Abstracts

Rationale: Decreased release probability and increased failure rate for monosynaptic inhibitory postsynaptic currents (eIPSCs) are present in presynaptic GABAergic terminals of the undercut (UC) model of posttraumatic epileptogenesis. We used selective blockers of P/Q and N-type Ca2+ channels to test the hypothesis that changes in these channels, induced by either lesion or excessive activity, alter physiological levels of GABAergic transmission.Methods: We recorded whole cell eIPSCs from layer V pyramidal (Pyr) cells in sensorimotor cortical slices from na ve and UC rats and locally applied 1 M ?-agatoxin IVa or ?-conotoxin GVIA to block P/Q- or N-current channels, respectively. Immunoreactivity (IR) of P/Q- and N-channels colocalized with the vesicular GABA transporter (VGAT) was quantified in the perisomatic region of layer V Pyr cells with confocal microscopy and Volocity software. Perisomatic cholecystokinin (CCK) and parvalbumin (PV)-IR were also measured from Z-stacks. Another group of P21 na ve rats was used to assess the effects of 2 hours of acute epileptiform activity, induced by epidural application of 100 m gabazine and 100 m 4-aminopyridine (4AP), on expression of P/Q- and N-channel-IR in inhibitory terminals 5 d later. Control rats were treated with epidural saline.Results: ?-conotoxin reduced mean peak eIPSC amplitude to 47.7 6.3% of baseline in control cells, but UC cells were insensitive to N-channel blockade by the toxin (104.5 5.1% of baseline, unpaired t-test; P < 0.01). ?-agatoxin completely suppressed eIPSCs in 5 of 6 injured cells, compared to a ~50% reduction of peak IPSC amplitude in control cells, suggesting that GABA release in injured terminals depends solely on P/Q-channels. These results were confirmed by analysis of P/Q- and N-channel IR. The volume of N-channel IR co-localized with VGAT around control Pyr cell somata was 20 2.2% (n = 13) of the total perisomatic region vrs 5.4 1.8% in UC Pyr cells (n = 8) (unpaired t-test: P < 0.0001). No change was present in P/Q-channel protein co-localized with VGAT. Similar decreases in N-channel-IR co-localized with VGAT, without changes in P/Q-channel-IR, were found when gabazine/4AP-treated sections were compared to control (saline = 18.9 3.1 %, n = 9; gabazine/4AP = 9 2.5 %, n = 12, P < 0.05) Normalized percentage of perisomatic PV-IR in UC cells (13.8 1.9%, n = 11) was significantly larger than control (9.1 0.7%, n = 12; unpaired t-test; P < 0.05), while the percentage of perisomatic CCK-IR in control Pyr cells (2.1 0.4%, n = 12) was similar to that in UCs (2.9 0.4%, n = 11; NS). Conclusions: Decreases in the expression of N-type channels are present in chronically injured UC cortex and following acute epileptiform activity in na ve cortex. These changes might result in decreased GABA release and contribute to hyperexcitability. The altered N-channel-IR and function is not due to loss of perisomatic PV- or CCK-containing terminals, but could involve abnormalities in expression of N-channels within existing interneurons.