Protective antigen (PA) is the most extensively studied component of anthrax toxin. The gene for PA is at the
pag locus on the plasmid pXO1 of
B. anthracis [
10]. The gene has been cloned [
11] and sequenced [
12] and found to contain a 2319-base pair (bp) open reading frame, of which 2205 bp encode an A/T-rich (69%), cysteine-free, 735-amino-acid (83-kDa) secreted protein.
The structure of PA was solved by X-ray diffraction [13]. PA is a long flat molecule (100 x 50 x 30 Å) mainly organized into anti-parallel β-sheets. It consists of four distinct domains. Domain 1 (amino acids 1–258) contains two tightly bound calcium ions and a large flexible loop (amino acids 162–175) that encompasses the sequence RKKR167, which is recognized by proteases such as trypsin and furin [14]. Cleavage at this site generates two fragments of 20 kDa and 63 kDa (PA20 and PA63, respectively) that do not dissociate easily in solution [15]; cleavage also exposes a large hydrophobic surface on the remainder of domain 1, called domain1′, which serves as a site for binding LF and EF (edema factor, the third component of anthrax toxin that was not used in these studies).
Domain 2 (amino acids 259–487) contains several long β-strands and forms the core of the membrane-inserted channel [16,17]. It also has a large flexible loop (amino acids 303–319) implicated in membrane insertion [16]. Treatment of PA with low concentrations of chymotrypsin causes a single cleavage at residues 313FFD315 producing 47-kDa and 37- kDa fragments [15,18]. Consistent with a role for this region in membrane insertion, chymotrypsin-nicked PA retains the ability to bind cells and internalize LF but is unable to translocate LF to the cytosol [19]. More recently, studies have shown that the amino acid residues Pro260 [20] and Lys397, Asp425, and Phe427 [21] are also involved in proper membrane insertion by PA63.
Domain 3 (amino acids 488-595) is the smallest of the domains and is thought to be involved in intermolecular interaction during oligomerization and/or LF binding [13,22,23,24]. Domain 4 (amino acids 596–735) is loosely associated with the other three domains and is involved in receptor binding [25,26,27,28,29,30].
