Journal of Applied Bacteriology 1992, 72, 21-20
Bacillus anthracis but not always anthrax P.C.B. Turnbull’, R.A. HutsonlY2,Mandy J. Ward133,Marie N. Jones’, C.P. Quinn’, N.J. Finnie’, C.J. Duggleby“, J.M. Kramer5 and J. Melling‘ Divisions of ’Biologics and 4Biotechnology, Public Health Laboratory Service Centre for Applied Microbiology & Research, Porton Down, Salisbury and 5F00d Hygiene Laboratory, Central Public Health Laboratory, London, UK 3634/04/91: accepted 1 June 1991
P.C.B. T U R N B U L L , R.A. HUTSON, M.J. WARD, M . N . JONES, C.P. QUINN, N.J. FINNIE,C.J. DUGGLEBY, J.M. K R A M E R A N D J . MELLING. 1992. Gram-positive bacilli isolated during epidemiological investigations which, on the basis of conventional tests, resemble Bacillus anthracis but which fail to produce the capsule or to induce anthrax in test animals have long been dismissed in clinical and veterinary laboratories as B . cereus or simply as unidentified Bacillus spp. and thereupon discarded as inconsequential. In this study, the application of newly available DNA probe, polymerase chain reaction and specific toxin antigen detection technology has revealed that a proportion of such strains are B . anthracis which lack the plasmid carrying the capsule gene (pX02). While these techniques cannot, of course, be used to confirm the identities of strains resembling B . anthracis but which also lack the plasmid carrying the toxin genes (pXOl), the likelihood that these also are bonaJde B . anthracis becomes more acceptable. (As yet no naturally occurring p X O l - / 2 + strains have been found.) At this point, the significance of the presence of such avirulent forms of B. anthracis in specimens can only be a subject for speculation, but the possibility that they may be indicators of virulent parents somewhere in the system being examined must be considered.
INTRODUCTION
The classification of Bacillus anthracis and its taxonomic separation from the closely related B . cereus has long posed a dilemma for bacteriologists. While, in practical terms, it has never been a problem since M’Fadyean published his capsule staining method in 1903 (M’Fadyean 1903a,b) to identify the typical large square-ended capsulating Grampositive bacillus isolated from a human or animal with classical symptoms of anthrax as B . anthraczs, it is clear from reports of strains given names such as B . anthracozdes (or simply ‘anthracoids’), B . anthracis similis and B . pseudoanthracis, that, even in the early days of bacteriology, problems of identity arose (Ivanovics & Foldes 1958). The situation was summed up by Stein (1943) in his statement that ‘Pathogenicity constitutes the principal point of difference between typical strains of B . anthraczs and those of anthrax-like organisms’. Present addresses :’AFRC Instrtute of Food Research, Shinjield, Berkshire RC2 9AT; ’Department of Microbiology, University of Southampton, General Hospital, Southampton SO9 4XY, U K . Corresponding author: P.C.B. Turnbull, Division of Biologics, Public Health Laboratory Service Centrefor Applied Microbiology 6 Research, Porton Down, Salisbury, Wiltshire SP4 OJC,U K .
The beginnings of true order in the classification of Bacillus species was first brought about by Smith et al. in 1952 with further clarification in the related manual of Gordon et al. (1973) which remains the basis of today’s identification systems for this genus. On the reasoning that, when a strain of B . anthraczs lost its virulence it became indistinguishable from B . cereus, Smith et al. concluded that ‘ B . anthracis is taxonomically a pathogenic variety of B . cereus’. Gordon et al., despite objections that had obviously been raised in the intervening years, felt justified in retaining this status and listed the anthrax bacterium as B . cereus var anthracis. At the grass roots level, the issue has remained unresolved. Recent considerable advances in the understanding of the pathogenicity of B . anthracis at the molecular and genetic levels also have not immediately answered the taxonomic questions. The two known virulence factors, the toxin and the capsule, are encoded on separate plasmids, pXOl and pX02 respectively (Robertson et al. 1990) and the loss of either or both of these results in an avirulent form which is essentially indistinguishable from B . cereus by conventional biochemical tests. As was the case in the first half of the century, little difficulty is now experienced in identifying as B . anthracis
22 P.C.B. T U R N B U L L E T A L
the typical non-haemolytic, non-motile, penicillin- and gamma phage-sensitive, capsulating Gram-positive bacilli isolated from animals and (rarely nowadays in the West, of course) humans with symptoms of anthrax. T h e difficulties arise in epidemiological or other studies concerned with detection of B. anthracis in environmental samples such as dust, soil, water, sewage sludge or feedstuffs. Frequently in these types of samples, relatively small numbers of B. anthracis are being sought against a background of relatively large numbers of B. cereus. Within the spectrum of B. cereus frequently found in such specimens may be strains which are non or weakly motile, non or weakly haemolytic, partially or totally sensitive to penicillin, possibly sensitive to the gamma phage and resembling B. anthraczs in colonial appearance though not, of course, producing a capsule. Most of the several attempts at designing anthraxspecific detection systems have not been widely put to the test under field conditions and are not readily available for applying to problem strains. Such systems include immunofluorescence (Phillips & Martin 1982a, b, c; Phillips et al. 1983; Phillips & Martin 1988; Phillips & Ezzell 1989), immunoradiometric (Phillips & Martin 1983, 1984) or enzyme-linked lectinosorbent (Graham et al. 1984) assays based on polyclonal antibodies to whole cell preparations or (Phillips & Ezzell 1989) to extracted cell antigens or on monoclonal antibodies to a B. anthraczs spore (Phillips et al. 1988) or cell wall polysaccharide (Ezzell et al. 1990) epitope. Faced with a number of problem strains in the past few years but now aided by newly available molecular and genetic techniques such as nucleic acid hybridization (Tenover 1988; Hutson & Duggleby 1990; Macario & de Macario 1990; Sayler & Layton 1990; Walker & Dougan 1990) and the polymerase chain reaction (PCR) (Fekete et al. 1990; Gibbs 1990; Innis et al. 1990), supplementing the conventional identification tests, we have concluded that careful follow-up of query B. anthracis strains which fail to produce a capsule in vitro or to produce anthrax in test animals may still be bona fide B. anthracis. Some speculations on the origins of these strains and the significance of their presence in specimens is offered.
MATERiALS AND M E T H O D S Strains: in vivo and in vitro data
The strains included in this study and their isolation histories are listed in Table 1. Wild type strains ASC 50-70, 148 and 178 were isolated from bonafide cases of anthrax. Ability to produce the polypeptide capsule (phenotypic expression of the capsule [CAP] genes on plasmid pX02)
was examined in all the strains as one of the routine tests for confirming the identity of virulent B. anthracis. The potent characteristics of ASC 7, 68 and 69 have been well established in vaccine protection studies (Ivins et al. 1986, 1990; Turnbull et a/. 1986, 1988, 1990b; Ivins & Welkos 1988) and their LD,, values have been characterized (Ivins et al. 1986, 1990; Turnbull et al. 1986; Ivins & Welkos 1988). Strains ASC 45 and 47 reached this laboratory by a circuitous route and are on record as being descendants of the vaccine strain isolated by Sterne (1937) which was subsequently shown to be pXOl+/2-. ASC 79, 80 and 152 were identified by standard laboratory criteria (Carman et al. 1985; Turnbull 1990; Turnbull et al. 1990b; Turnbull & Kramer 1991) as capsulating B. anthracis. Numbers 79 and 80 were isolated from samples taken on separate occasions from the same site at a disused tannery prior to demolition and re-development of the site. Number 152, from an old giraffe bone, is listed in Table 1 as an environmental isolate since the bone was dug up from soil in which it had been buried. These strains were not tested in animals but are regarded as fully virulent B. anthracis. Numbers 100-103, isolated from a set of soil samples, were regarded as ‘anthrax-like’ having the correct colonial morphology and in being non-motile, nonhaemolytic, penicillin- and gamma phage-sensitive but capsule negative. Strain 108 was isolated from old bones in the Etosha National Park, Namibia, strains 121-126 from sludge from a sewage treatment works in Wales, 127 and 128 from a farm drain leading into the same sewage treatment works and 129 and 130 from cattle feed ingredients. These were all, like 100-103, originally recorded as ‘anthrax-like’ strains under the same diagnostic criteria. Numbers 121-128 were all isolated during an investigation following an outbreak of anthrax on the same farm (Edginton 1990). The feed ingredients were from the investigation of an outbreak of anthrax in cattle in 1972 (Hugh-Jones & Hussaini 1974). D r Hugh-Jones had kept these samples and requested their re-examination in September 1989. T h e Etosha National Park is enzootic for anthrax (Turnbull et al. 1989) but the bones from which ASC 108 was isolated were not derived from an animal known to have died of this disease. As a result of the importance of a correct identification for strains 121-126, these were tested for virulence by the intraperitoneal injection of 1 x 10, to 5 x lo5 spores in mice. Toxigenicity tests : phenotypic expression of pXO1
Rocket electrophoresis (REP) was carried out at 7-10 V/cm in a 1% barbitone agarose containing 13-18pl/ml of rabbit anti-purified protective antigen (PA) (Fig. 1); 5 p l of
_
~
_
_
_
_
Onpal ID
Pen Res
1261/7-88
Hippo B
32/70
I48
178
_
I.}
45 47
_
7B
_
_
EB2/83 Ames NH
62 68 69
_
_
STI STl5
_
Vellum
^
{
_
1
_
L9 lXo5i l262jS-88
EB1/83
61
so 79 152
27 LA1183
55,489 58
: :;}
no.
ASC
_
_
see text
Sec text
_
_
_
_
_
_
_
_
_
Caw, Oxford, England. 1944. B. anthrnrir type stram Cured of pXO2 Cured of pXOl and pX02
U6 ( # 4 ) U7(#5) U8 ( X 1) 64-1 (X7)
&2(#8) AA(#2) S6U, ( X 3 )
S6U, (X6) MHJ SI
MHJ S2
121
122 123 124
125 126 127
128 129
130
i1
of farm that had recently experienced an episode of anthrax. 1989 ?Mineral feed supplements 1989 (held from anthrax outbreak investigation 1972)
%Ex drain
TDigested sewage sludge. All one rite. I989
Namibia, 1987
Old bones, Etosha National Park,
ljTmnery dump rite, 1988. Same site as ASC 79, 80 above
t
REP, rocket electrophoresis; EIA. enzyme immunoassay 'Expected' on the basis of phenotyv. 1 From Turnbull e l a/. (1986). 4 Mice injected intraperitoneally with 1-5 x 10' spores. 7 Set text for further details and referenccs w. weak reaction; n. not done.
I08
I
lRO5jZB 1805/3A Q78
102
103
strains
l805/ZA
101
environmental
_
Zambia, 1988 One tannery dump site, England. Isolations 23 days apan, 1988 Old giraffe bone, Namibia, 1988
_
_
+ n
+
+ +
+
+
+ + +
+ +
n
+
+(w)
_
_
_
_
_
-
-
_
+ + + +
n
pXOl+/2'
pXOl'/2+
PXOI+/2+
pXOl+/2A pXOli/2' pXOli/2'
pXOl*2'
pXOl+/2+ pXOl+/2' pXOli/2+ px01+/2+
Expectedt
n
n
PXOl+/2+
n pX0li/2' pX0li/2'
n
n
n n n
Direct plasmid rreen
-
_
+ + + +
+
-
+ -
-
-
-
-
-
-
_
_
MLD50487
~
-
.
.
-
-
pXOl+/2pXOI'/Z-
-
px01+/2pXolrj2-
pXOl'/2+
pxo1+/2+ PXOl'/2*
pXOli/2f
~
n n
_
~
_
-
-
pxo1+/2 pxolr/2-
pXO1'/2'
pXOl'12' PXOl'/2+
px01+/2+
n
-
+ +
+ "
-
-
+ -
+ + -
-
-
+
-
-
-
+
+
-
"
n
-
-
-
-
-
-
-
-
OjlO
OjlO
7/10 0110
OilW O/lO OjlO
n px01*/2n
pXO1-/2-
n
n
n
pxo1-/2px01-/2pXOlrj2 px01+/2-
-
pXOl*/2pXol'j2-
pxol-jz pxo1rj2px01rj2 px01+/2pX0l 12pxOl*/zpXO1-/2-
pxo1+/2pxoI'/2-
pxo1+/2PXOl+/2
~
p x 0 1 * 12 px01*/2px01-/2-
p X 0 l i 12 pXOl*/2pxo1*/2~
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
-
+
-
n n -
-
+
+
_
+ +
+
+ +
"
+ +
+
f
MI.DSO I071 MLDSO 50.7
Virulence
Genotype
. . . . . . . . . . . . . . . . . . . . . . . . .
+
+
+
+ + +
+
+
+
+
+ +
EIA
+ +
-
REP
+ +
+
+
Capsule
_ _ _ _ _ _ _ _ _ _ _ _ _
Etosha National Park, Namibia. 1983
Burchell's 7xbra Caw, Iowa, 1980 Human pulmonary anthrax. New Hampshire. USA, I957 Human, pulmonary or intestinal, fatal, Northampton, England. 1974 Hartmann's Zebra, Etosha National Park, Namibia, 1988 North Luangwa National Park,
Burchell's Zebra
Elephant
Human isolates, Zimbabwe I982
History
-------~---______-______~___________~ 'Anthrax-like' 100 IROSjl
Vaccinr
_
Pre- and post-curing
strains
Wild type environmental
Wild type strains from human or animal cases of anthrax
Class
Tested phenotype'
Table 1 Identities, histories, phenotype and genotype results of tests of the strains examined
-
-
_
+ f
+ +
+ + +
+
+
+
+
+
+
+
n
+ +
n
+ + +
+
+ +
_
+ +
+ -
+
I219 (PA)
+
975 (PA)
+
1.9 kb (PA)
n
+ + +
+
+ +
_
+
1236 (LF)
Oligonucleotide probes
+ + + +
-
204 bp (PA)
DNA probes
+ + +
+
+ +
+
n
-
977 (EF)
-
+ +
n
+ + +
+
-
1234 (CAP)
n
+
+ +
+
+
+ + +
97516 (PA)
PCR
n
+
+ +
+
+
+ +
123617 (1.F)
n
+
+
+
t
+
-
+
+
1234/1301 (CAP)
24 P . C . B . TURNBULL ET A L
Plasmid analysis
Fig. 1 Rocket electrophoresis for detection of the protective antigen (PA)component of anthrax toxin. Wells 1 4 : 14, 28, 55, 1 lOpg/ml purified PA; Wells 5-12: cell-free culture filtrates of ASC 45 (wells 5, 6), ASC 68 (7, 8), ASC 32/70 (9,10) and the Sterne vaccine strain ( 1 1, 12)
sample were added to the wells in the agarose. REP was subsequently superseded by capture enzyme immunoassay (EIA) on 96-well plates coated by adding 50pl/well of rabbit anti-purified PA diluted to 20 pg/ml. in carbonate coating buffer, p H 9.5, and either drying out at 37°C overnight or holding in the refrigerator for > 2 h. T h e purified PA used for generating the antisera was prepared and kindly supplied by D r S.H. Leppla (Leppla 1988) or by D r C.P. Quinn (Quinn et al. 1988). Production of toxin for REP and EIA was done in overnight static cultures (37°C) of the strains in the casamino acids medium of Belton & Strange (1954). This was carried out in glass screw-capped 25 ml bottles containing 5 ml of casamino acids base medium with 0.6ml growth nutrients giving the critical depth of 15-20mm. Supernatant fluids were filtered through 0.22 pm FP030/3 low-protein-binding filters (Schleicher & Schuell, Dassel, Germany) before testing by REP and/or EIA.
Plasmid profiles were prepared for a proportion of the strains only by methods modified from those of Green et al. (1985) and Reddy et al. (1987). Strains were grown at 37°C for 16 h in a rotary shaker at 200 rev/min in 250 ml conical flasks containing 25 ml volumes of brain-heart infusion broth supplemented with 3 % heat-inactivated horse serum (Tissue Culture Services Ltd., Botulph Claydon, Bucks, UK). Cells were harvested at 12000g for 10min at 4°C and the supernatant fluid removed by aspiration. The cell pellet was resuspended by vortex mixing in 1 ml of 40 mmol/l Tris-acetate p H 7.9, 2 mmol/l EDTA, 15% sucrose and then 1.0 ml of cells was pipetted into 2-0 ml of lysis buffer (50 mmol/l Tris-base, 15% sucrose containing 3% w/v sodium dodecyl sulphate and 150 mmol/l NaOH) in 15 ml polypropylene tubes. Cells were lysed at 60°C for 30 min with occasional gentle mixing. Cell lysate was chilled on ice for 10 min and then 0.5 ml of 2 mol/l Tris-HC1 (pH 7.0) was added. The neutralized lysate was mixed by gentle inversion and the incubation on ice continued for a further 10 min. Whole cell lysate was then extracted by inversion with 6 ml of a 1 : 1 phenol: chloroform solution (w/v). Phases were separated by centrifugation at 12000 g for 10 min at 4°C in a Sorvall SS34 centrifuge rotor. T h e aqueous phase was removed to clean polypropylene tubes and the nucleic acids precipitated with ethanol/ammonium acetate as described by Maniatis et al. (1982). T h e pellet was dissolved in 100 pl of sterile distilled water and 10-20 p1 used for agarose gel analysis. For agarose gel analysis, the 10-20 pl volumes of plasmid DNA extracts were electrophoresed at 70 V through a 0.7% agarose gel (Agarose NA, Pharmacia-LKB) in TBE buffer (Maniatis et al. 1982). DNA bands were visualized by staining in ethidium bromide in water (0.5 pg/ml) followed by exposure under short-wave U.V. light (Fig. 2).
Fig. 2 Plasmid analysis. (a) Strains as indicated. Pen R, penicillin resistant strain (ASC32/70). (b) From left to right: 1, ASC 7 (pX01+/2+); 2, nil; 3,4, ASC 124 (pXO1+/2-); 5, ASC 122 (pX01-/2-); 6, ASC 127 (pXO1-/2-); 7, ASC 129 (pX01+/2-); 8, ASC 7 (repeat). See Table 1 for strain details
B. A N T H R A C I S BUT N O T A L W A Y S A N T H R A X 25
Direct plasmid analysis was subsequently displaced by DNA probes and PCR for identification of various segments of the virulence factor genes on pXOl and pX02. DNA probes
A number of approaches were taken. I n the first, strains ASC 121-128 were lifted directly from plate cultures by the use of nitrocellulose discs (Schleicher and Schuell BAS 85). The cells were then lysed and the DNA denatured on the filters followed by binding to the nitrocellulose by vacuum baking (Maniatis et al. 1982). Immobilized DNA was probed using a y3'PdATP end-labelled PA gene-specific oligonucleotide. In a second method, colonies of the various isolates were also grown directly on the surface of nylon (Amersham, Hybond N), Immobilon N (polyvinylidene difluoride [PVDF], Millipore) or nitrocellulose membranes. After the DNA had been released from the colonies, denatured and immobilized, it was probed with a 204 bp fragment of the PA gene non-radioactively labelled by incorporation of digoxigenin-1 1-dUTP (dig-dUTP). Hybridizing bands were detected with an anti-digoxigenin conjugate linked to alkaline phosphatase with a subsequent colour reaction involving nitroblue tetrazolium salt and 5-bromo-4-chloro3-indolyl phosphate, toluidinium salt. Positive hybridizing bands yielded a blue-purple colour. The main approach was slot-blot analysis of both small and large scale total DNA preparations from the strains. The DNA was normally released by an enzyme/detergent system by simply vortex mixing the cells in a suitable buffer together with 40-mesh glass beads. T h e DNA released was denatured using l/lOth volumes of 3 mol/l NaOH and heating for 1 h at 65-70°C to form singlestranded DNA. The samples were neutralized by the addition of 5 mol/l ammonium acetate or 2 mol/l ammonium acetate if an equivalent volume of saturated NaCl was first added to the samples to precipitate proteins. T h e denatured DNA was then transferred to the membranes by a Millipore Milliblot-S apparatus; usually 100 pl of sample was applied to slots in triplicate. T h e membranes were washed briefly with 2 mol/l ammonium acetate after blotting, airdried and the DNA cross-linked to the membrane by longwave U.V. illumination for 3-5 min. Probes used to examine the DNA extracted from the cultures were the 204 bp and 1.9 kb PA gene fragments, both non-radioactively labelled with dig-dUTP by random prime labelling and two PA gene-specific oligonucleotides (975 and 1219), one lethal factor (LF) gene-specific oligonucleotide (1236), one oedema factor (EF) gene-specific oligonucleotide (977) and a capsule (CAP) gene-specific oligonucleotide (1234). All the oligonucleotide probes were labelled with dig-dUTP by end-tailing using terminal transferase with incubation at 37°C for 90 min.
Polymerase chain reaction
Three pairs of oligonucleotide primers, directed at PA, LF and CAP genes were used for PCR. T h e intervening nucleic acid fragments were amplified in a loop1 reaction mixture consisting of 50 mmol/l KC1, 10 mmol/l Tris-HC1 (pH 9-0), 0.01% gelatin, 0.1% Triton X-100, 200 pmol/l each of dTTP, dCTP, dATP and dGTP, 1 pl of each primer (0.2 pmol/l), and 3.0 mmol/l MgCI, for primer pair 1236/7 (LF) or 4-5 mmol/l MgCI, for primer pairs 975/6 (PA) and 1234/1301 (CAP). The samples were denatured for 4 rnin at 95°C after which 2.5 units of Taq polymerase were added and 100 p1 mineral oil added as an overlay to prevent reflux. T h e mixtures were then subjected to 35 cycles of amplification in a programmable thermal controller (M.J. Research, Cambridge, MA, USA) using the following parameters: denaturation at 95°C for 2min; annealing of primers 975/6 for 1 rnin at 55"C, 1234/1301 for 1 rnin at 58°C and 1236/7 for 1 rnin at 45°C; and primer extension for 3 min at 72°C. After the last cycle, samples were incubated for a further 7 rnin at 72°C. Analysis of 10 p1 of the reaction mixture was performed by standard submarine gel electrophoresis (1% agarose in TBE) and the ethidium bromide-stained reaction products examined under U.V. light (Figs 3, 4).
RESULTS AND DISCUSSION
For the purposes of comparison, all results are combined in Table 1.
Fig. 3 PCR amplification products of Bacillus anthracis virulence factor genes. 1, 1 kb ladder; 2, 1.9 kb PA fragment (primers 975 and 976); 3,850 bp CAP fragment (primers 1234 and 1301); 4, 800 bp LF fragment (primers 1236 and 1237)
26 P . C . B . TURNBULL E T A L
probe work was done before the strains in the collection were reorganized into ASC numbers and it is thought that the discrepant probe and PCR results may have resulted from the use of different culture slopes. Anthrax-like strains
Fig. 4 PCR using primers 1236 and 1237 (LF). From left to
right: ASC 45, 47, 50, 51, 55, 58, 61, 62, 68, 69, 79, 80, 100 and 101. Fragment size 800 bp Wild type strains
Among the wild type strains from anthrax cases, DNA hybridization and PCR results for the most part correlated with the phenotypic expression of the virulence factors, the toxin and the capsule. T h e exceptions to this were the direct oligonucleotide probes to CAP in ASC 50 and to PA, LF and CAP in ASC 51 and in no PCR amplification by the LF and CAP primers in ASC 55, or by the LF primer in ASC 148 and 152. One explanation for these exceptions may be sequence alterations, degeneracy or a deletion and this seems possible in ASC 148 and 152 which exhibited discrepant PA and LF PCR results. Plasmid isolation followed by mapping and sequencing of virulence genes would be required to verify this. Similarly, the only discrepancies between phenotype and genotype in the two environmental wild type B. anthracis strains, ASC 79 and 80, were the failure in both cases of oligonucleotide 1234 (CAP) to hybridize. In each of these discrepancies, the negative result was counterbalanced by a positive in the other oligonucleotide system or in the direct nucleic acid PA probe hybridizations. Culture 55, when re-examined in view of the discrepant PCR results, had clearly changed radically in its morphology from the original isolate and was toxin negative. A new storage slope ASC 189 was therefore made from the original isolation slope which was still available; this was toxin positive, though somewhat weakly so, but still PCR negative. Vaccine strain derivatives
Gene probe and PCR profiles matched the phenotype of ASC 47, a derivative of the Sterne strain which is known to lack plasmid pX02. ASC 45 was also listed as being a Sterne strain derivative but when PCR repeatedly indicated amplification of the CAP gene, the phenotype was checked and this culture was indeed found to produce a capsule. There were a number of cultures labelled S T 1 in the collection deriving from miscellaneous sources. T h e direct
Of particular importance are production of toxin (phenotype), probe/PCR results and, where done, the plasmid profiles of the ‘anthrax-like’ cultures. The ability to screen for toxin production is a relatively new innovation and, together with the genotypic profiles, provides firm evidence not available previously from conventional identification and confirmation techniques that at least a proportion of these ‘anthrax-like’ species are B. anthracis lacking pX02. Clearly strains 10G103, 108, 124, 126 and 129 are examples of this. Of these, only 124 was tested in mice (Table 1) and this proved to have the same low level virulence that is exhibited by the Sterne vaccine strain (Welkos et al. 1986). Some anomalous results are apparent in the oligonucleotide hybridization results with 121-123, while with strains 127 and 128, the PCR and hybridization readings obtained disagree with each other and with the genotypes expected from the phenotype profiles. Problems of discrepancies between PCR and probes may have arisen because they were applied to separate DNA preparations. Ideally DNA isolated from suspected anthrax or anthrax-like strains should be analysed with both PCR and probes for comparative purposes. T o date there have been no comparative DNA sequence analysis studies on the various plasmidborne virulence genes from a range of wild type, vaccine and environmental strains. Some sequence variations may exist at positions at which the primers anneal and hence PCR or hybridizations may give negative results where phenotypes show they should have been positive. Polymerase chain reaction also persistently indicated the presence of capsule genes in ASC 130. Absolute confirmation that this PCR product has homology with the CAP genes awaits sequence analysis. Assuming the positive here is false, there remains no way at present of being able to categorically identify this and other isolates like ASC 121-3, 125, and 127 as B. anthracis which have lost both plasmids. It seems a reasonable belief, however, that they are. T h e origin of avirulent or non-anthrax-inducing B. anthracis is speculative. It must be stressed that the specimens from which the strains described here were isolated were all examined because there was some reason to suspect they might contain virulent B. anthracis. T h e hypothesis that the avirulent strains were modified variants of the virulent counterparts being looked for would not seem unreasonable. Fully virulent (pXO1+/2+) B. anthracis strains are readily cured of either or both plasmids in the
B. ANTHRACIS BUT NOT A L W A Y S A N T H R A X
laboratory (Mikesell et al. 1983; Uchida et al. 1985; Ivins et al. 1986; Ezzell 1988) and, long before such things were known about, Pasteur (1881) in preparing his vaccine was inadvertently curing his strains (pXO1-/2+) by passage at elevated temperatures (Ezzell 1988). Similarly, Sterne (1937), by subculturing his strains on 50% horse serum nutrient agar under a 30% COz atmosphere, inadvertently obtained spontaneous p X 0 1 +/2- derivatives. Additionally, as mentioned previously, it was recognized by Smith et al. (1952) that B. anthrucis could lose its virulence. I t is, therefore, wholly possible that, under conditions encountered in tannery effluent (strains 100-103), sewage and sewage treatment processes (121-128), feed processing (129, 130) or under the harsh environmental conditions in the Etosha National Park (log), that either or both plasmids may be spontaneously lost. Strain differentiation or marker systems have not been successfully developed for B. anthrucis, so such a hypothesis would b e hard to prove. Anecdotal evidence suggests that these types of strains have been recognized for many years but, in being avirulent or unable to induce anthrax, have been discarded, usually unrecorded, as inconsequential. Now, apart from the academic interest in correctly classifying them, the possibility should perhaps be considered seriously that they may serve as indicators of virulent parents somewhere in the system being examined.
REFERENCES B E L T O N , F . C . & S T R A N G ER,. E . (1954) Studies on a protective antigen produced in vitro from Bacillus anthracis: medium and methods of production. British Journal of Experimental Pathology 35, 144152. C A R M A NJ ., A . , H A M B L E T OPN. , & M E L L I N G J, . (1985) Bacillus anthracis. In Isolation and Identrfication of Microorganisms of Medical and Veterinary Importance ed. Collins, C.H. & Grange, J.M. Society for Applied Bacteriology Technical Series No. 21. pp. 207-213. London: Academic Press. E D I N G T O N , A.B. (1990) An outbreak of anthrax in pigs: a practitioner's account. Veterinary Record 127, 321-324. E Z Z E L L , J . W . (1988) Anthrax-Pasteur to the present. Clinical Microbiology Newsletter 10, 113-1 16. ABSHIRE, T.G., LITTLE, S.F., EZZELL, J . W . , LIDGERDINB G .,C . & B R O W NC, . (1990) Identification of Bacillus anthracis by using monoclonal antibody to cell wall galactose-N-acetylglucosamine polysaccharide. Journal of Clinical Microbiology 28, 223-23 1. FEKETE, A,, BANTLE, J.A., HALLING, S.M. & S A N E O R NM, . R . (1990) Preliminary development of a diagnostic test for Brucella using polymerase chain reaction. Journal of Applied Bacteriology 69, 216227. G I B B S , R . A . (1990) DNA amplification by the polymerase chain reaction. Analytical Chemistry 62, 1202-1214. G O R D O NR, . E . , H A Y N E SW , .C. & P A N GC , . H - N . (1973) The genus Bacillus. Agriculture Handbook No. 427, United
27
States Department of Agriculture, Agricultural Research Service, Washington, DC. G R A H A MK, . , K E L L E R ,K . , E Z Z E L L J, . & D O Y L E ,R . (1984) Enzyme-linked lectinosorbent assay (ELLA) for detecting Bacillus anthracis. European Journal of Clinical Microbiology 3,210-212. G R E E NB , . D . , B A T T I S T IL, . , K O E H L E RT, . M . , T H O R N E , C . B . & I V I N S , B.E. (1985) Demonstration of a capsule plasmid in Bacillus anthracis. Infection and Immunity 49, 291297. H U G H - J O N E SM , . E . & H U S S A I N IS, . N . (1974) An anthrax outbreak in Berkshire. Veterinary Record 94, 228-232. HUTSONR , . A . & D U G G L E B YC. , (1990) DNA probes for anthrax detection. In Proceedings of the International Workshop on Anthrax ed. Turnbull, P.C.B. Salisbury Medical Bulletin, No. 68, Special supplement, pp. 27-28. I N N I S , M . A . , G E L F A N D ,D . H . , S N I N S K Y ,J . J . & W H I T E ,T .J . (1990) P C R Protocols : A Guide to Methods and Applications. London : Academic Press. J . (1958) Problems concerning the I V A N O V I CG S ,. & FOLDES, phylogenesis of Bacillus anthracis. Acta MicrobiaIogica Academiae Scientiarum Hungaticae 5, 89-109. I V I N S ,B.E. & W E L K O SS, . L . (1988) Recent advances in the development of an improved human anthrax vaccine. European Journal of Epidemiology 4, 12-19. I V I N S , B.E., E Z Z E L L ,J . W . , J E M S K I , J., H E D L U N D , K . W . , R I S T R O P HJ ,. D . & L E P P L AS, . H . (1986) Immunization studies with attenuated strains of Bacillus anthracis. Infection and Immunity 52, 454-458. LITTLES , . F . & KNUDSON, I V I N S ,B . E . , W E L K O SS.L., , G . B . (1990) Cloned protective activity and progress in development of improved anthrax vaccines. In Proceedings of the International Workshop on Anthrax ed. Turnbull, P.C.B. Salisbury Medical Bulletin, No. 68, Special supplement, pp. 8688. L E P P L A ,S . H . (1988) Production and purification of anthrax toxin. Methods in Enzymology, 165, 103-116. M A C A R I OA.J.L. , & D E M A C A R I OE,. C . (1990) Gene Probes f o r Bacteria. London : Academic Press. M A N I A T I ST, . , F R I T S C H E , .F. & SAMBROOK J ., (1982) Molecular Clontng : A Laboratory Manual. Cold Spring Harbor Laboratories, USA. M ' F A D Y E AJN. , (1903a) A peculiar staining reaction of the blood of animals dead of anthrax. Journal of Comparative Pathology 16, 3541. M ' F A D Y E A NJ, . (1903b) A further note with regard to the staining reaction of anthrax blood with methylene-blue. Journal of Comparative Pathology 16, 360-361. M I K E S E L LP, . , I V I N S ,B . E . , R I S T R O P HJ, . D . & D R E I E R , T .M . (1983) Evidence for plasmid-mediated toxin production in Bacillus anthracis. Infection and Immunity 39, 37 1-376. P A S T E U RL,. (1881) De I'attknuation des virus et de leur retour a la virulence. Comptes Rendus Hebdomadaires des Siances de I'Acadimie des Sciences 92, 429435. P H I L L I P S ,A.P. & E Z Z E L L J, . W . (1989) Identification of Bacillus anthracis by polyclonal antibodies against extracted vegetative cell antigens. Journal of Applied Bacteriology 66, 4 19432. P H I L L I P S ,A . P . & M A R T I N ,K . L . (1982a) Assessment of
28 P . C . B . TURNBULL ET A L .
immunofluorescence measurements of individual bacteria in direct and indirect assays for Bacillus anthracis and Bacillus cereus spores. Journal of Applied Bacteriology 53, 223-23 1. P H I L L I P S ,A.P. & M A R T I N ,K . L . (1982b) Evaluation of a microfluorometer in immunofluorescence assays of individual spores of Bacillus anthracis and Bacillus cereus. Journal of Immunological Methods 49, 271-282. P H I L L I P S A.P. , & M A R T I NK , . L . (1982~)Variations on the staining method in quantitative indirect immunofluorescence assays for Bacillus spores, and the use of fluorescein-Protein A. Journal of Immunological Methods 54, 361-369. P H I L L I P S ,A.P. & M A R T I N ,K . L . (1983) Comparison of direct and indirect immunoradiometric assays (IRMA) for Bacillus anthracis spores immobilised on multispot microscope slides. Journal of Applied Bacteriology 55, 315-324. P H I L L I P SA.P. , & M A R T I NK , . L . (1984) Radioactive labels for Protein A : evaluation in the indirect immunoradiometric assay (IRMA) for Bacillus anthracis spores. Journal of Applied Bacteriology 56, 449-456. PHILLIPS, A.P. & M A R T I N ,K . L (1988) Investigation of spore surface antigens in the genus Bacillus by the use of polyclonal antibodies in immunofluorescence tests. Journal of Applied Bacteriology 64, 47-55. P H I L L I P SA.P., , M A R T I NK , . L . & BROSTER, M . G . (1983) Differentiation between spores of Bacillus anthracis and Bacillus cereus by a quantitative immunofluorescence technique. Journal of Clinical Microbiology 17, 4147. P H I L L I P S A.P., , C A M P B E L LA, . M . & QUINN, R . (1988) Monoclonal antibodies against spore antigens of Bacillus anthracis. F E M S Microbiology Immunology 47, 169-178. QUINN, C . P . , S H O N E , C . C . , T U R N B U LP.C.B. L, & M E L L I N GJ,. (1988) Purification of anthrax-toxin components by high-performance anion-exchange, gel-filtration and hydrophobic-interaction chromatography. Biochemical Journal 252,753-758. R E D D YA , . , B A T T I S T IL, . & T H O R N C.B. E , (1987) Identification of self-transmissable plasmids in four Bacillus thuringiensis subspecies. Journal of Bacteriology 169, 5263-5270. R O B E R T S O N ,D . L . , B R A G G , T . S . , S I M P S O N , S . , K A S P A R ,R . , X I E , W . & T I P P E T T SM, . T . (1990) Mapping and characterization of the Bacillus anthracis plasmids pXOl and pX02. In Proceedings of the International Workshop on Anthrax ed. Turnbull, P.C.B. Salisbury Medical Bulletin, No. 68, Special supplement, pp. 55-58. S A Y L E RG.S. , & L A Y T O NA.C. , (1990) Environmental application of nucleic acid hybridization. Annual Review of Microbiology 44,625-648. S M I T H , N.R., G O R D O N ,R.E. & C L A R K ,F . E . (1952) Aerobic Sporeforming Bacteria. US Department of Agriculture, Agriculture Monograph No. 16, Washington, D.C. US Government Printing Office. S T E I N ,C . D . (1943) Studies and observations on the laboratory diagnosis of anthrax. Veterinary Medicine (Chicago) 38, 13&139 (cited by Smith et al. 1952).
STERNE, M . (1937) The effects of different carbon dioxide concentrations on the growth of virulent anthrax strains. Pathogenicity and immunity tests on guinea-pigs and sheep with anthrax variants derived from virulent strains. Onderstepoort Journal of Veterinary Science and Animal Industry 9, 49-67. T E N O V EFR . C, . (1988) Diagnostic deoxyribonucleic acid probes for infectious diseases. Clinical Microbiology Reviews 1, 82-101. T U R N B U LP L . C,. B . (1990) Anthrax. In Topley and Wilson’s Principles of Bacteriology, Virology and Immunity, Vol. 2, ed. Smith, G.R. & Easmon, C.R. pp. 364-377. Sevenoaks, Edward Arnold. T U R N B U L LP,. C . B . & K R A M E RJ,. M . (1991) Bacillus. In Manual of Clinical Microbiology 5th edn. ed. Balows, A,, Hausler, W. J., Herrman, K.L., Isenberg, H.D. & Shadomy, H.J. pp. 296303. Washington, DC: American Society for Microbiology. T U R N B U LPL . C, . B . , BROSTER,M.G., C A R M A NJ.A., , M A N C H E ER.J. , & M E L L I N GJ, . (1986) Development of antibodies to protective antigen and lethal factor components of anthrax toxin in humans and guinea pigs and their relevance to protective immunity. Infection and Immunity 52, 356363. T U R N B U LPL. C, . B ., L E P P L AS, . H . , B R O S T E R M , .G., QUINN, C . P . & M E L L I N GJ, . (1988) Antibodies to anthrax toxin in humans and guinea pigs and their relevance to protective immunity. Medical Microbiology and Immunology 177, 293303. TURNBUL P .L C ,. B . , C A R M A NJ,. A . , L I N D E Q U EP, . M . , JOUBERT, F . , HUBSCHLE,O.J.B. & SNOEYENBOS, G .H. (1989) Further progress in understanding anthrax in the Etosha National Park. Madoqua 16, 93-104. T U R N B U L LP.C.B., , K R A M E RJ, . M . & M E L L I N G ,J . (1990a) Bacillus. In Topley and Wilson’s Principles of Bacteriology, Virology and Immunity, Vol. 2, ed. Parker, M.T. & Duerden, B.I. pp. 187-210. Sevenoaks: Edward Arnold. T U R N B U LP.C.B., L, QUINN, C . P . , HEWSON,R., S T O C K B R I D GM E ., C . & M E L L I N GJ, . (1990b) Protection conferred by microbially-supplemented UK and purified PA vaccines. In Proceedings of the International Workshop on Anthrax ed. Turnbull, P.C.B. Salisbury Medical Bulletin, No. 68, Special supplement, pp. 89-91. U C H I D A ,I . , S E K I Z A K IT., , HASHIMOTO, K. & TERAK A D O , N . (1985) Association of the encapsulation of Bacillus anthracis with a 60 megadalton plasmid. Journal of General Microbiology 131, 363-367. , (1990) DNA probes: a new role W A L K E RJ,. & D O U G A NG. in diagnostic microbiology. Journal of Applied Bacteriology 67, 229-238. W E L K O S ,S . L . , K E E N E RT.J. , & G I B E S ,P . H . (1986) Differences in susceptibility of inbred mice to Bacillus anthracis. Infection and Immunity 51,795-800.
Bacillus anthracis but not always anthrax P.C.B. Turnbull’, R.A. HutsonlY2,Mandy J. Ward133,Marie N. Jones’, C.P. Quinn’, N.J. Finnie’, C.J. Duggleby“, J.M. Kramer5 and J. Melling‘ Divisions of ’Biologics and 4Biotechnology, Public Health Laboratory Service Centre for Applied Microbiology & Research, Porton Down, Salisbury and 5F00d Hygiene Laboratory, Central Public Health Laboratory, London, UK 3634/04/91: accepted 1 June 1991
P.C.B. T U R N B U L L , R.A. HUTSON, M.J. WARD, M . N . JONES, C.P. QUINN, N.J. FINNIE,C.J. DUGGLEBY, J.M. K R A M E R A N D J . MELLING. 1992. Gram-positive bacilli isolated during epidemiological investigations which, on the basis of conventional tests, resemble Bacillus anthracis but which fail to produce the capsule or to induce anthrax in test animals have long been dismissed in clinical and veterinary laboratories as B . cereus or simply as unidentified Bacillus spp. and thereupon discarded as inconsequential. In this study, the application of newly available DNA probe, polymerase chain reaction and specific toxin antigen detection technology has revealed that a proportion of such strains are B . anthracis which lack the plasmid carrying the capsule gene (pX02). While these techniques cannot, of course, be used to confirm the identities of strains resembling B . anthracis but which also lack the plasmid carrying the toxin genes (pXOl), the likelihood that these also are bonaJde B . anthracis becomes more acceptable. (As yet no naturally occurring p X O l - / 2 + strains have been found.) At this point, the significance of the presence of such avirulent forms of B. anthracis in specimens can only be a subject for speculation, but the possibility that they may be indicators of virulent parents somewhere in the system being examined must be considered.
INTRODUCTION
The classification of Bacillus anthracis and its taxonomic separation from the closely related B . cereus has long posed a dilemma for bacteriologists. While, in practical terms, it has never been a problem since M’Fadyean published his capsule staining method in 1903 (M’Fadyean 1903a,b) to identify the typical large square-ended capsulating Grampositive bacillus isolated from a human or animal with classical symptoms of anthrax as B . anthraczs, it is clear from reports of strains given names such as B . anthracozdes (or simply ‘anthracoids’), B . anthracis similis and B . pseudoanthracis, that, even in the early days of bacteriology, problems of identity arose (Ivanovics & Foldes 1958). The situation was summed up by Stein (1943) in his statement that ‘Pathogenicity constitutes the principal point of difference between typical strains of B . anthraczs and those of anthrax-like organisms’. Present addresses :’AFRC Instrtute of Food Research, Shinjield, Berkshire RC2 9AT; ’Department of Microbiology, University of Southampton, General Hospital, Southampton SO9 4XY, U K . Corresponding author: P.C.B. Turnbull, Division of Biologics, Public Health Laboratory Service Centrefor Applied Microbiology 6 Research, Porton Down, Salisbury, Wiltshire SP4 OJC,U K .
The beginnings of true order in the classification of Bacillus species was first brought about by Smith et al. in 1952 with further clarification in the related manual of Gordon et al. (1973) which remains the basis of today’s identification systems for this genus. On the reasoning that, when a strain of B . anthraczs lost its virulence it became indistinguishable from B . cereus, Smith et al. concluded that ‘ B . anthracis is taxonomically a pathogenic variety of B . cereus’. Gordon et al., despite objections that had obviously been raised in the intervening years, felt justified in retaining this status and listed the anthrax bacterium as B . cereus var anthracis. At the grass roots level, the issue has remained unresolved. Recent considerable advances in the understanding of the pathogenicity of B . anthracis at the molecular and genetic levels also have not immediately answered the taxonomic questions. The two known virulence factors, the toxin and the capsule, are encoded on separate plasmids, pXOl and pX02 respectively (Robertson et al. 1990) and the loss of either or both of these results in an avirulent form which is essentially indistinguishable from B . cereus by conventional biochemical tests. As was the case in the first half of the century, little difficulty is now experienced in identifying as B . anthracis
22 P.C.B. T U R N B U L L E T A L
the typical non-haemolytic, non-motile, penicillin- and gamma phage-sensitive, capsulating Gram-positive bacilli isolated from animals and (rarely nowadays in the West, of course) humans with symptoms of anthrax. T h e difficulties arise in epidemiological or other studies concerned with detection of B. anthracis in environmental samples such as dust, soil, water, sewage sludge or feedstuffs. Frequently in these types of samples, relatively small numbers of B. anthracis are being sought against a background of relatively large numbers of B. cereus. Within the spectrum of B. cereus frequently found in such specimens may be strains which are non or weakly motile, non or weakly haemolytic, partially or totally sensitive to penicillin, possibly sensitive to the gamma phage and resembling B. anthraczs in colonial appearance though not, of course, producing a capsule. Most of the several attempts at designing anthraxspecific detection systems have not been widely put to the test under field conditions and are not readily available for applying to problem strains. Such systems include immunofluorescence (Phillips & Martin 1982a, b, c; Phillips et al. 1983; Phillips & Martin 1988; Phillips & Ezzell 1989), immunoradiometric (Phillips & Martin 1983, 1984) or enzyme-linked lectinosorbent (Graham et al. 1984) assays based on polyclonal antibodies to whole cell preparations or (Phillips & Ezzell 1989) to extracted cell antigens or on monoclonal antibodies to a B. anthraczs spore (Phillips et al. 1988) or cell wall polysaccharide (Ezzell et al. 1990) epitope. Faced with a number of problem strains in the past few years but now aided by newly available molecular and genetic techniques such as nucleic acid hybridization (Tenover 1988; Hutson & Duggleby 1990; Macario & de Macario 1990; Sayler & Layton 1990; Walker & Dougan 1990) and the polymerase chain reaction (PCR) (Fekete et al. 1990; Gibbs 1990; Innis et al. 1990), supplementing the conventional identification tests, we have concluded that careful follow-up of query B. anthracis strains which fail to produce a capsule in vitro or to produce anthrax in test animals may still be bona fide B. anthracis. Some speculations on the origins of these strains and the significance of their presence in specimens is offered.
MATERiALS AND M E T H O D S Strains: in vivo and in vitro data
The strains included in this study and their isolation histories are listed in Table 1. Wild type strains ASC 50-70, 148 and 178 were isolated from bonafide cases of anthrax. Ability to produce the polypeptide capsule (phenotypic expression of the capsule [CAP] genes on plasmid pX02)
was examined in all the strains as one of the routine tests for confirming the identity of virulent B. anthracis. The potent characteristics of ASC 7, 68 and 69 have been well established in vaccine protection studies (Ivins et al. 1986, 1990; Turnbull et a/. 1986, 1988, 1990b; Ivins & Welkos 1988) and their LD,, values have been characterized (Ivins et al. 1986, 1990; Turnbull et al. 1986; Ivins & Welkos 1988). Strains ASC 45 and 47 reached this laboratory by a circuitous route and are on record as being descendants of the vaccine strain isolated by Sterne (1937) which was subsequently shown to be pXOl+/2-. ASC 79, 80 and 152 were identified by standard laboratory criteria (Carman et al. 1985; Turnbull 1990; Turnbull et al. 1990b; Turnbull & Kramer 1991) as capsulating B. anthracis. Numbers 79 and 80 were isolated from samples taken on separate occasions from the same site at a disused tannery prior to demolition and re-development of the site. Number 152, from an old giraffe bone, is listed in Table 1 as an environmental isolate since the bone was dug up from soil in which it had been buried. These strains were not tested in animals but are regarded as fully virulent B. anthracis. Numbers 100-103, isolated from a set of soil samples, were regarded as ‘anthrax-like’ having the correct colonial morphology and in being non-motile, nonhaemolytic, penicillin- and gamma phage-sensitive but capsule negative. Strain 108 was isolated from old bones in the Etosha National Park, Namibia, strains 121-126 from sludge from a sewage treatment works in Wales, 127 and 128 from a farm drain leading into the same sewage treatment works and 129 and 130 from cattle feed ingredients. These were all, like 100-103, originally recorded as ‘anthrax-like’ strains under the same diagnostic criteria. Numbers 121-128 were all isolated during an investigation following an outbreak of anthrax on the same farm (Edginton 1990). The feed ingredients were from the investigation of an outbreak of anthrax in cattle in 1972 (Hugh-Jones & Hussaini 1974). D r Hugh-Jones had kept these samples and requested their re-examination in September 1989. T h e Etosha National Park is enzootic for anthrax (Turnbull et al. 1989) but the bones from which ASC 108 was isolated were not derived from an animal known to have died of this disease. As a result of the importance of a correct identification for strains 121-126, these were tested for virulence by the intraperitoneal injection of 1 x 10, to 5 x lo5 spores in mice. Toxigenicity tests : phenotypic expression of pXO1
Rocket electrophoresis (REP) was carried out at 7-10 V/cm in a 1% barbitone agarose containing 13-18pl/ml of rabbit anti-purified protective antigen (PA) (Fig. 1); 5 p l of
_
~
_
_
_
_
Onpal ID
Pen Res
1261/7-88
Hippo B
32/70
I48
178
_
I.}
45 47
_
7B
_
_
EB2/83 Ames NH
62 68 69
_
_
STI STl5
_
Vellum
^
{
_
1
_
L9 lXo5i l262jS-88
EB1/83
61
so 79 152
27 LA1183
55,489 58
: :;}
no.
ASC
_
_
see text
Sec text
_
_
_
_
_
_
_
_
_
Caw, Oxford, England. 1944. B. anthrnrir type stram Cured of pXO2 Cured of pXOl and pX02
U6 ( # 4 ) U7(#5) U8 ( X 1) 64-1 (X7)
&2(#8) AA(#2) S6U, ( X 3 )
S6U, (X6) MHJ SI
MHJ S2
121
122 123 124
125 126 127
128 129
130
i1
of farm that had recently experienced an episode of anthrax. 1989 ?Mineral feed supplements 1989 (held from anthrax outbreak investigation 1972)
%Ex drain
TDigested sewage sludge. All one rite. I989
Namibia, 1987
Old bones, Etosha National Park,
ljTmnery dump rite, 1988. Same site as ASC 79, 80 above
t
REP, rocket electrophoresis; EIA. enzyme immunoassay 'Expected' on the basis of phenotyv. 1 From Turnbull e l a/. (1986). 4 Mice injected intraperitoneally with 1-5 x 10' spores. 7 Set text for further details and referenccs w. weak reaction; n. not done.
I08
I
lRO5jZB 1805/3A Q78
102
103
strains
l805/ZA
101
environmental
_
Zambia, 1988 One tannery dump site, England. Isolations 23 days apan, 1988 Old giraffe bone, Namibia, 1988
_
_
+ n
+
+ +
+
+
+ + +
+ +
n
+
+(w)
_
_
_
_
_
-
-
_
+ + + +
n
pXOl+/2'
pXOl'/2+
PXOI+/2+
pXOl+/2A pXOli/2' pXOli/2'
pXOl*2'
pXOl+/2+ pXOl+/2' pXOli/2+ px01+/2+
Expectedt
n
n
PXOl+/2+
n pX0li/2' pX0li/2'
n
n
n n n
Direct plasmid rreen
-
_
+ + + +
+
-
+ -
-
-
-
-
-
-
_
_
MLD50487
~
-
.
.
-
-
pXOl+/2pXOI'/Z-
-
px01+/2pXolrj2-
pXOl'/2+
pxo1+/2+ PXOl'/2*
pXOli/2f
~
n n
_
~
_
-
-
pxo1+/2 pxolr/2-
pXO1'/2'
pXOl'12' PXOl'/2+
px01+/2+
n
-
+ +
+ "
-
-
+ -
+ + -
-
-
+
-
-
-
+
+
-
"
n
-
-
-
-
-
-
-
-
OjlO
OjlO
7/10 0110
OilW O/lO OjlO
n px01*/2n
pXO1-/2-
n
n
n
pxo1-/2px01-/2pXOlrj2 px01+/2-
-
pXOl*/2pXol'j2-
pxol-jz pxo1rj2px01rj2 px01+/2pX0l 12pxOl*/zpXO1-/2-
pxo1+/2pxoI'/2-
pxo1+/2PXOl+/2
~
p x 0 1 * 12 px01*/2px01-/2-
p X 0 l i 12 pXOl*/2pxo1*/2~
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
-
+
-
n n -
-
+
+
_
+ +
+
+ +
"
+ +
+
f
MI.DSO I071 MLDSO 50.7
Virulence
Genotype
. . . . . . . . . . . . . . . . . . . . . . . . .
+
+
+
+ + +
+
+
+
+
+ +
EIA
+ +
-
REP
+ +
+
+
Capsule
_ _ _ _ _ _ _ _ _ _ _ _ _
Etosha National Park, Namibia. 1983
Burchell's 7xbra Caw, Iowa, 1980 Human pulmonary anthrax. New Hampshire. USA, I957 Human, pulmonary or intestinal, fatal, Northampton, England. 1974 Hartmann's Zebra, Etosha National Park, Namibia, 1988 North Luangwa National Park,
Burchell's Zebra
Elephant
Human isolates, Zimbabwe I982
History
-------~---______-______~___________~ 'Anthrax-like' 100 IROSjl
Vaccinr
_
Pre- and post-curing
strains
Wild type environmental
Wild type strains from human or animal cases of anthrax
Class
Tested phenotype'
Table 1 Identities, histories, phenotype and genotype results of tests of the strains examined
-
-
_
+ f
+ +
+ + +
+
+
+
+
+
+
+
n
+ +
n
+ + +
+
+ +
_
+ +
+ -
+
I219 (PA)
+
975 (PA)
+
1.9 kb (PA)
n
+ + +
+
+ +
_
+
1236 (LF)
Oligonucleotide probes
+ + + +
-
204 bp (PA)
DNA probes
+ + +
+
+ +
+
n
-
977 (EF)
-
+ +
n
+ + +
+
-
1234 (CAP)
n
+
+ +
+
+
+ + +
97516 (PA)
PCR
n
+
+ +
+
+
+ +
123617 (1.F)
n
+
+
+
t
+
-
+
+
1234/1301 (CAP)
24 P . C . B . TURNBULL ET A L
Plasmid analysis
Fig. 1 Rocket electrophoresis for detection of the protective antigen (PA)component of anthrax toxin. Wells 1 4 : 14, 28, 55, 1 lOpg/ml purified PA; Wells 5-12: cell-free culture filtrates of ASC 45 (wells 5, 6), ASC 68 (7, 8), ASC 32/70 (9,10) and the Sterne vaccine strain ( 1 1, 12)
sample were added to the wells in the agarose. REP was subsequently superseded by capture enzyme immunoassay (EIA) on 96-well plates coated by adding 50pl/well of rabbit anti-purified PA diluted to 20 pg/ml. in carbonate coating buffer, p H 9.5, and either drying out at 37°C overnight or holding in the refrigerator for > 2 h. T h e purified PA used for generating the antisera was prepared and kindly supplied by D r S.H. Leppla (Leppla 1988) or by D r C.P. Quinn (Quinn et al. 1988). Production of toxin for REP and EIA was done in overnight static cultures (37°C) of the strains in the casamino acids medium of Belton & Strange (1954). This was carried out in glass screw-capped 25 ml bottles containing 5 ml of casamino acids base medium with 0.6ml growth nutrients giving the critical depth of 15-20mm. Supernatant fluids were filtered through 0.22 pm FP030/3 low-protein-binding filters (Schleicher & Schuell, Dassel, Germany) before testing by REP and/or EIA.
Plasmid profiles were prepared for a proportion of the strains only by methods modified from those of Green et al. (1985) and Reddy et al. (1987). Strains were grown at 37°C for 16 h in a rotary shaker at 200 rev/min in 250 ml conical flasks containing 25 ml volumes of brain-heart infusion broth supplemented with 3 % heat-inactivated horse serum (Tissue Culture Services Ltd., Botulph Claydon, Bucks, UK). Cells were harvested at 12000g for 10min at 4°C and the supernatant fluid removed by aspiration. The cell pellet was resuspended by vortex mixing in 1 ml of 40 mmol/l Tris-acetate p H 7.9, 2 mmol/l EDTA, 15% sucrose and then 1.0 ml of cells was pipetted into 2-0 ml of lysis buffer (50 mmol/l Tris-base, 15% sucrose containing 3% w/v sodium dodecyl sulphate and 150 mmol/l NaOH) in 15 ml polypropylene tubes. Cells were lysed at 60°C for 30 min with occasional gentle mixing. Cell lysate was chilled on ice for 10 min and then 0.5 ml of 2 mol/l Tris-HC1 (pH 7.0) was added. The neutralized lysate was mixed by gentle inversion and the incubation on ice continued for a further 10 min. Whole cell lysate was then extracted by inversion with 6 ml of a 1 : 1 phenol: chloroform solution (w/v). Phases were separated by centrifugation at 12000 g for 10 min at 4°C in a Sorvall SS34 centrifuge rotor. T h e aqueous phase was removed to clean polypropylene tubes and the nucleic acids precipitated with ethanol/ammonium acetate as described by Maniatis et al. (1982). T h e pellet was dissolved in 100 pl of sterile distilled water and 10-20 p1 used for agarose gel analysis. For agarose gel analysis, the 10-20 pl volumes of plasmid DNA extracts were electrophoresed at 70 V through a 0.7% agarose gel (Agarose NA, Pharmacia-LKB) in TBE buffer (Maniatis et al. 1982). DNA bands were visualized by staining in ethidium bromide in water (0.5 pg/ml) followed by exposure under short-wave U.V. light (Fig. 2).
Fig. 2 Plasmid analysis. (a) Strains as indicated. Pen R, penicillin resistant strain (ASC32/70). (b) From left to right: 1, ASC 7 (pX01+/2+); 2, nil; 3,4, ASC 124 (pXO1+/2-); 5, ASC 122 (pX01-/2-); 6, ASC 127 (pXO1-/2-); 7, ASC 129 (pX01+/2-); 8, ASC 7 (repeat). See Table 1 for strain details
B. A N T H R A C I S BUT N O T A L W A Y S A N T H R A X 25
Direct plasmid analysis was subsequently displaced by DNA probes and PCR for identification of various segments of the virulence factor genes on pXOl and pX02. DNA probes
A number of approaches were taken. I n the first, strains ASC 121-128 were lifted directly from plate cultures by the use of nitrocellulose discs (Schleicher and Schuell BAS 85). The cells were then lysed and the DNA denatured on the filters followed by binding to the nitrocellulose by vacuum baking (Maniatis et al. 1982). Immobilized DNA was probed using a y3'PdATP end-labelled PA gene-specific oligonucleotide. In a second method, colonies of the various isolates were also grown directly on the surface of nylon (Amersham, Hybond N), Immobilon N (polyvinylidene difluoride [PVDF], Millipore) or nitrocellulose membranes. After the DNA had been released from the colonies, denatured and immobilized, it was probed with a 204 bp fragment of the PA gene non-radioactively labelled by incorporation of digoxigenin-1 1-dUTP (dig-dUTP). Hybridizing bands were detected with an anti-digoxigenin conjugate linked to alkaline phosphatase with a subsequent colour reaction involving nitroblue tetrazolium salt and 5-bromo-4-chloro3-indolyl phosphate, toluidinium salt. Positive hybridizing bands yielded a blue-purple colour. The main approach was slot-blot analysis of both small and large scale total DNA preparations from the strains. The DNA was normally released by an enzyme/detergent system by simply vortex mixing the cells in a suitable buffer together with 40-mesh glass beads. T h e DNA released was denatured using l/lOth volumes of 3 mol/l NaOH and heating for 1 h at 65-70°C to form singlestranded DNA. The samples were neutralized by the addition of 5 mol/l ammonium acetate or 2 mol/l ammonium acetate if an equivalent volume of saturated NaCl was first added to the samples to precipitate proteins. T h e denatured DNA was then transferred to the membranes by a Millipore Milliblot-S apparatus; usually 100 pl of sample was applied to slots in triplicate. T h e membranes were washed briefly with 2 mol/l ammonium acetate after blotting, airdried and the DNA cross-linked to the membrane by longwave U.V. illumination for 3-5 min. Probes used to examine the DNA extracted from the cultures were the 204 bp and 1.9 kb PA gene fragments, both non-radioactively labelled with dig-dUTP by random prime labelling and two PA gene-specific oligonucleotides (975 and 1219), one lethal factor (LF) gene-specific oligonucleotide (1236), one oedema factor (EF) gene-specific oligonucleotide (977) and a capsule (CAP) gene-specific oligonucleotide (1234). All the oligonucleotide probes were labelled with dig-dUTP by end-tailing using terminal transferase with incubation at 37°C for 90 min.
Polymerase chain reaction
Three pairs of oligonucleotide primers, directed at PA, LF and CAP genes were used for PCR. T h e intervening nucleic acid fragments were amplified in a loop1 reaction mixture consisting of 50 mmol/l KC1, 10 mmol/l Tris-HC1 (pH 9-0), 0.01% gelatin, 0.1% Triton X-100, 200 pmol/l each of dTTP, dCTP, dATP and dGTP, 1 pl of each primer (0.2 pmol/l), and 3.0 mmol/l MgCI, for primer pair 1236/7 (LF) or 4-5 mmol/l MgCI, for primer pairs 975/6 (PA) and 1234/1301 (CAP). The samples were denatured for 4 rnin at 95°C after which 2.5 units of Taq polymerase were added and 100 p1 mineral oil added as an overlay to prevent reflux. T h e mixtures were then subjected to 35 cycles of amplification in a programmable thermal controller (M.J. Research, Cambridge, MA, USA) using the following parameters: denaturation at 95°C for 2min; annealing of primers 975/6 for 1 rnin at 55"C, 1234/1301 for 1 rnin at 58°C and 1236/7 for 1 rnin at 45°C; and primer extension for 3 min at 72°C. After the last cycle, samples were incubated for a further 7 rnin at 72°C. Analysis of 10 p1 of the reaction mixture was performed by standard submarine gel electrophoresis (1% agarose in TBE) and the ethidium bromide-stained reaction products examined under U.V. light (Figs 3, 4).
RESULTS AND DISCUSSION
For the purposes of comparison, all results are combined in Table 1.
Fig. 3 PCR amplification products of Bacillus anthracis virulence factor genes. 1, 1 kb ladder; 2, 1.9 kb PA fragment (primers 975 and 976); 3,850 bp CAP fragment (primers 1234 and 1301); 4, 800 bp LF fragment (primers 1236 and 1237)
26 P . C . B . TURNBULL E T A L
probe work was done before the strains in the collection were reorganized into ASC numbers and it is thought that the discrepant probe and PCR results may have resulted from the use of different culture slopes. Anthrax-like strains
Fig. 4 PCR using primers 1236 and 1237 (LF). From left to
right: ASC 45, 47, 50, 51, 55, 58, 61, 62, 68, 69, 79, 80, 100 and 101. Fragment size 800 bp Wild type strains
Among the wild type strains from anthrax cases, DNA hybridization and PCR results for the most part correlated with the phenotypic expression of the virulence factors, the toxin and the capsule. T h e exceptions to this were the direct oligonucleotide probes to CAP in ASC 50 and to PA, LF and CAP in ASC 51 and in no PCR amplification by the LF and CAP primers in ASC 55, or by the LF primer in ASC 148 and 152. One explanation for these exceptions may be sequence alterations, degeneracy or a deletion and this seems possible in ASC 148 and 152 which exhibited discrepant PA and LF PCR results. Plasmid isolation followed by mapping and sequencing of virulence genes would be required to verify this. Similarly, the only discrepancies between phenotype and genotype in the two environmental wild type B. anthracis strains, ASC 79 and 80, were the failure in both cases of oligonucleotide 1234 (CAP) to hybridize. In each of these discrepancies, the negative result was counterbalanced by a positive in the other oligonucleotide system or in the direct nucleic acid PA probe hybridizations. Culture 55, when re-examined in view of the discrepant PCR results, had clearly changed radically in its morphology from the original isolate and was toxin negative. A new storage slope ASC 189 was therefore made from the original isolation slope which was still available; this was toxin positive, though somewhat weakly so, but still PCR negative. Vaccine strain derivatives
Gene probe and PCR profiles matched the phenotype of ASC 47, a derivative of the Sterne strain which is known to lack plasmid pX02. ASC 45 was also listed as being a Sterne strain derivative but when PCR repeatedly indicated amplification of the CAP gene, the phenotype was checked and this culture was indeed found to produce a capsule. There were a number of cultures labelled S T 1 in the collection deriving from miscellaneous sources. T h e direct
Of particular importance are production of toxin (phenotype), probe/PCR results and, where done, the plasmid profiles of the ‘anthrax-like’ cultures. The ability to screen for toxin production is a relatively new innovation and, together with the genotypic profiles, provides firm evidence not available previously from conventional identification and confirmation techniques that at least a proportion of these ‘anthrax-like’ species are B. anthracis lacking pX02. Clearly strains 10G103, 108, 124, 126 and 129 are examples of this. Of these, only 124 was tested in mice (Table 1) and this proved to have the same low level virulence that is exhibited by the Sterne vaccine strain (Welkos et al. 1986). Some anomalous results are apparent in the oligonucleotide hybridization results with 121-123, while with strains 127 and 128, the PCR and hybridization readings obtained disagree with each other and with the genotypes expected from the phenotype profiles. Problems of discrepancies between PCR and probes may have arisen because they were applied to separate DNA preparations. Ideally DNA isolated from suspected anthrax or anthrax-like strains should be analysed with both PCR and probes for comparative purposes. T o date there have been no comparative DNA sequence analysis studies on the various plasmidborne virulence genes from a range of wild type, vaccine and environmental strains. Some sequence variations may exist at positions at which the primers anneal and hence PCR or hybridizations may give negative results where phenotypes show they should have been positive. Polymerase chain reaction also persistently indicated the presence of capsule genes in ASC 130. Absolute confirmation that this PCR product has homology with the CAP genes awaits sequence analysis. Assuming the positive here is false, there remains no way at present of being able to categorically identify this and other isolates like ASC 121-3, 125, and 127 as B. anthracis which have lost both plasmids. It seems a reasonable belief, however, that they are. T h e origin of avirulent or non-anthrax-inducing B. anthracis is speculative. It must be stressed that the specimens from which the strains described here were isolated were all examined because there was some reason to suspect they might contain virulent B. anthracis. T h e hypothesis that the avirulent strains were modified variants of the virulent counterparts being looked for would not seem unreasonable. Fully virulent (pXO1+/2+) B. anthracis strains are readily cured of either or both plasmids in the
B. ANTHRACIS BUT NOT A L W A Y S A N T H R A X
laboratory (Mikesell et al. 1983; Uchida et al. 1985; Ivins et al. 1986; Ezzell 1988) and, long before such things were known about, Pasteur (1881) in preparing his vaccine was inadvertently curing his strains (pXO1-/2+) by passage at elevated temperatures (Ezzell 1988). Similarly, Sterne (1937), by subculturing his strains on 50% horse serum nutrient agar under a 30% COz atmosphere, inadvertently obtained spontaneous p X 0 1 +/2- derivatives. Additionally, as mentioned previously, it was recognized by Smith et al. (1952) that B. anthrucis could lose its virulence. I t is, therefore, wholly possible that, under conditions encountered in tannery effluent (strains 100-103), sewage and sewage treatment processes (121-128), feed processing (129, 130) or under the harsh environmental conditions in the Etosha National Park (log), that either or both plasmids may be spontaneously lost. Strain differentiation or marker systems have not been successfully developed for B. anthrucis, so such a hypothesis would b e hard to prove. Anecdotal evidence suggests that these types of strains have been recognized for many years but, in being avirulent or unable to induce anthrax, have been discarded, usually unrecorded, as inconsequential. Now, apart from the academic interest in correctly classifying them, the possibility should perhaps be considered seriously that they may serve as indicators of virulent parents somewhere in the system being examined.
REFERENCES B E L T O N , F . C . & S T R A N G ER,. E . (1954) Studies on a protective antigen produced in vitro from Bacillus anthracis: medium and methods of production. British Journal of Experimental Pathology 35, 144152. C A R M A NJ ., A . , H A M B L E T OPN. , & M E L L I N G J, . (1985) Bacillus anthracis. In Isolation and Identrfication of Microorganisms of Medical and Veterinary Importance ed. Collins, C.H. & Grange, J.M. Society for Applied Bacteriology Technical Series No. 21. pp. 207-213. London: Academic Press. E D I N G T O N , A.B. (1990) An outbreak of anthrax in pigs: a practitioner's account. Veterinary Record 127, 321-324. E Z Z E L L , J . W . (1988) Anthrax-Pasteur to the present. Clinical Microbiology Newsletter 10, 113-1 16. ABSHIRE, T.G., LITTLE, S.F., EZZELL, J . W . , LIDGERDINB G .,C . & B R O W NC, . (1990) Identification of Bacillus anthracis by using monoclonal antibody to cell wall galactose-N-acetylglucosamine polysaccharide. Journal of Clinical Microbiology 28, 223-23 1. FEKETE, A,, BANTLE, J.A., HALLING, S.M. & S A N E O R NM, . R . (1990) Preliminary development of a diagnostic test for Brucella using polymerase chain reaction. Journal of Applied Bacteriology 69, 216227. G I B B S , R . A . (1990) DNA amplification by the polymerase chain reaction. Analytical Chemistry 62, 1202-1214. G O R D O NR, . E . , H A Y N E SW , .C. & P A N GC , . H - N . (1973) The genus Bacillus. Agriculture Handbook No. 427, United
27
States Department of Agriculture, Agricultural Research Service, Washington, DC. G R A H A MK, . , K E L L E R ,K . , E Z Z E L L J, . & D O Y L E ,R . (1984) Enzyme-linked lectinosorbent assay (ELLA) for detecting Bacillus anthracis. European Journal of Clinical Microbiology 3,210-212. G R E E NB , . D . , B A T T I S T IL, . , K O E H L E RT, . M . , T H O R N E , C . B . & I V I N S , B.E. (1985) Demonstration of a capsule plasmid in Bacillus anthracis. Infection and Immunity 49, 291297. H U G H - J O N E SM , . E . & H U S S A I N IS, . N . (1974) An anthrax outbreak in Berkshire. Veterinary Record 94, 228-232. HUTSONR , . A . & D U G G L E B YC. , (1990) DNA probes for anthrax detection. In Proceedings of the International Workshop on Anthrax ed. Turnbull, P.C.B. Salisbury Medical Bulletin, No. 68, Special supplement, pp. 27-28. I N N I S , M . A . , G E L F A N D ,D . H . , S N I N S K Y ,J . J . & W H I T E ,T .J . (1990) P C R Protocols : A Guide to Methods and Applications. London : Academic Press. J . (1958) Problems concerning the I V A N O V I CG S ,. & FOLDES, phylogenesis of Bacillus anthracis. Acta MicrobiaIogica Academiae Scientiarum Hungaticae 5, 89-109. I V I N S ,B.E. & W E L K O SS, . L . (1988) Recent advances in the development of an improved human anthrax vaccine. European Journal of Epidemiology 4, 12-19. I V I N S , B.E., E Z Z E L L ,J . W . , J E M S K I , J., H E D L U N D , K . W . , R I S T R O P HJ ,. D . & L E P P L AS, . H . (1986) Immunization studies with attenuated strains of Bacillus anthracis. Infection and Immunity 52, 454-458. LITTLES , . F . & KNUDSON, I V I N S ,B . E . , W E L K O SS.L., , G . B . (1990) Cloned protective activity and progress in development of improved anthrax vaccines. In Proceedings of the International Workshop on Anthrax ed. Turnbull, P.C.B. Salisbury Medical Bulletin, No. 68, Special supplement, pp. 8688. L E P P L A ,S . H . (1988) Production and purification of anthrax toxin. Methods in Enzymology, 165, 103-116. M A C A R I OA.J.L. , & D E M A C A R I OE,. C . (1990) Gene Probes f o r Bacteria. London : Academic Press. M A N I A T I ST, . , F R I T S C H E , .F. & SAMBROOK J ., (1982) Molecular Clontng : A Laboratory Manual. Cold Spring Harbor Laboratories, USA. M ' F A D Y E AJN. , (1903a) A peculiar staining reaction of the blood of animals dead of anthrax. Journal of Comparative Pathology 16, 3541. M ' F A D Y E A NJ, . (1903b) A further note with regard to the staining reaction of anthrax blood with methylene-blue. Journal of Comparative Pathology 16, 360-361. M I K E S E L LP, . , I V I N S ,B . E . , R I S T R O P HJ, . D . & D R E I E R , T .M . (1983) Evidence for plasmid-mediated toxin production in Bacillus anthracis. Infection and Immunity 39, 37 1-376. P A S T E U RL,. (1881) De I'attknuation des virus et de leur retour a la virulence. Comptes Rendus Hebdomadaires des Siances de I'Acadimie des Sciences 92, 429435. P H I L L I P S ,A.P. & E Z Z E L L J, . W . (1989) Identification of Bacillus anthracis by polyclonal antibodies against extracted vegetative cell antigens. Journal of Applied Bacteriology 66, 4 19432. P H I L L I P S ,A . P . & M A R T I N ,K . L . (1982a) Assessment of
28 P . C . B . TURNBULL ET A L .
immunofluorescence measurements of individual bacteria in direct and indirect assays for Bacillus anthracis and Bacillus cereus spores. Journal of Applied Bacteriology 53, 223-23 1. P H I L L I P S ,A.P. & M A R T I N ,K . L . (1982b) Evaluation of a microfluorometer in immunofluorescence assays of individual spores of Bacillus anthracis and Bacillus cereus. Journal of Immunological Methods 49, 271-282. P H I L L I P S A.P. , & M A R T I NK , . L . (1982~)Variations on the staining method in quantitative indirect immunofluorescence assays for Bacillus spores, and the use of fluorescein-Protein A. Journal of Immunological Methods 54, 361-369. P H I L L I P S ,A.P. & M A R T I N ,K . L . (1983) Comparison of direct and indirect immunoradiometric assays (IRMA) for Bacillus anthracis spores immobilised on multispot microscope slides. Journal of Applied Bacteriology 55, 315-324. P H I L L I P SA.P. , & M A R T I NK , . L . (1984) Radioactive labels for Protein A : evaluation in the indirect immunoradiometric assay (IRMA) for Bacillus anthracis spores. Journal of Applied Bacteriology 56, 449-456. PHILLIPS, A.P. & M A R T I N ,K . L (1988) Investigation of spore surface antigens in the genus Bacillus by the use of polyclonal antibodies in immunofluorescence tests. Journal of Applied Bacteriology 64, 47-55. P H I L L I P SA.P., , M A R T I NK , . L . & BROSTER, M . G . (1983) Differentiation between spores of Bacillus anthracis and Bacillus cereus by a quantitative immunofluorescence technique. Journal of Clinical Microbiology 17, 4147. P H I L L I P S A.P., , C A M P B E L LA, . M . & QUINN, R . (1988) Monoclonal antibodies against spore antigens of Bacillus anthracis. F E M S Microbiology Immunology 47, 169-178. QUINN, C . P . , S H O N E , C . C . , T U R N B U LP.C.B. L, & M E L L I N GJ,. (1988) Purification of anthrax-toxin components by high-performance anion-exchange, gel-filtration and hydrophobic-interaction chromatography. Biochemical Journal 252,753-758. R E D D YA , . , B A T T I S T IL, . & T H O R N C.B. E , (1987) Identification of self-transmissable plasmids in four Bacillus thuringiensis subspecies. Journal of Bacteriology 169, 5263-5270. R O B E R T S O N ,D . L . , B R A G G , T . S . , S I M P S O N , S . , K A S P A R ,R . , X I E , W . & T I P P E T T SM, . T . (1990) Mapping and characterization of the Bacillus anthracis plasmids pXOl and pX02. In Proceedings of the International Workshop on Anthrax ed. Turnbull, P.C.B. Salisbury Medical Bulletin, No. 68, Special supplement, pp. 55-58. S A Y L E RG.S. , & L A Y T O NA.C. , (1990) Environmental application of nucleic acid hybridization. Annual Review of Microbiology 44,625-648. S M I T H , N.R., G O R D O N ,R.E. & C L A R K ,F . E . (1952) Aerobic Sporeforming Bacteria. US Department of Agriculture, Agriculture Monograph No. 16, Washington, D.C. US Government Printing Office. S T E I N ,C . D . (1943) Studies and observations on the laboratory diagnosis of anthrax. Veterinary Medicine (Chicago) 38, 13&139 (cited by Smith et al. 1952).
STERNE, M . (1937) The effects of different carbon dioxide concentrations on the growth of virulent anthrax strains. Pathogenicity and immunity tests on guinea-pigs and sheep with anthrax variants derived from virulent strains. Onderstepoort Journal of Veterinary Science and Animal Industry 9, 49-67. T E N O V EFR . C, . (1988) Diagnostic deoxyribonucleic acid probes for infectious diseases. Clinical Microbiology Reviews 1, 82-101. T U R N B U LP L . C,. B . (1990) Anthrax. In Topley and Wilson’s Principles of Bacteriology, Virology and Immunity, Vol. 2, ed. Smith, G.R. & Easmon, C.R. pp. 364-377. Sevenoaks, Edward Arnold. T U R N B U L LP,. C . B . & K R A M E RJ,. M . (1991) Bacillus. In Manual of Clinical Microbiology 5th edn. ed. Balows, A,, Hausler, W. J., Herrman, K.L., Isenberg, H.D. & Shadomy, H.J. pp. 296303. Washington, DC: American Society for Microbiology. T U R N B U LPL . C, . B . , BROSTER,M.G., C A R M A NJ.A., , M A N C H E ER.J. , & M E L L I N GJ, . (1986) Development of antibodies to protective antigen and lethal factor components of anthrax toxin in humans and guinea pigs and their relevance to protective immunity. Infection and Immunity 52, 356363. T U R N B U LPL. C, . B ., L E P P L AS, . H . , B R O S T E R M , .G., QUINN, C . P . & M E L L I N GJ, . (1988) Antibodies to anthrax toxin in humans and guinea pigs and their relevance to protective immunity. Medical Microbiology and Immunology 177, 293303. TURNBUL P .L C ,. B . , C A R M A NJ,. A . , L I N D E Q U EP, . M . , JOUBERT, F . , HUBSCHLE,O.J.B. & SNOEYENBOS, G .H. (1989) Further progress in understanding anthrax in the Etosha National Park. Madoqua 16, 93-104. T U R N B U L LP.C.B., , K R A M E RJ, . M . & M E L L I N G ,J . (1990a) Bacillus. In Topley and Wilson’s Principles of Bacteriology, Virology and Immunity, Vol. 2, ed. Parker, M.T. & Duerden, B.I. pp. 187-210. Sevenoaks: Edward Arnold. T U R N B U LP.C.B., L, QUINN, C . P . , HEWSON,R., S T O C K B R I D GM E ., C . & M E L L I N GJ, . (1990b) Protection conferred by microbially-supplemented UK and purified PA vaccines. In Proceedings of the International Workshop on Anthrax ed. Turnbull, P.C.B. Salisbury Medical Bulletin, No. 68, Special supplement, pp. 89-91. U C H I D A ,I . , S E K I Z A K IT., , HASHIMOTO, K. & TERAK A D O , N . (1985) Association of the encapsulation of Bacillus anthracis with a 60 megadalton plasmid. Journal of General Microbiology 131, 363-367. , (1990) DNA probes: a new role W A L K E RJ,. & D O U G A NG. in diagnostic microbiology. Journal of Applied Bacteriology 67, 229-238. W E L K O S ,S . L . , K E E N E RT.J. , & G I B E S ,P . H . (1986) Differences in susceptibility of inbred mice to Bacillus anthracis. Infection and Immunity 51,795-800.
