Limbic interactions of the dorsal medial prefrontal cortex
By Vanessa NyBlom and Jim Stellar
The modern understanding of how the neocortex supports cognition is that the 6-layered neocortex builds internal models of the world from input, rather than simply passively processing that input into an image. For example, if you see a ball rolling across the floor and it goes behind a screen, our eyes jump to where it will come out, and if it does not appear at the right time, we humans express surprise, even as a child. That means the brain is not just eye-tracking the ball like a lizard (it does that too with deep brain structures), but it is building an internal model of the ball as having a trajectory. This capability is what neuroscience calls the inside-out-brain, rather than the outside-in-brain, and it confers superior processing on all species who possess this trait with a significantly large 6-layered neocortex.
The point of that story is that the deep brain limbic systems structures (e.g. accumbens, amygdala) do not have this 6-layered structure and do not appear to model the external world in the same way. They do learn and after a tasty treat of, let’s say chocolate, the accumbens and its dopamine circuitry do predict when a stimulus that led to the first treat is about to lead to another. This is basic learning, and it is incredibly important and representational. But it is not inside-out brain modeling of the external world.
We have written about the prefrontal cortical-limbic connections in a past few blogs, but mainly from the point of view of top-down control. This time we focus specifically on the dorsal medial prefrontal cortex (dmPFC) and its interaction with incentive reactions generated in reward-related limbic structures like the accumbens from both a control and a cortical representation perspective. This could be important in humans, especially with the prefrontal cortex’s expanded structure compared to lower animals, like the ordinary laboratory rat. As one of us likes to say, “rats and humans may visit art museums for different reasons.”
Following the theme of this blog, the dmPFC might be the one of the places in the brain where college students who have had direct experiences like an internship extract the emotional value of that experience when considering their overall academic plan of study and their ultimate career choice.
Where is the dmPFC? It is located on the top medial part of the prefrontal cortex, as shown below.
The dmPFC has long been suggested to have many high level functions, including ones connected to emotions. One of them is reading the mental (emotional?) states of others, which is part of the well-known theory of mind brain network in humans and is suggested to result in empathy, cooperation, etc. We know the dmPFC’s role in the general PFC to be in reward-prediction and the incentive aspect of these kinds of reactions. For example, one study which we looked at was done on rats and suggests the dmPFC can directly control neural firing in the accumbens from the dopaminergic “reward” projection that comes from the ventral tegmental area (VTA). In this specific study, they first trained rats to do an operant response to get a liquid sucrose reward. Then, they trained them that one auditory cue signaled that the lever worked and a different auditory cue signaled that the lever did not work. For the test the rats were hungry after a week of restricted food and water access. Finally, they implanted recording electrodes in the nucleus accumbens (where dopamine projects and seems to produce reward) and small cannulae for microinfusion into the dmPFC of a mixture of two GABA agonist drugs that tend to inactivate neurons. This way they could record the behavior and neural activation in the accumbens after injecting the druts (or the saline vehicle) into the dmPFC. Some of the behavioral and accumbens recording results are shown below for a few aspects of their findings.
The figure above shows in part A the inhibiting effect of the GABA agonist microinfusion of the dmPFC in reducing the behavioral response as compared to the baseline in the condition (DS). Part B above shows the effect of increasing the latency to respond after that same microinfusion.
The figure above shows the effect on the accumbens neuron firing under the same conditions. Notice the bigger response under saline dmPFC infusion on the right compared to when the GABA agonists were microinfused at low (25 ng) or high (50) ng doses. This result indicates that inactivating the neurons of the dmPFC reduces the firing of the accumbens neurons to the incentive cue, paralleling the changes in behavior just discussed. We think that is the first step of the dmPFC in controlling the accumbens as we considered in our last blog.
So how does the dmPFC represent such limbic processes at the cortical level? To open the discussion here, we switch to an fMRI study in humans that looked at value trade-offs between immediate and delayed rewards that in the above studies were likely processed in the accumbens. Here they found that, and we quote from the article, “brain activity in the posterior dmPFC was modulated by the amount of immediate options, whereas the activity in the adjacent anterior dmPFC, … was modulated by the amount of delayed options (underlining is ours). In the figures below, taken from their paper, the posterior dmPFC fMRI activation is shown on the left and the anterior dmPFC is shown on the right.
There is much complexity in the PFC and its regions that is represented in this paper from which the two figures above were taken and combined here for the purpose of this presentation. What we think is important here is that the dmPFC is representing value (from immediate and delayed rewards directly) and thus has a much more sophisticated representation of the kinds of reward processes mentioned earlier in the accumbens, over which it also appears to have control, at least in rats.
In future blogs on the firing of the dmPFC and other related areas, as discussed in the beginning, we will look for other ways the PFC in both the human and primate cortex builds models of reward processes. Those operations would then allow for the limbic system to “advise” the rest of the cortex and allow various neocortical regions to talk to each other. For example, as stated above, that might allow college students to process the emotional contributions of their direct experience (e.g undergraduate courses and internships) as they relate that input to their academic curriculum and future career plans, which is the overall focus of the many blogs in this blog series.