American synthetic biologist Sheref Mansy is working on constructing artificial cells that mimic — and “talk” to — biological cells. In this fascinating conversation, Mansy weaves through the question of what does and does not constitute “life,” the possible practical applications for his work, and how conversations with artists have opened up concepts that feed into his research.
How would you describe what you do?
My laboratory — the Mansy Lab at the University of Trento — builds artificial cells, or cellular mimics. What distinguishes us from other people in this field is that, typically, people begin with something that is already alive and then they try to change the behavior of that already-living thing by changing its genes. What my lab does is a bit different. We start with things that are not alive. So, protein by itself is not alive, DNA by itself is not alive — but somehow, when you put these things together, under the right conditions, you get life. Nobody knows how that is, and so that’s what we’re trying to figure out. I guess you’d say we are exploring the boundary between living and non-living. What does it mean for something to be alive?
So if DNA in itself is not alive, what is it?
It’s just a molecule, which scientists can build in the lab. That was one of the nice things that the Venter Institute showed: you can build an entire synthetic genome, and it’ll function like a natural one. But what his work didn’t really show was what exactly is needed to get that DNA to sort of kickstart life. It’s nice work, I have no criticisms of it. But it leaves a lot unknown.
For example, the genome that was synthesized by the Venter group was relatively small. But still, approximately one-third of the genes of this genome provided unknown function. So it’s a very black-box approach. We know this stuff is needed, but we don’t know why. With the approach I’m taking, in which we build from the bottom, piece by piece, I hope that we will, in the end, get to a sort of cellular lifelike system in which we know why every single component is needed.
Can’t you go backwards and take out one gene at a time?
Many people do that as well. It’s a legitimate approach, but there are a couple of limitations. First of all, evolution is extremely complex. It’s not a linear process that can always be easily traced. You can go back a bit, but it’s hard to really get back to the beginning.
I think building an artificial cell with biological parts and studying the origins of life are two separate fields. I try not to talk about the two at the same time, because I think it’s confusing. A lot of times people assume that by going backwards — going back in time by knocking out genes — you’re making a simpler organism. But in fact, you’re perhaps only going a bit back in time. The simplest organism you could theoretically reach with this approach is still extremely complex, too complex to reach in a single step from a collection of molecules on prebiotic Earth. There must have been much simpler, living or lifelike things in between that we don’t have a trace of, and we haven’t figured out yet how to build in the lab.
What is the difference between life and not life?
This is actually something that there’s no clear answer to. “Life,” in many ways, is not a scientific term, or at least there is no scientific definition of it. This is one of the obstacles or challenges that we face in this field. I often contrast what we are doing with traditional engineering, because people like to say that our field is a field that engineers life. If you ask engineers to make a plane, or to build a car, they know exactly what to build. To build life — it’s an ambiguous term. To give one classic answer, many people say life is something that can self-replicate. But there are sterile animals, and none of us would say they’re not alive. There are salt crystals. Even table salt, under the right conditions, can form multiple salt crystals. So it’s a form of replication, but nobody would say that table salt is alive.
What about plants that can’t pollinate themselves?
You can find lots of examples like that. This is what happens with every definition: Every time somebody says “life is” the following, you can find living things that don’t fit, and non-living things that do fit.
So DNA, by itself — I mean, you could say DNA plus maybe some proteins can replicate, but, you know, is it self-sustaining? There are so many different characteristics of life we could go through. To be honest, I find it both interesting as well as frustrating, because when you go to conferences in this field, sometimes you get in a situation where people are just fighting: “Well, but that’s not really life! Why are you building that?” The thing I find frustrating about that attitude is that there’s not enough good science in this field, so one wants to sort of systematically try to make progress. That’s hard to do if all you do is fight over definitions.
What is the field called?
A very good question! I think, broadly, we do fit into synthetic biology, but typical synthetic biologists do what I said before, which is start with something that’s alive, and they tend to really focus on technological applications, such as how can you engineer E. coli or yeast to make diesel for your car. Practical things like that.
People in my field do try to do some practical things, but we’re often more concerned about fundamental science questions. We are broadly within the field, but we’re not in the mainstream part of it. Some people like to use other terms, like artificial life, for example. If prebiotically plausible molecules are used, then it falls under origins of life, of course.
Are you thinking about applications for what you do?
I do think that there are applications. So right now in my lab, we are building artificial cells in a way inspired by the artificial intelligence community. I guess I like to refer to it as trying to build a cellular Turing test. All living things communicate chemically, so we’re trying to build an artificial cell that can speak the same chemical language as a natural cell, and then we want to ask whether natural cells understand that ours are artificial, or do they think that they are talking to other natural friends, a natural neighbor? So can we trick E. coli, for example, into thinking it’s talking to another E. coli.? If we keep getting better and better at this, then perhaps they will become indistinguishable, not only to a bacterium, but to us.
But while this is a fundamental science question, I think it could have technological applications. Cancer cells secrete different molecules from non-cancer cells. So, in the future, we should be able to not only identify target cells, but perhaps we even could get our artificial cells to synthesize drugs in response to it. Perhaps it could even use local resources that are already provided in the body, so you wouldn’t have to actually, you know, take medicine, but have a little artificial cell that just manufactures it when needed, in the proper place, so you’re not flooded with toxins. Those are the kinds of things that I could see as applications.
To read the full interview, visit the TED Blog >>>