Received Date: July 02, 2011; Editor Accepted Date: August 12, 2011; Editorial-Office Accepted Date: October 12, 2011; Published Date: October 13, 2011
Citation: Hugh-Jones ME, Rosenberg BH, Jacobsen S (2011) The 2001 Attack Anthrax: Key Observations. J Bioterr Biodef S3:001. doi: 10.4172/2157-2526.S3-001
Copyright: © 2011 Hugh-Jones ME, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of Bioterrorism & Biodefense
On August 6, 2008 the FBI announced its conclusion that Bruce Ivins, a scientist at the US Army Research Institute for Infectious Diseases (USAMRIID), was the sole perpetrator of the 2001 anthrax letter attacks. Nine days earlier, Ivins had killed himself. The FBI’s announcement met with widespread doubts among scientists and other experts who considered the evidence to be, at best, highly circumstantial. The doubts remain in spite of subsequent FBI briefings and document releases. In September 2008 the FBI issued a formal request to the National Academy of Sciences (NAS) to conduct an independent review of the scientific approaches used during the anthrax investigation. Before that review had been completed, the FBI formally concluded its investigation and released an “Amerithrax Investigative Summary” on February 19, 2010. The NAS committee issued its Report on Feb. 15, 2011, after two delays and an additional meeting at the FBI’s request. When the NAS Report came out the FBI released some 10,000 pages of scientific documents, much of which consisted of plans and reports of laboratory work on unidentified, coded materials. By agreement, all the information that had been available to the NAS committee is now in the public domain. The major finding of the committee, which had no access to classified information, was “It is not possible to reach a definitive conclusion about the origins of the B. anthracis in the mailings based on the available [to them] scientific evidence alone”1. Perhaps their most celebrated finding is that “the scientific link between the letter material and flask number RMR-1029 [located in Bruce Ivins’ laboratory at USAMRIID] is not as conclusive as stated in the DOJ Investigative Summary”2.
Potential production sites of the attack anthrax
Genetic evidence from the attack anthrax itself is the prime indicator of laboratories at which genetically-matching B. anthracis could have been accessed for growing the spores sent in the letters. The spores were found to be Ames-strain B. anthracis and to include multiple colony morphotypes, some apparently unique. The latter provided the basis for four specific molecular assays that were developed for use as indicators of close relationship to the anthrax in the letters3. The FBI assembled a repository4 (likely incomplete5) of 1070 Ames-strain B. anthracis samples from 17 laboratories that were identified as possessing that strain. Assays on 947 of the repository samples6 detected samples from 10 laboratories that tested positive in one or more of the specific assays7. Eight samples from two laboratories were positive in all four assays: seven samples from USAMRIID (which possessed flask RMR 1029, an Ames-strain B. anthracis spore preparation later shown in genetic tests to be indistinguishable from the attack anthrax) and one sample from Battelle Memorial Institute (BMI) in Ohio8 (which had been sent samples from flask RMR 1029 on May 8 and June 18, 2001)9. Samples with partially-matching assay results cannot be disregarded, for a variety of reasons including variability in the assay results10, lack of repeat analyses, false negatives, and the fact that, prior to assay, each sample had been through two rounds of cultivation11. Among the eight laboratories that submitted a total of 63 samples with 1-3 positive assay results12 are: Dugway Proving Ground, the Naval Medical Research Center, Northern Arizona University, the Canadian Defense Research Establishment at Suffield (DRES)13, and a second sample from Battelle14. The submitters of the other three samples have not been revealed.
A priori, the most likely sites of production of the letter anthrax are laboratories that work with dry spores: Battelle, Dugway, and DRES, and their associated institutions and subcontractors. Battelle, for example, is well-known for its aerosol study capabilities and biodefense activities, for which dry spores are routinely needed15. USAMRIID, on the other hand, has always insisted that dry spores are never used in the work there. The FBI says that, prior to the attacks, no US laboratory had Ames anthrax spores in powder form16; however, powdered anthrax spores are known to have been produced at Dugway in the last few years before the attacks17. The FBI recognized that Dugway had the know-how18, and also, the strain—Dugway had produced the bulk of the B. anthracis in USAMRIID’s flask RMR 1029 in 199719. Furthermore, Battelle has an operation at Dugway and some other government locations20, and might have transferred material there. Also note that, according to the US Department of Justice, “Upon the receipt of RMR-1029 spores, the private research laboratory [defined earlier in the same document as “a private laboratory operated by Battelle”] was allowed to provide aliquots to other laboratory facilities for legitimate research purposes21”.
The FBI ruled out Battelle as the source of the attack anthrax on the implicit, and unwarranted, assumptions (1) that the anthrax spore preparations in the letters must have been made covertly, and (2) made by the perpetrator(s) of the attack. They say that every minute in the “Midwest” Battelle laboratory is accounted for, and no researcher was ever alone in the laboratory; background investigations of everyone who had access to the RMR 1029 material received from USAMRIID gave unremarkable results; and the great distance of the Battelle laboratory in Ohio from Princeton, NJ, where the anthrax letters were mailed, “preclude any reasonable possibility that the mailings came from there22”. Dugway and DRES, being much farther from Princeton than Battelle, may have been eliminated on that basis alone. However, there is no publicly available information to rule out the possibility that the anthrax spores in the letters were made somewhere in the normal conduct of authorized laboratory operations, and later acquired by the mailer(s) at the same or some other location.
The FBI has also routinely assumed23 (3) that the attack spores were prepared during the short interval between 9/11 and the mailings of the letters on Sept. 18 and Oct. 9, 2001; but there is no publicly known evidence for that assumption. Battelle received its first shipment of material from the USAMRIID flask RMR 1029 on May 9, 200124. A preparation of dry anthrax spores could have been grown at Battelle from that material any time thereafter, for some authorized (and probably classified) purpose such as vulnerability/response studies25; the spores could then have been provided to a distant client or other authorized person(s), who, following 9/11, decided to conduct the anthrax attacks. The same can be said of Dugway, and perhaps also of the unidentified laboratories that submitted repository samples that tested partially-positive.
The FBI has released a small amount of information on Battelle’s Ames-strain B. anthracis. Battelle received a sample of RMR 1029 in May, 2001 and used it, in part, to test aerosol equipment26; they used part of the second sample, received in June, to grow a new B. anthracis preparation, produced at Battelle on July 17, 200127. In 2002 Battelle submitted 19 samples (in duplicate) to the FBI repository, including samples of all the Ames-strain preparations it is publicly known to possess28. It is interesting that a document that may have contained information related to Battelle’s repository samples was withheld by the FBI because of its security classification29, and that the FBI executed a search warrant at the Battelle laboratory in 2004 to make certain that all its Ames-strain stocks were represented in the repository30.
As for USAMRIID, which the FBI has examined closely31, there is no material evidence that the attack spores were made there, and no direct evidence that an individual at USAMRIID made the anthrax, or mailed it. The FBI cannot point to specific equipment, facilities or materials that would have been needed at the site(s) where the spores were produced, or the time required for their production; in communicating with the NAS Committee, the FBI hedged on all these aspects of the production process, indicating uncertainty while seeming to know more than they were willing to reveal32. They did not tell the Committee the scientific rationale for the choice of the various surrogate preparations studied, and would not divulge the preparation procedures for some of the preparations. Perhaps the motive was to protect proprietary defense information; nonetheless, the lack of information raises questions about the adequacy of the investigation. The NAS Report recommended that a review should be conducted of all classified materials that are relevant to the investigation33.
Silicon and tin additives in the attack letter powders
Silicon34 has been recognized as a “key component35” of the attack anthrax from the beginning36, although there has been much controversy over its nature and purpose. The presence of silica (SiO2) nanoparticles, classically used to increase the dispersibility of biological agents, was eventually ruled out. Six months after the attacks, an unusual chemical was said to be found in the letter spores37, and in November 2003 the journal Science published an article claiming the spores contained a “polymerized glass” component –a silane or siloxane compound that “leaves a thin glassy coating38”. This information was said to have been provided by US Intelligence to officials of two NATO countries. Richard Spertzel, former Deputy Commander of USAMRIID and Senior Biologist of the UN Special Commission for Iraq, later confirmed that this was known to the German Foreign Ministry39. At the end of this section we propose a process, consistent with all the known evidence, by which the attack spores may have been given a polysiloxane (silicone) coating using a tin catalyst.
Meanwhile, Dwight Adams, the Chief FBI scientist, stated at a private FBI briefing of Senators Tom Daschle and Patrick Leahy in late 2002 that the letter anthrax contained no additives, but did contain silicon which occurred naturally in the spore coat, not on the surface of the spore (the exosporium). This information was leaked by “sources on Capitol Hill”40 but did not become fully known until Adams testified in January, 2006 in the Hatfill vs. Ashcroft et al. lawsuit. X At that time Adams said, in a sworn deposition, that scientific information obtained by the FBI about the letter anthrax is too sensitive to reveal to either the public or the Senate, Congress or their staff 41.
An ongoing concern of the FBI may have been to protect what may originally, before all the leaks and speculations, have been valuable security information about anthrax spore preparation. In an August, 2006 scientific article by an FBI scientist42, ostensibly about the procedures used to search bags of Congressional mail, a carefullyworded paragraph was inserted to imply, once again, that the attack anthrax consisted simply of spores, without additives, and was not weaponized. The article was widely touted as “the most expansive public comment on the nature of the powder by any FBI official43.” When the scientific journal printed a reader’s letter objecting to the absence of evidence in the FBI article to support its conclusions44, the journal’s Editor invited a reply but the FBI refused.
Two years later, in connection with the announcement in 2008 that Ivins was the perpetrator of the anthrax attacks, the FBI held a science briefing in which they reiterated that the attack spores contained no additives and that silicon was naturally present in the spore coat. When asked for the percent silicon in the spores, the FBI spokesman refused to specify it45. A few weeks later, FBI Director Robert Mueller was asked the same question at a Congressional hearing on FBI oversight. Mueller said he was unable to answer at that time; but he eventually did so in an April, 2009 letter46 to the Chairman of the Committee on the Judiciary, saying that the powder in the Leahy letter contained 1.4% silicon by weight. He added that insufficient material made it impossible to analyze material from the NY Post, Daschle or Brokaw letters. In fact, however, the FBI had known the percentage of silicon in the NY Post powder since October, 2002, when an FBI laboratory measured silicon content of 10.77%47 (this was not publicly divulged until the FBI document dump of February, 2011). The FBI also analyzed the Leahy sample in 2002, and again in 2004, and found it contained 1.84 and 1.44% silicon48.
Even earlier, measurements made by the Armed Services Institute of Pathology (AFIP) in October, 2001, when they first discovered the presence of silicon, must have been available to the FBI. AFIP released its data in response to a Freedom of Information Act request in April, 201049. Using the AFIP data, one of us (SJ) calculated50 silicon content of the order of 3% in the Daschle powder and 30% in the NY Post powder.
It was already known from earlier work that silicon can be naturally incorporated into Bacillus spores during their formation51. Consequently, the FBI contracted scientists at Sandia National Laboratory in December, 2001 to determine the location of the silicon within the spores by examining thin sections with high-resolution scanning and transmission electronmicroscopy (SEM and STEM)52. The Sandia findings became public in September, 2009: silicon was localized on the spore coat, “not on the spore surface,” in the letter samples, as well as in surrogate samples that contained naturally-incorporated silicon53. The weight-percent silicon “at the spore level” was the same in the Leahy, Daschle and NY Post samples, approximately 1.5%. Sandia did not report the percentage at the bulk level54.
Thus, the silicon content per spore in the Leahy letter, as determined at Sandia (1.5%), matched the bulk silicon content in the Leahy powder, as determined by the FBI (1.4-1.8%). But for the NY Post letter, as the NAS Report notes, “there was a substantial difference between the amount of silicon measured in bulk [by the FBI (10.77%), and by AFIP] and that measured in individual spores [1.5%, by Sandia]. No compelling explanation for this difference was provided to the committee55”. Neither the FBI nor the NAS committee mentioned the extra-sporular material visible in AFIP’s images of the NY Post sample, but not in images of the Daschle sample. Extra-sporular material would be included in measurements made on bulk samples, but not in measurements using higher-resolution SEM on single spores or spore slices, the focus of the Sandia work. We will come back later to the extra-sporular material in the NY Post sample.
The high levels of silicon found in the attack powders are extremely unusual. The largest amount that FBI contractors at Lawrence Livermore National Laboratory could find56, in 26 pre-existing Bacilllus samples submitted to them by the Department of Homeland Security (DHS) and collaborating laboratories57, was about 0.4 wt%, and the highest level they were able to achieve themselves through growth of B. anthracis Sterne spores in the presence of silicate (the mechanism of incorporation postulated by the FBI58), even in media approaching silicate saturation, was 0.3%. Similarly, the Sandia group found that the B. anthracis Ames spore samples sent to them by the FBI59 (other than the attack samples) had either no silicon in their spore coats or considerably lesser amounts, and a much smaller fraction of spores that contained any silicon60. In surrogate samples containing silicon, according to Sandia, “the details are different.”
In an effort to ‘reverse engineer” the attack anthrax, the FBI had asked Dugway early on to make 36 surrogate spore preparations by various methods61. Bulk elemental analysis was carried out by the FBI in 200262 on only ten of these surrogates (none of which contained added “dispersant,” i.e., silica nanoparticles). All ten were found to contain between 0.2% - 5% silicon (in bulk); four of these contained an amount of silicon in the same range as the attack samples. The NAS committee wrote that the bulk silicon analyses indicate that it is possible to prepare spores with high silicon content without adding a dispersant, but that, when analyzed for silicon in the spore coat, none of the FBI samples were similar to the attack samples with respect to either the amount of silicon per spore in the coat, or the fraction of spores containing any silicon in the coat63. Thus, it is possible that most of the silicon in the four samples with high silicon content was extra-sporular, like that in the NY Post sample64; there is no evidence that it was naturally incorporated. The methods of preparation of these four samples, including addition of any substances other than silica particles, were not described to the NAS65. The NAS Report noted that they were not provided with all of the preparation procedures used66; they “sought, but could not obtain, a detailed explanation of the thought process that went into selection of the DPG [Dugway] methods… it was not clear to the committee how the subset of surrogate preparation methods was selected and whether these choices were based on an understanding informed by the investigation or on other assumptions about the approach taken to produce the evidentiary materials67.” In light of all the open questions, it is strange that a “Red Team,” convened at Quantico on March 13, 2007, recommended that the FBI pursue no further research on the silicon in the letter powders68.
A new book69 edited by Bruce Budowle of the FBI and others suggests a reason why the FBI may have wished to avoid the silicon question: “if the estimates of silicon concentrations in the Amerithrax spores are correct, they are not consistent with our current understanding of silica deposition, or those materials must have indeed been produced under an unusual set of conditions. If the latter were true, the silica [sic] evidence might provide a significant bound on the credible growth and production scenarios that would be consistent with the prosecution narrative in this case.”
It has usually been assumed that any intentional additive containing silicon must have been intended to increase the dispersibility of the attack spores70. In reporting their work to the FBI the Sandia scientists wrote that “the silicon and oxygen found on the spore coat are difficult to explain as an intentional addition to the spores, mainly because of their location on the spore coat, which surrounds the spore core and is surrounded or encased by the exosporium71”. However, the spore coat of Bacilli, not the exosporium, is the effective surface of the spore; it is at the spore coat, not the exosporium, that large molecules (mol. wt. about 16,000 and above) are totally excluded from the spore72, and the coat is an important factor in spore resistance to toxic chemicals73. The effects that a silicon compound on the spore coat may have on spore properties, including dispersibility74, cannot be assumed.
Presumably for this reason, an effort to determine the dispersibility of the attack spores by direct measurement was undertaken by Michael Kuhlman of Battelle Memorial Institute. Sometime between October 17-23, 2001, he measured the particle size distributions of an aerosolized Daschle sample and later, of a Leahy sample and of several B. subtilis globigii spore samples made at Battelle using standard methods, with no milling or other processing75. He found them all to be similar. The particle sizes in all cases had bimodal distributions; for the Daschle sample, surprisingly, only 0.05 % of the mass had a diameter of 2 micrometers or less, and 0.9% had a 10 micrometer diameter or less; the Leahy sample had ten times more particles in this respirable range. The NAS Report took the Battelle data to indicate that “powders with dispersion characteristics similar to those of the letter materials could be made without the addition of a dispersant76.” However, there is reason to question whether the attack samples were in pristine condition when these measurements were carried out, or whether Battelle had autoclaved them first, which might have caused clumping. Richard Preston’s The Demon in the Freezer77 describes an argument at a meeting on October 22, 2001 involving the FBI laboratory, scientists from the Battelle Memorial Institute, and scientists from the Army. The Army scientists were telling the FBI that the attack powder was “extremely rarified and dangerous,” while Michael Kuhlman of Battelle “was allegedly saying that the anthrax was ten to fifty times less potent than the Army was claiming….One Army official is said to have blown up…at the meeting, saying to the Battelle man, ‘Goddamn it, you stuck your anthrax in an autoclave, and you turned it into hockey pucks.’” The FBI’s conclusion that the silicon content of the attack anthrax had nothing to do with its dispersibility remains unproven.
Returning to the question of the chemical form of the silicon in the attack anthrax, AFIP summarized some of its October, 2001 SEM-EDX findings as follows: “Significant findings for the SPS02.03 [Daschle] sample included the presence of silicon and oxygen, which is indicative of silica [SiO2] or, more likely, silicates (SiO4, etc.). Distinguishing among the various possibilities would require additional work with standard materials...The SP02.88.01 [NY Post] samples had regions which exhibited the same set of elements found in SPS02.03 [Daschle], but these tended to be on ‘large’ pieces within the sample. Many of the smaller pieces within the sample exhibited the main peak associated with silicon. It appears that silicon (not bonded to oxygen or other elements) is present in many areas of this sample78”. When the NAS committee asked the FBI about this last point, the FBI answered that the presence of “reduced silicon” was “just an observation79”. Exactly what AFIP meant by “not bonded to oxygen or other elements” is uncertain, but a possible guess is “not bonded to oxygen or other elements we looked for.” Surely they did not mean to imply that elemental silicon was present. In any case, an explanation is needed for the fact that the NY Post sample contains two different chemical forms of silicon.
Since we now know from the Sandia work that the NY Post and Daschle spores, in isolation, are elementally indistinguishable, the “large” pieces in the NY Post sample that exhibited the same elements as the Daschle sample in the AFIP analyses must have been spore aggregates. The “smaller pieces,” which contain most of the silicon in the NY Post sample, probably include cellular debris, absent in the more highly-purified Senate samples but present as extra-sporular material in the NY Post preparation. More on this below.
Besides silicon, an even more unusual element was found in the attack anthrax: tin (Sn). The Sandia group found tin80 in the 10- 20 nm continuous Si-O layer surrounding the spore coats of all the attack samples, but not in any of the surrogate samples they studied, including ten made at Dugway, or in samples that contained naturallyincorporated Si-O on the spore coat81. The presence of these unusual elements may be a fingerprint, they wrote. But the NAS committee “was never shown any evidence to indicate that this possibility was pursued further or that these discussions led to any conclusions about the source of material or production methods82.”
FBI documents released in February, 2011 reveal that in early 2002, the FBI laboratories also carried out extensive elemental analyses of the attack samples and of various other spore preparations, including the same ten surrogates made at Dugway. Using inductively coupled plasma optical emission spectroscopy (ICP-OES), a sensitive method, they found tin in the attack samples, but no tin in other spore preparations or in media83. Table 1 contains some of the FBI data. The amounts of silicon and tin appear to be related: the more of one, the more of the other. This suggests a process that involved both silicon and tin84. There is no tin in the simulant to which silica particles had been added, and no silicon or tin in the simulant without added silica, or in the contents of flask RMR 102985. The levels of other elements are reflective of media components (which do not include tin), as can be seen by comparing Table 2, which shows elemental analyses determined by ICP-OES for nine Bacillus spore samples prepared by different methods86. Neither silicon nor tin was detectable in any of the samples. Comparing Tables 1 and 2 it is immediately obvious that the attack powders are in a league of their own as far as the elemental forensic evidence is concerned. Any meaningful reverse engineering studies would have to deliberately include compounds containing silicon and tin if there were any hope of reproducing the attack powders. Using protocols that do not add silicon or tin would be quite meaningless. It is essential to learn whether the FBI labs tried to add tin to their preparations.
Sample 1 (Q12)
Sample 2 (Q1801)
| B. subtilis with silica fluidizer
Table 1: Elements (ppm) detected by FBI labs in attack powders and simulants measured by
inductively-coupled plasma atomic emission spectroscopy (ICP-OES) (data from FBI document B1M7).
**B. thuringiensis israelensis spore samples
n/d = none detected
Other elements reported: Rb, Ba, Sr, Mo, and Pb: all values were below 12 ppm except for 2 samples with Pb between 12-20 ppm and 3 samples with Sr between 16-105 ppm.
Table 2: Elements (ppm) in B. thuringiensis samples prepared and processed by various methods measured by inductively-coupled plasma atomic emission spectroscopy (ICP-OES) (data from Velsko, S, Bioagent sampling matching using elemental composition data: An approach to validation, ICR:-TR-220803, Lawrence Livermore Laboratory, 2006).
All the evidence in the public domain is consistent with the concept that the spore coats of the attack anthrax were silicone-coated. Silicone polymers are typically formed by hydrolysis of a silicon compound such as dimethyldichlorosilane87 (or other silanes with similar substituents)88, which contains no oxygen. Hydrolysis replaces the chlorine atoms with oxygen to form dimethylsilanol, which polymerizes spontaneously to form polydimethylsiloxane, containing silicon and oxygen in equal amounts. The polydimethylsiloxane89 chains can then be cross-linked (“cured”) to form a three-dimensional silicone coating for encapsulation. This step requires an organotin catalyst90,91 such as a dibutyltin dicarboxylate92.
A procedure of this kind can be envisioned for encapsulating B. anthracis spores. Silane monomers like dimethyldichlorosilane are low-molecular-weight liquids that probably can penetrate the exosporium93, the loose-fitting membrane sac that encloses the spore. If silane monomers were added to a suspension of dry spores in an organic solvent, the silane would not contact moisture until it reached the spore coat, where residual moisture diffusing from the core inside the spore94 would cause hydrolysis, followed by polymerization at the spore coat. The polysiloxane chains that would be formed at the spore coat could then be cross-linked to encapsulate the spore. This step would require continued diffusion of moisture from inside the spore, as well as an organotin catalyst. Organotins have low solubility in water but, like silanes, are soluble in organic solvents such as ether, carbon tetrachloride, etc95. The ratio of tin to silicon in the attack spores is “about right” for a tin catalyst used to produce a silicone coating, according to a chemist in the field96.
Before contacting moisture, chlorosilane compounds react readily with various functional groups in organic materials (e.g., amino groups, alcohols, carboxylic acid groups, etc.) that would be found in cellular debris. If the cellular debris contained no moisture, the result would be “silylation” of the material, i.e., addition of a silicon atom (but no new oxygen atom) at each silylated site. The silylated material would be found in the extrasporular material such as observed in the NY Post sample – the “smaller pieces.” Alternatively, if the cellular debris in the dried NY Post sample retains some moisture, chlorosilane molecules would react as they do at the spore coat, hydrolyzing and then polymerizing to form clumps of polysiloxane, a high-molecular-weight material that would not be soluble. The extrasporular material may be a mixture of these possibilities. The smaller pieces that “exhibited the main peak associated with silicon,…not bonded to oxygen or other elements” would be the silylated areas, containing Si-O-C or Si-N-C bonds, not the O-rich Si-O-Si-O bonds that form the backbone of polysiloxane polymers. These different bonds, which can be distinguished by spectroscopic methods97, would explain the AFIP observations.
Although a process of this type is not known to have been applied to bacterial spores, methods have been developed to encapsulate biologicals with silicone polymers that confer high stability while protecting biological activity98. Polydimethylsiloxane has been used in some pharmaceutical coating materials for years, even though a toxic material, organotin, is employed as catalyst in its production99.
Microencapsulation can provide protection from the environment and better dispersibility and flowability100. The polydimethylsiloxanes have a low surface tension that produces “very hydrophobic films and a surface with good release properties, particularly if the film is cured… surface tension is also in the most promising range considered for biocompatible elastomers101.”
Inorganic tin compounds are genotoxic to bacteria102; organotins are increasingly toxic; triorganotins are used as general biocides103. The potential toxicity of tin compounds underlines the probability that tin was not added to the attack spores during their growth, but rather, during post-production treatment of the spores.
Whether or not the hypothesis just presented is accurate in detail, it is obvious that the attack spores must have been treated with some procedure that produced the silicon-tin signature observed on their spore coats. That signature, alone, shows that the attack spores were grossly over-qualified for their job. The FBI has adequately proclaimed and demonstrated that surrogate samples, containing no silicon or tin, can be sufficiently dangerous. Perpetrators who were inspired on 9/11 and ready to go on September 18 would not need to bother with unnecessary complications like coatings. If they could, they would take advantage of pre-existing spore preparations, regardless of their sophistication.
It would be difficult not to conclude that the spores in the attack letters were prepared for some purpose other than terrorism. Potential procedures that might be applicable for silicone coating of spores, barely touched on here, are complex, highly esoteric processes that could not possibly have been carried out by a single individual. They would require a laboratory with specialized capabilities and expertise not found at USAMRIID, in addition to the possession of the correct strain of B. anthracis Ames associated with flask RMR 1029.
Further analyses of the attack samples to determine the chemical forms of the silicon and tin they contain should have been a high priority for the FBI. As AFIP pointed out, “distinguishing among the various possibilities would require additional work with standard materials” using electron microscopic/spectroscopic methods. The attack powders should have been analyzed using a wider arsenal of analytical techniques, including time-of-flight ion mass spectroscopy (TOF-SIMS). Such studies can and should be done now, focusing on precisely what chemical species and chemical bonds involving silicon are present in both the spore coats and the extra-sporular material in the NY Post powder. Reverse engineering experiments are also essential using various tin-catalyzed silicone formulations aimed at producing coatings on B. anthracis spores, and testing any other hypotheses that may link silicon and tin content in the attack spores. The work needs to be undertaken both by government bodies and independently by disinterested scientists. Critical information also remains to be uncovered in records pertaining to silicon and tin compounds, their purchase, possession, use, testing, storage and so forth, at locations of interest.
B. subtilis contamination in some of the attack letter powders
A bacterial contaminant that could be an important institutional “fingerprint” was found in the attack powders in the NY Post and Brokaw letters, but not in the more highly purified material in the letters sent to the Senate104, which were mailed later. The genetic sequence of the NY Post contaminant had 98% genetic similarity to B. subtilis 168105, a standard laboratory strain, but it was not identical to any presently recognized strain of B. subtilis, based on assays for three genetic markers106. The Brokaw B. subtilis contaminant has never been sequenced, but its similarity to the NY Post contaminant was demonstrated at 23 genetic loci for which multiple PCR assays were developed107. An environmental sample from the office of American Media, Inc. (AMI) in Florida –the only AMI sample for which any analyses have been reported –contained bacilli that were phenotypically (i.e., observationally) indistinguishable from the B. subtilis in the NY Post and Brokaw letters108. Cell suspensions of the B. subtilis and B. anthracis isolates from the AMI environmental sample were prepared in 2006 and sent for genetic analysis109, but evidently they were never analyzed; in 2011 the FBI told the NAS Committee verbally that the U.S. Attorney’s Office advised that further characterization of those samples “would not be undertaken110.” No explanation for this decision has been offered.
Samples submitted to the FBI repository by all known possessors of B. anthracis Ames appear to contain no B. subtilis contamination that perfectly matches that in the NY Post powder, as shown by negative results in assays for ID65111, one of three rare markers in the NY Post strain112. An FBI subpoena, issued in 2002, had required that samples submitted to the repository must be taken from stock B. anthracis cultures, which are usually handled with great care to prevent contamination. Thus it is surprising that 30% of the 1070 repository samples tested positive113 in a genetic assay for sboA, a pan-B. subtilisspecific marker used by the FBI as a common marker for B. subtilis114. On the basis of the ID65 results on repository samples the FBI concluded, and the NAS concurred, that the B. subtilis contaminant did not provide useful forensic information115.
But this conclusion was unwarranted. Flask RMR 1029 at USAMRIID, the putative B. anthracis stock culture from which the spores in all the attack letters were derived and the source of the more interesting repository samples contains no B. subtilis of any kind116. Because the letter spores were probably contaminated during growth or processing, the repository of stock B. anthracis cultures was not an ideal place to look for a contaminant of the letter spores. The NY Post B. subtilis strain is an extraordinarily efficient sporulator, which suggests that it may readily persist in the environment117. It is noteworthy that “military” strains that originated at USAMRIID and Dugway for use as B. anthracis simulants have evolved over time to become hypersporulators118. The FBI should have collected environmental and other samples from all laboratories that were potential production sites of the letter spores—i.e., laboratories known or suspected to possess B. anthracis similar to the attack anthrax.
FBI documents released in February 2011 contain assay results on about 400 samples consisting mainly of swabs, etc., grouped under the titles “Environmental Samples” and “Location Searches119.” The samples are not identified. Some or all of them may have originally been collected for B. anthracis analysis along the suspected paths of the letters and at contaminated worksites120, perhaps including some at USAMRIID. The samples were not assayed for B. subtilis until 2007-2008, when the procedures became available. A few of the 400 samples were found to contain B. subtilis, but none of them was ultimately shown to contain B. subtilis identical to that in the NY Post letter. The assay history of these samples raises questions about the reliability of the procedures used, however121.
Because of the various inconsistencies that have been encountered in the B. subtilis assays, a re-evaluation of the assay procedures is needed; future work should preferably be based on full genome sequencing. Since B. subtilis has been shown to possess high genetic variability122, perfect agreement with the NY Post strain should not be sought. For this reason it would also be desirable to re-examine some of the repository samples.
There is no evidence that relevant samples were ever collected at Dugway, Battelle or other potentially suspect sites. Samples of microbial production runs that took place in the appropriate time frame at those locations, and environmental samples, especially in areas involved in production and other activities with Bacillus species (aerosol testing, for example), should have been collected and examined for the presence of B. subtilis resembling the NY Post strain. Some environmental sampling may still be worthwhile today, and samples of stock cultures and products, as well as records of past activities, should still be available.
The NAS Report, in the words of the Washington Post, “is not satisfying - nor is it conclusive123”. Many scientists who contributed to the investigation have expressed reservations about the FBI’s conclusions124. Senator Patrick Leahy, a target of the anthrax letters and Chairman of the Judiciary Committee, says he has “extreme doubts” about the case; “I’ve expressed those concerns to the FBI, and this report adds to those concerns125”. Senator Chuck Grassley, the ranking member of the Judiciary Committee, says the NAS report “shows that the science is not necessarily a slam dunk. There are no more excuses for avoiding an independent review and assessment of how the FBI handled its investigation in the anthrax case126.” Legislation127 to establish a Congressional commission for that purpose has been reintroduced by Rep. Rush Holt128, a scientist and Princeton professor who represents the area in New Jersey where the attack letters were mailed.
The National Academy of Sciences was the right body to examine the scientific issues in the anthrax case, and the scientific issues were the right target. Once the method of preparation of the attack anthrax is understood, the questions of who made it, and where, will be rapidly resolved. But the NAS committee was held back by its dependence on the FBI for information and sponsorship. To quote the Washington Post again, the committee “should not be blamed for nitpicking over the test results; that are essentially what it was tasked to do by the FBI, which commissioned its report129.” The committee had no access to classified information130, and they met with a number of impediments in trying to interpret the data they were given131. Answers to questions posed to the FBI for clarification, they wrote, “were sometimes minimal or terse, or were deflected as … beyond the purview of the committee despite the committee’s explanation that the questions were of a scientific nature132”. Clearly, the committee’s purview should have been broader.
Critical scientific questions, some of which have already been indicated in this paper, must be answered before the anthrax case can be laid to rest. That will require scientific expertise and political neutrality, ideally with full access to all that the FBI knows, and with the resources to commission additional work at other laboratories if the existing scientific information is inadequate. Indeed, further scientific investigation may be the only way to bring the facts of the case to light.