A Trained Eye Finally Solved the Anthrax Puzzle

[Source New York Times, By NICHOLAS WADE] - When the Federal Bureau of Investigation announced it had cracked the long-unsolved anthrax case, the turning point cited by the bureau was its identification of a laboratory flask as the source of the anthrax.

The dots, or in this case more than a thousand separate anthrax samples, were connected with the help of a group of scientists working secretly for some seven years. They succeeded by using a combination of new techniques not even invented in late 2001 when the anthrax-laced letters were sent, and that most old-fashioned attribute of expert scientists and detectives: a trained eye.

Now, in their first interviews, after being released this week from their vows of silence, several scientists explained how they charted a new frontier in microbial forensics, one that could have the same evidentiary power as DNA fingerprinting in criminal cases.

The scientists say they are confident the F.B.I. has identified the source of the anthrax, a flask in the custody of Bruce E. Ivins, whom the F.B.I. considers to have been the perpetrator of the attacks. But almost a hundred other people were known to have had access to cultures from the flask, and the scientists say they have no opinion as to whether Dr. Ivins, who committed suicide last month, was the culprit. Some former colleagues and other experts have questioned whether the government was right in suspecting Dr. Ivins, a researcher at the Army Medical Institute of Infectious Diseases in Fort Detrick, Md. But the technical feat of matching the attack anthrax to its source is itself a gripping tale of scientific detection.

The scientific chase began in late 2001 as the first person to contract anthrax from powder in a letter lay dying in a Florida hospital. The victim, Robert Stevens, 63, a photo editor at The Sun, a tabloid, was suffering from pulmonary anthrax, and the F.B.I. needed to know whether the anthrax in the attacks, which began a week after Sept. 11, was natural, or a biological weapon.
A sample of anthrax from Mr. Stevens’s body was flown to Paul Keim, a biologist at Northern Arizona University who two years earlier had developed a test for distinguishing the various strains of anthrax found in nature and in biological weapons laboratories. Of the anthrax strains used as weapons, the most virulent was one known as the Ames strain.

Dr. Keim confirmed the fears of intelligence agencies: it was the Ames strain that infected and eventually killed Mr. Stevens.

Dr. Keim’s test could tell two strains of anthrax apart but it could not tell the bureau what it needed to know next, which of the many cultures of Ames anthrax around the world the attack anthrax might have come from.

The F.B.I. decided to go back to basics and to try decoding the entire DNA sequence — some five million units — of the anthrax genome to see if some clues to its source might be developed. For this job it turned to the Institute for Genomic Research or TIGR, a leader in decoding the genomes of microbes. Its director was then Claire Fraser-Liggett, who is now at the University of Maryland. The F.B.I. asked her to form a group, with as few people as possible, to decode an anthrax genome, without telling her it was the one that had killed Mr. Stevens.

In contrast to the way science is usually done, the research overseen by the F.B.I., which took seven years to finish, was highly compartmentalized. The scientists, who work at academic institutions, say they did not understand important details of the case until a news conference on Monday — which they and their scientific directors in the F.B.I. attended.

For the bureau, the compartmentalization was an essential safeguard against the nightmare that one of their many advisers might turn out to have prepared the attack anthrax. “It may have been in people’s minds that someone in the room could have been one of the perpetrators, which ended up being the case,” said Dr. Chris Hassell, the F.B.I. laboratory director.

By early 2002, the TIGR team had completed the genome and were able to compare it with a culture of Ames anthrax maintained at Porton Down, the British biological weapons establishment. Anthrax is a highly stable organism that changes very little from one generation to another. But the scientists found several differences between the Stevens and Porton Down genomes, raising the possibility that the source of the attack strain might be distinguishable from other cultures of Ames anthrax.

“The finding was very good news for the investigation by giving hope that molecular forensics might bear fruit but, if so, large numbers of samples would need to be analyzed,” Dr. Fraser-Liggett said.

All Ames anthrax is derived from a cow that died in Texas in April 1981. The F.B.I. acquired a sample from Fort Detrick of the original strain, known as the Ames ancestor. Decoding began on that, with the idea of constructing a genealogy that would show the Ames ancestor begat the source culture which begat the attack strain.

The TIGR team decoded the Ames ancestor and then turned to decoding the anthrax from one of the attack mailings. Each decoding took three to four months and cost about $250,000, said Jacques Ravel, a leading member of the TIGR team who is also now at the University of Maryland.

But when the decoding of the attack genome was finished in 2002, the TIGR scientists had a major surprise and disappointment. In virtually all of its five billion units, the attack anthrax was identical to the Ames ancestor. There were no differences that could tie the attack strain to any of the many known cultures of Ames held in laboratories around the world.

At the regular meetings the TIGR scientists held with the F.B.I., they received very little information or feedback. But they could tell that their counterpart scientists in the bureau were as discouraged as they were, Dr. Fraser-Liggett said. Anthrax genomes looked like a dead end.
Then an Army microbiologist from Fort Detrick made an unexpected discovery. Using an old-fashioned microbiological technique, he spread out some attack spores on a bed of nutrient and let each form its own colony. All the colonies looked identical except one, which, to his trained eye, seemed very slightly different. Different-looking colonies are called morphotypes or just “morphs.”

“Had that task been assigned to someone less experienced, these morphotypes might never have been seen or their significance never realized,” Dr. Fraser-Liggett said.

Because of the obvious possibility that the morph might look different because its genome was different, the F.B.I. asked the TIGR team to decode its genome. Four months later, the TIGR scientists were elated when they discovered the morph had a major genetic change in its genome, known as an indel, short for insertion or deletion of DNA. “We were extremely excited,” Dr. Fraser-Liggett said.

With the morph, the attack strain was at last developing a genetic signature of its own. Though 99 percent of its spores were identical with the Ames ancestor, some 1 percent or less were morphs.

Dr. Ravel was asked to decode seven more morph genomes, a task that took two years. He could do only one at a time for fear of cross-contamination in his laboratory. Dr. Fraser-Liggett said she did not know why the F.B.I. did not ask other laboratories to share the task and speed up the critical process.

One of the many mysteries the TIGR team had to live with under the bureau’s management was the puzzle of why the attack spores contained as many morphs as they did. At the news conference they learned why, when an F.B.I. scientist explained that the flask in Dr. Ivins’s custody, known as RMR-1029, held the product of 13 production runs of anthrax made at the Army’s Dugway Proving Ground and 22 spore preparations made at Fort Detrick. Some 160 liters of material, the scientist said, had been concentrated into the liter held in the RMR-1029 flask.

The vast number of spores, and the many different culturing procedures, Dr. Keim said, “guarantees you will see these mutants, and when you mix them together you will have a characteristic signature.”

Other scientists chosen by the F.B.I. selected four of the morphs as having the most reliable indels. All the attack letters contained these four morphs as well as the predominant form of Ames ancestor-type spores. The bureau at last had a signature of the attack strain.

Hoping for just this breakthrough, the bureau had been building a repository of Ames anthrax samples, taken under subpoena from laboratories around the world. As the morphs became available, the F.B.I. started testing samples. At first, some had one or two of the morphs. None had three of the morphs.

By late 2005 to 2006 it became clear that just eight of the 1,070 samples collected included all four morphs. And one of the samples was the ancestor of the other seven. The seven samples came from Fort Detrick and one other laboratory in the United States, F.B.I. scientists said at the Monday news conference, held at F.B.I. headquarters.

The source of the seven was a master flask of Ames anthrax known as RMR-1029 which was kept by Dr. Bruce Ivins. “That’s when the genetics caught up with the investigators,” a Department of Justice prosecutor said.

There, the scientific conclusions end. The bureau then began a second phase of the inquiry, that of ascertaining who had access to the flask and its seven descendants. The F.B.I. investigated almost 100 scientists who had had access to cultures from the flask or were in some way associated with them.

At the news conference, it emerged that Dr. Ivins had in fact submitted two samples of RMR-1029, one in February 2002 and a second in April 2002. The second tested negative. The F.B.I. rejected and destroyed the first sample because it had not been prepared according to a strict protocol that the F.B.I. says Dr. Ivins helped in devising.

A duplicate of the first sample was later located in Dr. Keim’s laboratory, where all duplicates were sent, and tested positive. Asked why Dr. Ivins would submit a true sample of his flask in February but a false one in April, the F.BI. scientists said they could not speculate about his motives.

Dr. Keim said he believed the bureau had correctly identified the source of the attack anthrax. “The science on that is pretty solid,” he said. As to whether Dr. Ivins was the perpetrator, Dr. Keim said that only a jury could make that decision. He said Dr. Ivins had been a friend and he faulted the F.B.I. for not having prevented his suicide. “Whether Bruce did it or not I prefer not to think about,” he said.

Dr. Fraser-Liggett said, “I am absolutely convinced the F.B.I. has the right source flask,” but added that she had no opinion as to who the perpetrator might be.
Scott Shane contributed reporting from Washington.