Reversible inhibition of GABAA receptors by alpha7-containing nicotinic receptors on the vertebrate postsynaptic neurons

Abstract

Nicotinic acetylcholine receptors (nAChRs) are expressed throughout the central nervous system and influence a variety of higher order functions including learning and memory. While the effects of presynaptic nAChRs on transmitter release have been well documented, little is known about possible postsynaptic actions. A major species of neuronal nAChRs contains the alpha7 gene product and has a high relative permeability to calcium. Both on rodent hippocampal interneurons and on chick ciliary ganglion neurons these alpha7-nAChRs are often closely juxtaposed to GABAA receptors. We show here that in both cases activation of alpha7-nAChRs on the postsynaptic neuron acutely down-regulates GABA-induced currents. Nicotine application to dissociated ciliary ganglion neurons diminished subsequent GABAA receptor responses to GABA. The effect was blocked by alpha7-nAChR antagonists, by chelation of intracellular Ca2+ with BAPTA, and by inhibition of both Ca2+-calmodulin-dependent protein kinase II and mitogen-activated protein kinase. A similar outcome was obtained in the hippocampus where electrical stimulation to activate cholinergic fibres reduced the amplitude of subsequent GABAA receptor-mediated inhibitory postsynaptic currents. The reduction showed the same calcium and kinase dependence seen in ciliary ganglion neurons and was absent in hippocampal slices from alpha7-nAChR knockout mice. Moreover, alpha7-nAChR blockade in hippocampal slices reduced rundown of GABAA receptor-mediated whole-cell responses, indicating ongoing endogenous modulation. The results demonstrate regulation of GABAA receptors by alpha7-nAChRs on the postsynaptic neuron and identify a new mechanism by which nicotinic cholinergic signalling influences nervous system function.

Figure 1

Nicotine application depresses GABA-induced responses in freshly dissociated chick CG neurons Nicotine (20 μm, 3 s) and GABA (100 μm, 1 s) were individually applied. A, stable GABA responses were elicited at 1 min intervals. The GABA-induced response was completely blocked by the specific GABA_A receptor antagonist gabazine (10 μm) and fully recovered after washout. Pairs of hash marks here and below indicate 1 min intervals. B, GABA induced responses were attenuated after a 3 s nicotine application (left); allowing an 8 s lag after nicotine application allowed full recovery of the GABA response (right). C, compiled results showing the time course of recovery for GABA-induced current amplitudes after nicotine application (mean ±s.e.m.; n = 8 neurons). The response was significantly reduced by nicotine when the interval between nicotine and GABA application was ≤ 2 s *P < 0.05, one-way ANOVA, Dunnett's post hoc test. D, nicotine-induced depression of GABA responses (control) was abolished by the α7-nAChR antagonist MLA (50 nm). The residual nicotine-induced current in the presence of MLA (lower trace) results from heteromeric nAChRs containing α3 and other subunits (Vernallis et al. 1993).

Figure 2

Intracellular calcium, CaMKII and MAPK are required for α7-nAChR-mediated depression of GABA_A receptor responses in CG neurons A, paired traces as in Fig. 1B showing GABA-induced responses 1 min before and 2 s after applying nicotine for 3 s using control solution (left), or solutions containing either 11 mm BAPTA (middle) or the CaMKII antagonist KN-93 plus the MAPK antagonist U0126 (right), both at 10 μm. B, compiled results obtained for GABA responses in the presence of nicotine after infusing the indicated compounds through the recording pipette; values are normalized to the response obtained in the absence of nicotine, and the error bars indicate s.e.m. (7–10 cells tested per condition). CaMKII blocker KN-93 (10 μm), MAPK blocker U0126 (10 μm), protein kinase C blocker chelerythrine (10 μm), PI3K blocker wortmannin (100 nm), negative control KN-92 (10 μm), and K + U (KN-93 + U0126, 10 μm). *P < 0.01 compared to control, one-way ANOVA, Dunnett's post hoc test; error bars indicate s.e.m.

Figure 3

Nicotinic activation of α7-nAChR depresses GABAergic IPSCs in mouse hippocampal CA1 interneurons A, diagram showing electrode positions to elicit IPSCs by electrical stimulation in SR (Stim) while recording from an SR interneuron (Record). Nicotine 100 μm, 200 ms was delivered to the dendritic field of the neuron by glass pipette (Nic) 2–3 s before stimulation. SP, stratum pyramidale; SLM, stratum lacunosum moleculare. B, SR-elicited IPSCs before (SR-IPSC) and after (Nic → SR-IPSC) nicotine application in the absence (top traces) or presence (bottom traces) of 50 nm MLA. C, compiled results (mean ±s.e.m., n = 10 neurons) of IPSCs elicited in the presence of nicotine (± MLA) expressed as a fraction of the response in the absence of nicotine. *P < 0.01, Student's t test.

Figure 4

SO stimulation to activate cholinergic fibres depresses IPSCs elicited subsequently by SR stimulation A, diagram showing positions of stimulating (Stim) and recording electrodes (Record) in the SO and SR. B, pairs of IPSCs were induced by stimulating first one input (e.g. SR) and then the other (SO) after a 200 ms delay, and then repeating after a 1 min delay, and averaging 6-8 such trials. The stimulation order was then reversed and the process repeated. C, compiled results showing the mean amplitude (±s.e.m., n = 11 neurons) of an IPSC when it was elicited as the second of a pair, normalized for its amplitude when elicited first. When SO stimulation preceded SR stimulation, the SR-elicited IPSC (SR-IPSC) was significantly smaller than when the stimulation order was reversed. Stimulation order had no effect on SO-elicited IPSCs (SO-IPSC). *P < 0.01, one-way ANOVA, Dunnett's post hoc test.

Figure 5

Knockout mice and nicotinic antagonists demonstrate that α7-nAChRs mediate the inhibitory effect of SO stimulation on SR-elicited IPSCs A, MLA (50 nm, upper traces) prevented SO stimulation from affecting the amplitude of SR-elicited IPSCs. SO stimulation also had no effect on SR-elicited IPSCs in hippocampal slices from α7-nAChR knockout mice (bottom traces). B, compiled data (mean ±s.e.m., n = 6–11 neurons) showing no effect of SO stimulation on SR-elicited IPCSs in the presence of MLA (MLA) or in α7-nAChR knockout slices (α7 KO), in contrast to wild-type slices in the absence of MLA (control). *P < 0.05, one-way ANOVA, Dunnett's post hoc test.

Figure 6

Infusing a calcium chelator or blockers of CaMKII and MAPK into the patch-clamped neuron prevents SO stimulation from depressing SR-elicited IPSCs A, traces showing SR-elicited IPSCs alone (control) and after pairing with SO stimulation (SO-paired) as in Fig. 4. The cells were infused either with 11 mm BAPTA (upper traces) or 10 μm KN-93 + 10 μm U0126 (lower traces). B, quantification of changes in SR-elicited IPSC amplitude induced by pairing with a preceding SO stimulation, normalized for IPSC amplitude in the absence of pairing. Cells were infused with normal solution (Control) or solution containing either BAPTA or KN-93 + U0126 as indicated. Both treatments completely abolished the effect of SO stimulation on SR-elicited IPSCs, compared to controls. *P < 0.01, one-way ANOVA, Dunnett's post hoc test. C, traces showing paired SR-elicited IPSCs with (lower) or without (upper) a preceding stimulation in SO. D, quantification showing that SO stimulation did not affect the paired-pulse ratio of SR-evoked IPSC amplitudes. P > 0.05, Student's t test.

Figure 7

Spontaneous activation of α7-nAChRs by endogenous agonist enhances rundown of GABA_A receptors in hippocampal CA1 interneurons A, superimposed traces showing the currents induced by application of GABA (100 μm, 100 ms) at 10 s intervals (arrows). B, time course of changes in the mean IPSC amplitude normalized for the initial value seen in the presence (•) and absence (^) of 50 nm MLA. Time to 50% depression (τ) was 10.3 ± 0.6 s and 14.8 ± 0.9 s for control and MLA conditions, respectively. The curves represent the best fits of the data using the calculated time constants and standard exponential simulation. C, MLA significantly increased the τ-value (Control) normalized for that obtained in the absence of MLA, unless the slice was pretreated with αBgt (αBgt) or was obtained from α7-nAChR knockout mice (α7 KO). The results represent the mean ±s.e.m. with n = 8–13 neurons. *P < 0.01, Student's t test.