Early Roles of SYNGAP1 in Neurodevelopment: Insights from a Top Autism Risk Gene

Speaker A: This podcast is a presentation of University of California television. Like what you hear, consider making a donation at UCTV tv. Donate so we can continue to bring you more great programs.

Speaker B: Really the biggest thank I want to do today is to the family in the audience. What I really want to say is that your resilience and ah, your commitment to improve the life of your loved ones is really the biggest source of inspiration for the work that we do the do. So really thank you. And today I'm going to tell you about the effort, our effort in improving modeling of human brain development and disease. And so in the last 70, uh, years or so we really as a field we've been putting an incredible effort in trying to understand neurodevelopmental and neuropsychiatric disorders. However, we really haven't been able to come up with effective treatment for, for this disease. And why is that? There are several reasons, but for me one of the most important is the incredible complexity of the genetics that underlie these disorders. For example, if we think about autism, most subtypes of autism are polygenic, which means the same genetic variant in two different individuals can lead to completely different clinical manifestation. And so this is obviously as complicated a lot our understanding of these disorders and our ability to come up with effective treatments. What we really need to do to understand these disorders is really to being able to dive into the genetic of these disorders and basically to have access to human tissue to do so. What we ideally would like to do is to go back in time and look at how the development of the brain of a specific patient happened in the womb. Obviously we cannot do that. But something that is quite close to this process is um, basically culturing 3D, um, brain organoid derived from human pluripotent stem cells. So most of you may know that um, we are now able to take somatic cells from a patient, then revert these cells to pluripotency and being able to generate the so called brain organoids. There are different protocols for generating brain organoids. Some are um, called unguided protocols and give us the opportunity to generate organoids that contain many region of the brain. Then there are some guided protocols that allow us to generate region specific brain organoids. Now how do these organoids look like? Well, you can see here there is um, a graduate student in my lab holding a flask with brain organoids. And this is how they look like. They are um, about 1,2 millimeter big in diameter and they look like lentis. Basically. Um, as I mentioned, we can generate organoids from different brain region. Most of the model that is mostly used to recapitulate neurodevelopmental disorders are cortical organoids. Why cortical? Well, we know that the human cerebral cortex is the structure in the brain that is responsible for the execution of higher order, uh, function, including movement, speech, communication, intellect, sensory perception. These are the very function that make us human and are also the function that are often disrupted in developmental disorders. Now there are multiple protocol to generate cortical organoids. There is um, one that um, we use in the lab that has been published by the Arlotta lab and really um, builds on the work on Yoshiki Sasai that is the pioneer of this field. Basically what you can see is that when we generate cortical organoids, we are able to generate basically the majority of the cell types that we see in the human fetal cortex. These organoids are pretty reproducible and so they really make for a good model system to understand these disorders. In the lab we have been focused on um, studying SYMBAP one that is one of the top autism risk genes. We know that de novopathogenic variants in SynGAP1 lead to intellectual disability, epilepsy, global developmental delay. We are really focused on this gene also because we have very close ties with the SYNGAP research fund. We know these families and we really care about understanding more about the biology of this gene. And what um was known about syngap1 was mostly known by work and research done in excitatory neurons of mouse. And uh, so when we started working on this protein, we wanted to understand a bit more about what SYNGAP does in human neurons. So Marcela Virtele and Ashley Del Doso, um, they're wonderful postdoctoral fellow and PhD student in the lab. They decided to look at the expression of syngap1 in the human fetal cortex. So they look at expression of syngap1 in different cell type of the cortex. And they were very surprised to see that syngap1 is not only expressed in neurons, but they found also expression of SYNGAP1 in progenitor cells. This was never been described and so they wanted to make sure that this observation was indeed correct. And so they decided to look at of SynGAP1 in cortical neurons. They really wanted to look at expression of the protein, not only of the gene transcript. And so here you can see a section of a brain organoid. You can see that much like it happened in our uh, brain, these organoids have ventricles here that are fluid filled Cavities surrounded by progenitors and then neurons. And you can see that syngap for sure is expressed in neurons, so in this black part of the section, but it's also indeed expressed very strongly here in the progenitor region. Basically, we, uh, discovered that syngap1 is indeed expressed at the apical end feet of these radial glial cells. In collaboration with, uh, Marcelo Cobb at USC, we were also able to do proteomics for SynGAP1 and look also at the complex, um, and protein interaction of SynGAP1 in progenitor cells in organoid day 7 that mostly, uh, contain only progenitor cells. And so we were able to confirm the expression of syngap1 in progenitors. This was also confirmed, um, by looking at expression of SYNGAP1 in human fetal tissue. So you can see the Singaporean has a very similar pattern of expression with tjp1 that is as well expressed at the apical length fit of these progenitors. Then obviously the question was, what is syngap1 doing in this progenitor? It's not supposed to be there. So we did, uh, bulk sequencing of organoids generated by patient with synGAP1 up insufficiency. So they have 50% of the level of synGAP1 that we see, um, in control. And so we were able to repair, um, this variant and then look at what processes were dysregulated. And we found that mostly the processes that are dysregulated are related to cytoskeleton remodeling or cell addition. Also, looking at SYNGAP1 interactors into progenitors, we saw basically the same picture. These progenitor cells, radial glial cells, are very special cells. They, um, have two processes, one epical one and one basal one. The epical process reaches the ventricle

Speaker C: of

Speaker B: the human brain and then the basal process extends towards the peel surface. We really wanted to have a model, um, to understand whether the cytoskeleton remodeling of these processes was altered. We, um, team up with Randy Ashton at University of Wisconsin Medicine. Yes, this rosette system, you can see, in this rosette system, we are able to see how these radial gale cells extend from the apical to the basic basal with this epicobasal polarity. And sure enough, we found that in patients with SynGAP1 upright insufficiency, we see that there is a disruption in the way this, um, rosette form. So here you can see that in control lines we see that these ventricles form and here basically they are mostly absent. When they are there you can see that the morphologies is pretty altered. This is a quantification again showing this reduction in the number of rosette and when they are there the perimeter and circularities is actually altered. Um, as I mentioned before it's uh, um these disorders are very much dependent by the genetic background. And um, so we decided to um, edit this variant in another genetic background. So in a control line and sure enough we found that the same phenotype was recapitulated also in a different genetic background. We also look at um, another variants. Actually we uh, disrupted the rasgap domain of the syngap protein and we found that the phenotype was again recapitulated suggesting that the RAS gap enzymatic domain of SynGAP1 is involved in this early developmental process. Now SynGAP1 is a very complex protein. So here you can see syngap1 has different domains and um, what we are interested in doing now is to extend our investigation to other variants and really um, try to understand how different, other different domains, what is the function of other different domains during early development. So this is something that we are now um, trying to do in the lab going back to the organoid model. That is a model that is more advanced compared to this 2D rosette that I showed you before. Normally again here is a section of a brain organoid and you can see there is a separation between progenitors and neurons in control organoids. In syngap upper insufficient organoids you can see that this separation is not as clear anymore. You can see that some of the neurons are actually still into the progenitor levels. And then you can see that the morphology overall is disruptive. We look um, at the division mode of this radial glia. We found that in apling sufficient lines we saw a disruption in the way these radial cells that are the progenitors of the entire cortex, um, they do not divide properly. There is an increase in differentiative division as you can see in here. Overall when we look at ah, with single cell sequencing, specifically at the number of these cells in the organoids and the transcriptional profile of these cells, we found that there was a decrease in number and then upregulation of terms associated with differentiation, downregulation of thermals associated with mitosis, suggesting that these radial GL cells have an impaired developmental trajectory. If we look at then the structural features, structural and functional features of neurons in these apple insufficient organoids we found that indeed there is um an evidence of faster maturation of neurons in this culture. So this is um, something that. So this increase and faster maturation in appliinsufficient organoids is actually something that is um, a very good biomarker to test potential therapies. Um we found that. So now um, basically we have tested some antisense oligonucleotide that are able to, to rescue the level of syngap gene transcript and protein. And so we found that these antisense uh, oligonucleotides are indeed also able to rescue the level of network, network activity. So now we are partnering with several, some biotechs to try to um, test in our system therapeutics that can using this network activity as basically biomarker. So we are very excited about this more translational spin for this project. Um, to summarize this first part of the talk, what I show you today is that we were um, able to discover a function for SynGAP1 and the presence of SynGAP1 in radial glial progenitors. In human radial glial progenitors and here Singap1 has an important function as scaffolding protein up insufficiency of syngap1 leads to disruption in the division mode of the radial glia and in overall in accelerated maturation of cortical projection neuron. Now um, when we look at syngap expression in these progenitor cells, we also look at as I mentioned before, the proteins that are interacting with syngap in these progenitors. And we found that several other synaptic proteins that interact with syngap in neurons are also present in progenitors. So now we are uh, very interested in testing these hypothesis of whether a complex similar uh, to the complex that we see in neurons for um, preserving the um, integrity of the uh, scaffolding of the postsynaptic density of the neuron could also be present to preserve the scaffolding of radial glial cells. And this concept has also been um, actually um, materialized in a sculpture by this local artist that is called Jen Kwoke. He's a local artist here in Los Angeles. She has represented, as you can see here, this concept showing that a complex that is used in neurons um to regulate synaptic transmission is actually recycled here in this other cell type, in this case these radial glial cells. And so now I want to switch gear for the last part of my talk and talk about a new direction that we have taken in the lab, mostly inspired by conversation with syngap families. So when we talk with SYNGAP families what they told us was that obviously cognitive impairment in their children is um, obviously heartbreaking and disruptive. But there is something else that is also very disruptive for their family which is gastrointestinal symptoms. So Marcella Birtele, that um, wonderful postdoc in the lab, she actually just got a faculty position, accepted a faculty position at UC Irvine. So she will be starting her own lab next year. Marcella Bertella really got passionate about this question about understanding whether UM syngap1 could also affect another type of neuron so um, enteric neurons. Um she did a first analysis looking at the expression of autism gene in um the adult enteric nervous system. She found that syngap1 is indeed expressed in enteric neurons along with a bunch of other gene that you can see here highlighted. The enteric neurons come from the neural crest developmentally. M. Uh they come mostly from the vagal neural crest. And some of those come uh, some of these come also from the sacral neural crest. And so what Marcella um um has then um, basically so she basically became very interested in understanding a bit more about these cells. We also found um, that indeed um, um these autism risk genes are also found not only in adult intestinal neurons but also in fetal enteric neuron. This is actually an analysis done by um, the Willsay lab showing that you can see that when you look at different type of intestinal cells there is this um, increase in the expression of autism genes in enteric progenitors and in enteric neurons. This was very interesting for us. Here you see some data collected by the Simons foundation and Citizens in which you see but um, this is not unique to SYNgap1. So many other genes associated with autism also show the same show that um basically um, patient cohort with autism gene variants have this uh, gastrointestinal dysmotility problem. In the case of syngap1 and also other genes the main um issue was actually with constipation. So Marcella decided to um, develop um, a system to um, look at um dysfunction in enteric neurons. She really build on the work of Lawrence Studer. She was able to generate these enteric spheroids that um, are able to um, generate um the cellular diversity that we see in the enteric nervous system. And um, she found. So this is all unpublished, very preliminary. But what she found is that much like we see for cortical neur also enteric neurons with SynGAP1 Apple Insufficiency show this faster maturation so this is pretty interesting. Um, then the question now is try to understand if this increased maturation has also effect on um, the intestinal epithelium. And so in collaboration with Jason Spence, we are trying to understand uh, a bit more about um, how syngap affect for example the niche or other, uh, intestinal cell type. Um, but something that Marcel has been also very much focusing on has been trying to develop, uh, um, this assembloid model. That is an assembloid model, um, a fusion between human smooth muscles and enteric neurons. Um, what she has found is that indeed she's able to see contraction in this, um, spheroid of human m. Small muscle. And so what she' now doing is try to understand how um, contractility is also altered in co culture with syngap cell lines. And then Marcella's idea is to bring this project to her own lab and screen many other genes associated with autism spectrum disorders. And so just to conclude the second part, I show you that autism genes are associated, um, um, are enriched in the enteric nervous system. I also show you how syngap1 is able to regulate MA deterioration not only in cortical neurons, but also in these uh, peripheral um, neurons. And then we um, really think that this assembloid between enteric nervous system and smooth muscle can also be a very simple model to screen for gut motility and for therapies to improve gut motilities. And so with this I like to thank, um, all my lab members. I think I highlighted the people that were involved in this specific project, our collaborators, and then specifically the SYNGAP Research Fund, um, NIH and all other, uh, funding, uh, agencies for uh, support. And then thank you for listening.

Speaker C: Hi Georgia, I don't know if you can see me. This is Lilia, ah, from ucsd.

Speaker B: Hi Lilia. Hi.

Speaker C: Uh, thank you for the very interesting talk. Um, I have uh, two questions basically. So, um, when you see this disorganization of this ventricular zones of the rosettes in the syngap organoids, um, and faster maturation into radial glia into neurons, um, do you think also migration is involved? So do you know whether, ah, radial glia matures into neurons, uh, faster in mutant organoids and then neurons cannot move out, right? So they're kind of stuck in the radial glia, uh, area. Is that the case? And then following that, my second question is, when you did your aso, ah, rescue, what exactly did you do it in organoids? And what exactly, exactly phenotypes were you able to rescue uh, were you able to rescue the um, you know, slow down the uh, maturation then or migration or like what if you know what. Exactly. Phenotypes. Thank you.

Speaker B: Thank you. Thank you. Lilia. These are great questions as always. So, um. Yeah, um, you're right. So radial glial cells are very important for guiding the migration and the generation of the cortical layers. Um, we uh, have seen that. I show you this image in which basically we see that neurons are actually stuck in the um, progenitor layers. And um, so we think that we haven't really got very deep into that type of analysis, but we think that that's probably the case. And in fact um, we think that um, it could be that this displacement of neurons in humans could also lead to problem in network connectivity circuit assembly. So that's certainly the case. The organoids are not the best model to investigate positioning in the cortical layer because as you know these cortical layers are not really well formed and maintained. So we didn't really go there much. But there is work from Gavin Rambaugh showing that uh, even in mouse they have seen some um, problem with the positioning of neurons. So I think you're absolutely right and that's a feature. So in terms of um, antisense oligonucleotide, what we have done so far is um, we have really look at whether we can rescue level of uh, protein and transcript and we can um, um. And then we have been focused mostly, mostly on network activity, uh, sort of bring down the level to the control, uh level. We wanted to have an assay that is uh, sort of high throughput, um, and easy. We have done that with calcium imaging and now we are um. But we want to scale it up with microelectral arrays. So the idea is to have something that is very high throughput. But you know, it's very interesting question to see if all the other phenotypes can also be, are also rescued. We haven't done it, but we should be doing it. We'll do it. Thank you.

Speaker D: Have your study participants been solely individuals with autism with coexisting cognitive disabilities, I.e. intellectual disability, or have you also sampled individuals with autism with average or above average intelligence?

Speaker B: Yeah, so this is um, a great question. In fact I have to say our studies have been only focus on seeing up insufficient patients. And so these patients have um, seizures, they have global developmental delay. Um, yeah. So uh, no, we have only done. We have been really focused on Singapore so far. We don't exclude to we want to expand our analysis to syngap interactors though. This is what we are going to. To do. And um, so, yeah, we are mostly focused on intellectual disability.

Speaker E: Uh, hi, Georgia, this is Jonathan Sebot at UCSD. Um, as you mentioned, uh, syngap1, uh, seizures are common in syngap1. Um, and do you know if they respond well to treatment or are they treatment resistant? And um, what do you know about how organoids respond to anticonvulsants? And I have a second part to the question as well, but finish that one first.

Speaker B: So, um, what I can say is that there is a lot of heterogeneity in this patient. And so there are patients that respond well. Patient that don't respond at all. Uh, we haven't, uh, test these in. In uh, in organoids. We should do it. And sorry, the second part of the question.

Speaker E: The second part of the question is relating some. So in terms of the synaptic effects of SynGAP1, um, and the epilepsy in particular, um, is there a way to relate that to the enteric nervous system is gut mo. So can you say something about gut motility that we might learn. So learning from the brain and then trying to extrapolate that to what's going on in the gut?

Speaker B: M. Okay, I really not sure how to address this question. I. I think, um, this is exactly what we really want to understand, um, how um, this faster maturation may affect motility. Because, um, I have to say we don't really know in the first place how this process happened and, uh, how they happen in a normal physiological condition. Um, there is really not much known. And so, yeah, so this is. We are trying to study this neuro factor junction and try to understand a bit more in control condition and then what happens. Um, yeah, in, in the case of syngap, specifically we want to understand if also for other autism genes there is some sort of convergence. Um, but yeah, that's a very, you know, I love the question. Very creative and very. I don't have an answer, but yeah, thank you. I'll be thinking more about it. I like this angle.

Speaker D: We have an online question. What was responsible for the accelerated maturation of cortical projections?

Speaker B: What's responsible? Yeah, so, um, what basically we think is happening is that there is, um, the radial glial cells, the division mode is altered and there is this increase in differentiative division. And um, so in general the developmental trajectory is altered. And so this leads to the faster maturation. There is, I have to say, um, I haven't stressed it, but there is also a number of, um, issues that are related to neurons and, uh, to, to the synapses. So there are alterations that are, we believe, are, may not be development related. So they are also synaptic dysfunction that, um, happen. And so, um, for example, there are changes in the morphology of dendritic spines, composition of the postsynaptic density, and, um, so this is something that is neuronal. But then what we think is happening is that the phenotype we see is really dependent by basically the radial glia, uh, the division model of the radial glia.

Speaker E: Thank you, Dr. Cuadrado, for a wonderful

Speaker B: presentation, for taking time to be with us today. Thank you.

Speaker F: You've been listening to a podcast by University of California Television. For more information about this program or uctv, visit us online at. Ah, UCTV TV.