# CARTA: Neanderthalizing Brain Organoids with Alysson Muotri

## Метаданные

- **Канал:** University of California Television (UCTV)
- **YouTube:** https://www.youtube.com/watch?v=RZkQ4_vYP34
- **Источник:** https://ekstraktznaniy.ru/video/33979

## Транскрипт

### Segment 1 (00:00 - 05:00) []

Thank you. I'd like to start by explaining what Neanderthalization of brain organoids mean. This is a new concept that we start working a couple of years ago. Neanderthalization is a derivative of archealization. Basically what we are proposing here is the resurrecting of the extinct genetic variants that is found on archaic genomes in brain organoids. It's a combination of paleogenomics, genome editing, and stem cell biology technology, mainly the brain organoid. Why we focus on the Neanderthals? Basically, I think we all know at this stage that humans have been sharing the world with extinct relatives, but the Neanderthals are the ones that we do have more rich fossil records. When we look through time, there is really no evidence that the Neanderthals acquire arts, technology and adaptation as we do today. Modern humans likely started in the same way. But in a very short period of time, have developed such a rich history of arts, technology and adaptation. We are basically everywhere. We are even in orbit at this stage. What actually makes modern humans so dramatically distinct from the Neanderthals, which is very close to us and even other species. There is no other species in the world that's able to manipulate tools and technology in the way we do. We postulate the following hypothesis that positive selection on specific genes might have resulted in the modern human brain. Our approach is very unique. We start with paleogenetics by aligning the genome of different modern populations with archaic genomes. What we find out a couple of years ago is that there is a catalog of 61 unique, modern nonsynonymous genetic variants in protein coding genes. These are the catalogs of mutations that distinguish us from the Neanderthals and other archaic genomes. When you look into the 61 genes, there is one that caused the attention, and it's a gene called NOVA1, neuro-oncological ventral antigen 1, which is a master regulator of alternative splicing that is highly expressed during the developing corticogenesis, which is exactly the model system that we have with brain organoids. I thought that NOVA1, being a master regulator of gene expression could control a cascade of gene expression changes between us and other archaic hominids. What we have here is some analysis on the haplotype suggesting that the human specific NOVA1 genetic difference, which is basically a change in isoleucine to valine amino acid change. In the second three K-homology domains in the NOVA1, these are the domains that are responsible for the alternative splicing. The NOVA1 actually interacts with the RNA and we have cryo structure here. It doesn't show directly the protein RNA contact with that specific position. However, the analysis of NOVA1 is showing that the target RNA can bind simultaneously to these two domains that gets together to complete the genome, the splicing factor. The haplotype distribution is consistent with a relative recent emergency of this haplotype followed by spread to fixation. We measure this Tajima D here on the left of the human specific haplotypes around all human specific synonymous and non-synonymous difference and we found that this is consistent with a purifying selection of the NOVA1 haplotype which is quite unique. I'll touch later on how this selection might actually happen. Once you find the differences in those genes and even alterations in proteins

### Segment 2 (05:00 - 10:00) [5:00]

changing amino acid, it's a long way from demonstrating that this actually affects neurodevelopment or how the brain develop. The proposed work here is in the interface between all the genetic variants in neurobiology. We need to test if the genetic variants are relevant. Here comes the brain organoid, which is our stem cell derived neurotissue, also called in the media as mini-brain. We don't fully like this terminology, but each one of these white dots here will grow up to become one of these neurotissues. We can actually guide them to become different brain regions. Here you have cortical tissue. They end up with 2. 5 million neurons in each one of these spheres, 5 million of cells in total. They can survive for several years. Most impressive is the electrical activity that can be generated from these brain organoids. To measure that in a non-invasive way, basically what we do is to plate some of these organoids on top of dishes containing electrodes. These are multielectrode arrays where the electrons are printed in the bottom of the dish and the organoid is on top. All the time, when there is an electrical activity passing by, it will be captured by the electrodes generating an activity map. The beauty of this technology is that you can perform a network analysis longitudinally. We measure the evolution of these networks over time over a full year or period, comprising prenatal and postnatal stages. What we found is quite remarkable. We found that in the beginning, these organoids produce random spikes that starts to get organized when they are about two months of age. By four months, we start to see neural oscillations. This is suggesting that there are microcircuitries in the cortex being formed. By six months, these oscillations are quite stereotype and coincide with the emergency of astrocytes where we have an almost exponentially maturation of synaptogenesis. After that, we start to see a discrete population of GABAergic neurons becoming inhibitory neurons inside these organoids, inducing a more complex network that will increase the complexity as we mature these organoids over time. All these oscillatory events are dependent on both glutamatergic and GABAergic signaling to suggesting that what we have here are all the basic cellular components of the human brain. But how do we show that this is exactly what happens in the human development it's another story. For that, we had to compare the emergency of these neural oscillations over time with EEG or the lateral cephalogram recorded from human brains starts prenatally. For that, we use premature baby brains, where we place this EEG cap on them and we record overtime. We use a dataset of different ages of premature baby brains to actually compare the potential ages of both the organoids and the EEGs using a machine learning algorithm. That reveals that the organoids actually follow the same trajectory of complexity as the human brain does. Of course, the human brain has different brain regions that are not represented in the organoids. But nonetheless, for the features that can be compared, they actually follow the same trajectory as the human brain. The idea would be, if you can make brain organoids from Neanderthals and compare to modern humans, that would be something quite useful. However, to generate brain organoids, you need to reprogram live cells, and there is no live cells coming from the fossil records. Even the brain does not fossilize, so we have to think about a different strategy. That's where it comes this idea of Neanderthalizing human pluripotent stem cells to give rise to brain organoids carrying Neanderthal genetic variants. We do that by genome editing, using CRISPR-Cas9 to replace the modern version of the gene or the genetic variant for the ancestral or archaic genetic variant. We are resurrecting extinct genetic variants back into

### Segment 3 (10:00 - 15:00) [10:00]

a modern human background and then we can compare how these brain organoids look like. Here is a summary of our data. On top, you have the organoids carrying the archaic version of NOVA1, and in the bottom, the modern version of NOVA1 that we all have. On the electrical activity, when we record this over time, we notice something quite interesting. The organoids carrying the archaic version of NOVA1 actually display an accelerated maturation compared to the modern humans. Modern humans are known to have a very slow neurodevelopment, even compared to other species. We are an outlier in terms of neurodevelopment. Our brains really take time to develop. This is clear with the analogy of a baby chimpanzee that can out smart a human baby. The maturation for a chimpanzee brain is much faster compared to the human neuron maturation. However, over time, what we see is that a higher complexity coming from the modern version of the NOVA1 when we compare to the archaic version of the NOVA1. There is an accelerated maturation at early stages, however, a lower complexity of these networks at later stages. Then, again, these places the archaic version of NOVA1 as a factor that might contribute to the slow developmental timing in modern humans, which is quite interesting to know that. But if that happens, it must have a strong selective pressure because we all have the modern version of NOVA1. But this is likely a sporadic mutation that was selected over time. What could be the factor that selected the modern version of NOVA1? We got some cues based on neural progenitor cells that has been exposed to lead. Lead has been one of the environmental stressors in these days, and we know that lead is bad for neurodevelopment. Several groups have been trying to understand how lead contamination might actually affect the development of the human brain. This working particularly, find out that NOVA1 is the gene that respond to lead contamination. That made us think that maybe NOVA1 is being manipulated or selected somehow by lead exposure. It looks counterintuitive to imagine that lead contamination might do that. But our colleagues at Mount Sinai, this is Manish Arora and Renaud Joannes-Boyau. They both show us that lead contamination is spread out during evolution in different primates. This is not a recent event. It has been contaminating humans for the past 2 million years, humans and other primates. How do they know that? They use a specific type of laser ablation on the fossil tooth of these different species, and by ablating the enamel of the tooth, they release some of the components of the enamel revealing event or contaminations, including lead. That's how they showed that the lead contamination was truly spread out through evolution is not like a recent event from modern days. If we were all contaminated by lead, and NOVA1 is a key gene that respond to lead, it seems like an obvious experiment to expose our brain organoids, both carrying the archaic genome, as well as the archaic genetic variant, modern NOVA1 genetic variant to lead. This work was done by Nina Sena, post doc in the lab. When she exposed both types of organoids to lead, she find out that both of them react by expressing genes that are related to neurodevelopment. However, the archaic version of NOVA1, when in contact with lead actually reduces the amount of FOXP2-positive cells. These are FOXP2 neurons that we can detect it using single cell levels. Consider the observed decrease in FOXP2 expression in the NOVA1, archaic cortical organoids, the further reduction of this gene expression following exposed to lead and we

### Segment 4 (15:00 - 20:00) [15:00]

decided to explore what kind of proteins might be creating this toxicity or event. That's when we turn into single cell proteomics. This was done by Aline Martins, a project scientist on the ISSCOR center here at UCSD. Aline was able to isolate these neurons from the organoids and run the single cell proteomics. She find out that most of the pathways are related to axon guidance, nervous system development, and the signaling by ROBO. The alterations in the ROBO family identified to the single cell proteomics suggests that both ROBO might be involved in pathways that are related to the cortico-thalamic axonal projections. Because of that, we decided to also create thalamic organoids. Nina went on and generated these thalamic organoids and exposed them to lead. Again, upon exposure to lead, different from the cortical organoid, the thalamic organoid overexpress the expression of FOXP2. Both ROBO1 and FOXP2 has been previously associated with language and speech disorders. FOXP2 plays a clear role in exonal outgrowth from the cortico-thalamic projections. The fact that they are both in the same complex, I'm showing here FOXP2 and NOVA1 suggests that this complex is actually involved not only on the cortico-thalamic projections, might be specific pathway during the emergency of modern human language. We ask if it is possible that the NOVA1 would be like a gene or a variant that help us to develop such a complex language. Of course, the brain organoids doesn't have that complexity and it's not linked to a body, so it's hard to study language development in a brain organoids. While this study was being done, we were surprised by the work of Robert Darnell, who actually did the opposite. What he did was to swap the modern human version from the mouse allele, from the mouse genetic variant in a transgenic model system. He figured out that the NOVA1 substitution, the same that we did with the organoids really change the contribution to neurodevelopment, as he can start to see different vocalizations in this animal model suggesting, again, that maybe these new NOVA1 and substitution, the modern version that we all have are probably contributing to the development of spoken language to the differential RNA splicing regulation during brain development. In a way, is an independent contribution, an independent validation of our hypothesis. Lead and language. The in vitro system has shown that the ancient or the archaic NOVA1 gene has struggled against lead, likely disrupting FOXP2-positive neurons that are participating on these cortico-thalamic projections. Would that give us an edge for survival? Because if the version of modern humans that was selected allow us for a complex human language, that would be a dramatic advantage over other hominids. Language is our superpower. This is what allows us to develop ideas to plan the future. We propose a very bold hypothesis, suggesting that lead poisoning actually helps to drive human evolution by favoring the modern version of NOVA1, in accelerating the development of complex language. It's a bold hypothesis. It's out there for other groups to help us prove or disprove that hypothesis. I'll end here by saying that there are a couple of take home messages and limitations here. We find out that there are 61 genetic variants that distinguish modern humans from other archaic hominids. The archealization of brain organoids does not generate Neanderthal organoids, but rather a hybrid system. Remember that we are Neanderthalizing single genes one at a time, but the rest of the genome remains modern humans.

### Segment 5 (20:00 - 21:00) [20:00]

That's one limitation. The archaic version of NOVA1 definitely accelerates modern human development. But we don't know if there is epistatic contribution of other genes that might compensate for that factor remains to be seen. The single nucleotide alteration in NOVA1 was likely selected by lead exposure because of its protective effect on FOXP2 cortico-thalamic neurons, allowing modern human language to flourish. These are the messages that we have so far. We are definitely continuing on these experiments. One of the things that I'm quite excited about is to start using a closed loop feedback between organoids and a robotic interface to start probe for memory and learning. We previously showed that these organoids are able to, for example, manipulate a robot and teach the robot or remember commands to avoid obstacle. The robot can use plasticity of these organoids to actually interact with different environments. We do that as well as trying to map the sequitries that are inside of these organoids or participating in these organoids using high density multi-lectin array, such as the ones that are showing here on the right panel. I'll finish here by thanking members of my lab, especially Nina Sena, who led the work and our collaborations at the Manish Arora lab at the Mount Sinai. Of course, the funding for this type of work, and here's my contact information. Thank you very much.