Media Register to Attend 

What: Scientists create tiny biological robot "healers" assembled from human cells

When: November 29th, 2023. 3:00 PM ET


  • Michael Levin, PhD, Distinguished Professor at Tufts University, School of Arts and Sciences
  • Gizem Gumuskaya, Synthetic Biologist & Architect, PhD Researcher at Tufts Biology Dpt. & Harvard Wyss Inst.

Where: Newswise Live Zoom Room (address will be included in follow-up email)


Scientists have created tiny moving biological robots from human tracheal cells that can encourage the growth of neurons across artificial ‘wounds’ in the lab. Using patients’ own cells could permit growth of Anthrobots that assist healing and regeneration in the future with no need for immune suppression.  Lead researchers Prof Michael Levin and Gizem Gumuskaya from Tufts University will provide a brief commentary on the science and potential impact of this discovery, followed by Q&A with reporters. 



Thom Canalichio 0:00
Hello, and welcome to this Newswise Live Event. Members of the media are invited to submit their questions on the chat. And we will also be providing a recording and a transcript of the session. The news and the news release and the paper being discussed in this session are under embargo until tomorrow, Thursday, November 30, at 3pm. So please note that nothing can be published about it until that time. And I will go ahead now and turn it over to Michael Silver, Associate Director of Media Relations at Tufts University, and also share his contact information in the chat so that you can get in touch with him if you have further questions after the session. Thanks very much, Mike. Please go ahead.

Michael Silver 0:40
Okay. Thanks, Tom. Thanks so much. I'd like to welcome all of you to this press question and answer session. It's on some very exciting research coming from the laboratory of Professor Mike Levin. Mike is the is the director of the Allen Discovery Center here at Tufts. And he's the associate an associate faculty member at the Harvard Wyss Institute. Together, Mike with Dr. Gizem, Kooskia. at Tufts, they've been exploring ways in which cells can be reconfigured, and reassembled to do things that they might never have done in their natural environment and the body. In this case, they've created these tiny multicellular assemblies from human tracheal cells, called anther bots that explore their surroundings and even encourage the growth of neurons. So I'd like to open up this session by inviting Mike in his attempt to give us a quick overview of the research to tell us something about what antibodies are, what they might be able to do. And then we'll follow that with questions that you could submit in the chat window. So Mike, do you want to take it away? Sure.

Michael Levin 1:51
Yeah, thank you, everybody, for coming. I would like to introduce you today to an interesting story that we're developing about the untapped potential of your own body cells. And I want to just start with a very brief little video that you can see here. So here's this, here's this little creature swimming around. And if I showed you this with no particular preamble, and I might ask, What do you think this is, and you might think that this is a some sort of primitive organism found in a pond somewhere. And I could tell you that we know what the genome of this organism looks like. And the genome is actually 100% homosapiens. And so what you're looking at here is a kind of reboot of adult human patient cells taken donated by people undergoing biopsies of of their tracheal epithelium. And what we found is basically that in particular environments and give them we'll talk about how that how that works, they can self assemble into what we think is a bio robotics platform for both understanding the kind of untapped qualities of cellular collectives in our bodies, and the different uses to which they can be put. So that's, that's my introduction, and Gizem will say a few words.

Gizem Gumuskaya 3:07
Hi, my name is Dr. Gumuskaya, as of two hours ago. I am Gizem Gumuskaya, I'm a PhD graduate from Mike's lab at Tufts University and Harvard based Institute. I'm the lead, first author of the study. And one of the things we really wanted to explore here is to see how we might be able to sort of treat nature as a design medium. We tend to think of biology nature as this thing sitting outside waiting to be understood. But we believe in the lab that we can actually reprogram it to create couple of new architectures and functions that are not found in nature unnecessarily, but could still be very useful to us humans. And this is what we'll talk about today in the context of Anthro. 

Michael Silver 4:02
That's great. Thank you. Thank you. Does anyone like, want to open up with maybe a question that I wanted to introduce here. So part of the magic of antibodies is that they can self assemble from a single growing cell, and the dividing cells organized into their tiny body plan in a way that achieves a function like movement? So as you look to the future, of building different types of antibody architectures, what kinds of tools and approaches would you use to encourage them to grow into different shapes and new functions? I mean, for what you know about how cells grow into, you know, tissues and organs? What do you use to kind of make new antibodies?

Michael Levin 4:50
So, this is this is part of the larger question that we asked in our group, which is, how do you convince groups of cells to build the things that you need the various origins for Regenerative Medicine, novel bio robotics, in vitro tissues for transplantation and so on. And what we've been looking at is the role of various kinds of interfaces to the cells that you can use to control their behavior. So this includes electrical signaling, which we especially study in our lab. This includes various chemical signals properties of the environment that gives them manipulated. In this case, it might include all sorts of modalities such as light and vibration, and just all all the different ways that you can manipulate the environment of cellular collectives that we are now trying to learn what kinds of stimuli convinced the cells to build things with different form and function.

Gizem Gumuskaya 5:47
And for answer about specifically, I guess I could add is that we have learned that even though Anthro bots build themselves from the single cell, and even though every single single cell is treated in the same way, the resulting architecture comes in different flavors. So in the paper we discuss, specifically in figures, three, and four, how we've uncovered that there are three different types of Anthro robots and that how that maps on to their functionality as well. So our first step would be to understand what makes them because they're all coming from the same progenitor cell, what makes them differentiate in their developmental trajectory into these three different architectures, and then deploy that sort of amplify that in that initial single cell to perhaps create populations that are entirely one flavor or the other. So understanding is the first step towards being able to engineer it.

Michael Silver 6:53
So that's really interesting. I mean, is this, the you have these variety of types from antibodies are basically grown from one type of cell? So imagine that, you know, you in the future looking, I'd be looking at constructing antibodies combining different types of cells with different functionalities, right?

Gizem Gumuskaya 7:17
Right. It's very similar to the human body. I mean, we have all these different types of cells, right, kidney cells, ears, yourselves, like I cells, but they're all coming from one progenitor because we all grew from single cells. But during that developmental process, the differentiation happens giving rise to different cell types and different tissue architectures. So we essentially observed the same phenomenon in the context of Anthro bots, which is very surprising, because it's this synthetic construct. Yeah, it's recapitulating, some of these hallmarks of biological morphogenesis. Yeah, but

Michael Levin 7:54
just Just to add, Mike, I think I think the other thing you were you were indicating is that we could in theory, add other cell types here. And that's absolutely true. Right? So in the future, we can think of more more and more complex constructions that include other types of cells. But the thing that that is really amazing, I think is M was just pointing out is that even with just this pure cell source, you're already getting several types of behavior, several types of, of shape. You know, there's, there's incredible variety, even even even already in this in this system here.

Michael Silver 8:30
Right? And the the so the, the tracheal cells that were used here, did they actually differentiate the different types within the Autobots different kinds of functions or types?

Gizem Gumuskaya 8:45
Right, so each answer about starts from a single cell, and that single cell gives rise to three main types of cells, one being misleaded cells, those are the lokmat of appendages that make the answer about moving to first place. Second being goblet cells. Those are the cells that secrete different molecules, predominantly mucus. Since the precursor cells are coming from the human airway. The mucosal escalator, mucus producing cells are significantly represented. And the third type of cell is the progenitor cells. So those are the sort of semi stem cells that live at the bottom of the human human airway epithelium. And those are the types of cells that we would start with the single cell. So as that single cell develops and gives rise to these different cell types, it makes use to enrich its own population as well.

Michael Silver 9:47
Great, thanks. So there's a possibility for multicellular assemblies like anther bots, becoming a new therapeutic model. already from what we're reading in your research. So this would be alongside other categories like small molecule drugs, larger biological drugs, gene editing, of course, each has their own applications and limitations. So what sort of niche might you think arthropods would have attributes would have? And how does that compare to what these other types of therapies can do and can't do?

Michael Levin 10:24
Yeah, let's see, one of the most interesting things to think about across the spectrum of different Therapeutics is the kind of level of competency of your therapeutic agent. So you might have a drug and drugs have a very specific function that they perform. And that's and that's it, you know, that's there's a molecular interaction that they have with other cells. And that's it. And you might have things like insulin pumps, or various kinds of pacemakers. And eventually we'll have various implants that are AI powered, and are fairly intelligent in terms of taking different actions under different circumstances. So Anthro bots offer an amazing possibility, which is that the cells being being living already have a huge amount of machinery for sensing Well, in this case for locomotion for signal amplification, for making decisions for, in some cases, some cells can form memories, and so on. And so this is this is really the ability to take what is potentially an extremely sophisticated, and this is, this is why we can think of these as bio robots. That's not the only way to think about them. But one way to think about them is as bio robots, because that lens emphasizes thinking about what they can be coaxed to do that's useful, right? What controllable functions do they have. And so you can imagine, all sorts of all sorts of possible biomedical uses in the body, where we really take advantage of their competencies, right, not only do they move, but the cells have just a whole panoply of sensors and, and other things that they can do that is really hard to try to duplicate with, with either nanotechnology or, or some other large scale, large scale engineering. And they also, they also occupy a really interesting size scale, because they're much bigger than the kind of nano nano technology that's been talked about for decades. But they're quite a bit smaller than what we can actually achieve by traditional engineering. And because I'm getting amplify some of that.

Gizem Gumuskaya 12:29
Yeah, one of the sort of most exciting things about Anthro bots is that menu, look at programmable, sort of medical applications out there. They're usually at the single cell or single molecule level. But here we're harnessing a higher order multi multicellular architecture. And that brings up the ability of, for example, more complex behaviors like motility. A single cell can't really reach to the motility levels that we're able to accomplish here by deploying multiple multi isolated cells on the same plane. So a single cell can't reach those speeds are the sort of vigorous for example bulldozing. As an example, behavior that ends robots are demonstrating severe, one of the hallmarks is that very leveraging the multicellularity in our future treatments in human body.

Speaker 1 13:39
Right, it sounds like, you know, you'll be able to get a lot more functionality by combining whatever the cells collectively can do, or having different types of cells a lot more than what you can get out of a single cell. Are there limitations in terms of, you know, obviously, the small molecule drug that can perfuse throughout the entire system of the body? cellular therapies can reach some, you know, most tissues? Is there like an access issue with the antibiotic? How were we what would they be able to access in the body?

Michael Levin 14:19
Yeah, well, it's so so we don't know what the limits of it are. But certainly, some of our first attempts are going to be in areas where it's, they were, it's clear, they have room to operate. So this might include some kind of retinal repair application that might include some sort of spinal cord, and that we, you know, when we start talking about some of the the neural repair functionality that we show in this paper, those are the things that come to mind immediately, and not to mention, maybe maybe arteries, maybe, maybe the gut, there are there are lots of places where they should be able to operate, but actually we don't know what the limits are. There may be other other regions of the body that we haven't even thought.

Speaker 1 15:00
All right. So here's a good question. This is from from Lindsay Brownell at the Wysee Institute. So she asked if robots were created by taking adult cells out of their normal environment and culturing them in a specific medium. If they were to be introduced into a human patient to treat a given condition, how likely is it that biomolecules or other signals inside the body could influence or change the attributes capacities in unexpected ways?

Speaker 2 15:36
Yeah, that's an excellent question. And so I think our core culture, experiment, figures five and six, start to scratch the surface of this question, because what we do there, we take ants robots and inoculate into an environment that's completely optimized for neuronal growth. And we see that as robots are not only sort of not affected and not losing from their own properties, but also that they are kind of adapting into that environment and able to do useful work in that tissue. So to answer that question, more comprehensively would ultimately boiled down to testing the answer robots in the growth environments for the different tissues of interest. So far, we've only done this in neurons and haven't seen any significant downside. But, of course, that's something that we would first do a pilot study for before going ahead and introducing the bots. And the second thing there is that once an anthro, about develops and really matures, like when it's moving it's shape is its, it still has some plasticity, but it's pretty much locked in Miami sees that it's more susceptible to environmental perturbation in different regions, and all of that during its development in the matrix. And to a degree during its development after the matrix while it's becoming more tile during that seven Navy now. But once it's smooth tile and moving around. Again, the features are pretty much locked in. And it's we've seen so far, that it's it's pretty compatible with other tissue types. Right.

Michael Silver 17:27
Thank you. We have a question from Elizabeth. Can you see? She's asking what is the initial matrix that the tracheal cells are put into what does that made of

Gizem Gumuskaya 17:40
its extracellular matrix harvested from rat fibroblasts. It's a pretty generic region that is used in the laboratories across the world. And primary goal is to mimic the basal laminate in the human body.

Michael Silver 18:03
So the paper outlines some of the aspirational and applications of antibiotics, like spinal cord regeneration, clearing out arteries of atherosclerosis, it will be likely a long path to get those applications. What are some of the near term? What might some of the near term applications or trials might be? What might you try first given say their ability to encourage cell growth?

Michael Levin 18:31
Yeah, I mean, I think I think some of the short term applications are looking at just based on what we've seen, what we show in this paper, looking at the things like various kinds of wounds, so you know, is neural wounds, for example, maybe retinal defects, maybe various other kinds of lesions in epithelial and things like that. I think it's also important to think about the long term impact here, which is that in addition to whatever immediate work, these functional bots are going to be put to do it this whole platform has has also one other kind of major potential impact, which is the use of it as a almost like a sandbox in which to learn to understand the Morpho genetic code. That is how do we convince groups of cells to do one thing versus another that is critical for repair of birth defects, regeneration, after injury, normalization of cancer, prevention of degenerative disease, all of those aspects of regenerative medicine, hang on our ability to control what it is that cells build. And I think these kinds of synthetic synthetic constructs to help us do the experiments that really show us what what the plasticity of the cells look like, what can they do besides what they normally do in the body? And that helps us to develop novel interventions that really won't necessarily have anything to do with bio robots. But someday in the future, when we have regenerative medicine, where we know how to When do cells how to how to coax cells to grow various missing organs or replace aging structures in the body, and so on. It will be because we worked out those rules of stimulus and response in these kinds of these kinds of models. So I see, you know, sort of short short term application and neural repair and sensing bacteria and, you know, working with cancer cells and so on. And then, and then there's the longer term of using these kinds of synthetic systems to work out a way to communicate our regenerative medicine goals to collections of cells.

Michael Silver 20:34
It's very interesting, because, you know, we, science has been studying how, you know, developmental biology, how tissues and organs organize. And that's in the context of an entire system. And really, what you're talking about, is like, is taking this out and kind of breaking it down into maybe as simple elements, and then maybe inspecting how the simple, simpler rules of that morphogenetic code apply to create different shapes and such that helps you get a tour to get to that big picture. Right. So we understand that you originally were trained as an architect before coming to biology. And, and so I'm really cuz we're talking about, you know, cells and how they assemble and build and create more larger, more complicated structures to do interesting functions. I'm really curious to hear more about your insights into the parallels between building structures in which we live and work and building structures out of cells.

Gizem Gumuskaya 21:42
Yeah, parallels, parallels and contrasts. So yes, I did my undergraduate studies in architecture and started out in masters in architecture. And a lot of the things we were trying to accomplish in the field of architecture, like lowering carbon footprints of the construction processes, or trying to make buildings that are more responsive to their environment, sort of smart buildings, are trying to find ways to mass fabricate in a rapid pace of the constituent building blocks, or even start thinking about ways to sort of heal like the cracks on the concrete. So a lot of there's a lot of research going on in the field of architecture for this. And at the time, I came across synthetic biology. And that helped sort of shift my view of biology from again, this thing, sort of sitting outside for scientists to go and investigate in a magnifier and sort of map out to a design medium, something we can design with. And the really attractive thing about biology is that it already has a lot of the properties we're trying to build into our architectural sort of structures I'm in biology can so cells can self replicate right in an exponential manner. So the constituent Building Blocks can self replicate, it can heal in the face of damage, it can sense the environment and respond. And it can let alone in a carbon neutral way, in a carbon negative way, carry out these constructions that literally sucks carbon dioxide from the, from the environment. So I thought that it would be fantastic. Try to try to bring these qualities into architecture, by learning how to work with nature, to reprogram the types of self constructing structures towards a desired outcome, just as an architect and sit downs and designs a piece of building, it has an end goal in mind. So Mike talked about cracking them or for genetic code. So our understanding of how nature builds these structures ultimately gives us the ability to get under the hood and edit them and steer them towards designed ends. And that's Anthro bots is sort of one example of that as the first step. And from here, what would be really fantastic is to find ways to scale up this quality of self construction towards a new synthetic architecture and really bring it into the scale of centimeters. Maybe build self constructing books are self replicating bricks, and sort of being this paradigm shift into architecture informed by biology.

Michael Silver 24:41
Right. So again, I want to encourage folks to put in any questions they might have, as we're discussing if something comes to mind, something you maybe want clarification on on any terms or, or topics, please, please submit your questions. I just want to throw in another quick one. So we know, you demonstrated that the antibodies were able to encourage the growth of neurons do what do you? Do you have any notions about what they might be doing? Because I know you mentioned that it's not, it's not just because you're putting something on top of it's not just physical contact, you could put up like a piece of plastic or something on top, and it doesn't do that. The antibodies are doing something special to have courage. What do you what do you think is might be happening to do that?

Michael Levin 25:39
Yeah, so we don't know, the number one thing to stress is that we don't know yet. We know it's not a passive process, because we well, good because then we'll talk about the specific controls that we did for that. But we don't actually know there's a range of things that could be and I'll just since you asked, I'll throw out my favorite possibility. But I want to be clear that we do not know that this is the case, this is just a guest. For me, I have a feeling that what might be happening is that the presence of the bug is making it easier for one side to know that there's something on the other side for it to connect to, I have a feeling that it's serving as some kind of a single signal processor, which passive materials would wouldn't do. That allows the two sides to at least be aware of each other's presence. That's a hypothesis. It's easily testable. We aren't going to test it. But that's what I think I think there's a it's it's it's part of the information exchanged between the two ends. But but we don't yet know if that's the case. And we don't know if that's happening chemically. If it's happening electrically, this we will be figuring this out, and then again, in the coming year.

Gizem Gumuskaya 26:43
Right? Yeah, like one supporting experiment, just wanting to add to that point, that Mike was talking about one side, getting to know the other side through the presence of Anthro robot. So if we put that bridge into a neuronal scratch, and only have to touch one side, but not the other, the healing does not happen. So the bot needs, even if you have a perfectly sized bridge bot, you know, large enough and all of that, if it's not docked on both sides, healing does not happen. So it does look like there is some sort of connectivity requirement going on in there. And interestingly enough, if we put a piece of Agros, on top, even if it's touching the both sides, we again, don't get the healing. So so far, what we're seeing is that it has to be a even even if not answer, but at least up on the living tissue. So simple mechanical loading won't do it. And that it needs to be touching the both sides. We have tested some other cell types just made for urges that look very similar, using off the shelf, sounds like Hek cells and sort of generic mammalian cells. And they did not give us the same result, either. So this doesn't mean that it has to be an anthro bat. But so far, all we know is that it's not simple mechanical loading, and also not any biological cell can do it. But it's entirely possible that there might be other types of solar assemblies that are not necessarily as robots that couldn't help with the healing.

Michael Silver 28:40
That's, that's interesting. So there's like a, it's almost as if they're up in communication between the two sides of the artificial wound. It's possible Yeah, yeah. So have I mean, I think you partially answered this. If you've tried looking at whether bots heal other tissues, this course will people are interested. I mean, you can imagine skin cells for like lacerations or cartilage for arthritis or something like have you ever been looked at that? Are you planning to look at that in the in the lab anytime soon, how they might encourage growth of Condor sites for cartilage or other types of cells?

Michael Levin 29:16
Yeah, well, you know, all of these things are on the list. We have a very long list of things that needs to be done next. And yeah, checking checking their effects on other cell types other scenario damage scenarios, which might be mechanical damage, they might be chemical damage, it might be genetic damage, it might be the view all the all the, you know, different kinds of inflammation, degenerative disease, there's a million different kinds of things. All of that needs to be screened through. Yeah.

Michael Silver 29:45
Right. Okay. Great. Anyone else have any questions that they might want to throw out there? I think we've kind of gone through much of what we have here. Oh, yeah. If I could throw another one out there. I mentioned about understanding how antibody is helping us understand how cells might organize into tissues and and larger structures like Oregon's maybe you could tell tell us a little bit more about what what is? What are some of the things that are known about that, quote, Morpho genetic code. They know that software that cells use to communicate with each other and say, Okay, you go, there I go, here, you go there, let's make a kidney. I mean, what, what kinds of things are known, and that might be applied as you construct? antibodies?

Michael Levin 30:50
Yeah. Well, one of the one of the most basic aspects of it is that it is, it is highly reprogrammable, at least in many organisms. So there's, there's often a perception among people that the genome of an organism in some way fixes it toward a specific set of structures and functions. And so when you say, Okay, why do we have five fingers on each hand? Why do we have a certain body plan, you say, well, the gets the DNA, the DNA, determines it. But the interesting thing is, as we're finding out now is that the DNA specified, the DNA doesn't actually specify morphology directly, or anatomy directly, the state of the DNA specifies the tiny hardware, the proteins that every cell gets to have. And so what we are now starting to understand is the software that's actually able to be implemented by cells bearing that hardware. And the amazing thing is that evolution has produced hardware that can do more than one thing. And our lab and various other I mean developmental plasticity, something that people have studied for a long time, in other words, different outcomes in different environments. But we, in our lab, have really been focusing on this idea of where is the information stored for large scale shape in the in the form of pre patterns, and some other people study pre patterns that are chemical we study, in particular bio electrical pre patterns. And these are very analogous to memories that you might have in your brain about where you're going to physically go. And in similar ways we've shown and others have shown that there are pre patterns, there are electrical pre patterns in cell sheets that say, where are these groups of cells going to go in terms of anatomy, what are they going to make, and so a lot of that is reprogrammable. And this is why we've made tadpoles that we can induce the eyes, on the guts and on the tails of tadpoles, we can make flatworms with heads that belong to other species, you can make all these changes, because the hardware is reprogrammable. It's willing to do other things besides its standard default. And that that plasticity, that reprogram ability is really critical. And for that we have to go beyond kind of where all of modern molecular medicine is, which is aimed at the the genome editing and and, you know, pathway engineering and things like that. And really try to understand the collective intelligence, the decision making of groups of cells, how do they use that hardware to make decisions on their journey through all the possible anatomies and functions?

Michael Silver 33:20
Right. So let's talk about this sure that these are human cells that we're creating possibly as a therapeutic modality. They are organized together to kind of act on their own kind of like organisms. Can you talk to us about, you know, the safety features, any, you know, any ethical issues that come into play or not come into play? Can you tell me about that?

Michael Levin 33:54
Because then you're on a sir. Oh, yeah,

Gizem Gumuskaya 33:55
I can go. So first of all, these are not genetically modified. So let me talk about creating new structures from precursor cells. A lot of what has been done in that field, which is general the field of synthetic morphogenesis, has so far relied on insertion of genetic circuits, which in and of itself is a really exciting area, because then we get to test our hypotheses about how form develops in nature in very sort of organized way. For example, like a really good example of that is, Alan Turing had this hypothesis for how certain paths or patterns arise in nature called Turing patterns. He had a mathematical theory about this, but he's not a biologist, so he never really got to prove this. And a few years ago, scientists took Alan Turing's mathematical theorem and basically created a genetic circuit out of that. So genetic circuits are very similar to electrical circuits. Instead of transistors, you just have genes interacting with one another and creating complex Boolean logic. So they encapsulated Turing's theory into the strength of circuit and put it into living cells to bacterial cells. And bacterial cells did indeed create these patterns that are known as training patterns. So without the strength ik circuit approach, we would have no way of testing this. So that's a really versatile and really interesting way to approach. But that dictates the use of exogenous DNA, so that essentially Crescentic modified organisms. So one of the questions we asked is, could we bring the synthetic morphogenesis design of creating new structures with biological cells, without using any genes and into robots, again, is an example of that, we have only changed the environmental parameters and how we grow these I mean, that's five years, we've started from tracheal cells, because we want to make mo tiles spheroids. And, you know, tracheal, cells already know how to build cilia, you don't need an exogenous genes or circuits to teach them how to do it. And by just growing them in different conditions, we're able to steer the entire system towards our target architecture of interest. And that's how we accomplished this kind of synthetic morphology. So that lack of genetic circuitry is makes it is one of the big features that make it inherently safe. Because if we want to create an answer robot from a given patient, we would just take the cells from that patient and build the answer robots that way. And at the end, that answer about would have the exact same DNA as the patient. And when we inject it into the patient, it wouldn't be any different than the existing cells in that patient's body. So this is a big win for potentially preventing any type of inflammation in the body or immune response. So that's our sort of safety gear number one. And number two, is the fact that answer robots after some time on naturally degrade. So even if you just, you know, let them live forever, they don't. And the way they degrade is by becoming individual cells. So in the lab, and let's just by becoming individual cells, so the multicellular and robot will just degrade into individual cells, which can just be then sort of, depending on the tissue that it's inoculated can be expelled from the body through natural bass. So these are the two main safety features that are already built in.

Michael Levin 37:49
Yeah. In addition to that, you know, one of the main areas of concern for safety profile of various kinds of bioengineered constructs is the ability to replicate out in the wild. So when people make new types of bacteria and new types of virus, and novel yeasts, genetically modified organisms of mosquitoes and plants and so on, the question is always okay, how is this going to interact with the rest of the biosphere and in the ecology that it's already sort of prepped to live in So bacteria and various plants and so on, all of these things already have a niche in the in the ecosystem, and they can proliferate and expand and so on. So the thing about Anthro bots to remember is that they can only be created and maintained in a very specific environment. So this is not something that can live outside the lab, they do not make 1000s and 1000s of copies of each one the way that bacteria and viruses will, they cannot proliferate through the ecosystem, they cannot live outside the lab, or that same patient's body. So the idea is that here, we can have personalized medicine, with the patient's own cells without the risk of having this, you know, in some way, propagate through the through the outside world the way that various other technologies need to be thought about.

Michael Silver 39:17
Right, thanks. So I have another question here from Elizabeth Konishi. And this is a good maybe a good one to close on. It's what is your pie in the sky vision for antibodies say a decade from now. So we can think of well, you know, what's the science fiction view of how this could play out in a longer term future? I mean, I know this is all speculation. So we have to put that in that context. So we're not making any claims for what it can do now, but just you know, what do you might like to see it? Go?

Gizem Gumuskaya 39:54
Yeah, I think we do envision as a platform technology I mean, the neuronal healing was just sort of first application that we want to specifically show. But sort of think of it as your smartphone, your smartphone becomes a calculator or a camera, or a, essentially, whatever you want it based on the application that you're running on it. So we kind of envision as robots in a similar way, it's a platform technology, it's a machine, if you will, that you can put in different applications and then repurpose accordingly. So if you wanted to, as an example, as a patient, if you wanted to open up the arteries in the body, you could put an A, you could program it to kind of pull those that adipose tissue in your arteries. Or if you wanted to chase down bacteria in your gut, you can program it to do that. So ultimately, or even petrol, some of these sort of at risk tissues in the body for the onset markers of tumor formation or cancer formation. So the pie in the sky version of these is that they, they could become your sort of smart bot and can do what you program them to do.

Michael Levin 41:18
Yeah, I really agree with that I envision and I don't know if 10 years is if it'll be sooner, or it'll take longer, it's very hard to put the exact numbers on things because this is cutting cutting edge science. But I really envision a kind of symbiosis where we not only we Yes, we have, we have programmed the bots to do useful things in the body. But also I'd like to see us take advantage of sort of a bi directional relationship. In other words, I would like to be able to ask the bots through reading their various bio electrical and other other kinds of properties, I would like to be able to benefit from what they know about the tissue surrounding them. So we want them to be to be helping to do things that we already know about. But we also would like to use them as a, as a kind of much more web as a much smarter scanning technology. You know, when we scan bodies, we can we only measure one of a few different metabolites, or one of a few different markers, or maybe we look at, you know, do an x ray, or a CAT scan or something like that, these these bots, I think, and even as you said, pie in the sky. So I'm speculating, I think that the amount of receptors and the information processing machinery in these bots is so extensive that they probably know way more about what's going on with not only within themselves, but within the organism or the surrounding environment, then then our current abilities to, to data to to read that information. And so I would like to really have a two directional symbiosis where we find out what they know, we encouraged them to take various kinds of beneficial and pro regenerative actions. And it's very much kind of a two way symbiosis for them for the brief period of time that they're in the body. That's what I would like to see.

Michael Silver 43:09
Great. That was fascinating, because you actually demonstrated a simple way proof of principle with the Xena bots last time, when you had them, basically recording information of whether they were exposed to light and they would glow a different color. I mean, that's, of course, a very simple recording of what they're detecting, and say the collective could with all its receptors and molecular makeup can kind of take a record of where it's been. and report back. It's interesting. Yeah,

Michael Levin 43:40
I mean, it's something that I guess we hadn't mentioned yet, today is that this is also a new platform for the field of diverse intelligence. So there's this amazing, super interesting, multidisciplinary field called diverse intelligence with in which my lab in many other good labs operate. Where we're really trying to understand basal cognition, we're trying to understand what are the information processing capacities of really unconventional forms, and these could be various kinds of synthetic synthetic life forms and cells and tissues and so on. And so trying to understand what do Anthro robots what do they measure? What are their preferences? What are their behavioral repertoires? What can they sense? And these are all things we're doing with with Senate bots, as well, is to ask, you know, in a sense, what we're really looking at here is a new kind of, whether it's a new kind of proto organism, it's not a full fledged organism, but it's the beginnings of one. And it probably has competencies and behaviors that we do not know. Because we've never seen one before and because these have never existed before. So really, really just an understanding what they can measure what how they make decisions, both physiologically and sort of in terms of their multirole behavior. I think it's a whole it's a whole new frontier for the area of diverse intelligence.

Michael Silver 45:00
Great, fantastic. So thank you all. Thank you, everyone. That was awesome. Thank you, Mike, Gizem. Thank you, Tom for hosting this. With a good amount of questions. If anyone has questions come to mind later on, feel free to contact me. We'll be do our best to to provide you your answers by email offline. I just also want to have a reminder that the embargo for the story will lift tomorrow, Thursday, November 30, at 3pm Eastern time, we're looking forward to reading your stories. And of course, we're here to help with any other information you may need. I put in a link to images and videos in the Dropbox link. In addition, what you Mike showed you just lots and lots of really cool videos and images of the act of bots in work in play. Let's see. Alright, so yeah, that's that's all I have for today. But thank you so much, everyone. And, Tom, do you have anything else to add?

Thom Canalichio 46:06
I've just made sure to put your contact information into the chat once more for everyone. And, again, to remind everyone that this is under embargo, so nothing can be published about it until 3pm. Tomorrow, Thursday, November 30. Thank you very much for everyone's participation. Thank you, doctors Levin and Kumar. skya. And thank you, Mike silver there at Tufts. Very happy to have you all today. Thanks very much. Stay safe. And good luck.