What is neocortex?

Written by Akiya Watakabe

1. Neocortex is a thin layered structure surrounding mammalian brains

Neocortex is a thin layered structure surrounding mammalian brains.  It is the hallmark of mammalian brains and not present in birds or in reptiles.  It is also the most divergent part across mammalian species (Fig. 1-1).  It is called "neo", because it is evolutionarily the newest part of the cerebreral cortex.  

Some people prefer to call it the "isocortex", because it sounds more neutral.  "Cerebral cortex" is almost synonymously used as neocortex, although this term includes hippocampus and rhinal cortex, in addition to neocortex (they are allocortex).  Just "cortex" usually means the same thing as the cerebral cortex.  

If you compare the brains of mouse, monkey and human (Fig. 1-1), you can see that the size and outlook is very different among the three brains.  The different outlook is attributable to neocortex (it is colored in Fig. 1-2).  

Mammalian brains

Fig. 1  Comparison of mouse, monkey and human brains

These are the sideviews of the whole brains. The left side of the  panel is to the front and the right site is to the back.  (Adapted from Comparative Mammalian Brain Collections)

In the mouse brain, the neocortical surface is smooth (it is called "lissencephalic", lizard-brain) and covers only the top part of the brain.  In the monkey and human brains, neocortex has many sulci or  wrinkles ("gyrencephalic") and covers almost the entire brain.  The expansion of neocortex is one of the most characteristic features of primates (especially humans).  The "cortical" or skin-like structure is apparent when you cut the brain in half (white lines).

neocortex is blue

Fig. 1-2 Comparison of neocortex among mouse, monkey and human.

The neocortical surfaces are colored blue.  (Adapted from Comparative Mammalian Brain Collections)

In the photo below, the thin-cut sections of mouse and monkey brains are stained by "Nissl staining", which stains the neurons in blue.  Neocortex is the outermost sheet of tissue (shown by blue) that surrounds the brain.

coronal sections

Fig. 1-3 Comparison of neocortex between mouse and monkey.

Neocortex is shown by blue.  The white matter, which consists of nerve fibers that travels from and to the cortex is shown by red. (The Nissl photos are from Comparative Mammalian Brain Collections)


2. Neocortex consists of multiple areas

In an essay written by Dr. Oliver Sacks, there appears one patient who lost color sight in a traffic accident ("The Case of Colourblind Painter" in "An Anthropologist On Mars").  This patient appeared to retain other visual functions, and the authors speculate that the cortical area dedicated for color vision was hurt in this patient.  

Like this patient, some people suffer from partial wounds in the brain caused by accidents, war, strokes, cancers, etc.  If the neocortex is hurt, the patients may lose varioius cognitive abilities.  The patients may lose speech capabilities, space recognition, eyesight, motor control, socialized behavior and so on.  What happens depends on which part of the neocortex loses its function.

What these observations show is that the neocortex consists of various "subunits" that perform different functions.  Such substructure of neocortex is called "area".  Certain functional areas, such as  those shown in Fig. 2-1, are present throughout mammalian species, suggesting that the ancestral mammal already possessed some prototypical area structure.  On the other hand, some areas, such as those required for color vision, or for speech, evolved only in selected species.  

The acquisition of higher cognitive function in primates accompanied the differentiation  of existing areas for more specialized function as well as emergence of novel functional areas.

rat areas

Fig. 2-1  Cortical areas of the rat.
  (The rat brain photo is from Comparative Mammalian Brain Collections)


3. Connectivity makes functional differences

The function of the nervous system depends on connectivity.  For example, the visual area can perform visual function because it receives visual information from the visual organ, the eye.  Similary,  the auditory area receives auditory information and the somatosensory area receives body sensations (Fig. 2-2).  Such area-specific connectivity forms the basis of functional localization.  

sensory areas

Fig. 3-1  Sensory areas of the cortex are connected to the peripheral sensory organs.

This figure shows that different sensations (vision, sound and tactile information) reach different cortical areas via segregated pathways.  The yellow circle means that each sensation passes through different relay "nuclei (aggregates of neurons)" before arriving the neocortex.  
  (The rat brain photo is from Comparative Mammalian Brain Collections)

Buried deep under the neocortical sheet lies a nucleus (aggregates of neurons) called "thalamus".  Thalamus has very tight connections with the neocortex.  

Like the neocortex, thalamus consists of many subnuclei.  Overall, the subnuclei of thalamus and the cortical areas exhibit one-to-one topological relationships.  This is especially evident for the sensory regions.  

The environmental inputs received by the peripheral sensory organs are transmitted directly (for visual pathway) or via several nuclei (for auditory and somatosensory pathways) to the thalamic subnucei and then relayed to respective sensory cortical areas.  The cortical areas that receive the thalamic inputs send back projections to the thalamic nuclei (Fig. 3-2).

Thus, there are segregated functional units for the processing of different sensations.  Thalamocortical relationships play central roles in such functional segregations of the neocortex.  

Fig. 3-2  Thalamus is the gateway to the neocortex

Neocortex is indirectly connected to the peripheral sensory organs.  Thalamus is the most important relay station that connects the neocortex with the outside world.


4.  Brodmann's "Localization in the cerebral cortex"

In the early 20th century, Brodmann described, in his famous book "Localization in the cerebral cortex", how the cerebral cortex of various mammals look like.   

His conclusions, based on the examination of 55 species, can be summarized as follows.

(1) All the neocortical areas derived from a prototypical six-layered structure.
(2) Neocortex can be subdivided into multiple areas, each of which differentiates into a unique structure.
(3) Despite specialization in each species, there are some homologous areas that exhibit structural similarity.  

What these conclusions imply is that different functional areas develop unique structures to cope with the need for different information processing.   Brodmann firmly believed that there must be an anatomical basis for any physiological phenomenon (e.g., see the last paragraph of Brodmann's book).  The homology of the structure, thus, meant the homology of function to him.  

Actually, he made several errors in identifying the homologous areas across species, which demonstrate the subtlety of the relationship between structure and function.  He was later criticized for subdividing human cortex into so numerous areas (~50), based on what appeared to be subtle evidence.  

Nevertheless, it is now believed that there should be more functional areas than Brodmann thought.  Although Brodmann made some errors, what he envisaged at the time was overall correct.  Even after 100 years, his monologue provides deep insight to the organization of the mammalian cerebral cortex.

Below are the photos of three different areas of the monkey neocortex stained by Nissl method (Fig. 3-1).  This is the same method used by Brodmann.  These three areas were chosen because they exhibit very different morphology.  Do you see the similarity or difference among them?

layer structure in various areas

Fig. 4-1  Different areas exhibit different structures.

Nissl staining of three different areas of the monkey neocortex, namely, primary motor area, temporal association area (area TE) and primary visual area.  

You can see the typical 6-lamina pattern for the staining of the "association area"; sandwiched between the cell-poor layer 1 and white matter (wm) are neuron-rich layers 2-6.  Among these layers, layers 2, 4 and 6 are identified as the more cell-dense "granular" layers.  The most characteristic feature of the motor cortex is the lack of granluar layers.  It is also characterized by the presence of particularly large pyramidal cells in layer 5, which are called "giant pyramidal cells of Betz".  In stark contrast, the visual cortex exhibits many layers, with very dense layer 4.  Actually, layer 4 of visual cortex is so well developed that it is considered to consist of multiple sublayers.  

One problem of Brodmann (or whoever) is that any single technique cannot show all the anatomical features that exist.  The use of new techniques has always broadened our view of the brain.  By using ISH technique, we can now visualize what Brodmann could not see.  
In the sections that follow, I am going to introduce some of such data.


5. Area structure of neocortex

Introduction to the work in Yamamori Lab (part 1)

Yamamori Lab has been interested in how the primate neocortex is different from and/or similar to that of rodents.  

Here is a number.  The proportion of the neocortical areas of the mouse, monkey and human brains is approximately 1:100:1000.  The expansion of area means the increase of neurons, which means much much more complicated neural connections.  Can the same organizing principle really account for the formation of both primate and rodent neocortex?

To answer this question, we need to know the primate neocortex better.  We started to investigate the monkey neocortex using molecular biological techniques.  Specifically, we started to look for genes that are differentially expressed in monkey cortical areas.  

"Gene" is DNA and stays in the nucleus of each cell.  It is transcribed into mRNA and translated into protein for function.  That process is called "gene expression".  Each neuron has the same set of genes (which are estimated to be approximately 30,000).  But which genes are expressed and to what extent differ greatly among individual neurons.  Because the expressed gene products determine the properties of each neuron, we reasoned that the genes expressed abundantly in a particular area would reflect the properties of neurons that reside in that area.  

Before starting the project, we had no idea what we will get.  There was even a possibility that we find no such genes.  But the search resulted in the finding of two genes with striking patterns of expression.  

One gene, occ1, was highly abundant in the visual cortex (V1), whereas another gene, RBP, was abundant in the association areas.  Both genes showed characteristic lamina patterns, which differed across areas.  Interestingly, the V1-specific expression of occ1 gene was observed only in the primates.

Our finding demonstrated that the monkey neocortex evolved some area-specific gene regulatory mechanism and implies that such regulation plays an important role in organizing the complex brain of primates.  For more details, see the reviews by Yamamori 2006 and Watakabe 2006 and other references.

ISH images of area-specific genes

Fig. 5-1 ISH of two area-specific genes in the monkey cortex
     See Yamamori and Rockland 2006.


6. Lamina structure of neocortex

Introduction to the work in Yamamori Lab (part 2)

Ever since Brodmann,  the mammalian neocortex has been viewed as having a six-layered bauplan.  Yes, you can clearly see layers.  But for untrained eyes, it's not so easy to identify which layer is which.

The work described below started with the desire to actually "see" the layers with the untrained eyes.  When our lab started to work on necortex, several genes were already reported to be "layer-specific" and the list of such genes were expanding.  Considering the conservation of lamina structure, I expected that these genes would serve as good markers to identify layers across areas and species.  

This belief was partly right and partly wrong.  Fig. 6-1 shows the expression patterns of four genes in the mouse and monkey cortices.  From this figure, it is obvious that the basic patterns of lamina distributions are conserved across species.  

Nevertheless, when their expression patterns were examined across areas and species, it became apparent that the specificity of "lamina" is compromised here and there.  For example, ER81 mRNA is expressed in layers 5 and 6 in the monkey cortex, whereas it is quite specific to layer 5 in mice (see below).

ISH of layer-specific genes

Fig. 6-1 ISH of four layer-specific genes in the mouse and monkey cortex

Five sections were stained for Nissl, and four different genes.  The color panel was synthesized by artificially coloring the ISH patterns of the four genes and overlaying them. ER81, Nurr1 and CTGF genes are most abundant in layers 5, 6 and 6b, respectively, in both mouse and monkey cortex.  You can also see some area and species differences of their expression patterns.

Another example is shown in Fig. 6-2.  This shows the ISH patterns of Semaphorin 3E (Sema 3E) gene in the mouse and monkey cortices.  In the monkey cortex, Sema3E mRNA is highly specific to layer 6 throughout neocortical areas.  In mice, however, Sema3E mRNA is expressed also in the lower half of layer 5 in addition to layer 6.

Fig. 6-2 ISH of Semaphorin 3E (Sema3E) gene in the mouse and monkey cortex

Fig. 6-3 shows a different kind of example, in which the layer-specificity is violated.  This figure shows the double ISH of Nurr1 and ER81 genes in the mouse cortex.  In this area of the mouse cortex, the expression of Nurr1 mRNA is not restricted to layer 6 and reaches even to layer 4.  Such "intrusion" occurs frequently at the lamina borders for other genes as well.  

What is intriguing in this case, however, is that Nurr1 and ER81 mRNAs are not expressed in the same neurons, despite extensive intermingling (Fig. 6-2, lower panels).  

Fig. 6-3 Double ISH of Nurr1 and ER81 genes in the mouse cortex

Nurr1 mRNA is expressed beyond the border between layers 5 and 6, even to layer 4 in the laterocaudal areas.  Thus, ER81 and Nurr1 positive neurons intermingle in layer 5. Nevertheless, these two mRNAs are not expressed in the same cells. The neurons that express these mRNAs belong to different subpopulations.  

At present, we do not completely understand what such expression profiles mean. One thing is evident, however.  With the new technique, we can now see what Brodmann would have loved to see.  

For more details of this lamina story, see Watakabe 2006, 2007.