Sometime in February or March, officials at the Justice Department held a closed-door meeting with a federal judge in Washington, where they laid out what the government knows about the anthrax-letter attacks of 2001.
Nobody is exactly sure what the department told the judge during that meeting. People familiar with the presentation say it was held under top-secret conditions; documents were escorted to the courtroom under the supervision of the U.S. Marshals Service, and the judge was not even allowed to keep copies of the papers that were shown to him. But at a hearing on March 29, the judge, Reggie B. Walton of the U.S. District Court for the District of Columbia, gave a hint of what he'd seen. The two-and-a-half-year-old investigation into the anthrax letters is at a "critical" stage, Walton said. Based on the evidence he'd been shown, a breakthrough in the case might be just around the corner.
Walton is presiding over the defamation lawsuit filed against Attorney General John Ashcroft by Steven Hatfill, a scientist once named by the department as a possible suspect in the attacks. The government presented its evidence to Walton in an effort to persuade him to delay the Hatfill case for six months; after seeing the documents, Walton granted the government's request, saying that he wanted the investigation to proceed without interference from Hatfill's civil suit.
Many close observers of the anthrax investigation dismissed Walton's pronouncements about the case -- the government has been periodically advertising impending breakthroughs in the case since just about the time the spore-ridden letters were mailed out, some critics of the Justice Department pointed out. But a few scientists believe that this time, the government may be telling the truth. That's because, for the last two years, the FBI has been building a more formal, standardized scientific method for dealing with the investigation and prosecution of bioterrorism -- and this work might finally be paying off in the anthrax investigation.
The new science, which the FBI calls "microbial forensics," aims to improve the way law enforcement officials and forensic labs across the nation deal with crimes involving microbes -- whether these are bacteria, like anthrax, or viruses like HIV or smallpox. Microbial forensics is a complex, multidisciplinary effort that has two main goals. The first is to build new forensic techniques to help investigators quickly track down bioterrorists. By examining both the genetic and atomic structures of different strains of dangerous microbes, researchers are trying to build something like a "fingerprint" for microbes -- a method of tying a suspect to a biological crime in the same way that a fingerprint ties a thief to the scene of a robbery.
The other part of microbial forensics has to do with perception; the FBI, which has established a working group of government and academic researchers to help develop the science, wants to make sure that microbial forensics will stand up in court. Law enforcement officials are often wary of bringing new scientific methods before a jury: While scientists can often use advanced forensic tools to unearth hints pointing to the likely cause of a bioterrorist attack, some of those clues don't provide for the kind of open-and-shut cases that prosecutors appreciate, especially in high-stakes cases of bioterrorism.
Because the FBI declined to comment on the science, it's not clear how microbial forensics is being used in today's anthrax investigation. It was certainly inspired by the anthrax attacks, though, and many researchers say it will become very useful in future bioterrorism cases. But microbial forensics will also help in areas beyond bioterrorism, says Paul Keim, a biologist at Northern Arizona University who is a member of the FBI's working group and a world expert on anthrax. "Its application might be greatest in what we call 'biocrimes,'" he points out. "For example, the deliberate infection of someone with HIV, or having unprotected sex when you know you're HIV-positive, which is a crime. Terrorism or crime, there could be hundreds of cases where we see people using this in court."
When forensic researchers talk about the perils of bringing new scientific methods into the courtroom, the name that occurs to many of them is O.J. Simpson. DNA identification took center stage during the football star's infamous murder trial, but as we all remember, nobody could agree on what the DNA evidence meant. Blood matching Simpson's and the victims' DNA was found at the crime scene, in his Bronco, and on that famous bloody glove. Prosecutors maintained that these samples proved Simpson's guilt, but defense attorneys charged that detectives and law enforcement forensic scientists had, sometimes mistakenly and sometimes intentionally, seriously mishandled the blood evidence.
The outcome of the Simpson case served as a wake-up call for those scientists who believed that juries would have no choice but to accept the conclusions of a forensic method that, to the scientists, seemed so obviously accurate. Instead, though, the Simpson jury apparently believed the defense's theories of DNA evidence -- that it could be "contaminated" by poor practices, and that, by itself, DNA didn't matter if you were dealing with a police force thought to be corrupt and racist.
The Simpson case could have dealt a deathblow to DNA forensics. But in the aftermath of the case, says Paul Keim, "crime labs went to great lengths to standardize and protect their evidence-collecting techniques and to institute quality assurance and quality control." The national effort was spearheaded by the FBI, which developed a set of guidelines for DNA crime labs to follow in order to present DNA evidence that would appear incontrovertible to juries. And the effort worked, Keim says. "The reason why DNA fingerprinting of blood and semen is such a great and useful tool today is because the FBI has been so great in assuring the quality."
The FBI scientist who was at the forefront of the effort to maintain the quality standards of DNA evidence is a well-regarded geneticist named Bruce Budowle. Budowle, who did not return calls for comment, is now the chairman of the FBI panel working on microbial forensics; the aim of this panel, Budowle and other members have said, is to develop a set of guidelines that will allow scientists to present microbial evidence that can stand up to defense tactics in court. In other words, they want to prevent what happened in the Simpson case from occurring in the anthrax case.
Members of the FBI first met with prominent scientists to work on microbial forensics in Burlington, Vt., one weekend during the summer of 2002. Abigail Salyers, a microbiologist at the University of Illinois at Urbana-Champaign who was at time the president of the American Society for Microbiology, recalls the meeting as a productive melding of the best minds in law enforcement with the best scientists in academia.
"It's the first time we ever had the FBI present at one of our scientific meetings," she says, "and they kept asking the kind of questions that gave people a reality check. They were able to point out to us the limitations they underwent in the process of doing this work -- the main limitation of not knowing how you deal with a bioterrorist crime scene like this. This isn't like going to a murder scene where you have a book telling you how to collect samples and look for bullets and things like that. Here, there are cases where you won't know what you are looking for. They have to make up the book about what they were going to do."
Since then, the working group has been creating that book, and it has published portions of it in respected scientific journals. In September, the panel published in the journal Science a strict set of quality guidelines that laboratories looking to perform microbial forensics must follow.
"When the anthrax incident occurred, much of the investigative work was referred to laboratories that were external of the FBI or the government," says Joseph Campos, the director of the microbiology lab at the Children's National Medical Center in Washington and a member of the FBI panel. "There was a concern that those laboratories were not following procedures that were as standardized as they might be." With the new guidelines in place, Campos says, work done in different labs around the nation will result in similar findings.
In addition to developing rules for handling biological evidence at crime scenes and in labs, the other main aim of microbial forensics is to build a database of known microbial threats, and to analyze those microbes genetically. This is a daunting task; there are dozens of known microbes that could be used in crimes and in terrorist attacks, and there are thousands (or more) of known genetic strains of many of those agents. But the FBI has an innovative plan to build this database -- outsourcing to non-governmental labs. While the government will maintain the main microbial database at a new bioforensics lab at Fort Detrick, in Frederick, Md., the lab will merely lie at the center of a large network of governmental and non-governmental laboratories, all sharing knowledge of the latest threats, members of the FBI microbial forensics panel say.
This setup is a departure for the FBI. Historically, the bureau has not been very enthusiastic about working with scientists from outside the government. But all that changed with the investigation of the anthrax attack; the agency has acknowledged that it was unprepared to deal with the complexity of that investigation, and shortly after the attack it quickly contacted outsiders to help the with the sophisticated genetic analyses needed to solve the case. "This may well be standard practice in future cases," Budowle wrote in Science in September.
Paul Keim, the anthrax expert, was one of the scientists that the FBI contacted in the anthrax letters investigation. Keim is prohibited by a nondisclosure agreement he signed with the FBI from talking about the case, but he has acknowledged that it was his team that pinned down the strain of the anthrax bacterium found at the first attack site, the American Media headquarters in Boca Raton, Fla. Keim determined that the anthrax used was of the "Ames strain" (which, curiously, is not from Ames, Iowa but instead from a cow in Texas.)
What does Keim's conclusion about the strain of anthrax used in the attacks actually prove? This question goes to the heart of the most difficult issue in microbial forensics. Human beings have unique DNA fingerprints, making it possible for blood found at a crime scene to be matched to only one suspect. Viruses and bacteria, though, can be genetically cloned. This leads to the possibility that a single genetic strain of anthrax, which is a naturally occurring bacterium, could be literally everywhere; and if law enforcement officials ever find a suspect in the anthrax letters case, they might have trouble convicting him even if they can prove that the suspect's belongings -- his house, or his car, perhaps -- are covered in anthrax spores of the Ames strain.
The chance of conviction, Keim says, would depend on how widely the Ames strain was disseminated. "If you have a situation where there's one lab in the world that has the Ames strain," Keim says, then you'd have a pretty good clue as to where the agents used in the attack came from. But if many different labs have identical versions of the Ames strain, it'll be more difficult to prove its origin, especially if you're facing a defense attorney who'll call such a situation a "reasonable doubt." (Keim says that that FBI has not given him any clue as to how common the Ames strain found in the first anthrax letter actually is.)
For anthrax, the difficulty in determining the origin of an attack is compounded by the organism's slow mutation rate, Keim says. "Anthrax is one of the slowest-evolving organisms on earth," he says. "Almost no changes occur when it's in spore form. When it's active, we know what the mutation rate is -- we can actually predict that it takes about 300 generations before a mutation will occur. In 300 generations, you can make a whole lot of anthrax." What this means, Keim explains, is that attackers who pick a common strain of anthrax can produce weapons from them with a pretty solid assurance that the weaponized anthrax will have an identical genetic sequence to the common strain. The weaponized anthrax will thus be anonymous, untraceable to a specific attacker.
Because of the difficulties Keim points out, some scientists are working on nongenetic methods to determine the origins of microbes used in attacks. For example, in a paper published in the Proceedings of the National Academy of Sciences in February 2003, a team at the University of Utah suggested a way to pinpoint where a microbe like anthrax was cultured by examining its atomic structure and measuring the level of one oxygen isotope -- oxygen-18 -- present in the bacteria.
The team predicted that the level of oxygen-18 in the bacteria would be similar to the level of oxygen-18 in the water in which the bacteria was cultured. And because the level of oxygen-18 in water is known to vary with geography -- seawater has more of it than fresh water, for instance -- counting oxygen-18 can give you a pretty good idea of where a microbe was grown, they determined. The group was able to verify its hypothesis using a blind test. They asked friends in North Carolina, Ohio, Louisiana and New Mexico to ship them spores of anthrax grown in those areas, and the team added its own spores grown in Salt Lake City. After analyzing the oxygen-18 levels, they were able to correctly guess the origin of three of the five microbes. (Because the microbes in North Carolina and Ohio were grown in water with identical levels of oxygen-18, the group could not say which was which.)
If one hurdle to investigating anthrax is its slow mutation rate, the problem scientists encounter when looking at HIV is just the opposite -- the virus changes extremely rapidly, making it difficult to say how one strain of the virus is related to another. This problem came up in the 1998 case of Richard J. Schmidt, a Louisiana doctor who was accused of drawing blood from a patient he knew was HIV-positive and injecting that blood into the arm of his former girlfriend, infecting her with the virus (she believed she was just receiving a vitamin shot).
When forensic scientists compared the genetic makeup of Schmidt's ex-girlfriend's strain of HIV with that of his patient, they could not find a perfect match; the HIV virus changes too quickly to allow such a thing. But they did find that the two had strains of the virus that were significantly more similar to each other than to others in the area. The prosecution argued that this proved Schmidt's guilt, but the defense called several experts who dismissed the evidence. "It was very contentious," says Keim, who was not involved in the case but considers it an important landmark in microbial forensics. "There were excellent experts on both the defense team and the prosecution. And one of the defense experts was claiming that because the prosecution was using the lab at a university you couldn't trust this. That shows you why having a standardized lab doing this analysis is important."
Still, despite the expert defense witnesses, the jury convicted Schmidt of attempted murder, and several appellate courts upheld the conviction. The trial provided the first U.S. courtroom victory for this kind of HIV analysis, and Keim believes that it set a precedent for future inquiries; he says we should expect many criminal and civil cases involving HIV -- for instance, wrongful death suits aimed at a sexual partner suspected of giving you the disease.
"Once you get one court to accept it," Keim says of the HIV test specifically but also microbial forensics in general, "then more and more courts are likely to accept it in the future, and we'll see a lot more of this in use."