In an elegant piece of deadly biology, the anthrax-causing bacterium, Bacillus anthracis, produces three proteins that can kill its host. One of these proteins, called protective antigen (PA), is a little like a Trojan horse in that it smuggles the other two proteins — lethal factor (LF) and oedemal factor (OF) — into the host cell. Once in the cell, LF and OF set in motion a train of potentially fatal events.

Two papers to be published in 8 November Nature (and currently online ahead of publication), reveal new details of the biology of the anthrax toxin and thereby provide academics and industrialists alike with information that will help them devise strategies to defeat anthrax.

In the first paper Kenneth Bradley and colleagues at the University of Wisconsin-Madison report the identification and isolation of the genetic code for anthrax toxin receptor (ATR) on the surface of eukaryotic cells that binds PA. They then sequenced the ATR gene and looked for clues to its specific function by comparison with similar sequences from other species.

Excitingly, the results of this study provide researchers with new avenues to pursue in search of novel therapeutic agents to combat anthrax. "Soluble ATR vbreceptor can be used like a sponge to mop up free toxin (PA), and this receptor can now be used as a tool to discover drugs that block toxin-receptor interactions," says John Young a co-author from the University of Wisconsin, Madison.

ATRs are common to a number of species, so Bradley and colleagues were able to work with a standard Chinese hamster ovary (CHO) cell line. Though the CHO cell surface is known to have a high concentration of ATRs, the receptors genetic code was unknown. In addition it was unclear how that code translated into the physical and chemical aspects of the receptors that enabled PA to bind to and infect the host cell.

To find answers to this mystery, they disrupted the CHO genetic code by inducing mutations and deletions in the code in a random manner. They then identified those resulting cell lines that no longer exhibited behavior typical of cells with ATRs. From among these, they selected one of the ATR^-/- cell for further analysis. They mixed it with fragments fragments of human DNA until they found a combination that restored the behavior typical of cells with ATRs.

They then sequenced the coding region of DNA that had succeeded in restoring ATR function and compared the sequence with known DNA sequences from a number of species, searching for patterns that would reveal the structural and functional properties of the protein.

The next step will be to work with ATRs "to follow precisely the mechanism of toxin binding and uptake," said Young.

In the second study Andrew Pannifer and colleagues at the University of Leicester, describe details of the crystal structure of LF bound to a portion of a protein, MAPKK-2, that is a member of the cell signaling family of proteins (mitogen-activated protein kinase-kinase – MAPKK).

These proteins play an important role in eukaryotic cell proliferation, differentiation, movement and death. LF can cleave members of the MAPKK family, thus disrupting cell signaling. Consequently, insight into the structural detail of LF and of LF bound to a member of this protein family is important for drug development.

Pannifer et al. observed that LF has four distinct regions (domains). Domain I, which sits on top of the other three interconnected domains, binds PA. Domain IV houses the enzymes destructive catalytic activity.

Domains II, III, and IV together form a long deep groove that holds the one terminal of MAPKK-2. It is the structural and sequence properties of this pocket that are the reason for the high specificity of the interaction between LF and MAPKK family members.

Robert Liddington, co-author and Director of the Cell Adhesion-Extracellular Matrix Biology Program at the Burnham Institute, describes as rather unexpected "the way in which domains II and III cooperated with domain IV to create this extended binding pocket for the target molecule."

Another unexpected finding was "that domain I was a copy of the catalytic domain [IV]" and had apparently evolved from an enzyme to a binding module. Though the two domains share no detectable similarity in amino acid sequence, they have similarities at a structural level.

Furthermore, domain III appears to have arisen from a structural element of domain II. The authors conclude that this protein complex evolved through a process of gene duplication, mutation, and fusion to form an enzyme with unusually high specificity.

The groove formed by domains II, III, and IV is 40Å in length and contiguous with the active site of LF. The team introduced peptides corresponding to the N-terminal 16 residues of MAPKK-2 into LF crystals to generate what they describe as "the first structural example of a protease in complex with its uncleaved natural substrate."

Drug design strategies based on these findings could focus on compounds that bind in the groove, targeting either the specificity region or the catalytic site, or on inhibitors of the interaction between LF and PA.

For an in depth description of the fundamental biology and basis of toxicity of five of the organisms most likely to be used as bioweapons, click through to:

Anatomy and actions of microscopic agents of terror