杏吧原创

Synthetic biologist Drew Endy is leading efforts to make the natural world programmable. He tells Douglas Heaven why succeeding means blurring the distinction between information and matter

How significant is this idea that living things can be engineered?
With Darwin and the theory of evolution came a sea change in perspective. We moved from an idea of the natural world as something that doesn鈥檛 change to something that does. I think biotechnology, at its heart, represents a similar transition. Biology is not just a science. Biology is a technology.

Do you think we fully grasp the magnitude of this transition?
We鈥檝e heard the word 鈥渂iotechnology鈥� so much that we don鈥檛 really wrestle with the depth of its meaning. In part, that鈥檚 because most of biotechnology hasn鈥檛 yet been imagined, let alone made a reality.

鈥淢ost of biotechnology hasn鈥檛 yet been imagined, let alone made a reality鈥�

We鈥檙e not very far along in learning how to partner with biology to make stuff. That is typically a last resort. If I could cure cancer with nuts and bolts, I鈥檇 use nuts and bolts because we figured them out in the 1800s. But for the Industrial Revolution to happen, somebody had to standardise parts, making it easy to build mechanical systems. Now someone has to do the not-so-glamorous work to make it easier to build biological systems.

Biotechnology includes synthetic biology and genetic engineering, what is the distinction between these?
In as much as anything else, synthetic biology is a community. Over a decade ago, we posed a question: how could we get better at engineering biology? Genetic engineers were happy using arbitrary, ad hoc methods over and over again. Synthetic biologists pushed three key ideas. First, we wanted to decouple design from manufacture and advance both computer-aided design and DNA synthesis. Second, we realised we needed standards to coordinate labour over time and across the world. And third, we borrowed an idea from computer science: that low-level details should be hidden. This means that you should be able to design a high-level organism 鈥� like a tumour-destroying microbe 鈥� without needing to know that DNA is made up of four bases.

Have these ideas been successful?
They鈥檝e led to real disruptions. As DNA design and synthesis become more powerful, matter and information become interchangeable. This is a big deal. It鈥檚 why, for example, biosecurity policies have had to change. You cannot rely on locking up smallpox in a laboratory when somebody could take the sequence information, print the genome and try to recover the virus. But equally, this allows vaccines to be transmitted at the speed of light over the internet 鈥� encode the DNA needed to make the vaccine, send that across the world and synthesise it at the other end.

So the distinction between living material and the information used to create it is collapsing?
Yes, it鈥檚 collapsing. When that happened with entertainment media it was a shock. People are still struggling with who owns different TV shows or music, for example. Am I allowed to download Game of Thrones from HBO, or not? If I can鈥檛 access it legally, am I going to pirate it? Apparently, a lot of people say 鈥測es鈥�. So when matter and information become interchangeable, it impacts all sorts of things. And it鈥檚 now happening at the heart of biotechnology, with genetic material.

Whether this is good for music depends on who you talk to. Would that be true for biology?
Some outcomes would be celebrated, others would be harmful, but it may take time to tell which is which. It鈥檚 interesting what Apple did with iTunes, for example. They took experiments in digital music downloads 鈥� early file-sharing services like Napster 鈥� and turned them into a way to make money. Are we getting better or worse entertainment as a result? I don鈥檛 know.

Are there now licensing issues in biology?
I co-direct the , the world鈥檚 first facility for producing general-purpose biological parts. Now that we鈥檙e shipping reusable parts, we鈥檙e experimenting with property rights. We鈥檝e released everything that has come out of the project under a public agreement, a collection of promises by the inventor not to assert property rights. It was established by the BioBricks Foundation, a charity working to advance biotechnology to benefit all of the people on the planet.

The agreement means that everybody is free to use these parts, so long as they agree to terms and conditions, such as not breaking the law. There鈥檚 no requirement to pay us or to give something back.

Is there a race between organisations like BioBricks and private companies who might want to lock down new inventions?
You mean EvilCo? Most genetic functions are not lottery tickets to billions in licensing revenue. Much like words in the dictionary, their value arises over time as they鈥檙e reused in innumerable novel combinations. It鈥檚 not straightforward to make money that way.

Are there lessons from the open-source movement in software?
In software, it鈥檚 mostly copyright, which applies when the work is run or downloaded and has no upfront cost associated. And a copyright doesn鈥檛 prevent you from independently developing code that does something similar.

In biotechnology we have patents as the property right. With patents, there is a high transaction cost, you have to file and get approved. And the claims, once granted, may not be fully understood until they鈥檙e litigated later, if there鈥檚 a conflict. Because this is new territory, it is not as simple as applying one or the other. So we have not yet been able to deploy lessons of free and open software to develop an open language for programming life, so to speak.

Do we need to create this open language?
Imagine if Abraham Lincoln at Gettysburg had had to seek a licence to say 鈥渇our score and seven years ago鈥�. Imagine if someone owned the rights to those words and you had to check before using them. And it鈥檚 worse than that. You also have to check similar words 鈥� 鈥渇our spore and seven beers ago鈥�. It may sound crazy but in biotechnology that鈥檚 what it鈥檚 like, you have to seek licences to use genetic words. So a long-term goal of the BioBricks Foundation is to create a free-to-use language.

How long will it take to create this language?
Many decades. But on a longer timescale languages that are successful tend to be the ones that are free to use. This is true in computer programming 鈥� and in communication between humans. There are some private languages and codes for the security world, but most languages are free to use. I鈥檓 not paying to use these words right now. In biotechnology we could do more to partner with the public domain to create things that are free to use.

You have spoken before about the necessity of taking a long-term view. Why is that?
We鈥檙e making investments for the next 70 to 100 years. It鈥檚 great to celebrate what鈥檚 happening right now, but it鈥檚 important to remember that there鈥檚 a longer arc.

I also want to avoid total industrialisation of nature. That sounds horrifying to me. We鈥檝e got to figure out a way to make our civilisation dance with the planet, including the natural living world. I鈥檓 interested in changing how we work with the environment and reimagining our material supply chains.

Is part of that change taking inspiration from nature鈥檚 own manufacturing processes?
How could you not? Look at trees. They grow self-assembling solar panels that recycle themselves. There鈥檚 a pine tree in my front yard that鈥檚 growing about a thousand pine cones. They鈥檙e growing from what, exactly? Atmospheric carbon, trace things from the earth, water and photons.

What do I do with this state-of-the-art manufacturing capacity? I wait for the pine cones to fall, then rake them up and put them on the compost heap. I鈥檓 an engineer and that鈥檚 what I do with this manufacturing platform in my front yard. It鈥檚 irresponsible. How could I not react to that?

This article appeared in print under the headline 鈥淒ownloading DNA鈥�