Connectome by Sebastian Seung
Introduction
Sebastian Seung introduces brings us to the realization that we still have a lot to learn about the brain, because of its’ massive complexity. He also introduces us to C. elegans, a type of worm. The worm has 300 neurons, all of which have been mapped and given individual names. However, despite us having the connectome of the C. elegans, having this information has not told us much about the worm’s behavior. Some believe it is because the neurons of this worm are not confined to one area also, some neurons do not have synapses.
The connectome is a sum total of all of the connections, neuron to neuron, in the brain. Dr. Seung believes that having this neuronal map will help explain our behavior and who we are. He talks about the importance of the 4 R’s, connectome: reconnections, rewiring, regeneration, and reweighting, ad how they hugely impact our connectome. He uses the rest of the book to convince us of this hypothesis.
My Opinion: To say that the connectome explains our total being is interesting. I do believe that it is necessary for us to find the connections of neurons in the brain, however, once you do this, all you are doing is finding the connections of a whopping 15% of the brain. What about the other 85%, which are glia. I believe that glial connections also feed into this hypothesis, as disturbances in glial cell connections can cause the destruction of some neuronal connections, or even, can cause extreme abnormalities.
Part I: Does Size Matter
Dr. Sebastian eases us into the connectome concept by first giving us a brief summary of the brain as a whole, then begins to break it up into parts. He starts with a look into history and who people have come to identify different areas of the brain.
The brain has 3 main parts, the cerebrum, the cerebellum, and the brain stem. First, the brain was characterized by phrenological maps. Then we began to learn more about the brain areas based on patient brain damage studies. For example, based on epileptic seizure locations, we neuroscientists have been able to understand the functions of the temporal lobe. With stroke, victims may lose the ability to speak, but can still comprehend. Otherwise known as Broca’s Aphasia, has to do with damage to Broca’s area, which is in the frontal cortex in the left hemisphere.
Brain Anatomy through pictures here {brought to you by <span class="s2">pixabay.com} {Dr. Penfield’s brain mapping}
We are also introduced to the idea of brain size. Scientists use to believe that having a bigger brain correlated to being smarter. However, measuring intelligence of a size scale is not sufficient for determining mental capacity. Dr. Seung argues that it is a lot more deeper than that.
Part II: Connectionism
“No neuron is an island. Neurons are polyamorous [40].” This statements alludes to the fact that no neural response relies solely on a signal neuron, but rather the connections of that neuron. If you look under a microscope you will see that a single axon can be responsible for many synapse. Because the brain is densely packed, many neurons touch many neurons. Neurons talk to each other through chemical - electrical responses via either releasing or collecting neurotransmitters. Neurotransmitters are the chemicals used in neuron to neuron communication. Neurons release neurotransmitters into the synaptic cleft through their axons. Electrical signals, called action potentials, are sent down the axon terminal in order for it to release chemical messengers, neurotransmitters, to the next cell via dendrites.
Dr. Seung explains neural connections by using the example of the snake. He explains what invokes the “fight or flight” response explaining how this is the result of neurons being close to each other and how signals propagate to each other, along a specific pathway, in order to induce a physical action or lack thereof. The action or lack of action really depends on whether an excitatory or inhibitory neurotransmitter is being sent. Nerves that are at the surface of the brain or brain stem or part of the central nervous system, CNS, while the neurons that extend to the periphery or surface of the body, are part of the peripheral nervous system, or PNS [51].
In the fourth chapter, we are introduced to the Jennifer Aniston neuron. Dr. Itzahk Fried, as Dr. Penfield, operated on people with epilepsy and used electrodes, before surgery, to map out the brain [61]. one day he flashed a picture in front of one of his patients of Jennifer Aniston, and one signal neuron lit up, or spiked. It did not light up for any other picture, but Jennifer Aniston’s. Dr. Seung believes that the neuron could be reacting in response to other responses. He offers an explanation, by saying to break the picture of Jennifer Aniston into several parts, meaning, the blue eyes, the hair color, etc. He suggests that each feature may have a specific neuron that responds to it, and if you trigger all of them, then you can get a specific neuron to spike. He then responds, as his title suggests, “it’s neurons all the way down”, arguing that the spiking of the Jennifer Aniston neuron is in response to the connections it has with other neurons that are encoded for different features.
Dr. Sebastian Seung believes that given a connectome, we can map the first kiss. He goes into a discussion about memory and what shapes it. Synapse strength can be strengthened by continual activation, which alludes to one of the 4 R’s, reweighting. Synaptic strength can either strengthen or weaken depending on different factors that include activation and connections. It can become stronger by “getting bigger, releasing more neurotransmitter, or becoming more sensitive to neurotransmitter. Synapses can also be created and destroyed [78]”. The later refers to the second R word, reconnection. These phenomena occur throughout our entire lives.
He then talks about Hebb’s rule, which was first introduced by Donald Hebb. Hebb’s rule or postulate says that if neurons are repeatedly activated in a sequence, then the strength of the synapse as well as its’ connections strengthens. This was further summarized by Siegrid Lowel, “neurons that fire together, wire together”.
Part III: Nature and Nurture
Dr. Sebastian Seung brings up the Nature vs Nurture debate. If we inherit our physical genes, then why not our behaviors as well, or even our IQ? As he mentions, “the psychologist, Eric Turkheimerahs promulgated the First law of Behavior Genetics, saying, ‘All human behavioral traits are heritable’[101]”. Studies to prove this had been done in the study of monozygotic twins.
One of the ways we learn about brain function is through our examination of people with brain damage or neurological disorders. Dr. Seung talks about how the brain is created, after the sperm has penetrated the egg. He gives us the four stages of neuronal development.
Neurons are created through the division of progenitor cells
Cells migrate to their proper places
Extend branches
Make connections
Then he teaches us about disorders that can arise if there is a disturbance in any of these stages. The first disorder he talks about is congenital microcephaly, which is when a child is born with an small head, due to an abnormally small brain. This is due to defects in the first stage of neuronal development. It has to do with a defect in proteins used to help the creation of neurons.
Lissencephaly occurs due to a disruption in the second stage of neuronal development, migration of cells to their proper places. Lissencephaly is a brain formation disorder characterized by a smooth brain, a brain lacking many ridges.
Disruptions in the third area, branch extension, refers to defective axonal growth. As Sebastian says, if axons don’t grow properly, “miswiring” can happen. An example he gives is the corpus callosum, in rare individuals, is either partially missing or completely missing.
As far as making connections, as neurons or created, and branches are extended, neurons begin to form connections through synapses. According to Dr. Seung, synapses are created at a staggering rate in infant brains [107]. However, the brain may make a lot of synapses initially, but eventually a lot of those synapses are destroyed later. Sebastian compares this phenomenon like a rough draft. You write everything down first, only to go back and refine what you wrote. He also says that “the theory that synapse elimination, in the adult brain, is driven by weakening, which is driven by experience”, likewise, he believes that this may also be “the main driver in elimination in the developing brain [109]”.
He also offers an explanation of Autism and Schizophrenia, saying that these disorders may be a result of disruptions in the developing brain. The reason he brings up Autism, especially Schizophrenia, is because they could actually be caused by a connectopathy, or faulty neural connections. many people have tried to find a genetic marker for these diseases, such as was done with Huntington’s Disease. With Huntington’s, one gene was found to be faulty. This allows scientist to be able to diagnose the disease with precision very early. In the case of Autism and Schizophrenia, many genes can be affected. This could be a reason why studying the neuronal connections could lend some help in early diagnostics, and eventually cures.
Dr. Sebastian Seung ends this section with the discussion of neuroplasticity in children and adults and the possibilities of learning in the adult brain. First we learn about the concept of Connectome Determinism, meaning that “denies the possibility of significant personal change after childhood [117]”. With this comes the debate over “reconnection denial” versus “rewiring denial”. Sebastian asks the question if “the brain could rewire, then it should be able to recover from injury or disease, however, if the brain does not rewire, then the adult brain is fixed [120]”.
He addresses these hypotheses in the study of Genie, (a pseudonym for a feral child). Genie was a feral child, who due to childhood abuse, lived in isolation, tied to her potty, in her room, for the first 13 years of her life. When she was found, she could not talk, nor did she have any control of her bowel movements or urinary excretions. Dr. Susan Curtiss began working with her to see if genie could learn to speak after missing the “critical period” of natural language development, (between 2 to puberty), for learning, your first language. If she could this would mean that her brain was able to form new connections and still learn language beyond the “critical” linguistic period.
Interesting 2003 Interview with Dr. Susan Curtiss about Genie: <a href="https://www.youtube.com/watch?v=zvu0MA2CF2A"><span class="s4">https://www.youtube.com/watch?v=zvu0MA2CF2A</a>
Eventually funding was lost, and Genie regressed to her old self.
Another case of a feral child was Victor who was discovered in 1800. He was found in the woods in France. Attempts were made to assimilate him into civilization. Unfortunately, he was not able to learn any linguistic or social behaviors after his initial capture.
Dr. Seung offers many examples that challenge the zero to three movement and whether reconnection or rewiring can happen, such as in vision, in adult brains.
Vision Study: Antonella Antonini and Michael Stryker on rewiring of axons in the visual cortex.
Debate: Dr. Pasko Rakic of Yale University and Elizabeth Gould of Princeton University
This debate is between whether new neurons can be added to the mammalian brain or not. Dr. Rakic argued that it was not possible, while Dr. Gould argued that it was possible. However, studies of neocortical neuronal development have been reported.
Sebastian then argues that we will not have a certain answer unless we study connectors. He says “that’s why connectomics will be important for figuring out whether and how regeneration serves learning [131]”.
Part IV: Connectomics
First we take a trip down tool lane. In “Seeing Is Believing”, Dr. Seung makes the argument that the advancement of technology is what fuels biological discovery. First we learn about Lawrence Bragg, the inventor of the X-Ray. We learned that with this piece of technology people were able to save more lives during WWI and it became a new tool for biological exploration.
My Opinion: I really agree with Dr. Seung on this point. Throughout history, from the discovery of staining by Golgi, to the discovery of the microscope, scientists have been able to see more and learn more. Before staining or the microscope, anatomists and other scientists thought of the brain as a reticulum, giving it the name the neural reticulum, meaning it was one mass. After Golgi accidentally discovered staining, we were introduced to the structure, the cell body, even though this just solidified is belief in the neural reticulum. This, however, is because he could not see the separations of the axons and dendrites, however, Ramon y Cajal did. using a microscope, he was able to see that each neuron was a separate entity, but that there were junctions in between each neuron.
With the advancement of technology comes a look closer into the biology of different structures. This is why I believe labs such as Dr. Karl Deisseroth and Dr. Ed Boyden’s are important. We need scientist dedicated to inventing new tools for neuroscience research. Karl Deisseroth’s lab, at Stanford, helped create Optogenetics. Dr. Ed Boyden and Dr. Feng Zhang were both students in that lab at the time (Both now at MIT). Dr. Boyden’s lab has now invented a new technique called Expansion Microscopy, where you apply a polymer is applied to a piece of tissue through which when water is added, it expands. Dr. Zhang has had a huge role in the creation of CRISPR, which stands for clustered regularly interspaced short palindromic repeats.
TED: <a href="https://www.ted.com/speakers/ed_boyden">Dr. Ed Boyden </a>
YouTube Talk: <a href="https://www.youtube.com/watch?v=yT5JeLVUgdo">Dr. Feng Zhang</a>
Back to the book…
Dr. Seung also talks about slicing. Each of the tissues has to be sliced the correct way with the correct dimensions in order to be imaged. Before a light microscope would not be able to pick up some of the structures, but with the invention of two photon microscopy, by Dr. Winfried Denk, and electron microscopy the resolution is even greater. We are introduced to Dr. Denk who also invented an automated system with an ultramicrotome inside of the vacuum chamber of the electron microscope called SBFSEM. SBFSEM stands for serial block face scanning electron microscopy. At Harvard, the ATUM was “first prototyped”, which stands for automated tape-collecting ultramicrotome. This was actually invented by Dr. Ken Hayworth in his garage.
Next, in “Following the Trail”, Dr. Seung takes us further into the necessary technological advances that are necessary for mapping the connectome. He first takes us through the works of Dr. Brenner, Dr. Leeuwenhoek, Dr. Nichol Thomson, and others and their struggles in finding the connectome of the C. elegans. The easy part was getting the images, the hard part is actually mapping the connectomes. We learn that only one connectome has been mapped from the C. elegans and it’s the ability to swim away from a touch. Finally we are introduced to the problem of computers. Computers today are good, but not good enough. We need technology that is more computationally heavy and that can handle computer vision problems. This is when we are introduced into Machine Learning and some of the work that is being done in Dr. Seung’s lab. \r\n We are also introduced the the concepts of AI and IA. AI stands for artificial intelligence, while IA stands for Intelligence Amplification. Dr. Marvin Minsky is a researched of AI, while Dr. Doug Engelbart is a researcher and creator of IA. Dr. Minsky wants to make machines smarter, Dr. Engelbart was machines to help make people smarter, and Dr. Seung wants a combination of both.
Dr. Sebastian Seung’s <a href="http://eyewire.org/explore">EyeWire Link</a>
Fun Fact: in 2005 Dr. Olaf Sporns and his colleagues coined the term “Connectome”
Dr. Seung continues by giving different hypotheses on how we could and should study connectomes. He offers a way of studying “synaptic chains” in the HVC Connectome and “cell assemblies in CA3”. He takes us through this journey by giving an introduction into memory. Let’s take start by looking at human memory regions. That is to large, so let’s focus on the Medial Temporal Lobe (MTL). However, lets then narrow that down to the hippocampus, and further narrowing it down to CA3, a region of the hippocampus. However, as Dr. Seung said, CA3 is still too large of a region to find a connectome, so studying small pieces may be more sufficient. He then gives an example of the zebra finch bird and its memory of a song. It learns this song when it is young, and as an adult recites the same song over and over without any deviation. this is where HVC comes to play. he offers if we could find the connectome for this song in the bird’s memory, then we would be at an advantage. He then says we may be able to find the connectome for it, however, this does not mean we would be able to make this bird sing. \r\n When it comes to study human connectomes, besides size, there is a variation in the approach. Are we studying regional connectomes or neuronal connectomes? Dr. Seung believes in neuronal connectomics, although he says “it is about means rather than ends”. \r\n Next, we are introduced to techniques such as dMRI that is very low resolution, but could be sufficient for mapping out connections across white matter. White matter only contains axons, while gray matter contains everything else from some axons, neurons, etc. \r\n Lastly, Dr. Seung ends Part IV by offering the hypothesis that understanding the connectome can help us find drugs for developmental disorders, or can help us find gene therapies that “could prevent connectopathies [221]”. He talks about creating therapies that rely on converting undifferentiated cells into neurons, which can be done. However, once injected into the brain it would “require enhancing the integration of new neurons into the connectome, promoting rewiring, reconnection, and reweighting [222]”. Currently most of the cells die once they are in the host’s brain. If we now connectomes we could offer ways of enhancing healthy brains such as with a connect based nootropic. He believes that in order to correct abnormalities, a connectome must first be established.
Part V: Beyond Humanity
Dr. Seung brings up two areas: Cryonics and Uploading. Cryonics is freezing the body so that one day in the future, scientist can unfreeze you and you will live amongst the future society. This is usually done when someone dies. The hypothesis is that the future society may also have a cure for death. There are many complications with freezing. One such problem is that the brain needs oxygen to stay alive, so if the brain has been dead for over an hour, then the brain could be dead, meaning neuronal death, and therefore loss of connectomes. He offers a way of using the connectome as a way to test out the truths of cryonics. He says that a few hours after respiratory/ circulatory death, there is mitochondrial damage and DNA abnormalities. However e says that those types of damages are irrelevant for connectome death, and what really matters is synaptic wiring.
Next he talks about uploading. Mind uploading or just uploading is creating a computational conscious copy of your mind/ self. Dr. Seung elaborates on this topic asking questions about uploading consciousness. Could the computer self truly be conscious. He then offers a way of doing this by using the connectome. you could find the connectome or partial connectome of a person and then computationally crate the rest. Then if the computer you could answer questions like you or think like you, then it is possible that this is a success. However, is it really a satisfying model. He tackles questions such as this in this chapter. I believe it is all about preserving information about the past, and using that data to infer things about the future.
Terminology
Neural Darwinism
dMRI
Hebbian Strengthening
Human Connectome Project