Chapter 13 DIY Biology

The ultimate dream of the Fab Age is just that—the universal fabricator. Just like the Star Trek Replicator, it’s a machine that can make almost anything on command. This idea has fired the imagi­nation of science fiction for decades.

In the beginning was an empty chamber, a diamond hemisphere, glowing with dim red light. In the center of the floor slab, one could see a naked cross-section of an eight-centimeter Feed, a central vacuum pipe surrounded by a collection of smaller lines, each a bundle of microscopic conveyor belts carrying nanome­chanical building blocks—individual atoms, or scores of them linked together in handy modules.

The matter compiler was a machine that sat at the terminus of a Feed and, following a program, plucked molecules from the conveyors one at a time and assembled them into more compli­cated structures.⁵¹

That’s fiction, but something similar is not impossible. MIT pro­fessor Neil Gershenfeld thinks it’s just twenty or thirty years away.

How will we get there? The path, Gershenfeld argues, will not be simply making Ç-D printers and other CNC machines faster and more precise. The problem with those approaches, he says, is that they just “smoosh stuff around.” They may squirt it or cut it or heat it, but they’re just moving material or changing its state (hardening it). The material itself has no intelligence or sense of what it’s supposed to be. Your fabrication machine has to do all the work; the material isn’t “helping.”

Contrast that with simple Lego blocks. When a child plays with Lego, the blocks correct the child’s mistakes—they only fit together if they’re lined up right. The larger Duplo blocks guide the child to the correct orientation with beveled edges that exert a force to rotate the parts in the right direction to fit when they’re pushed together.

Programmable matter

In a sense, even Lego blocks are “intelligent matter.” They carry with them their own rules of assembly and have pre-assigned functions, such as hinges and wheels.

Sounds crazy? It’s not—it’s already all around you. That’s the way nature works. Crystals, after all, are made of atoms self-assembling into incredibly complex structures, from snowflakes to diamonds. Your own body is made up of proteins assembled under the instruc­tion of your DNA/RNA from amino acids, which themselves are made up of self-assembled atoms. Biology is the original factory.

“Intelligent materials” describes some of the basic building blocks of life. Gershenfelds favorite example among those are the ribosomes found in your cells. A ribosome is a protein that makes proteins—a biological machine that makes other biological machines. But as Ger- shenfeld sees it, it’s a model of an advanced fabricator.

In your cells, genes coded in DNA are translated into RNA, a sort of mirror image. The ribosome is the “organelle” that reads the RNA and follows that code to assemble amino acids to make up spe­cialized proteins. Once made, those proteins automatically fold into complex shapes, driven by no more than the electrical charges and the attractive and repulsive forces that come from their atomic bonds. Those shapes, self-assembled in the billions, make up the structural elements of your body, from cell walls to bones.

This is a case of a one-dimensional code (DNA, four chemical “let­ters” in different combinations, strung together in a long monodimen­sional chain) creating a three-dimensional object (proteins). Because the material DNA is working with—first RNA, then ribosomes, then proteins—are not just smooshed together but instead have their own chemical and structure rules and logic, a little information can create incredible complexity.

In Gershenfelds MIT lab, students have taken some baby steps toward that, with tiny electronic components that can be plugged to­gether and automatically make the right connections. But researchers elsewhere have taken the concept even further. The most promising programmable matter is DNA itself.

The new field of “structural DNA” uses the material not as a ge­netic code, but as the building material itself, with no biological func­tion. Some sixty labs around the world today are now working on this, and researchers can synthesize strands of DNA that will form squares, triangles, and other polygons.⁵² Some of the structures are made by “tiling” many two-dimensional DNA shapes into one sheet. Others program the DNA to fold into three-dimensional shapes, a process called “DNA origami.”

Three-dimensional DNA structures can be programmed to as­semble into “scaffolding,” making structures like a box. Other se­quences can be programmed to respond to a chemical stimulus to open, making a door. The idea is that a drug could be placed in a structural DNA box, with the door shut, and transported by the body to a place where the drug was needed. Then a chemical trigger could be sent to open the door, and the drug would flow out, precisely where intended.

We’re a long way from such programmable nanomachines creating large-scale objects of any material. For one thing, DNA is not very rigid, so researchers have experimented with bonding other materials, such as gold nanoparticles, to DNA to strengthen it. Even then, they haven’t made anything big enough to be seen without a microscope. Other researchers have experimented with doing similar things with special polymers and other chemical compounds, which have advan­tages in stiffness but are harder than DNA to program.

It’s all been mostly proof-of-concept research so far.

Making with DNA

Just before midnight on an April Friday in 1983, Kary Mullis, a chem­ist with a streak of surf bum, was driving along California’s Pacific Coast Highway 128 between Cloverdale and Booneville sometime when he had an idea that would eventually win him the Nobel Prize. At that time, one of the biggest problems in genetics was that there was never enough DNA to study, and what DNA could be found was often contaminated.

As he drove, Mullis was noodling over various ways to analyze mutations in DNA when he realized that he had stumbled on a way to reproduce any DNA region with the use of a special bacterial en­zyme called DNA polymerase and a process of applying cycles of heat. Others had thought of using the polymerase for copying DNA sections of interest, but Mullis realized that cycles of heat would lead to a chain reaction by which each cycle doubled the number of copies, quickly reaching the millions.

In combination with a version of that enzyme which was derived from bacteria called extremophiles that live in hot springs and are heat-resistant, this led to an automated process of copying DNA that created the modern genetics research industry. Called Polymerase Chain Reaction (PCR), it won Mullis the 1993 Nobel Prize in Chemistry.

Today, PCR machines, also known as thermal cyclers, are a staple in any genetics lab. Once costing nearly $100,000 each, they can now be found commercially for as little as $5,000. PCR is one of the mir­acles of the genetics revolution and a cornerstone of the New Biology.

But Josh Perfetto, a young Californian researcher, still wasn’t sat­isfied. PCR machines were still too expensive, and what’s worse, they were all closed, proprietary systems. What if you wanted to use them powered by batteries in Africa? How about with kids in the class­room? What if you wanted to experiment with the machines them­selves, not just what went into them?

In short, he wanted to hack the PCR machine and open it to the world.

A tiny spare bedroom is not an ideal space for a high-tech biofab­rication facility. To get to the one Perfetto is putting together, visitors must walk all the way to the back of his mostly unfurnished house in Saratoga, California—through the kitchen, past some empty rooms, across a den with a lone couch—then climb a poorly lit staircase and round a corner. The room itself is about 120 square feet and has one big window with a view of an adjacent roof. There’s an eight-foot-wide gap in the middle; the rest of the room is for science. “I thought about moving the lab to the empty living room downstairs,” Perfetto says. “I really need more space. But that’s right by the front door. I don’t want to freak people out.”

He laughs a little awkwardly, and it’s easy to see why he’s worried. With its Pyrex containers on metal racks and other clinical-looking equipment, the bedroom looks perfect for cooking crystal meth. A mass of wires spills out of a wooden box; on top sits a metal plate punched full of holes. A table holds several laptops, test tubes, a box of purple surgical gloves, a rack with pipettes in various sizes, rubber tubes connected to vials, an orange plastic box with a blue light in the bottom, and a centrifuge that looks like an oversized rice cooker. The wooden box is actually a homemade PCR device. And the orange plastic thing runs gel electrophoresis, a way to sort DNA strands by size. Perfetto, an engineer, built a few of the gadgets himself.

‘Tve been sleeping in here,” says Mackenzie Cowell, Perfettos business partner. “And who knows what kinds of chemicals have soaked into this rug!” He flew out to California from Boston a week earlier and has been working with Perfetto on a DIY genomics kit to sell through their new business, CoFactor. The problem is, right now extracting and amplifying DNA at home still takes too many steps. The guys are worried that people won’t enjoy the process if it’s too complicated.

And the home audience is their target market.

Science is all about coming up with smart ways to answer hard questions. But sometimes getting those answers requires expensive machines. Physicists looking to understand the universe don’t just set up a pendulum anymore—today they build multibillion-dollar un­derground particle accelerators. PCR machines, critical to genetics- powered biology, start at around $5,000. And those machines, with their intricately tuned bits and pieces, aren’t friendly to the kind of void-your-warranty hacking at the heart of the maker movement (not to mention creative experimental design). In short, no amateur is going to drop tens of thousands of dollars to get a lab running, and many scientists don’t understand the inner workings of their expen­sive, grant-funded gadgetry well enough to whimsically crack the ma­chines open and see how they can be modified. But thanks to the DIY revolution and Arduino, the open-source circuit board, big thinkers like Cowell and engineers like Perfetto (whose OpenPCR device sells for just $599) are reverse-engineering the big-budget tools. And then they’re sharing their methods with the world.

Ask people inside the biohacker movement where they think it will have the biggest impact and they talk about education—being able to do genetics in classrooms. They regularly bring up Sushigate, the 2008 case of New York City high school students who used DNA testing to discover that sushi restaurants and supermarkets were mis­labeling their fish. The results may be cool, but for now the machines are where the real action is. Behind the scenes, engineers and science enthusiasts are teaming up to mod tools and technologies and then sell their inventions—or simply share tips on how to build them—to anyone interested. Homemade PCR devices are drawing the most attention, since anyone who wants to work with DNA has to put it through a PCR machine first.

That’s what has drawn a few hundred people to the online com­munity surrounding Perfettos OpenPCR project. Polymerase chain reaction is really just a process of heating and cooling genetic ma­terial. Weill Cornell Medical College researcher Russell Durrett, cofounder of New York City’s community lab GenSpace, built one using a lightbulb, an Arduino board, an old computer fan, and some PVC pipe. Biotech advocate Rob Carlson also has a version, called the LavaAmp, that he says could be easily mass-manufactured just like any consumer product. “It doesn’t have to be a big company,” Carlson says. “The manufacturing is set up so that if anybody wanted to make 100,000, they could do that, and the quality of the resulting molecules would be just fine.”

But PCR machines are only the beginning. Keegan Cooke, a for­mer microbial fuel cell researcher, has been selling a home-built battery called a MudWatt kit. The MudWatt creates energy by capturing elec­trons released when mud-borne bacteria eat sugars. The kit comes with an anode, a cathode, and an LED light. Users fill the box with about two cups of muck—any sort has the right microbes, though stinkier stuff seems to work better—and some leftovers from the fridge (to feed the critters). The microbes generate electricity as they eat, and the elec­trodes capture it to power the light. This microbial fuel cell tech isn’t good enough to be scaled up to wide use yet, but the open-source model for distribution means that people can start making advances in their backyards. Cooke has already updated and modified his kit based on user feedback. “Another customer found that the mud in his nearby river generated almost double the power. He also recommended that we try a different material for our electrodes, and we found that it also produced double the power. It’s nice, this process of people giving us feedback and evolving the technology,” Cooke says.

Another example: Cathal Garvey’s DremelFuge. The centrifuge— a device for rapidly spinning substances to separate lighter compo­nents from heavy ones—is essential in many fields of science, but a professional-grade version can cost thousands of dollars. Garvey, a biologist in Cork, Ireland, designed one that can be made by 3-D printers—either with a MakerBot or by the Ç-D print-on-demand company Shapeways (for $57). The DremelFuge is a small round disk with slots that hold standard microcentrifuge tubes. It’s designed to fit snugly onto a rotary tool, which can spin the tubes at 33,000 rpm, producing up to 51,000 Gs. (A standard professional centrifuge pro­duces only about 24,000 Gs.) Garvey gave his DremelFuge a Creative Commons Attribution ShareAlike license, meaning it can be used or remixed by anyone.

Perhaps the biggest toolmaking success so far, however, comes from the world of neuroscience. It’s a hand-sized box made of translucent neon-orange plastic with some electronics inside and a wire sticking out. For just ninety dollars, the device—called a SpikerBox—does something remarkable: It records and makes audible the sound of neurons firing. Connect two electrodes to the leg of a live cockroach (included); every time the bug twitches, its neurons emit an electrical spike that translates into a loud click. It sounds simple, but the ability to reveal a spike for such a small amount of money is a bit of a revolu­tion in the study of neurobiology.

The SpikerBox grew out of the frustration of its inventors—engi­neers Tim Marzullo and Greg Gage—with the high cost of their lab equipment. As students at the University of Michigan’s Neural Engineering Laboratory, they came to feel that their work wasn’t hav­ing enough impact given the money being spent. “In that lab you do silly things like design electrodes that cost thousands and thousands of dollars,” Gage says. “And then they would sit on a shelf afterward, because they were someone’s Ph.D. project.”

Eventually, Marzullo says, they realized that “if the objective is to show spikes, you don’t need a million dollars and a clean room.” They started on what Gage calls a self-imposed engineering chal­lenge: make electrodes for $100. They told each other that if they pulled it off, it’d be “funny.” Then they wrote an abstract and brought prototypes of their hack to the Society for Neuroscience conference in 2008. At their poster presentation, the prototype SpikerBox didn’t work—yet it still caused a stir. “We were flooded with neuroscientists and educators who said they’d been waiting for this for years,” Mar- zullo says. (Shortly thereafter they built a box that worked.)

In 2009, Marzullo and Gage started a company, Backyard Brains, to sell the box. They’ve shipped over 550 kits and demo’d the de­vice for more than six thousand people (including passengers on a Delta Airlines flight, where they stuck a sign to the bathroom door reading FREE NEUROSCIENCE LESSONS AT SEAT 33A AND â). Though they’re keen to earn a living from their company, they’re fine with the relatively small amount of money they’re making now—as long as

someone out there is learning about neuroscience. (A recent $250,000 grant from the National Institutes of Health also helps.) Which is why they’re strong believers in keeping the device and its intellectual property open-source. It can be purchased in various stages of assem­bly, and detailed instructions for building the SpikerBox are available for free download on their website, along with software for interpret­ing the intensity and duration of the spikes.

SpikerBoxes are even making their way into real science labs. W. David Stahlman, a professor of psychology at UCLA, recently pur­chased one to use in his research on hermit crab behavior. Tradi­tionally, he says, psychologists don’t focus on neuroscience. But while studying attention, learning, and distraction in the hermit crabs, he grew interested in how those behaviors are exhibited in the crabs’ brains. A SpikerBox means that doing this kind of experimentation is no longer cost-prohibitive for a young professor. Even better, there’s no warranty to void. “When you’re working with equipment that’s much more expensive, you’re more hesitant to open it up and tinker with it,” Stahlman says. And as a nifty bonus, he can capture the data right to an iPhone or iPad.

The companies that sell professional-grade PCR machines to laboratories are predictably unimpressed by all this fuss. Jeff Ros­ner, head of research and development for the PCR division at Life Technologies and a former engineer at Hewlett-Packard, says he’s in­trigued by the OpenPCR movement, but he hardly sees the gadgets it’s producing as competition. “What they’re doing is really sort of PCR 101,” he says. “The thermal-cycling machine is only a small piece of what’s important about PCR and what’s required to do it. You need so many other things, including access to chemistry that’s way harder to hack than the machinery itself.” (Most chemical re­agents are proprietary, and every manufacturer has a unique process for making the chemicals work. Perfetto and Cowell have been mix­ing and matching chemicals and protocols to simplify the process for their home kits.)

Still, it’s difficult for Rosner to conceal his excitement over the fact that hackers are getting interested in his technology—and he admits that he actually has a machine shop in his own backyard. “There are some real barriers for them,” Rosner says. “The reality is that costs have declined from hundreds of thousands of dollars to tens of thou­sands, but they have to get down lower before they’ll be accessible to hackers.” Then, after a pause, he adds, “I hope that happens.”

What if it does? Right now, the biohackers are largely just re­inventing the wheel, creating DIY versions of equipment and tech­niques already found in standard professional and academic labs. But elsewhere the DIY credo is taking a different twist. Underground synthetic chemists are creating variations of illegal drugs that have the same effect as those drugs but are chemically different enough to be legal. In a game of cat-and-mouse with the regulators, they have proven able to invent new compounds in their DIY labs faster than the regulators can identify and ban them. Synthetic powders that have similar properties to the THC in marijuana are sold legally in head shops in the United States, despite evidence that they can cause more harm that real marijuana.

And that’s just chemistry. What happens when the tools get pow­erful enough to extend to biology and genetics, too? Today we can amplify and identify DNA at the kitchen table. Tomorrow we’ll be able to sequence it, too. But after that comes synthesizing it, modify­ing it, and the rest of genetic engineering. The day when only a small number of professional labs could do this, checking and screening every request that comes in, will soon end. At that point, people will start hacking life. We’ve been doing that for thousands of years with cross-breeding and agricultural genetics, but that was always within the bounds of nature. In the lab, there are fewer such bounds.

When my family and I built a new workshop in the lower floor of our house, the architects asked what I wanted to do to future-proof it. The answer seemed clear: make it Biohazard Ï-compliant. It would need the sort of ventilation that you might find in a standard univer­sity biology lab.

Why? Γm not sure. But I have a dream that before my youngest daughter has left the house, we’ll have created a new organism in the basement. Nothing fancy, mind you: just some variant of E. coli that lights up in some sort of microbial fireworks display, or maybe one that smells stinky to gross out her friends. But it’s a start. Nature has created the most powerful factories of all. Who knows what can hap­pen when we can command them, too?