Bacillus anthracis edema factor substrate specificity: evidence for new modes of action

Abstract

Since the isolation of Bacillus anthracis exotoxins in the 1960s, the detrimental activity of edema factor (EF) was considered as adenylyl cyclase activity only. Yet the catalytic site of EF was recently shown to accomplish cyclization of cytidine 5'-triphosphate, uridine 5'-triphosphate and inosine 5'-triphosphate, in addition to adenosine 5'-triphosphate. This review discusses the broad EF substrate specificity and possible implications of intracellular accumulation of cyclic cytidine 3':5'-monophosphate, cyclic uridine 3':5'-monophosphate and cyclic inosine 3':5'-monophosphate on cellular functions vital for host defense. In particular, cAMP-independent mechanisms of action of EF on host cell signaling via protein kinase A, protein kinase G, phosphodiesterases and CNG channels are discussed.

Keywords: Bacillus anthracis; adenylyl cyclase toxin; anthrax; edema factor; edema toxin.

Figure 1

Structural formulas of the second messenger molecules (A) cAMP; (B) cGMP; (C) the potential novel messenger cCMP.

Figure 2

Entry mechanism and synergistic mechanism of action of Bacillus anthracis exotoxins according to [29,104]. Upon binding of protective antigen (PA) to anthrax receptors (ANTXR, CMG2 or TEM8), cell-associated furin proteolytic activity cleaves PA into PA₂₀ and PA₆₃, which self-associates into ring-shaped heptamers binding three molecules of edema factor EF and/or lethal factor LF. The ANTXR-PA-EF/LF complex is endocytosed into an acidic compartment where the high proton concentration mediates conformational rearrangements of the PA prepore allowing EF and LF translocation into the cytosol. In order to achieve a significant impact even at low toxin concentrations, LF and EF synergistically compromise host defense. LF degrades members of the MAPKK family causing inhibited proliferation and cytokine production in T cells, decreased maturation, mobility and cytokine release in macrophages, manipulated cytokine levels in dendritic cells as well as decreased cytokine production and proliferation in B cells. EF produces exceedingly high cAMP concentrations manipulating gene expression via CREB and cell signaling via protein kinase A (PKA). The consequences are decreased cell motility and cytokine production in macrophages, impaired cytokine release from dendritic cells and inhibited chemotaxis in T cells. Interestingly, EF also targets MAPK signaling via PKA. This crosstalk allows the enzymatic activities of EF and LF to synergize in inhibiting MAPK cascades resulting in effectively preventing T cell activation.

Figure 3

Manipulation of host defense by EF adenylyl cyclase activity (A) and cytidylyl cyclase activity (B). Compared to mammalian membranous ACs (mACs), EF possesses a much higher specific adenylyl cyclase (AC) activity. The excessive cAMP accumulation activates cAMP-dependent protein kinase (PKA), guanine nucleotide exchange factors (GEFs) and ion channels. As a result, host defense is compromised. EF cytidylyl cyclase activity may effect further impairment of immune response by causing intracellular cCMP accumulation. cNMPs are degraded by phosphodiesterases (PDEs). The role of endogenous cCMP as a messenger molecule, the existence of cCMP forming signaling enzymes, PDEs degrading cCMP as well as cCMP transport by multidrug resistance proteins (MRPs) are unknown.

Figure 4

HPLC chromatograms of reaction mixtures consisting of EF and the substrates CTP (20 nM EF and 20 nM CaM) (A); UTP (120 nM EF and 120 nM CaM) (B); ITP (300 nM EF and 300 nM CaM) (C). Samples were withdrawn at indicated reaction times. Chromatograms of standard substances were moved vertically in order to prevent overlapping of the lines. (D) Chromatograms with substrate CTP after 60 min reaction time for active and heat-inactivated enzyme. IS, internal standard. Data were taken from [163].

Figure 5

Structure of EF and interactions with ATP and CTP. Models are based on the PDB structure 1xfv [177]. If not otherwise indicated, atoms are colored as follows: C and some essential H of ligands—grey, O—red, N—blue, P—orange, Mg²⁺—magenta, Ca²⁺—green. (A) Domain organization of EF, adapted from [178]; helices are drawn as cylinders, β-sheets as arrow ribbons, ATP and CaM (uniformly grey) as Connolly surfaces. (B, C) Detailed interactions of ATP (B) and CTP (C) with EF; the nucleotides and the side chains of amino acids within a sphere of ~3 Å around the ligands and Mg²⁺ ions (space fill) are drawn as sticks, heteroatoms as balls, backbone traces as lines; colors of C, H atoms and backbone traces of EF correspond to the domain: C_A—green, C_B—greenblue, switch A—blue, switch B—yellow; w—suggested water molecule; dashed red lines—hydrogen bonds. For details of model generation with the software SYBYL 7.3 (Tripos, L.P., St. Louis, MO, USA), see [163].