By Ruud van den Brink & Tobias Donner.
A class of neurotransmitters known as catecholamines (norepinephrine and dopamine) play an important modulatory role in brain function. These catecholamines are synthesised by neurons clustered in small nuclei of the brainstem, which have diffuse projections to the forebrain. This is particularly true for norepinephrine, which is released across the entire cerebral cortex and most other parts of the brain. Due to this widespread release, current models of norepinephrine function assume a uniform impact on neural activity across their brain. However, norepinephrine and dopamine exert their modulatory effects on their target regions via a variety of receptors; and these receptors, in turn, are non-uniformly distributed across the brain and vary in terms of their effect on neural activity.
In our new paper, we first show that catecholamines modulate intrinsic neural activity in a non-uniform way across the brain. We then go on to establish that this spatial heterogeneity of catecholaminergic neuromodulation is explained by the spatial heterogeneity of catecholamine receptors.
We gained these insights by re-analyzing an existing data set, in which we measured brain activity with fMRI and boosted catecholamine levels in healthy humans with atomoxetine (van den Brink et al, 2016). We exploited the well established observation that neural activity fluctuates continuously, even at times when we are not faced with external task demands. What’s more, the intrinsic activity fluctuations that occur during such periods of ‘rest’, are correlated across brain regions. We quantified these intrinsic correlations by computing the covariances of the activity fluctuations for all pairs of brain regions, which were then collected in the brain-wide “covariance matrix”. We computed one covariance matrix for the atomoxetine condition and another for the placebo condition. A previously established analysis approach (Donner et al, 2013) enabled us to decompose covariance matrices into spatial patterns that captured either an atomoxetine-induced increase, or decrease, in the strength of intrinsic correlations (Figure 1a). We refer to these patterns as “spatial modes”. We then correlated these spatial modes with maps of genes that encode catecholamine receptors. We took those receptor maps from a unique post-mortem dataset provided by the Allen Institute for Brain Science (Hawrylycz et al. 2012).
Consistent with the notion that the effect of catecholamines on intrinsic correlations varies across the brain, we found two spatial modes in which the effect of catecholamines was opposite (Figure 1b). While one spatial mode captured an increase in correlation strength, the other captured a decrease. The modes reflected drug effects that were both reliable across participants, and manipulation-specific.
Importantly, each of the two spatial modes was associated with the distribution of a specific set of catecholamine receptors (Figure 1c). The mode that captured an atomoxetine-induced increase in correlation strength was associated with dopamine ‘D2-like’ receptors, while the mode that captured a reduction in the strength of correlations was associated with noradrenergic α1 receptors. Moreover, both spatial modes were correlated with the noradrenergic β receptor, but with opposite signs. Finally, neither pattern was correlated with a set of control receptors (acetylcholine and NMDA).
Our findings indicate that catecholamines have remarkably heterogeneous effects on large-scale intrinsic correlations in brain activity. This heterogeneity is, at least in part, explained by the heterogeneous distribution of catecholamine receptors. Our findings link system-level effects of catecholamines to low-level properties of the underlying neurotransmitter systems. Looking forward, these results provide important constraints for the development of realistic models of neuromodulatory effects on large-scale brain network dynamics.
Reference: van den Brink RL, Nieuwenhuis S, & Donner TH (2018). Amplification and Suppression of Distinct Brain-wide Activity Patterns by Catecholamines. Journal of Neuroscience Early Release.