dependentcomputation, or the pattern separation of the same cues once familiar. Althoughwe observed a decrease in firing rate with increasing familiarity of theenvironment in control rats, the firing rates stayed the same across days in animalswithout neurogenesis. This resulted in a significantly lower firing rate incontrol animals already on the second, more familiar day (p=0.0011,Kolmogorov-Smirnov test). The difference in CA3 network firing rate betweenGFAP-TK⁺ rats and GFAP-TK⁻ controls was not accompanied by a significantdifference in field size, number of fields per cell or spatial information. Ourresults suggest that adult born neurons modulate the excitability of thedownstream CA3 network, but are not critical for dentate-dependent CA3 networkpattern separation.