Anthrax

by Bill Suarez

Anthrax is deadly biochemical toxin. A small amount of the toxin is known to be fatal when inhaled. This toxin is airborne so that makes it easy to transmit. The Soviet Union experienced the deadly effects of the easily transmitted toxin in 1979. The activity of Bracillus anthracis with different concentrations of carbon dioxide and analysis of secondary structure at different pH values are shown in experiments that are discussed later. The two different reactions resulting from the binding of the two types of toxins with a protective antigen gene will also be discussed.

Bacillus anthracis is a spore forming bacillus, it is the etiological agent of anthrax. Anthrax is primarily a disease of herbivores, but B. anthracis is a highly pathogenic bacterium which can infect all mammals, including humans. Cutaneous anthrax infections are treatable with antibiotics. However, B. anthracis can be invasive by spreading throughout the body from the point of entry and produce massive terminal bacteremia.

The virulence of B. anthracis is due to the production of an antiphagocytic capsule and a tripartite exotoxin(cya,lef,and pag). The virulence factors are plasma encoded and plasmid pXO2 harbors three genes necessary for synthesis of d-glutamyl polypeptide capsule, capA, capB, and capC. It was proposed that these genes encode membrane associated proteins that mediate polymerization of D-glutamic acid. The structural genes for the three toxin proteins pag(PA) protective antigen, cya(EF) endema factor, and lef(LF) lethal factor are located on pXO1. The three proteins(PA, EF, and LF) act in binary combinations to produce two distinct reactions of PA-EF and PA-LF. The experiments on these proteins will be discussed later. None of the three proteins has a biological effect when administered alone.

Anthrax toxin and capsule synthesis are enhanced in vitro during growth under elevated levels of carbon dioxide or in bicarbonate solutions. Optimal production of toxin and capsule occurs when the bacterium is grown in a 5% carbon dioxide atmosphere in a medium containing 0.8% sodium bicarbonate. This carbon dioxide-bicarbonate effect on toxin and capsule synthesis is specific, and is not due to the buffering capacity of dissolved bicarbonate during bacterial growth.

The production of virulence factors by other organisms is also stimulated by carbon dioxide and bicarbonate at concentrations similar to those found in particular host environments. The synthesis of toxic shock syndrome toxin 1 by Staphylococcus aureus is enhanced at carbon dioxide concentrations present in the 5% carbon dioxide-8% bicarbonate environment. Cholera toxins are also stimulated by sodium bicarbonate. The mechanisms for the effect of carbon dioxide is not known for all systems described so far. B. anthracis remains the only organism in which synthesis of two distinct virulence factors is enhanced. Little is known regarding genetic control of the virulence factors of B. anthracis.

The presence of trans-acting regulatory elements on pXO1 has been suggested by a variety of observations. In B. anthracis, bicarbonate-enhanced expression of cloned pag and cya genes is observed only in the presence of pXO1. The carbon dioxide concentration required for capsule synthesis of pXO2+ strains increases 5 to 20% in the absence of pXO1. From these observations it can be concluded that pXO1 has a trans-acting regulatory element involved in toxin and capsule syntheses. In Bacillus subtilus, cloned PA gene is expressed well in the absence of pXO1, and PA synthesis is unaffected by carbon dioxide levels.

Experiments were conducted to demonstrate the trans-acting locus on pXO1 which is required for carbon dioxide induced transcription of pag. There is also evidence for basal level transcription of pag from a second transcription start site, that is not influence by carbon dioxide or the trans-acting locus.(1)

In order to isolate a B. anthracis mutant harboring a pac-lacZ fusion and to identify other carbon dioxide regulated genes, approximately 1,500 isolates from two different insertions were screened for mutants exhibiting carbon dioxide-bicarbonate enhanced beta-galactosidase expression. Eleven mutants were identified which produced beta-galactosidase (Lac+) when grown on LBC agar containing X-Gal in 5% carbon dioxide and when grown on a LB agar containing X-Gal in air (Lac-) was grown. This indicates that all of the carbon dioxide-bicarbonate enhanced Lac+ mutants carried insertions on pXO1 rather than on the chromosomes. Two of these mutants harbored insertions generating a pag-lacZ fusion.

To monitor the expression of a pag-lac fusion in UT4(a strain of B. anthracis), beta-galactosidase specific activity was measured during growth in 5% carbon dioxide and in air. A comparison of beta-galactosidase activity of UT4 with the PA production by UM44(another strain of B. anthracis) grown in the same conditions. The growth curve obtained showed that the growth rates of the UM44 and UT4 strains were similar under both conditions.(Koehler)

To determine whether stimulation of pag expression was due to increased carbon dioxide concentration or to nonspecific anaerobiosis, UT4 was grown in a variety of atmospheric conditions and the cells were assayed for beta-galactosidase. The results of the assays confirm the stimulation of pag transcription by elevated concentrations of carbon dioxide. The activities of the cells shows that by increasing the concentration from 0.03 to 20% resulted in a 13 to 19 fold stimulation of enzyme activity. In another test the concentration of carbon dioxide was kept constant while the concentration of oxygen was decreased. The activity was not significantly affected, indicating that the stimulation of expression is not due to nonspecific anaerobiosis.(Koehler)

An unusual outbreak of human anthrax occurred in the USSR in April 1979. A publication written in 1980 stated that the cause of anthrax was from eating Bacillus anthracis contaminated meat. Epidemiological studies strongly suggest that the release of B. anthracis spore aerosol from a nearby facility was responsible for the deaths of humans. A report describing examination of the samples of autopsy reports on infected patients were consistent with the results that B. anthracis was the causative agent. This led to the investigation of DNA from the tissues of the victims to verify that B. anthracis was the cause of death.(2)

The polymerase chain reaction(PCR) can be used to sensitively identify pathogens and their remnants in medical samples. DNA oligomers that prime PCR to amplify pathogen sequences can be used to determine the presence of a particular organism if the primers are complimentary to only pathogen specific sequences. Sample DNAs was used as template in a PCR containing primers that amplify portions of B. anthracis capA, capB, capC, cya, lef, and pag genes. The genes are located on the two large B. anthracis plasmids, pXO1 (pag, cya, lef) and pXO2 (capA, capB, capC) required for pathogenicity. Results of PCR based studies confirm that tissues from all victims contained B. anthracis and further analysis with primers suggests that the victims were apparently infected by a mixture of different strains of B. anthracis.

PCR analysis was conducted by using primers that amplified segments of each of the structural genes found on pXO1 and pXO2. Two sets of primer pairs were used to amplify fragments encoding portions of the pag, lef, cya, and cap A genes. Primer pairs used in initial PCR amplified a larger DNA fragment that contained the sequences complementary to the second set of primers. DNA products of the first amplification were used as template in a second reaction containing the second primer set. This method is known as nested primers. Nested primers increased sensitivity and maintained the specificity of the reaction. B. anthracis specific nested primers were not available for capB and capC genes, so the second reaction contained the same primers as the first. The results of PCR amplification using double PCR to amplify portions of capA(pXO1), and lef (pXO1) genes from the 13 tissue samples show that the anthrax victims were infected by virulent B. anthracis containing both plasmids.

The PCR primers amplify portions of known B. anthracis genes found on pXO1 and pXO2. PCR results using a primer set that specifically amplifies a randomly selected portion of the B. anthracis chromosomes were also done. These results clearly show that B. anthracis was responsible for the 1979 epidemic.(Jackson 1225)

There are two bacterial protein toxins produced by the bacterium Bacillus anthracis toxins that are structurally organized into distinct effector(A) and receptor(B) binding domains. The anthrax toxins differ from the diphtheria toxin (DT), single B domain interacts with two A domains which produce different effects. When the B domain (PA) protective antigen binds to one of the A domain (EF) edema factor the PA-EF complex is known as the edema toxin. The edema toxin causes edema when injected intradermally. When (PA) is bound to the other A domain (LF) lethal factor the PA-LF complex is designated a lethal toxin, and this causes rats and other animals to die. The exact role of the PA-LF complex is still unclear. Both complexes have clearly distinct biological effects on animals, but neither the EF or LF can bind to cells unless PA is bound to the cell surface, this means that PA plays a major role in cellular toxicity.

PA binds to a specific mammalian cell surface receptor (Mr 83 kDa). The bound PA is then cleaved into two fragments by proteolytic cleavage. The small fragment (20 kDa piece) is released outside the medium, and the larger piece which is 63 kDa (PA63) remains to attached to the receptor. The PA63 can then bind to either EF or LF and the complex is endocytosed by receptor mediated endocytosis on the cell surface.

The pH also has an effect on the toxicity. An acidic pH is crucial for toxicity, since cells are completely protected from B. anthracis edema toxin and lethal toxin by pretreatment with ammonium chloride and chloroquine that dissipates intracellular proton gradients and raises the pH of intracellular vesicles. When lowering the pH the protection has a reversible effect.

Both EF and LF act on intracellular targets. EF has been shown to be calcium dependent adenylate cyclase, that causes an elevation of cAMP within the cells. Evidence that a protease inhibitor prevents the toxicity of macrophages by lethal toxin. The lethal toxin is fully inactivated by mutagenesis of a zinc binding site which suggests that LF is a metalloprotease, but the target cell has not been determined.

From the studies of DT where both A and B domains are involved with the interaction of the endosomal membrane in a pH dependent manner, an analogous experiment has been used to study the interaction of all three anthrax toxin proteins with a lipid bilayer. The three proteins were independently inserted into the asolectin(mixed soybean phospholipid) lipid bilayer that destabilizes the membrane at low pH. It was also shown that PA63 is able to form cation selective channels in planar phospholipid bilayers, and to induce the release of markers from vesicles and cells at an acidic pH.

In DT, low pH has been shown to induce a conformational change in the protein structure which exposes surface hydrophobic domains. The hydrophobic domains have not been identified in any of the three anthrax proteins, but studies have shown that by lowering the pH, the hydrophobicity increases in PA, PA63, and LF.

In order to gain further insight on the mechanism of interaction of the lethal toxin complex ( PA-LF) with a lipid membrane the results of an experiment with the pH dependence of the binding of PA and LF, and the reversibility will be provided. An experiment showed the pH dependence of lipid association of anthrax lethal toxin proteins. The anthrax lethal toxin belongs to the group of the A-B type toxins. After internalization by the target cells, the A-B toxins are believed to interact with the lipid membrane of an endocytotic compartment. The toxin will then cross that membrane to reach their target in the cell cytoplasm. This interaction was characterized by the study of pH dependent association of the components of B. anthracis lethal toxin with lipid vesicles. The association of the PA-LF complex were tested at several pHs, and the amount of proteins and lipids was determined in each fraction.

At pH 7.2 there is no association of either PA or LF with the lipid membrane. When the pH was lowered to 6.0 or 5.0, both proteins associated with lipid vesicles. The yield of both PA and LF is strongly dependent on the pH. A test on the percentage of association of PA with the lipid membrane at several pH values was conducted and the results are: 64% at pH 6.0, 96% at pH 5.0, and 65% at pH 4.0. The data shows that the association of PA increases when the pH was lowered, with a maximum at pH 5.0, this suggests the possibility that as the pH is lowered there is competition between protein self aggregation and binding to the lipid vesicles. For LF there is 100% lipid association when the incubation was performed a pH 6.0 and below.

The test of reversibility was also conducted. When the pH was increased to 5.0 and 6.0 the percentage of association were the same at each pH. When it was raised up to pH 7.2, (which showed no association earlier) the protein remained bound, which proves that the PA binding to the lipid vesicles is irreversible.

A test on the effects of different pHs on the structure of anthrax lethal toxin proteins was also done. An attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR) was used to investigate the structure of soluble and membrane proteins. The analysis of the vibration bands of proteins and the amide I band (C=O), whose frequency of absorption is dependent upon the secondary structure. The FTIR-ATR deuterated spectra of PA were recorded at different pHs (7.2, 6.0, 5.2, 4.0). All four spectra looked similar and this suggests that no significant secondary structure change took place when the pH was lowered. At any pH the main absorption band was within (1637-1635 cm^-1) and this region is associated with the beta-sheet structure, suggesting that this structure predominates in PA. A curve fitting of the data gave the percentages of 28(+/- 5)% alpha helix and 38(+/- 5)% at all pH values.

The same experiment was conducted with LF and the shape of the deuturated amide I was very similar, indicating that the secondary structure is not dependent on pH. With the LF protein the main absorbancy was at (1652 cm^-1) and indicates that LF is predominantly alpha-helical. The curve fitting data gave percentages of 39% alpha-helical and 18% beta-sheets. Therefore PA is characterized by high content of beta-sheet structure and LF with alpha-helical structure.(3)

Many questions have been answered from the experiments conducted. The analysis of the tissue samples from victims in the1979 incident confirmed that they were exposed to B. anthracis . Several strains of anthrax were discovered along with evidence that B. anthracis was definitely the cause of death. The pH and the carbon dioxide test show that the activity for anthrax was irreversible in each case. Finally, the toxicity proteins of B. anthracis were shown to have different affects on mammalian cells, when bound to two different types of complexes.(PA-LF and PA-EF)

1. Koehler T M. Regulation of the Bacillus anthracis protective antigen gene: CO-2 and a trans-acting element activate transcription from one or two protomers. Journal of Bacteriology. 176(3) pgs. 586-595: 1994

2. Jackson P J. PCR analysis of tissue samples form the 1979 Sverdlovsk anthrax victims: the presence of multiple Bacillus anthracis strains in different victims. Proc Natl Acad Sci U S A. 95(3) pgs. 1224-9: February 3 1998

3. Wang X M. Secondary structure of anthrax lethal toxin proteins and their interaction with large unilamellar vesicles: a fourier-transform infrared spectroscopy approach. Biochemistry. 35(47) pgs. 14939-46: November 26 1996.

Other sources:

Copyright © 1998 Bill Suarez and Koni Stone

BACK to the May 1998 Table of Contents