Abstracts

Rationale: Neurogenesis persists in the adult hippocampal dentate gyrus (DG) and is stimulated by seizures. In rodent temporal lobe epilepsy (TLE) models, dentate granule cells (DGCs) born after status epilepticus (SE) develop aberrant connections that may contribute to pathophysiology. To examine how SE affects adult-born DGC integration into pre-existing circuitry, we use a dual-virus approach to identify afferent inputs onto neonatal- or adult-born DGCs. A retrovirus (RV-Syn-GTR) expressing human Synapsin1 (Syn) promoter-driven green fluorescent protein (GFP), an avian retroviral receptor (TVA), and rabies glycoprotein (Rgp) is used to selectively infect dividing cells in the rat DG. An avian envelope protein (EnvA)-pseudotyped rabies virus, with Rgp replaced by mCherry (RbV-mCh), is later injected into the same DG to selectively infect RV-Syn-GTR transduced neurons through EnvA binding to its cognate receptor, TVA. RbV-mCh complements with Rgp provided by RV-Syn-GTR and retrogradely crosses synapses, labeling first-order presynaptic neurons with mCh. As traced presynaptic inputs lack Rgp, RbV-mCh spreads no further. We use this approach to examine presynaptic inputs onto neonatal- vs. adult-born DGCs in rats undergoing SE. Methods: We injected RV-Syn-GTR into postnatal day (P) 7 or P60 rat DG to label (and birthdate) dividing DGC progenitors. Rats underwent pilocarpine-induced SE at P56, received bilateral DG injections of RbV-mCh 10 wks later to retrogradely label presynaptic DGC inputs, and were euthanized after 7 d. Controls received saline instead of pilocarpine. We identified ‘starter' neurons as GFP+/mCh+ and their presynaptic inputs as mCh+/GFP-. The identity of mCh+ neurons was determined by triple-label immunofluorescence for neuronal subtype-specific markers. Results: P60-injected controls showed mCh+ inputs onto adult-born DGCs from neurons in the hilus, entorhinal cortex (EC), and supramammillary nucleus (SMN), well-known DGC afferents. We also found mCh+/GFP- inputs from the gyrus fasciolaris (GF). Triple-labeling revealed that hilar inputs included parvalbumin (PV)+ interneurons and calretinin (CR)+ putative mossy cells. In P60-injected SE animals, fewer hilar PV+ interneurons and CR+ cells co-expressed mCh while more GF neurons did. We also found more mCh+/GFP- inputs from DGCs indicating that after SE, adult-born DGCs receive recurrent connections from other DGCs. Surprisingly, increased mCh+/GFP- inputs arose from CA3 pyramidal cells after SE, indicating increased back-projections formed onto adult-born DGCs. Conclusions: These findings suggest that monosynaptic inputs onto adult-born DGCs are altered by SE and contribute to recurrent excitatory networks. Inputs from other DGCs and CA3 pyramidal cells increase while hilar inputs decrease. Ongoing studies are aimed at comparing P7-injected animals and determining whether some mCh+/GFP+ ‘starter' DGCs are instead retrogradely labeled cells synapsing onto other DGCs. Grant Support: Epilepsy Foundation Predoctoral Research and Training Fellowship and NIH NS058585.