Though researchers emphasize that it will be years before scientists can demonstrate the true potential of these techniques to engineer life, they're now using the experiments to increase the fundamental understanding of cell biology.

During a tour last week of the institute's Rockville, MD, facilities, where the experiments are taking place, scientists explained that the synthetic cell was created as a result of a project to learn how to make a cell with the minimum number of genes possible to live. "The hope is that by understanding the basic principles of cellular life, we will be able to make cells make more stuff," says John Glass, a professor at the institute. Cells designed to make a particular chemical will make it more efficiently if researchers can eliminate every other nonessential metabolic process. "I've always wanted to know how cells work, and now we have the tools," he says.

One way to figure out what the minimal genome is would be to start with a microbe that's easy to work with in the lab, such as yeast or E. coli, and knock out each gene one at a time. But this process takes a very long time. Instead, the Venter scientists start out with a genome sequence from a bacterium that already has a very small genome, alter it on the computer to add or delete genes, then synthesize it from chemicals, transplant it into a cell, and see how the changes affect the cell's function. Building a genome in this manner speeds up the institute's research on minimal genomes, says Daniel Gibson, associate professor at the Venter Institute.

The first step in making a synthetic cell is to design the genome. This process happens on the computer, with the sequence of 1,077,947 letters that make up the genome of the bacterium Mycoplasma mycoides. In their initial demonstration experiments, the researchers deleted 15 genes. And to create a watermark that distinguishes the synthetic genome from the natural one, they also made additions. The researchers encoded their names, a URL, a few quotes, and an e-mail address into the four-letter alphabet of DNA and added it to the genome.

It's expensive and time-consuming today to synthesize long pieces of DNA in the lab. So the researchers used a computer program to chop the genome up into 1,100 pieces, each about 1,080 base-pairs, or letters, long. The computer program added sticky sequences at either end of each slice that would enable the pieces to be put back together again. Researchers then sent these 1,100 designs to a DNA-synthesis company.

Institute researchers then enlist yeast cells to stitch the 1,100 fragments into a single, circular piece of DNA that makes up the completed synthetic genome. Before the yeast can do their work, Gibson's group must make the DNA fragments yeast-friendly. Gibson's group first adds to each set of DNA fragments a short sequence of DNA that pulls the fragments into a loop and makes the fragments friendly to yeast cells that have been treated to make them amenable to gobbling up DNA.

Gibson combines the yeast cells in solution with ten types of DNA fragments, each of which make up a consecutive sequence of the Mycoplasma genome. The yeast cells do the work of putting the fragments back together. This process is repeated until the yeast are putting together larger and larger pieces of the genome. Eventually, some of the yeast cells will have put together a complete synthetic genome. After testing to verify that a colony has the entire bacterial genome, the researchers grow the yeast in a flask to allow them to multiply and produce the genome in large quantities.