The sum of the parts: synthetic biologists string genes into living machines.

In the 30 years or so since its inception, genetic engineering has created quite a legacy, ranging from glow-in-the-dark bunnies to bacteria that churn out life-saving drugs. This now-common lab technique gave biologists their first taste of custom designing living things by tinkering directly with their genomes. But for all its utility, genetic engineering hasn't met scientists' most optimistic expectations. Because the enzymes that slice and splice DNA aren't entirely predictable, swapping out an organism's genes can be a hit-or-miss prospect.

Sometimes, it can take years to get a single gene into an organism and make it do what the scientists had planned, says quantitative biologist Adam Arkin of the University of California, Berkeley. "People try something, and it doesn't really work, so they try something else, and that doesn't work either. It's a lot of hunting and pecking, and not a lot of upfront design," he says.

Rather than randomly altering a few genes in a cell's DNA as in old-school genetic engineering, some researchers are now breaking genomes into collections of parts and precisely reassembling them to do a scientist's bidding. The fruits of the approach are taking many different forms: bacteria that can count or form patterns in a petri dish, a virus redesigned to make its genes easier to study, microbes programmed to seek out and destroy tumors, and bacteria that spit out great quantities of a rare and complicated malaria drug.

ENGINEERING LIFE "Genetic engineering isn't engineering in its true form," says Samir Kaul of Codon Devices, a Cambridge, Mass.-based company that synthesizes chunks of DNA to scientists' specifications. Unlike engineers who develop and construct electrical circuits or bridges, for example, genetic engineers initially didn't craft plans, design models, and then build DNA structures to accomplish precise goals.

In the past decade, however, several technologies have begun to bring genetic engineering in line with other engineering disciplines. The time and money needed both to decode DNA and to assemble new combinations of its four-letter alphabet--subunits known by A, T, C, and G--have drastically decreased. In the meantime, there's been an enormous increase in knowledge of what individual genes do and in scientists' ability to model how genes interact.

Therefore, rather than moving a gene from one creature into the DNA of another and hoping that the transplant will be effective, many scientists are now designing sections of DNA to fit their particular needs. Companies such as Kaul's manufacture the desired DNA sections according to the scientists' design and ship them out ready to slip into cells. To distinguish themselves from researchers who practice genetic engineering the old-fashioned way, those in the new field coined a name for themselves: synthetic biologists.

As a further nod to engineers in other disciplines, synthetic biology is amassing a collection of standard parts that function in predictable ways. These parts include sensors, such as a receptor that binds a particular protein; devices that build, say, a drug or a glowing protein; and what the scientists call biological circuits, collections of genes that act on each other to produce a chain of signals between the input and out put. By hooking together the bits of DNA that encode these parts and then introducing them into a cell, scientists can tinker, as an electrical engineer might, to make tiny, living machines.

For the most part, synthetic biologists are still exploring the process by playing with gene combinations. However, the few applications already on the scene showcase the enormous potential of this field, says computational geneticist George Church of Harvard University. "If you interviewed people when you first put two transistors together in a circuit, it would have been crazy if they said, 'Oh, we're going to be doing spreadsheets, word processing, and making an Internet for online shopping.' But that's the feeling today" with synthetic biology, he says.

GENE MACHINE A few researchers are using this new field to study life's basic mechanisms. For example, Pamela Silver of Harvard University studies how cells age, so she and her colleagues count how many times a yeast cell divides. "We quickly realized how cumbersome that was," she says. "Someone had to sit there and watch [yeast cells] divide under a microscope."

To speed her research--and save the eyesight of her team members--Silver is constructing a biological circuit that will prompt yeast cells to essentially count how many times they've divided. Her strategy relies on proteins produced in a daughter cell when it buds off from its mother cell. Sensors inserted into the cells would pick up this signal. Then, through a Rube Goldberg-esque chain of signals and responses, genes in the circuit encode proteins that would produce an output, such as a flash of light in a particular color, whenever an individual cell had divided. An automated system would record each flash of color.

"If you had enough different colors, you'd have a way of knowing how many times [each] cell has divided" Silver says. By the end of the year, she expects to have a cell that, she says, "can count to two"

Ron Weiss of Princeton University has created bacteria that, when placed in a petri dish, can display distinct patterns such as a bull's-eye. "Patterns are very common in biology," he says. With pattern-making batches of bacteria, researchers could investigate how tissues form patterns during development, Weiss notes.

Working with Escherichia coli, his team constructed three elaborate biological circuits that produce a glowing protein only if they receive a chemical signal in certain concentration ranges. Weiss and his collaborators used one group of engineered cells to release the chemical signal. A batch of the team's E. coli containing one of the circuits responds to a high concentration of signal, another to a medium amount, and a third to a low concentration. Each of the three batches fluoresces in a different color.

Weiss and his colleagues placed the signal-generating cells in the middle of a lab dish where they produced a concentration gradient that diminished toward the dish edges. The researchers next spread a mix of the other cells over the rest of the dish area. After several hours, the bull's-eye pattern appeared. The fluorescing bacteria formed glowing bands when they received their signature concentration of the signal chemical.

Drew Endy of the Massachusetts Institute of Technology (MIT) is using synthetic biology for basic science in a different way. He says that he's "rewriting" the genetic code of a bacteria-infecting virus known as T7. His team's aim is to convert the virus into a form that's convenient for scientists to study.

"Living systems themselves, as provided by nature, might not be optimized to be easy to understand and interact with and predict," Endy says.

To work out nature's kinks, he and his colleagues designed a new version of the T7 genome that separated genes that normally overlap--where one gene "takes up some of the same real estate" on the genome as another. The team then put buffers between the detangled genes so that each gene could be cleanly removed without affecting its neighbors, an advantage for future studies aimed at understanding each gene's function.

Although severely altered, the resulting virus, when placed in bacterial cells, behaved much as the original virus did. Further tests will be necessary to show whether the rewritten T7 is easier than its natural cousin for scientists to study and manipulate in the lab, says Endy.

CONSTRUCTING HEALTH Many scientists predict that synthetic biology will eventually yield new ways to treat deadly diseases. For example, Arkin and Chris Voigt of the University of California, San Francisco are assembling a collection of biological parts that might turn E. coli bacteria into cancer fighters.

E. coli normally inhabits the gut of many animals. But when it's injected into the bloodstream, it preferentially settles inside various cancerous tumors, where it doesn't disrupt growth. The researchers are now constructing a system to transform the bugs into chemotherapy-delivering machines.

The scientists are working with a strain of the bacterium that doesn't make people sick. First, they're building a biological circuit that's kicked off by two sensors: one that detects a tumor's inherent low-oxygen environment and another that detects large numbers of congregating bacteria. When both sensors in the bacteria respond, a situation that would exist only inside a tumor heavily infiltrated by the bacteria, the circuit would flip on a gene that makes the bacterial cell produce a rigid protein rod that juts from its surface. Many species of bacteria sport similar rods.

The cancer cells would engulf bacteria after their rods adhere to the cells' surfaces, as other cells typically take up other rod-carrying bacteria. As their work progresses, Voigt and his team plan to have the bacteria release a chemotherapeutic drug once additional sensors tell the E. coli that it's inside a cancer cell.

Jay Keasling and his colleagues at the University of California, Berkeley are applying synthetic biology to fight malaria. They're creating E. coli bacteria that pump out artemisinin, a potent antimalarial drug typically harvested from wormwood plants.

Since at least 150 B.C., people have boiled wormwood leaves to extract the drug. However, wormwood produces far too little artemisinin to meet the needs of the millions of people around the world who are infected with malaria but who have little access to other antimalarial drugs.

Researchers producing insulin-making bacteria decades ago simply slipped an extra gene into E. coli. This process won't work with artemisinin because in the plants, a complicated, 12-step process builds the chemical, Keasling notes. Instead, he and his colleagues needed to figure out which genes are responsible for each step of the process in wormwood and then assemble them in a form that could direct the bacterium.

So far, the researchers have constructed a biological circuit encoding 9 of the 12 steps. "We're missing genes in the metabolic pathway, but we anticipate we'll be able to find these genes very shortly and start producing the final product," Keasling says.

ETHICAL DESIGN As many new technologies do, synthetic biology comes loaded with ethical concerns. "It could have extraordinary use, but it could also be used for extraordinary harm," notes Laurie Zoloth, a bioethicist at Northwestern University in Evanston, Ill.

Describing a doomsday scenario, Zoloth points out that nefarious researchers could use synthetic biology's tools to incorporate genes from a killer bacterium or virus into a deadly new design. Alternatively, bumbling scientists could by accident create a harmful bug, which might wreak havoc if released from the lab.

Synthetic biologists and others have proposed a bevy of preventive measures to implement as this new field takes off. One such measure would be to require DNA-synthesis companies such as Codon Devices to notify a regulatory body whenever scientists order bits of DNA that have the potential to cause harm.

Furthermore, researchers, including Endy, have suggested that scientists be required to tag a few signature letters of non-coding DNA on each biological part to track where it came from. For example, in the catalog of "standard biological parts" that he and his colleagues maintain at MIT, each part carries that sort of DNA barcode. If a design could be tracked by a notation on its parts, scientists might feel more accountable for their work, Endy says.

One of the best protections against wrongdoings and mishaps, asserts Voigt, is to get students involved in meetings such as the annual intercollegiate Genetically Engineered Machine (iGEM) competition, which was inaugurated last November at MIT.

In line with other engineering competitions, in which students construct elaborate bridges and towers out of Popsicle sticks or build containers to protect eggs from high falls, iGEM is intended to stoke students' creative fires. At iGEM, there are no winners or losers--only students keen on wowing each other.

Competitions such as iGEM are "going to be increasingly important to educate young [synthetic-biology] scientists so that they understand the risks and have a set of tools to counter them," Voigt says. The competitors receive instruction on ethical aspects of the technique as well as learning tips and tricks for their future work.

IGEM is also about boosting student interest in the field. Last year, groups from seven schools competed with projects. For example, Jeff Tabor and other students at the University of Texas in Austin collaborated with Voigt to create a mat of bacteria that captures images as photographic film does. They published the details of their picture-taking bacteria in the Nov. 24, Nature.

This year, Tabor and his teammates pitted an updated version of their picture-taking microbes against synthetic-biology designs submitted by 12 other schools. "Being around competition, you can really see it pull in lots of young people to do these cool new things," Tabor says. "My competitors now will probably be my future colleagues."