• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of iaiPermissionsJournals.ASM.orgJournalIAI ArticleJournal InfoAuthorsReviewers
Infect Immun. Oct 2001; 69(10): 6532–6536.
PMCID: PMC98792
Note

Purification of Anthrax Edema Factor from Escherichia coli and Identification of Residues Required for Binding to Anthrax Protective Antigen

Editor: J. T. Barbieri

Abstract

The structural gene for anthrax edema factor (EF) was expressed in Escherichia coli under the control of a powerful T5 promoter to yield the 89-kDa recombinant protein that reacted with anti-EF antibodies. Recombinant EF was purified to homogeneity by a two-step procedure involving metal chelate affinity chromatography and cation-exchange chromatography. From 1 liter of culture, 2.5 mg of biologically active EF was easily purified. This is the first report of purification of anthrax EF from E. coli. EF purified from E. coli was biologically and functionally as active as its Bacillus anthracis counterpart. The recombinant protein could compete with lethal factor for binding to protective antigen. Sequence analysis revealed a stretch of seven amino acids, Val Tyr Tyr Glu Ile Gly Lys, present both in EF (residues 136 to 142) and lethal factor (residues 147 to 153). To investigate the role of these seven residues in binding to protective antigen, the residues were individually mutated to alanine in EF. Mutations in residues Tyr137, Tyr138, Ile140, and Lys142 of EF specifically blocked its interaction with anthrax protective antigen. The adenylate cyclase activity of the mutants remained unaffected. The results suggested that residues Tyr137, Tyr138, Ile140, and Lys142 are required for binding of EF to anthrax protective antigen, which facilitates its entry into susceptible cells.

Bacillus anthracis, the causative agent of anthrax, is a highly pathogenic bacterium. It primarily affects animals. Humans acquire the disease via infected animals or contaminated animal products. Two major virulence factors of this bacterial pathogen are antiphagocytic poly-d-glutamic acid capsule (8) and the three-component protein exotoxin called the anthrax toxin complex (17). The three proteins that form the complex are protective antigen (PA; 83 kDa; 735 amino acids), edema factor (EF; 89 kDa; 767 amino acids), and lethal factor (LF; 90 kDa; 776 amino acids). Individually, all three proteins are nontoxic. However, a combination of PA and LF, called the lethal toxin, causes death in experimental animals (24) and lysis of mouse peritoneal macrophages and macrophage-like cell lines (1, 3, 6, 7), whereas a combination of PA and EF, known as the edema toxin, induces an increase in intracellular cyclic AMP (cAMP) levels in eukaryotic cells (16) and elicits skin edema after subcutaneous injection (25).

Anthrax toxin fits the A-B model of classification of toxins, according to which the A (activity) moiety and the B (binding) moiety reside on different proteins that interact during intoxication of cells. Anthrax toxin utilizes a single B moiety (PA) to deliver one of the alternative A moieties (EF or LF) into the cell cytosol (17). PA binds to cell surface receptors and is cleaved by a cell surface protease such as furin (15). Proteolysis releases an N-terminal 20-kDa fragment, PA20, from the cell surface and exposes a high-affinity binding site on the 63-kDa fragment, PA63, still bound to the receptor. PA63 then binds to EF or LF, and the entire complex undergoes receptor-mediated endocytosis (6). Acidification of the endosome results in insertion of PA63 into the endosomal membrane and translocation of EF and LF into the cytosol, where they exert their toxic effects. It appears that ion-conductive channels, formed upon oligomerization of PA63 fragments, facilitate the translocation process (19).

EF is a bacterial adenylate cyclase which, upon activation by its eukaryotic cofactor, calmodulin, causes a rapid increase in the intracellular cAMP levels of host cells (16). Anthrax edema toxin can differentially regulate lipopolysaccharide-induced production of tumor necrosis factor alpha and interleukin-6 by increasing intracellular cAMP in monocytes (12). The disruption of the cytokine network may contribute to clinical symptoms of anthrax, such as edema formation.

The genes for PA, LF, and EF (called pagA, lef, and cya, respectively) reside on a 185-kb plasmid, pXO1, of B. anthracis. Strains lacking pXO1 do not produce toxin and are essentially avirulent (13). All three genes have been cloned and sequenced (4, 5, 21, 26). Efforts have been made to express and purify PA and LF from other expression hosts, such as Bacillus subtilis (14) and Escherichia coli. Large amounts of biologically active PA and LF can now be obtained from E. coli with relative ease (10, 11). Unfortunately, attempts have not been made to overexpress and purify full-length EF from other expression hosts. Therefore, culture supernatant of B. anthracis remains by and large the major source for the purification of EF (18, 20). Purification of EF from B. anthracis cultures requires P-3 containment facilities. Moreover, B. anthracis cultures give poor yields of EF that are often contaminated with other proteins. This has discouraged researchers from undertaking studies of EF (23, 27).

The aim of the present study was to develop a rapid and efficient system for purification of EF. EF was expressed in E. coli as a six-histidine-tagged protein and was purified to homogeneity using a two-step procedure involving metal chelate affinity chromatography and cation-exchange chromatography. Further, we have used this system for cloning, expression, and purification of mutant proteins of anthrax EF to investigate the role of a stretch of seven amino acids (9), Val Tyr Tyr Glu Ile Gly Lys, that is present both in EF (residues 136 to 142) and LF (residues 147 to 153).

Plasmid construction.

The full-length structural gene for EF was amplified by PCR using pXO1 as a template and primers that added BamHI and KpnI sites to the 5′ and 3′ ends of the PCR product, respectively. The sequences of the forward and reverse primers were 5′ GAT GGC GCG GAT CCA TGA ATG AAC ATT ACA CTG AGA G 3′ and 5′ GAT GGC GGG GTA CCT TAT TTT TCA TCA ATA ATT TTT TGG 3′, respectively. The amplified PCR product was digested with the restriction enzymes BamHI and KpnI and then ligated to the BamHI- and KpnI-digested plasmid pQE30 (Qiagen) to generate the construct pPN-EF. Cloning of the gene was verified by restriction analysis and sequencing (22). The construct pPN-EF has a six-histidine coding sequence at the 5′ end of the gene. The expression of the EF gene is under the control of a powerful T5 promoter. There are two lac operator sequences, which in combination with the lac repressor protein ensure tight regulation of gene expression.

Expression and purification of EF.

For high-level expression of the gene, pPN-EF was transformed into E. coli SG13009(pREP4) competent cells. Cells bearing pPN-EF were grown at 37°C and 250 rpm in Luria broth containing 100 μg of ampicillin and 25 μg of kanamycin per ml. When the A600 reached 0.8, isopropyl-1-thio-β-d-galactopyranoside (IPTG) was added to a final concentration of 0.5 mM. After 4 h of induction, the cells were harvested by centrifugation. The sodium dodecyl sulfate-polyacrylamide gel electrophoresis profile of these cells showed EF migrating at 89 kDa. The protein reacted with rabbit polyclonal anti-EF antibodies on immunoblots. The periplasm, cytosol, and inclusion bodies were checked for the presence of EF, which was found to be mainly localized in the cytosol.

For the purification of EF from the cytosol of SG13009 cells, all procedures were performed at 4°C. The pellet obtained from 1 liter of culture was resuspended in 30 ml of sonication buffer (50 mM potassium phosphate, pH 8.0, 300 mM NaCl, 2 mM β-mercaptoethanol, and 1 mM phenylmethylsulfonyl fluoride). Lysozyme (1 mg/ml) was added, and the cell suspension was incubated on ice for 30 min. The cells were disrupted by sonication at 4°C. Cell debris was removed by centrifugation, and the supernatant was mixed with 5 ml of 50% Ni-nitrilotriacetic acid resin previously equilibrated with sonication buffer. The matrix was washed with the sonication buffer containing 20 mM imidazole until the A280 of the flowthrough was less than 0.01. The protein was then eluted with 300 mM imidazole chloride in elution buffer (20 mM potassium phosphate, pH 7.0, 50 mM NaCl, 1% glycerol, 2 mM β-mercaptoethanol, 1 mM EDTA, and 1 mM phenylmethylsulfonyl fluoride). Affinity chromatography resulted in 456-fold purification of EF (calculated by dividing the specific activity of the protein eluting from the column by that obtained for the total cell lysate).

For further purification of EF, the fractions containing EF were pooled and dialyzed against binding buffer (20 mM potassium phosphate, pH 7.0, 1% glycerol, 2 mM β-mercaptoethanol, and 1 mM EDTA) to remove imidazole and NaCl. The dialyzed sample was loaded onto an SP-Sepharose (Sigma) cation-exchange column previously equilibrated with binding buffer. The protein was eluted with a linear gradient of 0 to 500 mM NaCl in the same buffer. EF eluted from the exchanger at approximately 200 mM NaCl concentration as a homogeneous preparation. This was evident from the single band obtained for EF on both native and denaturing gels. Purified EF was dialyzed against 10 mM HEPES buffer containing 50 mM NaCl and was stored in aliquots at −70°C.

One liter of culture yielded 2.5 mg of purified EF with a specific activity of 230 μmol/min/mg. It was found that the two-step purification procedure resulted in a total of 1,045-fold purification compared to the cell lysate.

Having purified recombinant EF to homogeneity, further experiments were done to examine whether the biological activity of the recombinant protein was comparable to that of native protein purified from B. anthracis. In order to cause toxicity in susceptible cells, EF must first interact with PA, which facilitates its transport to the cell cytosol. The binding of E. coli-purified EF to PA was demonstrated both in solution and on cultured cells. Next, to evaluate the biological activity of recombinant EF in cultured cells, the cAMP response generated by the protein in CHO cells was studied. Competition assays with LF were also done to establish that the recombinant protein was biologically and functionally active.

In vitro binding of EF to PA.

To study the binding of EF to PA in solution, PA was cleaved with trypsin, and the nicked PA was incubated with EF (1 μg/ml) for 15 min in 20 mM Tris, pH 9.0, containing 2 mg of CHAPS {3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}/ml. Analysis of these samples on a nondenaturing 5 to 10% gradient gel demonstrated that EF purified from E. coli could bind to trypsin-digested PA and retard its mobility.

Binding of EF to receptor-bound PA.

To study the interaction between E. coli-purified EF and receptor-bound PA, J774A.1 macrophage-like cells plated in 12-well plates were cooled and incubated with PA (1 μg/ml) and radioiodinated EF (1 μg/ml) at 4°C, as described in detail previously for LF (3). PA, when added to the cells, binds to its receptors on the cell surface. Nicking by cell surface proteases leads to the exposure of binding sites to which EF rapidly binds. Unbound protein was removed by washing the cells with cold Hanks balanced salt solution. The cells were then solubilized in 0.5 ml of 0.1 N NaOH, and radioactive counts associated with the cells were measured to determine the amount of EF that bound to receptor-bound PA. E. coli-purified EF (specific binding, 12,759 ± 180) and B. anthracis-purified EF (specific binding, 12,467 ± 131) were found to be equally capable of binding to receptor-bound PA.

Biological activity of purified EF.

To evaluate the biological activity and the potency of E. coli-purified EF, the cAMP response generated by the protein in CHO cells was studied. Initial experiments showed that EF, like other agents that increase intracellular cAMP, induces characteristic morphological changes in CHO cells. Treatment of the cells with purified EF (even at concentrations as low as 10 ng/ml) along with 1 μg of PA/ml caused elongation of the cells and a 20-fold increase in the intracellular cAMP level (measured with a Biotrak cAMP EIA kit [Amersham Pharmacia]).

In fact, cAMP levels in cell cytosol rise proportionally in response to increasing doses of EF (with PA held constant). However, beyond a saturating dose of EF (1 μg/ml), no further increase in cAMP was observed. PA or EF alone can neither increase cAMP nor induce morphological changes in CHO cells. EF purified from E. coli and B. anthracis was found to be equally potent in generating cAMP response in cultured cells (data not shown).

Competition with LF.

EF or LF binds with high affinity to the site exposed on domain 1 of PA upon cleavage with proteases, such as trypsin and furin. It is expected that the binding of the alternate activity moiety to this site would make the site unavailable for the other moiety. This competition between LF and E. coli-purified EF for the binding site of PA was demonstrated on both J774A.1 and CHO cells.

CHO cells were incubated with edema toxin in the presence and absence of LF. The cAMP response of CHO cells to edema toxin was found to decrease when the dose of LF was increased (Table (Table1).1). A 10-fold excess of LF over recombinant EF effectively reduces intracellular cAMP to basal levels.

TABLE 1
cAMP response in CHO cells treated with anthrax toxin proteins

Intoxication of J774A.1 cells by lethal toxin was similarly affected by the presence of recombinant EF. Complete cell death occurs within 3 h when J774A.1 cells are incubated with lethal toxin (1 μg [each] of PA and LF/ml). However, in the presence of EF, J774A.1 cells are protected against the cytotoxicity of the lethal toxin (Table (Table2).2). Complete protection is achieved when the concentration of EF is 10 times that of LF.

TABLE 2
Cytotoxicity produced in J774A.1 cells in response to anthrax toxin proteins

Role of residues 136-Val Tyr Tyr Glu Ile Gly Lys-142.

To explain the competition between EF and LF for the binding site on PA, the sequences of the two proteins were compared. The N-terminal 300 amino acids of EF and LF have significant homology and are considered to be the domains that bind to PA63. The most notable feature of this region of homology is a stretch of seven amino acids, Val Tyr Tyr Glu Ile Gly Lys, which is present in EF (residues 136 to 142) as well as LF (residues 147 to 153). To investigate the roles of these amino acids in binding to PA63, each of the seven amino acids was individually replaced with Ala. Stratagene's Quikchange site-directed mutagenesis kit was used for introducing mutations. pPN-EF was used as the template for PCR with primers that contained the desired mutation. Seven mutant constructs were made. Sequencing of the mutant constructs confirmed that only specific mutations were introduced (V136A, Y137A, Y138A, E139A, I140A, G141A, and K142A). The constructs were then transformed into E. coli SG13009(pREP4) competent cells for expression of the mutant protein. The EF mutants were purified to homogeneity as detailed before for the wild-type EF.

To evaluate the biological activities of the mutant proteins, the cAMP responses generated by the mutants in CHO cells were studied. Mutants Y137A, Y138A, I140A, and K142A (when added along with PA) failed to elicit an elongation response in CHO cells, unlike wild-type EF or the other three mutants, V136A, E139A, and G141A. Quantitative estimation of the intracellular cAMP of the treated cells revealed that the abilities of the mutant proteins Y137A, Y138A, I140A, and K142A to increase intracellular cAMP levels were impaired (Fig. (Fig.1).1). The cAMP response elicited by the mutant E139A in CHO cells was also slightly reduced in comparison to that generated by the wild-type protein.

FIG. 1
cAMP response of CHO cells to EF mutants. Cells plated in 96-well tissue culture plates were treated with PA (1 μg/ml) and the indicated concentrations of EF mutant or wild-type EF (EFWT). After 2 h, the intracellular cAMP of the treated cells ...

Next, the enzymatic activities of the mutant proteins were determined. It was observed that the adenylate cyclase activities of all of the mutant proteins were comparable to that of the wild-type EF. This suggested that the impaired cAMP response generated in CHO cells by the mutants Y137A, Y138A, I140A, and K142A was not due to reduced adenylate cyclase activity but was due to a defect in their ability to enter the cells.

To gain entry into susceptible cells, EF binds to receptor-bound PA. To study the interaction of EF mutant proteins with PA, the mutant proteins were allowed to compete with radiolabeled wild-type EF for binding to receptor-bound PA. The results from the competition assay showed that cold wild-type EF, as well as the mutants V136A, E139A, and G141A, was able to compete with radiolabeled wild-type EF and inhibit its binding to receptor-bound PA. However, the mutants Y137A, Y138A, I140A, and K142A could not compete with the radiolabeled wild-type protein in binding to receptor-bound PA (Fig. (Fig.2).2). This suggested that the mutants Y137A, Y138A, I140A, and K142A were defective in the ability to bind PA.

FIG. 2
Competitive binding of EF mutants to receptor-bound PA. J774A.1 cells were incubated with 100 ng of radioiodinated wild-type EF (EFWT)/ml along with 1 μg of PA/ml at 4°C. Various concentrations of cold wild-type protein or the mutants ...

Similar results were obtained in vitro when the mutant proteins were allowed to bind to PA in solution. Mutants V136A, E139A, and G141A were able to bind to trypsin-digested PA and retard its mobility on native polyacrylamide gel electrophoresis. However, mutants Y137A, Y138A, I140A, and K142A did not bind to trypsin-digested PA in solution (Fig. (Fig.3).3).

FIG. 3
Binding of EF mutants to PA in solution. Wild-type EF (EFWT) or its mutants (1 μg) were allowed to incubate with trypsin-nicked PA (1 μg) for 15 min, and the samples were analyzed on a nondenaturing 5 to 10% gradient gel. The gel ...

Thus, it was concluded that the residues Tyr137, Tyr138, Ile140, and Lys142 are required by EF for binding to anthrax PA, which facilitates its entry into susceptible cells.

To summarize, a system for expression and purification of B. anthracis EF has been presented above. The procedure does not involve handling of the pathogenic bacteria, since the expression host is E. coli. From 1 liter of culture, 2.5 mg of purified EF can easily be obtained, which is a significant improvement over the yields (0.4 to 0.8 mg/liter) obtained from B. anthracis cultures. The recombinant protein is purified to homogeneity. Furthermore, we have demonstrated that EF purified from E. coli is biologically and functionally comparable to its B. anthracis counterpart. The rapidity of the procedure (just two chromatographic steps) ensures minimal loss of biological activity. The high reproducibility and simplicity of the method make it a convenient system to adopt routinely. We hope that availability of this system will encourage researchers to undertake the study of EF. We have used this system for cloning, expression, and purification of mutant proteins of anthrax EF (V136A, Y137A, Y138A, E139A, I140A, G141A, and K142A). Studies with the mutant proteins show that the residues Tyr137, Tyr138, Ile140, and Lys142 are required for the binding of EF to anthrax PA. Mutations in these residues disrupt the interaction between the two proteins, thereby affecting the entry of EF into susceptible cells.

EF merits study not just because it is an important virulence factor of B. anthracis but also because it is the only toxin known so far which enters susceptible cells through receptor-mediated endocytosis, where it displays intrinsic adenylate cyclase activity. Bordetella pertussis adenylate cyclase is another toxin with inherent adenylate cyclase activity, but unlike its B. anthracis counterpart, it enters cells through direct penetration of the plasma membrane. Most other toxins increase intracellular cAMP levels by modulating the adenylate cyclase activity of the host cell.

Edema toxin provides an interesting model to study how cAMP regulates basic cellular and metabolic processes. It is through its ability to influence intracellular cAMP levels that edema toxin modulates cytokine synthesis in the infected host (11). Disruption of cytokine networks impairs the host's ability to defend against the invading bacteria and contributes to clinical symptoms of anthrax, such as edema formation.

Hitherto, biochemical investigations of the structure-function relationship of anthrax EF were hampered by the difficulty of purifying protein free from the other two toxin components. The availability of homogeneous protein preparations will trigger research on this subject. Insights into the structure-function relationship of the protein can provide useful information. For example, delineation of the minimal sequence of EF required for binding to PA and its translocation across the heptamer can help in developing an anthrax toxin-based protein translocation system to deliver heterologous proteins into the cytoplasm of mammalian cells. Such a system could serve as a tool by which cellular processes can be modulated.

Acknowledgments

This work was partly supported by a grant from DBT, Government of India. Both Praveen Kumar and Nidhi Ahuja are financially supported by UGC, Government of India.

We thank R. N. Saini for photographic assistance.

REFERENCES

1. Bhatnagar R, Friedlander A M. Protein synthesis is required for expression of anthrax lethal toxin cytotoxicity. Infect Immun. 1994;62:2958–2962. [PMC free article] [PubMed]
2. Bhatnagar R, Ahuja N, Goila R, Batra S, Waheed S M, Gupta P. Activation of phospholipase C and protein kinase C is required for expression of Anthrax lethal toxin cytotoxicity in J774A.1 cells. Cell Signal. 1999;11:111–116. [PubMed]
3. Bhatnagar R, Singh Y, Leppla S H, Friedlander A M. Calcium is required for expression of anthrax lethal toxin activity in the macrophage-like cell line J774A.1. Infect Immun. 1989;57:2107–2114. [PMC free article] [PubMed]
4. Bragg S T, Robertson D L. Nucleotide sequence and analysis of the lethal factor gene (lef) from B. anthracis. Gene. 1989;81:45–54. [PubMed]
5. Escuyer V, DuFlot E, Sezer O, Danchin A, Mock M. Structural homology between virulence-associated bacterial adenylate cyclase. Gene. 1988;71:293–298. [PubMed]
6. Friedlander A M. Macrophages are sensitive to anthrax toxin through an acid dependent process. J Biol Chem. 1986;261:7123–7126. [PubMed]
7. Friedlander A M, Bhatnagar R, Leppla S H, Singh Y. Characterization of macrophage sensitivity and resistance to anthrax lethal toxin. Infect Immun. 1993;61:245–252. [PMC free article] [PubMed]
8. Green B D, Battisti L, Kohler T M, Thorne C B, Ivins B E. Demonstration of a capsule plasmid in Bacillus anthracis. Infect Immun. 1985;49:291–297. [PMC free article] [PubMed]
9. Gupta P, Singh A, Chauhan V, Bhatnagar R. Involvement of residues 147VYYEIGK153 in binding of lethal factor to protective antigen of Bacillus anthracis. Biochem Biophys Res Commun. 2001;280:158–163. [PubMed]
10. Gupta P, Batra S, Chopra A P, Singh Y, Bhatnagar R. Expression and purification of the recombinant lethal factor of Bacillus anthracis. Infect Immun. 1998;66:862–865. [PMC free article] [PubMed]
11. Gupta P, Waheed S M, Bhatnagar R. Expression and purification of the recombinant protective antigen of Bacillus anthracis. Protein Expr Purif. 1999;16:369–376. [PubMed]
12. Hoover D L, Friedlander A M, Rogers L C, Yoon I K, Warren R L, Cross A S. Anthrax edema toxin differentially regulates lipopolysaccharide-induced monocyte production of tumor necrosis factor alpha and interleukin-6 by increasing intracellular cyclic AMP. Infect Immun. 1994;62:4432–4439. [PMC free article] [PubMed]
13. Ivins B E, Ezzell J W, Jemski J, Hedlund K W, Ristroph J D, Leppla S H. Immunization studies with attenuated strains of Bacillus anthracis. Infect Immun. 1986;52:454–458. [PMC free article] [PubMed]
14. Ivins B E, Welkos S L. Cloning and expression of the Bacillus anthracis protective antigen gene in Bacillus subtilis. Infect Immun. 1986;54:537–542. [PMC free article] [PubMed]
15. Klimpel K R, Molloy S S, Thomas G, Leppla S H. Anthrax toxin protective antigen is activated by a cell surface protease with the sequence specificity and catalytic properties of furin. Proc Natl Acad Sci USA. 1992;89:10277–10281. [PMC free article] [PubMed]
16. Leppla S H. Anthrax toxin edema factor: a bacterial adenylate cyclase that increases cAMP concentration in eukaryotic cells. Proc Natl Acad Sci USA. 1982;79:3162–3166. [PMC free article] [PubMed]
17. Leppla S H. The anthrax toxin complex. In: Alouf J E, Freer J H, editors. Sourcebook of bacterial protein toxins. London, United Kingdom: Academic Press; 1991. pp. 277–302.
18. Leppla S H. Purification and characterization of adenylyl cyclase from Bacillus anthracis. Methods Enzymol. 1991;195:153–168. [PubMed]
19. Milne J C, Furlong D, Hanna P C, Wall J S, Collier R J. Anthrax protective antigen forms oligomers during intoxication of mammalian cells. J Biol Chem. 1994;269:20607–20612. [PubMed]
20. Quinn P C, Shone C C, Turnbull T B, Melling J. Purification of anthrax toxin components by high performance anion-exchange, gel filtration and hydrophobic interaction chromatography. Biochem J. 1988;252:753–758. [PMC free article] [PubMed]
21. Robertson D L, Tippetts M T, Leppla S H. Nucleotide sequence of the B. anthracis edema factor gene: a calmodulin dependent adenylate cyclase. Gene. 1988;73:363–371. [PubMed]
22. Sambrook J, Fritsch E F, Maniatis T. Molecular cloning: a laboratory manual. 2nd ed. Cold Spring Harbor, N.Y: Cold Spring Harbor Laboratory Press; 1989.
23. Singh Y, Klimpel K R, Goel S, Swain P K, Leppla S H. Oligomerization of anthrax toxin protective antigen and binding of lethal factor during endocytic uptake into mammalian cells. Infect Immun. 1999;67:1853–1859. [PMC free article] [PubMed]
24. Smith H, Keppie J. Observations on experimental anthrax; demonstration of a specific lethal factor produced in vivo by Bacillus anthracis. Nature. 1954;173:869–870. [PubMed]
25. Stanley J L, Smith H. Purification of factor I and recognition of a third factor of anthrax toxin. J Gen Microbiol. 1961;26:49–66. [PubMed]
26. Welkos S L, Lowe J R, Eden-McCutchan F, Vodkin M, Leppla S H. Sequence and analysis of the DNA encoding protective antigen of Bacillus anthracis. Gene. 1988;69:287–300. [PubMed]
27. Zhao J, Milne J C, Collier R J. Effect of anthrax toxin's lethal factor on ion channels formed by protective antigen. J Biol Chem. 1995;270:18626–18630. [PubMed]

Articles from Infection and Immunity are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links