Crystallographic studies of Pyrenophora tritici-repentis ToxA.

摘要

Tan spot of wheat is an economically significant disease caused by the fungal pathogen, Pyrenophora tritici-repentis. Certain races of the fungus secrete Ptr ToxA (ToxA), a 13.2 kDa proteinaceous host-selective toxin that is responsible and sufficient to cause disease in susceptible wheat varieties. Disease symptoms develop only when the ToxA gene in the fungus and a single gene in the wheat host are expressed. The understanding of this gene-for-gene interaction could be instrumental towards control of the disease and is also being developed as a model system for understanding host-pathogen interactions. Here, this effort is given a solid structural foundation through crystallographic analysis of the ToxA structure.; The ToxA structure was solved at 1.65 A resolution using the anomalous signal from inherently present sulfur atoms. The monomeric toxin adopts a beta-sandwich fold of two anti-parallel beta-sheets composed of four strands each. The mapping of existing mutation data onto the structure reveals that a sequence of Arg-Gly-Asp (RGD) and surrounding residues required for activity are present on a solvent-exposed loop thereby making them potential candidates for recognition events that are required for ToxA activity. Unexpectedly, after a simple circular permutation, the ToxA structure is topologically identical to the classic mammalian RGD-containing fibronectin type III (FnIII) domain, and furthermore the RGD residues are topologically equivalent. These results support the hypothesis that ToxA, like FnIII, interacts with an integrin-like receptor on the host plant cell surface.; There has been a renewed interest in the method of using the anomalous signal from sulfur atoms to solve protein structures. As a spin-off of the structure solution work, the data were systematically analyzed to study the effects of crystal decay, resolution and data redundancy on the ability to locate the sulfur positions and subsequent phasing of the protein. The analyses show that the choices made about data redundancy and resolution limits may be crucial for the structure determination and that anomalous correlation coefficients are helpful indicators in making these choices.