Adolescent cannabinoid exposure permanently suppresses cortical oscillations in adult mice

Abstract

Regular marijuana use during adolescence, but not adulthood, may permanently impair cognition and increase the risk for psychiatric diseases, such as schizophrenia. Cortical oscillations are integral for cognitive processes and are abnormal in patients with schizophrenia. We test the hypothesis that adolescence is a sensitive period because of the active development of cortical oscillations and neuromodulatory systems that underlie them. The endocannabinoid system upon which marijuana acts is one such system. Here we test the prediction that adolescent cannabinoid exposure alters cortical oscillations in adults. Using in vitro local field potential, in vivo electrocorticogram recordings and cognitive behavioral testing in adult mice, we demonstrate that chronic adolescent, but not adult, cannabinoid exposure suppresses pharmacologically evoked cortical oscillations and impairs working memory performance in adults. The later-maturing prefrontal cortex is more sensitive to adolescent exposure than the earlier-maturing, primary somatosensory cortex. These data establish a link between chronic adolescent cannabinoid exposure and alterations in adult cortical network activity that underlie cognitive processes.

Figure 1

Robust oscillations are pharmacologically induced in vitro in adult mouse neocortex in adolescent vehicle but not WIN-treated mice. (a) Experimental time course: comparisons of human and rodent development are modified from (Andersen, 2003). The CB1R agonist WIN55,212-2 (WIN; 0.25 or 1 mg/kg) or vehicle were administered to adolescent (P35–P55) or adult mice (P70–P90) once daily for 20 days. LFPs were recorded in brain slices from adult mice (>P100). (b, c) 1-s epoch of in vitro LFP from mPFC slice from an adult mouse administered vehicle (b) or 1 mg/kg WIN (c) during adolescence. LFPs were recorded before (baseline) and during kainic acid (KA; 400 nM) and carbachol (CCh; 20 μM) perfusion (KA+CCh). (d, e) Fourier transform of 10-s LFP recordings in (b) and (c), respectively. KA+CCh (black trace) markedly increases power at all frequencies compared with baseline conditions (grey trace). Frequency ranges: θ=4–7 Hz; α=8–12 Hz; β=13–29 Hz; γ=30–80 Hz. Inset in (e) shows a magnified view of the FFT in (e).

Figure 2

Chronic adolescent WIN administration attenuates pharmacologically evoked oscillations in vitro in mPFC and SCx. (a, b) Spectrogram (left) and power spectral density (right) of a 1-s representative KA+CCh LFP recorded in mPFC of an adult mouse administered vehicle (a) or 1 mg/kg WIN (b) during adolescence. Minimum–maximum scales are the same in a and b. (c–f) Box and whisker plots (box: 25th percentile, median, 75th percentile; whiskers: adjacent value to 25% or 75% values) of power from FFTs of LFPs in mPFC of adult mice with KA+CCh perfusion. Mice were administered WIN (0.25 or 1 mg/kg) or vehicle during adolescence (Figure 1a). KW tests determined significant differences between the three treatment conditions and pairwise comparisons were performed using MWU tests (significant P< 0.05). (g, h) Spectrograms and power spectral densities of 1-s example KA+CCh LFP recorded in SCx of adult mouse administered vehicle (g) or 1 mg/kg WIN (h) during adolescence. Minimum–maximum scales are the same in g and h. (i–l) Box and whisker plots of power from FFTs of LFPs in SCx of adult mice in the presence of KA+CCh. Mice were treated and statistics were performed as described for (c–f).

Figure 3

Chronic adolescent THC administration suppresses oscillations in vitro in mouse mPFC but not SCx. (a) THC (5 mg/kg) or vehicle was administered to adolescent mice (P35–P55) and LFPs were recorded in slices from adult mice (>P100). (b–e) Box and whisker plots of power from FFTs of LFPs in mPFC of adult mice with KA+CCh perfusion. Power from adolescent THC or vehicle-treated mice were compared using MWU tests (significant P<0.05). (f–i) Box and whisker plots of power extracted from FFTs of LFPs in SCx of adult mice in the presence of KA+CCh. Mice were treated and statistics were performed as in (b–e).

Figure 4

Chronic adolescent administration of WIN attenuates cortical oscillations in vivo. (a) Representative time course of γ power in frontal ECoG before and after injection of 20 mg/kg ketamine. Mice were treated with 1 mg/kg WIN (grey trace) or vehicle (black trace) from P35 to P55 and ECoGs were recorded from adults (>P100). γ Power after ketamine injection was normalized to γ power after saline injection. (b) Box and whisker plots of ECoG γ power in adolescent vehicle (n=5) or WIN-treated (n=7) mice after injection of 10 mg/kg (light grey) or 20 mg/kg (dark grey) ketamine, normalized as above. B=minutes 0–10 of recording; 1=minutes 10–20 of recording; 2=minutes 20–30 of recording; 3=minutes 30–40 of recording; 4=minutes 40–50 of recording. Post-injection power was compared with baseline power with MWU tests (significant P<0.05). (c) Representative time course of α power in frontal ECoG before and after injection of 20 mg/kg in an adolescent vehicle (black trace) or WIN-treated (grey trace) adult mouse. Data are presented as in (a). (d) Box and whisker plots of ECoG α power in adolescent vehicle (n=5) or WIN-treated (n=7) mice after injection of 10 mg/kg (light grey) or 20 mg/kg (dark grey) ketamine, normalized to power before and after saline injection. Time segments are indicated and statistics were performed as in (b).