33. Future Robots: Human Cells and Synthetic Morphogenesis.
Available now
-
Available now -
Episode TranscripT
This transcript of the podcast was auto-generated and may include typos.
Gizem Gumuskaya [00:00:00]:
Like trees self construct, like humans self construct, flowers self construct, but skyscrapers don't, or cars don't, or computers don't. So human made objects never have had this ability to build themselves, like, until now. Which is why I think synthetic morphogenesis is so exciting, because it is giving us access to properties in human designed products that so far had been exclusively reserved to nature.
Mizter Rad [00:00:38]:
Welcome to the Mizter Rad Show, where I talk to the most interesting global personalities about the future of humanity. Hello, beautiful humans. In the future, if you are diagnosed with, say, clock arteries, also known as artery plaque, we might be able. And listen to this. We might be able to go to the doctor and get a tiny little robot, clean those arteries for us. Now, these little robots are called anthrobots, and they're made out of human cells. Today I have Harvard synthetic biologist and architecture Gizem Gumuskaya with us, who's working in this exciting topic right at the forefront of human progress. Gizem Nazil sin, how are you?
Gizem Gumuskaya [00:01:33]:
I am Taisakhi lai. I'm good, thank you. How are you?
Mizter Rad [00:01:37]:
I'm very good. Gizem. Gizem. Let's start by you telling me in very simple words, what are antrobots? And now I want you to imagine I'm like ten or 15 years old. How would you explain that to me?
Gizem Gumuskaya [00:01:52]:
Yeah. So anterobots are a new type of robot, except that unlike traditional sort of concept of robotics that we're familiar with, they don't have any electrical parts or mechanical parts, no wires. They're completely made up of biological cells. So each antibiotic grows from an individual single human cell and builds itself in the course of two weeks into this multicellular motile structure. Because of their programmable synthetic anatomies, they are considered a type of robot. And then putting those two together, we call them biobots. An answer is coming from the prefix. Answer is coming from a reference to their human origin.
Mizter Rad [00:02:43]:
So in a nutshell, so tell me something. You just said that they're built in around two weeks.
Gizem Gumuskaya [00:02:48]:
Correct.
Mizter Rad [00:02:48]:
You take around two weeks to build it?
Gizem Gumuskaya [00:02:50]:
Correct. So this is the pipeline I've developed at Tufts University and Harvard University is this institute, as a part of my PhD in the laboratory of Professor Michael Levin. And our goal was to create these motile robots that had cilia on their surface. And there are a lot of different ways to do that. But another thing we wanted to accomplish is to make sure that these bats are building themselves so very much like a plant seed, you know, plant a seed and a flower tree or whatever you're trying to grow grows. So we're trying to recapitulate that self construction idea in a completely synthetic system. So when you.
Mizter Rad [00:03:30]:
So what? When you say. Let me interrupt you right there. When you say that, when you compare them as a plant seed, a plant seed, you put it in the soil, and that grows into a tree or a flower, whatever it is, what you're saying is you're taking a cell from the human body and you putting it somewhere in the lab right now. And that would grow into something else.
Gizem Gumuskaya [00:03:55]:
Exactly. Yeah.
Mizter Rad [00:03:56]:
Okay. And so what does it grow into?
Gizem Gumuskaya [00:03:59]:
So it grows into an anstrobot, and we are planting it into this extracellular matrix, which I can explain what that is in a second. But think of it as a proxy for soil. It's soil for human cells. It is a sort of nutrient rich environment that mimics the human body. So these cells that are originally coming from the human body now in the lab going to an environment that is very similar, except that it has these minute differences, which is our synthetic morphogenesis protocol, that at the end of the day, at the end of two weeks, they build themselves into something that has never been observed in or outside of the human body. So we're using the human DNA intact. We're not performing any genetic engineering, but we're accessing an additional layer of information by editing the morphogenetic code that changes their architecture.
Mizter Rad [00:04:51]:
What do you mean with changing the morphogenetic code, changing the architecture. So does that mean that that cell, that initial shape that that cell had is changed into something else?
Gizem Gumuskaya [00:05:03]:
Yeah. So morphogenetic code. So we're all very familiar with the genetic code. Right? That's our DNA, what's written in our DNA, except that when you look at the development of form in nature, we know that DNA cell is not sufficient. There are additional layers, like environment, like the information and signals the cells getting from the environment. On top of that, you have the epigenetics, which essentially determines which genes get expressed or not. So in addition to what's already written in your DNA, all these environmental inputs together as a package determine your form, your shape, your structure, and all of these determinants together are what sort of making up the morphogenetic code for different organisms, which essentially means that you can take a cell from an organism, don't touch its DNA, so do not add or delete any genes. But by just changing the environmental factors, like the soil it grows or what kind of chemicals it interacts or what kind of other sort of non chemical inputs the system is supplied with, you can change its morphogenetic code and get it to build itself into a completely new architecture.
Mizter Rad [00:06:13]:
And why is that important to me?
Gizem Gumuskaya [00:06:15]:
So that's important. That actually has implications for several different disciplines. In the constants, we've started a conversation for medicine. So in the context of medicine, if you wanted to create a bot, that we're trying to have it do a certain task. So we did say that biobots are, unlike traditional bots, do not have mechanical parts or chemical wiring, but also, very much like the traditional bots, they do execute tasks. They are useful, they do useful work. So, in biology, which is sort of the opposite of maybe what we say in design, we say form follows function. In biology, very much, function follows form, meaning, by changing the anatomy, the shape, the structure of a tissue, you can change its function.
Gizem Gumuskaya [00:07:07]:
So if we wanted to use these bats for different functions, different tasks in the body, they would need to have specific architectures. But those architectures might not be the ones that evolution happened to generate, but might be the ones that a human design or engineer wants to devise. So, for that reason, it's critical, it's really, actually groundbreaking to engage in morphogenetic engineering and to change the type of structures that these cells can build themselves into.
Mizter Rad [00:07:39]:
When you talk about the different implications that this could have in medicine, for example, what kind of use cases do you see the anthrobot can have in the near future, but also a little bit on the longer future in the medical sphere first, and then. I know there's other implications in other fields that I also want to touch after this, but please start with the medical field.
Gizem Gumuskaya [00:08:06]:
Yeah, so, short term. So, one of the applications we've showed in the first paper, which came out a couple months ago, we've showed that anthrobots can traverse, sort of move across tears in live tissues. So this is a great example for how morphology impacts function, how functional follows form, because these bots are able to engage in that sort of locomotion due to these specific structures on their surface, called cilia. So these are sort of like, think of it as like a rowing boat. If you have a boat in water, and then if you're rowing, you will propel forward. If only one side of your boat is rowing, you'll go in circles. If both sides are rowing at equal sort of thrust, you're going to go straight. If no one's rowing, you're not going to go anywhere.
Gizem Gumuskaya [00:08:58]:
So, very much like that principle, these bots have these cilia on their surface, and that's how they're able to move around. So that, in and of itself, could be useful. So if you are trying to say, well, you gave the example of clearing plaque, so you can imagine it by sort of bulldozing the adipose tissue from an artery, thanks to that ability to sort of thrust itself, you can additionally engineer it to, while doing that, secrete certain molecules that maybe melt out the adipose tissue, depending on the severity of the plaque. That's an example for how you can engineer to do different tasks. We have showed in that first paper, in addition to moving through tears in live tissues, we've showed that once we have these anthrobots interact with one another and aggregate into larger structures, we realize that they have the ability now, which is something they didn't have when they were in the format of single bots. They now have the ability to act as a bridge across that tear. One side of the bot docks on one side of the tear, the other side of the bot docks on the other side of the tear, very much like a bridge over water. What happens is, in the course of three days, this bridge enables this tear.
Gizem Gumuskaya [00:10:14]:
The particular tissue we tried was neurons, human cortical neuronal tissue, this tear to heal. So we saw neurons to heal themselves in the course of three days, which is really good news, because neurodegenerative diseases, like a lot of the Alzheimer's, cerebral palsy and stroke, a lot of those nursery diseases, are due to the neurons lost ability to heal. So we're hoping that this could be just like a very sliver. Right? We're just scratching the surface, but could be used perhaps for treating diseases down the line.
Mizter Rad [00:10:48]:
So you're saying that single anthro, like, at first you had single anthropods, and you didn't know how they would work together with other anthropots. And when you put them together, they generated this bridge that fixed what? Exactly correct.
Gizem Gumuskaya [00:11:04]:
Fix this. So, imagine a piece of tissue like monolayer of neurons, where there's a tear, there's a scratch. When you put the spots longitudinally over that scratch, it acts as a bridge, and it fixed the tear so that the neurons sort of would fill in that gap and then sort of, like, heal and regenerate.
Mizter Rad [00:11:26]:
Right. So it helps the body heal itself, almost.
Gizem Gumuskaya [00:11:31]:
Correct. We are currently trying to figure out, is it whether it's due to, like, migration of those neurons or whether new neurons, you know, have formed? I mean, doesn't. It's more like a scientific investigation, doesn't really matter. Maybe for the purposes of gap closure. So you can call it regeneration, but sometimes regeneration implies creation and proliferation of new cells. So we're not sure if that's exactly what's happening. But at the end of the day, the gap has closed, so there's healing.
Mizter Rad [00:12:02]:
The problem is fixed.
Gizem Gumuskaya [00:12:03]:
Yeah, but we prefer to the term healing for that reason, because when you say heal, like, you're not sure if it's because, like, a zipper getting close, or is it because new tissues being formed? So it's a more umbrella term.
Mizter Rad [00:12:14]:
I see. Okay, take me through the process of creating an anthrobot and then also applying it into, say, for example, a tissue or something.
Gizem Gumuskaya [00:12:26]:
Yeah, yeah, yeah. And before we move on, actually, I wanted to highlight something you actually had underlined, which is very exciting and interesting. Like, a group of antrobots were able to accomplish a task that a single antrobot couldn't. So they're also able to leverage collective intelligence in that way. When I say collective intelligence here, I'm not saying that they're intelligent beings and then they're making decisions. I'm just saying that collective behavior is perhaps better. And this is actually something that this behavior is seen in nature with ants. So there are these things called ant bridges.
Gizem Gumuskaya [00:13:00]:
So you can have an ant colony sort of form a trail, and then all of a sudden, they encounter a giant gap. No single ant can go across that big gap for the trail to continue. And then what they do is they hold onto each other with their bodies. They form a bridge. So, if you just google ants bridges, you'll see them, and then that's how they're able to sort of get across. So that's all sands are. Example of sort of collective behavior. And it's really exciting to see these concepts that naturally emerge in nature with, you know, different animal groups to also be observed with these synthetic, you know, constructs like the anthropots.
Mizter Rad [00:13:39]:
Let me ask you something. What? Why do you call it synthetic if, at the end of the day, you're not messing up with the genetic code of it, you're just sort of changing its form, as far as I understand, or changing the environment? That is it?
Gizem Gumuskaya [00:13:55]:
Yeah, that's an excellent question. So, with this work, I am also prompting us to reconsider what exactly synthetic morphogenesis mean. So, traditionally for synthetic morphogenesis. And that's also the. So, before Tufts and Harvard, I was a master's student at MIT, and that was sort of the education I received at MIT for synthetic biology. It has to be genetic circuits, there has to be exogenous genes, you have to insert new genes and promoters, and that's great. And there's a lot we can do with that. In my PhD, I realized that a.
Gizem Gumuskaya [00:14:31]:
That's really cumbersome, because, say, I want to create a completely new architecture and I want to express Cilia. Like, you know, in the case of anthro bats, we don't even yet understand 100% how Celia genesis happen in nature. That is a, like, it's a very complicated process involving like dozens and dozens of cells, a massive pathway, like encapsulating that in a genetic circuit and putting it into a new cell for it to grow cilia. Currently, with the sort of genetic engineering tools we have, it's not possible. So I'm also realizing it's not necessary. We can instead start with a. And this will connect back to the question that you asked, and I haven't forgotten, how do we make them? But instead we can tap into a cell type that already knows how to build the cilia, just not in the architecture natively that we wanted. So what we do then is that we don't really need to it to express this gene.
Gizem Gumuskaya [00:15:26]:
We're just changing the environmental sort of factors for it to express it in the sort of architecture that we want it to. So to sort of recap and like to say that why we call it synthetic, even though there are no trans genes, one of the theses is that to do synthetic morphogenesis, you do not have to use synthetic circuits, because synthetic morphogenesis is a result of the morphogenetic code, which is more than just the genes. So it's an alternative, new approach.
Mizter Rad [00:15:58]:
But. Sorry, when you say synthetic circuits, did you say synthetic circuits? What do you mean with that? Said the genes. The genetic code, exactly.
Gizem Gumuskaya [00:16:06]:
So why circuit? Why not just genes? And that is the sort of difference between the seventies, like genetic engineering, and the 21st century synthetic biology. In seventies, what they were doing. Okay, I'm going to take this one gene from this organism, put into this other organism, like I'm going to take the fluorescent protein gene from jellyfish and put it into mice, and I'm going to make fluorescent mice like it's transgenes, whereas in synthetic biology, you know, when we do use genes, what we're saying is that we're not just going to transfer one gene, we're going to create a complex gene circuit, very much like the electrical circuits that run everything, except that instead of transistors, we're going to use genes, and instead of copper wires, we're going to have these chemical wires. But at the end of the day, what doing that we're going to be able to accomplish is to create this combinatorial logic and complex behavior using the genes. So we're now engineering at the level of system, as opposed to at the level of individual genes. So that is sort of, in general, what is done in the field of synthetic biology. How can we put these gene circuits together to create complex behavior? And sort of through anthrobats, what we're saying is that that's amazing. Let's do that.
Gizem Gumuskaya [00:17:18]:
But that's also really difficult when we're trying to create complex, complex architectures like new biobats, and we can try to do that, but we might not need to. We have this now sort of alternative and complementary approach, where by bringing the environmental engineering, you could trigger radical change in the system towards design that you wanted, which is the case in anthrobots.
Mizter Rad [00:17:45]:
Interesting. All right, all right. So now tell me the process. Take me through the process of creating that anthropoid.
Gizem Gumuskaya [00:17:51]:
Yes. So this is a great sequel to that, because here's how it works in the lab. So, first of all, when we talk about synthetic morphogenesis, which is a sub field, let's call it, of synthetic biology, where we specifically focus on form. Using synthetic biology, you can do a lot of different things. People are trying to create new drugs, people are trying to classify cells to see whether there's cancer cells or not. Like, the sky's the limit. A sub field of that, that specifically focus on morphology from anatomy is the field of synthetic morphogenesis. So the number one thing when you are starting a synthetic morphogenesis project, let's call it, is, what is your end goal? What is your target design? What are you trying to accomplish? Which is a very sort of engineering mindset, right? Like, this is not how evolution generates structures.
Gizem Gumuskaya [00:18:48]:
Evolution is all trial and error. And at the end, you end up with a form, a shape that works, and then that gets inherited across progeny, across, like, generations. When we talk about doing synthetic morphogenesis, we very much bring our humanness, our engineering mindset into it, our goal orientedness, and we ask, what are we building? What are we trying to accomplish? So, for anthrobots, the architecture of the anthrobots were actually very much informed by a previous biobot called xenobots. So these were already sort of worked out. So these are biobots that are developed from frog embryos, and they were already sort of worked out in the lab when I arrived for my research. And what's being done there is that you just take cells from frog embryos and sort of sculpt them in a certain way, and then they end up as these pieces of motile tissue. And that was very interesting. So this came out, like, few years ago, a few years before anthrobots.
Gizem Gumuskaya [00:19:59]:
And that was very interesting for people, because that's really where we started talking about the concept of a biobot, where the system is fully cellular. But the biobots prior to that were always some sort of hybrids of a gel and cells, or some sort of scaffold that was leveraging both the inert materials properties as well as the biological cells properties. But if you're manufacturing it that way, you cannot have it self constructing. You cannot have this, like, okay, I'm going to seed it, plant. I'm going to plant the seed, and then the thing will grow. You can't do it because you're making these hybrid structures. So those biobots already existed. And then the xenobots after that were the first fully cellular biobots, except that the xenobots, the problem there was that they were derived from sort of frog embryos.
Gizem Gumuskaya [00:20:56]:
And when you try to make a mammalian, especially human, biobot, that means that you cannot use the same method, because you're not going to work with human embryos. So, for me, the goal was, how can I create a mammalian human biobot that looks like xenobots, which are, again, sort of ciliated, multicellular, but that is carrying 100% human DNA? So that's really tricky, because, again, you can just simply take the xenobot method, where you take frog embryos and, like, mess with them, and then create these motile tissues, because you can't touch the human embryos in that way. So then the really challenge for me was that, okay, I need to start because I.
Mizter Rad [00:21:43]:
Sorry, sorry. Why? Why can't you touch it in that way? Is it because the environment is very important for the human cells that you cannot really change their shape for ethical reasons?
Gizem Gumuskaya [00:21:54]:
For ethical reasons, yeah. So you can't really, like, there is a limit to. I believe it's day 14, I don't work with embryo, so you can fact check me on that. But I believe it's, like, beyond day 14, you can't really continue culturing human cells unless you're in IVF clinic, and that's your business. But for ethical reasons, you can't just experiment on human embryos. Yeah, that's a big topic in the US right now. You should have thought on that. You should bring an expert on that.
Gizem Gumuskaya [00:22:26]:
I think there's a lot there.
Mizter Rad [00:22:28]:
I talked to Professor Carlo Buletti. He talks about ectogenesis, and he has experimented with human embryos, I think a bit borderline still in the legal side.
Gizem Gumuskaya [00:22:41]:
But kept them alive till the last hour.
Mizter Rad [00:22:45]:
Exactly.
Gizem Gumuskaya [00:22:46]:
Yeah.
Mizter Rad [00:22:47]:
Last second.
Gizem Gumuskaya [00:22:48]:
Yeah, exactly. No, yeah. So that wasn't gonna work for us because, I mean, especially if we're, like, making biobots. Uh, really, you don't want to start with human embryos. So, for me, the challenge was, okay, I need to start with a human adult cell and still get this adult cell to undergo significant morphological reorganization as if it's an embryo, because that's what embryos do, right? Embryos start with, as I go, a like, fertilized egg cell, and then in the course of, like, few days, it undergoes a massive morphological reorganization and creates the more mature embryon than from their own adult organism. So that was the big challenge for me. That was my synthetic morphogenesis question. I want to create a multicellular spheroid that have cilia on the surface so it can move like xenobots do.
Gizem Gumuskaya [00:23:40]:
So that's the architecture we're targeting. I needed to carry human DNA so we can sort of deploy it in medical contexts. And I needed to come from a human adult tissue. And, you know, like we were discussing before, it is really not as simple as I'm gonna take all of these properties and put them into a genetic circuit, and then, you know, put that in a single cell, and the magic will happen. I mean, I think this will happen, but maybe in, like, 100 years. So not even 50. I think in a hundred years, we'll get there, but we get where?
Mizter Rad [00:24:16]:
Sorry, can you repeat that? In 100 years, we'll get where?
Gizem Gumuskaya [00:24:19]:
I think in 100 years, we will come up with a completely arbitrary description for what we want an organism to look like, a living tissue to look like, and we will be able to compile it, very much like compiling code into a series of genes in the context of a gene network, and then put it into a single cell and have that cell build itself into that final shape completely. But we'll have complete control at the genetic level. We do not have that yet, but still, it's not easy to, you know, interbots have a non trivial shape, like, it's still multicellular and silly on the surface. So that was still a significant challenge. And because our understanding of morphogenesis is not sufficient yet to do this compilation, basically, we had to come up with a whole new approach which doesn't rely on genetics, but also rely on environmental inputs. So the hypothesis was that let's look at human body, right? If we're trying to create a seleted tissue made up of human cells, first let's look at what evolution did for us, right? What's in the store? And so we already have cilia evolved in our body in a lot of different tissues. In the lungs, cilia run the in trachea, they run the mucosilia escalator. So when you inhale particles or dust, they sort of are sent back up.
Gizem Gumuskaya [00:25:58]:
It's a reverse working, like, anti gravity escalator, which is actually really cool. We have it in the ovidactyl epithelia. So in the ovaries that help with the propelling of the egg, we also have it in the brain ventricles. So, like, there are parts in the body where cilia already exists. So if you're trying to make a cilia tissue, and then if you want to use adult cells, then real, the question becomes, well, can I start with one of these cells that already know how to build cilia? These cells already know how to build this thing. Why am I trying to get a random cell to build cilia from scratch using genetic circuits? I mean, we don't even have that technology yet. Can't I just use a cell from the body that already have built the cilia? And then can I then engineer that cell to build it into the architecture that I wanted it to? Because here's the catch. When we look at these different tissues, none of them magically have spheroids with cilia coating, because that's just something evolution doesn't need.
Gizem Gumuskaya [00:27:02]:
Like evolution, I guess, didn't need biobats, but we do. So the question then became, okay, let's pick one tissue. I picked the human trachea because there are just. That's for, like, logistical reasons. More cells are available from the lungs than brains or oviductyl epithelia. And here's a really fascinating thing about, I mean, really, like, natural architectures in general, but particularly human body. So in most epithelium, most organs. So epithelia is sort of the think of it as a skin tissue, but lining the organs.
Gizem Gumuskaya [00:27:37]:
So we have the skin tissue lining our body, but sort of extension of that in the lungs. We also have epithelium, our sort of skin. Epithelium is specialized to have hair, to have pores to protect against the elements. Well, the skin tissue in the lung has sort of evolved to do different things. And one of those things is to generate cilia and also generate mucus to run that mucosa escalator. At the base of this epithelium, there exist semi stem cells. So these cells are not fully stem like, they're not embryonic stem cells. Embryonic stem cells can become whatever the hell they want.
Gizem Gumuskaya [00:28:21]:
I mean, that's another really fascinating thing about the human body. The cell, say, that became your eye and your kidney, they are coming from the same common ancestor. So that cell knew. The embryonic stem cell knew how to make a kidney or how to make eye, and then it differentiates in the course of development, which is, by the way, another testament to the morphogenetic code. Right? Because different cells in our body have completely different architectures, but they have the same genome because everyone is coming from that first fertilized egg.
Mizter Rad [00:28:57]:
So the cells in the eye, let's say they come from the same egg as the cells in your teeth, but the shape or the environment, the environment is the one that determine how the eye or the teeth behave or act or the kind of functions they have, how they look.
Gizem Gumuskaya [00:29:16]:
Yes, yes, 100%. So your eye cells and your, like, I don't know why I keep going. Eye and kidney, your eye cells and like fingertips, like, those two cells have exact same DNA, like exact same, no difference. But they have epigenetic differences, meaning they have the same genes, but they don't expand the same genes because genes alone don't make the world go around. What really makes the world go around are the proteins. What genes encode for. Cells divide and differentiate. And in some cells, certain genes get silenced and certain others get expressed.
Gizem Gumuskaya [00:29:54]:
And that's how the morphogenetic code creates this variability that we see in our bodies. That is fascinating. And.com. that's the power of embryonic stem cells. But here's the interesting thing. Even in adult tissue, no longer an embryo, at the bottom of these epithelia, there are semi stem like cells, meaning these cells can still become a couple different cell types, only within their own organ. Though if you take lung, there are semi stem cells. They're called progenitor cells.
Gizem Gumuskaya [00:30:28]:
They're semi stem in that they can't just magically become in the body and cell from another organ, but within their organ, they can become different types of cells. So they can become like secretory cells, or they can become ciliated cells, or they can replicate themselves and stay semi stem. So why this is important? Because those cells are the basis for those cells. Know how to make cilia, theoretically. So those are the cells. We start making the antrobots with each one of these cells. So they're isolated from the lungs, these semi stem like cells. We know that these guys know how to make cilia, and we're going to take them and then we're going to do a bunch of different experiments on them to see.
Gizem Gumuskaya [00:31:16]:
So, I mean, I think that is the sort of dark year of my PhD, where, like, every day it was a different experiment, and the result was, nope, nope, nope. You can culture them in all kinds of different environments to see how the environmental inputs are impacting them. If you're starting with these cells for them to generate cilia, what else do you want them to do? You want them to create a spheroid. You want your biobot to be multicellular because a single cell is too small. You want it to be a multicellular construct. Then I need to get the cell to become a multicellular cell. How can I do this? I can force it to kind of merge with its cells of different. Sorry, cells of similar.
Gizem Gumuskaya [00:32:00]:
Basically. Like, I can try to aggregate them, I can try to differentiate them in different ways. So there are a lot of protocols for this with the goal of getting that single cell to self replicate and then build itself into a spheroid, and then in the process, express the cilia, because, again, it all comes back to, what's my goal? My goal is to create a multicellular spheroid that has cilia on the surface so it can move around.
Mizter Rad [00:32:29]:
Okay, so the cilia is important because it allows these. The bot to move around.
Gizem Gumuskaya [00:32:36]:
Exactly.
Mizter Rad [00:32:36]:
So it's like the little feet.
Gizem Gumuskaya [00:32:38]:
100%. Yes, yes.
Mizter Rad [00:32:40]:
Okay. And so those feet, they need some kind of energy to move around. So where is that energy coming from?
Gizem Gumuskaya [00:32:50]:
Well, that is the other fascinating thing about biology. It's just coming from glucose. I mean, why biology is literally magical is because it's taking these simple building blocks, like glucose, and it's creating massive complexity and order from them. So that's why I'm saying in the course of, like, two weeks that an anthro grows, all I'm doing is, like, feed it twice. And what I'm feeding it was sugar. Sugar? Yeah.
Mizter Rad [00:33:18]:
Okay, so wait, so you got to know that you wanted to get a multicellular.
Gizem Gumuskaya [00:33:25]:
Spheroid bot?
Mizter Rad [00:33:27]:
Yes, spheroid with, you know, starting with a semi stem cell from the lung or the trachea and ending in this, again, multicellular spheroid called now antrobot that had cilia so that it could move around and you figure that if you give them glucose or you know that if you give them glucose, it will. That will be used as the fuel. But I want to know now what now you know that. And now how do you put that at work? How do you potentially put this in a real human environment, in the human body?
Gizem Gumuskaya [00:34:07]:
Right. So once in.
Mizter Rad [00:34:08]:
Or is there something that I'm missing? Of course, this is very general. Right. It's a very simplistic way of putting it, but I want to know exactly what's next then.
Gizem Gumuskaya [00:34:18]:
Yeah, yeah. And then sort of to, like to close the loop on that. The experiment that ended up working out of the semi stem cells from human lung that did end up giving us the architecture we wanted, was this method of growing them into something called lung organoid. So in the lung organoid. Right? And I know you had an episode on that. So in the field of organoids and how it's different from synthetic morphogenesis is actually, again, the goal in the field of organoids, your goal is to create something that's as similar as possible to the organ that the cell is coming from. So if you're creating a three dimensional lung organoid, you're trying to make it look like the lung as much as possible. Trachea in this particular case.
Gizem Gumuskaya [00:35:01]:
So when we grow it into a long organoid, what we saw was that the single cell did form the spheroid, and it also did form the cilia. But here's the catch. The cilia was stuck inside. Cilia was sort of looking inward. Think of it as like earth, and the cilia was, like, facing the magma, not the sky. Why? Well, because this is how it is in the body. Like, the trachea is a tube with cilia inside, and there's a lumen in there that the cilia is moving the mucus around so that particles and pathogens that get into your body can be kind of sent back up. And it's really funny, almost comical.
Gizem Gumuskaya [00:35:42]:
And then we're like, okay, we gotta. I was like, I need to get this to Philip. And, yeah, it's literally trying to make lung organize Philip, because I need celia on the outside. If the cilia is inside, it cannot help me with locomotion.
Mizter Rad [00:35:54]:
Right.
Gizem Gumuskaya [00:35:55]:
Uh, and. And this is where sort of that morphogenetic code and, like, environmental engineering is coming from. I'm like, why are these cells looking inside? Like, what's the hypothesis? They're looking inside because in the body, like, the gross environment, it's very thick. So the cilia is conditioned to look out away from sort of the thick matrix environment. That's just something we know from lung biology. Well, so if I like, remove the matrix around them and instead put really like liquid media that might get them to flip. So that was the hypothesis, and that's exactly what happened. So I grew the cells and then remove, so they grew in that matrix.
Gizem Gumuskaya [00:36:38]:
So, so that's the soil, if you will. And then once there's spheroid, you take them out from the soil and then you put them into a more liquid environment and bombard them with this thing called retinoic acid that also cilia, like to trick cilia to migrate outward. That was, at the end of the day, the protocol that ended up working. And that way I was able to achieve the final architecture that I was trying to get these sauce to build in a way that they self construct. Then what happens? How do we go from here to humans? The first step is really to test the cinema. Human tissue, like a tissue. In fact, the very first step is to not even tissue taken out from humans, but some sort of a monolayer, ideally human monolayer, or some tissues from rats. So I've tried different things in parallel.
Gizem Gumuskaya [00:37:32]:
I did put these into tissues from rats, like esophagus, or the branching aorta or the jugular. The problem was, once it's inside that dense tissue, it's real difficult to image it and want to do more imaging and experiments. So just decided to switch to a simpler tissue type, just like two dimensional tissue, but at least with human cells, so it's more relevant. And yeah, ran the experiments in that context, and that's where the neuronal healing happened. What would need to be done from here is to start putting these into ex vivo tissues. So if you're taking cells and then growing it into laboratory, that's called in vitro. So that's the sort of first step, and that's where we've tested these and saw the healing behavior. From here it would be ex vivo.
Gizem Gumuskaya [00:38:21]:
So those would be the tissues that you would take from humans, either with surgery or like biopsy, and you would put the bots into those tissues. So that's what I was trying to do with rats. But there were some imaging problems, so those would need to be, I mean, they're not difficult, they're solvable, but those would need to be solved. And then once, you know, based on what we see there, the step after that would be to move on to animal or like human trials. And it's all sort of like building on top of each other. We would first need to see how they behave.
Mizter Rad [00:38:55]:
Xv well, all right, so, you know, if everything goes well and if they behave well when they are tested in a human body, how do you imagine this eventually rolling out in a mass market situation? And now I want to jump a bit ahead and look up to the future. I'm trying to understand, how would you imagine this in a context where, let's say I have an injury in my skin or I have an injury in my lung, actually, and the doctor knows that with antrobots, this could be solved. So I go to that doctor and what. How would that be applied? How to does the antrobot or the army of anthropods get into my lung and fix it? Can you. Can you imagine that? Can you explain that to us? Yeah, I know this is not happening yet, but how to make that clear?
Gizem Gumuskaya [00:39:55]:
Yes, this is not happening yet, but this is sort of what we aspire to, let's say. So you would go to a doctor, and then if it turns out that and, you know, if answer bots are the answer to a problem, you would simply. You would not give a tissue from your lungs, because if they're like, you know, cutting you up to get tissue from your lungs like that, a little bit, like, defeats the purpose. Because the whole point of intro bots is that, in an invasive way, using, you know, synthetic bots that carry your DNA without your body, you know, getting cut up or getting inflammated, we can do things. So no one's gonna, like, cut up the lungs. You could just give a skin biopsy or something from the cheek. Cheek, like, cheek swab. Because, remember, all tissues in your body carry your DNA 100%.
Gizem Gumuskaya [00:40:42]:
So there is this thing that's established. There's like, super cool thing, and we've known this for a while. Any cell of yours can be turned to an embryonic state of yours and can be redifferentiated, so it can be de differentiated into an embryonic version. Embryonic stem cell. They're called induced pluripotent stem cells, which means that that skin cell, now, because it's in an embryonic state, has the potential to become a lung cell. We would de differentiate into a hipsc in this pluripotent stem cell and then differentiate into a lung cell as if we took it from your lungs. And then that would be the basis for then differentiating them into anthro bots. That, again, carrier DNA.
Gizem Gumuskaya [00:41:28]:
I mean, throughout this entire thing, we're not changing the DNA, we're just changing the environmental input, because all we need to change is the morphology. We don't care about the DNA, we just care about the architecture. So that's how your enterbaskary urdna would be manufactured. And the hypothesis is that, well, the immune system cares about DNA because the DNA is the basis for the proteins, which are like personal signature markers at the surface of our cells. When immune system encounters with this, I mean, technically foreign object, because synthetic tissue that we're putting into your system, it did not evolve or develop with you, but it will stay under the radar is the hypothesis, because it will have the exact same surface markers as the surface markers your immune system is accustomed to seeing in your tissues.
Mizter Rad [00:42:22]:
And does that mean their hypothesis that, say, our body won't reject it because the genetics of it are unchanged, identical.
Gizem Gumuskaya [00:42:34]:
Correct.
Mizter Rad [00:42:34]:
But the morphology is the one that is changed.
Gizem Gumuskaya [00:42:38]:
Exactly. The morphology is different and that accounts for the functionality. But for the immune system, that's not what the immune system is looking at. The immune system is looking at surface marker proteins, which is informed by the DNA.
Mizter Rad [00:42:52]:
So if you haven't changed DNA, okay, so you're bypassing.
Gizem Gumuskaya [00:42:55]:
Yeah, exactly.
Mizter Rad [00:42:56]:
You're bypassing the immune system. You're staying under the radar and then you kind of. But how do you inject that? Like, would you imagine that this is injected into your bloodstream or is this directly, somehow dropped into your, in this case your lung somehow, or does that matter?
Gizem Gumuskaya [00:43:15]:
It would depend on the application. If it's something where say, you want to use answer bots. And these are again, like the only application we've so far like showed is the neural application and other things are being worked on. So I just want to emphasize that these are currently hypothetical scenarios. But imagine you want to use antrobots to patrol for cancer formation. You want them to sort of let you know in a certain way that there is a cancerous tissue forming. So if you want to use it as diagnostic, you would just put it into the parts of the body almost like live with you symbiotically, that are prone to cancer. So in that case you would want to inject it into the tissue directly, but otherwise, or you can inject into blood, I mean, it depends on type of cancer.
Gizem Gumuskaya [00:44:02]:
You can also inject it into the bloodstream and have it just sort of circulate in your body and, you know, detect the sort of onset markers that way. Or another example for. So if you wanted it to, for example, bulldoze a like, sort of almost like arteritis, like at the joints, there is some sometimes like built up layers of plaque. Like, if you wanted them to kind of bulldoze those sort of build up, then you would inject it into the joint directly. But if you wanted to chase some sort of bacteria in the gut, then you would inject it into more the bloodstream. So it just really depends on what it's going to be deployed as.
Mizter Rad [00:44:49]:
Tell me something. Once you deploy them, let's say, in your knee, you inject a bunch of arthrobots in your knee with the objective to bulldoze the plague or whatever residue has formed around your knee. And that is painful. Do you, after the deployment, do you have any sort of control of those bots? Can you pull back in case something goes wrong? Are you considering that and working on that as well?
Gizem Gumuskaya [00:45:19]:
So what we've seen with anthro bots is that they degrade after sort of like two months. So it is. And we've, like, surveyed like thousands and thousands. And I. It's so far zero case that I've seen anything become sort of like turmeric. And I. Girl, like, never dies. Like the killer bot immortal.
Gizem Gumuskaya [00:45:39]:
That never happened. So the hypothesis there is that you would deploy this in the body, it would do what it needs to do and then degrade and sort of become debris, which is then, you know, not different than the derby that your body is getting rid of in the. In the synthetic. In the synthetic biology field in general. Um, there, this is a big, like, safety is a big, you know, concern, especially if you're using, you know, exogenous genes, because then you don't know how those genes would interact with your own genes. What people generally do are these things called kill switches. So the, you know, the circuit would have sort of a suicide gene, so to speak.
Mizter Rad [00:46:25]:
Like a trigger to pull it.
Gizem Gumuskaya [00:46:27]:
Yeah, exactly. That you would be able to trigger with a very specific drug. So if you want to sort of abort mission, you would just take that drug and it would shut down the circuit in your body. Because synthetic biology, like in general, there's a lot of interest in personalized medicine, which is what, like antibodies is a great platform for. So, yeah, that's something people think a lot about.
Mizter Rad [00:46:49]:
Tell me something, do you have an idea of how much it costs to manufacture an anthrobot today?
Gizem Gumuskaya [00:46:58]:
So, I mean, if you already have a biolab, and assuming you have infrastructure, the, you know, cells, I mean, we get them commercially sourced. So actually, anybody like, any lab can get them sourced. It's about thousand dollars. But, you know, we're not like doing any surgery. We're not like it's actually pretty clean for $1,000. There's like nothing in biotech, because usually these experiments are really expensive, but, and from there it's just like some reagents, like setting up the experiments and doing the creating. The bots themselves are really not, they don't really have any additional extraneous costs. What's cost is to like run a biolab in general, or like run a hospital, but interbots themselves, like the protocol doesn't really cost much.
Gizem Gumuskaya [00:47:57]:
And here's also why. So the self construction, right? Like, I think it's really exciting to create a self construction constructing synthetic structure, because that is just something we have never, right in the sort of history of like civil engineering or architecture, like in the history of humans building things, that is something we have never gotten our hands on. That's sort of that ability, that superpower is almost exclusive to nature, right? Like trees self construct, like, humans self construct, flowers self construct, but skyscrapers don't, or cars don't, or computers don't. So human made objects never have had this ability to build themselves until now. Which is why I think synthetic morphogenesis is so exciting, because it is giving us access to properties in human designed products that so far had been exclusively reserved to nature. So for that reason, I'm really interested in self construction and autonomous growth. But what that also really helps with is scalability and driving the coast down, because if you are just throwing thousands of cells there, and if each cell is building itself to an anthropot, that's a massively scalable pipeline compared to what we usually do, which is to build things one by one by hand, or like 3d print them one by one. Then, like, you have a bottleneck, your 3d printer is your bottleneck, or your like, human who's sculpting them is your bottleneck.
Gizem Gumuskaya [00:49:38]:
Or if you're using like moles, then well, before like manufacturing your stuff, you need to manufacture the molds. Like it's, it's, it's really helping us tap into that, like scalability. And that's what's gonna drive the, like, cost down. Because at the end of those two weeks, you don't just have one antibiotic thing, thousands of them, and you didn't put in any additional effort to go from one to thousand.
Mizter Rad [00:50:01]:
Tell me something you mentioned in a different interview, that nature can be a design medium. What do you mean with that?
Gizem Gumuskaya [00:50:10]:
So, I mean, this is what I mean by that. I think so far especially, I think I know because of the education system we are sort of conditioned to think about nature as this thing that's just sitting outside, waiting to be investigated by scientists and understood and mapped and written books about. But when we look closely, we're realizing that it's not just that passive entity waiting to be understood, but it's an active, real living design medium that we can engineer and get it to build things that are useful for us and not limit us to what evolution happened to generate. So in that sense, my starting point was design. I studied architecture in my undergrad and learned my masters, too. And then realizing that this sort of, kind of the ultimate architecture is out there in nature and is governed by this code, because we think of nature as magic, right? It's just black box. It's just doing its own thing. Well, it turns out underneath that magic, there's tremendous amount of logic, and we're really good as humans.
Gizem Gumuskaya [00:51:12]:
Engineering with logic, right? Like, look at everything we built. So why not bring that engineering sense, that design sense, into nature, and conceive it as a design medium?
Mizter Rad [00:51:23]:
And so when someone approaches you, someone with a more conservative background, and tells you, hey, gizem, you're playing to be got, nature is to be let alone. What would you tell them?
Gizem Gumuskaya [00:51:38]:
I don't want to get into, like, religion, but I think, like, I would probably, as I do when I get really, you know, religion or politics related questions, I would just say, well, if, you know, God didn't want us to edit it, why did she make it so editable? Why is there just, like, massive logic? But of course, some think that is part of God's plan. People believe in whatever they want to believe in, and however they want to explain things. But there are a lot of thoughts out there that don't think bioengineering or evolution is not, you know, antithesis of, like, God. So I just want to be, like, respectful of, you know, respectful of everyone's beliefs. And I think people should just decide for themselves. But I think I would just. My humorous response would be, well, you know, if she didn't want us to edit it, why is it so editable?
Mizter Rad [00:52:37]:
Tell me something, Gizem. Do you fear in any way that one day we end up in a fully synthetic world, or do you already think that we're kind of in a synthetic dimension which is not aware of it?
Gizem Gumuskaya [00:52:55]:
I mean, I think that, like, yeah, the boundary between, like, culture and nature is really blurry right now, because some believe, like, engineering to be an extension of nature, because it's an extension of humans, which is an extension, which is a product of nature, right. Very evolved within the earth's ecosystem that we, I guess, refer to when we say nature and beyond. So already, like, it is not clear where nature starts and, you know, culture begins. But I think the main difference here is this sort of intentionality. So there are already, like, a lot of, like, when you think about agriculture, humans have already affected the, you know, nature a lot. And I'm not just talking about global warming, but, like, when you think about, like, husbandry practices, right? We are, like, breeding, like, selective breeding or, like, you know, crossing different animals. Like, these are all actually sort of, to a degree, biotechnological practices. It's just that they would take a very long time, and they're very indirect.
Gizem Gumuskaya [00:54:04]:
So I think we've been doing this for a very long time, like, since the dawn of human civilization. I think that's when biotechnology started. Probably with agriculture or even just with, like, building things before agriculture. I think we're just sort of getting better at it and better and faster and more intentional. So although this is going to give us dimensions that we didn't necessarily have access to before, like, building self constructing objects, I don't think it's, like, fundamentally different in terms of how we're interacting with our planet. I think we're sort of exerting our agency to nature, which is something we've been doing, which is almost like the.
Mizter Rad [00:54:47]:
Definition of civilization when you talk about self constructing objects. We already touched the implications that this could have in the medical scenario. But what else can this antibodies improve or help with?
Gizem Gumuskaya [00:55:05]:
And also. Right. And not just anthropots. I think, like, by the way, antrobots were just sort of, like, one example for a biobot. I think we will be able to, like, develop all kinds of different biobots. So there it's just like, that's just, like, example number one. But in general, like, biobots and the field of synthetic morphogenesis. So other sort of sectors, like, beyond medicine, I think really the number one is going to be the construction industry.
Gizem Gumuskaya [00:55:30]:
That is another field I am very passionate about, just like, from my background. And that's the reason why I did my PhD in biology, because I truly want to get under the hood and understand how can I make this a reality. Because so look at global warming. I mean, like, close to half of those greenhouse gases or result of construction industry. Humans just trying to build things and, like, pumping CO2 into the atmosphere in the process. And then you look at nature. There's already this, like, amazing construction paradigm. And instead of pumping CO2 it actually sequesters CO2 from the environment.
Gizem Gumuskaya [00:56:09]:
And now we're understanding that, you know, there isn't built in code and that we could edit it. So one of the things I think really will happen in the next decades is that we're going to bring this idea of synthetic morphogenesis into creating more sustainable building materials. So, like, creating self constructing buildings is something I think we'll be able to do. I mean, 50 years, I think that we would hopefully 50 to 100, we'll get there. Depending on how much funding I get, it could be faster, but, yeah, so, like, I think that is really exciting because when you look at. We're building amazing things, but when you look at, like, humans construction practices, when you look at human construction practices, like, it's a lot of these sort of external machinery, external tools, raw materials, like descriptive blueprints, skilled builders, it's all about imposing shape onto the matter. Whereas, yeah, nature generates the form in a bottom up way. So I think that by leveraging that morphogenetic power in the types of things we might want to create, we'll have this whole new paradigm for building things.
Mizter Rad [00:57:34]:
This morphogenetic power that you talk about can be used to, for example, build a house out of genetic code on an organic, unorganic matter. Out of nothing, basically.
Gizem Gumuskaya [00:57:52]:
Well, out of a seed.
Mizter Rad [00:57:54]:
Out of a seed, yeah, not out.
Gizem Gumuskaya [00:57:56]:
Of nothing, but, yeah, out of a seed that's sort of engineered to build. Because the goal here is to, like, sort of hardness that morphogenetic power to build whatever you want to build. This could be a biobot to put it into humans. This could be a sort of more robust bot than, say, like anthropots to put into the rivers. Right. Because you wouldn't be able to put that with, like, you wouldn't be able to. That would, like a mammalian bot, because it would get you, you know, mammalian bots. Human bots require, like, sterile environment, but you can create a more kind of robust bottom, but to put into rubber streams, you can work with plant cells or other sort of.
Gizem Gumuskaya [00:58:36]:
Maybe more from the. I'm thinking, like, more calcified tissues to create, like self constructing bricks or self constructing building materials. It's really like this idea of nature is becoming your palate, because across the nature, across all kingdoms, morphogenetic code just, you know, sort of behaves in a similar way. It's all bottom up construction. So, yeah, I mean, I do think that we're going to be able to build things. I mean, it might not be like in that comical sense of plant the seed in like a house with sort of like classic, like, roof groves. But I'm thinking it would be more like building blocks that we would like sort of harvesting bricks, and you can have like a brick field, and then, you know, you could just sort of, instead of using cement, you would be able to just grow your bricks. That is an example.
Gizem Gumuskaya [00:59:41]:
I mean, I think that as more people get involved, there will be a lot of different ideas how we might be able to tap into this power.
Mizter Rad [00:59:50]:
Interesting. Super interesting. Gizem, is there anything else you would like to share? How can people reach out to you? How can people learn more about what you're doing? Where can they connect with you?
Gizem Gumuskaya [01:00:04]:
Yeah, so I am on Twitter. Gizamgumskaya is my handle. I'm also on Instagram. Doctor Dr. Gizam Gumuskaya. So, yeah, feel free to reach out. I think that the more people get involved, the more ideas will emerge. And I think it's really important that people who are interested in design and synthetic biology really try to get under the hood.
Gizem Gumuskaya [01:00:34]:
Biology. I mean, these days, just like learning things is really becoming easier and easier. It's just if you have the curiosity and the passion, I think that a lot of these things are completely learnable.
Mizter Rad [01:00:48]:
Amazing. Well, Gizeh, it was a pleasure to have you. I'm really happy and grateful that we managed to talk and record this episode, and I hope to see you anytime soon, maybe with a. With an antrobot inside our heads, cleaning up whatever mess is in there.
Gizem Gumuskaya [01:01:08]:
Yeah, well, I'm gonna need a swan for that. But.
Mizter Rad [01:01:15]:
Here at the Mizter Rad show, we provide firsthand information straight from the original source of knowledge. The personal opinions of our guests don't necessarily reflect those of Mizter Rad. This show is brought to you by the Rad House, an unbiased, transparent agenda, less independent media house. Our theme music is written and produced by Marco Mello.