''I know there are a number of people who would love an excuse to get after Iraq,'' said a top federal scientist involved in the investigation.
From the start, agents searched for clues in domestic industry, academia and terror groups. But while investigators were racing to link the Ames strain to Iraq, they have only recently begun examining government institutions and contractors in this country that have worked with that strain for years.
In hunting for a culprit in the attacks that killed five people, agents have chased tens of thousands of tips in the past two months and conducted thousands of interviews, law enforcement officials said.
They have traced prescriptions for the antibiotic Cipro, on the chance the perpetrator took the drug to guard against the disease. They have also checked the language and block-style handwriting on letters sent with the anthrax against digital databases of threatening letters maintained by the Federal Bureau of Investigation, Secret Service and Capitol Police.
But officials said no likely suspects have emerged, and they are settling in for what they fear could be a long haul.
The most promising evidence is still the anthrax itself, which federal scientists and contractors are studying for clues to its origin. The government tried to find links to Afghanistan and Iraq in the substance as well.
One discovery early in the inquiry seemed to undercut the foreign thesis. The anthrax used in the first attack, in Florida, and in subsequent attacks turned out to be the Ames strain, named after its place of origin in Iowa. While investigators found that this domestic variety of anthrax had been shipped to some laboratories overseas, none could be traced to Baghdad.
Nevertheless, government officials continued pushing the Iraq theory, scientists and officials involved in the inquiry said. They saw an intriguing clue in reports that Iraq had tried hard to obtain the Ames strain from British researchers in 1988 and 1989, raising suspicions that it had eventually succeeded.
Federal scientists hunted down records and biological samples from an investigation of Iraq's biological arms program, which was conducted by the United Nations in the 1990's. Those samples were analyzed in laboratories run by two biologists, Paul S. Keim of Northern Arizona University and Paul J. Jackson of the Los Alamos National Laboratory, in New Mexico.
But in the end few samples from Iraq's arsenal were found, and those that were turned out to have nothing in common with the Ames strain, officials said.
A different line of inquiry sought to re-examine seven anthrax strains that the world's largest germ bank, the American Type Culture Collection, in Manassas, Va., sold to Iraq in the 1980's, before the government banned such exports.
None of the strains were identified as Ames. But scientists inside and outside the government speculated that mislabeling might have inadvertently put the potent germ in Baghdad's hands. More laboratory tests were ordered.
Raymond H. Cypess, president of the germ bank, said recent investigations had disproved the mislabeling idea. ''We never had it,'' he said of the Ames strain, ''and we can say that on several levels of analysis.''
The Iraq inquiry also looked for chemical clues. An early focus was bentonite, a clay additive that is one of the few substances identified publicly that can help reduce the static charge of anthrax spores so they float more freely and potentially infect more people.
Richard O. Spertzel, a retired microbiologist who led the United Nations' biological weapons inspections of Iraq, told investigators that Iraq had explored using bentonite in its germ weapons programs. But Maj. Gen. John Parker of the Army's biological research center at Fort Detrick, Md., said in late October that tests had turned up no signs of aluminum -- a main building block of bentonite.
''If I can't find aluminum,'' General Parker told reporters, ''I can't say it's bentonite.''
Despite the scientific findings, the sophistication of the anthrax found in the letter mailed to Senator Tom Daschle, the majority leader, has kept Dr. Spertzel and others convinced that Iraq or another foreign power could be behind the attacks.
Richard H. Ebright, a microbiologist at Rutgers University who closely follows the anthrax inquiry, recently said that the Baghdad thesis ''should not be dismissed as a desperate reach for a casus belli against Iraq'' and is still worth investigating.
Publicly, White House officials have made no mention of the failure to find an Iraqi connection, but they have noted the inquiry's intensified focus on the United States. ''The evidence is increasingly looking like it was a domestic source,'' the White House Press secretary, Ari Fleischer, said on Monday.
Tom Ridge, the director of homeland security, said in a statement that he initially assumed that the culprits were foreigners. ''Like many people, when the case of anthrax emerged so close to Sept. 11, I couldn't believe it was a coincidence,'' Mr. Ridge said. ''But now, based on the investigative work of many agencies, we're all more inclined to think that the perpetrator is domestic.''
It remains unclear whether the focus on Iraq diverted investigators from the domestic inquiry. But some scientists say a decision made early on suggests that it might have.
In early October, the F.B.I. raised no objections when officials at Iowa State University, where the Ames strain was discovered, said they planned to destroy the university's large collection of anthrax spores for security reasons. Many biologists now say that step may have destroyed potential genetic clues to the culprit's identity.
Two months later, the investigation is largely focused in the United States. As the scientific inquiry into the anthrax itself continues, the F.B.I. is also employing more traditional forensic and investigative techniques to find out who sent the lethal letters.
Agents have compiled lengthy lists of who might have manufactured, tested, transported or stored anthrax. They have questioned manufacturers and marketers of biochemistry equipment and specialized machinery needed to make the material. They have inspected scientific literature, which could provide clues about who has knowledge to make anthrax.
But few clues have emerged. So far only three letters -- those sent to NBC, The New York Post and Mr. Daschle -- have been analyzed. A fourth letter, sent to Senator Patrick J. Leahy, Democrat of Vermont, is undergoing painstaking analysis by a number of laboratories, officials said.
All of the letters were photocopies and none appeared to contain any fingerprints. The plastic tape on the envelopes was a mass marketed variety. The paper on which the letters were written was an average size. The envelopes were prestamped and widely available. The marks left by the photocopier have been carefully studied but have revealed no clues.
One senior official said nothing the investigators had found had led to anyone who might remotely be called a suspect. Several people who seemed to fit the F.B.I.'s profile of a science loner have been aggressively investigated, but no one has emerged as a serious subject.
''Still, the more you are out there, the more things bubble up,'' the official said. But asked whether recent news reports of a possible suspect in the case were true, the official replied, ''I only wish that was true.''
Some tips have seemed encouraging, but only for a time.
''We run out every lead and we give these people a real hard look and real hard shake before we take them off the screen,'' the official said. ''There have been people who we have placed a little higher priority on than others.'' But then they fall off.
Some senior Bush administration officials have begun to worry privately that the case might take decades to solve, likening it to the Unabomber investigation that baffled investigators for nearly 20 years until David Kaczynski became suspicious of his brother Theodore and alerted the F.B.I.
Investigators have used various strategies to find suspects but have often been frustrated. When they tried to track down people who had sought prescriptions for Cipro in the weeks before the anthrax mailings, the effort quickly bogged down. ''Do you know how many people take Cipro in this country?'' an exasperated official said, explaining that Cipro is used to treat a variety of ailments.
Investigators also said they were continuing to examine the possibility that the culprit might have purchased stock in the company that makes Cipro in an effort to profit from the attacks.
The newest front in the search for culprits is the examination of government research institutions and contractors. The reason to look there is plain: Some of them have the Ames strain and know how to turn it into the kind of deadly powder used in the attacks.
But that has added yet another complication to the already challenging inquiry. After all, investigators have relied on these same experts for scientific advice from the earliest days of the investigation, back when Iraq was a prime suspect.
''It puts us in a difficult position,'' one senior law enforcement official said. ''We're working with these people and looking at them as potential suspects.''Continue reading the main story
Anthrax is a bacterial disease caused by the spore– forming Bacillus anthracis, a Gram-positive, rod-shaped bacterium (see chapter 6), the only obligate pathogen in the large genus Bacillus.
2.1.2. Sporulation and germination in the environment
(See also section 6.5.)
220.127.116.11. Nature of the spore
The bacterial spore is a resting form of the organism. The essential parts of the vegetative cell (the vegetative cell genome precursors in a dehydrated state, small acid-soluble proteins (SASPs) that bind to and protect the DNA as well as acting as amino acid sources during germination, and specific organic acids which act as energy sources for germination) lie in a “core” surrounded by a thick protective cortex, spore coats and a proteinaceous exosporium. The inner layer of the cortex is the precursor of the vegetative cell wall and the receptor for germinants lies in the interface between the cortex and spore coats. The proenzyme of a germination-specific cortex-lytic enzyme (GSLE) is activated when the germinant attaches to the receptor. The active GSLE allows uptake of water by the cortex for initiation of germination. The cortex also plays a role in the resistance of the spore to heat. The spore coats, which represent approximately 50% of the volume of the spore, supply the first line of resistance to chemicals and physical disruption. The function of the loose-fitting exosporium is not known but may have a role in adhesion to surfaces.
Spores are markedly resistant to biological extremes of heat, cold, pH, desiccation, chemicals (and thus to disinfection), irradiation and other such adverse conditions. The organism can persist in the spore state for long periods of time awaiting the moment when conditions favour germination and multiplication. The ability of anthrax spores to persist in the soil and other environments for decades is legendary (Titball et al., 1991; Quinn & Turnbull, 1998; de Vos & Turnbull 2004; section 18.104.22.168).
22.214.171.124. Sporulation and survival
The rate and extent of sporulation by vegetative cells shed from infected animals is affected in a complex manner by the environmental conditions into which they fall. Temperature, humidity, water activity (aw, available water within the microenvironment), pH, oxygen availability, sunlight and the presence of certain cations, particularly Mn++, are all influencing factors. Maintenance of the organism in the spore state and thus its persistence in the environment is also influenced by water activity, temperature, pH and the presence of nutrients and germinants (see section 126.96.36.199).
Although, in the laboratory, the vegetative forms of B. anthracis grow and multiply readily on or in normal laboratory nutrient agars or broths, the evidence is that, in natural circumstances, they are more “fragile” than the vegetative forms of other Bacillus species, dying more spontaneously in simple environments such as water or even milk, and being more dependent on sporulation for species survival (Turnbull et al., 1991; Bowen & Turnbull, 1992; Lindeque & Turnbull, 1994). See also sections 188.8.131.52, 184.108.40.206, and Annex 6.
220.127.116.11. Subsidiary cycles (“Sporulate or die”)
Spores will germinate outside an animal if conditions permit. For bacterial spores in general, these conditions include temperatures between about 8 °C and 45 °C, pH between 5 and 9, a relative humidity greater than 96% and the presence of adequate nutrients (Sussman & Halvorson, 1966). The extent to which they may then germinate, multiply as vegetative bacilli and resporulate, setting up subsidiary cycles in the environment, remains a topic of debate (Titball et al., 1991). While it has been shown that environmental cycling can be induced experimentally (Minett & Dhanda, 1941; Titball et al., 1991; Anon., 2004; see also section 18.104.22.168), the level of nutrient required for this to become possible is probably not reached very frequently under natural conditions. If the spores germinate, the emergent vegetative cells might generally be expected to die spontaneously (section 22.214.171.124) or as a result of competition from soil microflora (Sterne, 1959), or both. Then, unless further cases occur, B. anthracis is eliminated within a period of years (Sterne, 1959; section 126.96.36.199).
The conserved nature of B. anthracis as a species (see section 2.3) also belies the concept of frequent environmental cycling. Overall, it seems that the fragile vegetative forms shed by the dead animal die rapidly in most environmental conditions and depend (i) on sporulation in a proportion of their population for their survival, and (ii) on their next animal host for multiplication. This can be described as “sporulate or die”. For this reason, for all practical purposes, B. anthracis can be regarded as an obligate pathogen.
The rarity of environmental cycling is not universally accepted and Kaufmann (1990) believed that certain features of naturally-occurring anthrax are better explained by a pattern of B. anthracis multiplication in soil than by mere persistence of spores. He considered that the frequent association of outbreaks with rain ending a period of drought – “the onset of a distinct rainy season when animals disperse to graze the newly emergent vegetation” – is best explained by a concomitant burst of growth of B. anthracis in the soil. The conclusions of a recent study (Saile & Koehler, 2005) were that B. anthracis spores could germinate and establish populations of vegetative cells in the rhizosphere of grass plants, even supporting horizontal gene transfer. Certainly the fact that contamination levels at a carcass site can sometimes remain apparently undiminished for years after the death of the animal despite exposure of the site to wind, rain and sunlight (Turnbull et al., 1998) is difficult to explain in terms other than that localized multiplication has occurred. De Vos (1990) and Dragon et al. (1996) believed that the ecoepidemiological patterns of anthrax in, respectively, the Kruger National Park, South Africa, and bison in northern Canada were in line with the “incubator area” hypothesis of Van Ness (1971). (See section 188.8.131.52.)
184.108.40.206. Temperature, aw, pH and calcium
There is a surprising shortage of reliable data on germination, multiplication and sporulation of B. anthracis under different conditions of temperature, water activity and pH. Howie (1949) noted that a nutrient broth culture of B. anthracis formed spores within 6 hours at 37 °C as assessed by heat resistance and that, in an ice-chest (0–4 °C) bacilli did not spore but died in 6 days. In a recent study (Reyes, Turnbull & LeBron, unpublished data, 2006), vegetative cell preparations of five B. anthracis and two B. cereus strains were transferred to sporulation agar slants that were held at selected temperatures in refrigerated water baths. The critical temperature for three of the B. anthracis and the two B. cereus strains was found to lie between 9 °C and 12 °C. At 12 °C they grew and sporulated (as determined by heat resistance), albeit requiring up to 2 weeks before spores were detectable, while at 9 °C, none grew or sporulated and numbers declined to unrecoverable. The other two B. anthracis strains declined and failed to sporulate at 12 °C.
According to the experience of Turnbull (personal communication, 2002), spores, again as identified by heat resistance, appear in 37 °C blood agar plate cultures within 6–8 hours.
The laboratory study of Davies (1960) demonstrates the dramatic extent to which temperature and relative humidity affect sporulation and how temperature affects germination. Basically, as assessed from stained smears:
at 37 °C and 100% RH, sporulation was first apparent at 6 hours and was complete by 12 hours;
at 37 °C and 90% RH, spores were first seen at 12 hours and sporulation was complete at 16 hours;
at 37 °C and 80% RH, these time points were delayed to 16 and 24 hours;
at 37 °C and RH below 50%, variable sporulation was seen at 34 hours;
at 26 °C with RH 100% and 90%, sporulation was first seen at 24 hours and was complete at 28 hours; the corresponding time points were delayed with decreasing RH but at all RH values down to 20%, sporulation was complete by 60 hours.
In laboratory terms, germination is a much faster process than sporulation and, as assessed by loss of heat resistance, is apparently complete 2–10 minutes after exposure of the spores to germinants, such as alanine, tyrosine or adenosine (Sussman & Halvorson, 1966). In Davies’s study (1960) of the influence of temperature (as judged by visualization of 10% vegetative forms):
no germination occurred at 46 °C and 18 °C; while at
44 °C, germination was first seen at 12 hours and was fully present by 16 hours;
42 °C, germination was first seen at 3 hours and was fully present by 6 hours;
39 °C, germination was first seen at 2 hours and was fully present by 3 hours;
37 °C, germination was first seen at 2 hours and was fully present by 6 hours;
30 °C, germination was first seen at 4 hours and was fully present by 8 hours;
25 °C, germination was first seen at 10 hours and was fully present by 12 hours;
20 °C, early germination was not seen but was fully present at 16 hours.
Lindeque & Turnbull (1994) inoculated blood taken from wild animals just after death from anthrax into soil and water. In sandy soils, pH 7.5 to 8.2, sporulation (as detected by surviving 62.5 °C for 15 minutes) was first apparent at about 10 hours and was complete by 24 hours; in chalky karstveld soils (pH 7.9–8.1), sporulation only occurred reluctantly (total counts fell extensively and spore counts only equalled total counts at 96 hours) but was first seen at about 4–5 hours. The specific ambient temperatures are not given but appear to have ranged between about 20 °C and 30 °C over the course of 24 hours. In waterhole water similarly inoculated in the laboratory with blood collected just after death and held at 27–29.5 °C, sporulation had commenced by 15 to 24 hours; spores accounted for 100% of the total count at 68 hours. Although the pH of the water before addition of the blood was > 9, after addition of the blood it was effectively neutral.
Further comments are made in section 2.2.3 on the importance of temperature and relative humidity for the ecology of anthrax in relation to climate.
The belief of an association of high soil pH with “favourable sites” for anthrax persistence dates back to at least the statement of Higgins (1916) that “a suitable soil must be slightly alkaline”. Minett & Dhanda (1941) found that multiplication in sterilized soil of natural pH 6.69 was enhanced by the addition of slaked lime (CaCO3 9.45%, CaO 68.32%). A water content of ≥ 10% (optimum 25%) of dry soil and a temperature ≥ 17 °C (optimal 30 °C) were necessary for growth to occur (see section 220.127.116.11 for discussion on multiplication in soil under natural conditions). Analogy with information on B. cereus (Sussman & Halvorson, 1966) suggests that the rate and yield of germination may be influenced by temperature in a manner that varies with pH and that spores will not germinate at a pH of < 5, a temperature of < 8 °C and relative humidity of < 96%. Titball & Manchee (1987) showed that germination of Vollum strain spores was minimal at 9 °C and that the optimum germination temperature in the presence of germinant L-alanine was 22 °C.
Van Ness & Stein (1956) and Van Ness (1971) cite the results of Minett & Danda and those from a thesis by Whitworth (1924) as well as their own analysis of outbreaks in relation to pH in support of their hypothesis that “anthrax occurs in livestock that live upon a soil with a pH higher than 6.0, and in an ambient temperature above 15.5 °C”. On the basis of this, they mapped “the distribution of soils considered capable of supporting anthrax” in the USA (Van Ness & Stein, 1956). The “incubator area” hypotheses of Van Ness (Van Ness, 1971) have been cited frequently in the literature since the publication of his paper (it is sometimes forgotten that these were theories and were not demonstrated scientifically). These hypothesized “incubator areas” are depressions in calcareous or alkaline locations, collecting water and dead vegetation which, in turn, provide a medium suitable for germination and multiplication of anthrax spores.
The association of calcium and anthrax “favourable sites” seems to have been recognized in the Russian Federation over a century ago within the “chernozem concept”, the process of soil calcification from an underlying calciferous parent material, studied by Dokuchaev as early as the late 19th century (Smith, personal communication, 2003). In a retrospective analysis of anthrax in the USA over the 100-year period 1900–2000 (Smith, personal communication, 2003) found that the occurrence of the disease in livestock at the county level is firmly linked to chernozem soils (defined as a calcium rich, neutral-to-alkaline soil suitable for prairies, grasslands and cereal grain cultivation). Livestock anthrax mortality rates in areas with such soils, analysed for the period 1945–1955, were > 21-fold greater than other areas (Smith et al., 2000). Smith (personal communication, 2003), in the first quantitative study on the possible link between soil calcium and pH, and anthrax ecology carried out in the Kruger National Park, South Africa, also noted that areas where soil calcium was greater than 150 milliequivalents per gram and pH was greater than 7.0 had a greater than 7 times higher anthrax death rate than areas lacking these parameters. Furthermore Smith et al. (2000) demonstrated that different genetic types of B. anthracis may be “restricted” in their geographical ranges by an adaptation (or lack of it) to differing ranges of soil calcium and pH.
Calcium is integral to the dehydration of vegetative cell genome precursors necessary for effective long-term storage in the spore, and Dragon & Rennie (1995) proposed that exogenous calcium in calcium-rich soils may act as a buffer to leaching of calcium from the spore core, thereby enhancing preservation of the spore.
It may be that there is a variable ability to survive within a population of spores in the environment, but that a moist alkaline (pH 9) and calcium-rich environment will favour spore survival for long periods. This seems to be supported by evidence from observations in wildlife parks and reserves. It is possible that shorter-term survival in agricultural environments – as evidenced by the general experience over decades that eradication of the disease from an affected area can be achieved with a vaccination programme of about three years’ duration – may be at least partially explained in terms of germination and failure to survive and resporulate at the lower pH of most agricultural soils. (See also section 18.104.22.168.)
22.214.171.124. Physical movement of spores
Some differences of opinion are apparent in the literature as to the mobility of anthrax spores in the environment. Contaminated carcass sites in the generally dry, dusty soils of the Etosha National Park, Namibia, are noteworthy both for how discrete they remain and for how high levels of contamination can persist there for years despite seasonal winds (bacteriologically proven to move some of the spores) and rain, sometimes severe in nature (Lindeque & Turnbull, 1994; Turnbull et al., 1998b). It was proposed that strong attachment of spores to soil might be responsible for this, at least in part. Similarly, in the almost 40 years between the trials on Gruinard Island and predecontamination sampling, there had been negligible, if any, spread of contamination from the original detonation and testing sites (Manchee et al., 1990) although the highest levels of spores were recorded below the soil surface (Manchee et al., 1983). On the other hand, de Vos (1990) believed that epidemics of anthrax in the Kruger National Park arose from the concentration of spores in depressions as a result of run-off into the depressions following rain, reflecting the “incubator area” theory of Van Ness. The incubator and concentrator hypotheses have also been considered in relation to the persisting anthrax in the Great Slave Lake region of Canada (see section 2.2.5). As noted in section 126.96.36.199, however, the concentration/incubator theories have not been confirmed bacteriologically.
Historically, in the Netherlands and the United Kingdom (at least), the main enzootic livestock areas traditionally lay “downstream” from tanneries and the implication has been that watercourses have carried contaminated tannery effluent, depositing them in ditches and streams a few kilometres away. The occasional incidents or outbreaks that still occur in these areas are frequently associated with recent site disturbance (dredging or digging). Similarly, the large Australian outbreak in 1997 was believed to have been initiated by the movement of earth and the disturbance of old anthrax graves associated with the levelling of irrigation land on the index property. In support of the belief that this was an explosive point-source outbreak was the finding that all the isolates collected across the outbreak area were the same strain (Keim et al., 2000). Anthrax had not been recorded previously in the outbreak area since record-keeping began in 1914 (Seddon, 1953). One example was the outbreak in Washoe county, Nevada, in August 2000, one of the “persistent pockets” of anthrax in the USA, which followed ditch-clearing work (Hugh-Jones, 2000). Another good example is given in the report of Turnbull et al. (1996): two ponies died after grazing in a field which had been scarified and re-seeded and in which a bullock that had died of anthrax had been buried 50 years before. The first pony was not examined for cause of death but the second was confirmed as an anthrax case. Bacteriological mapping of the field pinpointed the burial site (the ponies had died in their stable). Although the circumstantial evidence for an association between soil disturbance and outbreaks is strong however, it is almost invariably anecdotal and unsupported by bacteriological evidence. The corollary is that farming operations frequently involve soil disturbance and that anthrax infection does not result.
The relevance of the infectious dose to the importance of physical movement of B. anthracis in transmitting the disease is addressed in chapter 3.
188.8.131.52. Persistence of anthrax spores
The ability of anthrax spores to remain viable for very long periods has become almost legendary but there is little well-documented information on this. Jacotot & Virat (1954) found anthrax spores prepared by Pasteur in 1888 to still be viable 68 years later, and Wilson & Russell (1964) reported that anthrax spores had survived in dry soil for 60 years. In 1992 Bowen & Turnbull (Turnbull, personal communication, 2002) found B. anthracis in samples of the plaster and lagging of London’s King’s Cross railway station roof space and dated this to infected horse hair used to bind the plaster when the building was constructed a century before (only in 1908 was the Horse Hair Act passed in the United Kingdom, requiring the sterilization of horse hair used in buildings). The longest survival claim is probably that of de Vos (1990) who recovered anthrax spores from bones retrieved during archaeological excavations at a site in the Kruger National Park, South Africa, that were estimated by carbon-dating to be 200 ± 50 years old. The condition that appears most to favour long survival is dryness. However, other conditions that discourage spores from germinating may also play a role in persistence. Manchee et al. (1990) noted that data from annual sampling between 1946 and 1969 of the contaminated site on Gruinard Island (where an estimated 4 x 1014 spores were dispersed by explosive means in 1942 and 1943 during the Second World War) predicted a decay to undetectable by 2050. The island, off the west coast of Scotland, has a wet cool climate and a highly organic soil with pH of 4.2–4.7; the low pH is probably the main factor inhibiting germination (Titball et al., 1991). Where germination occurs, the temperature may not permit growth and resporulation and the emergent vegetative forms probably die.
In contrast, Turnbull (personal communication, 2002) reports not infrequently finding that B. anthracis and other Bacillus species stored on agar slopes had died, particularly when the lids of the bottles containing the slopes had become loose, allowing the agar to dry out.
Turnbull et al. (1992b) observed that environmental isolates of B. anthracis from sites with a history of anthrax spore contamination in the distant past quite frequently lacked pXO2 and, less frequently, both pXO1 and pXO2. They hypothesized that, under stressful environmental conditions such as within sewage or in the harsh semidesert circumstances of the Etosha National Park in Namibia, B. anthracis could spontaneously lose one or both virulence plasmids. When first cultured, the Kings Cross isolates were a mixture of capsulating and non-capsulating cells, possibly representing a population in the transition stage. However, the precise causes and events responsible for the loss of one or other of the plasmids and the time or times during the germination, outgrowth, multiplication and resporulation at which these events occur is not known.