Full auto-generated episode transcript below. Beware of typos!
Nick Jikomes
Want to start by telling people who you are, what your lab studies and what you do at a very high level?
David Anderson 6:05
Sure. I'm David Anderson. I'm a professor at California Institute of Technology otherwise known as Caltech, in Pasadena, California. And I'm also Howard Hughes Medical Institute investigator and the director of the Chen Institute for Neuroscience at Caltech. And my lab tries to study brain circuits in animals that control behaviors that are very basic survival behaviors, like mating, and fighting and Predator defense, and which in humans are associated with emotion states, to try to understand what we can learn about human emotions by studying animal brains, and whether that will improve our ability to develop new treatments for mental illnesses in the long run. And we work mainly on mice, and also on fruit flies to see how evolutionarily ancient, the principles that we discover are.
Nick Jikomes 7:11
Yeah, one of the things we'll talk about is how sort of taking that evolutionary perspective or studying animals that are in many ways simpler or seem to be very different than humans, is nonetheless useful when we start thinking about things like developing new therapies for humans. And one of the things that you point out very early in your new book, is you say, quote, The sad fact is that there hasn't been a fundamentally new psychiatric drug approved in the last 50 years. All the so called new drugs being released are just variants on the same basic theme. So to sort of set the stage for a lot of the research that you and others are doing. Can you explain why that's the state of things today in psychiatry and why we've we've seen the sort of stagnation in the development of new therapies.
David Anderson 7:56
Yeah, there, there are a couple of reasons why that's true. And the first reason is that most of the drugs that are used in psychiatry widely, like selective serotonin reuptake inhibitors were discovered by accident. They were not discovered through any principled process of experimentation, that led to an understanding of a disease mechanism and suggested a way to treat it, such as we can do, or as was done in the case of diabetes, once people understood that diabetes was a disease due to insufficient insulin production by the pancreas, and they knew what insulin was, and they knew how to purify it or synthesize it, they discovered that you could make people with diabetes feel better by giving them more insulin. There's no psychiatric drug that is used today that was developed using that procedure. So that's, that's problem number one. Problem number two, is that the search for new psychiatric drugs is has gone on for several decades and led to a number of notable failures, such as drugs that attack the neuropeptide substance P for treating depression and anxiety that was a big failed clinical trial, or drugs that try to reduce stress by blocking the brain chemicals that cause cortisol release CRH, which is a particular neuropeptide. And many people blame those failures on the fact that the drugs were tested in rodent models, and that the rodent models are not predictive of human biology and that That's why there were all these failures. And therefore, going forward, we would have to use non human primates, as the animal models for discovery, drug discovery and drug testing. And that is very, very expensive. And even then it's not clear whether the results will immediately translate into humans. And so that that has those failures have made a number of the pharmaceutical companies and biotech companies so gun shy that they've closed down entirely, their neuroscience programs, and not just to try to look for new psychiatric drugs, but also drugs for neurological disorders like Alzheimer's, and Parkinson's disease, too many failures, and no clear knowledge based pathway for identifying new drugs that would be targeted to a particular psychiatric disorder?
Nick Jikomes 11:05
And is it fair to say that, you know, those two problems that you pointed out, what they seem to have in common to me is that both of them have to do with the fact that I think the brain is just very complicated, and we know relatively little about how it works compared to other organ systems in the body. And so there's just this basic knowledge problem.
David Anderson 11:21
That's right. That's exactly right. And until we're able to understand how the brain works, and how, when the brain is broken, it gives rise to various types of disorders, whether they're psychiatric disorders or neurological disorders, we won't really have a clear roadmap for how to develop a next generation of drugs, and will continue to rely on lucky accidents. And so, arguably, the only new drug that's been approved for psychiatry in the last 50 years is ketamine. And there's been a lot of buzz in, in the popular press about ketamine. And again, ketamine is properties as an antidepressant were discovered pretty much by accident, because people were using it as a party drug, we don't understand how it works. In many cases, we use it in the laboratory as a sedative, not as an antidepressant, which is kind of counterintuitive. And in fact, the only indication that ketamine has been approved for is a very narrow one that involves, I think, intranasal delivery with a spray, that's the only FDA approved use of ketamine. But there are lots of people paying to have ketamine infusions into their bloodstream every week, to the tune of, you know, an hour, a day each week, and many, many 1000s of dollars. And we don't know what the long term consequences of these treatments are, there have been no clinical trials to evaluate that. So just just to caution, your listeners that if they hear Oh, Kevin mean, is the great new psychiatric drug that certainly has some promise, but there's a lot we have to learn about it. And the same thing is true for psychedelics, there's been a lot of, of hype, some of it justified about the potential power of psychedelics for treating psychiatric disorders. But the the scientific studies that would be needed to approve get FDA approval to use that drug in this way are a long way from being completed.
Nick Jikomes 13:57
Yeah, so there's just a lot we don't know about how these particular drugs work. And in general, if we take a bird's eye view of neuroscience, historically, I think it's fair to say that for most of the history of neuroscience, a lot of the techniques and a lot of the studies that have been done, have been largely or even entirely correlational. In nature, we could measure certain things about brain activity, we could do imaging studies, or EEG studies, or what have you. And we could correlate, you know, different wiggles in an EEG plot or different imaging study outputs that we have with what people say they're feeling or what, what animals are doing. But a lot of that stuff has been correlational. And you start to talk in your book about this new era of causal neuroscience we're in where we can actually go in and study cause and effect using new technologies. Can you speak a little bit about what causal neuroscience is and what it's actually allowing us to do?
David Anderson 14:50
Sure. So I think we all know, we've all heard that correlation doesn't imply causation. Just because two things vary together in time or in space, doesn't mean that one of them causes the other. They, they could be completely independent, or they could be caused by a third thing that you're not measuring. And my favorite example of this is that there's a very strong correlation between ice cream consumption and violence in humans. And that's not because when people commit a violent act, it makes them hungry for ice cream, or eating ice cream makes people violent. It's because both violence and ice cream consumption increase in hot weather in the summertime, and that's why they're correlated with each other. So similarly, if a person is sitting in a brain scanner, and they report a subjective feeling of fear or anxiety, and the brain scanner shows, at the same time, that there is a hotspot of activity somewhere in the brain, that doesn't tell you whether the brain activity is causing the fear, the fear is causing the brain activity, or that neither of them is has a causal influence on each other. And there's some third thing that you're not measuring, that is independently causing the brain activity, and the feeling of fear in the way that hot weather independently causes violent crime to go up, and ice cream consumption to go up. And so that's really the limitation. And if you incorrectly infer a causal relationship, and try to design an intervention, based on that incorrect inference, you're going to fail. So if you decide that ice cream consumption, increases violent crime, and so the way to fight violent crime is to ban ice cream, you're going to make a lot of people unhappy, and you're not going to have any effect on violent crime.
And so that's really why simply having correlational neuroscience, which is primarily what you can do in humans, is not good enough, if we want to develop a new generation of therapies and treatments. And so causal neuroscience, by contrast, is a way of knowing that involves going into the brain and perturbing the brain, turning parts of the brain or certain cell populations, turning them on when they're not supposed to be on or making them turn on more strongly when they normally are turned on, or turning them off. And then asking what is the effect of that perturbation, the behavior of the animal. And so if a perturbation in a particular cell type changes of behavior, then you can say that there is a cause and effect relationship between the brain cells activity and behavior. For example, let's say that you found through correlational studies, that in mice, there is a certain population of brain cells that are active when the animal is anxious. And you can tell when a mouse is anxious, by putting it in a brightly lit arena, and it will sort of hug the corners and try to stay out of the center. And again, just like a person in a brain scanner, that's just a correlation. Now, if you were able to access that particular cell type, say it's in a brain structure, like the amygdala, and you were able to shut that cell off at will, you could ask whether that manipulation made the animals less anxious. And you could tell that because the animals would stop hugging the corner the walls of the arena, and wander out into the center. Or conversely, you could take a mouse that wasn't anxious that was happily wandering around in the center of the arena, and suddenly activate those cells and see if that caused the mouse to become more anxious, and to run out to the walls of the arena and stay circling out of the center. If you got results like that, that tells you that it's probably worthwhile to study that in more depth and understand how those cause and effect relationships happen in the brain. Because that might help you identify that cell type as a target. For new drug development, whereas if you did the same experiment based on a correlation between anxiety and the activities of some brain cell, and nothing that you did to the cells activity had any effect on the animals anxiety, you probably wouldn't want to spend a lot of time trying to develop a drug to turn that cell on and off as a treatment for anxiety because you know, it won't have any effect. At least it doesn't have an effect in a mouse. And so those sorts of causal neuroscience experiments, they they're not the whole answer, by any means. And they've been criticized, not without some justification for giving an overly simplistic view of causal relationships in the brain. But at least they tell you where to look, even if they don't give you the ultimate answer. And unfortunately, or fortunately, depending on your perspective, we can't do those experiments in humans, except in very, very rare cases. That is, we cannot stick something into a person's brain that is aimed at disrupting or increasing the activity of a particular brain region or group of cells. We can't do that in any arbitrary brain region, and ask how that affects the person's feelings or behavior. Because it's not medically justified. The only time we can do that is if there is a medical justification. So for example, if somebody is in the hospital because they have seizures, and a neurologist is poking electrodes in different parts of their brain, and stimulating to try to figure out where the seizure focus is, so that they can operate and cut it out. They can always ask the person as they're stimulating. Well, what did you feel when I did this? What did you feel when I did that, but they can only do that in a very circumscribed region of the brain, where they suspect the seizure focus site of the seizure may lie, they can't just move the electrode into a different region of the brain that's 10 centimeters away, just out of curiosity to see oh, well, what happens if they stimulate here, they'd have their medical license revoked, and the patient could sue them for malpractice. So we're very constrained by our ability to do directed site directed perturbations in the human brain. And furthermore, the tools that we have to do that are very crude, they basically involve sticking metal electrodes into the brain. And those electrodes inject current into a brain region in a way that probably activates or inhibits 1000s, and 1000s of cells, and comprised of many hundreds of different types of cells. And so we may get highly variable results from one person to the next person, or even in the same person from one day to the next, depending on which cells we're stimulating. So even when we can do there, those experiments, they're sort of blunt instruments. And that's what's really limited our ability to apply causal neuroscience to human neuroscience.
Nick Jikomes 23:38
So, so historically, even in animals, all you could really do is listen to what some neurons were doing in the brain, oftentimes, you wouldn't know exactly which neurons are even listening to, you can watch what an animal is doing, you can correlate the behavior that the animal displays with what you're recording in the neurons. But as you as you told us, You can't make inferences about cause and effect there, and you're really limited by those correlational observations. But now, there are new technologies that enable you people in your lab people and other labs to go in and very specifically manipulate individual circuits and neurons in the brain. So you can actually make neurons more active, less active, change the pattern of their activity, and do that in a very specific way. I think the principal technique there that some people will have heard of, but many will not have heard of is optogenetics. And so can you give people a quick one on one on optogenetics and how it works?
David Anderson 24:30
Sure, I am happy to do that. And also to mention a related technology, which is called chemo genetics. And that that is a different way of doing causal neuroscience experiments, which ultimately may have more immediate applications to humans, then as if on the therapeutic basis, then optogenetics will so first of all, what's the word? optogenetics up Though refers to light. And genetics refers to something having to do with genes. Why are those two words put together to make a new word. And I think that that term was coined by Carl dicer off at Stanford, who was one of the leaders in the development of this technology. And what it what it means is to manipulate using light. That's the Opto part, the activity of a particular group of neurons, and which neurons can be manipulated by light is dictated using genetic tools that allow you to restrict light sensitivity to a particular group of neurons of interest. So that's where the two words come from. But now I have to unpack how you can even make neurons responsive to light, we all know that we have nerve cells in our brains already that normally respond to light, they're in our eye. And they lie in the retina. And they're called photoreceptors. And those are sensory neurons that respond to light of particular wavelengths. By converting the energy of the photons they absorb into an electrical impulse, which they then transmit through a synapse to another cell in the retina. And their ability to do that is due to the fact that they express certain proteins that are called rhodopsin ons. And these are proteins in our eyes, we have them.
Many, many animals have them flies have them. Many mollusks have them. These are proteins that evolved to absorb a photon and somehow convert the energy of that photon into electricity. And I won't get into the details right now of how that happens. But the the idea is that if you could import into one of the cells that's deep in your brain that normally doesn't respond to light, like a cell in your cerebral cortex, or a cell in your hippocampus, if you could implant in that cell, one of these rhodopsin like genes, then in theory, you should make that cell susceptible or sensitive to flight. And that would allow you to change the electrical activity of that cell using light. And so that's really the concept behind optogenetics. And actually, the game or that the trick was to figure out which light absorbing protein which opsin should you choose from nature, to put into neurons that normally are not light responsive, to make them light response. And in fact, some of the first attempts to do this by Giro Meezan, Bach in Europe involve using the actual photo transduction light sensing machinery from the vertebrate eye involved putting opsins and some of the other proteins that they talk to, which are called transducer ions, all into the same neurons. And he was able to show in sort of proof of principle experiments, that you were then able to shine light on those neurons and make those neurons electrically active. The problem with that method is that it was very cumbersome because you have to put not one, not two, but three different genes into the same cell in order to make them light responsive in that way in order to to endow them with the ability to convert light energy into electrical energy. And the more genes you have to put into a cell, a neuron or any kind of cell to do an experiment, the more cumbersome and complicated the experiment is. And so that sort of implementation of optogenetics did not get very much traction in the community or wide adoption. What was really critical was the discovery of a different option from a different animal. In this case, it's a single celled alga called Chlamydomonas, which lives in ponds of water, and it has an option that allows it to detect blue light and swim to towards the light. And so it has a natural protein that detects blue light and converts that into electricity. And in this case, one protein encoded by one gene does the whole trick. And the reason is, because when the photon, the light energy particle hits that protein, which is sitting in the outer membrane of the cell, it causes that protein to change shape in a way that makes it create a pore, or a tunnel, through the cell's membrane that electrically charged ions can flow through to go into the cell from the outside of the cell. And when that happens, it changes the electrical balance across the cell membrane. And that's enough to make the cell electrically excited. And that in itself was a very important discovery. In fact, Peter hegman, who was the biochemist in Germany, that discovered that the the light sensor and the ion channel were part of a single protein in this channel, so called channel rhodopsin protein, share the Lasker prize in medical research last year with Karl Deisseroth, who put that channel rhodopsin gene into neurons and show that now putting just one gene into a neuron could allow you to make that neuron electrically excitable. So that's the that's the optical part. That's how we can make a neuron that is normally not sensitive to light, sensitive to light. Now, what is that good for? Or as my wife always asks, What are you going to do with that? And and what you're going to do with that is figure out a clever way to deliver the gene that encodes that opsin protein from the alga channel rhodopsin, into a specific group of neurons in the brain, using what I would call a genetic zip code, or address for those particular neurons. And the first thing that that your listeners need to know is that there, the brain is not made of one generic cell type called a neuron, it has probably many 1000s of neurons. In fact, the current estimates are about 5000 Different kinds of neurons. And those neurons are different from each other in the proteins that they have, and then their shape, and then their size, and then their connection, because they turn on different genes. And it's becoming clear that if you spent enough time understanding the genetic control of the identity of a particular kind of neuron, you may be able to identify a short piece of DNA, which is what I'm calling the genetic zip code of the genetic address. And if you physically glue or physically link that piece of DNA, with the piece of DNA that encodes this opsin, channel rhodopsin, the light sensitive protein, you now have a genetic embodiment of Opto. Genetics, that is you have a single piece of DNA that contains both the instructions for making a protein that can convert light into electricity. And that also contains the genetic instructions of where in what cell type, that gene should be turned on. And so if you do that, if you can introduce that into a brain using a virus, for example, as a vector, a defective virus has kind of a disposable molecular syringe is the way I like to think about it for inserting that gene into the brain, that gene will only that gene, meaning the opsin will only get turned on in the cells that whose address matches the genetic zip code that you put on that piece of DNA. So now you've got an animal that has in a specific set of cells in its brain, eight light sensing protein. And what Carl dicer off was able to show beyond the fact that you could then turn on such a cell with blue light in a petri dish is that in an intact brain, you could also So, turn that salon with light, if you could put an optic fiber cable into the brain and lower it right down to the area where the cells of interest are located, and then turn on blue light, so that you're now delivering a cone of blue light, deep in the brain. And when that cone of blue light hits the cells of interest, that have the opposite Gene in them, because they have the right genetic zip code, those cells will fire and they will activate an entire circuit of interconnected neurons, that they are linked to synaptic Li. And that may produce a particular behavior. That behavior could be feeding, for example, that behavior could be drinking, it could be fighting, it could be mating, many labs, including ours, have identified particular populations of cells that will make animals behave in a certain way, when you turn them on. An important point. And some neuroscientists even sometimes forget this, particularly systems neuroscientists is that optogenetics is not just about turning neurons on with light, you can also use it to turn neurons off with light by putting a different kind of opsin in the cell. And this is an opsin, that absorbs a different wavelength of light, yellow light. And when it absorbs yellow light, it pumps negatively charged ions, chloride atoms, basically chloride ions into the cell. And that makes the cell more electrically negative than the outside of the cell, and that silences the neuron. So it's the same sort of delivery method, it involves genetic zip codes and involves molecular biology to cut and paste pieces of DNA together, package them into these viruses, inject the virus into the brain, stick an optic fiber into the brain over the region where you injected the virus, and now flick on a blue light or flick on a yellow light to turn those neurons on to or to turn those neurons off.
Nick Jikomes 37:19
Wow. So yeah, this is an amazing technique. That's been around for a number of years now. So in other words, what you've just said is, there are ways today that scientists can take a light sensitive protein that nature evolved in other creatures for for natural reasons, we can take the the gene encoding this protein, we can package it in a way and literally inject or put it into an animal through through other engineering methods, such that specific neurons in the brain express this light sensitive protein. And then using a fiber optic or some other piece of hardware, you can shine light into the brain of an animal like a mouse, and thereby control these neurons by turning them up turning down or changing their pattern. And so you're using light in combination with molecular techniques to manipulate specific neurons in the brain.
David Anderson 38:09
Yep, that's exactly right. And there's, as I said, at the outset, there's a related technology called chemo genetics, which was independently developed by Brian Roth at the University of North Carolina. And that technology is conceptually similar to optogenetics, except that instead of engineering your favorite neuron to express a light sensitive protein, you engineer them to express a drug sensitive protein. And that drug sensitive protein has the property that when you feed the animal, the drug, and the drug binds to that receptor, that also changes the electrical activity of the cell. And there are different flavors of the receptor, some that when they bind the drug, turn the cell on, and others that when they bind the drug turn the cell off much as there are different types of options that detect different wavelengths of light that turn the cell on or turn the cell off. And the these methods have different strengths and weaknesses. Here, the optogenetics gives you millisecond time resolution, over the rate of firing of the neuron and you you can decide exactly how fast you want to fire the neuron. But it's also if you think about it. From the standpoint of human applications. It's very invasive. It requires drilling a hole in the brain, inserting a glass optic fiber into the brain, which is delicate, it could break off connecting that to a power pack and laser which is strapped to the patient and If nobody has done that, and no one is going to do that for a long time, whereas chemo genetics, there's still an invasive component in that you have to inject into the brain, the virus that encodes the drug sensing protein. But the advantages, once you've done that surgery once, you simply have to essentially give the animal a pill or give them an injection, to turn those neurons on or turn those neurons off. So there's no glass fiber in the brain, there's no battery pack, there's no laser, there's just delivering a drug just like you would deliver any other drug to an animal. Except that because you've genetically manipulated cells to have a special receptor for this drug, the only cells in the brain that that drug can act on are the ones that you have genetically engineered to be sensitive to it. And that technology will probably be the first or something like it will be the first technology that is used if and when we get around to developing what people pull circuit cures for brain disorders, that is where the target of the therapy is not a single molecule, as it often is with a drug. But the target of the therapy is a cell, or even a circuit involving several different cell types.
Nick Jikomes 41:36
And I would love to give people a concrete example of optogenetics in action. So David, if you don't mind, what I'd like to do now is, I will share a video that you supplied me, and maybe I'll describe what I see initially, as if I was watching this for the first time. And then we'll have you walk people through what's going on. And I'll loop it a couple more times as you're doing that. Sure. So I'm going to share this video here. Okay, so I'm going to hit play, we've see two mice, one of them is chasing the other one, they look like they're fighting each other one of them's attacking it, and this light came on, and they stopped suddenly. So David, I'm going to play a couple more times. But what exactly is going on here.
David Anderson 42:29
So what is happening here is that there's two mice in the cage, one of the mice lives in the cage. He's called the resident. And the other mouse is the intruder. It's a mice that we drop into the cage. And mice are very territorial, these are male mice. And so if you put a strange intruder mouse into a cage, with a resident in it, the resident will very quickly chase it and attack it sort of like if two people are walking their dogs on the street and the dogs meet each other. And they decide they don't like each other, they start fighting. That's what's happening here. And what we have done to the resident mouse is we've used optogenetics, to express the kind of opsin that will turn a cell off when it absorbs light. We've expressed it in a group of neurons that we discovered in a region of the brain called the hypothalamus that evidently control aggressive behavior. And so that mouse, the resident mouse, has already had a virus injected into its brain and has a fiber optic cable embedded in its brain. This doesn't hurt because there's no pain receptors inside the brain. And that opt that cable is connected to a laser that emits yellow light. And so we wait for the resident to attack the intruder. And when we flick on the yellow light, we are instantaneously shutting off this particular population of neurons, deep in the brain in the hypothalamus. And what that does is it stops the fight dead in its tracks. So imagine if you have something like this in your dog, and you were walking your dog and it encountered another dog being walked and decided to attack that dog, and you're trying to pull them apart by your leash their leashes, and you can't do it because the dogs are too strong. Imagine you could just flick a switch and turn on a laser and it would immediately stop your dog from attacking like that. That's basically what we've done here in a mouse.
Nick Jikomes 44:52
And so how did you so so you've got two mice. In the video we saw one mouse had you know a wire coming out of it. said and that's where the light is being delivered through fiber optic into the brain into the structure called the hypothalamus that you mentioned, where is the hypothalamus? How old is the structure in evolutionary terms, and how to neuroscientists think about what the hypothalamus does at a very basic level compared to something like, say, the cerebral cortex, right?
David Anderson 45:19
So the hypothalamus is at the base of the brain. It's, it's sort of, almost above the roof of your mouth, in in the brain, as it is in in rodents. And all vertebrate species have something like a hypothalamus. Fish have a hypothalamus burns have a hypothalamus, turtles have a hypothalamus, and so do mice, and monkeys and humans. And so it's a very evolutionarily ancient structure. And it's known to control a lot of important visceral functions, like feeding and drinking, and temperature regulation. And it does some of that by releasing hormones that go into the brain or into the bloodstream. And so at one point, the, the hypothalamus was sort of viewed as kind of the pancreas of the brain, sort of like a secretory structure in the brain that squirts hormones into your bloodstream that control whether you're hungry, or thirsty. And if you're hungry, they make you eat, if they're thirsty, they make you drink. But we now know that the hypothalamus is not just a secretory, Oregon, in the brain, but it's actually a complicated and sophisticated network of neurons that are connected to each other. And different neurons in that network control different types of behaviors, and they do so with enough specificity, that if you find people have found the neurons that control eating, and when they optogenetically stimulate those neurons, it can make a mouse that's completely satiated, eat ravenously, they found similar neurons that make mice drink when you turn them on. We've, as you've shown, we found neurons that are necessary for fighting behavior, and which can cause an animal to attack something that normally wouldn't. When you turn them on, we've done the same thing for mating. So there, there seem to be very dedicated behavior, specific groups of neurons in that part of the brain to control behaviors that are critical for the animal survival. And we like to think of these as the four F's feeding, freezing fighting and mating, let your listeners decide what the fourth, f is, because I can't say it on the air. But in contrast, you asked about the cerebral cortex, that prevailing view of the cerebral cortex is more like it's a sort of general purpose computer. It's like a central processing unit that you have in your laptop, it can run many different kinds of applications. It can be multi purposed to do many things. And it's a much more sophisticated device computationally in this in this view, than the neurons that are in in the hypothalamus that are in the sort of basement of your brain, if you will. And so that's the that that in a very, overly simplistic way, is, is the distinction between how people think about the cerebral cortex versus the hypothalamus, and roughly what the hypothalamus does.
Nick Jikomes 49:11
So, in that video, we watched, you've got light shining into the hypothalamus, you've got certain cells in the hypothalamus of the mouse that are expressing this light sensitive protein. And in the case of that video, you're shutting them off, the behavior, the aggression, behavior stops. So we're in a structure that's deep in the brain. It's very ancient, it's in all all types of animals. And it does, it's associated with behaviors that we think of as, as visceral or basic or instinctive things like feeding and fighting and mating, and so forth. Now, you mentioned that there's different kinds of neurons associated with different behaviors that you can stimulate or turn off. There's ones associated with feeding specifically with drinking specifically with aggression specifically, and so on and so forth. How did you identify i The particular population of neurons in this case that have to do with aggression, what was the molecular marker associated with those neurons? And why is that significant? Right?
David Anderson 50:09
So the way that we identified those neurons is first to do a correlational experiment, getting back to your first question where we could take mice that fought and look inside their brains. And we had to do this in slices through the mouse brain. So the mice had to be sacrificed for this experiment after they fought. But we stained those brain slices, with a a chemical that lights up individual neurons that have been active just before the mouse was sacrificed. So if the mouse was fighting, just before it was sacrificed, then you see the cells that were active during fighting, lighting up. And that showed us that in a particular region of the hypothalamus, there was a group of cells that got very active when the animal was fighting. It wasn't the only place in the brain where we saw that. But it was in a part of the brain that was suspected to be involved in this behavior from more classical studies that had been done decades before in cats and in rats. And so then what we tried to do is to ask, can we find a genetic zip code for those neurons, can we find a gene that marks those neurons and not other neurons that are intermingled with it in that region, which are not activated during aggression. And we did that by trial and error, we just combined staining the slices to visualize the active neurons, and with staining them with a different color tag that identified a particular gene. And so the gene tag was read the activity tag was green, when they're both in the same cell, the cell looks yellow. And so we tested lots of different genes, and asked which ones gave us the most yellow cells after the animal was fighting, and didn't show yellow cells in animals that weren't fighting or that we're mating, we're doing something else. And
surprisingly, at least to us, the gene that we turned up that looked to be the best candidate was an estrogen receptor, I say, an estrogen receptor, because there are two, this is the type one estrogen receptor. And I can come back later on, if you're interested to the question of why aggression, neurons should express an estrogen receptor. But right now, you can just think of it as a gene that marks those cells. And so through genetic trickery, then we were able to use the zip code from from that gene, the estrogen receptor gene to genetically address the light sensitive opsin to the neurons in that region of the brain. And when we did that, and when we used an option, actually, the first time we did the experiment, in instead of shutting the neurons off, we use the option that turned the neurons on. And we asked if that would make the mice attack something that they would not normally attack, like a female mouse, or even an inanimate object, like an inflated latex glove. And we found that it could do that. So these were very powerful cells. And here, you're showing a movie of stimulating those cells in a mouse, which is now attacking latex gloves. So when the light is on in the upper left corner, there in the movie, it means that the laser is on, and we are stimulating the cells. And the mouse goes from investigating the rubber gloves to actually biting it and attacking it, which is which is very striking. And so that's that's how we identified those cells. And importantly, in later experiments, we we showed, in a control experiment that if we deliberately tried to activate the neighboring cells that didn't have the estrogen receptor in them, then we didn't get fighting when we activated those cells. So if you think of these different cell types as a mixture of salt and pepper crystals in this region of the brain, and the estrogen receptor marks the salt and the Have the pepper neurons don't have the estrogen receptor, we only got attack when we activated the salt neurons, not we activated the pepper neurons. And that made it possible for us to very reproducibly and deliberately activate this group of neurons. And I should say that this, we, these experiments started over 10 years ago. And
it we sort of take for granted, I think many neuroscientists take for granted that, oh, yes, if you look for neurons in the brain that controls something like aggression, you're gonna find them somewhere. But that wasn't always the way that people thought about aggression. In fact, in the 1940s, and 50s, there was a lot of this so called nature nurture debate going on, specifically, with respect to aggression, and brain stimulated aggression. And the question was whether animals had to learn and have a part of their brain instructed by experience to make it aggressive, or whether animals were born with particular neurons in their brain that were pre programmed to produce aggression when they were stimulated. And while it was true, that in the late 1920s, experiments done in cats, by Walter Hess showed that you could make a cat aggressive by sticking metal electrodes into its hypothalamus and stimulating there were a couple of different regions where he could get a similar result. And because these were cats that were basically caught on the streets of Zurich, as far as I can tell, and brought into his laboratory, that's before there were any animal research regulations, you really couldn't, you didn't know what the life history of the cat was, and whether it had learned a lot of fighting while growing up on the streets of Zurich. And if so, whether that whether Dr has happened to be lucky enough to hit a region where the animal's brain have learned to control aggression. And what our experiments showed, really, if nothing else, is that that idea is wrong. And that every mouse has this group of estrogen receptor neurons in its brain. And that if you stimulate those neurons, you're going to get attack, it's not different from one mouse to the next. If you stimulate those neurons in a different mouse, it's not going to make them eat, instead of attacking, or it's not going to make them drink instead of attacking. It really is a hard wired population of neurons that appears to be dedicated to the control of aggression. And I think that is the sort of most basic conclusion from those types of studies. Yeah, I
Nick Jikomes 58:12
mean, it's just a dramatic result. I mean, you can turn aggression on at the flip of a switch, you can turn aggression off when it's happening naturally, you're doing it by manipulating this particular population of neurons in the hypothalamus that is marked by its expression of one of the estrogen receptors in the brain. And it's different from other neurons that are right next to it in the hypothalamus that are involved in other behaviors. I want to talk a little bit about the significance of the estrogen receptor here as a marker for these neurons. If I try and, you know, if we met, when we speak, in everyday speech, you know, people often reference hormones, they'll say things like, you know, teenagers are filled with hormones. And this is why they do crazy things or take high risk behaviors or erratic, they might reference you know, a woman who is going through pregnancy or different phases of her menstrual cycle, and that her behavior is quote, unquote, due to hormones, but what exactly are hormones? And how do they work to influence our behavior? And why is the hypothalamus sort of a very important brain structure for connecting those dots?
David Anderson 59:18
Right, those are, those are great and very important questions. So hormones are a type of chemical that circulates in our bodies, and they can come in different classes or different forms. people have probably heard of steroids, that's there's a group of steroid hormones and then there are also peptides, peptide hormones. So insulin is an example of a peptide hormone. And estrogen and testosterone are examples of steroid hormones. What these hormones do in the case of the steroid hormones, Is that they're able to cross the blood brain barrier and get into the brain. And then some of the cells in the brain, like the ones that express the estrogen receptor, what that receptor is, is a protein molecule in that neuron that is a docking site for the hormone. So the hormone goes into lots of different cells, it crosses their membrane, and it goes into the cell. And if the cell isn't expressing the right receptor, the hormone doesn't do anything, it just hangs around until it gets graded, or it goes back out of the cell. But if the cell has the receptor for the hormone, the hormone docks to the receptor, sort of like the little space shuttle docking with the space station at the beginning of 2001. And then the space station, which is the hormone receptor, turns on or turns off many genes in that same cell, and the turning on or off of those genes can make that cell more active, less active, it can make it grow, it can make it shrink, it can even make it survive and prevent it from dying, or in some cases, it could kill it. So these hormones are extremely powerful, because they affect very specific groups of neurons in the brain. And the effects that they have on those neurons are very profound in the way they're able to shape shape the activity of those neurons and therefore affect the behavior those neurons control.
Nick Jikomes 1:01:43
So hormones are literally going through the bloodstream, into the brain and directly affecting neurons in various ways. That's right. And was it surprising that the estrogen receptor and an estrogen receptor was the marker you found for these aggression neurons.
David Anderson 1:02:00
In in retrospect, it shouldn't have been surprising, have we been more sophisticated in our knowledge of this region of the brain. But like, many people in in, in the lay community, we assumed that if any hormone, steroid hormone has something to do with aggression, it must be testosterone, we assume that testosterone is the male hormone. When you have more testosterone, you're more aggressive. And estrogen is the female hormone. And if anything, it should make females less aggressive so that they can mate and take care of their young and offspring and raise them. So estrogen is responsible for all the good things that animals do and Testa and people. And testosterone is responsible for all the bad things that people do. And it turns out that that's way too simplistic, and many, it's been known for, for actually, for decades, that even in males, the actions a lot of the actions of testosterone, are mediated, not through its own receptor, which is called the androgen receptor. It's a receptor in, in, in that binds to testosterone, but testosterone can also be chemically converted in the brain to estrogen in a very simple chemical reaction, and then the estrogen can bind to the estrogen receptor and exert particular effects. So for example, it's, it's been known that if you chemically if you surgically castrate a rodent, a male, it will lose the ability to fight. And of course, if you give it back testosterone, it'll regain the ability to fight. But you can also give it back the ability to fight by giving it estrogen instead of testosterone because you're just bypassing the requirement for the brain to convert the testosterone into estrogen. So estrogen is very important for this is not all our work. This is work in many labs, including Niraj Shah, who's a former student of mine now at Stanford, estrogen, somewhat paradoxically, plays a key role in masculinizing, the mammalian brain. It's released shortly after birth in a peak and that release of estrogen is critical to set up the circuits in the male brain that control male beat specific behaviors and male or aggression. And then others have shown that in the adults even estrogen receptor in this region of the hypothalamus is necessary for mice to fight. If you genetically knock out that estrogen receptor in that region of the hypothalamus, the cells, the animals don't fight. So you need the estrogen both developmentally in order to construct a masculinized nervous system. And you also need it functionally in the adult to allow those neurons to function correctly in aggression. And conversely, in females, it's been known since the late 70s, from work from Donald faff that, in the region of the female brain, the analogous region of the female brain, there are cells that control mating behavior that also have the estrogen receptor in them. But there are different kinds of neuron from the neurons that control aggression, as we and others have recently shown. So even in the same brain region in there are not only in one sex, some neurons that respond to estrogen and some that don't. But if you compare the two sexes, males and females, there are different kinds of estrogen responsive neurons in the female brain, and then the male brain.
Nick Jikomes 1:06:30
And you know, one thing that I think that strikes me here about hormones, is they are probably not only controlling neurons, like the ones you just described in order to effect behaviors like this. But I imagine they're also coordinating physiological processes throughout the body. So for example, you know, when the aggressive mouse gets aggressive, it needs to have the motor output of aggression that we're watching in the video, but it also needs to, you know, be sending more blood flow to its muscles, it probably needs to have its pupils dilated things like this. So our hormones acting, how toxin are they acting locally versus sort of globally throughout different tissues to coordinate all of these different systems.
David Anderson 1:07:11
So there's definitely actions of the sex steroids outside the brain, for example, testosterone builds muscle tissue. That's why weightlifters often take testosterone to do that. And by the way, when they do that, they often develop female secondary sex characteristics, like breasts. And that's because they're taking all this testosterone, and a lot of it is getting converted into estrogen and estrogen makes breast cells grow in males as well as in females. So those weightlifters also take aromatase inhibitors, which are drugs that prevent testosterone from being converted into estrogen so that they can have big muscles and no breasts. And that indicates this intimate relationship between these two sex steroids, testosterone and estrogen, and the fact that we have drugs that can block this conversion. So yes, there are effects of these hormones in the body, on tissues like muscle tissue, breast tissue. And there are also indirect effects you mentioned, pupil dilation, heart rate, those sorts of things. Functions are mainly attributed to so called fight or flight hormones like adrenaline, which are released from the brain and also from the body during conditions where the animal is stressed. And they're released from a type of neurons called sympathetic neurons that are outside of the brain and the spinal cord, but which get input from the brain, including from the hypothalamus, and the regions of the hypothalamus that are sensitive to estrogen. So this is a way that estrogen acting on cells in the brain can affect cells outside the brain that release fight or flight hormones, like epinephrine, adrenaline, that can make your heart beat faster, increase your blood circulation, make your pupils dilate, and do all of the other things that you need to do if you're fighting or fleeing.
Nick Jikomes 1:09:33
Now, one always in science tries to resist the impulse to anthropomorphize. But on the other hand, you just can't help it when you see some of these videos. So when we think about the aggression neurons being turned on in the mouse that then attacks the latex glove. Is the mouse feeling pissed off? And how do you how do you even start to think about what might be going on there in terms of the emotional experience that might To be added by the mouse,
David Anderson 1:10:01
right? So that that is the guest these days, it would be the $64 million question, not the $64,000 question. And that is, does the mouse feel anything, when it is engaged in these behaviors? And if so, what is it feeling? Is it an angry mouse? Is it pissed off? Or is it a happy mouse because it has a chance to beat somebody up and it gets reinforcement from feeling like it's the champion.
We're just beginning to get glimpses of answers to those questions. And they're very hard to ask, first of all, the whole question of whether a mouse or indeed any animal is feeling something is something that we can't really ask right now experimentally, in an animal in an objective way. And that's because feelings are something that as humans, we determine or identify by introspection. If somebody asks you how you're feeling, you sort of think about okay, what, what, what am I sensing in my body? In my brain? Am I happy? am I sad? It's a very private experience, the subjective feeling, we can ask people how they feel because people can talk. But you know, somebody, a person can lie about their feelings, they can be angry and say they're not angry. If they're not fighting, you don't know what if they can keep a poker face. So even verbal report is flawed as a way of assessing feelings in humans. But animals can't talk. So we can't ask a mouse. Are you feeling angry when you are fighting, but we can ask whether there are things going on in the brain besides just the control of going through the motions of attacking. In other words, we can ask whether this attack behavior is just a robotic reflex, or whether there is some underlying state a drive state or mode of state that propels the animal to commit aggression, or perform aggression. And that may be that whether the mouse feels or is aware of that state or not, we don't know I'm not saying that mice can't feel anything, I'm just saying we don't have a objective scientific way to test that. But they may still have a state in them that we can measure. That tells us that there's something more than just motor reflexes going on the extreme examples, when I was a kid, there was a toy that you could buy called Rockem, sockem robots. And these were little robots, and you wound them up, and they would fight with each other. And I don't think anybody would attribute feelings of anger to those robots. And so clearly, one can observe fighting behavior in something that looks like an organism that's manmade, but which isn't feeling anything or having any internal state. So the way I think of the relationship between feelings, and internal states, or emotion states, is I imagined the whole state package as an iceberg floating in the ocean. And the tip of the iceberg is the feeling part, which is what we can measure in humans, but we can't measure it in animals. But there's a whole big part of the iceberg that is submerged below the tip, which is something that we have, and also that we think we share with a lot of animals, which is the the emotional state that is driving the behavior. And we can study that without having to worry about whether the animal is feeling a sense of anger. And so how do we do that? There are some indirect tests that people use to assess motivation. And so a classic test in experimental psychology is to test for motivation is to see if an animal will do work to obtain something. If it does that something is considered rewarding to the animal or it's considered a reinforcer. So, if an animal is starved, you can teach it to press a bar to get a food reward. And the more it presses the bar once it's learned to do that, you can control how many times the animal has to press the bar to get its food reward. And so when you start at the animal only has to press the bar once and it gets a pellet of food. That's very easy, and a very hungry mouse. And a not so hungry mouse will both learn to press the bar once. But if you change the rules of the game, and you make it necessary for the animal to press the bar five times, or 10 times, or 15 times, a not so hungry mouse will eventually give up presumably, in frustration, it simply doesn't. It's simply not motivated strongly enough to keep pressing the bar because it's not that hungry. But a mouse that's been starved for a long time will press the bar as many times as necessary in order to get that reward. So psychologists define a motivational state or in this case of food seeking state, as a state that you can measure by this type of bar pressing or nose poking, test. Now, you can't do that when an a mouse is fighting, because it needs a whole apparatus with bars and nose pokes and lights and sound. And you can interrupt the mouse, when it's in the middle of a fight and say, Excuse me, could you just stop fighting for a minute and go over here and poke your nose in this hole and tell me how many times you want to poke your nose and then I'll let you go back to fighting again, once it's begun fighting, it is sort of past the point of no return. But you can test the motivation of the animal to fight or to seek out the opportunity to fight. And this was done originally by Klaus meat check in a Tufts University and others who showed that mice will actually learn to press the bar or poke their nose to get the opportunity to beat up a weaker subordinate mouse. So you train the mouse, by having a light go on when the light goes on. Mouse knows that it is supposed to decide whether to poke its nose or not poke and you train it that every time it pokes its nose, you drop into the cage, a weak link, subordinate mouse, and you let the resident mouse of the cage Have at it and attack. And then you do it again and again and again. And eventually, the mouse learns that when the light comes on, if it pokes its nose, it will get the opportunity to attack another mouse now, so that's been shown in a number of studies that mice and rats will learn to do that. And ordinarily, we don't think of attacking as rewarding. We think of aggression as something that is negative that is associated with feelings of fear, frustration, harassment, anger, rage, but there are forms of aggression, that are reinforcing. Somebody. A prizefighter that wins a boxing match is is going to feel victorious and reinforced for having won the boxing match. And it looks like that if there is any internal drive or emotion state that is associated with that type of aggression, and mouse is a type, because there's many different types of aggression in mice, just like there are in people, but at least that type of aggression. If the mouse wins, seems to be rewarding to the animal. Now, we can't say from those experiments, whether the act of attacking and biting the other mouse is really rewarding to the mouse, or whether it's just the defeat of the other mouse and the establishment of dominance over that mouse, which is rewarding. I suspect it is probably the latter. But it's very hard to show that experimentally. But that the fact that the animal will learn and do work for the opportunity to engage in that kind of aggression, suggests that the animal that's something the animal wants to do, because he finds it rewarding. or reinforcing whether there are forms of aggression. Where the animal is in a state that is more analogous to what we would label as rage or anger is something that we're very interested in, and are trying to find out. And it has to do, it gets back to this issue of different types of aggression. So the type of aggression that the animal will learn to poke its nose for and that it finds rewarding, that we think of in the field as offensive aggression, or you can think of it as proactive aggression, the animal is starting the fight, because it wants to win. Whereas the type of fighting behavior that is we associate with anger and rage in humans, you can think of as defensive aggression, or reactive aggression, it's the type of aggression that we engage in, if we're threatened, or if somebody makes us angry, or if we're trying to protect or defend a resource, or a loved one. And there is good evidence in rats, that there are both proactive and reactive, offensive and defensive forms of aggression. It's been harder to define that in mice. But I think that if there is a clearly definable form of reactive or defensive aggression, that might be the type of aggression that has an internal state that is, has a negative valence that is unpleasant to the animal, and that the animal is fighting to get rid of that unpleasant state, rather than to gain a state that is a pleasant or reinforcing state. So it's a long winded question, answer, which we always give in science when we don't know the answer.
Nick Jikomes 1:22:09
So you mentioned a phrase a term a moment ago, valence, so you can have an experience as a negative valence that the animal doesn't like having it can have a positive valence, the animal does like having it at a fairly high level. Why? Why would nature want to evolve nervous systems that are coupling motor commands and motor outputs and motivated behavior? With these types of emotional states? What does that have to do with learning and memory? And how a brain is working in order to motivate an animal to do something or not do something?
David Anderson 1:22:43
Right. So that that is a is a deep? And an important question. And one, one answer to that question is, if you imagine that animals evolved to be robots, that were like a, the the unmanned rovers that we have on Mars, I mean, we can design robots that will drive around Mars, they will decide whether to turn right or to turn left, they will decide whether to keep trying to drive over a rock. Or if it's too big, they'll back off and try another route. So they're able to make decisions and exhibit motor behaviors, local motor behaviors, they stop, they go. But all of those behaviors are hardwired into the system by the engineers that built the robot, which means that the system doesn't have much flexibility, because all it can do is whatever the engineers programmed it to do. And if it encounters Wallace driving around Mars, a signal or an experience that the engineers didn't anticipate, it won't know with quotes, how to respond, and it could flip over and break or get stuck or something like that. So what I'm trying to say is that what internal states emotion and motivation states do is they give animals brains more flexibility in deciding how to respond to particular signals that they get in their environment. They are not just a series of hard wired connections, where if this wire gets triggered, than the animal is always going to do x. And if that wire is triggered, the animal is always going to do why. So one analogy I've used To think about emotion states and the brain is to think of them, kind of like one of those old fashioned telephone switchboard systems were, that you see in the movies now, where you have a bunch of operators sitting at tables. And when a call comes in, they unplug a wire from one socket, and then they plug it into another socket to make a particular connection. So there's a lot of flexibility in which wire can be plugged into which socket. And it's sort of a clearinghouse for taking information in, and then deciding if you will, what to do with that information, and how to route it. And so I think that's what these emotional states do is they endow animals with the capacity of their brains with the capacity of flexible routing, that is, they can take information that comes into their brain and flexibly route that information to produce different behaviors, according to where they find themselves at the moment, the context, what happened to them before their experience, what they think might happen to them in the future, and also their physiological state, are they hungry? Are they thirsty? Are they tired? Are they hot? Are there are they cold, etc? So that's really the best way I can answer that type of sort of teleological question. Why evolves? These sorts of states?
Nick Jikomes 1:26:31
Yeah, and I think an intimately related question is just, you know, how you start to think about why evolution would produce any animals that have any conscious awareness at all it gets, it gets at some of the things you were just saying. But you know, when you think about the correlation between the emotional states that you're talking about, and that you're trying to study in mice and other animals, and the need for behavioral flexibility, you know, when we just look at our own experience, as humans, the most intense conscious awareness is always associated with things that require a lot of attention, and a lot of choice between a menu of options. It requires a lot of behavioral flexibility. And so the conscious awareness piece seems to be very much tied up with the emotion piece of this. So do you how do you separate conceptually? Or do you separate, conceptually conscious awareness? And the feeling of emotion? And to what extent in what ways are these things related to behavioral flexibility? And how you think about the evolution of consciousness per se?
David Anderson 1:27:35
Okay, so that's, that's also a multi faceted question. And very profound one, I'm not sure I'm going to be able to answer all of it, but I'll try to touch on some of it. First of all, the relationship between conscious awareness and emotional feeling, I think about and many of my colleagues think about, as I said earlier, feelings are subjective states that we are consciously aware of. So from that perspective, emotional feelings require conscious awareness. But that doesn't mean that we can't have emotions that we're not consciously aware of. I mean, many of us have probably have the experience of, you know, a significant other, looking at us and saying what's wrong? And we say, What do you mean, what's wrong? Well, you have this concerned expression on your face? Oh, I do. I didn't think I had any expression on my face at all. That's, that's a sort of anecdotal example of what I would call an unconscious emotion. And that's been replicated and formally demonstrated under controlled laboratory settings now that you can find evidence of emotion states in the way that people will respond to a particular stimulus, say, an angry face or a fearful face, even if the people don't report being consciously aware of that state. So that makes it I think that to me is good evidence that, that supports the idea that when we use the word emotion, we're referring to the state. When we use the word feeling, we're referring to the conscious awareness or subjective experience of that state. Now, not everybody agrees with me, because in colloquial speech, and everyday speech, we use the word emotion and feeling interchangeably. So when we say emotion, in everyday speech, we mean feeling but scientifically, these things are distinct and right now feelings are something you can only study in humans because you can ask a human how you feel, but see Since you can have emotions without having any feelings, then there are ways you can study those in animals. And the emotion states, I think play a key role in controlling behavioral flexibility and behavioral decisions. And this gets into a very sticky, very sticky question of free will. When you when you do something, and you make a choice under a certain set of circumstances, you have a as a people, we have a feeling of agency, we made the choice, and then we behave that way. But there are data that suggest that our brains can actually make choices before we even have a conscious awareness of the fact that they have done that and made that choice is analogous to what I said earlier, that people can have emotions that they're not even aware of, or consciously feeling. And, and so I would say that you could view the sort of feeling part, the sense of agency and conscious awareness of a flexible decision making as an epi phenomenon. It's something that your brain makes you think you have control of, but you don't really have control of it. It's happening after the fact of the actual decision. And the decision is being made in your brain much the way the decision is being made in an animal that doesn't necessarily have a conscious awareness of its choices. Although, as I said, we should be agnostic about that, and not negate it. So that's one point of view. That is that feelings are just sort of epi phenomena, they don't guide flexible behaviors. But that sort of negates freewill if you if you accept the concept of freewill, at least in humans, and then then that incorporates the idea that conscious experience and conscious feelings are something that goes into decisions that we make when our brain is calculating, should we do x? Or should we do y is x going to make me feel good, or feel better than if I do y?
Or vice versa? And so it could be that there is a role for consciousness, at least in humans conscious awareness in guiding choices through a sense of agency. But that is not necessarily true. And if you you ask the question, well, why? Why should consciousness have evolved? And there again, there is a debate. And that is, is is consciousness, a biological function that arose in evolution through natural selection, like the function of your kidney, or the function of your heart, or the ability of your brain to control drinking, and eating? Or is consciousness sort of a inevitable byproduct of a neural system when it gets sufficiently complicated when it has enough nodes, enough wires, enough feedback, enough connections, and there are models of consciousness that have been proposed by Giulio Tononi that, basically define consciousness in terms of this complexity of the system? So you can ask, well, if it's a byproduct, and evolution didn't necessarily select for it? Why should we have it? Well, it's off. There are many examples where evolution will select for one function that an organ or tissue or a system in your body is doing. And then something else that is very tightly connected, comes along for the ride, as it were, and is also selected in evolution, not because it provided the immediate selective benefit to the animal but because it was tied to something else that provided the selective benefit to the animal. And that's, I know, that's a sort of abstract and complicated question. But if you think of systems in our body, whether they're physiological systems functional systems as as being chains of mechanisms that are connected to each other, and a particular mechanism, like your digestive system, or how you how you do a particular action has many different moving parts. And it may be that evolution and selection pressure acted on one of the of those parts in this interconnected chain, because that provided an advantage in a particular when that part played a role in another process, because it's connected into another chain as well. But once it is selected for all other things, that it's chained to get selected, as well, because they're sort of joined at the hip. And that's one way of thinking about the evolution of consciousness that it's sort of joined at the hip to having a very complicated brain with many, many wires and many, many connections. But that we might be there are animals that might be able to function perfectly well, without consciousness. And we can't answer that question right now. Because we have no objective way of knowing if an animal is conscious or not, are aware or not.
Nick Jikomes 1:36:31
Are there any good examples of an instance where you have something selected for in, in the physiology of an organism where this other thing comes along for the ride that isn't necessarily adaptive? Or wasn't the thing that was selected for?
David Anderson 1:36:48
There are, but I'm blanking on them right now. And person who has written about this extensively is Stephen Jay Gould, the evolutionary biologist. And so I would refer your listeners to his writings, too, if they're interested in this particular phenomenon, I'm sure it also has a technical term for it. I'm not an evolutionary biologist. So I don't remember that. But that would be that would be one place to find these examples. But there are, there are many examples that are believed to have evolved that way by sort of piggybacking or being joined at the hip to something that did have a selective advantage.
Nick Jikomes 1:37:33
Yeah, I mean, one thing that sort of comes to mind, I'm not quite sure how well this works is humans are very prone to choking. And the reason for that is, you know, what was very strongly selected for was our ability to speak to each other. And I'm not an expert in this, but that that involves a lot of evolutionary change in our throats to give us the the morphological capability to make speech sounds. And a natural consequence of, of creating an animal with the physical ability to speak like that, in addition to the the cognitive ability to do it, is it made us prone to choking, and we can, you know, we can breathe through our mouths, but really, our nose is what's for breathing. And mouth breathing, I guess looks like it would be selected for but it's basically just a side effect of making an animal with a throat capable of speech. And that just there's no way to hook that up without leaving us prone to these other things, or using our mouth in ways that aren't, aren't really what it was evolved to do.
David Anderson 1:38:33
Yep. The so I think that's a fair point. And whether that applies to consciousness or not, is really open to debate, I'm sure there are many people that feel that our our the consciousness that we have is, and our ability to have subjective feelings is critical to our decision making. And that if we didn't have those conscious feelings, then we wouldn't be able to function. And that may be true or not able to function at the level of sophistication that we do. But animals are pretty sophisticated. And which means that either you can, an animal can do a lot, even if it doesn't have conscious awareness, or it means that all animals are conscious. And that's or a large majority of animals are conscious. Again, this is a philosophical question at this point, not a scientific question that that reasonable people disagree on or agree on.
Nick Jikomes 1:39:36
One of the last things I want to ask you about David comes back to some of these new neuro technologies and the question that that you brought up at the beginning of the book about the sort of stagnation that we've seen in the development of new treatments to address human psychiatric conditions. And so if, if a lot of what your research and a lot of what other neuroscientists research are showing is that behavior errors and emotions and all of the stuff can be driven by specific patterns of activity and specific circuits in the brain. What does that say about the prospect of developing new and better psychiatric drugs, because, to my mind, it would seem pretty infeasible, that we're going to develop drugs pills that someone swallows that get into the bloodstream, go into the brain and affect only the specific circuits that we need to affect and only the specific ways we need them to be affected in order to treat psychiatric ailments. So is it possible to develop drugs that have that level of circuit specificity? And, you know, if it is, whether it is or isn't? Are we going to be using some of these other techniques like outdoor chemo genetics in human beings eventually?
David Anderson 1:40:42
Yeah. So to answer your first question, I would never say never, I think it might well be possible to develop drugs, maybe not drugs that you take by mouth is a pill, but drugs that are injectable that are maybe mRNA based nanoparticles like the vaccines that we've been getting. But they're not vaccines, they cross the blood brain barrier, they get into certain specific nerve cells by means of genetic zip codes, and they reprogram the cells to do certain things, that science fiction right now. But technology is developing so rapidly, that it's not impossible. On the other hand, the the use of of chemo genetics, in particular, to treat brain disorders is I think, something that is potentially on the near horizon. Because there if we can develop genetic zip codes that can be used to target therapeutic genes, or or genes that make a cell susceptible to a drug, if we can target them to particular cell types. And we know that those cell types are involved in, say, fear or anxiety, and they're hyper activated, for example, in patients with post traumatic stress disorder, or that have an anxiety disorder or something like that, then we could potentially endow those cells with the capability to be, let's say they are fear promoting cells, we could endow them with the capability to be shut off when we take a drug, which is the chemo genetic ligand that binds to the receptor that we have programmed them to express. So we're using a drug to change the electrical activity of a cell in the brain. And an entire circuit that it is connected to, because of its effect on behavior. And I, I call that chemical deep brain stimulation. So it achieves the same objective that deep brain stimulation doesn't do DBS is used now, where you put bundles of wires in the brain and the patient has much like they would have if they had a pacemaker, they have a battery pack, they have wires, and deep brain stimulation is being used to treat Parkinson's disease, as well as some psychiatric disorders like obsessive compulsive disorder. And there's some data that suggests it may be used useful in depression, as well. And the problem with DBS is that is kind of hit or miss because if you don't put the electrodes in the exact correct place in each person, then it may not work. And what chemical DBS would is, is it's doing the equivalent of DBS that is it stimulating or inhibiting some neurons in your brain but you have selected specifically which neurons to stimulate or inhibit. And you're achieving that stimulation or inhibition by taking a pill, not by having to put metal electrodes into your brain. Now, it's true that until a couple of years ago, if you were contemplating that kind of an approach, you would still need to do a one time brain surgery on that patient to inject the virus that encodes the the gene that will make the neurons that you want to activate or inhibit sensitive to the drug. But more recently, through work of, of Ben Devilman and Viviana Grata Naru. at Caltech, there have been developments in viruses that can actually cross the blood brain barrier, which is very surprising because viruses are huge by comparison to into Visual proteins and even individual proteins can't cross the blood brain barrier. Which means that if we were able to find the correct genetic zip code to address a, say, therapeutic chemo, genetic effector, to a cell type that we want to manipulate for therapeutic purposes, like a cell type that's involved in fear or anxiety, we wouldn't have to drill into the person's brain and inject the virus in their brain, we could just inject the virus into their bloodstream, and the virus would circulate. And it would go into the brain, and it would deliver its cargo into various neurons, and only the neurons that could read out its genetic zip code would would express the therapeutic agent. And I think that that is no longer science fiction, not just because we have now the ability to have viruses that can be administered in the blood and will cross the blood brain barrier. But also, because of our ability to identify genetic zip codes, for particular cell types, is growing by leaps and bounds, because of advances in technology for sequencing the DNA and RNA of individual cells in the brain. And I'm thinking of work that's being done at the Allen Institute, a nonprofit research institute in Seattle, that was founded by the late Paul Allen, and I sit on their scientific advisory board, there, they are making remarkable progress in being able to identify genetic zip codes that allow you to target these therapeutic genes to particular cell types in the brain. Now, this gets back to this question of, well, to do all that work, you're going to have to do it in mice, how do you know that it's ever going to translate to humans, and this is this is where another very important contribution of the Allen Institute comes into play. And that is their discovery, that among the myriad cell types in the brain, there are some cell types, not all, but some cell types that you can identify and recognize in a mouse in a monkey. And in a human, they have the same shape, they express many of the same genes, they're located in the same region of the brain, they have the same electrical properties. And if you identify the right genetic zip code, that genetic zip code can be used to address a therapeutic gene to that same cell type in a mouse or a monkey or human, which means you can do the sort of scut work, if you will, of finding the genetic zip codes, using a mouse, and then sift through those and find ones that will translate to monkey and to human. And I think the that ability, if it turns out indeed, that many brain disorders, like Parkinson's, and ALS, are disorders of specific cell types, if those cell types are evolutionarily conserved, and we can identify genetic zip codes of them by doing the work in mice, and they will immediately translate into humans, then combining that with these non invasive ways of delivering viruses into the brain that don't involve brain surgery, but just injection of something into the bloodstream, no different from a vaccine. I think that this goal of of taking the knowledge from causal neuroscience that is giving us an understanding of mechanism of the circuits that promote fear, aggressiveness, anxiety, and how those circuits may malfunction or function, maladaptive ly, in certain emotional disorders, we may have a prayer of developing new treatments for those disorders that are more along the lines of diet, insulin for treating diabetes, they're based on an understanding of the physiology and of what happens to the physiology in a disease and trying to reverse that pathophysiology. Well, David,
Nick Jikomes 1:49:41
in the final couple of moments we have here, do you want to talk to people about the title of your new book and what it's all about? Much of what we discussed today is obviously related to the content of that book, but there's a lot more in there. So if you want to take a moment, just direct people that way, I'm
David Anderson 1:49:55
sure thank you. So a lot of these ideas are discussed. In my recent book, which is called the nature of the beast, how emotions guide us. It's published by Basic Books It was published in March of this year. And the book is really focused on answering the question, how can we study emotions in animals? And what will it tell us about emotions in humans, and about the basis of mental illness? And will that help us develop new treatments for mental illness, and it focuses on work done in the areas that I know about, which are in fear, anxiety, aggression, mating behavior, and mostly worked on in mice and in flies, and it is emphatically not a book, like the Soul of an Octopus. It's not a book where it encourages you to attribute emotions to animals by anthropomorphize anthropomorphizing them, it's really describes a, an approach to identifying emotion states, and studying the parts of emotion states that we can get an objective grip on, in an animal in a mouse or even in a fruit fly, and to figure out how we can use that knowledge to translate into better understanding of human brain function eventually, and also mental health. So it is, it's it's really, almost a story of a scientific journey, and how thinking about a problem has evolved. It's not a light read. But I think if you're interested in brain science, and you're interested in how scientists are thinking about these tough questions about what's emotion, what's feeling, can we study? What can we study in animals? What can we not? How do we study them? And what does that mean, for developing treatments for mental illness? I think you'll find a lot that will be very thought provoking, at the least. It's like, like many things in neuroscience, it raises more questions than it answers, but it's at least trying to convert unknown unknowns into known unknowns. And that will sort of put us on a more direct path towards figuring out things like well, what is what is insulin for the brain, analogous to diabetes for the pancreas, if such a thing exists?
Nick Jikomes 1:52:52
Well, yes, I definitely recommend the book. I read it in the past few days. So I'll put a link to that in the episode description. And Professor David Anderson, thank you very much for your time. Thanks very much, Nick. I appreciate the opportunity.
Transcribed by https://otter.ai