Tucked away in the Smithsonian Institution's $624 million budget that Congress passed last year was a line item for $700,000 that right now is transforming Lee Weigt's laboratory at the institution's in-house research and supportcenter in Suitland, Md., into what could become a model for a global network of high-volume bar-coding facilities.
The process of determining the bar codes of millions of species and registering the information in databases only lays the groundwork. After that, users such as agricultural inspectors monitoring for pests, wildlife management officials sleuthing for poachers, and public health officials tracking mosquito movements presumably will be generating a relentless flow of samples for which they need bar codes. In each case, a 650-nucleotide-long stretch of the specimen's DNA will have to be extracted, amplified (with polymerase chain reaction, or PCR, methods), and sequenced.
"I'm now installing a new machine that can [get the DNA from] 400 samples in four hours with push-button, walk-away automation," Weigt says. Using a gentle mechanical and enzymatic digestion process on tissue specimens that are the size of pencil tips, the machine will free up the specimens' DNA within the first day of a bar-coding procedure. On the second day, Weigt says, the 400 little vials of DNA emerging from the first robot will feed into a PCR robot that will selectively amplify the key identifying segment of the CO1 gene (as described on page 28).
The selectivity derives from the PCR primers, which in this case are carefully designed strings of nucleotides with sequences that are complementary to those DNA regions just before and after the bar-code segment of the CO1 gene. "These primers tag the right spot among the entire encyclopedia of genes in the sample," Weigt explains. Once the sample DNA is so primed, this PCR robot amplifies the marked DNA a million-fold. That provides enough copies of the bar-code region for the third robot, a DNA sequencer that relies on four fluorescent labels corresponding to the four types of nucleotides that string together like letters into genetic sequences.
If the entire process is begun on a Monday morning, Weigt estimates, it ought to be completed by Wednesday midday. So if he keeps all of the machines running all of the time, the laboratory ought to be able to produce 2,000 or more bar codes each week, maybe even with aday off. The total count of genetic bar codes on hand in the world now is under 100,000, Weigt estimates, but "we're on our way to millions."