The rules of engagement

Plants are exposed to vast numbers of potential pathogens and symbionts throughout their life span and therefore must maintain mechanisms for identifying and appropriately responding to such microbes. Likewise, the microbial communities that interact with plants have developed several mechanisms to evade detection and/or suppress defense responses. The question of how these complex interactions are perceived and modulated by both the plant and the microbe remains central to this field. The XIII International Congress on Molecular Plant–Microbe Interactions assembled an array of experts whose interests range from host recognition and signal transduction to evolution of virulence to molecular dialogues of symbiotic interactions. At the expense of excluding a number of exciting topics, particularly in the area of symbiosis, this article focuses on several advances in our understanding of plant–pathogen interactions and how these are shifting the paradigms of disease resistance in plants. Pathogen recognition often leads to localized cell death (termed the hypersensitive response, HR), systemic resistance and inhibition of pathogen growth. Such recognition is most often triggered by ‘effector proteins’, which are secreted by the pathogen and translocated into the host cell where they contribute to virulence in susceptible hosts (Chisholm et al., 2005; Jones & Dangl, 2006). In resistant hosts, the HR is activated by a resistance (R) protein that detects the presence of the effector protein, either directly or indirectly. Using a novel trans-kingdom effector delivery system, two research groups led by Jonathan Jones (The Sainsbury Laboratory, UK) and Brian Staskawicz (University of California, CA, USA) have shown that R genes specific to oomycete effector proteins can inhibit the growth of bacterial pathogens when these bacteria are engineered to deliver oomycete effectors. To create their trans-kingdom delivery system, both laboratories spliced the amino-terminal portions, including the secretion and translocation sequence, of two different bacterial effectors (AvrRpt2 and AvrRps4, respectively) onto effectors from the oomycete Hyaloperonospora parasitica that lacked their own secretion and translocation signals (Mudgett & Staskawicz, 1999). These chimeric effectors were then delivered from Pseudomonas syringae into host plant cells via the P. syringae type III secretion system. Upon entry into the plant cell the portion comprising the bacterial effector is processed, releasing the oomycete effector. This vector system was dubbed the ‘effector detector vector’ by Jonathan Jones. Both groups used the ‘effector detector vector’ system to deliver the H. parasitica effector ATR13 into Arabidopsis carrying the matching R gene RPP13. Strikingly, RPP13, which normally confers resistance to H. parasitica, conferred resistance to P. syringae. This resistance was allele specific, whereby ATR13 from the H. parasitica isolate Emco5 (normally recognized by Colombia RPP13) was recognized, triggering an HR, whereas ATR13 from H. parasitica isolate Emoy2 (normally not recognized by Colombia RPP13) remained virulent. In addition, ATR13 conferred a growth advantage to P. syringae on plants that lacked RPP13 (Jonathan Jones, The Sainsbury Laboratory) and could activate viral resistance when delivered by Turnip Crinkle Virus (Brian Staskawicz, University of California). Therefore, the oomycete effector ATR13 could act as both a bacterial virulence and avirulence factor as well as a viral avirulence factor. These observations force us to consider how a single protein can function as a recognition factor for oomycete, bacterial and viral pathogens, and raise questions regarding the biological mechanisms that plants use to combat these diverse pathogens. Are the pathways identically activated by RPP13 regardless of the pathogen delivering ATR13, or does the plant use separate mechanisms to sense pathogen identity and detect effector presence, thus modulating the defense response? If the former, then is there a single defense mechanism that is effective against these diverse pathogens, or does RPP13 activate a large suite of defenses, some of which are effective against viruses, others against bacteria and yet others against oomycetes? In any case, it is clear that our understanding of pathogen recognition and defense response is far from complete. The effector delivery system defined here, in combination with the identification of many new putative oomycete effector proteins, will provide a novel means to dissect recognition pathways in plants. It is already apparent that multiple effectors can be monitored by the same R protein through indirect recognition (a.k.a. the guard model; Mackey et al., 2002). Further complexities were addressed by Jeff Dangl (UNC, NC, USA) as he posed the following questions: ‘can pathogen effectors target multiple plant proteins?’ and ‘can a single effector be simultaneously monitored by multiple R proteins?’. Several lines of evidence suggest that pathogen effectors target multiple host proteins. The AvrRpt2 and AvrRpm1 effectors from P. syringae are recognized by R proteins because of their modification of another host protein, RIN4 (Bisgrove et al., 1994; Mackey et al., 2002; Mackey et al., 2003). AvrRpm1 induces the phosphorylation of RIN4 and this modification appears to be recognized by the plant protein, RPM1, whereas AvrRpt2 cleaves RIN4, resulting in recognition by RPS2. However, in the absence of RIN4, both AvrRpm1 and AvrRpt2 exhibit virulence activity, suggesting that they target other host proteins to the advantage of the pathogen (Ritter & Dangl, 1995; Guttman & Greenberg, 2001). Furthermore, the AvrRpt2 cleavage sequence found in RIN4 is also present in several other Arabidopsis proteins, suggesting that these proteins may also be cleaved by AvrRpt2. Taken together, these data suggest that pathogen effectors target multiple plant proteins and that this contributes to both their virulence and avirulence activities. The second question, whether multiple unrelated plant proteins can recognize a single effector protein, was also considered during the meeting. Previous studies have shown that the P. syringae protein, AvrB, is recognized in the plant by RPM1 (Bisgrove et al., 1994). Like AvrRpm1, AvrB induces phosphorylation of RIN4 (Mackey et al., 2002). Interestingly, in the absence of RPM1, P. syringae expressing AvrB causes a weak chlorosis on Arabidopsis ecotype Mt-0, whereas the chlorosis is absent in ecotype Col-0. This suggested that the Mt-0 ecotype may possess a second protein involved in recognizing AvrB. Map-based cloning efforts have now identified this protein as TAO1. Plants with a functional copy of TAO1 can recognize AvrB, resulting in a moderate decrease in pathogen growth. Interestingly, TAO1 specifically recognizes AvrB, but not AvrRPM1, setting it apart from RPM1, which recognizes both proteins. Furthermore, TAO1 recognition of AvrB is independent of RIN4. Recent work has shown that previously identified residues, important for the interaction of AvrB and RIN4, appear to differ, at least in part, from the AvrB residues required for TAO1 recognition (Desveaux et al., 2007). These data suggest that Arabidopsis has evolved at least two independent mechanisms for detecting AvrB. Pathogen-associated molecular pattern (PAMP) receptors are important for recognizing pathogens and establishing basal resistance. The gains made in this field over the last few years were obvious by the sheer number of PAMP-related talks and posters at this 2007 congress, including the discovery of an Arabidopsis chitin receptor that is a member of the LysM receptor-like kinase (RLK) family (Jinrong Wan, University of Missouri-Columbia, MO, USA; Volker Lipka, The Sainsbury Laboratory). This finding is particularly interesting considering that the LysM RLK family in legumes is involved in recognizing rhizobial NOD factors, which have an acylated chitin backbone. Wan reported how loss-of-function mutations in the LysM-RLK1 gene of Arabidopsis blocked the induction of chitin-responsive genes and resulted in increased susceptibility to Erisyphae cichoracearum and Alternaria brassicicola. In contrast, Lipka reported how gain-of-function mutations in the LysM RLK, CIL1, resulted in aberrant interactions with Colletotrichum lagenarium and runaway cell death in response to powdery mildew challenge and compromised resistance to A. brassicicola. Perhaps the hottest PAMP-related topic was the identification of BRI1-associated kinase1 (BAK1) by three independent research groups as a critical component for several PAMP-mediated signaling pathways. BAK1 (also known as SERC3), a member of the somatic embryogenesis receptor kinase (SERK) family, was originally identified for its role in brassinosteroid signaling (Li et al., 2002; Nam & Li, 2002). Talks by Delphine Chinchilla (University of Basel, Switzerland), Silke Robatzek (Max Plank Institute for Plant Breeding Research, Germany), Thorsten Nürnberger (University of Tübingen, Germany) and John Rathjen (The Sainsbury Laboratory) described the role of BAK1 in PAMP-mediated defense. Part of this work was published shortly before the 2007 congress (Chinchilla et al., 2007; Kemmerling et al., 2007). The role of BAK1 in disease resistance was initially examined by Thorsten Nürnberger (University of Tübingen) and Thomas Boller (University of Basel) who observed its up-regulation in response to the functional flagellin- and elongation factor Tu-derived peptides, flg22 and elf18. Independently, John Rathjen also identified BAK1 in a screen for mutants that lost reactive oxygen species (ROS) production after flagellin treatment. Delphine Chinchilla and John Rathjen both reported how plants with null alleles of bak1 lost both early and late responses to flagellin treatment, as evidenced by loss of ROS burst and flg22-induced growth inhibition, respectively, whereas only early responses to the elongation factor Tu were lost. These phenotypes appear specific to bak1 because mutations in other SERC family members did not result in phenotypes similar to those of bak1. Interestingly, both Nürnberger and Rathjen showed that although bak1 plants had enhanced disease phenotypes, growth of P. syringae pv. tomato DC3000 (Pst DC3000) was not dramatically altered in these plants. Infection of bak1 plants with necrotrophic fungi such as Botrytis cinerea or A. brassicicola (typically weakly pathogenic on wild-type Arabidopsis plants) resulted in an increase in disease phenotypes, whereas bak1 plants showed increased resistance to the biotrophic pathogen, H. parasitica. BAK1 interacts with the flagellin receptor, FLS2, very weakly (Delphine Chinchilla, University of Basel) or not at all (John Rathjen, The Sainsbury Laboratory) before flg22 exposure; however, addition of flg22 (but not elf18) induced BAK1 and FLS2 interaction as early as 2 min after flg22 addition. Silke Robatzek (Max Plank Institute for Plant Breeding Research) presented data demonstrating that flg22-induced relocalization of FLS2 from the plasma membrane to endosomes is lost in bak1 mutants. The data presented support the following model; in the absence of pathogen, FLS2 and BAK1 are not complexed in the cell. Upon binding of flagellin to FLS2, BAK1 rapidly associates with FLS2, creating an active signaling complex. This association leads to activation of downstream signaling components, including activation of MAP kinases and receptor internalization. It is important to note that addition of brassinolide or mutation of the brassinosteroid biosynthetic genes does not affect flg22 binding or any of the documented pathogen responses, indicating that the role of BAK1 in disease resistance and brassinolide perception are separable (Delphine Chinchilla, University of Basel; Thorsten Nürnberger, University of Tübingen; John Rathjen, The Sainsbury Laboratory). John Rathjen (The Sainsbury Laboratory) reported that bak1 mutants exhibit a partial loss of responsiveness to elf18, csp22 and INF1, but not to chitin, suggesting that BAK1 plays a role in some, but not all, PAMP response pathways. Thorsten Nürnberger (University of Tübingen) also highlighted how BAK1 interacts with BIP89, and a bip89 mutant exhibits a similar phenotype to bak1, indicating that BIP89 may be another member of PAMP signal transduction pathways. Significantly, it appears that flagellin and elongation factor Tu-responsive pathways converge further downstream than previously anticipated, as several mutations have been found that affect elf18, but not flg22, perception and vice versa (Paul Schulze-Lefert, Max Plank Institute for Plant Breeding Research; Cyril Zipfel, The Sainsbury Laboratory). The majority of plant R genes encode proteins containing a nucleotide-binding site (NBS) and leucine-rich repeats (LRR). NBS-LRR proteins have long been thought to reside in the plant cytoplasm either as soluble proteins or as peripheral membrane proteins. This hypothesis is consistent with the cytoplasmic delivery of many pathogen effector proteins. Earlier this year it was demonstrated that two NBS-LRR proteins, MLA and N, localize, at least in part, to the nucleus and that this localization is necessary for activating resistance (Burch-Smith et al., 2007; Shen et al., 2007). The N protein recognizes infection with tobacco mosaic virus (TMV) by detecting a portion of the TMV replicase protein, termed p50 (Erickson et al., 1999). Exclusion of N from the nucleus eliminates N-mediated disease resistance; however, p50 nuclear localization is not required for resistance. To elucidate in greater detail the mechanism of N activation, Savithramma Dinesh-Kumar's research group (Yale University, CT, USA) performed a yeast two-hybrid experiment to identify proteins that interact with N. N-interacting protein 1 (NIP1) was identified in this screen. Curiously, NIP1 has a chloroplast targeting sequence and was localized to the plant chloroplast, bringing into question the biological relevance of the NIP1–N interaction. However, when p50 was co-expressed with NIP1, NIP1 was found in the cytoplasm and nucleus. Furthermore, NIP1 was found to be critical for viral resistance, as silencing of NIP1 resulted in reduced viral resistance. Yeast two-hybrid, bimolecular fluorescence and in vivo co-immunoprecipitation experiments showed that NIP1 and p50 interact, but that NIP1 only interacts with N in plant cells in the presence of p50. These data suggest that p50-mediated interaction between N and NIP1 may play a role in the activation of signaling. N was also shown, by fluorescence resonance energy transfer (FRET) analysis, to interact with SPL, a nuclear transcription factor required for N-mediated resistance. These exciting discoveries indicate the importance of understanding the cellular dynamics of NBS-LRR proteins if we are truly to understand their roles in disease resistance. It is clear from the large amount of exciting new data presented at The 2007 XIII International Congress on Molecular Plant–Microbe Interactions that the field is experiencing a growth spurt. Research on PAMP recognition, which once took a backseat to effector recognition, has moved to the forefront in generating exciting new concepts. The use of effectors from bacteria, fungi and oomycetes to dissect the basal defense pathways of plants is proving to be a powerful approach and is revealing new connections between PAMP-induced and R gene-induced defenses. Major questions remain, however, such as what are the physical and biochemical links between receptors (both PAMP receptors and R proteins) and downstream signaling components, what precisely are the downstream signaling components and, finally, what defense responses are key to preventing growth of a given pathogen? There is clearly plenty to look forward to when the XIV International Congress convenes in Quebec City, Canada, in July 2009.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.