When will the World become like Avatar?
Discussing Synthetic Biology with Tom Ellis, Professor of Synthetic Genome Engineering at Imperial College London
Tom Ellis is leading a research team in synthetic genome engineering and synthetic biology in the Department of Bioengineering. He has track record in synthetic biology, being author of over 50 publications in synthetic biology including work in Cell, Nature Methods, Nature Biotechnology, PNAS and Nature Reviews. He is leader of the UK-funded project to build a synthetic yeast chromosome for the international synthetic yeast project (Sc2.0). He co-leads the teaching of Imperial’s synthetic biology undergraduate module and has won multiple awards for teaching and for supervision of iGEM teams. His research focuses on developing the foundational tools for accelerating, automating and scaling design-led synthetic genomics and synthetic biology, focusing on research projects in yeast (S. cerevisiae) as well as applied projects in other industrially-relevant and medically-relevant microbes.
For more information about Tom Ellis and his lab, please visit the Tom Ellis Lab Webpage
This interview has been edited for length and clarity.
WP: So, biosensing — using biology to sense the world around — how well can biological sensing systems plug into the existing electro-mechanical signals that we have?
TE: That’s a very good question. Biosensing is a really promising area for the next several decades because if you think about the challenges of climate crisis, healthcare, emerging infectious diseases like coronaviruses, going to space and so on, being able to sense what is going on around in a living system — but also using biology to sense what's not living — frankly it's a superpower.
If we can ask a plant what's going on with the soil beneath you, why are you not happy, is the acid wrong in the soil? Or we can ask microbes in the upper atmosphere are you sensing too much ozone or not enough ozone. You could ask cells inside bats in a cave somewhere, is there some coronavirus you're dealing with that's going to be a problem for us in the future.
So if we can harness biology’s innate ability to do all of this sensing — it’s evolved to sense everything because it has to, both biological and non-biological and it does it all at the micro and nano scale — if we can capture that and communicate that back to us, ideally in a timely and non-invasive way, then that’s a super power. That would give us surveillance over everything going on in the world, both good and bad, in terms of how it’s going to affect us. That’s a really promising tech.
In our past work, particularly the AstraZeneca, Cambridge University work that Will Shaw did most of in the group, that was to take a specific type of sensory system, G-protein coupled receptor sensors, and show that we can come up with a way to control when they sense something, how that sends a signal that is easy to detect, which in this case is just yeast cells turning green with fluorescence. And that GPCRs are the majority way human cells sense stuff around them and mammals and other animals do as well.
We have hundreds of different GPCRs inside our nose and in our taste buds, dogs have many thousand more than us, mice and rates as well, they are much more about sensing things by smelling. But GPCRs also sense light, entire cells and organisms, they can sense anything, so they’re like this beautiful little module that has evolved to sense lots of different things.
But that’s just one of the millions of ways that biology senses. Now you’ve posed the question what about lining that up with everything else — wouldn’t everyone love the kind of that sensing that goes on in the plant in your garden to get to the internet, so we can monitor it?
That fits in with dream of internet of things, that there is stuff all around that is wired to the rest of the internet that you can look at. Internet of things is not a phrase used that much anymore, I would love to think that one day engineered biology and biosensing could enable us to to have an internet of very little things or internet of living things and we can plug into that. How that can be achieved is very difficult. There is very little work that’s producing useful results at the interface between stuff that goes on inside a cell and then electrical signals that could go into a computer
Ways I’ve thought you could do this — cells are good at releasing molecules, so you can imagine a cell that releases a small molecule that’s very specific and very rare but diffuses very easily, it’s very gaseous or something and them nearby you have things that then pick that up and convert that into electrical signal as a sensor. To some extent, at the moment this would be a stop-gap measure.
We have technology that can do very good sensing of certain molecules, like pregnancy tests, insulin sensors, diabetics these have electronic sensors, cheap little test strips, Bluetooth so it goes straight to your phone to record everything and they sense glucose, blood glucose. So you could modify that so it senses something else or you can engineer a cell so it senses something very specific like presence of cocaine in wastewater and if the cell senses that, it produces glucose and then I can sense that with my glucose sensor and I can get that off my mobile phone and I have this sort of half-way house approach.
Synthetic biology teams have done that — get glucose sensor, get a cell or some reaction that can sense SARS-Cov-2, the virus from Covid, and it will produce glucose and you can piggy back off that into a handheld glucose sensor and you can go around sensing for SARS-Cov-2.
WP: I can see how that would potentially evolve. What does it look like coming the other way, where the computing systems we that use would themselves become more biological, technologies like DNA computing?
TE: I don’t see DNA computing doing very well against current computing at the moment. I don’t want to do it a disservice because some of the initial stuff on DNA computing in the 90s was one of the first things that really ever grabbed me in science and made me think about what I might want do if I grew up.
In the 1990s they were saying that theoretically if you had enough DNA molecules in a bucket and you let them all compute with one another you could do things like solve the travelling salesman problem very quickly and I thought that was fascinating and that got me into DNA and that got me where I am. It hasn’t quite evolved much quicker than that — I think we’ve done square root of two — a few other things.
Basically it requires a lot of scale.
Where it works well is distribution — a single cell is never going to compete on the kind of computation we like to use now against a processor in your phone which has so many wires that can be physically separated so they don’t cross-react with each other. Our brain is obviously way better than what goes on inside a computer, but that’s actually really big thing, the mass of computational circuitry in our brain is bigger than a computer already and it’s not going to get any smaller.
So circuitry and computing done by cells or biomolecules has a use where DNA can do a few small calculations in distributed way. There are computational problems where that is a very efficient way of dong things. For example if you wanted to scan a large image and ask where are the contrast boundaries between light and dark parts of the image, for computer with a central processor that’s quite a lot of work — it has to go to every pixel make a recording of how strong it is, build a map and find out where the boundaries are. If you have distributed computation, 10 million small dumb processors, they can each pick a pixel or a subset, do the analysis and straight away you would get the answer. Bacteria could do it, just a few logic gates or something, by being able to have ten million of them in a tube. That is powerful for certain types of problems but not generally for the ones we’ve built computation to handle.
WP: Synthetic genomes — these custom and dynamic genomes. I think this is where it starts to get into the territory that starts to scare people. Are you going to be creating synthetic humans with modified genomes on the basis of that technology?
TE: It’s hard to say where things will go, we’ve got the all of the future of humanity to find out and I doubt I’ll be there to see it. We’ve got synthetic yeast close to being finished and it looks and acts and behaves just like yeast but its underlying DNA is different. We can customise it as well. We can have synthetic yeast and we can remove whole sections its abilities from it and add while other ones. At at point it doesn’t necessarily act and behave like yeast, it’s become something else.
Now imagine being able to do that with something like a tomato plant, you could strip out loads of stuff and have a tomato plant that also grows some other fruit in another area of the plant. Sounds a bit wacky, but actually if you think about it, a lot of these plants, through traditional crop breeding methods have been pushed in different directions and their underlying genomes have been changed from the traditional breeding methods.
I think as soon as we think about the unique and special case of humans it becomes very difficult. Dogs are not far from humans and look at what we’ve done to dogs, we’ve seriously messed around with them! It’s definitely possible and almost certainly going to happen in the lifespan of the planet earth that there is a synthetic human genome and potentially even synthetic people or organisms walking around with synthetic human genomes, but not in any research I’m going to do.
We’re going to focus on, for example, making a synthetic human chromosome. We’re only going to be doing that work as cells in tissue culture, a single cell like a cancer cell line that is typically used for research, modified DNA of that so it now becomes synthetic, and doing experiments to test if it still works, how can we learn what is naturally going on with the human genome by replacing it with sequences we’ve written. That tell us something about how much we know about the genome.
If you think about it, the genome being the computer code, the operating system code or even the book, whichever analogy you choose, fundamentally it’s a load of information encoded in a kind of language and all of these billions of dollars of fundamental research that’s done on humans and genetics and human cells, a huge amount of that is to work out — how do we read this language in this book?
The ultimate test is — write something in that language. If you spent all this money deciphering hieroglyphics using a Rosetta Stone like the Victorians did, prove that you know hieroglyphics by writing some and travelling back to ancient Egypt and seeing if people can read it. That’s what synthetic biology offers — the ability to write. You think you know everything about the way DNA is encoded? Try writing some and we’ll see how well it works.
WP: What do you think the potential for a new kind of language is. Looking at artificial cells, would there be the possibility of an improved, streamlined coding language for biophysical systems?
TE: For sure we can streamline compared to what nature has done, because nature’s evolution of genomes, particularly the large genomes like ours and those of higher plants and other organisms are just full of junk. People argue about how ‘junk’ it is but I’m not scared of saying junk. The evolutionary process involves lots of duplications, these things called transposons that jump around and add DNA to new places. It’s evolution so obviously it’s very hit and miss, stuff just happens and if it works, it works. So you could easily think about much more streamlined genomes if you were building from first principles and you knew the language.
Significantly different languages would involve adding extra bases to DNA above the four. So work in the field is showing that that’s possible. We’ve had six different base pairings, another group going up to eight, extra animo acids into proteins, which is totally possible.
WP: That seems to me that you’re starting to expand the definition of life?
TE: Yes and no. It depends what the definition of life is.
WP: That is very true. Does expanding the base pairs in that way, does it really give us more functional capabilities?
TE: It can do, but not necessarily yet. Expanding the extra bases of DNA allows us to put more information into a smaller amount of molecule. But that doesn’t seem to be that big a problem, because our own genome is 90% junk and it’s not getting used. So we’re carrying around ten times more DNA than we need to and we’re fine with that, it doesn’t cost much.
There’s no great advantage right now for compressing more information into DNA. Proteins on the other hand, adding extra animo acid capabilities — so at the moment we use twenty for all of our proteins — each time you add one more, you’re going to get a multiplication of how much more different forms of chemistry you can do. So you’re really giving evolution a whole new path to go down to make proteins that can do different kinds of chemistries.
So you can imagine, for example, enzymes are pretty good at doing all of this biology around us and instead we use some pretty nasty chemical processes to make other things like polymerising plastics or breaking down plastics, turning fossil fuels into things, you could think that adding extra animo acids into enzymes could unlock chemistry that we’re currently doing through very nasty methods. You would have this extra diversity of the kinds of reactions you do by having non-standard amino acids as part of the proteins that get made.
Engineered Living Materials (ELMs)
WP: Speaking of materials, it seems like engineered living materials, bacterial cellulose and such, seem to be the commercial bridgehead of synthetic biology at the moment, in terms of what consumers could experience. I know it’s used for synthesising certain chemicals but it seems like these new kinds of materials could be the first wave of what people get to see, touch and experience.
TE: That’s a fairly good point, people probably don’t appreciate how many ingredients in what they put on as cosmetics or in food supplements and medicines are made from engineered cells and engineered microbes.
Like vitamins for example, you think of vitamins ‘they’re good for you, they must be from plants and stuff’ but they’re actually mostly made from microbes in vats and those microbes have been heavily genome engineered to be able to produce loads of these vitamins. So that’s been going on for two to three decades and synthetic biology has helped push that a lot in the last fifteen years.
But it’s a very one dimensional product of synthetic biology — it’s basically saying you’ve got an entire cell here, take this specific molecule and make a lot of it. Do we need to do something with logic gates, or oscillations or pattern formation? No! Just get the cell to make as much of that thing as possible. So it’s very dumb in terms of what synthetic biology can achieve. All of these fancy research papers in synthetic biology show cells sensing how much light there is and drawing a pattern in response to it. But that’s completely pointless if you just want to get a vat of cells to make a load of the vitamin.
Materials on the other hand start to add in these new dimensions. At the molecular level, a material is just getting a cell to make a lot of something, but then that something has to be polymerised, and then put into a network in some sort of shape and then maybe modified, and it has to grow over time. So you’re starting to bring in lots of dimensions other than ‘I have a molecule, give me as much as possible.’ Now it’s like ‘I have a molecule, give me as much as possible, arrange it in this way, grow that around in this way, over time’ and then you get to materials.
So from an engineering and research point of view, it’s an exciting frontier to take on. In terms of something that can then be a product that’s sold to people — a material is something that gets big enough that you see and feel it whereas a purified molecule like a vitamin just ends up as an ingredient within something.
That’s good, I’m trying to think of other products. I mean, meat replacements, stuff that’s replacing meat and animal products. That’s another area where people are seeing increasingly there are consumer products that people want to get that come from synthetic biology work.
WP: Yes, and I think fashion and architecture.
TE: It will be a while before stuff is done at a scale that’s appropriate for architecture.
WP: Scale seems to be the challenge that comes up again and again in this area.
TE: I was kind of joking over lunch with my PhD students, thinking what organism would be cool to work with in space to be able to engineer to make things. Why is no one doing the synthetic biology of bamboo? It grows so fast, you can eat it, build things out of it. We want people to make products out synthetic biology bamboo, that would be very cool.
WP: That makes a lot of sense — Sufficiently Advanced Bamboo technology isn’t it.
TE: I think so right, we talk about growing wood in a dish and then that could start to be used in architecture. But wood takes a long time to make — whether you make it in a dish or grow a tree it takes a long time. Bamboo is quick, you can get enough to make a house in a year.
Scaling and Industrialisation
WP: One of the themes throughout the different sections in your review paper (Ten future challenges for synthetic biology) was about having standardised systems, modularity and to be able to actually scale up and design things in a much more industrialised way. Are we still quite far away from that?
TE: There’s an industrialisation going on in synthetic biology and Ginkgo Bioworks, for example, are a great test case for that. That is the industrialisation of the design and improvement process. You tell me your problem, you tell me the sort of DNA you’re wanting to encode into a cell system to then fix that problem. And we, because we have very highly skilled people, very good processes and lots of information about the best way to put DNA together and lots of robots to help us do that. We’ll do that for you and we can then give you a greatly improved yeast strain or bacteria strain or plant cell or mammalian cell that can then do the task. That is happening now in the field, that sort of professionalisation of designing and building cells is happening.
But there is still then the gap that if you want to make a product out of something that someone like Ginkgo has given you or that research students have slaved very hard on or that a biofoundry has made for you, you have to learn how to scale that up and make loads of it.
The dream would be that as well as there being Ginkgo who do the design, there is some company who are the kings of knowing how to scale up lots of different bacteria and things like that that have been engineered. There’s lots of problems with this. You can do all the recoding of DNA and assemble a cell which is the best cell at making stuff or doing something in the lab at a scale of say, 100ml and it’s ten times better than the one you had at the beginning that only produced 10ml. You then take both strains to a fermentation facility and you say, ‘now make me a hundred million litres of stuff’ and the one that was specially engineered to be ten times better just fails miserably and the other one somehow can do it.
There’s a bit of a lack of knowledge — not a bit — there’s a gaping lack of knowledge about the difference between doing something at lab scale and what then happens when you try and scale that up to a very large scale. A lot of people are working on that problem, but it’s an unsolved problem. It would be great for the field if there were, several companies that were very good at that problem. Then if you’d got your strain improved by Ginkgo you could then go to one of those companies to scale it up.
WP: Once progress has been made in that direction, it seems intuitive to me that it would be cheaper than current material supply chains. Am I underestimating the cost?
TE: Current material supply chains — depending on the material — can be exceptionally cheap. Let me give you an example. So we make products with bacterial cellulose. In terms of engineered microbes making a material that then grows and it’s actually something you can touch and feel, bacterial cellulose is orders of magnitude better than most. You produce within a few days enough to play around with and hold or put on your head or whatever.
Whereas if you work on something like spider silk, no way, you’re going to have to brew up for ages and then you end up with a little bit of spider silk. The difference is that ours is just made of sugar stuck together whereas spider silk is made of proteins and the cells have to do lots of fancy things to make a protein. So we’re making one of the easiest and cheapest to scale up engineered living materials.
So how cheap can we get it? We know that bacterial cellulose in its hydrated form is sold as a commodity product in the Far East, because people add it to make deserts. There’s places you can go online and place an order for a metric tonne of freshly made bacterial cellulose for a hundred U.S. dollars, shipping costs aside. Now you’re like, ‘Oh my god, I’ve got a thousand kilos’ and let’s just imagine it works just as well with our engineered bacteria doing something like becoming fluorescent when it touches your skin at night, or something like that, something fun. As long as the engineering is done right it wouldn’t cost much more you would still get the same sorts of yield for the same amount of input so it might only cost two hundred dollars for a thousand kilos. Great. But the things is when you buy that, ninety-nine of it is water because it’s a hydrated material.
Alright, you dehydrate it because you want to use it as a leather. We’re not at a thousand kilos now, we’re down to ten kilos of the stuff. So we’ve got ten kilos and we’ve paid a hundred dollars for it and that’s still cheap. But how cheap is that right? Because for a hundred US dollars I can buy forty kilos of printer paper, which is cellulose material that’s been beautifully finished and dyed and wrapped up and cut. It just makes you realise that existing materials like paper and polyester and nylon are insanely cheap.
One of the main reasons for that is we’re not really paying for the environmental cost of that. Those are all industries where they can go into forests somewhere and just smash all the trees down. They don’t pay for the environmental cost of doing that, they’re just companies that already own the forest. Nylon comes from fossil fuels and it pollutes like mad and they don’t get costed for that. So unless something dramatic happens from government to makes unsustainable practises like those have to pay a lot more, then it’s going to tough to make sustainably biologically grown materials be competitive.
The Cultural Promise of Synthetic Biology
WP: This is where we get to the cultural promise of synthetic biology. Part of the appeal of it is that it promises a closer relationship with nature than we have in the moment. Tied into all it are these ideas about sustainability and not just the technological paradigm we will have, but the political and social and so on. How optimistic are you about, let’s say, how synthetic biology as a technology acts as a spearhead that says to people that we can develop technologically without having to carry on with the same negative externalities.
TE: I would say it’s a thought leader in this area, but it’s not going to be the one that solves this issue. People in synthetic biology — particularly younger people getting into it — are much more passionate about the idea that by engineering cells and living systems, we can then build with biology and learn to live with biology in a way that’s better for us and better for the environment.
But that’s thought leading. I don’t think synthetic biology — or if it gets replaced by another term like ‘engineering biology’ — I don’t think it will ever be in a position to dictate to the world and bring the rest of the world along.
It’s a scientific and engineering idealism that we’re going to hack nature and recode things. People don’t want nature to change. People don’t like the idea of nature changing.
The whole idea of conversation and saving the planet is very much that we like how things are now in the wild we like how things used to be fifty years ago. It doesn’t fit that narrative to say great, we can save the rainforest, first thing we’re going to do is rewrite the all the genomes of everything in there, make new organisms that can help out in all of these different ways. That’s too different from what people’s mindset is to be the link that’s going to happen.
We can have these conversations and you can try to bring more people on board into realising that it’s part of future solutions of being able to work more with biology. If people who are not doing synthetic biology that are instead looking at what they buy and switching to things that are more plant-based in terms of food or use more sustainable versions of agriculture around them then great. I just wish that the whole organic thing would go away.
Organic makes no sense. Some of the practises are fine but others cause worse problems.
So organic cotton requires four times as much to make the same amount of product, that land is needed because of water. The amount of water needed for organic cotton, and all the stresses of taking and farming on that land, is four times as much as normal cotton and probably ten times as much as GMO cotton.
Organic — it’s like someone has come up with some black and white rules and refuses to budge when there’s so much evidence that some of those rules in certain circumstances are completely counter-productive.
WP: That can be the dark side of that kind of thought leadership.
TE: Things come and go in fads and hopefully organic will go out of fashion soon.
Regenerative farming sounds very promising, for example. I would love a movement where we creates GMO organic product, something that’s an oxymoron that would make people think about what’s going on.
WP: The context of this discussion was thinking about the technology itself and where it could go. One of the things you talk about was communication with the public, how you create a dialogue between the promises and the anxieties associated with the technology. By looking at this in a fictional way in the novel I’m hoping to address some of the anxieties we have at the moment. You could call it thought leadership, but it’s that cultural part of the equation.
TE: It’s a shame that there aren’t enough future visions of the world that incorporate the vision of building with biology instead of building with metal.
WP: That’s actually something which I very much want to address specifically. Imagining that future will help us build it, if people have those visions in their heads about how differently things could look.
TE: How many movies have either a bleak post-apocalyptic futures or shiny skyscrapers all made of metal. And then anything to do with new organisms, diversity of organisms or genetic modification in a movie set in the future, stuff’s gone wild and everything’s all over the place and it’s a big mess and it’s terrible. And that really dramatically influences public perception, what we see in movies.
There’s literally corners of the internet where people dream of life on the planet of Avatar, being able to talk to all of the organisms around the earth in some internet of living things, which goes right back to the beginning of our conversation.
WP: That is very true and visions like that do contribute. I get more and more of a sense of people bullish on fields like synthetic biology. I spoke to Alex Ayad recently, from Outsmart Insight and his predictions of the future are big on the synthetic biology aspect. I do think a big part of that is making the world like Avatar.
Engineered Living Materials (ELMs)
Scaling Up and Industrialisation
The Cultural Promise of Synthetic Biology