Acidic, basic, and aromatic residues in the CMG2/ANTXR2 vWA domain important for receptor function

Bacillus anthracis, the causative agent of anthrax, is a large, rod-shaped, Gram-positive bacterium. Its pathogenesis is dependent on two important virulence factors: an antiphagocytic capsule and anthrax toxin. Anthrax toxin refers collectively to three monomeric, plasmid-encoded proteins: lethal factor (LF), edema factor (EF), and Protective antigen (PA). Both LF and EF are enzymes which intoxicate cells. PA is nontoxic, but has a role in delivery of the catalytic moieties LF and EF into the cell. This is accomplished through receptor-mediated endocytosis. Tumor endothelial marker 8 (TEM8; antxr1) and capillary morphogenesis gene 2 protein are the anthrax toxin receptors (CMG2; antxr2). Both proteins are similar; the extracellular part of each contains an approximately 190 amino acid long von Willebrand factor A (vWA) domain that is related to integrin
-2 inserted (I) domains. Within this domain is a metal ion dependent adhesion site (MIDAS), which coordinates a divalent metal ion (Ca++, Mg++, or Mn++). Ample evidence has demonstrated that the vWA domain and its MIDAS motif are required for PA binding. However, fine structure-function mapping of this region has remained scant. It is not clear which specific residues within this region are crucial for receptor interaction with PA, or the receptors' function in any other manner for toxin entry.

The focus of this work was to elucidate the role of certain acidic, basic, and aromatic vWA domain residues in CMG2 functionality as an anthrax toxin receptor. The approach was alanine scanning mutagenesis. The mutants were derived using plasmid-encoded 488-residue isoform of CMG2, a strong PA receptor. Mutants were confirmed by sequencing and expressed in JCR65 for functional analysis. JCR65 is a hamster cell line that does not bind PA, and is therefore considered receptor-negative. The first functional assays aimed to assess each mutant receptor's capacity to support intoxication by anthrax toxin. Replacement of glutamate 71 with alanine abolished receptor function, but its replacement with aspartate had little or no effect. Surprisingly, lysine at position 71 made the receptor nearly 10-fold more efficient. D152A mutation conferred a high degree of resistance to the toxin, D152E has no effect, but D152K mutation abolished function, suggesting an acidic residue at this position is required. R72A mutation was dramatically disruptive, but the effect of R99A mutation was very small. The most disruptive aromatic residue mutations were F73A and Y158A; both rendered the receptor incapable of supporting intoxication. The precise mechanisms by which these mutations disrupt receptor function are not clear. However, all mutant receptors appear to bind PA either normally, or at a minimum show significant binding, suggesting the intoxication steps disrupted are after PA binding.