Neuro Basics: Part 1

As a preamble to some more discussion about artificial minds, I thought it would be good to review some very basic neuroanatomy and physiology. I think a lot of the misinformation about minds comes from ignorance of these basics.

Below, is a simple approximation of a neuron - the basic building block of the Central Nervous System (CNS). Neurons are found in the brain, spinal cord, ganglia, etc.

The neuron has a cell body (soma) and finger-like projections emanating from the body. Invariably, there can be many tendrils called dendrites, but only one tendril called an axon. The dendrites and other receptors along the soma, are part of the synaptic connections to the axon end terminals of many other neurons. Dendrites are one of the ways a neuron can receive a chemical signal from another cell. The axon carries the signal of the neuron to the dendrites of other neurons, or to regulatory connections with other cell types like muscle cells. In other words, the dendrites are the input jacks to the neuron and the axon is the output.

The single axon can have more than one end terminal complex allowing the single axon to interact with more than one other cell.

The axon is an amazing structure. A single axon may span from a ganglion in the spinal cord, to a motor end plate in the foot - a distance of more than a meter!

Neurons are supported in the CNS by specialized cells - the glia (glue). You may have heard of brain tumors called gliomas, glioblastomas, or astrocytomas. All originate from the support cells of the CNS. The ratio of glia to neurons is around 3 to 2 in the CNS. Importantly, there are no connective tissue cells in the CNS like the collagen producing cells everywhere else in the body. The advantage of course is that there are no extraneous cells taking up valuable real estate in the CNS. The disadvantage is that the neurons and their connections are fragile and susceptible to ‘shear injury’ because there is no elastic tissue in the CNS to prevent them from being displaced by changes in acceleration - like falling on concrete for example.

Glial cells aid in metabolic support of the extended axons and a subtype, the Schwann cells, form an insulating lining around the axon called the myelin sheath.

Myelin is an insulator which aids in the efficient transmission of electrical impulses along the axon. You may have heard of ‘demyelinating disease’ (like MS). Loss or damage to this sheath can interrupt the action potentials of neurons interfering with normal function. What’s an action potential? That’s the next part.

A neuron is regulated by connections to many other neurons through the dendrite and soma connections (synapses). Greatly simplifying, there are two types of actions going on in a neuron: neurotransmitter based, and electrochemical signal propagation. The inputs to the neuron are neurotransmitter based. Other neurons fire and release neurotransmitters that affect the target neuron. Not all receptors need to be stimulated at once (usually a good thing). What’s more, not all neurotransmitter signals are positive, ie stimulating. Some are negative or suppressive signals telling the neuron to ‘stand down’.

This is similar to a lot of the transmitter/receptor negative feedback loops found throughout the body. The sum total of positives with respect to negatives released by all the interlaced neurons determines whether the neuron fires or not. There are no half-way or partial firings. Neuron are said to behave in an ‘all or none’ manner. They are either on or they are off. Nothing in between. Neurons are therefore binary systems. The complex interplay of dendrite connections can be quite extensive, but in the end it’s either on or off. The subtlety comes from the magnitude of the interacting binary processes and the various transmitters they release. This is important to the discussion of minds. As an aside, the actions of many drugs can be explained in this way. In essence, they may alter the normal stimulus/feedback mechanisms allowing aberrant stimulation of normally inactive pathways.

If the thing fires, it does it at one output level. Any subtle effects it might have are due to the interplay of many such neurons, some firing, some not.

If the neuron fires, an electrochemical reaction propagates along the axon causing depolarization of the axon membranes and the generation of an electric action potential. This is how the chemical messages triggered by dendritic stimulation exceeding the action threshold, are relayed the great distance to the axonal end terminals.

The net effect of this propagating action potential is felt at the end terminus of the axon, which sits in a terminus / receptor complex called a synapse. The axon terminus is separated from an adjacent receptor on another cell, by a short space called the synaptic cleft. What happens at this point is illustrated by the next diagram of a specific example of a synapse - a motor end plate where a neuron connects to a muscle cell to regulate its contraction.

The action potential reaches the nerve terminus and causes a neurotransmitter to be released into the synaptic cleft (in this case, acetylcholine). Reversible binding of the acetylcholine to its receptor on the target muscle cell triggers it to respond. This is a short lived effect because the binding is reversible and the acetylcholine is quickly metabolized with the aid of enzymes in the synapse. The metabolized products are reabsorbed by the nerve terminus to be reformed into acetylcholine neurotransmitter molecules for use the next time the neuron fires.
As a complete aside, this is the process attacked by nerve agents.

The nerve agent binds (sometimes irreversibly!) to the esterase that breaks down the acetylcholine. The result is more acetylcholine sticks around and continuously stimulates the muscle cell, and any other cell type using this type of neurotransmitter. This effect isn’t always an agent of evil, however. Medications are used in certain diseases of the motor end plate such as Myasthenia Gravis, to make up for the diseased reduction in neurotransmitter by allowing that acetylcholine which is present, to persist longer and stimulate the muscle more effectively.

Back to the point at hand. In summary:
  • The CNS is composed of neurons and their attendant support cells, the glia.
  • Neurons are linked together in huge interlocking networks.
  • Neurons receive biochemical prompts from other neurons (and hormones too, but let’s ignore that for the moment, since the basic physiological responses are the same.)
  • These prompts can be to stimulate the neuron or to suppress its response.
  • These neurochemicals are essentially the only way a neuron receives a stimulus.
  • Once a neuron is stimulated beyond a threshold to activate, it transmits this activation using an electrochemical process that extends this activation to the business end of the neuron, the end terminus of the axon.
  • The end terminus of the axon releases its own neurotransmitters starting the process in another cell.
This is very rudimentary explanation but the basic biochemistry of all this is pretty well understood. This includes the actions of neurotransmitters and the ion pumps and channels that drive the action potential. Micro probes can measure this electrical activity around a single point on one axon. The actions of most medications and drugs can be explained by effects upon these processes and replicated. Interfering with any point in this process can create repeatable results. Though the brain links these basic building blocks together in many amazing ways, this is the heart of it. It's easy to see how some equate the binary nature of the all or none response of a neuron to the world of digital computing. There are similarities, and there are huge differences, we'll discuss later. But it's not completely crazy.

And there is absolutely no current evidence that anything else is going on in the CNS that isn’t derived from this basic unit and its biochemistry. There is no evidence that anything coexists with the neurons inside our skulls. Whatever we are, it comes from this. What ever we might think is going on, must account for the fact that nothing, not even a hint, of anything else, has ever been detected. The dream we each have of consciousness, is found not in mysticism, but somewhere in this biochemistry.

I’ll come back to that when we talk more about AI.


Jared said...

Professor Pliny, when do we get to the fun bits like the role of gene regulation in memory formation and the possible link between failures of information parsing systems and conditions such as schizophrenia and autism?

This is actually a remarkably concise overview of neurons. I like it.

Pliny-the-in-Between said...

Professor Pliny? Giving me a hard time again? Thank you, but as you can see, we have some basics to get the group through ;) I'd personally like to see your take on the gene regulation.

Jared said...

Eh, nothing you can't handle...

As far as the gene regulation parts go, I can hardly scratch the surface; there was tons more research to be done to that effect last I read. For Long Term Memory, I know CREBs and HDACs (and others) play a role, as do histone modifications and other forms of DNA modifications, but as far as the exact roles (specific mechanisms) played by these various pathways, I would have to do a bit of reading. My favorite experiment on the role of genes in memory formation involves Drosophila (of course) and temperature sensitive mutations that had flies that couldn't recall memories when one gene was deactivated and couldn't form memories when another gene was deactivated. I am reminded of a quote by John B. Watson, "Most of the psychologists talk, too, quite volubly about the formation of new pathways in the brain, as though there were a group of tiny servants of Vulcan there who run through the nervous system with hammer and chisel digging new trenches and deepening old ones." Guess the joke's on him...

The skinny of it from what I understand: the stronger or more frequent the stimulation of a neuron, the more changes in the expression of a number of genes which influence synaptic and neuronal properties. I'd have to reread some of the papers I have on my computer, because it's been about 5 years or so since I've really read anything on the subject.

Unknown said...

Love the images. Are they open source? Can I use them in my blog (with attribution?)


Richard Friesen

Pliny-the-in-Between said...

Richard, feel free to use my images with references.