Received Date: October 01, 2012; Accepted Date: December 12, 2012; Published Date: December 17, 2012
Citation: Hugh-Jones ME, Rosenberg BH, Jacobsen S (2012) Evidence for the Source of the 2001 Attack Anthrax. J Bioterr Biodef S3:008. doi: 10.4172/2157-2526.S3-008
Copyright: © 2012 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
The elemental composition of the 2001 attack anthrax presents critical clues that were not considered or were misinterpreted throughout the original investigation. Extensive experimental data released by the FBI after the anthrax case was closed make it possible to trace some of the implications of these clues: the substantial presence of tin, a toxic material that must have been added subsequent to growth, and a uniquely high content of silicon in the attack spores. No Bacillus spore preparations other than the attack anthrax have ever been found to contain such a high level of silicon, although some surrogate spore powders prepared at Dugway following FBI instructions have been cited as evidence that high levels of silicon can be reproduced; however, examination of the experimental data reveals that the silicon in these samples was unquestionably an artifact. The elemental evidence suggests that the attack spores had been coated with silicone (a polysiloxane) in the presence of tin, which catalyzes the cross-linking of polysiloxane chains needed to form an encapsulating coating on the spore coat. Microencapsulation helps protect biological agents from damage during atmospheric exposure and from the body’s defenses during infection, and would defeat some detection methods. Microencapsulation, which would explain the location and amounts of both tin and silicon in the attack spores, requires special expertise and sophisticated facilities. DOD-sponsored projects explicitly involving microencapsulation at DARPA, Dugway and perhaps elsewhere were spelled out publicly in budget documents in 1999 and thereafter, and executed at the very time of the anthrax attacks. Both the Dugway laboratory and Battelle Memorial Institute, a sub-contractor at Dugway, had extensive experience in making Bacillus spore powders; both had access to Bacillus anthracis genetically matching the attack spores; both could have made the attack spores legally for institutions conducting biodefense activities that required microencapsulated spores. Furthermore, a small but significant amount of tin, about 4% of that in the attack spores, has been found in some surrogate spore products made at Dugway. A measureable tin content has not been found in any other Bacillus spores except the attack spores. The tin in the Dugway surrogates may have been a remnant, indicative of earlier, classified work. Avoidance of governmentsponsored, classified research may account for some of the limitations of the investigation.
The FBI’s investigation of the 2001 anthrax attacks focused heavily on genetic evidence that strongly links the attack anthrax to a liquid suspension of Ames-strain Bacillus anthracis spores in a flask known as RMR 1029, found at USAMRIID (the US Army Medical Research Institute for Infectious Diseases). Prior to the attacks, spores from this flask − believed to be the parental source of the anthrax in the letters – had been sent to a number of other laboratories. Collection and analysis of an FBI repository of samples from all laboratories known to possess the Ames strain indicated that several laboratories had Bacillus anthracis that was identical or nearly identical to that in RMR 1029. There is no public information, however, as to whether any of those laboratories may have produced the attack anthrax, possibly as part of their authorized work1.
In addition to the genetic composition of the attack anthrax, the elemental compositions of the spore powders provide equally important clues: the presence of tin, a toxic material that must have been added subsequent to growth; and a uniquely high content of silicon, not in the familiar form of silica nanoparticles. Examination of the extensive (but incomplete) experimental data released by the FBI in February 2011 and of other evidence from government sources has made it possible to trace some of the implications of these clues2. It is remarkable that there is no evidence that the FBI tried to determine the chemical forms of the tin and silicon in the attack powders, although that could have been (and still could be) done straightforwardly by spectroscopic methods. Perhaps the FBI knows the answers but has concealed them in order to keep potentially dangerous information out of the hands of adversaries.
Significance of tin in the attack anthrax
Analysis by the FBI for tin in the attack samples, using ICP-OES (inductively-coupled plasma optical emission spectrometry), found 0.1979 wt% tin in the Leahy powder and 0.6511 wt% in the NY Post sample3. In contrast, many other Bacillus spore preparations4 from various sources that were subjected to elemental analysis by the FBI or its contractors were found to contain no tin at all (see data in Table 1, and summary of tin contents in Table 2).
|Sample Name/ (No. of samples)||SpeciesA||Source||Analysis Lab||MethodBC||Ref.||Tin wt%||Silicon wt% (No. of samples)||Spore coats with : silicon/tin (%)|
|Leahy||Ba||letter||FBIE||ICP-OES||B1M7 p.9, 14||0.1979||1.6|
|Ba||Dugway (rev. eng.)||FBI||SEM-EDXJ||B1M7
|NDLB SDLB||Ba||Dugway (rev. eng.)||Sandia||STEM-EDX||NAS p.70, 67||0
|Dugway SEM stubsL (12)||Ba||Dugway||FBI||SEM-EDXJ||B1M7 p.38, 54||n.d.||?(inter-ference)|
|Bc||?||FBI||ICP-OES||B1M7 p.7-10, 93||n.d.||n.d.|
|BuransN (12)||Ba||USAMRIID||FBI||ICP-OES||B1M7 p.23-25, 93||n.d.||0.010-0.019 (11), 0.0265 (1), av.0.016|
|Stewart et al. 1980||Bc||Stewart et al. 1980||Stewart et al. 1980||STEM-EDX||Stewart et al. 1980O||0.3|
|Somlyo||Bc||Somlyo, sample from Stewart et al.1980||Sandia||STEM-EDX||B1M6 p.367||+|
|Ba, Bt, Bg||DHS & collaborating labs (26) + other (6)||L. Liver-more||NanoSIMS||Weber et al. 2009P||0.002-0.004 (3), 0.009-0.13 (28),
new preps (23)
|Ba||L.Livermore||L. Liver-more||NanoSIMS||Weber et al. 2009||0.002-0.004 (5), 0.007-0.13 (17),
|RMR 1029||Ba||USAMRIID||Sandia||STEM-EDX||B1M1 p.109Q; NAS p.67||0 n.d.|
|RMR 1030||Ba||USAMRIID shake flask||Sandia||STEM-EDX||ibid.||6 n.d.|
|040255-1R fermentation||Ba||Dugway fermentation||Sandia||STEM-EDX||ibid.||22 n.d.|
|NBFACS (12)||not identified||NBFAC||Sandia||STEM-EDX||B1M6 p.432-472||0 (7),
ABa = Bacillus anthracis; Bc = Bacillus cereus; Bt = Bacillus thuringiensis; Bg = Bacillus globigii.
BICP-OES = inductively coupled plasma-optical emission spectroscopy; SEM-EDXj = scanning electron microscopy with energy dispersive X-ray analysis; STEM-EDX = transmission SEM on thin sections; NanoSIMS = nanometer-scale secondary ion mass spectroscopy.
CSensitivity of SIMS methods (NanoSIMS or TOF-SIMS (Time-of-Flight)) is greater than that of ICP-OES. Both SIMS and ICP-OES can measure an entire sample to better than 0.0001 wt%; SEM-EDX can see only the top layer and is therefore less reliable, except on thin-sectioned samples (see Brewer LN, Weber PK, Grant RP, Ghosal S, Michael JR (2006) Microanalytical Methods for Bio-Forensics Investigations. UCRL-CONF-21920, Lawrence Livermore Laboratory;Techniques for Metal Analysis at CEST (www.nd.edu~cest/news/inorgmetals. pdf; and Drioli E, Criscuoli A, Macedonio F (2011) Membrane-Based Desalination: An Integrated Approach (MEDINA) (IWA Publishing)).
D+ signifies that the element was detected on some spore coats, but quantitative data are not available.
EFBI analyses using ICP-OES are summarized at B1M1 p. 83 and at B1M7 p. 93.
FPreparation of the Dugway “reverse engineered” samples is given in B1M13; sample codes are on p. 72.
GSamples NDLB and SPVB are those for which the analyses were considered most questionable by the FBI laboratory (see text).
ISPOI is one of the 36 Dugway reverse-engineered surrogates not chosen by FBI for ICP-OES analysis.
JNote that SEM-EDX measurements on non-sectioned samples do not penetrate the samples, and therefore may not detect elements that are detectable by ICP-OES or STEM-EDX.
Kn.d. indicates not detected.
LSample identities not revealed.
MSixteen samples obtained by the FBI from Teresa Abshire of USAMRIID. Some information on their preparation is given in the FBI’s Elemental Analysis Summary Table at B1M7, p. 93.
NTwelve samples obtained by the FBI from Capt. James Burans of USAMRIID. Information on preparation of the samples is given at B1M7 pp. 23-24.
OStewart M, Somlyo AP, Somlyo AV, Shuman H, Lindsay JA, et al. (1980) Distribution of calcium and other elements in cryosectioned Bacillus cereus T spores, determined by highresolution scanning electron probe x-ray microanalysis. J Bacteriol 143: 481-491.
PWeber P, Viani B, Davisson L, Velsko S [Lawrence Livermore National Laboratory], “Nanometer-scale secondary ion mass spectroscopy for microbial characterization,” presentation to the National Academy of Sciences Committee on September 25, 2009.
QCodes for samples 1029 and 1030 are identified at B1M6 p.30; code for 040255-1, at B2M14 p.69.
RSee text and references for explanation of 040255-1, the sample code used to indicate a specific fermentation product made at Dugway.
SNBFAC = National Bioforensic Analysis Center.
Table 1: Tin and Silicon in Bacillus Spores.
|Number of Samples||Tin in Sample||Tin on Spore Coat||Analytical Method||Sample Types|
|10||0.0033-0.0266 wt%||ICP-OES||Dugway Reverse Engineered|
|3||n.d.||STEM-EDX||RMR 1029, 1030, Dugway fermentation|
Total 51 samples, tin not detected in any except the 10 Dugway Reverse Engineered
ABased on data in Table 1, excluding attack samples (shown by ICP-OES to contain tin) and negative SEM-EDX analyses of unsectioned samples, which reflect only the sample surface, not the entire sample.
Bn.d. signifies not detected.
Table 2: SummaryA: Tin in Bacillus Spore Preparations.
To explain the substantial presence of tin in the attack spores, which has never been addressed by the FBI, Hugh-Jones et al.5 have proposed that the spores were processed after growth by coating them with silicone, typically a polysiloxane formed by hydrolysis and polymerization of a silane compound, in the presence of tin. Tin catalyzes the cross-linking of polysiloxane chains, which would thereby form an encapsulating silicone coating on the spore coats6 – the location at which tin, and silicon as well, have been found7 in the attack spores. This process, which would explain the presence, location and amounts of both tin and silicon, would require special expertise and sophisticated facilities. It is an aim of this paper to explore the evidence that the spores used in the letter attacks may have been microencapsulated for legitimate biodefensive purposes before they fell into the hands of the letter sender(s).
“Reverse engineering” of the silicon content of the attack spores
Natural incorporation of silicon by Bacilli, from silicates in growth media, was first reported in 19808. Sandia National Laboratory has shown that the silicon in samples made at that time, as well as Bacillus samples more recently produced in other laboratories, is located at the spore coat9. All of the many Bacillus anthracis and other Bacillus spore preparations with naturally-incorporated silicon that have been analyzed by the FBI or its contractors have contained a maximum of 0.5 wt% silicon, usually much less, often zero (Table 1), with the sole exception of eight samples from a set of 36 agar-grown Bacillus anthracis spore powders prepared in 2003 at Dugway (DPG, Dugway Proving Ground) according to FBI instructions, in an effort to “reverse engineer” the letter powders (data in Table 1, silicon contents summarized in Table 3). The 36 samples comprised all permutations of: two media, two washing procedures, four drying procedures, and three milling procedures10. Based on the plan and the extensive experimental data provided by Dugway11, there appears to have been no intentional effort to reproduce the silicon or tin in the attack spores; the emphasis seems to have been on particle size and viability. However, at that time there was considerable public concern about the silicon in the attack spores (the presence of tin had not been divulged by the FBI).
|Number of Samples||Silicon in Sample||Silicon on Spore Coat (% spores)||Analytical Method||Sample Types|
av. 0.14 wt%
|2||0.25 wt%||ICP-OES||Dugway Rev. Engineered|
|(8)||artifactC||n.d. (2 examined)||STEM-EDX||Dugway Rev.
|8||0.3 wt.%||n.d.||STEM-EDX||RMR 1029 &
av. 0.03 wt%
|1||+ D||STEM-EDX||from Stewart 1980|
Total 118 samples: silicon 0-<0.3 wt% in 107 samples; 0.3-0.5 wt% in 3 samples;
shown to be an artifact in 8 samples.
ABased on data in Table 1, excluding attack samples.
Bn.d. signifies not detected.
CSee text: silicon detected by ICP-OES in 8 Dugway Reverse Engineered samples has been shown to be a milling artifact.
D+ signifies “detected”.
Table 3: SummaryA: Silicon in Bacillus Spore Preparations.
The FBI selected only ten of these “surrogate” samples for elemental analysis, using ICP-OES in an FBI laboratory. Eight of the ten samples were found to contain unusually high levels of silicon (0.5-5 wt%, average 1.4 wt%)12. These eight are the only known Bacillus preparations, other than the attack spores, containing more than 0.5 wt% silicon. These eight samples were all ball-milled, while the remaining two samples, which contained only 0.2 and 0.3 wt% silicon, were milled by hand in a porcelain mortar and pestle13,14. The FBI did not choose to examine any of the samples that had been milled by a third procedure – a manual stainless steel ball and sieve method.
The ball-milled samples also contained extremely high levels of aluminum (0.4-3 wt%), while the mortar-and-pestle-milled samples contained only 0.0087 and 0.0149 wt% aluminum15 (The NY Post and Leahy samples contained 0.0158 and 0.0220 wt% aluminum, respectively).
The ball mill described by Dugway consisted of a porcelain milling jar containing a grinding medium consisting of zirconium cylinders, rotated for approximately 16 hours. Porcelain is composed of silica and alumina. The presence of a high level of aluminum only in the ballmilled samples is telltale and would be hard to miss. Furthermore, the Dugway laboratory noted that all ball-milled samples (but not those milled by other methods) showed a high degree of clumping, as viewed microscopically, making spore viability determination impossible16, and the FBI laboratory noted that the Dugway samples were abnormal in not fully responding to the usual acid digestion procedure, as a consequence of which the concentrations of two of the samples in particular-those with the highest silicon content (5 and 2 wt%) – were likely inflated17. These observations signal that most or all of the silicon and aluminum in the eight ball-milled samples was an extra-sporular artifact. Analyses done at Sandia of two of the ball-milled Dugway preparations (those containing 5 and 0.7% wt% silicon, NDLB and SDLB in Table 1), using STEM-EDX (scanning electron microscopy with energy-dispersive X-ray analysis) measurements on thin sections, confirm that none of the silicon was in the spore coat18, where it is located both in the attack spores and in Bacilli containing naturally-incorporated silicon19.
The question arises as to why the FBI wanted any of the Dugway samples to be ball-milled, or milled at all, given that the purpose of making the samples was “to give some insight into how the materials in the [anthrax] letters were made20”, and given that the Dugway samples were made in 2003, well after Battelle (Battelle Memorial Institute, BMI, a defense contractor) had prepared unmilled Bacillus globigii surrogates that compared favorably to the Leahy attack spores with regard to dispersibility, as indicated by aerosol particle size distribution (see next section). Battelle’s report to the FBI on particle sizes, dated February 2002, noted that their result “indicates that neither milling nor other processing of a freeze-dried B. anthracis spore powder was required21”.
The FBI/Dugway sample preparation plan22 had called for some additional samples to be dried by an (undescribed) acetone method in use at Dugway, but this part of the plan was evidently set aside, although portions of pastes from two other samples were eventually acetonedried23 – but they were never named, further processed, described or analyzed, at least not in publicly-released documents. Acetone drying might have obviated any reason for milling: Bruce Ivins, who assisted the FBI in the early part of the investigation, later quoted one of his USAMRIID colleagues as telling him “he’d use an organic solvent like acetone or alcohol to pull water out of purified spores and then easily make them into powder24”.
Along with the 36 agar-grown samples, Dugway had also prepared two surrogate samples by fermentation in liquid Leighton- Doi medium25. This was an interesting experiment, in which one fermentation sample was grown in the presence of an antifoam agent containing polydimethylsiloxane (Sigma Antifoam C), while the other was grown with an antifoam containing no silicon (Sigma Antifoam 204). Did the spores that had been exposed to polydimethylsiloxane during growth contain a higher level of silicon than has been found in spores exposed only to silicates in media? If so, might the attack spores have been exposed to polydimethylsiloxane? The only relevant information released to the NAS Committee and the public was Sandia’s observation that a Dugway fermentation sample received from the FBI under the code name “040255-1,” for determination of silicon in the spore coats, contained some silicon on about 25% of the spore coats26. This information was used solely as evidence that some fermentation samples contain some silicon. Which of the two fermentation samples this was, and how much silicon was on the spore coats and in the bulk sample, were not revealed to the NAS or the public. Sandia noted, however, that, although some of the various surrogate samples they studied contained silicon on the spore coat, compared to the attack samples “the details are different…chemistry, distribution27”.
The silicon content of the eight ball-milled Dugway agar-grown surrogates has been wrongly cited as evidence that high levels of silicon can be reproduced in Bacillus anthracis preparations28. Similarly, the fact that small amounts of silicon can sometimes be incorporated naturally into the coats of Bacillus spores during their growth has been cited to imply that the much larger amounts of silicon in the attack spore powders must result from the same process29, even though no growth experiments have actually reproduced the high silicon content, nor has the chemical nature of the silicon in the two cases been determined and compared. This selective use of partial information may have convinced many who have not looked into the experimental details that the silicon question has been resolved.
Dispersibility of the attack anthrax spores as aerosols
Because of the suspicion that the silicon in the attack spores might have been there to increase their aerosol dispersibility, Battelle – known for its aerosol expertise – was asked to compare the aerosol particle size distributions of the attack spores to those of dry simulant spore preparations made at Battelle without any special post-growth processing30. The simulants were two Bacillus globigii powders, “washed” and “unwashed,” isolated only by centrifugation, water-washing (or not) and lyophilization31. The aerosol particle size distributions of the Bacillus globigii samples were found to be bimodal, with 1-2% in the single spore range; their distributions were similar to, and slightly narrower than, that of a sample taken directly from the Leahy letter and analyzed in February 200232. The similarity of the results for the Leahy and surrogate samples appears to confirm the FBI’s repeated insistence that the attack spores had not been treated in any way that affected aerosol dispersibility. These results can best be understood in light of the observations of Dr. Thomas Geisbert at USAMRIID, who measured particle sizes in the pristine Daschle sample by electron microscopy and reported that hydrated samples consisted of single spores – which accounts for the Daschle titer of 2.1×1012 cfu/g, the theoretical limit for pure, 100% viable, single Ames spores; but in dry samples (such as those used for aerosol studies) Geisbert found that the spores aggregated to form clumps of up to 105 spores33.
Battelle initially (October 2001) examined the aerosol particle size distributions of two Daschle powders provided by the FBI right after the attack, labeled SPS.57.0334 and SPS.57.0835, neither of which was taken directly from the attack letter itself36. These were the results that were most frequently cited by the FBI. Both the Daschle samples had very low titers (4.6×1010 cfu/g and 1.5×108 cfu/g, respectively) and showed other signs of contamination and aggregation37. It is therefore not surprising that the particle size results for these samples compare poorly with the Bacillus globigii results. Why did the FBI choose to send samples recovered from spills in Daschle’s office, rather than a pristine sample, for this important early test? Only the Leahy data, obtained some four months later and largely ignored, can be considered significant, allowing the tentative conclusion that the attack spores probably had no special advantage or disadvantage for aerosol dispersion.
Microencapsulation of pathogens in US biodefense research
Although there is no evidence to indicate that the tin and silicon content of the spores conferred any benefit for purposes of the letter attacks, their presence is meaningful if the attack spores had been prepared legitimately for other purposes. Silicone microencapsulation would have been desirable for increasing the resistance of the spores to inactivation by hazards such as UV light, ozone or toxic materials38, and for preventing detection of the spores by some methods. (It is at the spore coat, rather than the external membrane (the exosporium) of Bacillus anthracis that these functions occur39). These are properties of military concern. The use of microencapsulation for such purposes was already an old idea40 in the biodefense community in the years immediately preceding the attacks. Microencapsulation by special polymers to produce particles in the 1-10 micron range could protect microbes from environmental damage during aerosolization and delivery [e.g. via bomblets] and also from the body’s initial defenses during the infection process, according to an encyclopedia on weapons of mass destruction published in 2004, and could also help defeat some detection schemes41. The encyclopedia notes that the technology requires an advanced research and development infrastructure, unlikely to be available to terrorists, but “state-level CBW programs could certainly employ microencapsulation to produce highly effective weapons of mass destruction42”.
In the non-military literature it has recently been shown that single, living Bacillus spores can be encapsulated using layer-by-layer polyelectrolyte nanocoating43, and that individual cells, including Bacillus spores, can then be subsequently encapsulated with silica44; the silica encapsulation greatly enhanced viability by protecting the cell from harsh environments.
By 1999 and thereafter, DOD (Department of Defense) interest in pathogen encapsulation was explicitly spelled out in (unclassified) budget documents for the Biological Warfare Defense program at DARPA (the Defense Advanced Research Projects Agency) and the Chemical/Biological Defense program, which includes Dugway.
DARPA’s Biological Warfare Defense program was conducting a project, initiated in 1995, to develop a miniature time-of-flight mass spectrometer for rapid detection of a broad spectrum of chemical and biological warfare agents in aerosol form45. Johns Hopkins University’s Applied Physics Laboratory (APL) took the lead, with critical collaboration from USAMRIID to provide and test pathogens and develop a database of their mass spectral signatures; this work included preparation of both Sterne46 and Ames (inactivated) Bacillus anthracis powders – the latter grown from an RMR 1012 inoculum in late 200047. Dr. John Ezzell of USAMRIID, who acted as the FBI’s scientific advisor on anthrax, speaking from the floor at a seminar on the Amerithrax investigation on Nov. 29, 2010, described his production of (sterile) anthrax spores at that time, commenting that the powders produced in his lab were purer than any of the letter powders48 and that high purity was needed for determining unique mass spectral signatures49,50.
From 1999 to 2001, annual DOD budget item justifications for the DARPA project on detectors (called “sensors”) listed a plan to be carried out in FY 2001 to “evaluate methods for removing microencapsulation of disguised pathogens and/or sensing through the micro-encapsulation51”. The budget item justification sheet dated February 200252 listed that plan under “FY2001 Accomplishments”, indicating that at least one microencapsulated pathogen was studied by DARPA in 2001.
Also accomplished by DARPA in 2001 was the evaluation of “methods of cell stabilization for possible application to cell based sensors”. One such stabilization method is microencapsulation.
Concomitantly, the Chemical/Biological Defense Program (CBDP)53, covering work at Dugway, undertook in 1999 to identify and evaluate emerging threat agents by various means, and continued work on this project through FY2001, during which year they assessed the gaps in the threat agent data and the needs for improved simulants. Also in 2001, plans were made to initiate a program of synthesis, toxicology screening and characterization of new threat materials, to include Fourth Generation Agents [which include those altered for better survival54, e.g. by microencapsulation], and to initiate development of improved simulants for microencapsulated viruses and stabilized bacteria. Throughout this period the program provided controlled biosimulant aerosol challenges for Joint Service, DARPA, and DOE experimental equipment. Dugway tested DARPA’s detection equipment in 199955.
Because of Dugway’s emphasis on improved simulants and on stabilization/survivability, Dugway may be the source of the unique B. subtilis contaminant found in the early attack letters56. There is no evidence that Dugway or any other site was ever examined for the presence of the critical Bacillus subtilis strain. The strain was found to be a hypersporulator57, typical of the strains that originated at Dugway for use as simulants58. The Bacillus subtilis isolated from the NY Post letter was sequenced for the FBI59 and found to have greater than 98% identity60 with the widely-studied strain Bacillus subtilis 168 (sequenced in 199761), often used as a surrogate for Bacillus anthracis62 and as the standard model in studies of the molecular and genetic basis of Bacillus spore resistance to environmental stresses63. Research on Bacillus genetics was proliferating at the time of the anthrax attacks.
Dugway is the only place known to have made live, dry, weaponsgrade anthrax powder in the years before the attacks64. During the Amerithrax investigation the Dugway laboratory was the place the FBI asked to conduct experiments attempting to reverse-engineer production of the attack anthrax powders. The Dugway laboratory had supplied USAMRIID with most of the spores in flask RMR 1029, the putative parental source of the attack anthrax, in 1997.
A B. anthracis stock sample provided by Dugway to the FBI Ames anthrax repository tested positive in at least one of the four genetic assays used as indicators of relationship to the attack anthrax65, and the NAS committee believed that Dugway probably produced all four of the genetic markers used for assays66. A subcontractor67 working at Dugway in 2001, Battelle Memorial Institute, had twice received material from RMR 1029 from USAMRIID during 200168, upon receipt of which Battelle was given permission “to provide aliquots to other laboratory facilities for legitimate research purposes69”.
Battelle is well-known for aerosol expertise, and was working with dry anthrax simulant spores in the period before the attacks70. It would probably be difficult to distinguish whether Dugway or Battelle personnel were responsible for any anthrax-related work done at Dugway at that time.
A new clue
FBI documents released in 2011 show that the ten “reverse engineered” surrogate samples made at Dugway for the FBI contained a small but significant amount of tin. The FBI has never commented on this finding, nor on the tin content of the attack spores, and there is no evidence that the “reverse engineering” experiments71 or any other investigations were aimed at reproducing or explaining the presence of tin. No other Bacillus spore preparations except the attack spores contained tin (Table 2). The tin content of the Dugway samples, measured by ICP-OES in the FBI laboratory72, varied from 0.0033- 0.0266 wt% (average 0.0086 wt% tin). Omitting the two samples for which the FBI laboratory reported that the analytical results were most questionable73 gives a range of 0.0033-0.0084 (average 0.0058 wt%). In any case, the Dugway surrogates contained about 3-4% as much tin as was found in the Leahy spores (0.197 wt%) (data in Table 1). A “commercially available multi-element standard solution74” analyzed at the same time had 0.0069 wt% (69 ppm) tin75. The presence of tin cannot be attributed to ball milling; the two samples that were not ballmilled contained 0.0038 and 0.0064 wt% tin, consistent with the tin in the eight ball-milled samples. No tin was found in any culture medium components76 nor was tin exposure likely during growth or in any of the post-growth treatments77.
Random, trace contaminants are extremely unlikely to be found by ICP-OES in the concentration range observed in the Dugway samples. Analyses of a set of Bacillus thuringiensis samples at Sandia National Laboratory using TOF-SIMS (time of flight secondary ion mass spectrometry), a more sensitive method than ICP-OES78, found traces of tin, too small to quantify, in those samples79. One of the Sandia authors (J. Michael) is quoted80 as saying “we were surprised at first [at the traces of foreign elements], then we realized that the elements could have come from any number of sources − lab equipment, a residual cleaning solution, some other kind of contamination”. Contaminants that were not quantifiable by TOF-SIMS would certainly not be observable or quantifiable by ICP-OES81.
The small but real tin content found by the FBI laboratory in the Dugway samples appears to have been overlooked. Was it a remnant of tin used previously at Dugway in classified work? There is no evidence whatsoever to rule out the possibility that the attack samples had been legally made at Dugway82 or elsewhere. It must be recognized, of course, that making the spores is not synonymous with sending the letters. The “attack” anthrax could have been made for the use of US agencies or contractors conducting legitimate activities such as vulnerability and response assessment or testing detection devices such as DARPA’s. The letter sender(s) may have been one or more individuals who, whether legally or illegally, had access to the material at some point in the process.
Tin, found by the FBI in substantial amounts on the spore coats of the attack anthrax, has not been discussed or investigated. Tin is toxic to bacteria and therefore must have been acquired by the attack spores after their growth. The quantities are too high to be accidental contaminants. In approximately the amounts found83, tin is known to be a catalyst in the formation of silicone coatings for microencapsulation. Although there may be other possible explanations for the presence of tin, neither the FBI nor anyone else has put forward an alternative. The microencapsulation hypothesis is strengthened by the evidence that government-sponsored programs specifically involving microencapsulated pathogens were in progress at the very time of the attacks.
Silicon, in uniquely high amount, was also found on the coats of the attack spores. Although the FBI reported similarly high silicon content in a set of surrogate Bacillus anthracis spore powder preparations “reverse engineered” at Dugway, examination of the laboratory data (released after the case was closed) reveals that the silicon present was unquestionably a milling artifact and had no connection to the spores. Neither the Dugway surrogate spores nor any of the many other Bacillus spore preparations examined have ever been found to contain levels of silicon comparable to the attack spores (Table 3). These facts cast strong doubt on the FBI’s explanation that the silicon in the attack spores was naturally incorporated during growth. The FBI’s investigation of the silicon issue has been characterized by loose assumptions and sketchy data, raising fundamental questions about the investigation’s approach to the whole matter of additives in the attack spores.
The presence, shown by FBI analysis, of the two extraneous elements, tin and silicon, together in the attack spores favors the silicone microencapsulation hypothesis. Microencapsulation, a process that would require special expertise and sophisticated facilities, could explain the presence, location and amounts of both elements. At least two government programs, at DARPA and Dugway, had projects requiring microencapsulated pathogens or simulants. Both Dugway and Battelle, a sub-contractor there, had access to Bacillus anthracis from the presumptive parental flask RMR 1029. Both had the expertise to make anthrax spore powders, both – and perhaps other government- supported laboratories as well – could have made the attack spores legally for institutions conducting biodefense activities that required microencapsulated spores.
Other than the attack spores, no Bacillus preparations have ever been found to contain tin, with one exception: the set of ten Bacillus anthracis powders “reverse engineered” at Dugway in 2003 (Tables 1 and 3). These samples were produced by standard, well-described preparation procedures, with no known exposure to tin. Nonetheless, FBI analyses that became available after the case closed found a small but real amount of tin in the Dugway powders, about 4% of the amount in the attack spores, too large to be attributed to random, trace contamination; contamination from a previous use of tin appears to be the most likely explanation. The tin in the Dugway samples may be an indicator of previous, classified work with tin at Dugway to provide materials for biodefense activities.
Many US agencies were involved in biological antiterrorism activities at the time of the anthrax attacks. The Chemical and Biological Defense Program initiated in FY 2000 a “broad CB countermeasures program to enhance ability to recognize, prevent, respond to, mitigate, and recover from a CB terrorist incident84”. There had already been an epidemic of hoax anthrax letters. CIA scientists possessed Ames anthrax and were working with other government agencies and outside contractors [including Battelle85] in the defensive biowarfare program86; former biological weaponeer Bill Patrick wrote a classified paper about anthrax sent by mail, dated February 1999, for SAIC (Science Applications International Corporation) under CIA contract87. Nonetheless, the CIA issued a statement that they were unaware of any project to assess the impact of anthrax sent through the mail88; similarly, the FBI claimed to have been blindsided by a letter attack89.
The evidence indicates that live, microencapsulated Bacillus anthracis spore powders probably existed legitimately at the time of the attacks and were the likely sources of the anthrax in the attack letters. A number of individuals would have had access to such materials, whether legally or illegally. For the letter-sender(s), the presence of additives in the powders would have been irrelevant. For the FBI, a real investigation of the presence of additives may have been impossible without ‘off-limits’ intrusion into classified biodefense matters90.
1Evidence regarding the attack anthrax is discussed in: Hugh-Jones ME, Rosenberg BH, Jacobsen S (2011) The 2001 Attack Anthrax: Key Observations. J Bioterr Biodef S3: 001. (https://www.omicsonline.org/2157-2526/2157-2526-S3-001.pdf).
2Sets of FBI Documents identified by B (Batch) and M (Module) numbers were provided by the FBI to the National Academy of Sciences (NAS) Committee on the Review of the Scientific Approaches Used During the FBI’s Investigation of the 2001 Bacillus anthracis Mailings Investigation; the documents are listed in the NAS Report issued February 15, 2011, pp. 133ff, Index of Documents Provided by the Federal Bureau of Investigation; the documents were released to the public in February 2011 and were available from the NAS at that time.
6See Hugh-Jones et al. (op. cit.) for details of the proposed siliconization procedure, which requires moisture from within the spores and results in deposition on the spore coat, not on the exterior surface of the spore (the exosporium).
7Michael J, Kotula P (2009) [Sandia National Laboratory]. Elemental Microanalysis of Bacillus anthracis Spores from the Amerithrax Case, presentation in September 2009 to the National Academy of Sciences (NAS) Committee studying the FBI’s scientific approaches; also, FBI document B1M6 (Sandia National Laboratory Report), pp. 4ff. Tin and silicon were also found in extrasporular debris in the NY Post sample (see discussion in Hugh-Jones et al., op. cit.).
8Stewart M, Somlyo AP, Somlyo AV, Shuman H, Lindsay JA, et al. (1980) Distribution of calcium and other elements in cryosectioned Bacillus cereus T spores, determined by high-resolution scanning electron probe x-ray microanalysis. J Bacteriol 143: 481-491.
12Elemental analyses of the 10 samples are given in FBI Document B1M7 (FBI Laboratory Reports), Samples from Dugway, pp. 26-28; the data are repeated at B1M7, Elemental Analysis Summary, pp. 92-94. The ten Dugway samples studied by the FBI are labeled NDLB, NPLB,SDLB, SPLB, SDOB, NDOB, SPVB, NPVB, NDLM, NPLM, indicating their preparation methods (those ending with “B” were ball-milled; see FBI Document B1M13 (Dugway), p.72 or 27, for the significance of these labels).
14The mortar-and-pestle- milled samples, labeled NDLM (0.3 wt% silicon) and NPLM (0.2 wt%), can be compared to their ball-milled counterparts differing only in milling method: NDLB (5 wt% silicon) and NPLB (0.5 wt%). Note that NDLB was one of the two samples for which the analytical results were questionable, according to the FBI laboratory report (B1M7, p. 28); the other was SPVB (2 wt%). These two are the samples with the highest analytical results for silicon.
19Michael J, Kotula P (2009), op. cit.; also Sandia reports to the FBI in FBI documents B1M6 and B1M1 (in these reports, samples are coded and it is sometimes difficult or impossible to find their identities).
25B1M13 (Dugway), pp. 70 and 74-5. The two fermentation products, grown in 2003 at the same time and from the same stock as the agar samples, were stored until 2005, when they were irradiated and processed (methods not disclosed) under the designations “Lot 05AUG05” for one grown with Antifoam 204 and “Lot 01SEP05” for one grown with Antifoam C.
26FBI document B1M1(Technical Review Panels), Sandia report on elemental mapping of Amerithrax Samples, pp. 109, 110, 100, 106. The data for this sample, coded 040255-1, in Sandia’s Table on page 109 are the same as those for the “Dugway surrogate (fermentation using Leighton-Doi media)” in the NAS Report “Review of the Scientific Approaches Used During the FBI’s Investigation of the 2001 Anthrax Letters,” Feb. 15, 2011, pp 67 and 70, indicating that the FBI must have identified the coded material to the NAS Committee as a Dugway fermentation sample, but evidently did not specify which one.
31FBI Document B2M13 (Chemical and Physical Characteristics, Battelle), The Analysis of Surrogate Dry Powder Bacillus Spore Product, pp. 92ff. The “unwashed” preparation had a titer of 9×1011 cfu/g, an indication of its high quality (compare to the titer of the pristine Daschle sample, 2.1×1012, which had been determined at USAMRIID just after the material was received from the Capitol Police (FBI Document B1M2 (USAMRIID), Anaytical Test Report, pp. 36ff; note pp. 37 and 42). The titer of spores in the Leahy letter has never been reported.
32FBI Document B2M13 (Chemical and Physical Characteristics, Battelle), Summary of Sample Analyses, pp. 146ff. Analysis of the Leahy sample was done in February 2002, after the Leahy letter had been stored in a mail bag for a number of months; it is therefore possible that the Leahy spores were no longer in their original condition.
37FBI document B2M13 (Battelle), pp. 6, 31, 35; note that “Sample A” (p. 12) is sometimes incorrectly identified in the Battelle report as SPS.57.01, rather than SPS.57.03 (corrected on p. 105); “Sample B” is SPS.57.08 (p. 11).
38See, e.g., the Naval Surface Treatment Center, US Navy, Silicone Coatings. (https://www.nstcenter.biz/writeup.aspx?title=Silicone%20Coatings&page= TechResourcesSiliconeCoatings.html).
39Gerhardt P, Black SH (1961) Permeability of Bacterial Spores. II. Molecular Variables Affecting Solute Permeation. J Bacteriol 82: 750-760; Nicholson WL, Munakata N, Horneck G, Melosh HJ, Setlow P (2000) Resistance of Bacillus endospores to extreme terrestrial and extraterrestrial environments. Microbiol Mol Biol Rev 64: 548-572.
40See, for example, Biological and Toxin Weapons Today, ed. Geissler E, p. 32 (Oxford University Press, NY, 1986); and Report of the Secretary General on Chemical and Bacteriological (Biological) Weapons and the Effects of their Possible Use, p. 64 (UN Document A/75/75/Rev.1, 1969).
41Weapons of Mass Destruction: An Encyclopedia of Worldwide Policy, Technology, and History, ed. Croddy EA and Wirtz JJ, Volume 1: Chemical and Biological Weapons, ed. Croddy EA, pp. 184-5 (ABC-CLIO, 2004).
43Balkundi SS, Veerabadran NG, Eby DM, Johnson GR, Lvov YM (2009) Encapsulation of Bacterial Spores in Nanoorganized Polyelectrolyte Shells. Langmuir 25: 14011- 14016; Fakhrullin RF, Lvov YM (2012) “Face-lifting” and “make-up” for microorganisms: layer-by-layer polyelectrolyte nanocoating. ACS Nano 6: 4557-4564.
44Yang SH, Lee KB, Kong B, Kim JH, Kim HS, et al. (2009) Biomimetic encapsulation of individual cells with silica. Angew Chem Int Ed Engl 48: 9160-9163.
45Donlon M and Jackman J (1999) DARPA Integrated Chemical and Biological Detection System. Johns Hopkins APL Technical Digest 20: 320-325; Suter JJ (2005) Sensors and Sensor Systems Research and Development at APL with a View Toward the Future. Johns Hopkins APL Technical Digest 26: 350-355.
46Hathout Y, Demirev PA, Ho YP, Bundy JL, Ryzhov V, et al. (1999) Identification of Bacillus Spores by Matrix-Assisted Laser Desorption Ionization-Mass Spectrometry. Appl Environ Microbiol 65: 4313-4319. Note that dry B. anthracis Sterne spores from USAMRIID were used in developing this identification technique, which was employed in the DARPA project for analysis of aerosol samples collected on tapes.
47On August 28, 2000, forty ml were removed from flask RMR 1029 at USAMRIID “for DARPA mass spec project with JHU-APL” (inventory control sheet, www.vault.fbi.gov/ amerithrax, part 24, p. 8); see also, FBI interview believed to be of Dr. Joany Jackman, a major participant in the DARPA project working at USAMRIID under John Ezzell from 1997-2000 and then at Johns Hopkins APL: B. anthracis, grown from an inoculum (about 1012 cfu/ml) provided by Bruce Ivins and purified on a gradient, was used at USAMRIID for the aerosol work (https://vault.fbi.gov/Amerithrax/ part 21, pp 19-22; Jackman’s corroborating biographical details are at https://www.hopkinsmedicine.org/ medicine/std/team/Jackman.html).
51The plan is contained in DOD Budget Justifications issued in February 1999, February 2000 and June 2001: DOD RDT&E Budget Item Justification Sheet (R-2 Exhibit) for BA2 Applied Research, R-1 Item Nomenclature Biological Warfare Defense PE 0602383E, R-1 #14, dated February 1999; DOD RDT&E Budget Item Justification sheet (R-2 Exhibit) for BA2 Applied Research, R-1 Item Nomenclature: Biological Warfare Defense PE 0602383E, R-1 #15, dated February 2000; DOD Amended Budget Submission, RDT&E Budget Item Justification sheet (R-2 Exhibit) for BA2 Applied Research, R-1 Item Nomenclature: Biological Warfare Defense PE 0602383E, R-1 #16, p. 93, dated June 2001.
53Chemical/Biological Defense Program projects cited here are found in the following documents: DOD CBDP Budget Item Justification Sheet (R-2A Exhibit) for BA2 - Applied Research, 0602384BP Chemical/Biological Defense, Project CB2, dated February 1999; and DOD CBDP Budget Item Justification Sheet (R-2A Exhibit) for BA2- Applied Research, 0602384BP Chemical/Biological Research, Project CB2, dated June 2001.
58Gibbons HS, Broomall SM, McNew LA, Daligault H, Chapman C, et al. (2011) Genomic signatures of strain selection and enhancement in Bacillus atrophaeus var. globigii, a historical biowarfare simulant. PLoS One 6: e17836.
61The genetic sequence of B. subtilis 168 became available in 1997: Kunst F, Ogasawara N, Moszer I, Albertini AM, Alloni G, et al. (1997) The complete genome sequence of the gram-positive bacterium Bacillus subtilis. Nature 390: 249-256.
62Nicholson WL, Galeano B (2003) UV resistance of Bacillus anthracis spores revisited: validation of Bacillus subtilis spores as UV surrogates for spores of B. anthracis Sterne. Appl Environ Microbiol 69: 1327-1330.; Greenberg DL, Busch JD, Keim P, Wagner DM (2010) Identifying experimental surrogates for Bacillus anthracis spores: a review. Investig Genet 1: 4.
63Nicholson WL, Munakata N, Horneck G, Melosh HJ, Setlow P (2000) Resistance of Bacillus endospores to extreme terrestrial and extraterrestrial environments. Microbiol Mol Biol Rev 64: 548-572.
65FBI Document B2M10 (Statistical Analysis), especially Appendix V of the Report on Statistical Analysis, where the names of some specific laboratories that submitted repository samples of interest are handwritten next to the sample data: DPG (Dugway Proving Ground); BMI (Battelle Memorial Institute); DRES (Defense Research Establishment, Suffield, Canada); NMRC (Naval Medical Research Center); USAMRIID. See also the National Academy of Sciences (NAS) Report “Review of the Scientific Approaches used during the FBI’s Investigation of the 2001 Anthrax Letters,” February 15, 2011, pp. 110-112, and FBI Document B2M10, p. 25. From information in these documents it appears that seven repository samples from USAMRIID and one from Battelle tested positive in all four genetic assays; one sample each from USAMRIID and Northern Arizona University (an FBI contractor in the anthrax case) had three positives; a samples from DRES had two positives; and samples from six other laboratories, including Dugway and the Naval Medical Research Center plus four unnamed laboratories, tested positive in one or two assays, as did additional samples from some of the laboratories already mentioned.
66NAS Report (op. cit.), p. 108 gives reasons why all four markers probably originated at Dugway; the Report also discusses the probability of false negatives and presents a Table of assay results on 30 repeat samplings of flask RMR 1029 (the putative parental source of the attack anthrax) as an illustration (p. 117).
67Personal communication January 11, 2002 from David Lore, Columbus Dispatch Science Reporter, author of article “Labs deny use of letter anthrax: Powdered spores not part of stocks, Battelle official says” in Columbus Dispatch, January 9, 2002.
68FBI Document B3D16. One Ames repository sample from Battelle tested positive in all 4 assays; another from Battelle, known to have originated from an RMR 1029 sample, tested positive in 2 of 3 assays (FBI Document B2M10, p. 25).
69Maureen Stevens et al. vs United States of America: Notice of Errata, Document 162, entered on FLSD Docket 07/19/2011, submitted by the Defendant United States in US District Court, Southern District of Florida, Case Number: 03-81110-CIV-Hurley/Hopkins.
72Elemental analyses of the 10 samples, measured by ICP-OES in an FBI laboratory, are given in FBI Document B1M7 (FBI Laboratory Reports), Samples from Dugway, pp. 26-28; the data are repeated at B1M7, Elemental Analysis Summary, p. 93.
73Problems in the FBI laboratory’s analysis of the Dugway surrogates, particularly sample NDLB (5 wt% silicon, 0.0265 wt% tin), and sample SPVB (2 wt% silicon, 0.0132 wt% tin) are stated in FBI Document B1M7, p. 28 (see Table 1).
75FBI Document B1M1 (Technical Review Panels), p. 83, gives the elemental analysis of the Standard, together with averages for the elements in the ten Dugway samples (presented at a Chemistry Review Panel in August 2005).
78See Table 1, note c.
79Brewer LN, Ohlhausen JA, Kotula PG, Michael JR (2008) Forensic analysis of bioagents by X-ray and TOF-SIMS hyperspectral imaging. Forensic Sci Int 179: 98-106.
81For sensitivities of the analytical methods see Table 1.
82The finding, via stable isotope analysis (B1M9, p.44), that Dugway water is unlikely to have been used to grow the attack spores is probably not relevant; Dugway’s report (B1M13) on preparation of the surrogate samples mentions the use of “sterile water for irrigation” and “sterile water for injection,” which are generally purchased in small bottles from distant providers. The FBI laboratory, in analysing media components (B1M7), included “Baxter sterile water for irrigation”.
90An appropriately targetted investigation would seek to determine whether Bacillus anthracis had been microencapsulated prior to the letter attacks, and, if so, by whom, where, when, and the amounts, strains and dispositions of the resulting materials. Parts of such an investigation might still need to be classified.