What is microbial forensics?
Stained photomicrograph of Bacillus anthracis (anthrax) is a spore-forming, rod-shaped bacterium. Source: CDC.
You have probably heard of commonly used forensic methods such as the analysis of striations on bullets to identify the gun used to commit a crime. But what if a microbe is the weapon of choice, as can occur if a bioterrorist comes to town? Microbes as weapons is not a new topic. There have been reported cases, for example of HIV-infected people intentionally infecting others.
Moreover, microbes can be involved in cases of medical negligence in which a surgeon or nurse causes a patient to contract a post-surgical or other hospital-acquired infection due to inadequate hygiene. There is currently a lawsuit making its way through the Scottish courts against a hospital alleging that inadequate hygiene resulted in the death of a patient.1 It is also conceivable that outbreaks of foodborne disease could spawn lawsuits alleging either negligence or intentional contamination. Tracing the infecting microbe to the company and person(s) of origin will be critical in such cases.
The anthrax attacks in 2001 may have turned the spotlight once again on the necessity for having reliable and validated procedures for tracing microbial “suspects” and identifying their “hideout,” but the need for such procedures is much more general. How can one “fingerprint” a microbe, which does not have anything resembling a finger?
The answer currently in favor is to use DNA sequences that identify a microbe, not just as a member of a species, which can include a variety of different individual microbes, but as an individual microbe that can be traced to a particular person or a piece of equipment.
It is also possible to use carbohydrates or proteins found on the surface of a microbe as a fingerprint, but unlike human fingerprints, this type of fingerprint can change. So too can a microbe’s DNA sequence but most scientists agree that DNA sequence information carries the most promise for reliable identification of a microbe.
Given the increasing amount of microbial genome sequence data that is becoming available, it is possible to identify multiple sequences to make up a fingerprint, so that a change in one sequence can be detected if the other sequences remain the same.
Microbial forensics is the term that is applied to this new type of forensic analysis. Molecular techniques have been used for years to trace outbreaks of microbial diseases, a practice called molecular epidemiology. In fact, there are currently surveillance systems that store and make available DNA fingerprints for microbes that are likely to be involved in hospital-acquired infections and foodborne infections (e.g., PulseNet of the U.S. Centers for Disease Control [CDC], a surveillance system for tracking infections such as Salmonella).2 Although these surveillance systems are still only a few years old, they are rapidly growing in sophistication.
What distinguishes microbial forensics from molecular epidemiology is that microbial forensic data must hold up not only to the scrutiny of scientists in the health care community, but also to the scrutiny of judges and juries. Nonetheless, work done to date on microbial epidemiology will provide an invaluable starting point for the additional work that needs to be done to make microbial forensics ready for its day in court.
The bioterrorism connection
A good example of the problems that need to be solved is provided by the response to the anthrax attack of 2001, in which spores of the bacterium Bacillus anthracis, the cause of anthrax, were disseminated via the mail.3 The effects of this crime extended far beyond the deaths of the 5 people who died of inhalation anthrax.
Hundreds of people who were exposed to the spores experienced anxiety about developing the disease and suffered the physical discomfort associated with side effects of the long course of antibiotics prescribed to prevent them from incurring the disease.
Beyond that, the attack frightened millions of people, from postal workers to ordinary citizens, who feared that they might be exposed to the spores as a result of something as seemingly harmless as opening their mail.
If a suspect is ever arrested and charged with perpetrating the anthrax attack, spores will be an important part of the physical evidence.
- Prosecutors will have to prove that spores isolated from the suspect’s home or laboratory are in fact THE spores, the ones introduced into the envelopes and mailed.
- The problem with making such a case is that spores of B. anthracis are found widely in soil, especially farm soil in the southern U.S. So the prosecution will have to prove that any spores submitted as evidence were the spores used in the attack and not simply spores that had been tracked into a house or laboratory from a nearby field.
Proving the assertion that the spores introduced as evidence were the ones used to contaminate the envelopes used in the anthrax bioattack may not be as easy as a laboratory scientist, who is familiar with DNA-based molecular epidemiology methods, might think.
- Spores of different strains of B. anthracis and of the vegetative (actively dividing) form of the bacteria, unlike diatoms, look very much alike.
- In fact, different strains of B. anthracis are also very similar to each other at the genome sequence level.
- Given that the error rate for DNA sequencing is not zero, proving that a particular isolate of B. anthracis is the same as the strain used in the bioattack may prove to be difficult.
So, if the assertion to be proven is that the spores found in the home or laboratory of a suspect are the same as those mailed in the envelopes, questions about the meaning of slight variations in test results will have to be answered.
Scientists who are accustomed to using such techniques as DNA sequencing or other types of DNA-based tests that might be used to differentiate strains of B. anthracis are comfortable with their ability to interpret the results of such tests despite any inherent variation that might exist. Such limitations, however, when brought into a trial by expert witnesses hired by the defense or prosecution attorneys, might well evoke uncertainty in the minds of a judge and jury, who are not as familiar with these widely used methods of analysis.
All of us remember the debacle of the O. J. Simpson trial during which lawyers were trying to establish the validity and limitations of a DNA-based analytical method used for identifying the human source of a blood sample, and few would argue that an actual trial is the place to establish whether a scientific test is reliable.4 Now that years have passed since the Simpson trial and the tests about which there was so much argument at the time have been used in many courts cases, the public and the legal profession have become more comfortable with the use of such tests for identifying murderers and rapists. Reaching the same level of comfort with microbial forensics will not be nearly as easy.
This is so because one is not dealing with a single familiar species, Homo sapiens, but with a huge diversity of microbial species with names that are unfamiliar, in some cases even to the average microbiologists. In contrast to DNA-based tests now widely used to identify human suspects, a single set of tests and interpretations will not work for all microbial species. Moreover, although these tests are widely used by scientists, they have not been validated in a way that would give a nonscientist confidence that they are reliable enough to send someone to jail.
The beginning of a solution — the scientific community responds
The response of the scientific community to the challenge of the hoped-for anthrax court cases provides a good illustration of how scientists proceed in such cases.
- One thing was very clear from the outset — that too little attention had been paid previously to microbial forensics.
- This realization caused scientists to start back at the very beginning, by meeting to identify what information is available, what research remains to be done, and how to proceed as expeditiously as possible.
- Paul Keim, a leader in the early development of DNA-based methods for identifying individual strains of B. anthracis, set the process in action when he approached the American Academy of Microbiology (AAM) about convening a meeting of experts to consider the subject.5
The American Academy of Microbiology is an organization that has a long history of assembling small groups of experts and challenging them to define the future needs for work in a new area of microbiology and was thus the obvious organization to spearhead such an effort. A group of 35 scientists with expertise that might be able to help answer the question of what needs to be done to validate tests that could be put to forensic use was identified by the AAM and met in June 7-9, 2002, in a bucolic setting in Burlington, Vermont. The author of this paper was one of the attendees.
For the first time in the history of these AAM-organized meetings, three scientists from the FBI were included. The FBI scientists, all of whom had had direct involvement in investigation of the anthrax case, helped provide the occasional reality check, as other scientists not familiar with work in the field grappled with the question of how to establish standards for evidence collection and for analysis and interpretation of the plethora of new molecular tests, more of which are being published every month. The anthrax attack was not the only example of the possible use of microbial forensics considered by the group of AAM experts. Other examples included intentional contamination of others by HIV-positive individuals and outbreaks of hospital-acquired or foodborne disease. Understandably, however, the anthrax bioattack dominated the discussion.
“Microbial Forensics: A Scientific Assessment,” the report of the conclusions reached by this group of experts, has now been published by the American Academy of Microbiology.6 (See “references” at the end of this paper for information on how to obtain a copy of this report.)
Identification of key challenges and recommendations for finding solutions
This report listed a number of recommendations on how to prepare for possible bioattack court cases or for responding to future bioattacks. No attempt will be made here to discuss every one of these recommendations. Rather, this paper seeks to highlight some of the major challenges (see Table 1 below) being faced by those who will confront bioattacks in the future or participate in gathering evidence for the prosecution of any suspect who is apprehended and charged with the anthrax attack of 2001. Also, no attempt will be made to describe the variety of new tests that are currently developed, because the issues raised in Table 1 need to be resolved before any of the tests currently being developed can be evaluated realistically.
Challenge #1: Collecting specimens at the attack site
The first challenge is proper collection of evidence at a site where the release of an infectious microbe is suspected. There are three major goals.
First, the specimens must be collected from appropriate locations. In the case of the anthrax attack, because the spores remain airborne for hours, specimens of air needed to be taken as well as specimens from the surfaces of inanimate objects. But where to collect specimens was not so clear during the first days of the anthrax attack because of lack of knowledge of the likely location of the threat agent. That is, are the spores primarily found in air, or could there be additional contamination of desks, machinery or even water from sinks? How long were the spores likely to remain suspended in the air? Where were the spores coming from? For example, scientists collecting specimens found that they had to check air vents as well as objects in the room where the release was thought to have taken place to ascertain whether the spores might have moved into the air handling system. At least in the case of the B. anthracis spores, investigators had a lot of information about the properties of the spore powder. Deciding how and where to collect specimens may well be more different if a new biothreat agent, with which scientists have much less experience, is used.
A second goal is to preserve the chain of evidence. That is, specimens must be stored in tamper-proof containers and a scrupulous record kept of when, where and by whom the specimens were collected. This is important to eliminate questions about whether a specimen might have been contaminated, intentionally or unintentionally, during handling by investigators. A related problem is that the specimens must be collected and stored under conditions optimal for the preservation of the specimen for further testing. Spores are famous for their ability to survive under a variety of harsh conditions, but other types of microbes that do not form spores differ widely in their ability to survive drying or air or any of a long list of environmental conditions. Keep in mind that the first responders to an emergency will most likely be police officers, emergency medical technicians or firefighters, not trained microbiologists. The good news here is that well-trained first responders know the importance of following a protocol scrupulously. They are less likely to go off on investigative tangents than are laboratory scientists whose well-intentioned curiosity sometimes gets the best of them. An additional factor that looms large is making sure that whatever measures are taken not only insure the integrity of the specimens being collected but also the safety of people at the site. The first responders must be protected to the greatest extent possible from unnecessary exposure to the suspected infectious agent.
Unfortunately, as scientists discussing the issue soon concluded, there cannot be a single protocol for first responders that covers all cases of release of an infectious microbe. The appropriate response to a release of spores of B. anthracis was to quarantine a room or building, releasing people after surface decontamination and directing them to a source of treatment with antibiotics. However, in the case of smallpox (a disease which is caused by a virus that does not survive well in the environment and is transmitted from person to person), it would be more important to sequester and quarantine the people infected rather than the building in which they were infected. Once again, there is some good news. First responders are trained to be flexible, even when following strict protocols. They do not respond in the same way to a four-car pile up as they do to a house fire or a toxic waste spill. This flexibility will stand them in good stead as they deal with the challenge of different protocols for different bioattack threats.
Challenge #2: Recognizing that an attack is occurring and diagnosing the disease
In the case of the anthrax bioattack, the agent responsible was identified almost immediately. In other cases of intentional disease transmission, the identity of the microbe being used in the attack may not be apparent so quickly. This is where physicians and other health care workers come in. The physicians are the ones who will recognize, diagnose and treat infected patients. They face two immediate challenges:
recognizing that a bioattack is underway and
communicating their new knowledge to first responders who may have to encounter the next cases
Herein lies an important potential weak link in plans for response to a bioattack: the feedback loop. First responders need to be informed as quickly as possible of advances the health care workers are making and of concrete recommendations for action, recommendations that may well change as the response to the threat unfolds.
In the case of the anthrax attack, hospital workers in Florida, where the first case was seen and diagnosed had, fortuitously, just attended a workshop on bioterror agents, including the bacterium that causes anthrax. If this had not been the case and if the first professionals on the scene had not thought to include anthrax (a very rare disease) as part of their initial diagnosis, it might have been weeks before the disease and its cause were identified. Most of the viruses and bacteria that have been developed for biowarfare are uncommon causes of disease, especially in developed countries. Few physicians are trained to recognize and diagnose such diseases and even fewer would be likely to consider such unlikely causes of disease when diagnosing a patient’s problem.
An imaginative solution to the problem of how to train health care workers to recognize diseases they might not have seen before is being tested at the University of Louisville in Kentucky under the direction of Dr. Ron Atlas. Makeup artists and actors cooperate to give the actors the appearance of a person infected with the smallpox virus or with Ebola virus. Physicians who know the disease then train the actors to mimic the state of a person with the disease. Watching an actor in a wheelchair suddenly begin to bleed from the eyes and ears, as might occur in a case of Ebola virus infection, is a sight few people will forget. These surrogates will be used to train physicians to identify infected people who may be brought into an emergency room or clinic.
Challenge #3: Analysis of specimens
The next challenge is the analysis of the specimens collected by first responders and by microbiologists subsequently sent to the site. Here, once again, the spores of B. anthracis illustrate some of the difficulties involved in analysis of the specimens.
The first step is to establish whether the specimen contains spores of B. anthracis or of some innocuous Bacillus species such as Bacillus thuringensis (well known to organic gardeners as a natural insecticide). Currently, there are a number of well-established tests that make such identifications reliably in the case of B. anthracis, but such reliable tests may not always be available or, if available, known to non-specialists.
Suppose that the specimen is found to contain spores of B. anthracis. The second step is to establish whether the particular strain of B. anthracis present in the specimen is the one used in the bioterrorist attack, as opposed to some soil isolate that might have been tracked in from outside. Here, the degree of difficulty rises. B. anthracis, unlike many species of bacteria, is genetically quite homogeneous. That is, different members of this species (strains) have very similar properties and DNA sequences. At present, the genomes of a number of strains of B. anthracis are being sequenced in a search for reliable DNA sequence signatures. An ironic twist to this story is that scientists are now searching for as many different strains of B. anthracis as they can find. Yet, in the immediate aftermath of the attack, fears of possible prosecution under the Patriot Act or liability for strains stored in freezers caused many veterinarians who had long kept collections of B. anthracis strains for diagnostic and epidemiological purposes to destroy these stocks. The irreparable loss of these valuable collections is now being bemoaned, a lesson for the future: political pressure should not put scientists in the position of acting in a way that may compromise future investigations.
The biggest challenge of all, and one that is still stumping scientists, is to differentiate between different descendants of the Ames strain of B. anthracis, the strain that was used most widely for generation of weaponized spores (spores in a powder form that could be disseminated easily during an attack). Will it be possible to differentiate a derivative of the Ames strain that was cultivated repeatedly in military laboratory A from one that was cultivated repeatedly in laboratory B? If so, such evidence could help to pinpoint the location from which the spore powder was obtained. At question here is the mutation rate of the Ames strain, which appears to be very low. Genome sequencers are currently searching for possible mutation hot spots in the B. anthracis genome that might make such fine-tuned differentiations possible.
As mentioned at the beginning of this paper, not all analytical tests rely on DNA sequence data. Testing for traces of components of the medium used to grow B. anthracis, which might still be present in the spore preparation, has been suggested as a possible method for differentiating between preparations that came from different laboratories. But so far, such an analysis has not been done successfully.
Challenge #4: Validation — quality assurance and control
The next challenge is, in some ways, the most formidable one, rigorously validating each analytical method by establishing its limitations, its sensitivity, and its reliability. Also important is the robustness of a method — the assurance that the method can be used successfully in many different laboratories and field conditions, always giving the same results. The credibility of analytical results relies absolutely on proof that the analytical procedure has been thoroughly vetted by experts.
Fortunately, the clinical microbiologists, who routinely process specimens containing microbes in hospital laboratories, are well versed in the process of validating an analytical procedure. Since their results may affect the lives of patients and must also be able to survive the scrutiny of their colleagues, these microbiologists have learned how to assess effectively the validity of a new method and to train workers to perform the method in a consistent and accurate manner. Although the validation approaches developed by clinical microbiologists are not designed to be used as evidence in a criminal case, they can certainly aid in providing a model for forensic microbiologists to use as a starting point.
Is all this effort worth the cost?
You do not have to be an expert to realize that the research described in the forgoing material will cost a lot of money. Not only will it be expensive to solve these problems for B. anthracis but also, if full preparedness is the goal, it will be necessary to go through the same process for other agents that might be used in a bioterror attack. What is the taxpayer getting for this large expenditure of money, especially if no further attacks occur?
- One possible benefit is that having a well-prepared response plan in place might deter at least some potential terrorists.
- Perhaps the main benefit, however, is that much of the outcome would also be applicable to tracing natural outbreaks of disease.
- Also, there have been cases in which infected people have intentionally infected others and such cases may well end up in court.
True, specific tests for B. anthracis or Variola (the smallpox virus) would not themselves be of much use, but the development of procedures for reliable collection and storing of microbial specimens and for QA/QC (quality assurance/control) of new molecular tests for identifying and tracking a disease outbreak could be very beneficial in many different infectious disease situations. If, for example, there was a sudden cluster of cases of a disease like Ebola, having a plan for a rapid and effective response could quickly limit the spread of the disease.
Communicating research results
Application of research done on responses to bioattacks to natural disease outbreaks is not a guaranteed benefit, however. Scientists who work in the areas of epidemiology and diagnosis of infectious diseases have always had a tradition of free communication with each other through speeches at open meetings and publication of papers in widely disseminated journals. By contrast, the scientists and politicians who have controlled much of the research on bioterrorism and biodefense have become accustomed to a security system that controls information flow and classifies much of the information obtained.
If those who are responsible for protecting the public against bioattacks insist on keeping most of their discoveries out of the public domain, the public will not be well served. The free communication of scientific findings, free of government censorship, has been proven to be an essential precondition for scientific progress. Also, if an analytical method, for example, is going to be useful in gathering evidence that may be used in court, it must be made available to law enforcement personnel and to lawyers and juries. That is, it will have to become public information.
For these reasons, many scientists are concerned that classifying information about analytical methods will severely limit its use in disease situations that are not bioattacks. There is currently a debate underway about what types of microbiology research results should be published freely. Leading scientific societies such as the American Society for Microbiology and the National Academies of Science have come out in support of free publication of any scientific results that have not been classified, but not all people, especially politicians, agree with this stance.
As is evident from this paper, microbial forensics is a field that includes not only scientific research but also such nonscientific areas as the organization of human efforts in an emergency, the politics of secrecy, and the economics of how much research the government should fund. Scientists are not accustomed to having to deal with this breadth of involvement, but some of them are trying to learn to adapt.
© 2003, American Institute of Biological Sciences. Educators have permission to reprint articles for classroom use; other users, please contact firstname.lastname@example.org for reprint permission. See reprint policy.