cis-Acting elements that control expression of the master virulence regulatory gene atxA in Bacillus anthracis

Abstract

Transcription of the Bacillus anthracis structural genes for the anthrax toxin proteins and biosynthetic operon for capsule is positively regulated by AtxA, a transcription regulator with unique properties. Consistent with the role of atxA in virulence factor expression, a B. anthracis atxA-null mutant is avirulent in a murine model for anthrax. In culture, multiple signals impact atxA transcript levels, and the timing and steady-state level of atxA expression are critical for optimal toxin and capsule synthesis. Despite the apparent complex control of atxA transcription, only one trans-acting protein, the transition state regulator AbrB, has been demonstrated to interact directly with the atxA promoter. Here we employ 5' and 3' deletion analysis and site-directed mutagenesis of the atxA control region to demonstrate that atxA transcription from the major start site P1 is dependent upon a consensus sequence for the housekeeping sigma factor SigA and an A+T-rich upstream element for RNA polymerase. We also show that an additional trans-acting protein(s) binds specifically to atxA promoter sequences located between -13 and +36 relative to P1 and negatively impacts transcription. Deletion of this region increases promoter activity up to 15-fold. Site-directed mutagenesis of a 9-bp palindromic sequence within the region prevents binding of the trans-acting protein(s), increasing promoter activity 7-fold and resulting in a corresponding increase in AtxA and anthrax toxin production. Notably, an atxA promoter mutant that produced elevated levels of AtxA and toxin proteins during culture was unaffected for virulence in a murine model for anthrax.

Fig 1

Schematic representation of the atxA promoter region. Relative sizes and positions of the 5′ or 3′ deletion fragments cloned upstream of a promoterless lacZ gene in pHT304-18z (denoted as pUTE followed by a numeral) are shown, and DNA fragments used for electrophoretic mobility shift assays (denoted in bp) are depicted below the atxA promoter. A 9-bp palindrome sequence is indicated by two adjacent horizontal lines between positions +3 and +21.

Fig 2

SigA consensus sequence-dependent expression of a transcriptional PatxA-lacZ fusion. B. anthracis mutants containing transcriptional PatxA-lacZ fusions were cultured in CACO₃ plus 5% CO₂. β-Galactosidase activity was assessed at the early exponential (2 h), transition (4 h), and stationary (7 h) phases of growth. Specific mutations within the putative SigA −35 and −10 consensus sequences are denoted by lowercase bold letters. A representative growth curve is depicted by the hashed line with diamond symbols.

Fig 3

Specific binding of a trans-acting protein(s) to the atxA promoter region. B. anthracis sigH- and abrB-null soluble cellular extract was mixed with radiolabeled 171-bp PatxA. Soluble cellular extract was obtained from cells cultured in CACO₃ plus 5% CO₂. (A) EMSAs were conducted with increasing concentrations of PatxA (2.5- to 80-fold excesses) and PspoVG (10- to 50-fold excesses) unlabeled competitors. (B) EMSAs using 50-fold more specific and nonspecific unlabeled competitors than radiolabeled PatxA. Unlabeled competitors included 171-bp PatxA (white star), 171-bp PatxA with mutated SigA −35 sequence (gTCCCA; dotted star), and nonspecific 187-bp PspoVG (black star).

Fig 4

5′ and 3′ atxA promoter deletion analysis. 5′ PatxA deletion constructs (A) and 3′ PatxA deletion constructs (B) transcriptionally fused to a promoterless lacZ gene were subjected to β-galactosidase assays. B. anthracis ANR-1 harboring the PatxA-lacZ construct was cultured in CACO₃ plus 5% CO₂, and samples were obtained during the early exponential (2 h), transition (4 h), and stationary (7 h) phases of growth. Symbols: filled triangles, pUTE839/904-bp PatxA; squares, pUTE843/171-bp PatxA; circles, pUTE890/153-bp PatxA; ×, pUTE891/135-bp PatxA; inverted triangles, pUTE915/133-bp PatxA; open triangles, pUTE918/103-bp PatxA; diamonds, pUTE914/85-bp PatxA; solid gray line, empty vector.

Fig 5

atxA promoter sequences required for binding of a trans-acting protein(s). B. anthracis cellular extract obtained from cells cultured in CACO₃ plus 5% CO₂ was incubated with radiolabeled 3′ (A) or 5′ (B) PatxA deletion constructs. Refer to Fig. 1 for schematics of the 3′ and 5′ deletion constructs. Relative quantities of free probe are shown below each lane. Symbols: white stars, 100 ng specific PatxA unlabeled competitor; black stars, 100 ng nonspecific PspoVG unlabeled competitor.

Fig 6

The atxA promoter region contains a 9-bp palindromic sequence required for repressor binding. (A) atxA promoter sequences from −13 to +36 relative to the P1 transcription start site. The palindromic sequence is denoted by bold, underlined letters. Nucleotides mutated using site-directed mutagenesis are denoted by lowercase, gray lettering. (B) EMSA results for cellular extracts obtained from B. anthracis cultured in CACO₃ plus 5% CO₂ and incubated with radiolabeled PatxA probes of sizes 49 bp or 171 bp. Quantitative values of free probe are depicted below each lane. Symbols: white stars, PatxA unlabeled competitor; black stars, PspoVG unlabeled competitor. (C) β-galactosidase activity assays for atxA promoter activity in a parent strain (squares) versus promoter activity in a strain in which the native atxA promoter sequence from +14 to +22 was mutated (triangles). The empty vector control is denoted by diamonds.

Fig 7

AtxA and toxin levels produced by the parent strain and atxA mutants. Production of AtxA (A) and LF, EF, and PA (B) by parent and mutant B. anthracis strains. Culture samples were obtained during the transition phase (4 h) of growth. (A) Samples of cell lysates were subjected to SDS-PAGE and Western blot analysis with rabbit anti-AtxA antibody raised against B. anthracis AtxA or mouse anti-RNAP β (Pol β) antibody raised against E. coli RNA Pol β. (B) Samples of culture supernatants were subjected to slot blot Western analysis using rabbit anti-LF and rabbit anti-EF serum and goat anti-PA antibody raised against B. anthracis proteins.

Fig 8

Virulence of parent and atxA mutants. Survival curves of mice infected intravenously with vegetative B. anthracis are shown. A/J mice were injected i.v. with 1.5 × 10² CFU of the parent (circles; n = 6), 1.9 × 10² CFU of atxA-up (triangles; n = 6), or 1.5 × 10³ CFU of atxA-null (squares; n = 3) vegetative cells.