Neural circuit selective for fast but not slow dopamine increases in drug reward

Abstract

The faster a drug enters the brain, the greater its addictive potential, yet the brain circuits underlying the rate dependency to drug reward remain unresolved. With simultaneous PET-fMRI we linked dynamics of dopamine signaling, brain activity/connectivity, and self-reported 'high' in 20 adults receiving methylphenidate orally (results in slow delivery) and intravenously (results in fast delivery) (trial NCT03326245). We estimated speed of striatal dopamine increases to oral and IV methylphenidate and then tested where brain activity was associated with slow and fast dopamine dynamics (primary endpoint). We then tested whether these brain circuits were temporally associated with individual 'high' ratings to methylphenidate (secondary endpoint). A corticostriatal circuit comprising the dorsal anterior cingulate cortex and insula and their connections with dorsal caudate was activated by fast (but not slow) dopamine increases and paralleled 'high' ratings. These data provide evidence in humans for a link between dACC/insula activation and fast but not slow dopamine increases and document a critical role of the salience network in drug reward.

Fig. 1. Experimental design.

a Timeline of events. In each session, participants were given an oral dose of methylphenidate (MP) or placebo (PLA) at time 0; the [¹¹C] Raclopride bolus injection and simultaneous PET-fMRI scanning started at 30 min; an IV dose of MP or PLA was given at 60 min; and throughout the duration of the session participants used a button box in the scanner to self-report their experience of ‘high’ to the drug. b Session structure. Participants underwent three separate imaging sessions that were identical except for drug condition: Session A) oral PLA and IV PLA (black color); Session B) oral MP (60 mg) and IV PLA (3 cc saline) (pink color); Session C) oral PLA and IV MP (0.25 mg/kg in 3 cc sterile water) (red color). c Hypothesized cortical-striatal fMRI signal from a simplified model based on postsynaptic dopamine receptor stimulation. We hypothesized opposing fMRI signal patterns to the IV (fast brain delivery) versus oral (slow brain delivery) MP doses, based on previous work.

Fig. 2. Dynamic dopamine increases and associated brain activity to oral and intravenous (IV) methylphenidate (MP).

a Relative standardized uptake value (SUVr) showing the binding of [¹¹C]raclopride in the striatum for each drug condition. b Delta SUVr (i.e., Oral MP – Placebo and IV MP – Placebo) showing minute-by-minute differences in dopamine receptor occupancy to each drug administration. c The rate of dopamine increases to MP (derived from the derivative of the plot in panel b), which was used in subsequent analyses to estimate where brain activity paralleled speed of dopamine increases to MP. d Whole-brain analysis showing where brain activity was significantly associated with slow dopamine increases (oral MP) across time. e Whole-brain analysis showing where brain activity was significantly associated with fast dopamine increases (IV MP) across time. f Time courses of the BOLD fMRI signal extracted from the ventromedial prefrontal cortex (vmPFC) cluster that had significant decreases in activity to both oral and IV MP. g Beta values for each participant demonstrating the fit between the time-course of vmPFC activity and speed of dopamine increases (n = 20 biologically independent adults). h Time-courses of the BOLD fMRI signal extracted from the dorsal anterior cingulate cortex (dACC) cluster that had increases in activity selective to IV MP. i Beta values for each participant demonstrating the fit between the time-course of dACC activity and speed of dopamine increases (n = 20 biologically independent adults). Data in panels (a) and (b) were previously reported. In panels (a), (b), (f), and (h), the lines represent the mean of the 20 participants, and the shaded regions represent the standard error of the mean; the vertical dashed line denotes the time of the IV MP or placebo injection. The black color denotes the placebo session; pink denotes the oral MP session; and red denotes the IV session. fMRI time courses were temporally smoothed (for visualization only), and the y-axis units represent the percent signal change from the mean signal during the ‘baseline’ period, i.e., the first ten minutes at the beginning of the scan. Source data are provided as a Source data file.

Fig. 3. Dynamic functional connectivity (FC) in a dorsal corticostriatal circuit is selectively associated with speed of fast, but not slow, dopamine increases and subjective ‘high’ to methylphenidate (MP).

a Dynamic brain FC in association with the speed of dopamine increases (as estimated by PET imaging), to intravenous (IV) MP. The significant dorsal anterior cingulate cortex (dACC) cluster from the analysis in Fig. 2e was used as a seed region. In voxelwise whole-brain analysis, the dorsal caudate emerged as the only region significantly positively connected with the dACC in association with speed of dopamine increases (negative clusters are also shown in Supplementary Fig. 8). Results from dACC are shown here, but the left insula also showed very similar significant patterns of dynamic FC with dorsal caudate (see Supplementary Fig. 9). b Time course of dynamic FC between dACC and dorsal caudate, for all three sessions. The lines represent the mean of the 20 participants, and the shaded regions represent the standard error of the mean. c Beta values for each participant demonstrating the fit between the time-course of dACC-to-dorsal caudate dynamic FC and speed of dopamine increases (n = 20 biologically independent adults). d Self-reported ‘high’ ratings for each drug condition, which were previously reported in another study using a portion of these data. The lines represent the mean of the 20 participants, and the shaded regions represent the standard error of the mean. e Beta values for each participant demonstrating the fit between the time-course of dACC-to-dorsal caudate dynamic FC and subjective ‘high’ ratings (n = 19 biologically independent adults). The black color denotes the placebo session; pink denotes the oral MP session; and red denotes the IV session. Note: because in the oral MP session only 13 out of 20 participants reported feeling any ‘high’, we only show the results for the IV session for this analysis. Source data are provided as a Source data file.