Abstract
294
Objectives The metabotropic glutamate receptor 5 (mGluR5) is involved in addiction. Clinical PET studies have shown lower mGluR5 binding potential (BP) in addicted patients. However, it is still unclear whether this decrease is a state or a trait condition. Therefore, we longitudinally evaluated the mGluR5 BP and glutamate/glutamine (GLU/GLN) concentrations of rats in a cocaine addiction paradigm to investigate pre-exposure decreases in animals prone to addiction.
Methods Thirty-six animals underwent a 6 week self-administration (SA) protocol consisting of a baseline week, two acquisition weeks, two withdrawal weeks and one relapse week. Rats had free access to cocaine, via intravenous catheters, for 3 hours per day by pressing a lever during the acquisition and relapse weeks. Animals were scanned every week with a FOCUS-220 microPET for 60 min, following a bolus injection of 18F-FPEB (17 ± 1 MBq, > 35 GBq/µmol). BP values were extracted for the striatum (S), nucleus accumbens (NAC), hippocampus (HC), and frontal cortex (FCX). Eighteen animals additionally underwent 9.4T 1H-MR Spectroscopy in the NAC and pre-frontal cortex (PFC).
Results Self-administration of cocaine increased progressively within the two acquisition weeks and during relapse. Modelling this increase with baseline measurements, it was found that an increase of 1 in the 18F-FPEB baseline BP of the NAC decreased the number of lever presses (LP) with 64 ± 34, whereas in the FCX it increased the LP with 76 ± 48. Additionally, a 1 mmol/L increase in baseline glutamate in the PFC as measured by MRS increased LP on average with 10 ± 3. Considering all time-points there was a predictive effect of mGluR5 BP in the hippocampus, and both GLU and GLN in the PFC. Finally, the GLU concentration in the NAC was significantly associated with the increase in LP between the second acquisition week and relapse.
Conclusions This longitudinal multi-modal study of a cocaine SA model in rats shows that mGluR5 and glutamate levels are significantly associated with drug intake. Moreover, the results suggest that pre-existing neurochemical differences are related to the propensity of drug use. Figure 1. A mixed model using only baseline values can explain 55% (p<0.001) of variation in the observed SA.