There are several strains of fungi that have 'killing' phenotypes (e.g. yeast and Ustilago maydis. In other words, when these strains of fungi are placed with other similar strains of fungi, they will take over the culture by killing the competing fungi. In the case of the P4 strain of Ustilago maydis, this killing is due to the expression of a small (105 amino acid) protein called KP4. This protein is toxic to many other strains of Ustilago, but not the host (P4) strain. This virus is quite unusual compared to viruses that infect other types of hosts. This virus is not spread by leaving the cell to infect another cell. Instead the virus only spreads through cytoplasmic exchange (e.g. cell mating). Therefore, the virus depends on the host survival for its own survivial. In a rare example of symbosis between virus and host, this virus (UMV4) makes the toxin, KP4, to kill competing strains of fungi. This insures host, and therefore its own, survival.
We have now determined the structure of KP4 to 1.5 angstrom resolution (see example density below). The toxin forms a dimer in the asymmetric unit of the P21 crystal cell. However, in solution, the toxin seems to be monomeric. KP4 belongs to the alpha-beta sandwich family of proteins with a double split beta-alpha-beta motif. There are two beta-alpha-beta cross-overs in this protein. While this is very unusual, the function of such cross-overs in this protein is not clear. The major secondary structure elements consist of a 5-stranded, anti-parallel beta sheet with two anti-parallel helixes lying at about 45 degrees to these strands. About 80% the members of this family function by binding to other proteins.
	Other fungal toxins have been speculated to act by forming ion channels in the target cell. However, KP4 has 5 disulfide bonds that firmly prevent the exposure of the amphipathic helices from being exposed (green balls in diagram). This remarkably stable protein is resistant to solvents, high temperatures, and proteases. It is therefore unlikely that this protein unfolds like the bacterial toxin, colicin, to act on the target membranes.
Scorpion Toxin	KP4 Toxin
The amino acid sequence and atomic structure of KP4 did not align well with any other protein in the databases. Therefore, we had to examine other properties of KP4 that might suggest a mechanism of killing. From the charge distribution, the existance of a protrusion off the beta-sheet, and a C-terminal disulfide, we suspected that KP4 might act, like scorpion toxin, by binding to ion channels. This led to killing assays and patch-clamp studies that demonstrate that KP4 blocks both fungal and mammalian calcium channels. Unpublished mutagenesis studies have now shown that charges about this loop are indeed involved in killing. Further work is underway to characterize the types of channels affected to further define the active site of the toxin.

[Return to top]