Bioengineers Build Open Source Language for Programming Cells

ORIGINAL: Wired

By Daniela Hernandez

04.19.13

Image: Steve Jurvetson/Flickr.

Drew Endy wants to build a programming language for the body. 

Endy is the co-director of the International Open Facility Advancing Biotechnology — BIOFAB, for short — where he’s part of a team that’s developing a language that will use genetic data to actually program biological cells. That may seem like the stuff of science fiction, but the project is already underway, and the team intends to open source the language, so that other scientists can use it and modify it and perfect it. 

Photo: BIOFAB

The effort is part of a sweeping movement to grab hold of our genetic data and directly improve the way our bodies behave — a process known as bioengineering. With the Supreme Court exploring whether genes can be patented, the bioengineering world is at crossroads, but scientists like Endy continue to push this technology forward. 

Genes contain information that defines the way our cells function, and some parts of the genome express themselves in much the same way across different types of cells and organisms. This would allow Endy and his team to build a language scientists could use to carefully engineer gene expression – what they call “the layer between the genome and all the dynamic processes of life.” 

According to Ziv Bar-Joseph, a computational biologist at Carnegie Mellon University, gene expression isn’t that different from the way computing systems talk to each other. You see the same behavior in system after system. “That’s also very common in computing,” he says. Indeed, since the ’60s, computers have been built to operate much like cells and other biologically systems. They’re self-contained operations with standard ways of trading information with each other. 

‘In synthetic biology, the equivalent of a Java virtual machine might be that you could create your own compartment in any type of cell, so your engineered DNA wouldn’t run willy-nilly.’

— Drew Endy 

The BIOFAB project is still in the early stages. Endy and the team are creating the most basic of building blocks — the “grammar” for the language. Their latest achievement, recently reported in the journal Science, has been to create a way of controlling and amplifying the signals sent from the genome to the cell. Endy compares this process to an old fashioned telegraph. 

“If you want to send a telegraph from San Francisco to Los Angeles, the signals would get degraded along the wire,” he says. “At some point, you have to have a relay system that would detect the signals before they completely went to noise and then amplify them back up to keep sending them along their way.” 

And, yes, the idea is to build a system that works across different types of cells. In the 90s, the computing world sought to create a common programming platform for building applications across disparate systems — a platform called the Java virtual machine. Endy hopes to duplicate the Java VM in the biological world. 

“Java software can run on many different hardware operating system platforms. The portability comes from the Java virtual machine, which creates a common operating environment across a diversity of platforms such that the Java code is running in a consistent local environment,” he says. 

“In synthetic biology, the equivalent of a Java virtual machine might be that you could create your own compartment in any type of cell, [so] your engineered DNA wouldn’t run willy-nilly. It would run in a compartment that provided a common sandbox for operating your DNA code.” 

According to Endy, this notion began with a group of students from Abraham Lincoln High School in San Francisco a half decade ago, and he’s now calling for a commercial company to recreate Sun Microsystems’ Java vision in the biological world. It’s worth noting, however, that this vision never really came to fruition — and that Sun Microsystems is no more. 

Nonetheless, this is what Endy is shooting for — right down to Sun’s embrace of open source software. The BIOFAB language will be freely available to anyone, and it will be a collaborative project. 

Progress is slow — but things are picking up. At this point, the team can get cells to express up to ten genes at a time with “very high reliability.” A year ago, it took them more than 700 attempts to coax the cells to make just one. With the right programming language, he says, this should expand to about a hundred or more by the end of the decade. The goal is to make that language insensitive to the output genes so that cells will express whatever genes a user wants, much like the print function on a program works regardless of what set of characters you feed it. 

What does he say to those who fear the creation of Frankencells — biological nightmares that will wreak havoc on our world? “It could go wrong. It could hurt people. It could be done irresponsibly. Assholes could misuse it. Any number of things are possible. But note that we’re not operating in a vacuum,” he says. “There’s history of good applications being developed and regulations being practical and being updated as the technology advances. We need to be vigilant as things continue to change. It’s the boring reality of progress.” 

He believes this work is not only essential, but closer to reality than the world realizes. “Our entire civilization depends on biology. We need to figure out how to partner better with nature to make the things we need without destroying the environment,” Endy says. “It’s a little bit of a surprise to me that folks haven’t come off the sidelines from other communities and helped more directly and started building out this common language for programming life. It kind of matters.”