Alprazolam-induced EEG spectral power changes in rhesus monkeys: a translational model for the evaluation of the behavioral effects of benzodiazepines

Abstract

Rationale: Benzodiazepines induce electroencephalography (EEG) changes in rodents and humans that are associated with distinct behavioral effects and have been proposed as quantitative biomarkers for GABA_A receptor modulation. Specifically, central EEG beta and occipital EEG delta activity have been associated with anxiolysis and sedation, respectively. The extent to which nonhuman primates show the same dose- and topography-dependent effects remained unknown.

Objectives: We aimed at establishing a nonhuman primate model for the evaluation of benzodiazepine EEG pharmacology.

Methods: Four adult male rhesus monkeys were prepared with fully implantable telemetry devices that monitored activity, peripheral body temperature, and contained two EEG (central and occipital), one electromyography (EMG), and one electrooculography channel. We investigated daytime alprazolam-induced changes in EEG spectral power, sleep-wake states, EMG activity, locomotor activity, and body temperature. Alprazolam (0.01-1.8 mg/kg, i.m.) or vehicle was administered acutely, and telemetry recording was conducted for 1 h.

Results: Daytime alprazolam dose-dependently increased central EEG power (including beta activity), increased occipital EEG delta power, and decreased occipital EEG alpha, theta, and sigma power. There was an ~8-fold difference in the potency of alprazolam to increase central EEG beta vs. occipital EEG delta activity (based on relative EEG power). The highest dose, which increased both central EEG beta and occipital EEG delta relative power, induced sedative effects (increased time spent in N1 and N2 sleep stages) and decreased peripheral body temperature and locomotor activity.

Conclusions: Alprazolam induces dose- and topography-dependent EEG changes in rhesus monkeys and provides a valuable model for studying benzodiazepine pharmacology.

Keywords: Alprazolam; Benzodiazepine; EEG; Rhesus monkey; Sedation; Spectral power.

Figure 1.

Sleep-wake state analysis of selected doses of alprazolam after daytime administration in rhesus macaques. Sleep analysis was conducted with the central derivation following the first hour after i.p. injection of vehicle (V), 0.3 mg/kg alprazolam (a dose below the ED₅₀ for beta band increases, see Fig. 2), and 1.8 mg/kg alprazolam (the highest dose tested). Data are from individual monkeys (individual colored symbols) and mean + SEM (bars with error bars). A one-hour period of scoring with no injection and at approximately the same time as the alprazolam treatments served as baseline (BL). Note that *p<0.05 vs. vehicle, Dunnett’s test.

Figure 2.

Effects of alprazolam on frequency bands in central and occipital derivations. For clarity, only the 4 highest doses are shown. Left panel illustrates raw EEG power (referred to as “power” in uV2/Hz, on a logarithmic scale). Middle panel illustrates normalized power, which is the raw EEG power for an experimental condition compared to baseline raw EEG power. Right panel illustrated normalized relative power, or raw EEG power in each separate band as a percent of the absolute power summed over the five frequency bands for each 5-sec epoch, normalized to baseline relative power. Top panels show the three measures for the central EEG derivation, and bottom panels shows the same measures for the occipital EEG derivation. All data are shown as mean + SEM (N=4). Each bandwidth was analyzed with a linear trend model (repeated measures ANOVA), significant trends are indicated by the “v” symbol.

Figure 3.

Comparison of non-linear regression functions for alprazolam-induced increases in beta frequency bands (central derivation) and delta frequency bands (occipital derivation) in rhesus monkeys. Data are changes in EEG bandwidth, normalized to percentage of baseline (mean - SEM, N=4). Functions were obtained through non-linear regression analyses using the 4-parameter logistic equation, with minimal effect constrained to 100. The ED₅₀ values, shown as mg/kg and SEMs, were significantly different (F-test, p<0.05).

Figure 4.

Peripheral body temperature (top panel) and activity (bottom panel) following vehicle and doses of alprazolam in rhesus monkeys. Each symbol represents an individual monkey, identified by in the legend. The functions connect to horizontal bars that represent the means for each dose or vehicle (V). Data are percent of baseline (i.e., in absence of injections). Multiple comparisons, represented by the bars with p<0.05, were Dunnett’s tests.