JOURNAL OF BACTERIOLOGY, Oct. 2008, p. 6483–6492
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 190, No. 19
Dual Promoters Control Expression of the Bacillus anthracis Virulence
Cristina Bongiorni,§ Tatsuya Fukushima, Adam C. Wilson, Christina Chiang, M. Cecilia Mansilla,¶
James A. Hoch, and Marta Perego*
Division of Cellular Biology, Department of Molecular and Experimental Medicine, The Scripps Research Institute,
10550 North Torrey Pines Road, La Jolla, California 92037
Received 29 May 2008/Accepted 18 July 2008
The AtxA virulence regulator of Bacillus anthracis is required for toxin and capsule gene expression. AtxA is
a phosphotransferase system regulatory domain-containing protein whose activity is regulated by phosphory-
lation/dephosphorylation of conserved histidine residues. Here we report that transcription of the atxA gene
occurs from two independent promoters, P1 (previously described by Dai et al. [Z. Dai, J. C. Sirard, M. Mock,
and T. M. Koehler, Mol. Microbiol. 16:1171–1181, 1995]) and P2, whose transcription start sites are separated
by 650 bp. Both promoters have ?10 and ?35 consensus sequences compatible with recognition by ?A-
containing RNA polymerase, and neither promoter depends on the sporulation sigma factor SigH. The dual
promoter activity and the extended untranslated mRNA suggest that as-yet-unknown regulatory mechanisms
may act on this region to influence the level of AtxA in the cell.
The virulence of Bacillus anthracis, the causative agent of
anthrax, results from the production of a tripartite toxin and an
antiphagocytic poly-D-glutamic acid capsule (21). The toxin
consists of three proteins: protective antigen (PA), encoded by
the pagA gene; edema factor, encoded by the cya gene; and
lethal factor, encoded by the lef gene. The pagA, cya, and lef
genes are located noncontiguously on the pXO1 virulence plas-
mid, and their expression requires the product of another plas-
mid gene, atxA (9, 17, 22, 32). The AtxA regulator is also
required for transcription of the capBCADE operon for cap-
sule production, which is located on the second virulence plas-
mid of B. anthracis, pXO2 (7, 12, 19, 33). The requirement for
AtxA for cap operon transcription is mediated by the products
of two pXO2 genes, acpA and acpB, which share some simi-
larity with AtxA (11).
In addition to being required for virulence gene expression,
AtxA has also been shown to be a global regulator of gene
expression in B. anthracis. Expression profiling indicated that
the transcription of a variety of genes, located either on the
AtxA belongs to the wide family of phosphotransferase sys-
tem regulatory domain (PRD)-containing proteins (31). PRDs
are structural domains generally found in RNA-binding anti-
terminators or in DNA-binding transcription factors. The ac-
tivity of PRD-containing proteins is generally regulated by
phosphorylation on conserved histidine residues. The phos-
phorylation is carried out by enzymes of the phosphoenolpyru-
vate:sugar phosphotransferase system (for a review, see refer-
ence 10). Phosphorylation of the His199 residue in PRD1 of
AtxA was shown to be necessary for its activity, while phos-
phorylation of His379 in PRD2 was inhibitory. This suggested
that regulation of virulence factor production in B. anthracis
may be linked to carbohydrate metabolism (31).
Virulence factor gene expression is also affected by growth
under specific conditions; capsule synthesis and toxin protein
synthesis are induced when strains are grown in defined media
with bicarbonate at elevated atmospheric CO2concentrations
(?5%). Although this induction requires AtxA, the transcrip-
tion of atxA itself did not appear to be affected by growth under
CO2/bicarbonate conditions compared to growth in air (8, 18).
The atxA gene was reported to be transcribed from a pro-
moter (referred to here as P1) (9) located 99 bp upstream from
the ATG translational start site (Fig. 1). The AbrB protein,
characterized in Bacillus subtilis as a transition state regulator
that suppresses postexponential-phase gene expression during
the exponential phase of growth, was shown to bind to the atxA
promoter region (30) and to repress its transcription (28). In
agreement with these results, deletion of the abrB gene in B.
anthracis resulted not only in elevated atxA expression in ex-
ponential phase but also in earlier transcription and higher
levels of transcription of pagA, cya, and lef compared to the
parental strain (28).
Deletion of sigH, encoding the sigma factor for sporulation
initiation in B. anthracis, was shown to result in a sporulation
defect and to completely inhibit atxA and toxin gene expres-
sion. However, Strauch et al. (30) observed that transcription
of the atxA P1 promoter did not require SigH when the B.
subtilis model system was used.
Here we report that transcription of atxA initiates at an
additional promoter (P2) located within the coding sequence
of a gene on pXO1, orf118 (referred to here as the pXO1-118
gene), which is upstream of and divergently transcribed from
* Corresponding author. Mailing address: Division of Cellular Biol-
ogy, Mail Code MEM-116, Department of Molecular and Experimen-
tal Medicine, The Scripps Research Institute, 10550 North Torrey
Pines Road, La Jolla, CA 92037. Phone: (858) 784-7912. Fax: (858)
784-7966. E-mail: email@example.com.
† Supplemental material for this article may be found at http://jb
‡ Manuscript 19413 from The Scripps Research Institute.
§ Present address: Genencor International, Inc.—A Danisco Divi-
sion, 925 Page Mill Rd., Palo Alto, CA 94304.
¶ Present address: Instituto de Biologia Molecular y Cellular de
Rosario, Suipacha 531, 2000 Rosario, Argentina.
?Published ahead of print on 1 August 2008.
the atxA gene. This study showed that, although the Orf118
protein was not required for atxA expression, its coding se-
quence contained promoter elements critical for full activation
of this gene. Furthermore, our analysis revealed that SigH is
not required for atxA transcription under either air or CO2/
bicarbonate growth conditions.
MATERIALS AND METHODS
Bacterial strains and growth conditions. B. anthracis strain 34F2 (pXO1?
pXO2?) was used throughout this study. Cells were grown in LB medium in air
or in R medium (27) in an atmosphere containing 5% CO2. Antibiotics were
added at the following concentrations: kanamycin, 7.5 ?g/ml; chloramphenicol,
7.5 ?g/ml; spectinomycin, 50 ?g/ml; erythromycin, 5 ?g/ml; and lincomycin, 25
?g/ml. Escherichia coli strains DH5? and TG1 were used for plasmid construc-
tion and propagation. E. coli strains C600 and SCS110 were used for the pro-
duction of unmethylated DNA for transformation in B. anthracis. Antibiotics
were used at the following concentrations: ampicillin, 100 ?g/ml; kanamycin, 30
?g/ml; and spectinomycin, 100 ?g/ml.
Electrocompetent cells of B. anthracis were prepared as described by Koehler
et al. (17).
Construction of pXO1-118, atxA, and sigH deletion strains. The sequences of
the oligonucleotide primers used in this study are shown in Table S1 in the
supplemental material. Plasmid pORICm carrying a temperature-sensitive rep-
lication origin and a chloramphenicol resistance marker was used for construc-
tion of the pXO1-118 gene deletion strain (5). A 720-bp fragment downstream of
the pXO1-118 gene was PCR amplified using oligonucleotides Delta 118Kpn and
Delta 118Bam and cloned in pORICm at the KpnI and BamHI sites. An 860-bp
fragment upstream of the pXO1-118 gene was also PCR amplified using oligo-
nucleotides Delta 118Sal and Delta 118Pst and cloned in the plasmid indicated
above at the SalI-PstI sites. Finally, a blunt-ended spectinomycin cassette (5) was
cloned at the HincII site located between the two cloned fragments in the vector
multiple cloning site. The resulting plasmid (pORICm-?118) was transformed
into strain 34F2 and used to generate a deletion-spectinomycin replacement of
the pXO1-118 gene essentially as described previously (5).
Plasmid pORICm was also used for construction of the atxA deletion strain. A
fragment containing the entire atxA coding region and approximately 500 bp
upstream was PCR amplified using oligonucleotide primers Ba118delta and
AtxA3?Bam and cloned in the EcoRI-BamHI sites of pORICm. The resulting
plasmid, after transformation in dam E. coli strain C600, was digested with BclI
and EcoRV, and the 670-bp excised fragment was replaced by the spectinomycin
cassette as a BamHI-HincII fragment. The resulting plasmid (pORI-?atxA) (Fig.
1) was used to transform strain 34F2 and generate a deletion replacement of the
atxA gene essentially as described previously (5).
A deletion in the sigH gene (BA0093 in the B. anthracis Ames strain; accession
number NC_003997) was generated by using the protocol of Janes and Stibitz
(16) and the temperature-sensitive plasmid pORICm-SceI (3). A 1,090-bp frag-
ment generated by PCR amplification using oligonucleotide primers Basig
HpEco and BasigH3?Bam2 and digested with BamHI was cloned in the HincII-
FIG. 1. pXO1-118-atxA region on the pXO1 virulence plasmid: restriction map of the pXO1 plasmid carrying the pXO1-118 and atxA genes
with the restriction sites relevant for this study. The open arrows indicate the direction of gene transcription. The fragments cloned in the plasmids
are indicated below the restriction map. The directions of transcription of the spectinomycin resistance cassette used for gene deletion are indicated
by arrowheads. The positions of the oligonucleotide primers used in the RT-PCR and primer extension reactions are indicated above the restriction
map. The designations of the primers are shortened for clarity.
6484 BONGIORNI ET AL.J. BACTERIOL.
BamHI sites of pORICm-SceI. A spectinomycin cassette, purified as an EcoRV-
EcoRV fragment, was cloned in the pORICm-SceI-SigH plasmid digested with
BglII and treated with T4 DNA polymerase (New England Biolabs). The result-
ing plasmid, designated pORI-?SigH, generated a spectinomycin cassette inser-
tion in the sigH gene of strain 34F2, giving rise to strain 34F2?sigH.
Each mutant construct was verified by diagnostic PCR using genomic DNA.
All cloned fragments were fully sequenced to ensure the fidelity of the amplifi-
cation reaction. A diagnostic PCR was also carried out with genomic DNA using
atxA-specific primers to ensure that the pXO1 plasmid was not lost during the
Construction of lacZ fusion plasmids. The pTCVlac transcriptional fusion
vector was used for construction of the atxA-lacZ reporter plasmids (24). A map
of the cloned fragments is shown in Fig. 1. The fragment in pAtxA10 was
obtained by PCR amplification using oligonucleotide primers AtxA5?promEco
and AtxA3?Bam. The fragment was digested with EcoRI and EcoRV (the latter
naturally occurring within the atxA sequence), and the purified 200-bp fragment
was cloned in the EcoRI and SmaI sites of pTCVlac. Similarly, fragments in
pAtxA13, pAtxA15, and pAtxA12 were generated with oligonucleotide primers
Delta 118Eco2, Ba118delta, and AtxA5?upEcoRI, respectively, as the forward
primers and primer AtxA3?Bam as the reverse primer. The fragments were
digested with EcoRI and EcoRV and cloned in the EcoRI-SmaI-digested
pTCVlac vector. The fragment in pAtxA12 was also cloned in the pHT315 vector
digested with EcoRI and SmaI to generate pAtxA3 for in trans complementation.
The fragment in pAtxA19 was generated with oligonucleotide primers
AtxA5?upEcoRI and AtxApromBam2. The fragment was digested with EcoRI
and BamHI and cloned in similarly digested pTCVlac. Fragments in pAtxA20
and pAtxA21 were generated with oligonucleotides AtxA5?upEcoRI and
delta118Eco2, respectively, as the forward primers and AtxApromBam3 and
AtxApromBam2, respectively, as the reverse primers. The approximately 500-
and 200-bp digested fragments, respectively, were cloned in the EcoRI-BamHI-
digested pTCVlac vector.
Plasmid pAtxA18 was constructed in three steps. First, the product of PCR
amplification carried out with oligonucleotide primers AtxA5?upEcoRI and
1183?Kpn was cloned in the EcoRI-KpnI sites of the pUC19-derived multiple
cloning site of plasmid pHT315 (2). The resulting plasmid was digested with
KpnI and BamHI and ligated to the KpnI-BamHI-digested PCR product ob-
tained with oligonucleotide primers 1185?Kpn and AtxA3?Bam. The resulting
plasmid contained the entire pXO1 fragment shown in Fig. 1 with a KpnI site and
a base pair deletion engineered at the ATG translational start codon of orf118.
The plasmid was then digested with EcoRI and EcoRV, and the purified 900-bp
fragment was cloned in EcoRI-SmaI-digested pTCVlac to obtain pAtxA18.
In order to ensure that orf118 was no longer translated in this construct, a
translational lacZ fusion construct was generated by cloning a PCR fragment
obtained with oligonucleotide primers P1185?Eco and Ba1185?Sal in pJM115
digested with EcoRI and SalI. This resulted in fusion of the orf118 open reading
frame with the lacZ open reading frame at its internal SalI site. A ?-galactosidase
assay carried out with a B. subtilis strain transformed with this construct resulted
in essentially no activity, confirming that orf118 was not translated (data not
shown) after mutagenesis.
Every construct was fully sequenced to ensure fidelity of the PCR amplification.
mRNA purification, primer extension, and reverse transcriptase PCR (RT-
PCR). mRNA was purified essentially by the method described by Farrell in 1996
(12a). B. anthracis cells were grown in LB medium in air and in R medium in a
5% CO2atmosphere with 0.8% sodium bicarbonate to optical densities at 525
nm of 1.0 and ?0.4, respectively. Cells were collected by rapid filtration with a
0.22-?m Corning cellulose acetate filter system. The membrane with the cells was
washed immediately with liquid nitrogen, and the cells were collected in a conical
tube and stored at ?80°C. The cells were poured into a mortar containing dry ice
and broken by grinding with a pestle. Lysis was checked by phase-contrast
microscopy. The lysate was resuspended in extraction buffer (4 M guanidinium
thiocyanate, 25 mM sodium citrate, 0.5% sarcosyl, 100 mM ?-mercaptoethanol;
citrate buffer was added to 1? final concentration from a 10? concentrate at pH
7). The sample was vortexed until the powder was evenly dispensed in the buffer.
Then 0.1 volume of 3 M sodium acetate (pH 5.5) was added. Cell debris was
removed by centrifugation at 12,000 ? g for 10 min at 4°C. The supernatant was
recovered and extracted once with an equal volume of phenol-chloroform (1:1)
(pH 4.5). After centrifugation for 10 min at 4°C, the upper phase was recovered,
and the RNA was precipitated with 0.75 volume of cold isopropanol. After 1 h
of incubation at ?20°C, the sample was centrifuged for 20 min at 4°C, and the
pellet was washed with 70% cold ethanol and then resuspended in extraction
buffer in order to repeat the phenol-chloroform extraction and isopropanol
precipitation procedure one more time as described above. After the final cen-
trifugation and removal of the 70% ethanol, the pellet was resuspended in H2O
and stored at ?80°C. The concentration of mRNA was determined by spectro-
photometric analysis and by using a denaturing 1% formaldehyde gel (29).
In order to carry out the primer extension reaction, the oligonucleotide prim-
ers were labeled with [?-32P]ATP (4,500 Ci/mmol; MP Biomedicals) using T4
polynucleotide kinase (New England Biolabs) by following the supplier’s recom-
mendations. The primer extension reactions were carried out in 40-?l mixtures
containing approximately 60 ?g of mRNA and 1 ?M labeled oligonucleotide.
The reaction mixtures contained 1 ?l of RNase inhibitor (Perkin Elmer), each
deoxynucleoside triphosphate at a concentration of 0.5 mM, 5 ?M dithiothreitol,
1.5 ?l of Superscript III RT (Invitrogen), and 1? Superscript III RT reaction
buffer. The mixtures were incubated for 1 h at 48°C, and the reactions were
stopped with 5? formamide loading dye (95% formamide, 0.01% bromophenol
blue, 0.01% xylene cyanole). The mixtures were run on 5% acrylamide-8 M urea
gels using Tris-borate-EDTA buffer (29). The gels were dried and exposed to
PhosphorImager plates. The 1 Kb Plus molecular weight marker (New England
Biolabs) was also labeled by using [?-32P]ATP and polynucleotide kinase (New
The RT-PCRs were carried out with mRNA that was treated with RNase-free
DNase. Each extension reaction was performed with a 50-?l mixture that con-
tained 50 ?g of mRNA, 4.5 pmol of oligonucleotide primer AtxART2, each
deoxynucleoside triphosphate at a concentration of 0.5 mM, 5 mM dithiothreitol,
1? reaction buffer, and 0.5 ?l of Superscript III RT (Invitrogen). After 1.5 h of
incubation at 47°C, 5 ?l of each reaction mixture was used for PCR amplification,
using a 50-?l (final volume) mixture containing the AtxART3 and AtxART2 or
AtxA5?upEcoRI and AtxART2 oligonucleotides. Five microliters of each reac-
tion mixture was then analyzed by 1% agarose gel electrophoresis. The PCR
control reaction was an extension reaction performed with a mixture that did not
contain the Superscript III enzyme.
Sequencing reactions. Each sequencing reaction was carried out with a Se-
quenase DNA sequencing kit (USB) and ?-35S-labeled dATP (1,250 Ci/mmol;
MP Biomedicals), using oligonucleotide AtxART4 as the primer. The sequenc-
ing and primer extension reaction mixtures were run on a 5% acrylamide-8 M
urea sequencing gel with Tris-borate-EDTA buffer. The gel was dried and ex-
posed to a PhosphorImager plate.
?-Galactosidase assays. Cultures used for ?-galactosidase assays were grown
in 100 ml of LB medium in air or in R medium containing 0.8% sodium
bicarbonate in a an atmosphere containing 5% CO2. Duplicate samples were
removed at hourly intervals and processed essentially as previously described,
except that they were incubated with lysozyme for 1 h at 37°C and Triton X-100
was used at a final concentration of 0.5% (13).
Western blot analysis. Samples used for Western blot analysis of protective
antigen in culture supernatants, as shown in Fig. 2, were prepared as previously
described (31). Cells were grown in LB medium supplemented with erythromycin
and lincomycin, and supernatants were collected when the optical density at 525
nm of the cells reached 5.0 (which corresponded to approximately 3 h after the
transition to the stationary phase). The lanes contained volumes of supernatant
equivalent to the same cell optical density for all strains. The purified PA protein
used as control was purified from a Bacillus megaterium overexpression system
(MoBiTech Molecular Biotechnology, unpublished data).
Samples for the Western blot analysis shown in Fig. 7 were collected from
cultures grown in R medium with CO2/bicarbonate. Samples were collected at
the mid-exponential phase, at the transition phase, and 3 h into the stationary
phase. For PA analysis, 10 ?l of culture supernatant was run on a 12% sodium
dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis gel. For analysis of
AtxA, AbrB, and SigH, cells were collected by centrifugation, resuspended in 0.1
volume of water containing 5 mg/ml lysozyme and 50 U/ml mutanolysin, and
incubated at room temperature for 30 min. EDTA was added to a final concen-
tration of 0.1 M, and protease inhibitor (Roche) was added according to the
recommendations of the supplier. Samples were sonicated five times for 10 s
each time. Samples were mixed with SDS loading dye and run on 12% (AtxA and
SigH) or 16% (AbrB) SDS-polyacrylamide gel electrophoresis gels. Protein was
detected with an enhanced chemiluminescent light-based kit from Amersham
using a polyclonal antibody.
The production of the AtxA antibody was described by Tsvetanova et al. (31).
The B. subtilis AbrB protein and the B. anthracis SigH protein were used to
immunize rabbits using a standard protocol.
Deletion of the pXO1-118 gene reduces expression of the PA
gene. The pXO1-118 gene is 357 bp upstream of the atxA gene
and is transcribed divergently from it (Fig. 1). The pXO1-118
VOL. 190, 2008 TRANSCRIPTION REGULATION OF AtxA6485
gene encodes a virulence plasmid sensor domain with a high
level of similarity to the sensor domain of the BA2291 sporu-
lation histidine kinase (5, 34). Both the pXO1-118 protein and
its paralogue encoded on the pXO2 virulence plasmid, the
pXO2-61 protein, were shown to inhibit the activity of the
BA2291 kinase and thereby inhibit sporulation initiation in B.
anthracis. A deletion of the pXO1-118 gene was generated by
replacing the entire coding sequence with a spectinomycin
resistance cassette, resulting in strain 34F2?118. While analyz-
ing the role of the pXO1-118 gene in the physiology of B.
anthracis, we found that deletion of this gene resulted in a 50%
decrease in pagA expression (Fig. 2A), as measured by a lacZ
reporter fusion, and a corresponding reduction in the level of
PA in the culture supernatant (Fig. 2B, compare lanes 1 and 3).
In order to test whether this effect was due to the absence of
the pXO1-118 protein, we constructed the in trans comple-
menting plasmid pAtxA3 in the pHT315 replicative plasmid
carrying the pXO1-118 gene coding sequence and 371 bp of
the pXO1-118 gene-atxA intergenic sequence (Fig. 1) (2).
Quantitation of PA expression in culture supernatants of the
parental strain and strain 34F2?118 carrying plasmid pAtxA3
revealed no restoration of PA production in the mutant com-
pared to the parental strain (Fig. 2B, compare lanes 2 and 4).
These results indicated that the pXO1-118 DNA in cis, but not
its gene product or the DNA in trans, was required for full
expression of pagA.
Full expression of atxA requires the DNA region containing
the pXO1-118 gene. Because the expression of pagA was re-
duced by deletion of the pXO1-118 coding sequence but not by
the absence of the pXO1-118 protein, we reasoned that dele-
tion of this DNA region could have affected the expression of
atxA, whose product is required for pagA expression. Thus, we
constructed an atxA-lacZ transcriptional fusion in vector
pTCVlac (pAtxA10) and tested the level of atxA transcription
in the 34F2?118 strain. The fragment in pAtxA10 contains the
P1 promoter previously identified as the atxA promoter (Fig. 1)
(9). The results (data not shown) did not reveal any difference
in the level of transcription between the parental 34F2 strain
and the pXO1-118 gene mutant strain. Therefore, we consid-
ered the possibility that DNA regions upstream of the frag-
ment cloned in pAtxA10 and encompassing the pXO1-118
gene could be involved in atxA transcription.
To examine this, a series of atxA-lacZ fusion plasmids were
constructed that contained fragments extending upstream of
atxA to include the entire pXO1-118-atxA intergenic region
(pAtxA13), the intergenic region and part of the pXO1-118
gene (pAtxA15), and the entire intergenic region and the
pXO1-118 gene region, including the putative rho-independent
terminator that follows it (pAtxA12). When ?-galactosidase
assays were carried out with the 34F2 strains carrying these
atxA-lacZ fusion constructs grown in LB medium (Fig. 3A), it
was observed that the fusion carrying the entire pXO1-118
gene, pAtxA12, had atxA transcriptional activity that was ap-
proximately threefold higher than that of the pAtxA10 fusion.
Using the pAtxA13 and pAtxA15 fusions, which contained the
intergenic region without and with a portion of the pXO1-118
FIG. 2. PA expression in the pXO1-118 gene deletion strain. (A) ?-
Galactosidase assay of the pagA-lacZ reporter in parental strain 34F2
(F) and in the 34F2?118 mutant strain (Œ). Cells were grown in LB
medium containing kanamycin. Time zero is the time of transition
between logarithmic growth and the stationary phase. (B) Western blot
analysis of the PA levels in culture supernatants of parental and
34F2?118 mutant strains complemented with the pXO1-118 gene on
the multicopy vector pHT315. Samples were collected at the time point
indicated by the double arrow in panel A. Lane 1, 34F2/pHT315; lane
2, 34F2/pAtxA3; lane 3, 34F2?118/pHT315; lane 4, 34F2?118/pAtxA3;
lane 5, purified PA (0.01 ?g). Lane M contained Western blot standard
Magic Mark XP (Invitrogen). The molecular masses of bands (in kDa)
are indicated on the left. The bands were quantitated with the Image
Quant software (Amersham-Molecular Dynamics), and the pixel value
for each lane is indicated below the gel.
FIG. 3. Transcriptional analysis of atxA-lacZ fusion constructs in
parental strain 34F2. Cultures used for ?-galactosidase assays were
grown in LB medium, and samples were taken at hourly intervals
before and after the transition (time zero) from exponential phase to
stationary phase. Symbols: ?, pTCVlac; ?, pAtxA10; ?, pAtxA12; E,
pAtxA13; ƒ, pAtxA15; f, pAtxA18; }, pAtxA19; Œ, pAtxA20; F,
pAtxA21; ‚, growth curve for a representative strain (34F2/pAtxA15).
OD525, optical density at 525 nm.
6486BONGIORNI ET AL. J. BACTERIOL.
gene, respectively, resulted in a level of atxA transcription that
was intermediate between that of the fusion with the short
fragment (pAtxA10) and that of the fusion with the long frag-
ment (pAtxA12) (Fig. 3A). Similar patterns of transcription
were observed when cells were grown in R medium with CO2/
bicarbonate, but significantly smaller differences in the level of
activity were observed between the pAtxA12 construct and the
pAtxA10 construct (see Fig. 7B and data not shown).
The levels of AtxA expression observed with each fusion
construct (pAtxA10, pAtxA12, pAtxA13, and pAtxA15) were
identical in the parental 34F2 strain and the 34F2?118 strain
(data not shown), confirming the lack of a role for the pXO1-
118 gene product in controlling atxA transcription. Transcrip-
tion from the pAtxA10 and pAtxA12 constructs was also not
affected by deletion of atxA itself (strain 34F2?atxA) (see Fig.
S1A in the supplemental material), ruling out the possibility of
an autoregulatory mechanism. Furthermore, the level of ?-ga-
lactosidase activity measured in the 34F2 strain carrying plas-
mid pAtxA22, which contained the long fragment but in which
the pXO1-118 gene was replaced by the spectinomycin resis-
tance cassette, was twofold lower than the activity measured in
the strain carrying the pAtxA12 construct (data not shown).
The findings just described again suggested that the DNA
containing the pXO1-118 gene was required for full atxA tran-
scription. In order to confirm this, an additional atxA-lacZ
fusion construct was generated in pTCVlac carrying the long
fragment like pAtxA12 but with the start codon of the pXO1-
118 gene replaced by the GTA sequence in a KpnI restriction
site and a base pair deletion that shifted the translational
frame (see Materials and Methods). ?-Galactosidase assays
carried out with the 34F2 strain containing the resulting plas-
mid, pAtxA18 (Fig. 1), showed that the translational frameshift
mutation in the pXO1-118 gene, which eliminated expression
of the pXO1-118 protein, did not affect the level of transcrip-
tion of atxA, which remained identical to the level obtained
with plasmid pAtxA12 (Fig. 3B).
Altogether, these results indicated that full transcription of
atxA requires the ?900-bp fragment upstream of its translation
start codon, which contains the divergently transcribed pXO1-
The atxA 900-bp upstream region contains an additional
promoter. In order to distinguish whether the 900-bp sequence
upstream of atxA was required for full expression because it
contained a binding site for an activator or contained an ad-
ditional promoter(s), two more atxA-lacZ fusions were gener-
ated with the same 5? end of pAtxA12 but with truncation at
the 3? end (Fig. 1); plasmid pAtxA19 lacked approximately 170
bp at the 3? end and the P1 promoter was deleted, while in
pAtxA20 the entire pXO1-118-atxA intergenic region was de-
leted. A third plasmid, pAtxA21, was also constructed and
contained the intergenic region without the P1 promoter.
?-Galactosidase assays carried out with the 34F2 strains con-
taining these plasmids in LB medium showed that the tran-
scription generated from the pAtxA21 reporter was not signif-
icantly different from the transcription observed with the
control vector pTCVlac alone, indicating that this fragment is
unlikely to carry an active promoter (Fig. 3A). In contrast,
efficient transcription was observed with the pAtxA19 and
pAtxA20 constructs, suggesting that promoter activity inde-
pendent of the P1 promoter was present. The transcription of
atxA from the pAtxA19 and pAtxA20 constructs initiated at the
same time during growth, as observed with the pAtxA10 and
pAtxA12 constructs but with different rates. The initial rate of
transcription was only 20% lower for pAtxA19 than for
pAtxA12, but it was 80% lower for pAtxA20 than for pAtxA12.
After 4 h of growth, the level of transcription from the frag-
ment in pAtxA19 was approximately the same as the level
obtained with the long fragment in pAtxA12, while the level of
transcription from pAtxA20 was approximately the same as the
level of transcription from the short fragment in pAtxA10. The
same pattern of gene expression was observed for all constructs
when cells were grown in R medium under CO2/bicarbonate
conditions (data not shown).
Altogether, these results suggest that the 900-bp region up-
stream of the atxA gene, which includes the pXO1-118 gene,
most likely contains an additional promoter for atxA expres-
sion. Furthermore, the pAtxA21 results indicate that this pu-
tative promoter is in the pXO1-118 gene coding region and not
in the intergenic region between the pXO1-118 gene and atxA.
The atxA transcript starts within the pXO1-118 gene. In
order to confirm that an atxA mRNA transcript initiated within
the sequence of the pXO1-118 gene, RT-PCRs were carried
out with DNase-treated mRNA from strain 34F2 grown in LB
medium using primer AtxART2 in the RT reaction and
the oligonucleotide primer pairs AtxART2-AtxART3 and
AtxART2-AtxA5?upEcoRI for PCR amplification (Fig. 1).
The results shown in Fig. 4 indicated that an amplified product
was obtained with the AtxART3 primer, located within the
pXO1-118 gene, but not with primer AtxA5?upEcoRI, which is
located upstream of the putative terminator that follows the
These results confirmed that atxA mRNA extended into the
pXO1-118 gene region and supported the hypothesis that there
was a promoter within the gene that contributed to atxA
transcription in addition to the P1 promoter identified by
Dai et al. (9).
FIG. 4. RT-PCR analysis of the atxA mRNA. RT-PCR was carried
out with mRNA extracted from strain 34F2 and treated with RNase-
free DNase. The mRNA was first incubated with (?) or without (?)
RT and oligonucleotide primer AtxART2 (RT2 in Fig. 1). An aliquot
of the reaction mixture was then incubated with oligonucleotide prim-
ers AtxART2 and AtxART3 (RT3 in Fig. 1) (lanes 1 and 2) or oligo-
nucleotide primers AtxART2 and AtxA5?upEcoRI (5?up in Fig. 1)
(lanes 4 and 5). Lanes 3 and 6 contained the products of PCR ampli-
fications carried out with oligonucleotides AtxART2 and AtxART3
and with oligonucleotides AtxART2 and AtxA5?upEcoRI, respec-
tively, using 34F2 genomic DNA as the template. Lane M contained
the 1 Kb Plus DNA ladder (New England Biolabs). Molecular sizes (in
base pairs) are indicated on the right.
VOL. 190, 2008TRANSCRIPTION REGULATION OF AtxA 6487
Identification of the additional atxA transcription start site.
In order to precisely determine the additional start site of atxA
transcription, RT reactions were carried out with mRNA ex-
tracted from parental strain 34F2. Using the oligonucleotide
primer AtxART2, whose sequence was identical to that of the
primer used by Dai et al. (9), we noticed that, in addition to an
approximately 180-bp product that defined the P1 start site, a
very weak but consistent approximately 700-base product was
also generated (see Fig. S2 in the supplemental material). A
primer extension reaction was then carried out with primer
AtxART1, which overlaps the P1 start site (Fig. 1). A slightly
stronger, defined product consisting of two approximately 600-
base bands was consistently observed on 4% acrylamide-8 M
urea gels. This result was obtained with mRNA extracted from
cells grown in either LB medium or R medium (see Fig. S2 in
the supplemental material; data not shown). This confirmed
again that the atxA mRNA extended upstream of the P1 start
site. With a third oligonucleotide primer, AtxART4, located
within the pXO1-118 gene, two bands at approximately 180
and 250 bases were also clearly obtained (see Fig. S2 in the
The product of the RT reaction carried out with the
AtxART4 primer was then run on a sequencing gel alongside
the product of a sequencing reaction carried out with the same
primer (Fig. 5A). This allowed identification of the start site of
transcription of the longer product as an adenine base (P2),
which is preceded at the expected distance by a ?10 canonical
sequence for ?A-containing RNA polymerase (TATAAT). A
possible ?35 consensus sequence (ATGAAT) was also iden-
tified upstream and was separated by 17 bp from the ?10
sequence. The shorter and weaker product of the RT reaction
identified a guanine nucleotide as a putative start site (P3).
This residue is also preceded by a putative ?10 consensus
sequence (AATTAC), whose first four residues could suggest
involvement of a ?H-containing RNA polymerase in the tran-
scription initiating at this site. However, sequences with simi-
larity to consensus sequences for ?A?35 and ?10 sites could
be identified (ATGATC-18 bp-TTCAAT) upstream of this
site, raising the possibility that the product is indeed a product
of the RT reaction and not an artifact (Fig. 5B).
Altogether, these results showed that a second promoter, P2,
and perhaps a third promoter, P3 (Fig. 1), in addition to the P1
promoter identified by Dai et al. (9) contribute to full expres-
sion of the atxA gene.
Transcription of atxA is independent of the SigH sigma
factor. A requirement for the SigH sporulation sigma factor
for transcription of atxA was recently reported (14). Our find-
ing that multiple promoters contribute to atxA expression
prompted us to analyze whether the SigH requirement applied
to any specific atxA promoter. An inactivation/spectinomycin
insertion of the sigH gene was generated in the 34F2 strain (see
Materials and Methods), and, as previously reported (14), the
resulting 34F2?sigH strain was sporulation deficient (data not
FIG. 5. Mapping the P2 5? end of the atxA mRNA transcript. (A) Sequencing reactions (lanes G, A, T, and C) were carried out with plasmid
pAtxA12 using oligonucleotide primer AtxART4, and the mixtures were run on a 4% acrylamide-8 M urea gel alongside the mixture for a primer
extension reaction carried out with the same oligonucleotide (lane RT4). The nucleotide sequence surrounding the 5? ends (indicated by an
asterisk) of the noncoding strand is shown. (B) Nucleotide sequence of the pXO1 plasmid region containing the pXO1-118 gene and the
pXO1-118-atxA intergenic region. The ATG start codons for atxA and the pXO1-118 gene are enclosed in boxes, and the open arrows indicate the
direction of transcription. The stop codons for the pXO1-118 gene are also indicated. Putative ?10 and ?35 sequences are underlined. The region
protected by AbrB in DNase footprinting analysis (30) is indicated by overlining.
6488BONGIORNI ET AL.J. BACTERIOL.
shown) and did not produce any SigH protein as determined by
Western blot analysis (Fig. 6).
The atxA-lacZfusion constructs
pAtxA13, pAtxA15, pAtxA19, pAtxA20, and pAtxA21 were
introduced by electroporation into the 34F2?sigH mutant
strain, and ?-galactosidase assays were carried out in LB me-
dium in air or in R medium in a 5% CO2atmosphere. ?-Ga-
lactosidase assays carried out with the resulting strains grown
in LB medium showed no inhibition of atxA expression; rather,
they showed a slight, but consistent, increase in expression
(compare Fig. 3A with Fig. 7A and Fig. S3 in the supplemental
material). An increase in ?-galactosidase activity in the sigH
mutant compared to the parental strain was also observed with
the pAtxA10 and pAtxA12 constructs when cells were grown in
R medium with CO2/bicarbonate, indicating that this effect was
not specific to one growth condition (Fig. 7B). The same re-
sults were obtained with a sigH mutant generated in a different
background, Sterne strain 7702, indicating that the effect was
not strain specific (data not shown).
Notably, neither the time of induction of atxA transcription
(approximately 1 h before the transition from the vegetative to
stationary phase) nor the initial level of activity in R medium
(Fig. 7B) was affected by the sigH mutation in either the LB or
R medium growth conditions.
AtxA and PA protein analysis by Western blotting carried
out with cell lysates and culture supernatants, respectively,
confirmed that sigH inactivation did not eliminate the produc-
tion of these proteins (Fig. 6).
Altogether, these results indicate that inactivation of the
sigH gene in B. anthracis blocks sporulation initiation but does
not affect atxA gene transcription and, consequently, toxin
Toxin and capsule synthesis in B. anthracis requires the
product of the atxA gene. Consistently, an atxA-null mutant
was shown to be avirulent in mice and to elicit a lower antibody
response to toxin proteins than the parental strain (9). The
relevance of AtxA in the virulence of B. anthracis is likely to be
associated with complex mechanisms regulating its expression
and/or activity. The evidence that this is the case includes the
finding that AtxA is phosphorylated in vivo and the finding that
its activity is regulated by opposing effects of phosphorylation
on two conserved histidine residues within PRDs. Phosphory-
lation on His199 of PRD1 is required for AtxA activity, while
phosphorylation of His379 inhibits the regulator (31).
Notably, the virulence regulator Mga in the group A strep-
tococcus (Streptococcus pyogenes) shares structural domain
similarities with B. anthracis AtxA as it also contains two DNA-
binding domains at the amino terminus and two PRDs in the
central domain. Expression of mga is autoregulated, and its
promoter has two transcriptional start sites, one of which is
activated by the carbon catabolite regulator protein CcpA (1,
In this work, previous studies on transcription regulation of
the atxA gene were extended in order to better understand the
FIG. 6. Western blot analysis of protein levels in parental strain
34F2 and the 34F2?sigH strain. Samples were collected at the mid-
exponential phase (lanes 1, 4, and 7), transition phase (lanes 2, 5, and
8), and stationary phase (lanes 3, 6, and 9). Lanes 1 to 3 contained
samples from parental strain 34F2; lanes 4 to 6 contained samples from
the sigH mutant strains; and lanes 7 to 9 contained samples from the
abrB mutant strains. Lane M contained the Magic Mark XP standard
(Invitrogen); the molecular mass of the smaller band in this marker is
20 kDa, and that is why no band is present in this lane in the AbrB
panel. Lane H contained purified B. anthracis SigH (0.05 ?g). Lane A
contained purified B. anthracis AtxA protein (0.006 ?g). Lane P con-
tained purified B. anthracis PA (0.1 ?g). Lane B contained lysate of the
B. subtilis JH642 parental strain as a positive control for AbrB (10.6
kDa). The sizes (in kDa) of the bands for the Magic Mark XP standard
visible in each panel are indicated on the right.
FIG. 7. Transcription analysis of the atxA promoter in the sigH
mutant strain. ?-Galactosidase assays were carried out using LB me-
dium (A) or R medium under CO2/bicarbonate growth conditions (B),
and samples were removed at hourly intervals. (A) Strain 34F2?sigH
carrying the following plasmids: pTCVlac (?), pAtxA10 (?), pAtxA12
(?), pAtxA13 (E), pAtxA15 (ƒ), pAtxA19 (}), pAtxA20 (Œ), and
pAtxA21 (F). ‚, representative growth curve (pAtxA10). (B) Direct
comparison of pAtxA10 and pAtxA12 transcription in parental strain
34F2 and the sigH mutant strain grown in R medium. Symbols: ?, 34F2/
pAtxA10; f, 34F2?sigH/pAtxA10; ?, 34F2/pAtxA12; }, 34F2?sigH/
pAtxA12; ‚, representative growth curve (34F2/pAtxA10). OD525, optical
density at 525 nm.
VOL. 190, 2008 TRANSCRIPTION REGULATION OF AtxA6489
possible contribution of regulation of gene expression in B.
The first promoter identified as a promoter responsible for
atxA transcription (P1) (9) is approximately 100 bp upstream of
the translational start site and is presumably dependent on
RNA polymerase containing the ?Asigma factor. This pro-
moter is also subject to repression by the AbrB transition state
regulator, as shown by ?-galactosidase assays and DNase foot-
printing analysis (28, 30). The region protected from DNase
digestion by the AbrB protein overlaps the putative ?35 region
of P1 (Fig. 5B), and derepression of atxA expression in an abrB
mutant strain was observed in the B. subtilis model system
using a lacZ-reporter fragment extending as little as 200 bp
upstream of the P1 start site (30). This is consistent with the
hypothesis that AbrB acts as a preventer of RNA polymerase
binding to the atxA P1 promoter (15). The same study also
showed that this promoter was not affected by the absence of
the SigH sigma factor because, in the absence of the repressor
effect of AbrB, the P1 promoter was transcribed in a B. subtilis
abrB-sigH double mutant at the level observed in the abrB
single-mutant strain. This is in contrast to the later report by
Hadjifrangiskou et al. (14), which described the requirement
for the SigH sporulation sigma factor in atxA transcription in
Our analysis of atxA transcription in parental strain 34F2
and the sigH mutant revealed that two promoters are respon-
sible for this transcription, the proximal previously described
promoter P1 (9), whose start site is located 101 bp upstream of
the translational start, and a distal promoter, P2, which is
located 744 bp upstream of the ATG start codon. A possible
third promoter, P3, whose start site is 686 bp upstream of the
atxA first codon, may also contribute to atxA transcription
because a consistent, although weaker-than-P2, 5? end of
mRNA was detected with two different oligonucleotides in the
RT primer extension assay. P2 and P3 are within the coding
sequence of the divergently transcribed pXO1-118 gene. As
shown previously for the P1 promoter (9), P2 and P3 were not
affected by growth under CO2/bicarbonate conditions (see Fig.
S1B in the supplemental material). Nonetheless, the expres-
sion of atxA is consistently three- to fourfold higher in cells
grown in R medium than in cells grown in LB medium (see Fig.
S1 in the supplemental material; data not shown).
The presence of a canonical ?10 consensus sequence and a
?35 region with three bases conserved with the canonical
sequence TTGACA strongly suggests that P2 is also most likely
transcribed by ?A-containing RNA polymerase. Consensus se-
quences for P3 are more divergent but still compatible with ?A
recognition. Neither the P1 promoter nor the P2 and P3 pro-
moters were found to be dependent on SigH for transcription
regardless of the bacterial growth conditions used for the assay
(LB medium in air or R medium in CO2/bicarbonate) (Fig. 7
and data not shown).
These conclusions contradict the conclusions reached by
Hadjifrangiskou et al. (14) but are in line with the observations
made by Strauch et al. (30). Several lines of reasoning support
our conclusions. (i) The weak product of an in vitro transcrip-
tion assay carried out for the atxA promoter region with puri-
fied B. anthracis SigH protein and E. coli core RNA polymer-
ase suggested to Hadjifrangiskou et al. (14) that a possible atxA
sigH-dependent promoter could be located approximately 250
bp upstream of the translational start site, in the pXO1-118-
atxA intergenic region, despite the absence of canonical ?35
and ?10 consensus sequences for ?H-containing RNA poly-
merase. Our studies did not identify any 5? mRNA end in this
intergenic region and did not detect any significant transcrip-
tional activity greater than the activity of the negative control
plasmid pTCVlac when this intergenic region was tested using
construct pAtxA21 (Fig. 3A; see Fig. S2 in the supplemental
material). Nevertheless, the presence of this intergenic region
in the pAtxA13 and pAtxA15 constructs resulted in higher
?-galactosidase activity than in the pAtxA10 construct contain-
ing the P1 promoter alone in the short fragment. Similarly, the
construct in pAtxA19, which contained P2 and part of the
intergenic region but did not contain P1, generated more ex-
pression of the atxA-lacZ fusion than the construct in pAtxA20
containing only P2. Whether the effect exerted by this approx-
imately 170-bp region is intrinsic to the DNA sequence or
results from the presence in it of the pXO1-118 gene promoter
determinants (unpublished data) has not been determined.
The presence of a binding site of an activator that can act on
either the upstream or downstream promoter is another pos-
(ii) The seemingly total dependence of atxA transcription on
SigH reported by Hadjifrangiskou et al. (14) using an atxA-
lacZ reporter extending from the 5? end of atxA, like our
pAtxA12 construct, implies that all promoters (P1, as well as
P2 and P3) are recognized by this sigma factor. This would be
a very unusual situation as, to our knowledge, tandem SigH
promoters have never been identified. The activities of the P1
and P2/P3 promoters of atxA are additive but independent of
each other, as each promoter can be transcribed in the absence
of the other (cf. the activity of pAtxA10 and the activity of
pAtxA19, as shown in Fig. 3A), ruling out the possibility that in
the absence of SigH the lack of transcription from one of the
atxA promoters reduces transcription of the others.
(iii) The hypothesis proposed by Hadjifrangiskou et al. (14)
to justify the dependence of atxA transcription from SigH de-
spite the absence of canonical ?35 and ?10 consensus se-
quences similar to those identified in B. subtilis implied that the
B. anthracis orthologue may recognize different binding se-
quences and/or have a more relaxed specificity. Although this
hypothesis cannot be completely ruled out, the current knowl-
edge concerning sigma factors, and SigH in particular, suggests
that it is highly unlikely. The SigH proteins of B. subtilis and B.
anthracis share 75% identical residues and 20% conserved
substitutions. The lowest level of similarity was found in the
N-terminal 15 amino acids, suggesting to Hadjifrangiskou et al.
(14) that the divergence resulted in altered target specificity
and binding affinity in a manner similar to the one proposed by
Ramirez-Romero et al. (26) for ?70of Rhizobium etli. How-
ever, lax promoter recognition was proposed by Ramirez-Ro-
mero et al. to be associated with the amino acid differences in
amino-terminal region 1 between R. etli ?70(685 amino acids)
and E. coli ?70(613 amino acids), but this region is absent in
?H(213 amino acids), making the argument inconsistent. Fur-
thermore, two divergent residues in SigH of B. subtilis and B.
anthracis (G160S and V172S, B. anthracis numbering) were
also proposed to perhaps affect promoter specificity (14).
These residues are in region 4 of the sigma factor, but they are
not in helix-turn-helix region 4.2 that directly binds the ?35
6490 BONGIORNI ET AL.J. BACTERIOL.
sequence and provides sigma factor specificity; thus, they are
unlikely to affect the interaction with the DNA based on the
structure of region 4 of Thermus aquaticus ?A, a member of the
Although the reason for the conflicting results concerning
the role of SigH is unknown, it is clear that at least in the
genetic backgrounds used in this study (B. anthracis Sterne
strains 34F2 and 7702) the absence of the SigH protein results
in the expected sporulation-deficient phenotype but does not
affect AtxA and toxin gene expression.
Nevertheless, in the sigH mutant, the atxA gene is tran-
scribed at a consistently higher level than it is in the parental
strain, particularly when cells are grown in a rich medium (such
as LB medium) (see Fig. S3 in the supplemental material).
This may be due to the fact that in a rich medium, the parental
strain may have a regulatory mechanism that reduces atxA
expression if conditions conducive to sporulation are present.
It is becoming more apparent, in fact, that efficient sporulation
negatively affects the virulence potential of B. anthracis (5, 23).
In the sigH mutant this negative regulation might be missing
because the cells cannot commit to spore formation, resulting
in higher levels of atxA expression. In view of the hypothesis
that efficient sporulation is detrimental to virulence gene ex-
pression (23), the requirement for the SigH sigma factor,
which is necessary for cell commitment to the developmental
process, in toxin production seems paradoxical (14). In the
spo0H mutant, a negative effect on atxA expression by the
AbrB repressor is not expected because in B. anthracis, as well
as in B. subtilis, the lack of SigH does not have a significant
impact on the level of transcription of abrB and its product in
the cell (Fig. 6 and data not shown). This is because the level
of phosphorylated Spo0A protein necessary to repress the
transcription of abrB does not require the induction of spo0A
transcription brought about by SigH at the onset of sporulation
(25; M. Perego, unpublished data).
As observed for the Mga protein of S. pyogenes, the pres-
ence of multiple promoters transcribing the atxA gene is not
surprising. However, the distance of the P2 and P3 promot-
ers (650 bp upstream of P1 and within a divergently tran-
scribed open reading frame) is considerably greater than the
distance in the mga promoter (295 bp) (21), possibly pro-
viding additional space for a regulatory effector(s) acting at
the DNA or RNA level.
In contrast to the Mga system of S. pyogenes, the AtxA
protein does not have autoregulatory functions (see Fig. S1A
in the supplemental material), and it does not appear to
require the CcpA carbon catabolite regulator protein for
full induction (1, 21; unpublished data). Nevertheless, evi-
dence of additional regulatory mechanisms for atxA expres-
sion is emerging from a transposon mutagenesis analysis
(35; A. C. Wilson, M. Perego, and J. A. Hoch, unpublished
data), and characterization of these mechanisms is under
way in our laboratory.
This study was supported in part by grant AI055860 from the
National Institute of Allergy and Infectious Diseases, by grants
GM019416 and GM055594 from the National Institute of General
Medical Sciences, U.S. Public Health Service, and by grant
CI000095 from the Centers for Disease Control and Prevention and
the National Center for Infectious Diseases. Oligonucleotide syn-
thesis and DNA sequencing costs were supported in part by the
Stein Beneficial Trust.
We acknowledge Sophie Stephenson, Chandra La Clair, Joelle
Jensen, and Sharon Tang-Black for technical support and J. Michael
Green for PA purification. We thank Scott Stibitz for providing B.
anthracis strain 7702 and an anonymous reviewer who suggested the
?35 and ?10 possible promoter sequences of P3.
1. Almengor, A. C., T. L. Kinkel, S. J. Day, and K. S. McIver. 2007. The
catabolite control protein CcpA binds to Pmga and influences expression of
the virulence regulator Mga in the group A streptococcus. J. Bacteriol.
2. Arantes, O., and D. Lereclus. 1991. Construction of cloning vectors for
Bacillus thuringiensis. Gene 108:115–119.
3. Bongiorni, C., R. Stoessel, and M. Perego. 2007. Negative regulation of
Bacillus anthracis sporulation by the Spo0E family of phosphatases. J. Bac-
4. Bourgogne, A., M. Drysdale, S. G. Hilsenbeck, S. N. Peterson, and T. M.
Koehler. 2003. Global effects of virulence gene regulators in a Bacillus
anthracis strain with both virulence plasmids. Infect. Immun. 71:2736–2743.
5. Brunsing, R. L., C. La Clair, S. Tang, C. Chiang, L. E. Hancock, M. Perego,
and J. A. Hoch. 2005. Characterization of sporulation histidine kinases of
Bacillus anthracis. J. Bacteriol. 187:6972–6981.
6. Campbell, E. A., O. Muzzin, M. Chlenov, J. L. Sun, C. A. Olson, O.
Weinman, M. L. Trester-Zedlitz, and S. A. Darst. 2002. Structure of the
bacterial RNA polymerase promoter specificity sigma subunit. Mol. Cell
7. Candela, T., M. Mock, and A. Fouet. 2005. CapE, a 47-amino-acid peptide,
is necessary for Bacillus anthracis polyglutamate capsule synthesis. J. Bacte-
8. Dai, Z., and T. M. Koehler. 1997. Regulation of anthrax toxin activator gene
(atxA) expression in Bacillus anthracis: temperature, not CO2/bicarbonate,
affects AtxA synthesis. Infect. Immun. 65:2576–2582.
9. Dai, Z., J. C. Sirard, M. Mock, and T. M. Koehler. 1995. The atxA gene
product activates transcription of the anthrax toxin genes and is essential for
virulence. Mol. Microbiol. 16:1171–1181.
10. Deutscher, J., C. Francke, and P. W. Postma. 2006. How phosphotransferase
system-related protein phosphorylation regulates carbohydrate metabolism
in bacteria. Microbiol. Mol. Biol. Rev. 70:939–1031.
11. Drysdale, M., A. Bourgogne, S. G. Hilsenbeck, and T. M. Koehler. 2004. atxA
controls Bacillus anthracis capsule synthesis via acpA and a newly discovered
regulator, acpB. J. Bacteriol. 186:307–315.
12. Drysdale, M., A. Bourgogne, and T. M. Koehler. 2005. Transcriptional
analysis of the Bacillus anthracis capsule regulators. J. Bacteriol. 187:
12a.Farrell, R. E., Jr., (ed.). RNA methodologies: a laboratory guide for isolation
and characterization, 2nd ed., p. 67–84. Academic Press, London, United
13. Ferrari, E., S. M. H. Howard, and J. A. Hoch. 1986. Effect of stage 0
mutations on subtilisin expression. J. Bacteriol. 166:173–179.
14. Hadjifrangiskou, M., Y. Chen, and T. M. Koehler. 2007. The alternative
sigma factor ?His required for toxin gene expression by Bacillus anthracis. J.
15. Hoch, J. A. 1993. Regulation of the phosphorelay and the initiation of
sporulation in Bacillus subtilis. Annu. Rev. Microbiol. 47:441–465.
16. Janes, B. K., and S. Stibitz. 2006. Routine markerless gene replacement in
Bacillus anthracis. Infect. Immun. 74:1949–1953.
17. Koehler, T. M., Z. Dai, and M. Kaufman-Yarbray. 1994. Regulation of
the Bacillus anthracis protective antigen gene: CO2and a trans-acting
element activate transcription from one of two promoters. J. Bacteriol.
18. Leppla, S. H. 1988. Production and purification of anthrax toxin. Methods
19. Makino, S., I. Uchida, N. Terakado, C. Sasakawa, and M. Yoshikawa.
1989. Molecular characterization and protein analysis of the cap region,
which is essential for encapsulation in Bacillus anthracis. J. Bacteriol.
20. McIver, K. S., A. S. Thurman, and J. R. Scott. 1999. Regulation of mga
transcription in the group A streptococcus: specific binding of mga within its
own promoter and evidence for a negative regulator. J. Bacteriol. 181:5373–
21. Okada, N., R. T. Geist, and M. G. Caparon. 1993. Positive transcriptional
control of mry regulates virulence in the group A streptococcus. Mol. Mi-
22. Okinaka, R. T., K. Cloud, O. Hampton, A. R. Hoffmaster, K. K. Hill, P.
Keim, T. M. Koehler, G. Lamke, S. Kumano, J. Mahillon, D. Manter, Y.
Martinez, D. Ricke, R. Svensson, and P. J. Jackson. 1999. Sequence and
organization of pXO1, the large Bacillus anthracis plasmid harboring the
anthrax toxin genes. J. Bacteriol. 181:6509–6515.
23. Perego, M., and J. A. Hoch. 2008. Commingling regulatory systems following
VOL. 190, 2008TRANSCRIPTION REGULATION OF AtxA6491
acquisition of virulence plasmids by Bacillus anthracis. Trends Microbiol. Download full-text
24. Poyart, C., and P. Trieu-Cuot. 1997. A broad-host-range mobilizable shuttle
vector for the construction of transcriptional fusions to ?-galactosidase in
Gram-positive bacteria. FEMS Microbiol. Lett. 156:193–198.
25. Predich, M., G. Nair, and I. Smith. 1992. Bacillus subtilis early sporulation
genes kinA, spo0F, and spo0A are transcribed by the RNA polymerase
containing ?H. J. Bacteriol. 174:2771–2778.
26. Ramirez-Romero, M. A., I. Masulis, M. A. Cevallos, V. Gonzalez, and G.
Davila. 2006. The Rhizobium etli ?70(SigA) factor recognizes a lax consensus
promoter. Nucleic Acids Res. 34:1470–1480.
27. Ristroph, J. D., and B. E. Ivins. 1983. Elaboration of Bacillus anthracis
antigens in a new, defined culture medium. Infect. Immun. 39:483–486.
28. Saile, E., and T. M. Koehler. 2002. Control of anthrax toxin gene expression
by the transition state regulator abrB. J. Bacteriol. 184:370–380.
29. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a
laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold
Spring Harbor, NY.
30. Strauch, M. A., P. Ballar, A. J. Rowshan, and K. L. Zoller. 2005. The
DNA-binding specificity of the Bacillus anthracis AbrB protein. Microbiol-
31. Tsvetanova, B., A. C. Wilson, C. Bongiorni, C. Chiang, J. A. Hoch, and M.
Perego. 2007. Opposing effects of histidine phosphorylation regulate the
AtxA virulence transcription factor in Bacillus anthracis. Mol. Microbiol.
32. Uchida, I., J. M. Hornung, C. B. Thorne, K. R. Klimpel, and S. H. Leppla.
1993. Cloning and characterization of a gene whose product is a trans-
activator of anthrax toxin synthesis. J. Bacteriol. 175:5329–5338.
33. Uchida, I., S. Makino, T. Sekizaki, and N. Terakado. 1997. Cross-talk to the
genes for Bacillus anthracis capsule synthesis by atxA, the gene encoding the
trans-activator of anthrax toxin synthesis. Mol. Microbiol. 23:1229–1240.
34. White, A. K., J. A. Hoch, M. Grynberg, A. Godzik, and M. Perego. 2006.
Sensor domains encoded in Bacillus anthracis virulence plasmids prevent
sporulation by hijacking a sporulation sensor histidine kinase. J. Bacteriol.
35. Wilson, A. C., M. Perego, and J. A. Hoch. 2007. New transposon delivery
plasmids for insertional mutagenesis in Bacillus anthracis. J. Microbiol.
6492BONGIORNI ET AL.J. BACTERIOL.