Neurons and Glia

Topics

• Introduce the two principal cells of the brain
  • Neurons
  • Glia
• Neuron doctrine
• Projection neurons and interneurons
• Cortex
• Parcellation schemes for cortex

Preparation for class

• Reading: PN6 Chapter 1 (pp. 1–10, 13–16)
• Reading: PN6 Chapter 2 (Not Box 2A)
• Reading: PN6 Chapter 4
• Read the notes and watch the two videos for Lecture 3.01 (Historical Overview of brain Structure and Function).

Title Slide

Elizabeth Caroline Crosby (1888–1983) received her Ph.D. in Anatomy in 1915 at the University of Chicago. As a Professor of Anatomy at the University of Michigan, she made important contributions to the study of the comparative anatomy of the vertebrate brain.

Marian Diamond (1926–2017) received her Ph.D. at U.C. Berkeley in 1953. After working at Harvard and Cornell, she returned to Berkeley where she attained the rank of Professor of Anatomy. Diamond made seminal contributions to the study of brain plasticity and, with her colleague Rosenzweig and others, demonstrated that enriched environments in rodents led to changes in brain structure. Diamond was the subject of an award-winning documentary in 2017 entitled “My Love Affair with the Brain”.

Neurons

There are two principal cell types that make up the central nervous system: neurons and glia. Neurons are the primary cells for information processing in the brain. There are different types of neurons and we will discuss them in more detail in the following sections. There are also different types of glia cells and they play many different roles in the brain, from supporting communication among neurons, to regulating the microcirculation of blood, to immune function. We will discuss glia in more detail in a section below.

Many textbooks report that the human brain has ~100 billion neurons. Azevedo and colleagues (2009) performed a systematic count of the number of neurons in the human brain and have provided more precise numbers.

• The average human brain has ~86B neurons (range: ~78B–94B)
• Of those 86B neurons, ~16B are in cerebral cortex. While ~69B are in the cerebellum.
  • It is interesting to note that the cerebellum accounts for ~10% of the brain’s mass, but ~80% of its neurons. As we will learn, the cerebellum has a very small neuron called the granule cell that accounts for the majority of those neurons. Cerebral cortex accounts for ~82% of the brain’s mass, but only ~19% of its neurons.

For context, this same group has reported that the elephant brain has 257B neurons with 251B of those neurons in the cerebellum and 6B in the remainder of the brain. The gorilla brain is reported to have 33B neurons.

Anatomy of a prototypical neuron

The figure below is taken from a textbook that my colleagues and I wrote. The depicted neuron is prototypical. We will discuss different types of neurons in a section below.

The components of a neuron consist of the following:

• Cell body (or soma): like all cells, the neuron has a cell body with a nucleus and the other typical cell machinery and organelles.
• Dendrites: the dendrites are thin protoplasmic processes extending from the soma and branching extensively. The dendrites are the primary processes of the neuron that accept input from the axons of other neurons. We can say that the dendrites are the most important sites for integration of information from other neurons.
• Axon: the axon is a thin protoplasmic process that extends from the axon hillock region of the soma and carries the output of the neuron’s processing to other neurons. It is thus the neuron component involved in the transmission of information. There is only one axon leaving the soma, but it can branch and make connections to many other neurons.
• Axon hillock, or trigger zone: the part of the soma from which the axon projects. It is a critical region in determining whether or not a neuron will ‘fire’ and send information (in the form of an action potential) through the axon to other neurons. It is becoming clear that there may be other ‘trigger zones’ in some neurons; for example, at the base of the large apical dendrite of a pyramidal neuron (see below).
• Myelin: The myelin sheath is not an intrinsic component of the neuron and not all neurons are myelinated. Rather special glia cells (see below) called oligodendrites wrap the axons in layers of a fatty substance called myelin. These myelin sheaths are interrupted by gaps. Action potentials jump from gap to gap in myelinated neurons, speeding transmission.
• Synapse: A synapse (from the Greek, clasp) is the place where the terminus of an axon from the presynaptic neuron is physically opposed to the membrane of the postsynaptic neuron (usually on a dendrite, or spine). Synapses are the main sites for information transmission among neurons.
  • In chemical synapses, neurotransmitters are released from the axon terminus from the presynaptic neuron, cross a small cleft and affect the postsynaptic neuron.
  • In electrical synapses (not shown), the membranes of both presynaptic and postsynaptic neurons physically touch and share connecting channels such that transmission does not involve neurotransmitters.

Types of neurons

Neurons can be categorized in different ways using different criteria such as morphology and gene expression. Much of this is beyond the scope of the course. However, one main distinction we can discuss is between projection neurons and interneurons.

Projection or Principal neurons

Projection neurons (also known as principal neurons) integrate and process information within a brain region and then project the output to a different brain region via a long axon. There are different varieties of principal or projection neurons. For example, the medium spiny neuron is the principal neuron of the striatium, and the Purkinje cell is the principal neuron of the cerebellum. Perhaps the most iconic principal neuron is the pyramidal neuron which represents ~ 70% of all neurons in the cerebral cortex. A good summary of pyramidal neurons can be found in Bekkers, 2011 from which the figure below was modified.

The pyramidal neuron received its name from its pyramid (triangular) shaped cell body. The pyramidal neuron shown in the image above is some layer 5 of somatosensory cortex (the main output layer). The pyramidal neuron has a large apical dendrite with a tuft of dendrites as its most distal location. It also has many branching dendrites near its cell body known as the basal dendrites. Although not visible in the image above, the dendrites of pyramidal neurons are covered with small protruburances called spines which are sites of synaptic activity and plasticity that we will discuss in detail in our next two lectures.

There is one axon leaving the pyramidal cell’s soma (indicated by a red asterisk). In this image, the axon is cut. However, the axon may travel to distant subcortical targets. In the case of large pyramidal cells in motor cortex, the axon may travel down the spinal cord. In the giraffe, the longest axon is reported to be 15 feet in length.

Pyramidal neurons are excitatory, meaning that the synapses formed from the terminal axon branches release glutamate, the primary excitatory neurotransmitter in the brain. This depolarizes, or excites, the post-synaptic neuron making it easier to become activated. We will discuss excitation and inhibition and neurotransmitters in detail in the next lecture.

All projection neurons project output through long axons. However, we cannot generalize the specific characteristics of pyramidal neurons to other projection neurons. For example, the medium spiny neurons (MSNs) of the striatum (basal ganglia) release GABA and inhibit their axonal targets. As their name implies, MSNs also have spines. Purkinje neurons in the cerebellum also are inhibitory upon their targets and use GABA.

Interneurons

The second most common neuron in cortex is the interneuron which makes up ~20–30% of cortical neurons. In cortex, most interneurons are inhibitory and release GABA as its neurotransmitter. Inhibitory interneurons participate in local circuits that modulate the activity of the pyramidal project neurons by making synaptic contact with pyramidal neurons and other interneurons. In cortex, interneurons are smooth, meaning that they do not have spines like the pyramidal neurons. Interneurons within cortex can also connect to other interneurons via electrical synapses, which we will discuss in more detail below and in our next lecture.

Simple neural circuits

Neurons are connected to each other and thus create circuits that can provide a biological substrate for computations. The existence of both excitatory and inhibitory neurons provides a robust toolkit for neural computation. Indeed, if all neurons were excitatory, computation would be limited. This website at UTHSCH has many interesting animations of basic neural circuits made of excitatory and inhibitory neural elements. The figure below is taken from that website an illustrates how simple combinations of inhibitory and excitatory neurons can create a variety of computational motifs. We will encounter some of these motifs later in the semester.

Before leaving this topic, let me emphasize that while neurons can inhibit or excite their target neurons, the ultimate effect of the circuit depends upon the activity of those downstream targets. In example B (feedforward inhibition), an excitatory neuron excites an inhibitory neuron which then inhibits an excitatory neuron. Thus, the end result is inhibition of the activity of that final excitatory neuron. Consider this example (not shown) where an inhibitory neuron inhibits a second inhibitory neuron that is inhibiting an excitatory neuron. The inhibition of inhibition (referred to as disinhibition) releases that final excitatory neuron from inhibition. My point is not to make your head spin, but for you to appreciate how combinations of excitatory and inhibitory neurons can provide a powerful tools for complex neural circuits.

Cajal, Golgi and the Neuron Doctrine

The image above is a camera lucida drawings by the famous Spanish anatomist Santiago Ramon y Cajal showing several neurons in cerebral cortex. Cajal used a silver nitrate staining technique developed by Camillo Golgi in 1873. Golgi’s stain is idiosyncratic in that only a relative few neurons are stained, but these neurons are stained in their entirety revealing their dendrites and axons. This contrasts to other techniques, such as the Nissl stain, which stains all cell bodies, but not their dendrites or axons.

Although bitter rivals, both Cajal and Golgi won the Nobel Prize for Medicine in 1906 for their independent work on the nervous system. However, their perspectives were quite different, and Golgi used his Nobel address to rebuke of the work of Cajal. Golgi had a holistic view of the nervous system, and thought that all neurons were connected in a continuous reticulum or syncytium and rejected the idea of cerebral localization of function. Cajal believe that neurons were discrete cellular units and championed the “Neuron Doctrine”, an idea compatible with the localizationists and with the stimulation studies of Fritch and Hitzig, and David Ferrier .

The Neuron Doctrine and Functional Polarity

The “Neuron Doctrine” was formally proposed in 1891 by Heinrich Wilhelm Gottfried von Waldeyer-Hartz in a review of the literature that relied heavily upon Cajal’s observations. The four tenets of the neuron doctrine are as follows:

1. The fundamental structural and functional unit of the nervous system is the neuron.
2. Neurons are discrete cells which are not continuous with other cells.
3. The neuron is composed of three parts – the dendrites, axon and cell body.
4. Information flows along the neuron in one direction (from the dendrites to the axon, via the cell body).

One of the most important contributions of the Neuron Doctrine is that is assigned a functional polarity to neurons (input to dendrites, output to axons). This allowed for a functional analysis of brain regions and neuronal circuits.

While the general tenets of the neuron doctrine have stood up reasonably well over time, there are numerous examples that violate the neuron doctrine. One such example is the electric synapse, in which neurons share a membrane where excitation can flow bidirectionally. Reverse (or retrograde) transmission has been detected within neurons where the postsynaptic neuron sends chemical messages back to the presynaptic neurons. This process plays an important roles in learning that we will discuss in the lecture on plasticity. Finally, it is becoming clear that dendritic branches may themselves be sites of information processing that does not have an effect upon the cell body or generate signals carried by axons.

It is interesting to consider the Neuron Doctrine as an extension of the Cell Doctrine, first formalized in 1839 by Schleiden and Schwann.

Glia

Glia are the second principal cell of the brain. Glia means ‘glue’, and it was once thought that glia’s role in the CNS was minor, and that glia just held neurons together. However, it has become increasingly clear that glia play an important role in the brain, and may also be important in neurodegenerative diseases such as Alzheimer’s and Parkinson’s. There are several subtypes of glia, of which two are important to our interests.

• Astrocytes: There are about as many astrocytes as neurons in the brain (although the ratio differs in cortex and the cerebellum). Astrocytes support and influence synapses between neurons, modulate the extracellular environment around synapses, and help control local blood flow. One of the interesting aspects of astrocytes is that they connect to each other through gap junctions in a kind of syncytium so that molecular signaling in one astrocyte can spread rapidly to other connected astrocytes. This idea of how astrocytes are connected is similar to Golgi’s view of how neurons were connected (see section below on the conflict between Cajal and Golgi).
• Oligodendrocytes (CNS), Schwann cells (PNS) provide the myelin sheaths that speed communication in the axons of neurons. The disease Multiple Sclerosis attacks the myelin sheaths, and disrupts neural transmission.

The other types of glia are as follows:

• Oligodendrocyte precursor cells (OPCs) (5–8% of glia)
• Ependymal cells (microvilli)
• Microglia (immune system)

Research has recently suggested that microglia are involved in stripping synapses and may play a functional role in forgetting memories. Please see my blog entry that describes a recent paper on this topic.

Organization of neurons

We can simplify our picture of the central nervous system as consisting of neurons (gray matter) and their axons (white matter). The axons of interneurons are short as interneurons participate in local circuits. The axons of projection neurons are long, as they communicate information from one brain area to another. The axons of projection neurons are organized into bundles that are called white matter tracts, and are discussed in more detail in Lecture 2.01. The take-away point about white matter for our present purpose is simply that white matter tracts are organized. We all have the same white matter fiber bundles (although there is variation as in all things biological).

But what of the organization of neurons? Neurons participate in circuits with each other, and so group into clusters. These neuronal clusters accept input and produce output, so they are visibly set off from other clusters by the white matter that carries the input and output. There may also be mophological differences that characterize and differentiate these clusters based upon neuronal type and density and pigmentation. For example, the substantia nigra is a cluster of dopamine producing neurons in the midbrain tegmentum that is notable for its dark pigmentation due to the accumulation of melanin. Recently, it has become possible to differentiate neurons based upon the genes they express, and so additional factors can be used to differentiate neuronal clusters.

Laminated and non-laminated neural structures

We can discern two general organization principles for neurons: laminated and non-laminated. Lamination refers to an orderly layering in which different neuron types are stratified into different layers. Non-laminated structures do not exhibit that layered organization.

The image below is from a paper by Michelsen and colleagues (2007) in which a dorsal raphe nucleus in the brainstem has been stained with a Golgi silver stain. The dorsal raphe nucleus is a source of serotonin (5HT) in the brain, but the details don’t matter for our example. What should be clear is that the neurons are not organized into layers but are seemingly haphazardly distributed. This form of organization is typical of nuclei in the hindbrain.

The terminology for this type of non-laminated neural structure can be arbitrary and confusing. Clusters of neurons in the enteric and peripheral nervous systems are generally referred to as ganglia. Ganglia do not have a laminated structure. Non-laminated clusters of neurons within the central nervous system are generally referred to as nuclei. If one refers to a nucleus or nuclear group in the brain, it is usually safe to assume it is not laminated. However, some structures in the CNS labeled as nuclei do have a laminated structure — for example, the lateral geniculate nucleus that we will discuss in the visual system. That is where confusion arises.

The stratification and organization of laminated neuronal structures is obvious in the image below drawn by Cajal.

The leftmost two panels are Nissl stain images of visual (left) and motor (right) cortex from human brains. The Nissl preparation stains neuronal cell bodies, and so the layered stratification is obvious in the densities of neurons and the sizes of the cell bodies. These cortical regions are neocortical, and thus have six layers (with subdivisions) as indicated on the far left in Roman numerals. The righmost panel is a Golgi stain which fills the protoplasmic processes of individual neurons and provides information about dendrites and axons. Other staining techniques (not shown) emphasize stratified white matter distribution.

Cortex

Neurons that populate cortex are divided into layers that can be readily observed under a microscope in stained tissue samples as shown above. There are different numbers of cortical layers in different forms of cortex.

• Allocortex is a less developed and evolutionary older form of cortex. Allocortex has been divided into two types.
  • Archicortex (3 layers). In humans, Archicortex is found in the hippocampus
  • Paleocortex (4–5 layers). In humans, Paleocortex is found in the olfactory cortex, in parahippocampal cortex, and in cingulate cortex.
• Neocortex is found in all mammals and is found most of cerebral cortex. Neocortex (’new cortex’) consists of 6 layers (and some layers have been further subdivided). These layers are specialized for input, output, and intrinsic processing, and consist of different kinds of neurons that may be packed in differing densities.

Origin or cortical layers

As mentioned above, six-layered cortex (neocortex) is thought to be an evolutionarily new development restricted to mammals. In a recent paper, Harvey Karten has challenged as least part of this orthodoxy by pointing out that the neural circuits observed in neocortex can also be found in birds.

The prevailing and recurrent theme that emerged was that the microcircuits of the mammalian thalamus and cortex had evolved long prior to and independently of the developmental processes that led to the more notable cortical lamination in mammals. We suggested that this radial recurrent type of microcircuit, so characteristic of mammalian cortex, was probably common to all amniotes, and represented an antiquity of more than 200 million years.

Karten’s point isn’t that six-layered cortex is found in birds, it is not. However, the circuits represented in six-layered mammalian cortex can be found distributed throughout the avian brain.

Neocortical layers

The six layers of neocortex are organized similarly (with variation) in all neocortical regions, that is, in most of the cerebral cortex. We will have more to say about the particulars of this organization when we discuss sensory systems. However, now is a good time to introduce the topic.

The following figure is taken from a brief but excellent summary of cerebral cortex by Stewart Shipp (2007). The figure illustrates the main inputs in terms of sources and target layers; the main outputs from the different target layers; and the main internal connectivity. For example, we see that the main input to cerebral cortex from sensory thalamic nuclei terminate in layer 4; while the main output to subcortical structures originates in layer 6.

Vertical organization

The figure above from Shipp, 2007, uses color to imply a vertical organization within neocortex. That is, information arrives in different layers from different sources, ascends and descends within a column, and then information if projected from that column. Thus, a neocortical column can be conceptualized as an information processing unit. We will discuss the columnar organization of cortex in more detail when we discuss primary visual cortex.

Before we leave this topic, however, it is notable that pyramidal neuron’s apical and basal dendrites can extend vertically within all layers of cortex. Thus, in the schematic figure below, we see that the soma of a pyramidal neuron receiving sensory input may be located in layer 4, but the apical dendrite extends to layer 1 where it branches into a dendritic tuft that can be 1 mm in diameter. The basal dendrites ramify in layers 5 and 6. Thus, the shape of the pyramidal cell is ideal for integrating information within the cortical column.

Parceling cortex

Cerebral cortex (cortex is latin for ‘bark’) is formed of a continuous sheet 2–4 mm thick and 1.3 square feet (per hemisphere), but that was folded so that a larger surface area of cortex could be fit inside the skull.

The cortex is subdivided into lobes of which there are four: frontal, parietal, temporal, and occipital. Two other cortical regions are sometimes described as lobes: (1) The insula, which is hidden from view by the overhanging frontal and temporal lobes. (2) The limbic lobe, which refers to a ring of cortex along the medial surface of the hemispheres.

The pattern of folding of cerebral cortex is reasonably similar across individual humans (i.e., the major folds are usually recognizable, but there can be considerable variability). The in-folded parts are called sulci, while the exposed surfaces are called gyri. The major sulci and gyri have names (e.g., the middle temporal gyrus is a major gyrus in the temporal lobe. It is demarcated from the more dorsal superior temporal gyrus by a sulcus known as the superior temporal sulcus).

Cytoarchitecture

Histology refers to the microscopic study of bodily tissues. Histologists noticed that the architecture of cells (cytoarchitecture or cytoarchitectonics) that make up the cortical layers were different in different brain regions. Brodmann (1868–1918) was a histologist who studied variation in cell types and cortical layers in the cortical sheet, and gave them numbers that are still used today. The implication is that a change in structure (i.e., Brodmann number) reflects a change in function. For example, Brodmann’s Area 4 (BA 4) is known as primary motor cortex. It differs remarkably in cytoarchitecture from BA 17, which is primary visual cortex. Although visual and motor cortex are both six-layered, the thickness and cell types within each layer can differ. This is obvious in Cajal’s drawings of cerebral cortex in the preceding section when comparing the leftmost (visual) and the middle (motor) panels.

Brodmann’s parceling scheme is not the only such parceling scheme, and we will encounter others. However, his nomenclature is still widely used today, particularly in neuroimaging studies.

White matter tracts

Neurons give rise to protoplasmic processes called axons that communicate with other neurons in the brain, sometimes over long distances. It is this communication among neurons that is the basis for information processing in the brain. If we think of the axons as wires, we can think of these wires running together in distinct bundles, connecting different parts of the brain. These bundles are called tracts or fiber tracts.

We will encounter white matter tracts throughout the semester, and a more detailed overview can be found in Lecture 02.02.

Prerecorded videos for Fall, 2020

Neurons (59:16)

~~

Glia (8:57)

~ ~