Diazepam and ethanol differently modulate neuronal activity in organotypic cortical cultures

Abstract

Background: The pharmacodynamic results of diazepam and ethanol administration are similar, in that each can mediate amnestic and sedative-hypnotic effects. Although each of these molecules effectively reduce the activity of central neurons, diazepam does so through modulation of a more specific set of receptor targets (GABA_A receptors containing a γ-subunit), while alcohol is less selective in its receptor bioactivity. Our investigation focuses on divergent actions of diazepam and ethanol on the firing patterns of cultured cortical neurons.

Method: We used electrophysiological recordings from organotypic slice cultures derived from Sprague-Dawley rat neocortex. We exposed these cultures to either diazepam (15 and 30 µM, n = 7) or ethanol (30 and 60 mM, n = 11) and recorded the electrical activity at baseline and experimental conditions. For analysis, we extracted the episodes of spontaneous activity, i.e., cortical up-states. After separation of action potential and local field potential (LFP) activity, we looked at differences in the number of action potentials, in the spectral power of the LFP, as well as in the coupling between action potential and LFP phase.

Results: While both substances seem to decrease neocortical action potential firing in a not significantly different (p = 0.659, Mann-Whitney U) fashion, diazepam increases the spectral power of the up-state without significantly impacting the spectral composition, whereas ethanol does not significantly change the spectral power but the oscillatory architecture of the up-state as revealed by the Friedman test with Bonferroni correction (p < 0.05). Further, the action potential to LFP-phase coupling reveals a synchronizing effect of diazepam for a wide frequency range and a narrow-band de-synchronizing effect for ethanol (p < 0.05, Kolmogorov-Smirnov test).

Conclusion: Diazepam and ethanol, induce specific patterns of network depressant actions. Diazepam induces cortical network inhibition and increased synchronicity via gamma subunit containing GABA_A receptors. Ethanol also induces cortical network inhibition, but without an increase in synchronicity via a wider span of molecular targets.

Keywords: Benzodiazepines; Diazepam; Ethanol; GABAA receptors; Neocortex.

Fig. 1

Exemplary trace of a recorded cortical up-state. The early phase is dominated by a strong initial peak followed by a slow transient phase back to the baseline amplitude. Following this initial peak oscillatory activity develops. For our analyses, we excluded the initial peak segment and focused on the oscillatory phase. The black dots indicate the occurrence of action potentials

Fig. 2

Action potentials for a diazepam (DZP) and b ethanol (EtOH) relative to control conditions. Both diazepam (blue, left) and ethanol (red, right) decreased the number of action potentials in a concentration-dependent manner. a 30 µM diazepam had a strong effect on the spiking rate compared versus control conditions. Diazepam caused a decrease in the number of action potentials as indicated by Hedge’s g (g = 1.20 [0.70 2.59]) that was not significant after Bonferroni correction (p = 0.031, uncorrected). The decrease in spike rate from 15 µM to 30 µM diazepam was significant and strong (p = 0.0156; g = 1.20 [0.72 2.57]). b When compared to control conditions, 30 mM ethanol did not show a significant reduction of spike rate (p = 0.054, g = 0.43 [− 0.08 1.16]), but 60 mM significantly reduced the spiking rate (p = 0.003; g = 0.69 [0.31 1.26]). The change from 30 mM to 60 mM ethanol was weak and not significant after Bonferroni correction (p = 0.042, uncorrected; g = 0.22 [− 0.11 0.51]). *p < 0.05 Bonferroni corrected; ^#p < 0.05 uncorrected

Fig. 3

Cumulative probability plots for the action potential frequency distribution in the first 200 ms of each up-state for diazepam (left) and ethanol (right). Diazepam significantly affected this distribution, whereas ethanol did not. p < 0.001 for all comparisons between the diazepam groups (control vs. 15 µM; control vs. 30 µM; 15 µM vs. 30 µM). For the ethanol experiments the test results were p = 0.799 (cnt vs. 30 mM); p = 0.364 (cnt vs. 60 mM); and p = 0.867 (30 mM vs. 60 mM)

Fig. 4

Relative change in the duration of up-states for a diazepam (DZP) and b ethanol (EtOH). a Diazepam had a strong effect on the duration of up-states when compared versus control conditions (15 µM: (p = 0.031, uncorrected; g = − 0.88 [− 0.43 − 1.77]); 30 µM diazepam (p = 0.11; g = − 1.25 [− 0.54 − 2.75], and p = 0.578 and g = − 0.04 [− 0.80 0.47] for 15 µM vs. 30 µM diazepam. b While 30 mM ethanol had no effect on up-state duration when compared versus control conditions (p = 0.432; g = − 0.14 [− 0.82 0.32]), 60 mM ethanol had a weak, but significant (p = 0.0488, uncorrected) effect (g = 0.40 [− 0.11 1.05]) on up-state duration when compared versus control conditions. Further, 60 mM ethanol had a medium effect causing shorter up-states (p = 0.0195, uncorrected; g = 0.55 [0.28 1.05), when compared against 30 mM ethanol. ^#p < 0.05 uncorrected; ^§strong effect

Fig. 5

Relative changes in absolute power spectral density (PSD) or normalized PSD (nPSD) as induced by diazepam (DZP) or ethanol (EtOH). a DZP-induced changes in PSD: DZP concentration-dependently increases the power (i.e., the amplitude) in a wide range of frequencies. The grey and blue horizontal line indicate a significant effect of low (15 µM, grey) or high (30 µM, blue) DZP concentration vs. control (CNT). A horizontal bar in dark blue indicates a significant difference between 15 and 30 µM DZP. b DZP-induced changes in nPSD: The oscillatory composition did not change in a significant fashion, except for a narrow frequency range around 20 Hz for 15 µM DZP. c EtOH-induced changes in PSD: EtOH does not affect the power (i.e., the amplitude) in LFP oscillations. d EtOH-induced changes in nPSD: the oscillatory composition changed in a significant fashion towards a stronger contribution of higher frequencies above 10 Hz for the low EtOH concentration (30 mM) versus control as indicated by the horizontal bars. The solid trend lines indicate the median and the shaded areas the median absolute deviation. The horizontal bars indicate a significant difference (p < 0.05, Wilcoxon signed rank test) for the comparison indicated by the color of the bar

Fig. 6

Diazepam-induced changes of the action potential to local field potential phase relationships. Especially at the high diazepam concentration (dark blue) peaks in the distribution develop that are indicative of a strong spike to phase locking. DZP diazepam, CNT control conditions

Fig. 7

Ethanol-induced changes of the action potential to local field potential phase relationships. Application of ethanol (EtOH) leads to more uniform distribution of AP and LFP phase relationships. CNT control conditions