Fig.3 | Hypothetical model for Tin2 function. In the upper part the model for infection of maize with wild type U. maydis is depicted. Tin2 effector (red stars) is secreted, taken up by maize cells and is binding to the ZmTTK1 maize kinase (brown ellipsoid), which stabilizes the kinase and renders the kinase active. Active kinase is proposed to positively affect anthocyanin biosynthesis (purple) via the R transcription factor. Arrows and their thickness depict the activity of the anthocyanin and lignin pathway, respectively. The lower part depicts the situation that develops after infection with the tin2 mutant strain. In the absence of Tin2, ZmTTK1 is degraded and consequently more precursor may be available for lignin biosynthesis, leading to a reinforcement of plant cell walls (brown) which may be limiting fungal spread and nutrition.
The Cmu1 effector: The secreted chorismate mutase effector Cmu1 was the first effector of U. maydis for which we had unambiguously shown that the protein is taken up by host cells after secretion from fungal hyphae. This effector promotes virulence by lowering salicylate levels in infected tissue. Salicylic acid (SA) is a key plant defence hormone which plays an important role in local and systemic defense responses against biotrophic pathogens like U. maydis. As virulence of cmu1 deletion strains was only weakly attenuated compared to the wild type, we considered the possibility that U. maydis may possess alternative means to lower SA levels. We identified in the U. maydis genome a putative cytoplasmic salicylate hydroxylase, shy1, as one such candidate. shy1 was transcriptionally induced during the biotrophic stages of development, and the trigger for shy1 transcriptional induction was SA, suggesting the possibility of a SA-sensing mechanism in U. maydis. In addition, Shy1 activity was needed for growth on plates with SA as sole carbon source. However, in seedling infection assays, shy1 did not contribute to virulence. These results show that U.?maydis can use SA as carbon source and through this the fungus could conceivably lower the amounts of this plant defense hormone after infection. The lack of a virulence phenotype of ?shy1 strains most likely indicates that additional pathways exist that eliminate SA, i.e. enzymes that modify or conjugate SA, and that the corresponding genes are not expressed in axenic culture where we have seen the shy1 mutant growth phenotype.
With Cmu1 being the first translocated effector of U. maydis identified, we decided to map the translocation motif in Cmu1 to begin to analyze the translocation mechanism. To this end we generated a series of deletion mutants, which were first tested for their ability to complement an aro7 mutant of Saccharomyces cerevisiae. Mutants which were able to complement and thus had chorismate mutase activity, were subsequently tested for complementation of an U. maydis cmu1 mutant. This allowed us to map a motif of 20 amino acids downstream of the signal peptide that was not required for catalytic activity but was crucial for cmu1 function in U. maydis. Complementation could be restored when these 20 amino acids were replaced by a cell penetrating peptide from HIV Tat-protein. Orthologous proteins to Cmu1 from other smut fungi display only about 30 % identity with U. maydis Cmu1, but were nevertheless able to complement the virulence defect of the U. maydis cmu1 mutant, suggesting that they must also be taken up by plant cells. The sequence of the 20 amino acid stretch of Cmu1 downstream of the signal peptide also aligned very poorly and revealed only one conserved amino acid residue (Asp40) and two positions where amino acids with strongly similar properties are found (Asp37 and Leu39). To determine whether conserved amino acids in the 20 amino acid motif are required for function, Asp37 and Asp40 in Cmu1 were both substituted with Ala. The respective double mutant was unable to complement the virulence defect of the cmu1 deletion strain, suggesting that this region may in fact constitute the host uptake motif. Furthermore, microscopic studies indicate host membrane association of N-terminally tagged mCherry-Cmu1 fusion proteins in infected maize tissue, and this association depends on the presumed translocation motif. We suggest that Cmu1 may bind a membrane-associated receptor. Our current efforts lie in the identification of this molecule and in determining the structure of the Cmu1 domain that is implicated in binding and uptake.