Speaker: Randy M. Bruno
Department: Dept. of Neuroscience
Subject: The many layers of the neocortex
Location: EMC Rotterdam
Date: 5 October 2015
Author: Carolien Bastiaanssen
The neocortex is deeply involved in any human and animal cognition. Sensation, problem solving, executing motor problems, you name it, the neocortex plays an influential role in it. This six layer structure in the brain is the focus of Randy Bruno and his team, they want to know how the neocortex does all this. In this seminar mr Bruno presented their latest results about the different layers in the neocortex and how they are (not) dependent.
Our sensory systems transduce information in the periphery. Next it is propagated centrally to the thalamus, which relays this information to primary areas of the cortex. The conventional model is that axons from the thalamus go to layer four of the neocortex. The neocortex’s cells transduce the signal to layers two and three, that activate layers five and six. These last layers give feedback to “the outer world”.
Mr Bruno works on the rodent whisker system. This system is very similar to our sensory system of the skin. They observed that most branches of the axons from the thalamus end in layer four, but there are also many branches ending in the higher layers five and six. He asked himself why this would be. Could it be some sort of shortcut?
To answer this question they used rats and recorded the reaction of neurons from many layers after a stimulus was applied. The results showed that cells in layer four responded before cells in layers two and three responded, this was expected and it is in line with the theory mentioned before. Yet they also saw that half of the layer five cells depolarised at the same time or even before the layer four cells. And the other half of the layer five cells reacted later than layer four but earlier than the cells from layer two and three. Another experiment to study the connection between the different layers of the cortex was done in mice. A pipe filled with artificial cerebrospinal fluid (ACSF) was inserted in layer four of the neocortex and a patch was inserted a few micrometres away in the same layer. This allowed them to observe the action potential and random activity. When the ACSF was swapped for lidocaine, a local anaesthetic, there was no field potential anymore and just a flat line was observed. Changing back to ACSF allowed for a full recovery of the signal. Next the patch was moved to layer five. Lidocaine was used to deactivate layer four and the behaviour of cells in layer five was observed. The signal was normal, indicating that the layer five response does not require layer four.
Next mr Bruno wanted to know what the difference is in the function of cells in different layers of the neocortex. He and his team looked at the response of mice neurons to many different kinds of stimuli. They could move the whiskers of the mice separately and in any direction. The neurons of interest do not fire so often therefore they looked at the voltage and not at peaks above the threshold. This enabled them to find the pattern of stimuli needed to get the maximum response. For layer four and layer five and six cells the optimized stimulus gave better response than a random stimulus. Yet for layer two and three cells there was no difference. A second experiment again used mice, these mice were trained to identify a pole in the dark. The animal self-initiates the test by using a lever. Then a pole will sometimes come down and the animal has to move its whiskers to detect it. When it correctly identifies a pole and lets go of the lever it is rewarded with a drop of water. If the make a mistake they are punished with a time out. When the trial starts a peak in activity is observed at the tops of apical dendrites that rise from layer five to layer one. There is a second peak around the time the animal expects or gets its reward. Layer four cells however do not have these apical dendrites and therefor they do not have reward-related activity. So the reward-related signal in layer five is independent of the signal in layer 4.
Figure 1: The model on the left is the conventional model, in which a chain of events transforms the sensory information of the thalamus. The model on the right is the model mr Bruno proposed, with two strata. Source: Constantinople, CM. Bruno, RM. Deep cortical layers are activated directly by thalamus, Science, 2013 Jan 28
Mr Bruno’s findings indicate that instead of a chain of events, there are two independent strata and layer four is not the first stage of the cortical processing. This contradicts with the way people in the field of neurology have thought about the way the neocortex works for decades. It was no surprise that several individuals in the audience asked sceptical questions. The response from the audience shows that these ideas are revolutionary. It was interesting to see how mr Bruno did not just accept the common theory but that he looked at some small details which were ignored before.