Date Published: July 24, 2014
Publisher: Public Library of Science
Author(s): Johannes Mathieu, Simon Schwizer, Gregory B. Martin, Shengyang He.
The tomato—Pseudomonas syringae pv. tomato (Pst)—pathosystem is one of the best understood models for plant-pathogen interactions. Certain wild relatives of tomato express two closely related members of the same kinase family, Pto and Fen, which recognize the Pst virulence protein AvrPtoB and activate effector-triggered immunity (ETI). AvrPtoB, however, contains an E3 ubiquitin ligase domain in its carboxyl terminus which causes degradation of Fen and undermines its ability to activate ETI. In contrast, Pto evades AvrPtoB-mediated degradation and triggers ETI in response to the effector. It has been reported recently that Pto has higher kinase activity than Fen and that this difference allows Pto to inactivate the E3 ligase through phosphorylation of threonine-450 (T450) in AvrPtoB. Here we show that, in contrast to Fen which can only interact with a single domain proximal to the E3 ligase of AvrPtoB, Pto binds two distinct domains of the effector, the same site as Fen and another N-terminal domain. In the absence of E3 ligase activity Pto binds to either domain of AvrPtoB to activate ETI. However, the presence of an active E3 ligase domain causes ubiquitination of Pto that interacts with the domain proximal to the E3 ligase, identical to ubiquitination of Fen. Only when Pto binds its unique distal domain can it resist AvrPtoB-mediated degradation and activate ETI. We show that phosphorylation of T450 is not required for Pto-mediated resistance in vivo and that a kinase-inactive version of Pto is still capable of activating ETI in response to AvrPtoB. Our results demonstrate that the ability of Pto to interact with a second site distal to the E3 ligase domain in AvrPtoB, and not a higher kinase activity or T450 phosphorylation, allows Pto to evade ubiquitination and to confer immunity to Pst.
In the perpetual evolutionary arms race between hosts and pathogens, plants evolved two layers of inducible defense to protect themselves from infection . The first layer is now commonly referred to as pattern-triggered immunity (PTI). At its core are cell surface host receptors that detect common, highly conserved molecular features of microbes, referred to as microbe-associated molecular patterns. These receptors activate a relatively mild but effective defense response that includes the release of reactive oxygen species, changes in gene expression, and cell wall fortification , . While this response is sufficient to prevent many potentially pathogenic microbes from successfully establishing an infection, adapted pathogens have evolved large arsenals of ‘effector’ proteins –. These effectors are typically delivered into the plant cell cytoplasm, where they interfere with immune signaling to subvert PTI . The second plant immune response addresses this threat by monitoring for the presence of pathogen effectors and is consequently referred to as effector-triggered immunity (ETI) –. Because detection of an effector indicates the presence of a highly adapted and potentially devastating pathogen, the immune reaction is stronger and often includes localized cell death, referred to as the hypersensitive response –. Collectively, the ETI responses prevent the successful colonization by biotrophic or hemibiotrophic pathogens.
Testing variant forms of AvrPtoB for their interaction with Pto and Fen in a yeast two-hybrid system, revealed that Pto is capable of binding two distinct domains in the effector, both the N-terminal PID and the E3-ligase proximal FID, whereas Fen binds exclusively to the FID (Fig. S3). Furthermore, Pto that binds the FID is subjected to AvrPtoB-mediated degradation just as is Fen, showing that Pto is not intrinsically resistant to degradation. We also tested a construct in which the PID was fused directly to the E3 ligase domain and observed that Pto bound to this E3 ligase proximal PID was degraded. We initially interpreted these results as an indication that the position at which the kinases bind AvrPtoB relative to the E3 ligase domain might be the underlying cause for their apparent differential ability to phosphorylate T450 in AvrPtoB. We hypothesized that the unique ability of Pto to bind the more N-terminal PID allows Pto to efficiently phosphorylate AvrPtoB and thereby repress its E3 ligase activity. In contrast, Fen, which can only bind the more C-terminal FID, would not be in the optimal position and/or orientation to phosphorylate T450 and consequently be subject to ubiquitination and degradation.