The large, diverse, and robust arsenal of Ralstonia solanacearum type III effectors and their in planta functions

Abstract The type III secretion system with its delivered type III effectors (T3Es) is one of the main virulence determinants of Ralstonia solanacearum, a worldwide devastating plant pathogenic bacterium affecting many crop species. The pan‐effectome of the R. solanacearum species complex has been exhaustively identified and is composed of more than 100 different T3Es. Among the reported strains, their content ranges from 45 to 76 T3Es. This considerably large and varied effectome could be considered one of the factors contributing to the wide host range of R. solanacearum. In order to understand how R. solanacearum uses its T3Es to subvert the host cellular processes, many functional studies have been conducted over the last three decades. It has been shown that R. solanacearum effectors, as those from other plant pathogens, can suppress plant defence mechanisms, modulate the host metabolism, or avoid bacterial recognition through a wide variety of molecular mechanisms. R. solanacearum T3Es can also be perceived by the plant and trigger immune responses. To date, the molecular mechanisms employed by R. solanacearum T3Es to modulate these host processes have been described for a growing number of T3Es, although they remain unknown for the majority of them. In this microreview, we summarize and discuss the current knowledge on the characterized R. solanacearum species complex T3Es.

determinant of RSSC bacteria is the type III secretion system (T3SS), a "molecular syringe" that allows the translocation of several type III effector proteins (T3Es) directly into the host cell (Coll and Valls, 2013). These T3Es, referred to as Ralstonia injected proteins (Rips), are able to subvert the defences and modify the metabolism of the host to promote virulence.

| THE R SSC T YPE III EFFEC TOME , A L ARG E AND VARIED AR S ENAL
Since the first RSSC T3E genes were cloned in the 1990s (Carney and Denny, 1990;Arlat et al., 1994;Guéneron et al., 2000), different approaches have been conducted to systematically identify at the genome scale the full T3E repertoire of several RSSC strains. Two main strategies were undertaken: (a) sequence-based approaches, searching for sequence homology with previously described effector genes and/or for the presence of certain 25-nucleotide cis elements in their promoters, the hrp II box or the plant-inducible promoter (PIP) box motifs (Salanoubat et al., 2002;Cunnac et al., 2004a;Gabriel et al., 2006;Peeters et al., 2013b;Sabbagh et al., 2019), and (b) regulationbased strategies, exploiting that T3E gene expression is controlled by HrpB, an AraC family member of transcriptional regulators (Genin et al., 1992;Cunnac et al., 2004a). Regulation-based strategies include gene expression studies (Cunnac et al., 2004b;Occhialini et al., 2005) and genetic screens using random transposon-insertion mutagenesis (Mukaihara et al., 2004). Verification of the T3SS-dependency of the secretion or translocation is typically required to confirm the bona fide T3E status of in silico predicted or candidate T3Es . Most translocation analyses exploit the adenylate cyclase (Cya) reporter system (Cunnac et al., 2004b;Mukaihara and Tamura, 2009;Mukaihara et al., 2010). T3SS-dependent secretion analyses compare the secreted proteins, detected by immunoblotting or mass spectrometry, of wild-type compared to hrp mutant strains (Tamura et al., 2005;Solé et al., 2012;Lonjon et al., 2016;Sabbagh et al., 2019).
A recent genomic study on 140 RSSC strains identified the pan-effectome of the species complex, consisting of 102 T3E and 16 hypothetical T3E genes (Sabbagh et al., 2019). RSSC strains carry on average 64 T3E genes (minimum 45 in R. syzygii subsp. syzygii strain R24 and maximum 76 in R. pseudosolanacearum strain Rs-10-244). This contrasts with other plant pathogenic bacteria such as Pseudomonas syringae and Xanthomonas campestris, with an average of 31 (min. 3, max. 53) and 23 (min. 12, max. 28) T3E genes, respectively Dillon et al., 2019). The existence of several paralog families, such as the RipG (former GALA), RipS (SKWP), RipA (AWR), RipH (HLK), or RipP (PopP) families, can be considered as a remarkable feature of the RSSC. Not a single known RSSC strain does not carry multiple copies of these paralog T3E families. This contributes to the large size of the RSSC pan-effectome. The T3E repertoires of different RSSC strains are quite diverse, with only 16 core T3Es (i.e., T3Es present in at least 95% of sequenced strains), which represents 13.6% of the RSSC pan-effectome (Sabbagh et al., 2019).
This core-effectome is larger than in P. syringae (four core T3Es, 5.7% of its pan-effectome) or X. campestris (three core T3Es, 8.6% of its pan-effectome) Dillon et al., 2019). Several studies have tried to connect the T3E diversity to the host specificity of RSSC strains (Ailloud et al., 2015;Cho et al., 2019;Sabbagh et al., 2019). Although some host specificity determinants could be identified, the power of such studies has usually been largely limited by the lack of exhaustive strain host range empirical data.

| MANY T3E S , BUT FOR WHAT PURP OS E?
As model root and vascular plant pathogens, RSSC bacteria are among the pathogens with a larger number of functionally characterized T3Es.
Some effectome-scale experiments have tried to shed light on the function of RSSC T3Es through systematic determination of their ability to induce a hypersensitive response (HR; Wroblewski et al., 2009), inhibit plant defences (Nakano and Mukaihara, 2019a), or identify their plant targets . However, most of our current knowledge on effector function comes from smaller-scale experiments in which often one or a few T3Es are studied. To date, we have counted more than 50 different RSSC T3Es that have been characterized with varying degrees of detail ( Figure 1 and Table 1). One of the main factors complicating this task is the observed genetic redundancy among different RSSC T3Es (Angot et al., 2006;Solé et al., 2012;Chen et al., 2014). This redundancy is likely to ensure a more robust virulence strategy for the bacteria (Ghosh and O'Connor, 2017), although it makes the functional dissection of single effectors more complicated, particularly for the paralog families. Nevertheless, some members of these families can still have specific and nonredundant functions (Angot et al., 2006;Turner et al., 2009;Wang et al., 2016).
Similar to other pathogens, RSSC T3Es collectively contribute to the pathogen fitness in the plant through different and not always well-characterized mechanisms (Toruño et al., 2016). These include the interference with the plant basal defence responses, alteration of the plant metabolism, and avoidance of the specific recognition of other T3Es. However, some RSSC T3Es can also be recognized by specific plant genotypes and induce strong immune responses.

| Interference with plant basal immunity
The subversion of basal defences is one of the most studied functions of pathogen effectors. Several RSSC T3Es are known to interfere with different host cellular processes involved in these basal defence responses. RipP2 (former PopP2) relies on its acetyltransferase activity to acetylate the WRKY domain of the plant homonymous transcription factors, which prevents their association with DNA and subsequent expression of defence-related genes (Le Roux et al., 2015). RipAY is selectively activated by eukaryotic thioredoxins to degrade the host glutathione, which plays an important role in plant immunity (Fujiwara et al., 2016(Fujiwara et al., , 2020Mukaihara et al., 2016;Sang et al., 2018). RipAR and RipAW rely on their E3 ubiquitin ligase activity to inhibit plant defence responses (Nakano et al., 2017).  I G U R E 1 Ralstonia solanacearum species complex (RSSC) bacteria deploy an arsenal of type III effectors (T3Es) to alter the plant metabolism and interfere with plant immune responses. During the infection process, conserved bacterial molecules are recognized by plant pattern recognition receptors (PRRs) at the surface of the host cell. They activate basal defence responses to prevent pathogen proliferation. However, RSSC bacteria translocate T3Es into the plant cell to subvert the plant defences and accommodate the bacterial needs. T3Es act on different host pathways. RipAY and RipN alter the glutathione level and NADH/NAD + ratio, respectively. RipAY, RipR, RipAL, RipG1, and RipG3 target the hormone synthesis and signalling level. Different RipG family members, RipAR and RipAW, interfere with ubiquitination processes. The metabolism is also manipulated by RSSC T3Es. RipA5, RipTPS, and RipTAL are able to modulate certain metabolic pathways. RipTAL binds to the plant DNA, activating the expression of shorter and more efficiently translated transcripts of arginine decarboxylase (ADC) genes, key enzymes in the biosynthesis of polyamines. This boost in the polyamine level could prevent the proliferation of Ralstonia niche competitors. RipP2 relies on its acetyltransferase activity to acetylate defensive WRKY transcription factors, inhibiting their DNAbinding activities and preventing subsequent expression of defence-related genes. The nuclear T3E RipAB inhibits the expression of Ca 2+related defence genes. In addition to these functionally characterized RSSC T3Es, other effectors involved in dampening of basal defence through as yet unknown mechanisms have been identified: RipAR, RipAW, RipG family, RipAB, RipA5, RipAD, RipAF1, RipD, RipE1, RipI, RipQ, RipAC, RipAP, RipAU, RipH1, RipM, RipS1, RipAN, and RipB. RSSC T3Es can also be perceived in planta by intracellular immune-Nodlike receptors (NLRs), leading to the activation of specific defence mechanisms, often associated with an HR. RipE1, RipAA, RipP1, RipX, RipP2, RipAT, RipAV, RipA1-A5, RipTPS, RipAX2, RipAB, RipB, RipBN, and RipI also induce HR on several hosts. Some T3Es can modulate the activity of others and prevent their recognition by the plant surveillance system. Indeed, peroxisome-localized RipAK suppresses effectortriggered HR by inhibiting host catalase activities (CATs). RipAY and RipAC inhibit RipE1-mediated HR TA B L E 1 List of functionally characterized Ralstonia solanacearum species complex type III effectors d Indicated only when the ability to inhibit any classical PAMP-triggered immunity (PTI) response has been proven. In parentheses when only some members of a paralog T3E family members inhibit PTI responses.

TA B L E 1 (Continued)
(former PopB) down-regulates the calcium signalling pathway and inhibits the plant basal defences (Zheng et al., 2019). Finally, RipN contains a Nudix hydrolase domain required to alter the NADH/ NAD + ratio in planta and to inhibit the plant defence responses (Sun et al., 2019).

| Targeting plant metabolism
Plant pathogenic bacterial T3Es can also interfere with different host metabolic processes to promote the bacterial survival, release nutrients, and facilitate the infection (Macho, 2016 (de Lange et al., 2013). RipTAL induces the expression of plant genes involved in the synthesis of polyamines, evading their native translational regulation mechanisms . It is hypothesized that this RipTAL-induced boost of the plant polyamine levels prevents the proliferation of possible Ralstonia competitors . RipA5 acts as an inhibitor of the conserved target of rapamycin (TOR) pathway in yeast and plant cells . As a key regulator of the switch between growth and stress responses (Dobrenel et al., 2016), RipA5-mediated inhibition of the plant TOR pathway leads to reduced nitrate reductase activity . Lastly, RipTPS possesses trehalose-6-phosphate synthase activity in yeast (Poueymiro et al., 2014). As trehalose-6-phosphate is a key regulatory molecule in plant metabolism (Baena-González and Lunn, 2020), RipTPS could potentially interfere with this regulation but so far this activity has not been shown in planta.

| Contribution to virulence through (as of yet) unknown mechanisms
In addition to the beforementioned RSSC T3Es for which functional roles could be assigned, other T3E genes have been also identified as contributors to bacterial virulence on different hosts. These additional T3E genes have been identified through pathogenicity or competitive index assays with single or multiple gene mutants.
These tests allow us to pinpoint the involvement in virulence but do not provide further information about the underlying molecular mechanisms. This is the case for RipA2 and RipD, which contribute to virulence in tomato (Cunnac et al., 2004b), or RipAA and RipG7, important in the early and late stages of infection of the model legume species Medicago truncatula, respectively (Turner et al., 2009;Wang et al., 2016). RipAC, RipAF1, RipAK, RipAV, RipAY, RipD, RipP2, RipR, RipS4, RipY, and RipTAL contribute to bacterial fitness in eggplant (Macho et al., 2010). For RipD and RipP2, this contribution to fitness was also demonstrated in tomato and bean, and in the case of RipAA, exclusively in tomato (Macho et al., 2010). The RipA family members contribute collectively to virulence in both eggplant and tomato (Solé et al., 2012), and the RipH family members also contribute to virulence in tomato (Chen et al., 2014). RipAM, RipAN, and RipBH contribute significantly to virulence in potato (Zheng et al., 2019), and RipAC acts similarly in tomato .

| Effectors triggering plant immune responses
Through evolution, plants have evolved mechanisms to recognize specific RSSC T3Es and induce a strong defence response often associated with a hypersensitive response (HR) (Balint-Kurti, 2019).
RipAX2 (former Rip36) elicits immunity on wild and cultivated eggplants in a Zn-finger domain-dependent (Nahar et al., 2014) and independent (Morel et al., 2018) manner, respectively. RipAB triggers an HR in N. benthamiana but only when localized in the nucleus (Zheng et al., 2019). RipB induces chlorosis in different Nicotiana spp. in a Recognition of XopQ1 (Roq1)-dependent manner (Nakano and Mukaihara, 2019b). RipBN triggers resistance in tomato in a Pseudomonas tomato race 1 (Ptr1)-dependent manner (Mazo-Molina et al., 2019). RipE1 triggers immune responses mediated by both SA and JA in N. benthamiana and Arabidopsis . RipE1 also triggers an HR in N. tabacum and N. benthamiana in a Suppressor of G2 allele of skp1 (SGT1)-dependent manner for the latter (Jeon et al., 2020). Last, RipI triggers immune responses in tomato and cell death in yeast and N. benthamiana, the latter through interaction with the plant basic helix-loop-helix 93 (bHLH93) transcription factor (Deng et al., 2016;Zhuo et al., 2020).

| Effectors preventing other effectors to be recognized in planta
The recognition of RSSC T3Es and subsequent strong immune responses can also be counteracted through the action of other T3Es, sometimes referred as "meta-effectors" (Kubori et al., 2010). This could allow the bacteria to conserve effectors with potent virulence functions for which a given host has already developed specific recognition capabilities. This is the case for RipAY, which can inhibit the previously mentioned RipE1-triggered immunity . RipAY inhibits RipE1-mediated activation of the SA signalling pathway probably through degradation of the plant cellular glutathione (Mukaihara et al., 2016;Sang et al., 2018Sang et al., , 2020. It has also been proposed that RipAC suppresses RipE1-triggered immunity, inhibiting in this case SGT1mediated MAPK activation . RipAK is able to prevent Ralstonia-induced HR in N. tabacum by inhibiting plant catalase activity (Sun et al., 2017). Whether this HR is induced by RipAA, RipB, and/or RipP1, responsible for RSSC incompatibility in N. tabacum (Poueymiro et al., 2009;Nakano and Mukaihara, 2019b), is still unknown.  (Wicker et al., 2007), coffee plant (Lopes et al., 2015), fig tree (Jiang et al., 2016), African daisy (Weibel et al., 2016), and roses (Tjou- Tam (Angot et al., 2006;Remigi et al., 2011), RipR inhibits SA-mediated defence responses (Jacobs et al., 2013), RipAY degrades plant glutathione (Fujiwara et al., 2016(Fujiwara et al., , 2020Mukaihara et al., 2016;Sang et al., 2018), and RipAB downregulates the calcium signalling pathway (Zheng et al., 2019). These different processes, together with the unknown ones targeted by the other core T3Es, could represent the minimum plant processes that Ralstonia needs to modulate. This "basal arsenal" could be complemented with accessory T3Es that could have additive effects, targeting the same or different processes. However, this characterization might prove quite complex as these plant processes, and their modulation by Ralstonia T3Es, might vary substantially among different organs and host species. The diverse, and sometimes large, host range of RSSC strains and the functional diversity and redundancy of its effectome are therefore some of the causes of RSSC adaptability and aggressiveness, but also some of the major factors complicating its systematic and exhaustive study. A valuable tool that will open a wide variety of possibilities in the decipherment of RSSC T3E functions is the generation of a strain devoid of all its effectors, as has been performed on the P. syringae strain DC3000 (Cunnac et al., 2011). This should be completed soon on the RSSC strain OE1-1 (K. Onishi, Kochi

| CON CLUS I ON S AND PER S PEC TIVE S
University, Japan, personal communication). The fact that RSSC bac- ANR-11-IDEX-0002-02). None of the coauthors have a conflict of interest to declare. We would like to thank our colleague Stephane Genin for critically reading this work.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analysed in this study.