The Human Operating System Gets an Overhaul
The boldest science project currently in the works? It’s not Elon Musk’s race to Mars or the the next iteration of the Large Hadron Collider. Instead it’s something many people haven’t heard of.
Think of it this way — right now, around the globe, synthetic biologists are building novel organisms from scratch for an array of purposes in medicine, energy, agriculture, and other fields. The project I’m talking about, Human Genome Project-write (or GP-write, as the project is known), aims to use these same tools to build a much more familiar organism: a human cell, complete with all the DNA required to produce more human cells. Mastery of this technique could wipe out diseases and bring about other applications that we can’t yet imagine. It’s the ultimate engineering blueprint for life.
What follows is a lengthy conversation with synthetic biologist Andrew Hessel — and it’s lengthy for a reason. Hessel is the person who kickstarted the GP-write project, turning a controversial idea—in 2016, fellow scientists questioned GP-write’s aims and faulted what they saw as excessive secrecy—into a global movement with almost 1,000 people involved. This is the first time he’s told the story behind this moonshot to re-engineer human life.
This interview has been edited for clarity.
Before we dive into the nitty-gritty of the GP-write project, what’s the big practical benefit? Why should people care?
This is a project that will touch every human life — literally. If we’re successful, what it will really do is unlock the power of biology. It gives us the ability to heal disease, repair ecosystems and — because it gives us the ability to design and grow resources from scratch — the ability to sustain humanity in an environmentally friendly way.
Where did GP-write come from, and why are you one of the people to lead it?
I’m a little difficult to define, but the role that I seem to play in a lot of projects is a catalyst. I help to bring people together and explore new ideas and see if we can’t make something happen. I love to champion certain technologies in their early phase.
Training-wise, I’m a microbiologist and cell biologist, although it’s been years since I’ve really worked at the bench. I was involved in genome-mapping projects in bacterial genomes earlier in my career, and then I moved over to bio-pharma with a major pharmaceutical company [Amgen]. Spent seven years with them learning the ins and outs of drug development and the research behind that.
From there, and for almost 15 years now, I’ve defined myself as a synthetic biologist. Someone not taking apart genomes or working with organisms to understand their metabolism but someone who is starting to design and build organisms.
After I left pharma, I started to see the parallels between the computing world and the biotech world — the synthetic biology world where you’re coding, really writing code, to make your new applications. And I didn’t want to write proprietary code. I wanted to be able to use open-source code and have the entire community be able to look in under the hoods of the projects, to make collaborative efforts easier and really just respect the foundation of academia, which should be transparency and openness.
Most recently, I worked with Autodesk, essentially leading them in bio-strategy and in the prototyping of a novel cancer-fighting virus. I recently left, because our cancer virus work was successful, so I spun it off into a different company, Humane Genomics, which is what I’m currently doing besides helping to lead GP-write.
So, let’s use that as a transition into the origin story for GP-write, because it was an open-source project until it became a secret project. I’m just kidding.
Well, for me, the origin of GP-write actually goes way back to the closing of the genome project. The Human Genome Project was founded in the 1980s, launched in the 1990s, and the first draft was completed in 2000. Remember this is the race between the academic group and corporate groups to sequence the human genome, and it was exciting.
But when the first draft was done, the entire genome bubble, which had been inflated by $3 billion of cash and global activity, deflated. And suddenly, it was just a mop-up job of finishing the genome and closing up sections and winding down genome centers.
We now speak God’s language; now go do the dishes.
Yeah. Pretty much. Internally, at [my] bio-pharma company, we had just had this giant glut of information that was going to take years to sift through and validate. So, for me, it was a really dull period between 2000 and 2003. It was the perfect time to transition out and do something new.
I started focusing on, look, we’ve done reading, we’re moving into heavy-duty analysis, the next thing is writing. This is just basic language: Reading, writing, comprehension. And so I started focusing on writing DNA, trying to drum up activity internally in my bio-pharma company, and no one was interested. No one wanted to think about writing [a] genome. I kept thinking the academic community, or some groups, are going to come to the same realization that now we should be focusing as a scientific and engineering community on new technologies for writing DNA.
When synthetic biology got a name around 2003, I thought, “Oh wow! This is going to be the community that writes DNA. It’s going to be absolutely amazing. The entire scientific world will start to jump on it.” And instead it really got off to a slow start.
A lot of scientists weren’t even sure if there was any value in this, or that you could design a genome or even a section of a genome. The synthesis technologies were too expensive, and there was all, you know, “what’s the value of writing DNA versus cloning?” and all these arguments.
So, for me, the fact that the next genome project did not appear around 2003 was frustrating, and every year after that it just got continually more frustrating for me. By 2009, 2010, I’m still surprised that there’s no coordinated genome project, because Craig Venter in 2010 had made the first synthetic bacterial genome and published that.
Fast forward to 2012, I had been given this new job at Autodesk. I wrote a piece for Huffington Post asking the question, “Isn’t it time for a new human genome project?” But the response was basically, “Nah.”
Amazingly, through a series of events, which I can go into if you’re interested, I got to propose writing a human genome as a project in 2015 to the synthetic yeast community. Wrong audience.
Walk me through that a little bit.
Sure. So Craig Venter boots up the first synthetic bacterial genome, a prokaryote, no organized nucleus. They are simple organisms relatively, and let’s just say that most biologists don’t study the prokaryotes.
The academic community interested in writing genomes said, “Okay. Craig did the first prokaryote, let’s do the first eukaryotic genome.” Eukaryotes are cells that are similar to our own. They have a true nucleus. They are much more sophisticated in their organization compared to bacteria.
So Jef Boeke — at the time at John Hopkins — organized the synthetic yeast project, and it was a big deal. Craig’s [synthetic bacterium] genome was about a million base pairs. The challenge with yeast is it has about 12 million base pairs. The chromosomes are organized much more like human chromosomes, so they have much more structure compared to a bacterial chromosome.
So the [synthetic yeast researchers] created this international team of scientists to work on different chromosomes and do the foundational work in designing and synthesizing and assembling the yeast genome. The project was developing the core tools and technology, and it is the most sophisticated genome project in the world today.
They organize an annual meeting. I was at the fourth annual meeting in New York in 2015, invited to be on stage with an ethicist from John Hopkins; Nancy Kelley from the New York Genome Center; and the chairman of BGI, the big sequencing group in China. There’s about a hundred people in the room. A lot of them young. Everyone’s involved in the yeast project. Technicians and post-docs working on the project and a bunch of senior scientists. And Nancy Kelley asked me, “Well, what’s the next grand challenge for this community?” because they were seeing the end of the yeast synthesis project on the horizon.
I just looked at Nancy, and said, “Well, there’s only one grand challenge in synthetic biology to my mind, and that is to synthesize a human genome. Everything else is a project. Everything else will be interesting to some community of researchers, but really, the grand challenge is to synthesize the human genome.”
Hold on. I’ve just got to stop you there. You’re talking about synthesizing a human genome but not building an absolutely, completely new mammal from scratch, right?
You can’t build anything from scratch. You write a genome, you put it in a cell, and the cell divides and becomes the organism, right? We’re not talking building a cell from scratch. We’re talking about programming a genome that can lead to the development of the organism and putting it into an existing cell, an existing egg, so to speak.
You can synthesize a dog genome, a plant genome, a mouse genome. All of these would be interesting, but the only grand challenge I saw was to synthesize a [human] cell. To me, it just seems obvious. It’s not uncommon in other fields of engineering to see people get together — the Large Hadron Collider, for example — and spend billions and billions of dollars and decades of work by tens of thousands of people to stand up a new project, or in aerospace, too. Go to the moon or Mars.
But life science only has one big project that they can point to and that’s the Human Genome Project. Essentially, I proposed [another one] at the yeast meeting. I said, “Why don’t we take a page out of history? We know synthetic biology is happening. We know there’s a lot of activity. Rather than just going from taking a small incremental step from the yeast genome to another model organism, why not go all the way to the grand challenge and go and really push your limits and inspire people and do a big project and deal with all the thorny issues that most scientists just don’t want to deal with?”
And, honestly, it was the wrong place to propose that. But most panels are boring. This galvanized the room. I saw really shocked reactions on some people and just ear-to-ear grins on other people, and it was a lively, dynamic discussion, and I thought, “Wow, there is something here.”
What does that actually look like — the writing of genetic code?
It looks like computer design software — not very good software right now — that requires a lot of manual editing. The way to build DNA is to write the code, using bio-informatics software. Then the code is sent to a lab, where they chemically synthesize the snippets of DNA. These short fragments are then assembled into longer fragments and inserted into a cell. That’s the process.
Thanks. Now let’s jump back to GP-write. You brought in George Church next. Right?
I’d just gotten this ambassador position. The American Association for the Advancement of Science and the Lemelson Foundation were doing this ambassador program around invention, to kick-start invention again. So I had a mandate to go and kick butt, essentially, to go and inspire people to —
They gave Andrew Hessel what he’s been always wanting: the “go stir it up” mandate.
Exactly! And these are two of the biggest organizations in science and development. And it’s just like, “Wow! This is really cool.” Because I was emboldened by this ambassador program, I called up George and said, “I nominate you to lead the next genome project around synthesizing the human genome.”
George is so gracious and open, and he gave it serious consideration. He said, “You know, you’re right, the scientific community hasn’t really done anything to organize around synthetic biology in a serious way, in a global way.”
George was part of the first genome project, so, wow. George actually agreed to it reasonably quickly. But he put one condition on it: that Jef Boeke, the leader of the yeast genome project, be the co-lead.
Jef’s a geneticist but sort of the anti-version of both you and George, right?
I definitely like thinking I’m a provocateur. Jef is really, really conservative and thoughtful and doesn’t rush into things. He plans things out very, very carefully. I think Jef did the design phase of the synthetic yeast project. It took about a year and a half. So I knew I would probably rub Jef the wrong way.
It took about three months, but the more Jef investigated the idea, the more he saw the value in having a big stepping stone [for synthetic biology] after the yeast project was done.
Remember, this does not preclude doing any other synthesis project. It’s just, “can we set the bar a little higher?” Because these tools and technologies are coming fast.
The project generated a lot of attention early on — not all of it good — because you held one of your major meetings in secret. Why did you choose to go that route?
Yeah, our secret meeting to synthesize human babies, I heard about that. Except, the meeting wasn’t secret. The issue was that we [the leaders of GP-write] had our foundational paper under review at Science. And the magazine had embargoed the paper and its contents until publication. So we literally weren’t allowed to talk about it. But, I’ll tell you the truth, if you want to get a lot of attention really quickly — host a fake secret meeting. It really works.
All right. You’ve explained why GP-write is important in terms of inspiring the next generation of scientists. But let’s return to where we started and drill down into those practical benefits. What are they? Why does this project matter so much?
Oh man. Put it this way. We are at such early days of being able to design, synthesize, assemble, boot up, and test organisms, that we need a foundation to stand on. Imagine if we were trying to make electronic circuits, and there was no standardization in being able to describe your electronic circuit and no standardization in any of the components you were trying to solder together. Imagine if you were trying to build a network, and there were no standards for sending data across those networks.
Everyone that is working in synthetic biology right now is a pioneer. And they are all starting to build tools, technologies, robotics, software, etc. to do their projects. I think it is really important to start to build a foundation for everyone to be able to do this sort of work. It’s not just the synthesis technologies, which are still pretty crude, or how you get a new DNA into a cell and do all the testing of whether that DNA’s working.
There’s been some great work in the last few years with a synthetic biology open language. There are scripting languages for that, so you can actually do this faster. No one in electronics programs computers with zeros and ones, and yet most genetic scientists still work at the level of A, C, G, and T.
There is the legal framework to work in. What’s the intellectual property? What are the ethical boundaries that we should be paying attention to? What are things that we should not do? And how do we monitor for it?
There is this entire field right now that is just a big bubbling cauldron, and it needs structure. And no one had proposed anything large enough, any community large enough, to start even discussing what that structure might be.
So where we are right now? I know that people have really rallied around the creation of what’s called an ultra-safe cell. What is that and why is it important?
What George and Jef explained to me and educated me about was, “Look, we have to frame this around pilot projects.” In other words, scientifically interesting projects that are stepping stones to being able to synthesize a large genome like the human genome.
And the number one stepping stone, the one that was held up as the first shining beacon as a pilot project, was the ultra-safe cell line. [It would be] a platform for many biomedical applications.
Growing cells in the lab is really hard because they need rich media, they’re really finicky, and they can get contaminated with bacteria and viruses. That can ruin your experiment, it can corrupt your cell lines, it can totally destroy your manufacturing processes. The ultra-safe cell [would be] engineered to be highly resistant to most of the weaknesses of a human cell. There are ways to make cells virus-resistant. They want to make it prion-resistant. They want to make it transposition-free, so it doesn’t have any genetic components hopping around [and getting into other cells].
And then it’s radiation-resistant. They can make it cancer-resistant. Immuno-negative, engineered to minimize immune rejection. They have all sorts of multiple safely targetable sites. So if they want to add a new feature into the genome, they can direct it to a certain place where they know it’s not going to disrupt another important part of the genome.
Help me understand: is that going to happen all at once? Are we going to start with an ultra-safe cell that is the full menu — from prion-resistant to cancer-resistant — or do we start with the prion-resistant cell and then we add in the next layer and add in the next layer?
All of these properties are engineerable today. It could actually be divided up against 10 different groups to work on each individual feature. Then they can be recombined into a single cell line at the end.
The work can be parallelized without too much of a problem. So, some groups can work on the virus resistance and that’s it. That’s their expertise. You know, it would be great to be able to do all of this in a computational framework. Here are all the changes that we think are going to work, put them all through at once. But we just don’t have that level of sophistication today. So the project would be divided.
What’s something else of practical value that has been proposed beyond the ultra-safe cell?
Synthesizing a prototrophic mammalian genome.
This is fascinating because our cells do not make all the essential amino acids that are required to make our own protein. We can make many of them, but not all of them, which requires us to go and eat other organisms to get the basic building blocks to go and make our own protein.
And so, one of the proposed projects, by Harris Wang, was to go and put in the metabolic pathways to make all the essential amino acids into a human cell. Which is fascinating. If it could ever be put into the human body, it could virtually eliminate malnutrition. You would be able to make most of the foundational materials that we need to metabolize.
Talk to me about the challenges. Let’s talk about the funding challenges right now, then I want to hear about the technological challenges and the scientific challenges.
Okay. Funding is an interesting one. We knew going into this project that it was a different time than when the first genome project was stood up. When the human genome project started, eminent scientists went to Congress, made their arguments and Congress essentially allocated money for the project. As we know, it was $3 billion.
It’s a different time now. We thought it was unlikely that we were going to be able get a large block of money from the government to move a project like this forward, particularly because it could be politically toxic, [even though] there is nothing in the project that advances us toward making a human being.
Now, at some point, we realized money would probably come to the organization, but we just felt that at the beginning it was going to be a spectrum of funding. It’s going to be some small grants going into projects; it’s going to be some corporate sponsorship; it could be philanthropists. And Autodesk is a really generous company when it comes to sponsorship, so I figured there was a good chance of this. It was just one of the easiest conversations I’ve ever had. I told [Carl Bass, then Autodesk’s CEO], “you’ve got these eminent scientists coming together. They’re really interested in doing this project. But we need some money to start organizing the scientists.”
“Okay, how much?”
“Why don’t we start with a quarter million and take it from there.”
It was just like, “Great! Done!”
The CTO, Jeff Kowalski, said, “I’m really supportive of this, but I want to play devil’s advocate. How will this be perceived, and what happens when Fox News starts calling and saying that we’re involved in making synthetic babies?”
And Carl just said, “I’ll take the calls.”
So there was not only a generosity, but there was a fearlessness and an excitement that this was a project whose time had come.
By the way, you guys started out as Human Genome Project-write — HGP-write — and now it’s just GP-write. Was “human” just too controversial?
There are some people that think this is going to be toxic, and [calling it HGP-write] is way too provocative, and others, including myself, that just go, “Look, let’s wade into this, and just take the body hit early, and then we can go and have the real discussions about the substance.”
But after the first meeting in 2015, which was the “secret meeting,” there was a growing consensus amongst the participants that it should be de-tuned to just GP-write. HGP-write is still part of it, but it’s one of the projects under the GP-write dome. So this is about synthesizing large genomes, including the human genome.
And I said, “Guys, from a marketing perspective, I just want to say, you’re burying the lead, and now you’re complicating it. ”
Let’s talk more about the controversy, and the question of synthetic babies. Tell me how writing a genome doesn’t open the door for synthetic babies.
Well, I will say: “Yes. One day there will be synthetic babies.” But they will not come from this project. It won’t come, in my opinion, 20 years after this project is done. There is just way too much fundamental work that needs to be done before the risk and benefits of pioneering a synthetic human baby can be validated.
Look, we already have synthetic babies. We already do engineering of babies in the sense that we have IVF. We do cell manipulation, not necessarily genetic manipulation, to make babies.
Then there’s the world of CRISPR, the world of editing genomes, which has exploded over the last five or six years. That is really dealing with the issues of truly engineering a human genome in an embryo. Very quickly, scientists came to the consensus that, “you know, this is genetic surgery for repairing life-threatening diseases on individuals. If we can intervene even at an embryo stage, ethically, we’re okay.” That’s still controversial.
We don’t have the technology today to synthesize bacterial genomes quickly and inexpensively and reliably, so we’re not going to make a synthetic human genome that will lead to a baby tomorrow. But I can certainly see that this project should lay the foundations for faster, more reliable, more robust designs, builds, and tests of genomes, including the human genome. So we lower the economic barriers, build a technological foundation, and start to build a legal and ethical framework. That will take us to about 2026, if we estimate correctly. That just gets us to a starting line.
Now remember, just reading the first genome cost billions, the second one was substantially cheaper. So if we make the first synthetic human genome by 2026, it’s still going to be too expensive. This is not going to be off-the-shelf technology. Allow another 10 years to start making it cheap enough that you can actually start to experiment reliably and quickly.
So you start looking at timelines like that, okay, well maybe someone will make a synthetic human baby when standing on decades and decades of increasing complexity and work, you know, somewhere around 2050 or 2060. We’re certainly going to be editing genomes sooner than that, but when it comes to complete synthesis, which is the way I’m framing this, it’s going to be a long time.
I’m not worried about synthetic babies. But will synthetic babies come in the future? I think when the timing is right, the tools are available, the technology is affordable and reliable enough, someone will pioneer that particular project. I don’t know how you could stop it.
Also from NEO.LIFE: Bring on the Post-Human Future
Gregory Stock is ready for cloning, gene editing, and conscious AI. But first he’s trying to get you better health care.
I think long before we get to synthetic babies, we are going to get mutated humans, human-animal hybrids. That’s going to come out of subculture. It’s going to be the next thing for the body-modification crowd.
Oh, the grinders and hackers among us. Put it this way, there are very few restrictions, practically, on self-experimentation. If you’re an adult, and of sound mind, and have access to tools, and you want to go and modify yourself, there’s very little that society can do to put a stop to that. In fact, self-experimentation has been a key part of advancing research and development. So, if you wanted to add a tail or glow in the dark, or modify your eyes even, then tech is coming soon and I would not stand in your way.
We talked about the initial check from Autodesk. How funded are you right now?
Not enough. Funding is coming in for various pilot projects. People are starting their own synthetic biology centers around the world. The Genome Project-write is the community umbrella. They are using that as the validation for getting support, and it’s causing this really amazing network of scientists, engineers, and other community members — people like ethicists and people interested in grant writing and finance and other things — to come together.
The cool thing is, unlike the first genome project, where these centers all dissolved after the project was completed, these have the promise of becoming enduring centers that just become the homes and hubs for people engineering biology.
So what technologies would advance GP-write from being at a basic-science level? What are the tools that would really make a difference?
We need better synthesis technologies. The synthesis technologies we have today are really good compared to where they were 20 years ago, really good. But they’re not good enough to print a 50 million base-pair chromosome quickly and cost effectively.
Right now, [synthesizing] DNA, let’s just call it around 10 cents a base pair. That gets to be pretty expensive, and that’s working with relatively small segments of DNA. But it’s not just synthesis, it’s synthesis and assembly of the DNA. It turns out that if you’re starting to work with large constructs, a million base pairs, two million base pairs, the costs go up substantially in that manipulation. So it may turn out to be, all in, closer to a dollar a base pair.
Now that was the original estimate for reading the genome 30 years ago. Right? Like a dollar a base pair. So writing at a dollar a base pair, a whole genome is not out of line, but no one is going to spend that for writing the genome today. It just doesn’t make any economic sense. What really does make sense is taking that money and starting to prototype better DNA synthesis technologies. And that’s what we’re seeing.
This is starting to really be a hot area of R&D, not just for biological DNA applications, but for things like using DNA as data archiving in the computer industry, kind of like biological magnetic tape. But it will take a breakthrough to make synthesizing whole large genomes more cost-effective.
I have not yet seen the breakthrough technology appear in the literature that will push us over the line and make human genome synthesis really cheap or make large genome synthesis cheap. But it’s coming. I can feel it.
I want to be clear. Biology manufactures the most complex things in the universe for virtually no cost. Right? So it is the cheapest form of advanced manufacturing imaginable. So as we start to build interfaces to the cellular machinery, everything we know about designing biology, the economics of designing biology, falls essentially to zero.
That is the thing that is why GP-write is so important, because all life on this planet, all life from bacteria to you and me to blue whales, uses the same programming language, the same cellular architecture. It’s all conserved. So if I learn how to program a bacterium, I learn how to program you and me. And if I develop tools and technology for controlling the cellular processes to write DNA, it will work for encoding anything biological.
So, I know that a human genome is essentially free because, again, every time a cell divides, it writes a human genome. And it costs nothing. All we have to do is figure out that interface and then build design tools to be able to inject our intentions into that machinery. Then biology goes through a Cambrian explosion of human creativity. And that’s not that far away.