Direct actions of carbenoxolone on synaptic transmission and neuronal membrane properties

Abstract

The increased appreciation of electrical coupling between neurons has led to many studies examining the role of gap junctions in synaptic and network activity. Although the gap junctional blocker carbenoxolone (CBX) is effective in reducing electrical coupling, it may have other actions as well. To study the non-gap junctional effects of CBX on synaptic transmission, we recorded from mouse hippocampal neurons cultured on glial micro-islands. This recording configuration allowed us to stimulate and record excitatory postsynaptic currents (EPSCs) or inhibitory postsynaptic currents (IPSCs) in the same neuron or pairs of neurons. CBX irreversibly reduced evoked alpha-amino-3-hydroxy-5-methyl-4-isoxazole-proprionic acid (AMPA) receptor-mediated EPSCs. Consistent with a presynaptic site of action, CBX had no effect on glutamate-evoked whole cell currents and increased the paired-pulse ratio of AMPA and N-methyl-d-aspartate (NMDA) receptor-mediated EPSCs. CBX also reversibly reduced GABA(A) receptor-mediated IPSCs, increased the action potential width, and reduced the action potential firing rate. Our results indicate CBX broadly affects several neuronal membrane conductances independent of its effects on gap junctions. Thus effects of carbenoxolone on network activity cannot be interpreted as resulting from specific block of gap junctions.

FIG. 1.

Carbenoxolone (CBX) reduces α-amino-3-hydroxy-5-methyl-4-isoxazole-proprionic acid (AMPA) receptor–mediated excitatory postsynpatic currents (EPSCs). A: In this autaptic culture, we evoked AMPA receptor mediated EPSCs by brief (0.5 ms) depolarizations of the soma. Increasing concentrations of CBX applied sequentially to a neuron reduced the EPSC. The arrowhead indicates the action current (rendered in gray). B: CBX at 100 μM produced a greater decrease in AMPA receptor–mediated EPSCs than 10 or 50 μM (ANOVA, P < 0.05). C: CBX effect occurs quickly. Once CBX was applied, reduction occurred within the 7.5-s interstimulus interval. D: CBX did not reduce whole cell currents evoked by direct application of glutamate. In A and D, black traces are control recordings; red traces indicate recordings from the same cell in the presence of the indicated concentration of CBX. Recordings were done in d-AP5 (100 μM) or d-CPP (10 μM) to block N-methyl-d-aspartate (NMDA) receptors.

FIG. 2.

AMPA receptor–mediated EPSC reduction results from a presynaptic effect of CBX. We used the paired-pulse ratio (PPR) to assess the release probability. A: pairs of AMPA receptor–mediated EPSCs were stimulated at 50-ms interpulse intervals. We normalized the first EPSC of the pair to show that CBX reduced the 2nd EPSC to a lesser extent than the 1st EPSC. B: the PPR of the AMPA receptor–mediated EPSC was higher in CBX (100 μM) compared with control, consistent with a presynaptic site of action. C: NMDA receptor-mediated EPSCs (in 5 μM NBQX to block AMPA receptors) were also reduced by CBX (100 μM) as shown in this recording (50-ms interstimulus interval) D: similar to AMPA receptor–mediated EPSCs, normalizing the 1st EPSC showed that the 2nd EPSC was reduced less by CBX (50-ms interstimulus interval). Dashed lines in A and D indicate the difference in amplitude of the 2nd peaks in control and CBX. Vertical scale bars in A and D indicate the amplitude fraction from the normalized EPSC of the pair. In A, C, and D, black traces are control recordings; red traces indicate recordings from the same cell in the presence of the indicated concentration of CBX.

FIG. 3.

GABA_A receptor–mediated inhibitory postsynaptic currents (IPSCs) are reduced by CBX. A: the reduction of IPSCs by CBX was reversible, as shown in this example using 50 μM CBX. B: CBX reduced GABA_A IPSCs in a dose-dependent manner. Black traces are control recordings; red traces indicate recordings from the same cell in the presence of the indicated concentration of CBX.

FIG. 4.

CBX alters action potential firing. Current injections (200–500 ms; 50–300 pA) in whole cell current clamp resulted in trains of action potentials. The action potentials in A are averaged aligned action potentials in control and CBX (100 μM) from a single neuron. B: pairwise comparison shows that CBX (50 and 100 μM) increased the action potential width at half-height (arrowhead in A). We used only the 1st action potential in the train for analysis. C: CBX (100 μM) reduced the number of action potentials and increased the interspike interval in the 1st 200 ms of depolarizing current injection (red traces) compared with control (black traces). The mean spike times for this neuron are shown below the voltage traces by the thick black vertical hash marks, whereas the individual spike times (25 repetitions per condition) are show in thin vertical hash marks. D: a pairwise comparison shows the reduction in the number of spikes in the 1st 200 ms in CBX (50 and 100 μM) compared with control. E: CBX (100 μM) also increased the interspike interval of the 1st pair of action potentials in response to sustained current injections. Open red circles are 50 μM CBX; closed red circles are 100 μM CBX.