Biochemical and Biophysical Research Communications 307 (2003) 446–450

www.elsevier.com/locate/ybbrc

BBRC

Deletion mutants of protective antigen that inhibit anthrax toxinboth in vitro and in vivo

Nidhi Ahuja, Praveen Kumar, Sheeba Alam, Megha Gupta, and Rakesh Bhatnagar*

Centre for Biotechnology, Jawaharlal Nehru University, New Delhi 110067, India

Received 3 June 2003

Abstract

The anthrax toxin complex is primarily responsible for most of the symptoms of anthrax. This complex is composed of three

proteins, anthrax protective antigen, anthrax edema factor, and anthrax lethal factor. The three proteins act in binary combination

of protective antigen plus edema factor (edema toxin) and protective antigen plus lethal factor (lethal toxin) that paralyze the host

defenses and eventually kill the host. Both edema factor and lethal factor are intracellularly acting proteins that require protective

antigen for their delivery into the host cell. In this study, we show that deletion of certain residues of protective antigen results in

variants of protective antigen that inhibit the action of anthrax toxin both in vitro and in vivo. These mutants protected mice against

both lethal toxin and edema toxin challenge, even when injected at a 1:8 ratio relative to the wild-type protein. Thus, these mutant

proteins are promising candidates that may be used to neutralize the action of anthrax toxin.

� 2003 Elsevier Inc. All rights reserved.

The use of anthrax as a bioweapon has highlighted

the urgent need to understand the pathogenesis of the

disease and to design effective strategies to combat it.

Anthrax is caused by Bacillus anthracis, a large gram-

positive bacillus that in its spore form can persist in

nature for prolonged periods, possibly years. Depending

upon the mode of entry of the spores, anthrax takes oneof the three forms. Cutaneous form of anthrax is ac-

quired when B. anthracis spores enter the host through a

cut or abrasion in the skin. The gastrointestinal form of

anthrax is acquired upon ingestion of B. anthracis spores

in contaminated food and the pulmonary form of an-

thrax is acquired upon inhalation of spores. The intes-

tinal and pulmonary forms are regarded as being more

often fatal than the cutaneous anthrax. This is becausethey frequently go unrecognized until it becomes too late

for effective treatment.

During pulmonary anthrax, the inhaled spores are

rapidly and efficiently phagocytosed by the alveolar

macrophages that carry them to the regional lymph

nodes in the media stinum [1,2]. Here, the spores ger-

minate to produce the vegetative forms that multiply

* Corresponding author. Fax: +91-11-2619-8234.

E-mail address: rakbhat01@yahoo.com (R. Bhatnagar).

0006-291X/03/$ - see front matter � 2003 Elsevier Inc. All rights reserved.

doi:10.1016/S0006-291X(03)01227-0

rapidly. Soon, the phagocytic capacity of the lymph

nodes becomes overwhelmed and the infection extends

to successive lymph nodes. The bacilli then enter the

bloodstream to cause severe bacteraemia. The germi-

nation of spores is soon followed by transcription of

genes for the three anthrax toxin proteins (1) protective

antigen (PA; 83 kDa), lethal factor (LF; 90 kDa), andedema factor (EF; 89 kDa). The three anthrax toxin

proteins act in binary combination of protective antigen

plus edema factor (edema toxin) and of protective an-

tigen and lethal factor (lethal toxin). The edema toxin,

as the name suggests, is responsible for extreme tissue

edema that is associated with cutaneous anthrax [3]. Its

contribution to the virulence and pathogenesis of an-

thrax is not well understood, however, it is quite possiblethat edema toxin disrupts the bactericidal function of

immune effector cells, disables the host immune re-

sponse, and thereby facilitates the replication and sur-

vival of the invading bacterium [4]. On the other hand,

the lethal toxin is responsible for causing the death of

animals or humans inflicted with anthrax [5].

Anthrax protective antigen plays a central role during

intoxication by anthrax toxin [6,7]. It is the receptor-binding moiety that facilitates the delivery of the other

two components, LF and EF, into the cell. During

N. Ahuja et al. / Biochemical and Biophysical Research Communications 307 (2003) 446–450 447

intoxication, PA binds to its receptors on the surface ofsusceptible cells [8]. The cleavage of the receptor-bound

PA by the cell surface proteases [9], such as furin, results

in the release of a 20 kDa fragment from the N-terminal

of the protein. The 63 kDa fragment of PA (PA63) oli-

gomerizes to form ring-shaped heptamer [10]. LF or EF

binds competitively to the site exposed on release of

20 kDa fragment of PA [11]. This entire complex un-

dergoes receptor-mediated endocytosis [12]. The acidi-fication of the endosome causes major conformational

changes in the PA molecule, leading to the insertion of

the heptamer into the endosomal membrane [13,14]. LF

and EF are translocated across the endosomal mem-

brane to the cytosol through these pores [15]. After

reaching the cell cytosol, LF and EF exert their toxic

effects. EF is a calcium/calmodulin-dependent adenylate

cyclase that causes an increase in the intracellular cAMPlevels of the host cells [16]. Whereas, LF is a metallo-

protease which cleaves several isoforms of MAP kinase

kinases within mammalian cells [17–19].

Recent studies have shown that mutations in toxin

proteins may result in variants that disrupt the action of

the toxin in vitro [20–22]. In continuation of these

studies, we demonstrate here that the deletion of the

residues Asp425 or Phe427 of PA yields dominant-neg-ative mutants of PA that are much more potent inhibi-

tors of anthrax toxin than any other mutant tested thus

far (including the alanine-substitution mutants of these

residues). The mutants, 425del and 427del, inhibit an-

thrax toxin action both in vitro and in vivo, even when

injected at a ratio of 1:8 relative to the wild-type protein.

These mutants protect mice against both lethal toxin

and edema toxin challenge. Thus, the 425del and 427delmutants are promising candidates that may be used to

neutralize the action of anthrax toxin.

Materials and methods

Site-directed mutagenesis. For the oligonucleotide-directed muta-

genesis of the PA gene, PCR was performed using pMS1 [23] as the

template. Briefly, the mutagenic primer A was used along with primer

C (that introduced BamHI site at the 30 end of the PCR product) and

mutagenic primer B (whose sequence was complementary to that of

primer A) was used along with primers D (that introduced HindIII site

at the 50 end of the PCR product) to amplify two segments of the gene

in the first round of PCR. This was followed by a second round of

PCR using the purified products of the first round of PCR as template

with primers C and D. The amplified product, obtained after the

second round of PCR, was digested with HindIII and BamHI and then

ligated to the backbone obtained upon digestion of pMS1 with the

same enzymes. The ligation mix was transformed into Escherichia coli

DH5a competent cells, and the colonies obtained after plating the

transformed cells were screened for the recombinant plasmid. The re-

combinant constructs selected after restriction analysis were sequenced

to confirm that the desired mutation had been incorporated.

Expression and purification. The recombinant constructs (contain-

ing the desired mutations) were transformed into E. coli BL21(DE3)

competent cells and expression of the mutated genes was induced as

described in detail previously [24]. After the induction was complete,

the E. coli cells were harvested and the periplasmic proteins were iso-

lated by osmotic shock. Mutant PA was then purified to homogeneity

using ion exchange and hydrophobic interaction chromatography, as

detailed previously for the wild-type protein [24].

Mammalian cell culture. Macrophage-like cell line J774A.1 was

maintained in RPMI 1640 medium containing 10% heat-inactivated

FCS, penicillin (100U/ml), and streptomycin (100 lg/ml). Chinese

hamster ovary (CHO) cell line was maintained in EMEM supple-

mented with non-essential amino acids, 25mM Hepes (pH 7.4), peni-

cillin (100U/ml), streptomycin (100 lg/ml), and 10% heat-inactivated

FCS.

Cytotoxicity assay. J774A.1 cells were plated at a density of

105 cells/ml in 96-well tissue culture plates and grown to 90% conflu-

ence. At the start of the experiment, spent medium and detached cells

were removed by aspiration and replaced with RPMI containing 0.5 or

1 lg/ml LF and varying concentrations of the wild-type and/or the

mutant PA. The cells were incubated for 3 h at 37 �C in a humidified

CO2 incubator. After 3 h, the cell viability was determined with 3-(4,5-

dimethylthiazol-2-yl)-5-diphenyltetrazolium bromide (MTT) dye, as

described previously [25]. All experiments were done in triplicate.

Elongation response of CHO. The CHO cells were plated in 24-well

plates and grown to confluence. To begin the experiment, old media

were replaced with H199 medium containing 0.1 or 0.2 lg/ml each of

EF and PA (wild-type and/or mutant protein). After incubation for 2 h

at 37 �C, the cells were examined under the microscope for the elon-

gation response [11].

Binding of PA to cell surface receptors and its proteolytic cleavage.

J774A.1 cells were plated in 24-well plates and incubated at 4 �C with

400 ng of wild-type or mutant PA. After 15min, the cells were washed

with cold PBS. They were then scraped off and lysed. One hundred

micrograms of cell protein was heated for 3min at 95 �C and subjected

to 10% SDS–PAGE. PA was identified by immunoblot analysis with

anti-PA antiserum. To study the proteolytic cleavage of PA on cell

surface, the same procedure was followed except that PA was incu-

bated with cells for 2 h at 4 �C.

In vitro cleavage of PA and its binding to LF. To study the binding

of LF to PA in solution, PA was cleaved with trypsin (1 ng trypsin per

lg of PA) for 30min at 30 �C in 25mM Hepes, 1mM CaCl2, and

0.5mM EDTA. Trypsin was inactivated with 1mM PMSF and the

samples were analyzed on SDS–PAGE. To study the binding of PA to

LF, nicked PA was incubated with LF (1lg/ml) for 15min in 20mM

Tris, pH 9.0, containing 2mg/ml CHAPS. The samples were then

analyzed on a non-denaturing 5–10% gradient gel.

Toxicity of anthrax toxin proteins in Balb/c mice. For all experi-

ments with Balb/c mice (female mice, 25–28 g), groups of four animals

were taken for each set of conditions. To study the action of lethal

toxin, mice were intravenously injected with a mixture of PA and LF,

with or without the PA-mutants. The final volume of the dose injected

in mice was 100ll (PBS was used for making the dilutions). After

injection, the animals were kept under observation. To study the action

of edema toxin, mice were injected with a mixture of PA and EF, with

or without the PA-mutants. Final dose of 100ll was used for the

injection in the footpad of the mice.

Results and discussion

Site-directed mutagenesis, expression, and purification of

the mutant proteins

The codons for residues Asp425 and Phe427 of PA

were individually deleted by oligonucleotide-directedmutagenesis of the PA gene. The mutations were con-

firmed by sequencing and the mutant constructs were

Fig. 2. Proteolytic cleavage of the PA mutant proteins. The purified

mutant proteins were digested with trypsin (as described previously) at

30 �C for 20min. The samples were analyzed on 12% SDS–PAGE. The

gel was stained with Coomassie blue. Lane U, undigested PA; lane A,

425del after digestion with trypsin; lane B, 427del after digestion with

trypsin; and lane C, wild-type PA after digestion with trypsin.

448 N. Ahuja et al. / Biochemical and Biophysical Research Communications 307 (2003) 446–450

transformed into E. coli BL21(DE3) cells to induce theexpression of the mutated genes. Following induction

with IPTG, the cells were harvested and their periplas-

mic fraction was isolated. The PA-mutants were purified

to near-homogeneity by sequential chromatography on

DEAE–Sepharose and phenyl-Sepharose columns

(Fig. 1). Further experiments were then conducted to

evaluate the biological activity of the mutant proteins.

Biological activity of the mutant proteins

To study the effect of the deletions on the toxicity of

PA, cytotoxicity assays were done on sensitive cell lines.

Different doses of the mutant proteins, 425del and

427del (concentration tested: 0.1, 0.5, 1, 5, and 10 lg/ml), were added to macrophage cell line, J774A.1, along

with 0.5 lg/ml of LF. The viability of the macrophages

was determined after 3 h of incubation with the toxinproteins. It was observed that both the mutant proteins

were completely non-toxic and failed to cause the death

of J774A.1 macrophages at any of the concentrations

tested. On the other hand, even 0.1 lg/ml of wild-type

PA (when added along with 0.5 lg/ml LF) was sufficient

for causing complete lysis of the macrophages. The

toxicity of the mutant proteins was also evaluated on

CHO cells. It was observed that the mutant proteinfailed to elicit elongation response in CHO cells when

added (at concentrations varying between 0.1 and 10 lg/ml) along with 0.1 lg/ml EF. On the other hand, CHO

cells got elongated when treated with 0.1 lg/ml each of

wild-type PA and EF.

Further experiments were then conducted to under-

stand how the deletion of residues Asp425 or Phe427 of

PA abolishes its biological activity in vitro. The first stepof intoxication process is the binding of PA to the re-

ceptors on the surface of the host cells. To study the

binding of the mutant proteins to cell-surface receptors,

the mutant proteins were incubated with CHO cells at

4 �C. After 15min of incubation, the cells were washed

to remove the unbound protein. The cells were then

scraped and lysed. The cell lysate was then resolved on

Fig. 1. Electrophoretic analysis of the PA mutant proteins purified

from E. coli. The mutant proteins were purified from E. coli and an-

alyzed on 12% SDS–PAGE gel that was stained with Coomassie blue.

Lane M, molecular weight standard; lane 1, 425del; lane 2, 427del, and

lane 3, wild-type PA.

SDS–PAGE and immunoblotting was done with anti-

PA antiserum. The immunoblot profile of PA demon-strated that the mutant proteins could not only bind but

also get proteolytically activated on the cell-surface (just

like the wild-type protein) to yield the 63 kDa fragment

(data not shown). This cleavage of PA and the mutant

counterparts could be mimicked in solution by treating

the proteins with trypsin (Fig. 2). The trypsin-digested

proteins were allowed to bind to LF or EF in solution

and were later subjected to non-denaturing polyacryl-amide gel electrophoresis. It was observed that the

mutant proteins could bind to LF or EF and form a

high-molecular weight complex that migrated slowly on

the non-denaturing gel (Fig. 3). This indicated that the

mutant proteins were unimpaired in their ability to oli-

gomerize and bind to LF and EF. Thus, it was inferred

that the deletion of the residues Asp425 and Phe427 of

PA affects step(s) beyond the oligomerization and LF/EF binding and consequently make the mutant proteins

non-toxic when used in combination with LF/EF. It has

been previously demonstrated that alanine-substitution

Fig. 3. Binding of LF to PA mutants in solution. The wild-type PA or

its mutant proteins were treated with trypsin before incubating with

LF for 15min in 20mM Tris containing 2mg/ml CHAPS. The samples

were then loaded on a non-denaturing 5–10% gradient gel. Lane A,

wild-type PA; lane B, LF; lane C, wild-type PA that was digested with

trypsin and incubated with LF; lane D, 425del mutant that was di-

gested with trypsin and incubated with LF; and lane E, 427del mutant

that was digested with trypsin and incubated with LF.

Table 1

PA mutants 425del and 427del inhibit anthrax lethal toxin in Balb/c

mice

Quantity of protein (lg)a Number of survivors/

number challengedWT-PA LF 425del 427del

50 — — — 4/4

— 22 — — 4/4

— — 50 — 4/4

— — — 50 4/4

50 22 — — 0/4

50 22 50 — 4/4

50 22 25 — 4/4

50 22 12.5 — 4/4

50 22 6.25 — 4/4

50 22 3.12 — 0/4

50 22 — 50 4/4

50 22 — 25 4/4

50 22 — 12.5 4/4

50 22 — 6.25 4/4

50 22 — 3.12 0/4

a Female Balb/c mice (25–28 g) were intravenously injected with the

indicated mixture of anthrax toxin proteins. Final volume of the dose

injected in each animal was 100ll.

N. Ahuja et al. / Biochemical and Biophysical Research Communications 307 (2003) 446–450 449

of residues Asp425 and Phe427 blocks the ability of PAto form pore and mediate translocation of LF/EF [26].

Inhibition of anthrax toxin action in vitro

The observation that the mutant proteins were un-

impaired in their ability to bind to cell surface receptors,

get activated to oligomerize and bind to LF/EF,

prompted us to investigate if these mutant proteins could

affect the action of the wild-type toxin proteins in vitro.J774A.1 macrophages were treated with 1 lg/ml each of

wild-type PA and LF in the presence or absence of var-

ious concentrations of the mutant proteins (1, 0.5, 0.25,

and 0.125 lg/ml). It was observed that in the presence of

the mutant proteins, 425del or 427del, the lethal toxin

failed to kill the macrophages. The macrophages were

completely protected against lethal toxin action, even

when the ratio of the mutant protein was 1:8 relative tothe wild-type PA. The results presented above show that

the PA-mutants are defective in mediating LF/EF tox-

icity. It is quite possible that at 1:8 ratio, the single mu-

tant-PA molecule that gets incorporated in majority of

PA heptamers, inactivates these heptamers, and that the

minor population of the PA heptamers that is devoid of

mutant PA is not enough to mediate LF/EF toxicity.

Further, we investigated if the deletion mutants couldprotect the cultured cells against the action of edema

toxin. CHO cells were treated with 0.2 lg/ml each of PA

and EF in the presence or the absence of various con-

centrations of the mutant proteins, 425del or 427del. It

was observed that in the presence of the mutant proteins

(even at a ratio of 1:8 relative to the wild-type PA), the

edema toxin failed to cause elongation of the CHO cells.

These results demonstrate that the mutant proteins,425del and 427del, inhibit anthrax toxin action on cul-

tured cells.

Inhibition of anthrax toxin action in vivo

To test if the mutant proteins could protect animals

against anthrax toxin action, we challenged Balb/c mice

with anthrax toxin proteins in the presence or absence

of these mutant proteins. We observed that femaleBalb/c mice injected with a dose of 22 lg/ml of LF and

50 lg/ml PA died after 12–14 h of injection. However,

the animals survived when various concentrations (50,

25, 12.5, and 6.25 lg/ml) of the mutant proteins, 425del

or 427del, were injected along with the lethal toxin

(Table 1). This demonstrated that the mutant proteins

inhibit the toxicity of anthrax lethal toxin in animals.

We then proceeded to study the effect of the mutantproteins on the toxicity of anthrax edema toxin in an-

imals. We observed that the injection of edema toxin

(50 lg/ml of PA and 22 lg/ml of EF) in the footpad of

Balb/c mice produces characteristic edema at the site of

inoculation within 6–8 h of injection. However, when

the mutant proteins, 425del or 427del, were co-injectedwith the edema toxin (at a minimal ratio of 1:8 relative

to the wild-type PA), there was no edema formation in

the footpad of the mice. It was thus concluded that the

mutant proteins inhibit anthrax toxin action both in

vitro and in vivo.

Hitherto, several mutations have been identified in

PA that inhibit the action of anthrax toxin in vivo

[21,22]. However, the mutations studied thus far inhibitlethal toxin action in vivo at a ratio of 1:4 relative to the

wild-type PA. In continuation of these studies, we

demonstrate here that the deletion mutants, 425del and

427del, protect mice against both lethal toxin and edema

toxin challenge. Moreover, these mutants could inhibit

anthrax toxin action in vivo, even when injected at a

ratio of 1:8 relative to the wild-type protein. Thus, the

427del and 425del mutants are promising candidates toneutralize the action of anthrax toxin. Further studies

are underway to evaluate the therapeutic potential of

these mutants as an adjunct to antibiotics, so that both

toxinemia and bacteraemia associated with anthrax

infection may be curbed.

Acknowledgments

Both N.A. and P.K. have received financial assistance from UGC,

Government of India. M.G. has received financial assistance from

CSIR, Government of India.

References

[1] C. Guidi-Rontani, M. Weber-Levy, E. Labruyere, M. Mock,

Germination of Bacillus anthracis spores within alveolar macro-

phages, Mol. Microbiol. 31 (1999) 9–17.

450 N. Ahuja et al. / Biochemical and Biophysical Research Communications 307 (2003) 446–450

[2] S. Welkos, A. Friedlander, S. Weeks, S. Little, I. Mendelson, In-

vitro characterisation of the phagocytosis and fate of anthrax

spores in macrophages and the effects of anti-PA antibody, J.

Med. Microbiol. 51 (2002) 821–831.

[3] J.L. Stanley, J. Smith, Purification of factor I and recognition of a

third factor of anthrax toxin, J. Gen. Microbiol. 26 (1961) 49–66.

[4] P. Kumar, N. Ahuja, R. Bhatnagar, Anthrax edema toxin requires

influx of calcium for inducing cyclic AMP toxicity in target cells,

Infect. Immun. 70 (2002) 4997–5007.

[5] H. Smith, J. Keppie, Observations on experimental anthrax;

demonstration of a specific lethal factor produced in vivo by

Bacillus anthracis, Nature 173 (1954) 869–870.

[6] S.H. Leppla, The anthrax toxin complex, in: J.E. Alouf, J.H. Freer

(Eds.), Sourcebook of Bacterial Protein Toxins, Academic press,

London, UK, 1991, pp. 277–302.

[7] R. Bhatnagar, S. Batra, Anthrax toxin, Crit. Rev. Microbiol. 27

(2001) 167–200.

[8] K.A. Bradley, J. Mogridge, M. Mourez, R.J. Collier, J.A.T.

Young, Identification of the cellular receptor for anthrax toxin,

Nature 414 (2001) 225–229.

[9] K.R. Klimpel, S.S. Molloy, G. Thomas, S.H. Leppla, 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 89 (1992) 10277–10281.

[10] J.C. Milne, D. Furlong, P.C. Hanna, J.S. Wall, R.J. Collier,

Anthrax protective antigen forms oligomers during intoxication of

mammalian cells, J. Biol. Chem. 269 (1994) 20607–20612.

[11] P. Kumar, N. Ahuja, R. Bhatnagar, Purification of anthrax edema

factor from Escherichia coli and identification of residues required

for binding to anthrax protective antigen, Infect. Immun. 69

(2001) 6532–6536.

[12] V.M. Gordon, S.H. Leppla, E.L. Hewlitt, Inhibitors of receptor-

mediated endocytosis block the entry of Bacillus anthracis

adenylate cyclase toxin but not that of Bordetella pertussis

adenylate cyclase toxin, Infect. Immun. 56 (1988) 1066–1069.

[13] R.O. Blaustein, T.M. Koehler, R.J. Collier, A. Finkelstein,

Anthrax toxin: channel-forming activity of protective antigen in

planar phospholipid bilayers, Proc. Natl. Acad. Sci. USA 86

(1989) 2209–2213.

[14] J.C. Milne, R.J. Collier, pH-dependent permeabilization of the

plama membrane of mammalian cells by anthrax protective

antigen, Mol. Microbiol. 10 (1993) 647–653.

[15] C. Guidi-Rontani, M. Weber-Levy, M. Mock, V. Cabiaux,

Translocation of Bacillus anthracis lethal and oedema factors

across endosomal membranes, Cell. Microbiol. 2 (2000) 259–

264.

[16] S.H. Leppla, Anthrax toxin edema factor: a bacterial adenylate

cyclase that increases cAMP concentration in eukaryotic cells,

Proc. Natl. Acad. Sci. USA 79 (1982) 3162–3166.

[17] N.S. Duesbery, C.P. Webb, S.H. Leppla, V.M. Gordon, K.R.

Klimpel, T.D. Copeland, N.G. Ahn, M.K. Oskarsson, K. Fukas-

awa, K.D. Paull, G.F.V. Woude, Proteolytic inactivation of

MAP-kinase-kinase by anthrax lethal factor, Science 280 (1998)

734–737.

[18] G. Vitale, R. Pellizzari, C. Recchi, G. Napolitani, M. Mock, C.

Montecucco, Anthrax lethal factor cleaves the N-terminus of

MAPKKs and induces tyrosine/threonine phosphorylation of

MAPKs in cultured macrophages, Biochem. Biophys. Res.

Commun. 248 (1998) 706–711.

[19] R. Pellizzari, C. Guidi-Rontani, G. Vitale, M. Mock, C. Monte-

cucco, Anthrax lethal factor cleaves MKK3 in macrophages and

inhibits the LFP/INFc-induced release of NO and TNFa, FEBS

Lett. 462 (1999) 199–204.

[20] A.D. Vinion-Dubiel, M.S. McClain, D.M. Czajkowsky, H.

Iwamoto, D. Ye, P. Cao, W. Schraw, G. Szabo, S.R. Blanke, Z.

Shao, T.L. Cover, A dominant negative mutant of Helicobacter

pylori vacuolating toxin (VacA) inhibits VacA-induced cell

vacuolation, J. Biol. Chem. 274 (1999) 37736–37742.

[21] Y. Singh, H. Khanna, A.P. Chopra, V. Mehra, A dominant-

negative mutant of Bacillus anthracis protective antigen inhibits

anthrax toxin action in vivo, J. Biol. Chem. 276 (2001) 22090–

22094.

[22] B.R. Sellman, M. Mourez, R.J. Collier, Dominant-negative

mutants of a toxin subunit. An approach to therapy of anthrax,

Science 292 (2001) 695–697.

[23] M. Sharma, P.K. Swain, A.P. Chopra, V.K. Chaudhary, Y. Singh,

Expression and purification of anthrax toxin protective antigen

from E . coli, Protein Exp. Purif. 7 (1996) 33–38.

[24] N. Ahuja, P. Kumar, R. Bhatnagar, Rapid purification of

recombinant anthrax-protective antigen under nondenaturing

conditions, Biochem. Biophys. Res. Commun. 286 (2001) 6–11.

[25] R. Bhatnagar, N. Ahuja, R. Goila, S. Batra, S.M. Waheed, P.

Gupta, Activation of phospholipase C and protein kinase C is

required for expression of Anthrax lethal toxin cytotoxicity in

J774A.1 cells, Cell Signal. 11 (1999) 111–116.

[26] B.R. Sellman, S. Nassi, R.J. Collier, Point mutations in anthrax

protective antigen that block translocation, J. Biol. Chem. 276

(2001) 8371–8376.