Tony Tye, Post-Gazette Garth D. Ehrlich, left, executive director of the Allegheny-Singer Research Center at Allegheny General Hospital, senior research assistant Benjamin Janto and Dr. J. Christopher Post, medical director of the Center for Genomic Sciences with the 454 Genome Sequencer. Click photo for larger image.

There was a time, not so very long ago, when a researcher who wanted to know the genetic sequence of a microbe would have to invest months, if not years, and maybe a million bucks to find out.

Now, researchers at Allegheny General Hospital can unravel the entire genetic code of a bacterium in a matter of hours at a cost of $5,000 to $20,000 for various reagents.

This is possible thanks to a new generation of gene sequencing machinery developed by a Carnegie Mellon University graduate that promises to rapidly change biomedical research -- and ultimately medical practice.

"It allows you to ask questions you couldn't before," said J. Christopher Post, medical director of the Center for Genomic Sciences at the Allegheny-Singer Research Center, which earlier this year became one of the first groups in the country to install the $500,000 sequencer.

And the machine, made by 454 Life Sciences Inc., may soon usher in an era when a human patient's genome can be sequenced for about the cost of an MRI scan. Garth D. Ehrlich, the center's executive director, predicts AGH will be sequencing human genomes within two or three years.

That could help doctors realize the dream of truly personalized medicine -- using a person's unique collection of genetic mutations to guide both therapy and preventive health strategies.

Jonathan M. Rothberg, founder of 454 Life Sciences and developer of the new gene sequencer.

In the meantime, the technology is helping Drs. Ehrlich and Post develop a better understanding of how bacterial infections can sometimes sidestep antibiotics. And they also are using it to analyze the bugs that help cows digest grass in hopes of identifying bacterial strains that could improve the production of biofuels.

It was the human genome -- specifically his newborn son's -- that Jonathan M. Rothberg had in mind in 1999 when he first started pondering how to improve genetic sequencing.

His son developed breathing problems after birth and was moved to a newborn intensive care unit. Dr. Rothberg, who earned a degree in chemical engineering at Carnegie Mellon and a doctorate in biology at Yale University, worried that he might have cystic fibrosis, which is caused by a genetic mutation. And that in turn made him wonder why it cost so much to sequence the human genome.

His son recovered and is fine now, but the sequencing idea stuck. "Genes guide everything, from conception to death," said Dr. Rothberg, who had previously founded a company called CuraGen Corp. to use knowledge of the human genome to guide drug development.

The international Human Genome Project had spent upward of $2 billion and more than a decade to sequence the first human genome, he noted. The cost has since come down to about $1 million, but the process is so inherently complex that further cost reductions are unlikely, he said.

The decoding process has remained pretty much the same for 30 years, requiring bits of DNA to be grown in colonies of bacteria, extracted by robots and then processed through expensive machinery.

He envisioned a system in which samples of DNA could be easily prepared for sequencing and a sequencing machine that, like electronic computer chips, could be made increasingly efficient, small and cheap.

Base pair sequences

DNA carries genetic code in the form of a long sequence of four chemical bases, which act much like an alphabet. Genetic sequencing determines the order of those bases.

Dr. Rothberg's system involves isolating double-stranded DNA from a cell, breaking it into little, single-stranded bits and loading them into tiny wells on a credit-card-sized plate. Each plate contains 1.6 million wells, though not all get filled.

Sequencing reagents, each containing one of the four bases that comprise DNA, are then washed over the plate, with bases attaching themselves to DNA bits as necessary to rebuild the double-strands. As each base attaches to the strand, it triggers luciferase, the enzyme that allows fireflies to glow.

These tiny flashes of light are detected using a charge-coupled device, or CCD, a sensitive light detector similar to those used in astronomical telescopes. A computer keeps track of which bases are attaching themselves in which wells; once it knows the sequence of bases in each tiny well, the computer can reconstruct the entire DNA sequence.

Last summer, he and his colleagues reported the process in the journal Nature.

Soon after, a group at Harvard University reported their own quick, cheap sequencing technique in the journal Science. Though similar in some ways to the 454 method, the Harvard technique doesn't require a special machine, but can be performed using a computer-controlled digital microscope already available in some labs.

By increasing the length of the tiny bits of DNA in each well and by increasing the number of wells that get filled with bits of DNA, Dr. Rothberg said the method can become increasingly fast and efficient.

Dr. Rothberg noted that his company received a three-year, $7.5 million grant from the National Human Genome Research Institute, which expires next year, to reduce the cost of sequencing a human genome to under $100,000.

"We're well on track to do that. We can see the $10,000 genome," he added and a $1,000 human genome looks all but inevitable. At those costs, sequencing an individual's genome costs about the same as an MRI scan and could become routine, he said.

"We believe sequencing will be as big a market as MRI scanning," he added.

For now, sequencing of pathogens, such as the Haemophilus influenzae, Streptococcus pneumoniae and Pseudomonas aeruginosa studied by Drs. Ehrlich and Post, are the bread-and-butter of this new sequencing technology.

The Allegheny-Singer researchers have long explored why bacteria, such as those that cause ear infections, can sometimes resist antibiotics. One factor, they know, is that each infection can involve a variety of strains of bacteria -- genetically distinct forms of the same bacterium. And they know that bacteria can group into communities called biofilms; in this form, bacteria are more resistant to antibiotics.

What is becoming apparent now that they are able to sequence each strain is that each strain's genome -- all of the genes found in a bacterial species -- is larger than the genome of any individual bacterium.

That is, a Haemophilus influenzae bacterium -- which is involved in a variety of respiratory diseases -- might have between 1,500 and 1,600 genes, Dr. Ehrlich said. But when all of the different genes in the various strains are totalled, the researchers find H. flu genome contains about 3,300 genes.

It may be that the bacteria can rapidly develop resistance to antibiotics by swapping genes between themselves in the biofilm. This mixing and matching of genes could be a common strategy of bacteria, they suspect.

It's an idea they can test by infecting lab animals with strains of bacteria whose genomes already are known from sequencing. By removing samples of bacteria from those animals at a later time, they can repeat the DNA sequencing to see if new strains with new sets of genes have evolved.

"With this new technology, we have leapfrogged every other lab looking at this," Dr. Post said. "If you don't have this kind of stuff in your lab these days, you're no longer part of the game."