Full auto-generated transcript below. Beware of typos & mistranslations!
Sergiu Pasca 4:50
I'm a professor at Stanford University. I'm a neuroscientist physician by training. And my interest is in understanding the rules that underlying the development of the human brain and how disorders of the human brain arise.
Nick Jikomes 5:06
And so what is what's the significance of actually studying brain development in the sense of like, you know, there's so much that we don't know even about the adult brain? What's the significance of studying how it develops? Why not just focus on studying the fully mature brain?
Sergiu Pasca 5:24
Now? Well, I mean, I think, you know, to some extent, the majority of disorders of the human brains are disorder, this whole development. And these days we're finding and even some classic neurodegenerative disorders that have a late onset have a much earlier neurodevelopmental start, that even classic disorders such as schizophrenia. You know, we now know that the disease starts very early in development. So it's, it's very clear that understanding how the human brain builds itself. And how it does, of course, in the context of the genes that are associated with with disorders, it's quite important to understand the pathophysiology of this conditions. And, you know, the reality is that most of what we know about the human brain comes from studying rodents. And it comes at no surprise, I guess, to anybody that the human brain has a number of interesting features. You know, we don't have to call them unique features, but it has a number of interesting features when compared to rodents. So there's always obviously a debate, you know, to what extent, you know, modeling the human brain or studying the actual human brain is important to understand some of these conditions.
Nick Jikomes 6:43
It so so actually a very large percentage of brain disorders, where the symptoms start in adulthood, we've actually have learned that the problems are, you know, going wrong earlier in development, even though we don't really see them with the naked eye until much later.
Sergiu Pasca 7:02
Luckily, and, you know, especially when we're talking about disorders that are genetic, I mean, the genetic mutation is there from the beginning, think of ALS or, you know, think of Huntington's disease. And certainly, you know, cellular pathophysiology starts much, much earlier before the clinical onset. And in fact, there's evidence just in the last few years, then even in the case of Huntington's Disease, which is a severe neurodegenerative disorders that usually start in the third or the fourth decade, that the mutated protein actually plays an early role in development very early on not is that important to cause Huntington's disease, we still don't know. But it's clear that it causes defect defects already early in development, then, of course, when it comes to schizophrenia, you know, many of the genes that are associated with schizophrenia are known to have roles in development very early on. There are risk factors for schizophrenia there are associated with pregnancy. And so there are a number of genetic, epidemiological, you know, pathological studies that pinpoint that most of these disorders are starting very early on in development.
Nick Jikomes 8:15
And so you mentioned how a lot of what we know about human brain development actually comes from studies in rodents. And even though rodents and humans are two different animals, and there's many differences between them. A lot of that probably has to do with the technical and practical and ethical limitations around actually literally studying the developing primate brain. Can you talk a little bit about some of the major technological advancements from the last few years or decades that have allowed scientists to actually study the development of human or primate neurons developing?
Sergiu Pasca 8:50
Well, this is the what I like to sometimes call the unbearable problem of the inaccessibility of the human brain. You know, it is very obvious reason, you know, we not only that we cannot go to patients and directly, you know, measure electrical activity in their neurons. But we also don't have access to the human brain for most of its development, right. And I'm talking like the first few years of life when the human brain still continues to develop. So we know actually very little about most processes in human brain development and just think it's not just about how the cells are made. But how the cells migrate in the nervous system, how they form connections, and in humans, and in many primates, of course, why this process takes such a long time. So just think about myelination, or there is evidence that in humans, myelination ends in the third decade, right so and especially in frontal areas, it almost seems as this you know, ability of the human brain to stay plastic, you know, is one of the unique features of our brains just to stay plastic and open to change for much, much longer. And so the question is, obviously, how are we going to understand this processes in a systematic way so that we can ultimately understand the biology of these conditions. And it's always, you know, an issue. And it's always been frustrating to see how most of the other branches of medicine have easy access to their organ of interest. Right. And I think the classic example is, of course, oncology, where, you know, you want to remove the tumor anyway. And so you can bring it to the lab, or you can do a biopsy. And with the advent of molecular biology, we've really seen a revolution in oncology over the last few decades, where we've gone from knowing very little about, like, what causes cancer, to understanding some of the genes understanding the molecular biology, and today, actually, having cures definitive cures for many forms of cancer First, there's still a lot of work to be done. But, you know, 40 feet 50 years ago, this was like unconceivable. And the reason why this happened is because, of course, molecular biology was discovered. But also at the same time, we have access to the tissue of interest. And so that's why, you know, I, I often like to joke that I suffer from an oncology and V syndrome, because, you know, when I was going through medical school, it was so frustrating, just to see how, you know, fast and the advances were in oncology, and just how far we are. And we still are in psychiatry, where to a large extent, we still define psychiatric disorders. Behaviorally, we have no biomarkers, or very few biomarkers, and none of them today in the clinic, none of them are necessary for diagnosing a patient, you don't need a biomarker, all you need is to observe the behavior and put it in a context of a series of criteria that have been developed over the years. And of course, that's problematic, right? Because brain disorders are ultimately disorders, biological disorders, and we want to know exactly what causes them in order to treat them.
Nick Jikomes 12:01
And so what are some of the tools that had been developed and that are being used now to understand things like primate and human brain development in the lab in particular, one thing I'm interested to hear you talk about, is the role of induced pluripotent stem cells.
Sergiu Pasca 12:16
So, I mean, the some of the ways in which we can access, you know, brain tissue or the, you know, the main way in which we could access the brain tissue, you know, let's say 1314 years ago, was to use post mortem tissue, right, which is, which is still one of the ways in which, you know, this is done. Of course, the problem with that is, we don't have that much tissue, you know, you'll be surprised by although there are like banks of brain tissue from patients with autism in order, we're talking about a disease, that it's almost 2% of the population, we actually have a remarkably low number of brains available to study. And of course, even if we were to have more, that brain tissue is not functional. So it's like very difficult, right to study, you know, what is wrong with those cells? Or what are the circuits that are affected? What are the molecules and it's also very challenging to disentangle the history of the disease from the actual core pathophysiological process, most of these patients and again, I keep giving examples about autism, because that remains my main clinical and scientific expertise. But essentially, for patients with autism, I mean, many of these patients have epilepsy. So they'll take antiepileptic drugs for a very long time, they'll have seizures, you know, they'll they'll they'll take a number of medications over the years. And so it's very difficult to disentangle the, you know, the effects of those drugs are interventions, clinical interventions, and on the other hand, the actual pathophysiological process. And so that's why, you know, the discovery of cell reprogramming, which, you know, happened about 15 years ago, was, in my opinion, so revolutionary, and I was like, still doing my clinical training, where Yamanaka described these, you know, incredible ability of taking somatic cells, and these somatic cells, fibroblasts, blood cells, and being able with a trick to, you know, with a genetic trick to push them back in time to look like pluripotent stem cells, and therefore, there were named induced pluripotent stem cells or iPS cells. And of course, in the beginning that was just remarkable and unbelievable to some extent. But over the years, it was shown clearly that this is a robust process that you can take any cell and reprogram it. And of course, the promise of that is that once you make stem cells from any individual, if you were to have the tools to guide those cells, you could start making you can start making neurons from patients. And that's essentially what I decided to do very soon after Yamanaka made that discovery. And I came to Stanford as a postdoctoral fellow trying to make you know some of the first neurons from patients in a ditch.
Nick Jikomes 14:57
So, so induced pluripotent stem themselves, you take a cell from an adult individual, it could be human being. And you're saying that no matter what type of cell it is, could be a brain cell could be a liver cell could be anything, in principle, you can use molecular tricks in the lab to turn it back into a stem cell so that it's then able to sort of develop forward again.
Sergiu Pasca 15:19
Exactly. I mean, you, you know, of course, in theory, you can take any cell, but it should be a cell that is still capable of dividing. So we're very difficult with the neuron to do that, because you know, the cell will have to undergo a number of cell cycles to really be reprogrammed. Not that it would be impossible, but in general, and so skin cells were the first ones to be reprogrammed, because it's easy to take a skin biopsy, or some some cells in the blood that have a nucleus, so it would be very difficult to do like red blood cells, right? So reprogramming, but you can take white blood cells and reprogram those, but yes, exactly in principle, you could take almost any somatic cells that you have, and push them back in time, by just briefly expressing a number of genes associated with pluripotency. And almost through almost like, you know, cellular alchemy, a sales will go back in time. And, you know, half the features of a pluripotent stem cell, which means that they can self renew. So they can divide and make the same cell over and over again. And secondly, they have the ability to turn into any of the three major cell lineages so and other mesoderm activarmor.
Nick Jikomes 16:30
And so if you create induced pluripotent stem cells, let's say you take a skin cell from someone who has some kind of skin disease that they're born with. And you turn the cell into an induced pluripotent stem cell, and then you allow it to divide and turn back into skin cells in a petri dish in the lab, will it recapitulate the disease features from the person that you got the cells from originally.
Sergiu Pasca 16:54
So that was the promise in the beginning when this technology was developed. And certainly, there was a lot of skepticism and I still vividly remember the brutal rejections that I was getting to almost every single postdoctoral fellowship that I apply for at that time, where, you know, most reviewers thought that first of all, it would be like, technically difficult, and even if it was to work, then you will want to recapitulate the pathophysiology of the disease that the cells were just kind of like, on, you know, on their way to reprogramming and then on their way to differentiating and becoming a neuron lose some of the, you know, I guess, like features of disease. And certainly, with the example of the first neuron that we made, we showed that there was like, not actually the case. And, you know, we can go certainly on the details, but we focus on a initially on a disease, where the mutation was in the calcium channel that is present in excitable cell excitable cells. So, you know, unless you were to have heart tissue, or brain tissue from this patient, you wouldn't be able to study the, you know, the disease itself. And so it was, you know, because this channel is simply not expressed.
Nick Jikomes 18:06
And so, you said, so, you have worked with neurons, human neurons that are created from induced pluripotent stem cells, what can you describe the basic process there, you sort of hinted at it, but what, how does that actually work? And when you get the neurons, how are you been able to study them in the lab? Like, what types of experiments can you do?
Sergiu Pasca 18:28
So, the methods have evolved a lot over the past 13 years, since we've, you know, we've been doing this, and certainly they're more sophisticated today. And I'm, you know, we're certainly going into that discussion as well. But in the beginning, it was relatively straightforward. But it was known for a very long time, that if you have a pluripotent stem cell and embryonic stem cell, the, you know, and you let them differentiate, meaning that you remove some of the factors that you keep in the dish to keep the cells pluripotent, if you just remove them, the cells will naturally prefer to go towards ectoderm. And so towards making, you know, neural ectoderm, and, you know, parts of the nervous system and more specifically cortex. And so, in the beginning, the methods that we've had, you know, 2009 were quite primitive, we would get this iPS cells from patients through a very painful process. At that time, the programming was quite difficult and tedious and inefficient, and then take those induced pluripotent stem cells, and essentially remove the factors that keep them pluripotent, and just hope that they will kind of make neurons soon after the methods became slightly more sophisticated and certain signaling pathways were shown to be very important for guiding that differentiation. So for instance, if you block this map pathway from two sides, you can get almost 100% ectoderm of cells. And so, you know, essentially there was work done in learning how to guide the cells to go or go towards a certain phase. And in the beginning, the cells were essentially just sitting on a flat surface, right? So you just like have a dish, you know, with a coverslip and a glass coverslip. And the cells will be there, you'll see the neurons and you will be able actually to study them. First of all, look, do they look like neuron that they have the morphology of a neuron? And then starting to look at functional properties of those cells? Like, do they have calcium signals? Do you know if you patch them? If you put an electrode inside and you listen to their electrical properties? Do they? You know, do they they are they're really neuronal like. And so those are the ways in which we can we were testing them in the beginning. And that was very useful. I mean, it was, it was good enough for us to discover some fundamental. Well, I guess, to some extent, to validate that this method works. And the way we validated it, in the beginning, was we focused on this disease, this rare form of autism and epilepsy called Timothy syndrome. It's very, very rare. But the mutation is in a calcium channel. And this is a voltage gated calcium channel. And as the name implies, this is less calcium in the neuron, when the voltage of the cells changes. So when the neuron receives some sort of input, the calcium channel will sense that open and let calcium in. And the mutation was a gain of function mutation, in the sense that it would let the calcium channel stay open for slightly longer. That was the prediction. But of course, nobody knew because nobody ever looked in either heart cells or neurons from this patient. So what we did is we make neurons from patients with Timothy syndrome. In the beginning, we had two patients, we flew them here to Stanford, they're very few worldwide, by the way, they're about like 50 or so we flew them here to Stanford got their skin cells, generated iPS cells, and then we made neurons. And then the first, you know, test that we did was to look at calcium dynamic inside the cells. And I still remember the day, you know, very vividly, when I did that experiment, which summer of 2009, where we literally put this some of this Timothy syndrome neurons, and we depolarize them and look at calcium and rice. And you can literally see how calcium goes up in a control cell and it goes down, and then it patient cells went up. But then it took way, way longer to come down, which was, you know, was it Wow, in a wild moment for us. Showing that look, you can take the cells from patients, make cells that are inaccessible, and see a phenotype that to some extent, you would expect, but it was like, you know, unbelievable, to be able to do that in human cells. And I guess
Nick Jikomes 22:55
that. So you have this rare disease, a rare form of autism, it's got a genetic basis, there's something you know that there's something wrong with a particular type of calcium channel. But you could never confirm exactly what was wrong, because there was no way to go look. So what you guys did was literally fly in humans with Timothy syndrome, took some of their skin cells, turn them into stem cells, turn the stem cells into neurons. And then you can actually measure what was going on with these calcium channels and actually see exactly what was wrong.
Sergiu Pasca 23:27
Exactly. That's exactly, that's exactly what we did. And we started using it for multiple other disorders very quickly, this technology, but we started to realize that there were some limitations. And one of the main limitation was just like, simply how long it takes for the human brain to develop. So just just think about it. If you were to think about making the cerebral cortex, the outer layer of the brain, the one where cognition resides, I guess, like most of our work associated with, like most of what makes us human. But you know, if you were to look at, like how long it takes to build, to make all the neurons in the cortex of rat, you realize it just takes about a week, you know, maybe seven days in mice, 11 days in rats, or something like this. Now, if you were to look in humans, in order to make all the neurons that were we have in the cortex, you need 146 days, 150, let's say, right, that's a very, very long time, we're talking, you know, almost 20 times longer have a period to do it. And we realized very quickly, that we were just simply not able to keep the cells in a dish to get to those stages, you know, because you make the cells at the bottom of a dish and you keep them there for a week or two for five or six or seven, but at one point, they peel off, they come off of the 10 centimeter dish, and you know, through heroic effort, you try to replay them everything else, but neurons are not happy when you just move them between the plates. So no matter what we did, usually by the time they would Each 100 days, the neurons would be dead. So they would not even finish making all the cells in the cortex. And that's just the beginning of studying human brain development, you know, because themselves start to make astrocytes, glial cells, which are also important for disease, they start to migrate, they start to make connections. So it's just clear that we're just opening a window into the early stage of development.
Nick Jikomes 25:21
But it sounds like so if I'm hearing you correctly, you have this key limitation, which is you can't have these developing neurons in a dish for very long for more than like 100 days, because they just they're unhappy, and you can't culture. But within that time that you can study them, it does sound like they start to form structures that are not unlike what they actually do in, in actual development. They don't
Sergiu Pasca 25:43
have much, you know, they don't have much site architecture, you know, they, you know, the progenitors will start to organize, almost like an event chicken or like zone, but afterwards, there's not enough structure for them to organize. And they're, again, they're at the bottom of it, there's, they're on a flat surface, which is obviously not what the brain is. And so then, you know, more than 10 years ago, now, you know, sometimes around 2011, or so, came up with the, you know, very simple idea, I thought, Oh, why are we even keeping them, you know, attached to this dishes, why don't we just simply move them in another plate where they cannot attach, or they just be floating. So, you know, I bought this cell culture dishes, which are coated, so that sales don't sit down, you know, most of the plays that we use in the lab are most of the petri dishes that we use, are coded for the cells attached to them. But you can actually buy cell culture dishes that have a substance that prevents them from attaching. And so essentially, just use those and aggregated the cells. So make tiny, you know, spherical bowls of sales, and just put them onto those plates. And to be honest, I thought, they're just going, I thought, we're going to keep them for a week or two, and then they're going to disintegrate, like the balls of sales were like, kind of grow. And at one point, they're just like, breakdown. And the surprise was that it was not happening. In fact, they were very stable like that they grew up to three to four millimeters in size, so you can literally see them by eye, and then you could keep them and now we know, we can essentially keep the cells indefinitely, you know, we've kept the longest cultures that have ever been reported going, you know, 800 900 days, in a dish, so several years, but you could go for even longer than that, in principle, there is no limit to how long you can keep them. And we call those initial preparation steroids because there are spherical, but as you know, there are multiple efforts into building three dimensional cultures across multiple fields, some with using extrasolar matrices, some with like, you know, plating and replacing the cells, you know, slowly, the sculptures were known as organized because they resemble organs. Again, not the most, you know, fortunate name, I would say, because it implies that we're making an actual organ, but you know, that's the term that is used today. So this culture is we're collectively known as neural organoids, or brain organoids.
Nick Jikomes 28:09
I see these, these are neurons that you're culturing in a dish, and they're essentially just sort of floating suspended in liquid,
Sergiu Pasca 28:16
is that good. So they're, they're literally floating there bowls of cells that have, you know, maybe one to 2 million cells. And but the key aspect of it is that they're developing, they're self organizing. They're not just like cells that you aggregate, right, you take the stem cells in the beginning, you guide them just slightly, and then you let them do their job. And essentially, you will have, you know, a first wave of progenitors, those will generate other progenitors, those progenitor will start making neurons, at one point, progenitors will start making glial cells. And so there's development that happens in the dish. And now we could actually keep them for 150 200 days. And when we look inside, we realize, actually, cortical Genesis or the generator generation of the cerebral cortex does continuing. And in fact, what we've discovered, you know, over the past five, six years, is that, quite surprisingly, I would say that, it seems that the cells have a way of keeping track of tracking time of, of, of knowing where they are in development. So, you know, the experiments go something like this, you take this cultures, and you keep them for, let's say, 800 days or 600 days, and you keep collecting cells, you know, organs at multiple time points. And then you ask, Well, what did they resemble the most in human brain development, what stage and if you take them into the beginning, in the first couple of 100 days, it's very clear that they resembled the first trimester, second trimester human brain. But then something happens and when they reach nine to 10 months of keeping them in a dish, so counting the time of birth, they transition to a postnatal signature on their own of course, there's no birth not that birth would mean anything. It's just that it tells us that they progress at a certain pace that is similar to the pace in vivo that almost, if you want to say, you know, if you want to call it, they have a there's a clock, a brain clock that keeps track of time in the cells, and even very canonical switches that people have discovered, like, people have its classic examples in neuroscience, that there are NMDA receptors that have certain units before birth, and certainly units after birth. And it was unclear whether those are caused by activity by birth itself. But if you look in a dish, they will transition. And they will transition around nine to 10 months in a dish as well, which tells us that they're part of some sort of intrinsic developmental time tracking molecular machinery, that tells us tell us exactly where they're in time. And of course, nobody knows what the molecular basis for that clock is.
Nick Jikomes 30:55
I see so so we know that when you go from prenatal development, to postnatal development, so a human being is born, comes out of the womb, and there are changes in the brain that happened around that time. And 111 idea would be well, those changes are driven by the fact that now the baby's eyes are open, and there's information going in through your eyeballs, and you know, their sound. And there's all these experience dependent components to now being outside the womb. In the external world, what you're saying is that you see certain molecular changes in things like NMDA receptors in the brain, even in these neurons that are just floating in the dish.
Sergiu Pasca 31:31
Well, there is a fundamental program of maturation that is kept. Not every aspect of maturation is recapitulated. In addition, in fact, that's prompted the experiment that we've been doing and be publishing a few months ago, of trying to do some of these experiments in an in vivo setting precisely because some features are missing. And, of course, we can talk about that as well. But there is a basic fundamental program, the cells will know what to generate, and when. And this is species specific. Because chimp cells, if you take chimp iPS cells, and you do exactly the same experiment, they'll finish cortical Genesis much faster, or myself, they'll finish it much faster. So this is a species intrinsic clock that, you know, maintains the pace of development. But then, of course, and for us, of course, birth doesn't mean that much, right? I mean, certainly there are like some dramatic events that the brain the human brain has to deal with, right, like in terms of oxygen, and so on, so forth at birth. But in the grand scheme of things, if you think about it, it's you know, our birth is premature, right. And it's dictated to a large extent by the size of our of our head. And so that's why humans when they're born there aren't capable of walking, we're not yet myelinated. In fact, we now know that even interneurons, continue to migrate for about a couple of years after we're born. So it's not like development is done. And then so I think it's worth keeping in mind that, not that there's necessarily something special just about birth, there is that is not just birth, it is just the timing of development, that the cells will be able to just like track over time.
Nick Jikomes 33:13
And so this is an intrinsic thing. So that means it's largely determined by genetic factors.
Sergiu Pasca 33:18
I buy some genetic programs. Yes, absolutely.
Nick Jikomes 33:22
And how much how hard you guys have to work to mimic the environment. So when you're growing these organoids, in a culture, you have to work very hard to mimic the in utero environment in terms of all of the molecules and stuff floating around?
Sergiu Pasca 33:37
Well, I mean, so the way we do it is, you know, it depends a little bit on the brain that you're making, of course, we put them in cell culture media that provide some of the nutrients that the cells have, and there's a lot of work that needs to be done in improving some of these medias. But one of the things that we do have to do is we have to provide guidance to make very specific brain regions, meaning, meaning that you can essentially, you know, if you want to make autonomous or if you want to make a spinal cord, you have to provide very specific growth factors and molecules, and that we add very early in development that we spend a lot of time figuring that out. And but, you know, other than that, you know, we don't do that much. We don't provide a scaffold for the cells, we don't you know, we don't keep coming and instructing them. We just let them develop. And I think that's the beauty of development right? I mean, the human brain builds itself to large extent. And so once you provide some conditions for for the from brain development to take place, it will take place within those constraints, of course, and just to make it clear, the cultures that we make resemble specific regions of the nervous system, so they can resemble the striatum or the cortex. They're not miniature versions of the entire brain, which is You know, it's certainly a very important nuance, especially when it comes to misunderstanding what is being done.
Nick Jikomes 35:07
So, you know, this is obviously, you know, we mentioned how a lot of what we know about human brain development comes from studying rodents. But obviously, we can never understand everything by doing that, because we're just different from rodents, our development takes longer, et cetera, et cetera, our brains have very different features, that that end up developing, what are some of the major anatomical and cellular level ways that the human brain differs from rodents and non human primates, or I
Sergiu Pasca 35:33
mean, there, there are many features that have been described over time. And, of course, one of them is even just simply the timing of development, it takes much longer to make a brain of a primate than to make one of our broden, for instance, there are different types of progenitors. And the proportions are very different. So there are types of progenitors that are over represented in the primate brain, they're presumably responsible for the many neurons that we have in our in the expended cortex that we have, there are differences also in the timing of the formation of synapses, and even, you know, you know, observations about differences in the number of synapses for cells, but again, it is still very early days, truly understanding of the molecular cellular level, what are the differences, I mean, just in the last five years or so, we've been able to have atlases of the human brain at single cell level, right, so many of these comparisons up to this point, were done by morphological analysis, right, by essentially looking in tissues and comparing, but at the molecular cellular level, that has not been done. And functionally, there are also a few studies that have been done in the last few years that also show that some of the electrophysiological properties of human neurons are different, or at least of some of these neurons, they're different. But again, a lot of work needs to be done in this area, because the main limitation has been lack of access to the cells.
Nick Jikomes 37:04
So so a lot of the difference between humans and other mammals in terms of their brains is we have more neurons than than most other species that are presumably, physically connected to each other in somewhat different ways. Do Are there any like types of neurons or cells in the human brain that just aren't present in other mammals.
Sergiu Pasca 37:22
So we're not necessarily the species with the most cortical neurons. You know, there are other species that have, you know, way more neurons, dolphins, for instance, like have more neurons in the cortex. And so it's, it's not just about the, it's certainly the number of neurons will be important, but we should not just presumed that it's just a number of neurons is probably the way they're connected to each other, which makes a big difference. And certainly, there's been a big, you know, a, they a cert that has been going on for a long, long time to try to identify human specific or primate specific cell types. At this point, you know, certainly there are a few cell types that have been presumed, like the classic neurons, pyramidal neurons in the frontal cortex called foreign economic neurons, which were described a long time ago. You know, they do seem to be enriched in some mammals, but not in others. But again, it's very difficult to know exactly what they do. And there's not a cell type that is only present in humans. And so it may be that some of the cell types may have appeared multiple times during evolution. But there are, you know, there are other differences that are there are rising, for instance, you know, the ratio between excitatory and inhibitory cells in the cortex seems to be different in primates. And so for instance, if in rodents, for every single, you know, pyramidal cell, right, every single, you know, there's like one GABAergic inhibitory cell for every three pyramidal cells. So the ratio is like four, four to one three to one. In humans, it seems that it may be one to one. So it almost as if we have more GABAergic cells in the cortex. And that's obviously very important that that's one of the aspects that we can now study in addition, as well. Because what is interesting about this interneurons is GABAergic cells, which essentially put a brake on activity in the and refine activity in the cerebral cortex is that they're not born in the cerebral cortex. They are born actually in a deeper side of the forebrain. And then they have to migrate all the way up to the cortex and populate the cortex. And they do so over many, many, many months. They start that like mid gestation, and we now know that in humans, this continues up to the second year of life. And so to study actually this process, six, seven years ago, we introduced an approach to model migration of sales and connectivity that is called an assemblage. And so, in an assembler, essentially, you make these two brain regions separately. So you make a Cortex, and you can make a ventral forebrain. And at one point, you just essentially put them together. And, you know, see what happens. And it's interesting. I mean, it's always funny to look back of how we're developing this, but this was one of the goals of my lab, when I was starting, you know, we were, you know, we thought that it was going to make these brain regions who will be very difficult to fuse them that they want to stick to each other. So we started thinking of all kinds of like nano electrodes and biological glues that we will like put to link the two together. And so somebody in the lab came, said, look, it's very simple. You take the two, you put them at the bottom of an Eppendorf tube, you leave them there overnight, the next day, they're going to be fused. And not only they were fused, but you could literally see in there are starting to migrate and move. So this inter neurons will start to jump, because they're not crawling on a surface, but they do this peculiar jumps and moving populate the dorsal forebrain. So again, seeing that a life also in the lab was incredibly rewarding at that time. And of course, we've been using it now systematically to study disease. And in it's very well known that, that what, well, it's a very well known hypothesis that in autism interneurons have some sort of contribution. There is a so called excitation to inhibition hypothesis of autism, that claims that an imbalance between excitation and inhibition is was what deals to disease. And that's based on the fact that many patients with autism have epilepsy. So they have seizures, and, and a number of other observations. And so in a very recent study, which is still not not published, we've actually taken a large group of genes that have been associated with autism and other neurodevelopmental diseases, about 604 100 of these are expressing interneurons, and try to map them to see which ones of those genes contribute to stages of internal development in human cells. So essentially doing a CRISPR screen of stages of development that allowed us to pinpoint that about like 40 genes or so out of all this large group of autism genes that are interfering with inter neuron degeneration, and some of them were certainly very surprising and unexpected genes associated with disease that interfere with continual migration. But I think it speaks to the fact that we're finally getting at a time in neuroscience, where we have the ability to study neurodevelopmental processes in human cells, and even start to map some of these decisions, their contribution to the stages of development, of course, was, you know, difficult to imagine 1520 years ago.
Nick Jikomes 42:54
So you have the ability to create neurons from stem cells, you have the ability to grow them in a petri dish, so that they are forming these things called organoids, which are basically, you know, balls of cells that start to resemble at least in some ways, different parts of the brain, you mentioned these things called assemblages. So basically, you take two organoids, you put them together, and they start to grow together in ways that maybe mimic the way the brain normally develops. In some ways. You also did some recent experiments, I got a lot of attention where you were actually transplanting some of these organoids into animals, can you explain those experiments and why you did them and how you did them.
Sergiu Pasca 43:34
So certainly, with assemblages in Oregon, we've been, you know, able to model, you know, aspects of development that, you know, we never thought that we could model before including some early features of circuit formation. And, you know, I didn't have the chance to mention this. But, you know, for instance, we've tried, we built the entire cortical spinal tract pathway from parts. And so you know, within an assembly, you know, that cortical neurons project all the way to the spinal cord, connect with motor neurons motor and control muscle contraction. So, in a paper that we published a few years ago, we were able to make a cerebral cortex make a spinal cord like organized and a muscle, like organ, a muscle have literally a ball of human muscles, and then connect them all three together. And although we don't know the rules for the cells to assemble with each other, you know, we found that a specific specific types of cortical neurons will project all the way five motor neurons connect with them. And then motor neurons will go and stimulate muscle contraction. And you can literally have a circuit and stimulate the cortex optogenetically or electrically and trigger muscle contraction. And so although we've been getting more and more sophisticated, in terms of the circuits that we can build, it was also clear that there are a number of limitations moving forward. One of them I mean, there are two in particular there are many but but let me just focus on two Two major limitations that are perhaps more straightforward to articulate. One of them was that no matter how long we kept the cells in a dish, and despite the fact that they were undergoing this, you know, there were going through phases of development, as you would expect, the cells did not have the morphological and electrophysiological complexity that you would expect. And so even when you kept the cells for 250 days in a dish, they were still not at the size of a neuron, in the actual human cortex at that stage. And if you were to patch them, you know, the resting membrane potential would be like minus 50, but not minus 70, as you would expect. So it's like clear, you know, that something is missing, that, although they know where they're on development, there's a, you know, a secondary program that is not being activated. So that was part of the motivation, where we thought, well, maybe, you know, we can provide that in a living system in Iraq, if we were to transplant and integrate them, maybe some of those features will be provided there. But the other limitation, which, you know, of course, has been in the back of my mind for quite a while is that, in psychiatry, we define disorders. Behaviorally, as I was mentioning before, right, there are no, there are no biomarkers for these conditions, there are no cellular phenotypes to define them, you essentially look at the behavior of this patient. So when you find a cellular defect for a form of autism in a dish, how do you know that that is truly part of the you know, the core pathophysiology of that condition? Right? How do you know that that phenotype is, for instance, not compensated by at the circuit level in a living animal? Right. So I think it also becomes much more clear that we will need better ways of assessing whether there are consequences for the seller defects that we will see. And you know, just to give an example, not that this will be much easier. But in neurodegeneration is it's it's slightly more straightforward, in the sense that we know and we've known for a very long time that in Parkinson's disease, we know there's a loss of dopaminergic cells, what is the cellular phenotype phenotype loss of dopaminergic cells, I mean, that was seen in the 19th century, where you would do post mortem studies, and you will look and see substantial Niagra, which is supposed to be Black was gone, it was clear something was missing. But for autism, for schizophrenia, for intellectual disability very often is not clear what we should be looking for. And if we were to find something, how do we know that that truly causes disease that we should restore that phenotype in order to have a treatment? And, and, and it's not that this needs to happen? This is not necessary, perhaps. But it was something that we should consider moving forward. So until that there's two, there's two aspects, how do we get the cells to mature farther and integrated into circuits? And could we have more complex behavioral readouts and circuit readouts for the cells? And so that's why, you know, a number of years ago, and this this project started in the lab, you know, almost seven years ago, we started thinking, could we actually put some of the cells in the rat brain, but make sure that they integrate into the circuitry of the rats so that they receive meaningful input. So we thought about putting them in the somatosensory cortex of the rat, this is the part of the brain that receives information for the whiskers on the opposite side. Now, they're critical periods for the development of these regions. And that's very well known in the rap. That neurons in the thalamus that receives information from the whiskers will project into the cortex up to a point in development, and after that, they don't care anymore. So the critical period was close. So then one of our goals was to try to put human neurons before that critical period would close in the rat, which is after about a week that the rats are born. So essentially, we took immunocompromised rats, so rats that cannot reject human cells that would be transplanted, and then transplanted, and intact, organized right into the somatosensory cortex of the rat, and then closed and waited. And we discovered very early on that we could monitor the graft, the human organ that has been transplanted by MRI. So you could just run an MRI, and you can see very clearly the border of the organoid. And that allowed us non invasively to optimize the procedure, the surgical procedure and getting better, better at like how we run this experiment. So we discovered that if you do that, the graph grows relatively large within a few months. And in the end, after about four to five months, it covers about it represents about a third of a rap hemisphere on one side. Again, we're not removing the rat cortex, the rat cortex is still there. But the human cortex will grow within the rat cortex and ultimately be about 1/3 of the of the volume.
Nick Jikomes 50:08
How big is it when you first put it in? Well, it's
Sergiu Pasca 50:11
relatively tiny. I mean, it's, you know, less than a half a millimeter, or half a millimeter to a millimeter when you put it in, but then it grows because it becomes vascularized. Well, now we know. So very quickly, blood cells from the rat will grow in. And that's why it survives, otherwise, it would not survive. It's primarily human. So you can almost think about over a unit of human cortex into the rat brain, a 99% of it being human, the 1% of the cells that are in there are either microbiota from the rat that would come in and kind of populate the graph, or some of these endothelial cells, because again, the graft gets vascularized. And so one question was, well, does it is it connected to the host in any way, and so we use a rabies virus, to see what are the sources of input into the human graph, then we look throughout the entire brain, and found that there are two sources of input. One is the surrounding cortex. So the cortex of the rat on, you know, surrounding the graph will project inside. But then the main source of input is actually from the thalamus on the same side, which again, conveys information from whiskers. And so that prompted us to ask Will, would they actually respond to sensory stimulation. And so to do that, we put a calcium indicator into the human graph before we transplanted, waited for the graph to grow about 250 days or so. And then you can open the skull and image live the activity of human neurons, and then you can puff air onto the whiskers of the rat onto the opposite side because the pathway is crossed. And that we have shown that indeed, if you pass err, human neurons would respond, the responses that come right after that, telling us that they have integrated, you know, to a sufficient demand as to be able to receive input, which, again, is primarily coming from the, you know, from from the Rat Pack, but they're not building an entire human pathway, just to make it clear, it's just the cortex is now starts to receive input from the thalamus.
Nick Jikomes 52:14
So just to make it super clear for people, so you're grafting one of these neural organoids, composed of human neurons into a rat brain? So what exactly are you guys looking for? And what is that what is the purpose of doing that is to see if the neurons will integrate in a way that resembles ordinary rat neurons that would be growing in the cortex, that they are actually sort of finishing development, so to speak.
Sergiu Pasca 52:39
So the main goal was, you know, the cells that we were growing in a dish, were certainly not sufficiently mature in terms of their morphology in electrophysiological. properties. So the question was, you know, they're grown outside of the human brain, if they were to be grown in an environment where they receive actually input, the electrical input, would they actually develop much more, you know, you know, develop better features. And so the goal here was to integrate them into their nervous system, so that they can receive some of that sensory input early on. And you know, there are classic experiments, classic experiments done here at Stanford, by Carla Schatz years ago, were, you know, it was very clearly shown that if neurons in development do not receive input at the right time, you know, they won't mature, they won't develop, they won't develop the conductivity. And so, of course, in the dish, they wouldn't be unable to. So one question was, well, maybe they're not developing because they're not receiving, you know, the right sensory input or the right input at the right time. And, you know, just to make it clear, the integration with the rat brain is limited, in many ways, because the rat develops much faster than the human, we've discovered that even when we put human cells into the rat, they will still develop at their own pace. And so you can look at almost, you can almost look as a, at a competition, the rap, you know, finishes cortical Genesis in a couple of weeks, it's done, the human will take much, much longer, so the integration can not be perfect. And I think this also speaks to some of the ethical aspects, the integration cannot be perfect, but we wanted to see could we get enough integration, so that human neurons would receive input and perhaps mature much more. And indeed, that is the case, because you can take the graft out, after it has been there for 250 days. And when you look at neurons, now, they're about six times larger in size than they were before the resting membrane potential is much closer to minus 70. So that tells us that there are some factors inside the rat brain that are helping the cells, the human cells to mature, you know, farther in development, and in fact, we've shown that this helps us already to model disease. Just, you know, in, again, in a genetic form of disease, we've shown even in this paper that we were unable to, to identify differences in a dish, no matter how long we kept the cells, unless you were to put them in vivo, and the cells will grow much larger. And then you would start seeing the differences between patients and controls. So I think this tells us that there will be a, you know, a certain group of phenotypes or disease related processes that we may only uncover if we embed the cells into circuits, which again, it's not surprising, right? It's clear that that will be the case. But it was interesting to actually observe it firsthand.
Nick Jikomes 55:40
So, you know, when you do these transplantation experiments using these organoids, all the techniques that you've been talking about, do you see these primarily as tools to understand how development naturally unfolds? And to figure out the pathophysiology of different diseases? Like what goes wrong in development to give you something like Timothy syndrome or or another developmental disorder? Or do you see a future where some of these tools are actually used themselves in a clinical setting? Like, for example, could you imagine doing human to human grafting in the future to integrate healthy healthy neurons and healthy circuitry into someone that has some kind of developmental defect?
Sergiu Pasca 56:22
Well, I think at this point, the main utility of this platform is to model disease and to test drugs. You know, because you can think about today, if you know, first of all tomorrow, this is I was mentioning, but also think about like drug development, right? Let's say you have a potential therapeutic for a disease, right? What do you do next? With it, like, if you want to move it towards the clinic, you want to test it into an animal model. And sure, you can do that. But sometimes some of these animal models for disease don't recapitulate the disease. And so you still want to test it in an in vivo setting, you want to make sure that it actually works. So your your, your best next choice is actually to use a primate, right, a primate model, which, again, is limiting for many, many reasons. Well, the alternative with this method is that you can actually use this model where you inject the drug of interest or the therapy of interest into the animal, but you will look at the effect on human cells that have been transplanted into that, that that cortex, and we still don't have the ability to look at very complicated behavioral results. And I'll tell you what we've done so far about that. But it shows that this is, and we've already shown that actually, we have one conditions that we've been studying where we've done exactly that, and a promising therapy that have been working throughout like all the in vitro assays, and you wonder what is next, you know, what else should you do before you move into patients, and you want to do as much due diligence as you can, to show that this works into any viewer setting, you know, how many desert you know, how many, you know, rescues that are in feature that never translate in vivo? Right, as people say, in vivo Veritas, you know, ultimately, you know, truth is in an in vivo 70. And especially for circuits of disease. And that's exactly one of the applications of this platform that you can just test drugs, have cell transplantation. You know, I'm not sure whether that's the immediate application. And because, as I was mentioning before, if you transplant the cells at much later stages of development, integration is not ideal. You know, the host brain is no longer permissive. And so we will not make as many conditions as many connections. So you wonder, if you were to transmit sales after a stroke, you know, how much integration would you actually be able to get? I think if as as we're getting better and better at understanding how the sales get integrated, perhaps we're going to be able to play with some of these programs and open briefly periods of but that I think it's, it's it's a longer effort that will take many years to get there.
Nick Jikomes 58:51
I see. So, so doing these sort of grafting and transplantation experiments. One of the things one of the applications here is that it will allow scientists to sort of study the effects of drugs or interventions that are meant to be used ultimately, in humans, in human cells, even though it's in an animal model. So it's sort of the idea would be that it's going to tell you more than just putting in the animal model all by itself. But you could do this sort of faster and more rigorously than you could in actual human bigs. Exactly. Yeah. And so what are, you know, what are some of the major questions that your lab and similar labs are working on right now? What are some of the big questions in development, brain development that you think we're going to start to learn some answers to in the coming years?
Sergiu Pasca 59:40
Well, you know, building upon the discussion now about transplantation, you know, one fundamental question is like, why are their cells maturing much more in vivo? These are not trivial experiments to do and requires hundreds of days, a lot of expertise. And well, I'm sure that many labs will be able to do this. I don't think it will. it'll be a platform that hundreds or 1000s of labs will be able to, to apply just because of the expertise and the timing. So what we're trying to do now is to understand what are the reasons and the programs that underlie this fast maturation and then try to mimic it, you know, in an in vitro setting to try to obtain similarly mature neurons morphologically and functionally, but in an individual setting, I think that that is that I think it's a fundamental question, how is that happening? How do cells mature, and I wouldn't be surprised if some of the genes that regulate that are probably disease genes, because very often, we see this maturation effect in many of these conditions. So it's very likely that the same genes that control this are probably also causing disease, at times. And so I think that is one aspect. But, you know, going back to the discussion about the behavioral readouts of these conditions, I think it's worth keeping in mind that this psychiatric disorders are behavioral conditions, and they work primarily at the level of circuits. So we're going to need to be more sophisticated in how we're reading the effect of the cells. And in this paper, we started doing some experiments in this direction, trying to see whether human cells participate to the behave, can participate to the behavior of the rat, because we obtain a behavioral readout. And what we've done is, we've essentially put general rhodopsin, which is you may know a tool developed, pioneered by my colleague here at Stanford, Karl Deisseroth, where you can control the activity of neurons will blue with blue light. And so essentially, we put general adoption into human neurons. And then transplanted them waited for, you know, 150 days or so implanted an optic fiber that allows us to deliver blue light to the cells, and then train the rats. In a reward conditioning test. Essentially, we were teaching them to associate stimulation of human neurons with delivery of over reward, in this case, receiving water. And we shown that if you train the rat, you know, after about a couple of weeks, you can just, you know, stimulate human neurons and rats will go and seek the reward, in this case, water, telling us that they were able to actually associate human neurons participate that circuit, but now, just imagine, if you were to have, you know, two rods, one carrying cells from patients with a specific disease, and others with the control, would we be able from a behavior or tasks like the one that I described right now to distinguish which one cares, patient cells and which one cares control cells? Or to almost, you know, you can look at almost like, in this case, the animal as an avatar for the human cells as a readout as a complicated and complicated readout for the defect in human cells? And of course, then could you administer a drug and restore it, so you don't see that difference anymore? That's the promises for this as as we're dealing with evermore complex psychiatric disorders, right, because there's disorders that are more complex than others that are forms of intellectual disability that are caused by microcephaly where neurons are simply not there. You know, and that is relatively, you know, straightforward to understand. But schizophrenia, and other conditions are not that straightforward. We, you know, they're not major cells as missing from the brain, it's clear that the defects are probably in the way the cells are communicating with each other. And so we need to build more complicated readouts for that. So the
Nick Jikomes 1:03:31
idea would be you, you literally graphed two different types of neural tissue into something like a rodent. And, you know, one would be a control tissue, and there's nothing wrong with it. One would be something like, you know, maybe the neurons that you that you made from the patients with Timothy syndrome, and they've got a defect and a calcium channel, you allow them to integrate into the mature brain of something like a rat. And then you could actually look at not only how those cells integrate into the brain, but if there are behavioral readouts, that you can start to detect that point to something being wrong, in other words, that the the rat behavior sort of recapitulating some kind of human behavior that you know, goes wrong in some of these disorders, you can then go in, if that's true, and actually functionally characterize hopefully, exactly what's wrong with the cells.
Sergiu Pasca 1:04:21
Exactly. That's sort of like the promise if there's gonna be a lot of work to get up to that point. But I think this type of work is justifiable considering, you know, the immense burden that psychiatric disorders bring today, you know, like, almost one in five individuals suffers from a psychiatric disease. Most of these conditions are chronic conditions, lifelong, lifelong conditions for which we essentially have no cures. I mean, we have a number of classes of drugs that alleviate some of the symptoms, but to a large extent, this disorders are still mysterious and uncurable. And you If we think that psychiatric disorders are, you know, uniquely human, and there may be or they may not be, but certainly, if we think of them as related to some of the unique features of the human brain, it becomes very clear that we need models that mimic more closely the human brain. And certainly, the more human to models are, the more uncomfortable, we feel so about, you know, the meaning of those models, and, you know, their properties, and so on, so forth. And, and that's where some of the ethical, moral, legal implications of this type of work arise, which, you know, we've been very sensitive to and very actively involved in and proactively involved in outlining, what are some of the, you know, what are some of this ethical, moral and legal, you know, features of this type of work?
Nick Jikomes 1:05:51
Yeah, and what do you think some of those features are? Or to put it a slightly different way, some of your recent research got a lot of attention in the press, and on social media and stuff, and, you know, especially to non experts sort of reacting to this, you know, and seeing like, Oh, we're, you know, people are grafting human brain cells into rat brains, it starts to really sound like science fiction type stuff. How has that whole process been for you? And what are some of the biggest sort of misunderstandings and things that you've seen out there in terms of public reaction to some of this work?
Sergiu Pasca 1:06:24
Yeah, well, I mean, we've been thinking about this for a long time. And I think it's very important to know this experiments are not done. And then we look at them. And we think, Oh, what have we done? That's not how it works. You know, we've been both working with the, with the larger community of ethicist. And philosophers, actually, who are thinking deeply about this, as the experiments are ongoing. And even here at Stanford, I have colleagues in part of the Stanford brain Organogenesis, program for like, you know, thinking about every single step, they see this experiments, and we think carefully about what the next steps are. And in general, there are at least like three, I guess, categories of issues that one could think about. So one of them has to do with the animal. And the welfare of the animal, you know, is the animal suffering because of this transplantation? And you know, that we've looked very carefully from the beginning, because they have a large graph. And so we've looked at, you know, do they have epileptic seizures? Are they're under any pain? And, you know, to the extent that we've seen, we have not seen that, and another, you know, issue related to this is, are there any augmentation of like, features in this animal, right, like any changes, and to be honest, we were not thinking that the ROB, you know, will just, you know, have any, you know, novel features to improve their performance, we were more more afraid that they will be, you know, will perform worse in some of these tasks. So, we ran a battery of cognitive, you know, attention, anxiety tests, and found those differences between rats that were transplanted, and rats that were not transplanted. And, of course, we continue to do that for all the experiments that we perform. But, again, you know, just to make an analogy, which may not be an analogy, a perfect analogy, but you know, if you, we take a car, you take a Prius, or whatever you want, and then you change a part of it, by taking, you know, whatever you want out of a Ferrari, you know, the Prius will not become a Ferrari by just changing a piece in the entire car. Right. So there's, you know, we're not humanizing. And that's, that's one of the things that we prefer not to refer to the term humanization is not a very good term to describe this type of work. But we're just putting a few million cells into this rat in order to mature them and to study the disease. So one aspect has to do with like the rat, and then another has to do with the cells, the human cells, you know, were the human cells consented appropriately, like those who donated the cells were they aware of the fact that we be the cells will be transplanted into an animal and in general, you know, the constraints that we use specify that, you know, the cells will be transformed into other cells, and they will be transplanted into animals. And the third aspect has to do with the perception. How is the research presented? This can quickly capture people's imagination, right? And so one has to be very careful what type of words we use to even describe this type of work, right? And so that clarity, of clarity of the message is essential is essential to explaining the work and so that's why we put a lot of effort into explaining the work carefully to the public as well and understand exactly what has been done before there will be any exaggerations.
Nick Jikomes 1:09:58
Are there any final things that you Want to reiterate or let people know about your work and and what you've been researching?
Sergiu Pasca 1:10:06
I think we've covered a lot of ground. Yeah.
Nick Jikomes 1:10:09
Yeah. And so we covered a lot. Thank you very much for your time. I really appreciate it. This is very interesting research. For those listening. I did a previous episode with Karl Deisseroth, who we mentioned earlier. So if you're interested in optogenetics, and understanding that technology and how neuroscientists use it, and a little bit more detail, check out that episode. We've done a number of episodes about brain development with a variety of people but but obviously, this episode is in in that cluster of, of conversations. Sergiu Pasca, thank you very much for your time. This has been fascinating.
Sergiu Pasca 1:10:42
Well, thank you so much, Nick, thank you.
Comments