Science Focus - the home of BBC Science Focus Magazine
Ritu Raman on building machines with biological material © L'Oréal USA

Ritu Raman on building machines with biological material

Published: 20th July, 2020 at 08:00
Subscribe to BBC Science Focus Magazine and get 6 issues for just £9.99

Read the full transcript of our Science Focus Podcast interview with Ritu Raman on biohybrid design - listen to the full episode at the bottom of the page.

Sara Rigby: So in a few words, could you please just describe what it is that you study?


Ritu Raman: Sure. So I'm a mechanical engineer. And typically you think of mechanical engineers as people who build machines or some kind of device using materials like metals or ceramics or plastics. And I basically add another set of materials into the tool box and think about whether we can start building machines using biological materials like living cells.

So one example of a biological machine that I built is a robot that uses living skeletal muscle to move and walk around. And I show that machines that incorporate biological materials can do pretty interesting things like, you know, when you fall down and cut your skin, you heal. When you exercise, you get stronger. And it turns out that these robots can do the same thing. When they exercise, they can get stronger and they can also heal from damage.

SR: So what do you use these biohybrid machines for?

RR: That is the ultimate question. So, you know, a lot of times when you're doing this kind of very exploratory research, which is what I did during my PhD at the University of Illinois, that the work on these robots, a lot of what we were trying to do is just discover what are the design rules and principles of building with biological materials. You know, building a car is a complex mechanism, but for the most part, most trained mechanical engineers can do all or some part of that.

They know that approximately this stiffness of a steel beam is going to take this much force, or a motor that I construct in this way is going to produce this much energy and consume this much energy. So you have all of these numbers kind of in place and getting those design rules and principles in place took us thousands of years.

So a lot of what I'm trying to do in my initial work is develop what those rules and principles are for building with biological materials. And my focus is to think about whether we can make robots, implantable devices that integrate living biological materials and integrate with our bodies so that they can do things like sense what's going on in an individual system, in response to, say, a disease and adapt to their environment the way the biological materials are trained to do.

SR: So what benefits do biological machines have over just metal and plastic? Why can't you just build a metal or plastic machine that could learn from its environment, like with artificial intelligence or something like that?

RR: Yeah. You know, I think there's a place for innovation using lots of different kinds of materials. So I think there are some places, for example, inside the body in particular, where there's a strong case for using biological materials because they're designed to optimally function in that environment. They are soft. They function really well in human environments that are 37 degrees C, they operate at the pH that the particular environment that they're placed in is.

And so there's a lot of different things where they're already designed to function in this environment. And they also have all of these adaptive capabilities that synthetic materials don't have. And certainly we've gone some way with synthetic materials.

There's some composite materials that, if you scratch them, might be able to seal that scratch. But it's never as strong as it was before. And if you repeatedly scratch the same spot, it gets worse and worse and worse, which doesn't happen to your skin. You might have fallen and hurt your knee when you were five. And if you did that again now, maybe you take a little bit longer, because we're all a little bit older, but it's still going to heal and it's going to return the skin to its original condition, which it wouldn't necessarily do in some of those kinds of materials.

And similarly with AI, we made a ton of progress in being able to develop machines that can process a lot of different signals and compute a pretty complex response and maybe even learn or adapt. But we're still very, very, very far from what the human brain, or brains in general, can do in terms of processing a lot of different types of signals and making complex decisions or functional behaviours.

So I think there's a place. My interest level is in implantable devices. So if you're looking at something, what can I put inside the body that's safe, that can adapt to its environment, and that I can engineer in some way to have a predictive functional response? I still think there's a strong case to be made that biological materials are your best answer.

SR: So why not just use all biological materials? Is there still a place for metal and plastic?

RR: Yes, absolutely. I used to, when I was younger, talk about this as building with biology. I was very obsessed with the idea of building with biology.

And to an extent, I still am. I was like, it's a Holy Grail. They can adapt to their environment. There's a reason our whole world is made out of biological materials. It's wonderful. It's beautiful. And to an extent it is. But since then I've switched to this idea of biohybrid design or building with biology as well as other things.

And the reason why I think this is important is that we've built a really fantastic array of synthetic materials to work with. We have materials that are so strong, yet so light. We have things that can resist insane amounts of corrosion. We've developed all these technologies and there's no reason to throw all of that away.

It's just a question of adding a set of materials, in this case biological materials, to every engineer's tool box and thinking: for the specific environment that you're designing and for the specific functional application that you're designing, does it make sense to add living cells from an animal or an insect or a plant that might be able to satisfy this and either do it better or do it cheaper, faster or more sustainably?

SR: So specifically, what sort of biological materials do you use in these machine? Is it living tissue, like you'd find in a human body?

RR: Kind of similar. So let's stick with the example of the walking robot, for example. So when we are trying to make skeletal muscle that acts as an actuator that makes this robot walk, we designed something that kind of looks like a rubber band.

So we have this sort of mould or a 3D printer and we pattern cells and a mixture of different proteins that mimic their natural environment into this ring-like shape. And what this looks like at that point is like if you've ever made any kind of gelatine-based dessert or jello, it's kind of right before it's about to gel. It's kind of this gooey mixture. You can pattern it into a bunch of different shapes and then it sets. So that's the same thing that happens with these tissues. They essentially set into that ring-like shape.

And then because the cells are alive, when they go through this process, they're sensing and responding to their environment. So the way skeletal muscle works in our bodies, for example, is there are individual cells that fused together to form these long fibres or muscle fibres. And that muscle fibre is what's generating force and producing motion. When we stack a bunch of those in parallel, like in our legs or arms, we can generate pretty large forces.

So we can do the same thing with these tissues. What happens is the cells proliferate. They reproduce. They get in touch with each other in this ring-like construct, and they fuse to form fibres within that construct. So they essentially remodel their environment and then create the tissue that can generate force and produce motion. And once you get to that point, you can essentially take that muscle ring – which is contractile, or can produce force in response to a stimulus – and you can put it around what we call a skeleton, which is really just like a flexible exoskeleton that can produce different types of motion.

So we focus on something that has two legs and a flexible beam between them and around those two legs we wrap this muscle and every time the muscle contracts, it's able to deform that skeleton and the robot sort of inches forward, kind of like an inchworm. It walks or crawls across the surface.

SR: You said, I think that essentially you have living cells and then you pipe them into a shape. And the living cells, they form the actual structure themselves. Wow. OK. So where do you actually get these cells from?

Because, I mean, I can understand in the human body, cells reproduce themselves. But, you know, how do you create them? Or where do you get from?

RR: Yeah. So I'll actually clarify one point about how they remodel their environment and then I'll talk about where they come from. So in a sense, yes, the cells are sensing and responding to their environment and creating the contracted fibres by themselves. But we do provide external cues to guide that transition, both mechanical and biological and chemical.

So the robots or the tissue essentially is sitting in a liquid bath. That liquid bath has a lot of different sugars and proteins and amino acids. And you can sort of alter that composition to tell it, you know, proliferate a lot. Like you reproduce a lot and make a bunch of individual cells or start fusing these individual cells into the fibre format. You're done proliferating.

And similarly, with the mechanical stimulus, you know, if you stretch the tissues, we found that if you if you don't stretch the tissues, these fibres will form, but they'll form kind of a willy nilly, like not really in any particular alignment or orders. So even if they are contracting, they're not contracting in the same direction. And so you don't really see in that response. So you can do things like if you stretch the tissue during this process of growth and differentiation, all of the fibres form in the same direction.

And so when they generate force individually, they act as a unit and you can see really observable contractions. So it's a little bit of self-guided and a little bit of external cues, all of which would be naturally provided if that were happening inside the body of a living being. But we have to create differently inside the lab.

As to the question of where these cells come from, most of the time when you're building with living cells, you can have a few different cell sources or types of cells that you can derive from. I would say the three sort of primary types of cells are primary cells, which are cells that you extract from a living animal. So they have the advantage of being very close to the natural tissue.

If you take skeletal muscle tissue from the skeletal muscle of a rat, for example, the tissue that you engineer using those cells is going to look very similar to skeletal muscle. But this is problematic for a variety of different reasons. It's not a very sustainable thing to do because it would require sacrificing an animal every time we want to do something. Or maybe you just take a biopsy and it's not that big of a deal.

But there's still some ethical considerations related to that. And there's also some reliability constraints. I mean, we all know that not all humans are the same. And probably if I took some muscle from myself and some muscle from you, it wouldn't act the same way.

And similarly, if you're trying to think about making a robot, you want a pretty reliable cell source. So you're taking something from different animals might result in very different tissues. So we chose not to do primary cells. What we did choose to use is something called an immortalised cell line.

So this is a very useful tool that a lot of biologists and engineers use in their research.

Essentially, there's some type of cell that's been harvested at some point from a biopsy, say, from an animal or maybe even a human being, maybe their tumour. And they just happened to grow very well on a petri dish. And we're able to keep them alive for many, many generations. We can freeze them down when we don't need them, thaw them and keep growing them so they can reproduce for multiple cycles in the lab. And so they are very useful tools for sort of building with biology, and they work really well.

And the last type of main cell that people used to build are stem cells. I think most of time when people think of stem cells, they probably think of something that's derived embryonically. And it really doesn't have to be. So there's a whole bunch of research around induced pluripotent stem cells. Stem cells are cells that, rather than taking something that's a muscle cell, you start with something that can turn into muscle or skin or bone, and you differentiate it along a path that leads to muscle. So there are very powerful cells.

And somebody discovered about 10 years ago a really robust protocol where you can just kind of scrape some skin cells off of your body and reprogramme them to return to a stem cell-like state and then redifferentiate them back into something that looks like muscle.

So if that sort of muscle differentiation technology had existed or been as robust or mature as it was as it is now when I started my PhD eight years ago, I would probably have used those cells instead of an immortalised cell line, because that way you can, you know, if you were thinking about one day putting some sort of engineered muscle inside a person's body, you don't want them to have an immune response to it.

So you want it to be as similar to the cells in their body as possible. And you can just scrape some skin cells from them and turn that into muscle and then create something that you put inside them that would be the safest and easiest way to do that. So those are kind of the broad classes support the cells can come from.

SR: And so is this is that a problem that you you encounter a lot with these sorts of using these biological materials? Is them interacting with the body? Like, I could imagine if if there was something that you swallow, I could imagine. Organic material would be more likely to interact with, you know, stomach acid or something like that.

RR: Yeah, no, it's a great point that you mentioned. Biological materials have a lot of positives, but let's say they're just kind of finicky. They have to operate at a certain pH and temperature and humidity and air composition, and there have to be certain sugars and proteins available to them and they have to be cultured around biocompatible materials. And if you put them inside of a body, they have to be compatible with the body in which they're placed. So you actually raise an excellent point about the stomach, for example.

I recently wrote a paper with a gastrointestinal surgeon. He does a lot of surgeries for people suffering with obesity. So if you're struggling with obesity, one potential way to deal with that is to place a bariatric balloon inside your stomach. And this is exactly what it sounds like. It's a balloon that takes up space in your stomach. It makes you feel full. You don't eat as much and it helps you lose weight.

So if you have something like this, this could potentially help you lose weight over time and then deal with that process. But right now, these balloons are placed endoscopically for the most part, and also removed endoscopically for the most part. And an endoscope is that giant tube that they have to sort of put you under and stick down your body. Not a very pleasant process and also not a very cheap process. And so the surgeon was asking me, you work with a lot of materials that can adapt to their environment.

So could you make a balloon that when we don't need it anymore or the patient has responded to therapy or doesn't want to do this therapy anymore, rather than doing an endoscopy, we could just remove it from the patient's body? And initially, you know, he heard about the work that I had done with these sort of biological robots before. Why don't we use something like that to do it? And I think eventually you could do something like that if you create an exoskeleton or something for these muscle actuators.

But since that technology didn't exist yet, well, at some point the skeletal muscle is just kind of like meat. So it's just gonna be digested up by the stomach probably before it can do anything useful. So instead, what I ended up doing is designing a new synthetic material that sort of takes inspiration from biology and from nature and systems that degrade or respond to light in some way.

So I made a new type of polymer that breaks down when you shine light on it. And we made the balloon out of that material. We sealed it with that material such that you could swallow the balloon. It would expand in your stomach. And then when you wanted it to break down, you could swallow an LED pill that shines light on the balloon, causes that to break down, essentially just pass it out the way you would pass food in general.

So, you know, that's, I think, a great example of this idea of biohybrid design. There are times when biological materials make a ton of sense, and there are times when smart synthetic materials make a ton of sense and you kind of just have to pick what works best for a particular location.

SR: Right. Wow. So what else might you use biohybrid design for? Is it all medical?

RR: No, I certainly don't think it's all medical. I think my personal interest is medical. And so I focus on a lot of those applications. And as a result, most of the time when I build with cells, I build with cells derived from mammals because we are mammals and they function in the environment that we've created.

However, there are a few other people in this biohybrid design or biohybrid robotic space, and there's folks that focus on a lot of different types of materials. So they might use cells derived from insects, for example, which actually function pretty well at room temperature and they function really well even in air, not necessarily always needing to be immersed in a liquid medium.

So if you were thinking of maybe having an untethered robot that could, say, move in our ambient environment either to work in a factory or perhaps be going in a water stream to sense a source of pollution and neutralise whatever toxin it was in that environment. All these sort of like outdoorsy robotics applications. That's the sort of thing where you might want to think about using cells from an insect or maybe cells from bacteria. And those are all certainly very valid and useful applications.

I think from my perspective, it's just such a huge field. And there's so few people in it. So in a way, we've sort of self-selected into little niches and bubbles. So I am, you know, the skeletal muscle person. I'm the mammal person. But there's other people that focus on different applications.

SR: How do you go about designing and testing one of these machines?

RR: So in some ways, it's very similar to how you design and build anything in an engineering context, with some added difficulties or quirks, as it were. So, for example, for the robots, you know...

A lot of times when you are designing or using an actuator in an engineering context, you're basically looking at what is the energy source, how much energy does it take to produce a certain amount of force? How reliably can you do it? And what's the efficiency of that process? So it's pretty much the same for skeletal muscle.

Skeletal muscle is very similar to a lot of, you know, a pneumatic actuator or a shape-memory alloy-based actuator. You have a certain volume. You know it produces linear actuation. Just contract, relax, contract, relax. And you have to turn that into rolling or gripping or anything else. So, you know, the approximate size constraints, you know how much force you want to generate. You know how much energy you want.

And you build a system the same way that you would build with a synthetic material and follow the sort of typical engineering design, build test, reiterate, do that over and over again until you optimise for a certain amount of function and price and efficiency and whatnot. So all of that is pretty much the same. What's different, I would say, is that biology has inherently more variability and it's a little bit more of a black box.

And both of those things might be problems that are problems right now because we're still have a long way to go in understanding these systems and maybe some of them are problems that will persist. So I would say, you know. When you're I'm a mechanical engineer, as I mentioned, and, you know, when I was in classes, we would go and we'd say, you know, here's a bunch of different steels that were manufactured at different factories, but using the same recipe. Go test them for their stiffness and see how it is.

Are you going to see some variability between samples? Yes, but that variability is a tiny percentage of the total thing that you're measuring, and that's what you consider as normal. So when I first started making these muscle powered robots, I was frankly just stunned by the amount of variability that I had and the amount that the biologists that I talked to just completely accepted that they were just like, well, of course, you know, cells are just different.

They're just responding to a million different things in their environment. And you might think that you created all these 10 robots the same way. But there are probably minute differences in the media composition or the exact chemical that you infuse and what time it was, or maybe this, you know, tissue experience or mechanical stress, because you stretched it a little bit more when you put it around the skeleton. All of those little things, those processing things which might not have made a huge difference in a synthetic material, make a big difference in biology.

And there's also a certain amount of stochasticity, or statistical randomness, that goes into the process. There might be, you know, four or five stable states that a cell can end up in. And maybe that randomness that leads it to one versus another is something that we just haven't fully quantified and can't predict with reliability.

So I think there are some things that I've done over time, for example, like stretching to help the cells align better. Those are things that I sort of discovered through this intensive design, build, test, iterative process where I discovered how to get them fairly reliably functioning all along the same way. I still haven't gotten them to the point where, you know, they function with the same reliability that something made out of steel would happen.

But that same thing that makes them sort of difficult to build with is the thing that's great about them, because then I know when I have a final system, it's going to be extremely responsive to its environment and its output is going to tell me a lot about what its inputs were. And if I understand all of those rules and processes, maybe I can build something that has a far higher functional complexity than I could with synthetic materials.

 SR: And just one last question. So what do you hope this research will lead in the future? And what is the most exciting thing that's going on, in the current research for you?

RR: Yeah. So in the future, again, my focus always returns to medical implants because I think one of the biggest trends that we see in medicine today is this idea of precision medicine or personalised medicine. The fact that we all know from lived experience that a few different cancer patients can go into a clinic and get essentially the same therapy and some people respond really well and some other people don't.

Again, it goes back to this idea of biological variability. So rather than giving everyone the same medicine, why not give them an implant that can maybe generate the sort of chemicals that they need in their body, but that senses that this person needs a little bit more at this time or that person doesn't. And this is sort of the field of cell therapy at present, really deals with this context a lot.

If we place certain cells that secrete insulin in a diabetic, for example, they automatically could potentially regulate how much insulin they're secreting to the needs of that individual person at that time. So I think that's the most exciting context for me. Can I make implants that adapt to the needs of individual patients? Because that's what we need. We need to be able to to meet people where they are and give them the care they need at the time they need it.

I think the most interesting thing that I'm doing right now is working towards using the skeletal muscle that I've developed in the lab as potentially an implant to replace skeletal muscle that's lost inside the body due to disease or damage. This is something that I just think is a really meaningful cause. It matters a lot to me. I think mobility is something that a lot of us take for granted. And when you lose it, it can have a significant impact on the quality of your life.

And so, you know, a lot of a lot of people have tried to do this over the years. But I think we've understood a lot by building these robots. How do you tune it to respond to its environment and how you generate really, really large forces from compact volumes?

And I think if I could, you know, eventually do this successfully in animal trials and maybe one day in clinical trials. I would love to be able to contribute any small part to helping restore the mobility to somebody who's lost it.

This podcast was supported by, helping people build quantitative skills in maths, science, and computer science with fun and challenging interactive explorations.

Listen to more episodes of the Science Focus Podcast:


Sponsored content