# CARTA: Human-specific Alterations in Brain Cellular Proportions with Genevieve Konopka ## Метаданные - **Канал:** University of California Television (UCTV) - **YouTube:** https://www.youtube.com/watch?v=jjQ3c4BxZn4 - **Источник:** https://ekstraktznaniy.ru/video/33975 ## Транскрипт ### Segment 1 (00:00 - 05:00) [] Well, it's really great to be here. I'm glad everyone's participating. I'm excited to be one of the co-organizers and also have this opportunity to share with you some of my lab's own research that's relevant to this symposium. Today, I'll just be sharing with you one vignette on the topic of human-specific alterations in brain cellular proportions. One of the major goals of my lab is to try and understand what are the human-specific genetic and molecular features of the brain. If they exist, can we identify them, and can we use them to give us some insight into the behavioral features that set humans perhaps apart from some of our closely genetically related species, such as chimpanzee, as shown here. Now, as a card-carrying neuroscientist, I usually have to show one slide from Ramón y Cajal, as indicated here in this slide, and Ramón y Cajal and other neuroscientists now for over 100 years have been looking at the brain in one way. There are many ways to look at the brain, and that's with respect to trying to catalog and observe and give us information into the various cell types, not just in the neocortex, as indicated here in this slide, but throughout the entire brain. In more recent modern times, current times, there have been a lot of cell atlasing efforts using transcriptomic at the single cell level to build these atlases that define all of the various brain cells that are in the human brain, as well as in other species. If we can do that in the human brain, and in closely related species, then we might have some insight into cell types that might distinguish us from our closely related species, such as chimpanzees. This is just one effort through one of the atlasing efforts. I think Ramón y Cajal and his contemporaries would be surprised that by using transcriptomics, these atlasing efforts have identified thousands of cell types in the mammalian brain. Because for many years, we've thought of brain cell types in a simplistic manner, in that we have neurons and non-neurons. Over the years, various non-neuronal cell types have gone in and out of fashion. There's been some recent interest in microglia. Astrocytes tend to come and go within the neuroscience field, but I would like to make a pitch today for other cell types and subtypes of major cell types, and I'll be spending a lot of time in my talk focusing on oligodendrocytes. Which as shown in this cartoon, indicate that they are important for myelination of neurons, but I'd also like to show you or present evidence that they may be involved in other important aspects in brain function. This is just another cartoon to indicate to you that again, it's not just major cell types we care about, but subtypes of all of these major cell types as well. They're very varied, and we can look at them not just based on morphology, but also at the level of transcriptom as these atlasing efforts have done. Like many probably in the symposium, my work has been inspired by the work of Mary-Claire King and A. C. Wilson that was published now over 50 years ago, where they were looking at levels of proteins between humans and chimpanzees, found that they were very similar, and then came to the conclusion that their macromolecules are so alike that it must be regulatory mutations that account for their biological differences. Now we know, fast forward many decades later, after the sequencing of the human and chimpanzee genome that indeed at the level of the protein coating parts of genes, they're highly similar. But if we look outside of those parts of the genomes where there's less constraint, that's where we can see some variation. It's also how those genes are regulated, how they're turned on and off, how they're expressed, that may lead to the different phenotypes and perhaps functions and behaviors between species. For a number of years, we've decided now to make these comparisons across species at the gene expression gene regulation level. ### Segment 2 (05:00 - 10:00) [5:00] This is just an example of looking at one part of the brain, in this case, the neocortex, the dorsal lateral prefrontal cortex, among three species. Now, typically, we would want to have a minimum of three species, if not more, so we can assign changes in gene expression specifically to the human lineage. We would look for genes that are similar between chimpanzee and our outgroup, rhesus macaque, and then genes that are different between human and chimpanzee, as well as human and Macaque. In this case, we can then assign that gene expression change to the human lineage. We want to do this not just broadly, but try and really intersect this idea of many different types of cells in the brain with changes of gene expression. A number of years ago, we weren't quite able to refine it down to that subtype level, but we could get broad cell type comparisons by just simply looking at neurons and all the cells in the oligodendrocyte lineage. We could do this by taking tissue from humans, chimpanzee, rhesus macaque. This is all postmortem tissue. Isolate the nuclei from this tissue, sort using antibodies for cells that were neurons or cells from the oligodendrocyte lineage, and then sequence the RNA to get gene expression levels. If we look at just the first principal components of these analyses, either in the neurons or the oligodendrocyte lineage, this basically recapitulates phylogeny, which was good. However, what we really care about are changes in gene expression, specifically on the human lineage. That's our interest when we look between these two major cell types. Using the number of millions of years that separate these species from common ancestors, we could define genes changing specifically in neurons or oligodendrocytes on the human lineage. The changes in neurons were not surprising, and we were very pleased with those, but the fact that we got many changing in oligodendrocytes and perhaps an even more accelerated change in oligodendrocytes than neurons was surprising to us because this didn't follow what had been previously shown in the literature, and so we thought perhaps this was technical issues. What we did was we looked at the previously published studies, including work that came from my own research, and we intersected those with these newly identified human-specific neuronal genes and human-specific oligodendrocyte lineage genes. Everywhere you see red here indicates a significant overlap. As we expected, we found a significant overlap with the previously published neuronal genes. However, our new data in the human oligodendrocytes did not overlap with the previous published datasets as we had remembered. Again, we were still left with this feeling that, perhaps there's some technical difficulty in our approach because this hadn't been seen before. Right around this time, the field generally of genomics was able to isolate truly individual cell types and quantify the transcriptome within individual cells using what's called single-cell genomics. There are a number of ways of doing this. One way is through a microfluidics approach using a droplet-based capture of cells or nuclei. A number of groups had published the first beginnings of these atlasing efforts to identify cell transcriptomes in human and mouse brain, for example. We could take those ground truth cell type transcryptomes and use that to do what's called a deconvolution of our somewhat sordid or even a bulk tissue approach. When we did that, we found something very striking, in that the majority of signal that was in the published studies was really derived from excitatory neurons. We now know, many years later of doing this single cell approach over and over again, that neurons express more genes and they express them at higher levels. If one was to take a piece of brain tissue and just do gene expression on it, the majority of the signal would indeed be driven by neurons. What we could then do was take our sorted data and apply the same deconvolution approach, and we could see that the neuronal sorted nuclei were driven by gene expression from neurons, and the oligodendrocyte sorted nuclei were driven by signals from oligodendrocytes. The previously published studies and a lot of the literature was basically unable to detect this non-neuronal signature because the neuronal signature was so high, it was masking it. It wasn't a technical issue on this new approach ### Segment 3 (10:00 - 15:00) [10:00] it was more technical issue on what we had been doing previously in the field. This was where we were a number of years ago, and we wanted to intersect this approach of this comparison across species and really look, again, identify human-specific gene expression changes, but look at all of the cell types, not just neurons, and only the oligodendrocyte lineage. Although we're primed to think about this and keep in mind all these subtypes. We had a study moving forward where we had the same approach of comparing human chimpanzee, using rhesus macaque as our al group. But in this case, instead of using frontal cortex, we were using posterior cingulate cortex. The reason we used this part of the brain was, A, it was available. It's very hard to get these tissues from non-human primates. B, we could be certain we were comparing roughly the same region across all of the species. It hadn't been looked at before transcriptomically, to our knowledge across these species. We then took the tissue, we isolated nuclei, and we use this new approach, a single cell transcriptomic, not the sorting approach, but looking at every individual nuclei. Then in this case, we could identify many different subtypes of the oligodendrocyte lineage because we were already primed to be thinking about oligodendrocytes in human brain evolution. OPC is oligodendrocytes, precursor cell, COP, committed oligodendrocyte progenitor, newly formed oligodendrocytes, mature oligodendrocytes, et cetera. Before we even started thinking about genes, one of the first things we wanted to look at because again, had kept all the cells we didn't enrich for one or the other, was whether there were changes in proportions among all of these cell types across all of the species. What we found very strikingly to our eye here was that there was an increased relative proportion of human indicated here in blue of these immature oligodendrocytes, not just the OPCs but also the COPs, relative to the mature oligodendrocytes. We then quantified this and applied statistics to it, and found indeed a significant increase of immature oligodendrocytes and a significant decrease of mature oligodendrocytes on the human lineage. We found this not only at the level of RNA, but also chromatin accessibility because in the same tissue, we applied another approach of single nucleus, ataxic, which looks at open enclosed areas of chromatin. This was very exciting to us. But one of the caveats to this approach is that you have to take the tissue and dissociate it to do this experiment, and so being cautious, I thought perhaps we may have inadvertently only in the human samples taken perhaps more white matter or areas that would have more oligodendrocytes than the other species. A way to validate this result without having to dissociate tissue is to carry out an in situ hybridization at single cell level. In this case, Single Molecular Fluorescent In situ hybridization, smFISH we could use a probe for immature oligodendrocytes, PDGFRA, or mature oligodendrocytes, MOG. We did this in independent intact tissue from both humans and chimpanzees. If we quantified this, we could recapitulate what we found in the single cell approach with increased immature oligodendrocyte ratio in the humans compared to decreased mature oligodendrocyte ratio, humans compared to champ. This is a really great validation. We were very excited about this, and it struck me that this seemed unusual that this hadn't been observed previously, perhaps in the literature, maybe due to lack of technology or markers to distinguish these cell type but also perhaps it was just some vagary of the posterior cingulate cortex researchers hadn't looked at this closely. I thought, well, what if we looked at other areas of the brain? There was a publicly available data set from the anterior cingulate cortex from Philip Kokic and colleagues, where they had human chimpanzee and rhesus Macaque single cell data. Using that published data and re-analyzing it, we were able to find also in this brain region increased immature oligodendrocytes, decreased mature oligodendrocyte ratio. Now you may ask, why didn't these individuals report this change in this ratio? Well, this was because in the years since they had published this in the years where we were looking at this, it became apparent to the field and work out of my own lab that when one dissociates the tissue, the neurons which express more genes, more RNAs all the way out to the synaps, you open up those RNAs into the fluids ### Segment 4 (15:00 - 20:00) [15:00] it gets encapsulated and you have this issue of ambient RNA. We found ways to clean datasets. We cleaned their dataset and found this change in ratio. We also in this case, wanted to confirm, again, in intact tissue samples, this finding, and so we got independent human and chimpanzee tissue samples, performed the same smFISH experiment, quantified it, and again found this increased immature, decreased mature oligodendrocyte ratio in humans compared to chimpanzee. We were very excited about this. Again, being ever cautious about new findings, I thought, well, perhaps this is just something about the anterior singular. The cingulate cortex in general. We had looked in the anterior and the posterior. What about other parts of the brains? Were there other datasets that we could query to see whether this held up there? We took our sorted data that I started this presentation with, from the dorsolateral prefrontal cortex. We didn't have it at cell type resolution, but we could distinguish using this deconvolution approach, this ratio. Again, using this data, we could see this increased immature to mature oligodendrocyte ratio in the dorsolateral prefrontal cortex. There was also a publicly available data set from the Allen Institute folks looking in the primary motor cortex. They had human and rhesus macaque, and also human and marmoset. Again, in these datasets, cleaning it up for ambient RNA, we could again find this same change in oligodendrocyte ratio specifically in the human brain. This was very exciting. Everywhere we looked in the neocortex, we could see this change in ratio, specifically in humans. What about outside the neocortex? Well, at the time, there were two publicly available datasets, one in the caudate nucleus, one in the dentate gyrus, and in neither case did we see a change in this ratio in these non- neocortical regions. To summarize, what we found was that humans have this altered OPC and oligodendrocyte ratio, specifically in cortical regions. Across all these datasets, now we have four species. We looked in four neocortical regions, posterior anterior cingulate, dorsolateral prefrontal cortex, primary motor cortex, and two non-neocortical areas. In the neocortical areas, no in the other regions. Now as all of these atlasing efforts across the entire brain are moving forward, we'll see how this ratio does or does not hold up across other parts of the brain. Importantly, we're also hoping to look to see how this ratio may or may not dynamically change over development, and if there is a dynamic change over development, where it happens in the human brain development compared to other species' brain development. One may ask, well, if this is a ground truth about human oligodendrocytes, this change in immature to mature ratio, specifically on the human lineage, why would the human brains have evolved this? What could it be good for? Taking this image from a review now over 15 years ago, hopefully you can appreciate that mature oligodendrocytes are important for myelination. But as you can see along this whole lineage, there are other purposes for immature oligodendrocytes. Some of these red squares are indicating neurotransmitter receptors or signaling, and so we know that these immature oligodendrocytes serve other functions for at least interacting with neurons, let alone other cell types in addition to myelination. We've developed this hypothesis that the evolution could have occurred to have this increased potential, perhaps for experience-dependent generation, a mature oligodendrocytes, so there's this pool of immature oligodendrocytes that sitting here, this increased proportion, waiting for some signaling, either from neurons or other cells, telling it to either myelinate or to do other things. For example, OPCs have also been shown recently to be involved in pruning synapses. This next slide shows some of the other functions of oligodendrocyte precursor cells. They're involved in hypoxia, in injury. Importantly, one point I want to make is this idea of demyelination. With respect to thinking about how this might be important for brain evolution, how it may intersect with brain disease. There is one human demyelinating disorder, multiple sclerosis that has not been observed in any other species. This is something that could very likely, of course, be observed pathologically, up to the current time, we have not seen this in other species. It could be something about this change in ### Segment 5 (20:00 - 24:00) [20:00] the OPC to mature oligodendrocyte ratio in the human brain that makes the human brain at risk for multiple sclerosis and intersecting perhaps with other cell type evolve functions such as in the immune system. One other final point I'd like to make about these data is we're doing it all in postmortem tissues. We're making these important findings, developing these hypotheses. But then the question would be, how could we test the hypothesis? We can't really do that ethically or for a number of reasons in postmortem tissues. We can use in vitro models, and you've heard about stem cell models, or we'll be hearing about it in the symposium. In this case, we can take IPSC-derived cells from both humans and chimpanzees, and our lab has developed a fast protocol for differentiating IPCs into both immature and mature oligodendrocytes alone without the need for other cells. If we quantify that and compare between the two species, we can again see this change in ratio and increased immature as indicated by PDGFRA and decreased mature as indicated here by O4 and Olig2 positive cells ratio in a dish. The cell type model alone recapitulates what we were able to observe in the in vivo tissue. Now we can use this to test our hypothesis about whether this ratio is important for, for example, the experience-dependent plasticity of neurons, if we co-culture them with neurons and mix and match species, human OPCs with chimpanzee neurons, or vice versa. We can also test the contributions of individual genes to this alteration ratio between the species. It's very exciting times. I've given you some data and information about changes in the oligodendrocyte lineage in these species in these brain regions, but what data are we still missing? Well, all of this data, at least in the neocortex, were derived from the gray matter. We know the majority or a lot of oligodendrocytes are found in white matter parts of the brain. We need to be doing the same cell type transcyptomic profiling in those parts of the brain, in the white matter tracts. We also need to understand if there may be differences across different regions of the brain, just like we did in the neocortex, what about different areas of white matter? Are all oligodendrocytes and their precursors the same transcriptomically? How do the proportions and gene expressions of these oligodendrocytes change over development? As I already mentioned, that would be very exciting to see where and when this occurs, and does it change when we're comparing across species. If we can actually generate those data across closely related species such as non-human primates? To summarize some of the take-home messages, it's very important when you're looking across species and transcriptomically to treat all cell types equally. If you want to observe these changes in proportions, think about how proportional differences could occur across regions, across development, across species. I showed you some validation experiments using smFISH and how that validated the dissociation of the microfluidic single cell approach. But there are new technologies in play now spatial transcriptomics, where you can have the tissue intact and measure the transcriptome of every cell. I think that will eventually really be the best way to look at how these changes are occurring at the cell type level across tissues and across species. No matter what you're using to generate the data, validation in an independent way is great. In this case, I showed you smFISH, and now we can test the functional outcome, test our hypotheses in a system that's perturbable, such as the IPS-derived approach. I just like to here's a picture of my lab. From last almost two years ago now, thank everyone in my lab that participated. I have a great set of collaborators, many of whom contributed tissue, for example, Chet Sherwood and Bill Hopkins for the non-human primate Soojin Yi has been a great collaborator, my funders. I'm looking forward to the Q&A session at the end of this, and thank you so much for attending this symposium.