Permanent suppression of cortical oscillations in mice after adolescent exposure to cannabinoids: receptor mechanisms

Abstract

Marijuana use in adolescence, but not adulthood, may permanently impair cognitive functioning and increase the risk of developing schizophrenia. Cortical oscillations are patterns of neural network activity implicated in cognitive processing, and are abnormal in patients with schizophrenia. We have recently reported that cortical oscillations are suppressed in adult mice that were treated with the cannabinoids WIN55,212-2 (WIN) or Δ(9)tetrahydrocannabinol (THC) in adolescence, but not adulthood. WIN and THC are cannabinoid-1 (CB1R) and CB2R agonists, and also have activity at non-cannabinoid receptor targets. However, as acute WIN and THC administration can suppress oscillations through CB1Rs, we hypothesize that a similar mechanism underlies the permanent suppression of oscillations by repeated cannabinoid exposure in adolescence. Here we test the prediction that cannabinoid exposure in adolescence permanently suppresses cortical oscillations by acting through CB1Rs, and that these suppressive effects can be antagonized by a CB1R antagonist. We treated adolescent mice with various cannabinoid compounds, and pharmacologically-evoked oscillations in local field potentials (LFPs) in vitro in adults. We find that WIN exposure for six days in early adolescence suppresses oscillations preferentially in adult medial prefrontal cortex (mPFC) via CB1Rs, and that a similar CB1R mechanism accounts for the suppressive effects of long-term (20 day) adolescent THC in adult somatosensory cortex (SCx). Unexpectedly, we also find that CB2Rs may be involved in the suppression of oscillations in both mPFC and SCx by long-term adolescent cannabinoid exposure, and that non-cannabinoid receptors may also contribute to oscillation suppression in adult mPFC. These findings represent a novel attempt to antagonize the effects of adolescent cannabinoid exposure on neural network activity, and reveal the contribution of non-CB1R targets to the suppression of cortical oscillations.

Keywords: Development; Marijuana; Neural synchrony; Schizophrenia.

Figure 1. Kainic acid (KA) and carbachol (CCh) evoke robust oscillations in vitro in adult neocortex

(A) Experimental time course: comparisons of human and rodent development are adapted from (Andersen, 2003). Cannabinoids were administered to mice either for 6 days during early (P35–P40) or late adolescence (P47–P52), or for 20 days from P35–P55. LFPs were recorded in vitro from adult mice (>P80). (B) One-second example of LFPs recorded in vitro from mPFC of an adult mouse exposed to a vehicle solution from P35–P55. LFPs were recorded before (Baseline) and after perfusion of KA (400 nM) + CCh (20 μM), which together evoke robust oscillations in LFPs. Oscillation power is quantified in the accompanying Fourier transform of 10-second LFP recordings. KA + CCh (black trace) increases power at all frequencies [theta (θ; 4–7 Hz); alpha (α 8–12 Hz); beta (β; 13–29 Hz); gamma (γ 30–80 Hz)] relative to baseline conditions (gray trace). (C) Similar to B, but example LFPs and Fourier transforms are from adult SCx. Note that KA + CCh evokes gamma oscillations in SCx that have approximately double the power as those recorded in mPFC.

Figure 2. Exposure to WIN in early adolescence suppresses oscillations in vitro in adult mPFC, but not SCx, via CB1Rs

(A) Experimental time course. Animals were injected with the CB1R/CB2R agonist WIN (2 mg/kg) or vehicle + the CB1R antagonist AM251 (2 mg/kg) in early adolescence (P35–P40) once daily for six days. LFPs were recorded in vitro in slices of adult mPFC (B–D) or SCx (E,F). (B,C) Cumulative probability distributions of normalized oscillation power plotted on a log scale from LFPs recorded in mPFC of adult mice treated with vehicle (solid line) or (B) WIN (dashed line) or (C) AM251 (dashed line) from P35–P40. Kolmogorov-Smirnoff tests were used to compare the effect of cannabinoid treatment on normalized oscillation power. (D) Plots as in (B,C), except normalized power was compared between animals treated with WIN or AM251 + WIN, and between vehicle and AM251 + WIN with KS tests. (E,F) Plots as in (B,C) for LFPs recorded in SCx.

Figure 3. Long-term exposure to THC in adolescence suppresses oscillations in vitro in adult mPFC and via CB1Rs in SCx

(A) Experimental time course. Animals were injected with the CB1R/CB2R agonist THC (5 mg/kg) or vehicle + the CB1R antagonist AM251 (0.3 or 0.5 mg/kg) in adolescence (P35–P55) once daily for 20 days. LFPs were recorded in vitro in slices of adult mPFC (B,C) or SCx (D–G). (B, C) Cumulative probability distributions of normalized oscillation power plotted on a log scale from LFPs recorded in mPFC of adult mice treated with vehicle (solid line) or (B) THC (dashed line) or (C) AM251 (dashed line) from P35–P55. KS tests compared the effect of adolescent cannabinoid treatment on normalized oscillation power. (D,E) Plots as in (B,C) of LFPs from adult SCx. (F,G) Plots as in (D,E), except normalized power was compared between animals treated with THC or AM251 + THC [0.3 mg/kg AM251 in (F); 0.5 mg/kg AM251 in (G)], and between vehicle and AM251 + THC with KS tests.

Figure 4. Long-term exposure to WIN or AM251 in adolescence suppresses oscillations in vitro in adult mPFC and SCx

(A) Experimental time course. Animals were injected with the CB1R/CB2R agonist WIN (1 mg/kg) or vehicle + the CB1R antagonist AM251 (1 mg/kg) in adolescence (P35–P55) once daily for 20 days. LFPs were recorded in vitro in slices of adult mPFC (B,C) or SCx (D,E). (B,C) Cumulative probability distributions of normalized oscillation power plotted on a log scale from LFPs recorded in mPFC of adult mice treated with vehicle (solid line) or (B) WIN (dashed line) or (C) AM251 (dashed line) from P35–P55. KS tests compared the effect of adolescent cannabinoid treatment on normalized oscillation power. (D,E) Plots as in (B,C) of LFPs from adult SCx.

Figure 5. Long-term exposure to CB1R or CB2R antagonists during adolescence suppresses oscillations in vitro in adult mPFC, with milder effects in SCx

(A) Experimental time course. Animals were injected with the CB1R/CB2R agonist WIN (1 mg/kg) or vehicle + the CB1R antagonist AM4113 (1 mg/kg) or the CB2R antagonist AM630 (1 mg/kg) in adolescence (P35–P55) once daily for 20 days. LFPs were recorded in vitro in slices of adult mPFC (B,C) or SCx (D–F). (B,C) Cumulative probability distributions of normalized oscillation power plotted on a log scale from LFPs recorded in mPFC of adult mice treated with vehicle (solid line) or (B) AM4113 (dashed line) or (C) AM630 (dashed line) from P35–P55. KS tests compared the effect of adolescent cannabinoid treatment on normalized oscillation power. (D) Plot as in (B) of LFPs from adult SCx. (E) Plot as in (D), except normalized power was compared between animals treated with WIN or AM4113 + WIN, and between vehicle and AM4113 + WIN with KS tests. (F) Plot as in (C) of LFPs from adult SCx.

Figure 6. Long-term exposure to the putative inactive enantiomer WIN55,212-3 (WIN-3) in adolescence suppresses oscillations in vitro in adult mPFC, but not SCx

(A) Experimental time course. Animals were injected with WIN-3 (1 mg/kg) or vehicle in adolescence (P35–P55) once daily for 20 days. LFPs were recorded in vitro in slices of adult mPFC (B) or SCx (C). Cumulative probability distributions of normalized oscillation power plotted on a log scale from LFPs recorded in adult mice treated with vehicle (solid line) or WIN-3 (dashed line) from P35–P55. KS tests compared the effect of adolescent WIN-3 treatment on normalized power.