Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses

Report

Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses Highlights d

d

d

M. truncatula nodules can be infected by the pathogenic bacterium R. solanacearum M. truncatula nodules display weak defense reactions upon pathogen infection R. solanacearum encounters confinement in M. truncatula nodules

Authors Claire Benezech, Fathi Berrabah, Marie-Franc¸oise Jardinaud, ..., Pascal Ratet, Fabienne Vailleau, Benjamin Gourion

Correspondence [email protected]

In Brief Legume plants form root nodules, which host massive numbers of beneficial bacteria. Benezech et al. show that the symbiotic organs are vulnerable to infection by pathogen and that their defense ability is reduced. However, pathogens encounter confinement in the nodules, which prevents them spreading to non-symbiotic parts of the plant.

Benezech et al., 2020, Current Biology 30, 1–8 January 20, 2020 ª 2019 Elsevier Ltd. https://doi.org/10.1016/j.cub.2019.11.066

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

Current Biology

Report Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses Claire Benezech,1 Fathi Berrabah,2,3 Marie-Franc¸oise Jardinaud,1 Alexandre Le Scornet,1 Marine Milhes,4 Gaofei Jiang,1 Jeoffrey George,2,3 Pascal Ratet,2,3 Fabienne Vailleau,1 and Benjamin Gourion1,5,*

 de Toulouse, INRA, CNRS, 31326 Castanet-Tolosan, France Universite  Paris-Sud, Universite  Evry, Universite  Paris-Saclay, Baˆtiment 630, of Plant Sciences Paris Saclay IPS2, CNRS, INRA, Universite 91405 Orsay, France 3Institute of Plant Sciences Paris-Saclay IPS2, Paris Diderot, Sorbonne Paris-Cite , Baˆtiment 630, 91405 Orsay, France 4INRA, US1426, GeT-PlaGe, Genotoul, 31326 Castanet-Tolosan, France 5Lead Contact *Correspondence: [email protected] https://doi.org/10.1016/j.cub.2019.11.066 1LIPM,

2Institute

SUMMARY

Legumes have the capacity to develop root nodules hosting nitrogen-fixing bacteria, called rhizobia. For the plant, the benefit of the symbiosis is important in nitrogen-deprived conditions, but it requires hosting and feeding massive numbers of rhizobia. Recent studies suggest that innate immunity is reduced or suppressed within nodules [1–10]; this likely maintains viable rhizobial populations. To evaluate the potential consequences and risks associated with an altered immuni‘ty in the symbiotic organ, we developed a tripartite system with the model legume Medicago truncatula [11, 12], its nodulating symbiont of the genus Sinorhizobium (syn. Ensifer) [13, 14], and the pathogenic soil-borne bacterium Ralstonia solanacearum [15–18]. We show that nodules are frequent infection sites where pathogen multiplication is comparable to that in the root tips and independent of nodule ability to fix nitrogen. Transcriptomic analyses indicate that, despite the presence of the hosted rhizobia, nodules are able to develop weak defense reactions against pathogenic R. solanacearum. Nodule defense response displays specificity compared to that activated in roots. In agreement with nodule innate immunity, optimal R. solanacearum growth requires pathogen virulence factors. Finally, our data indicate that the high susceptibility of nodules is counterbalanced by the existence of a diffusion barrier preventing pathogen spreading from nodules to the rest of the plant. RESULTS AND DISCUSSION Legume Nodules Are Ralstonia solanacearum Infection Sites Root tips and lateral root emerging zones are described as the main entry points for the phytopathogenic bacterium

R. solanacearum strain GMI1000 in non-nodulated M. truncatula [19]. In order to determine whether legume nodules represent multiplication niches and/or can provide an access to the root system for R. solanacearum, we initiated a set of experiments using M. truncatula plants harboring mature nodules. We first evaluated the capacity of R. solanacearum to reach nodules after root tip inoculation without wounding. The phytopathogenic bacterium was detected in all analyzed symbiotic organs with populations rising to 105 to 2.106 pathogenic cells per nodule in most samples (19 out of 20) 21 days post-inoculation (dpi). Similar data were obtained after 7 and 14 dpi (Figure S1). To allow potential direct infection of the nodules by R. solanacearum, we then analyzed plants where the whole nodulated root system was inoculated with the pathogen by flooding. Upon such a treatment, nitrogenase activity was generally negatively impacted. However, for some plants with all nodules highly colonized by R. solanacearum, nitrogenase activity was found to be similar to plants non-inoculated by R. solanacearum (Figure 1A). Presence of the pathogenic bacterium in nodules, in root tips, in leaves, at zones of lateral root branching, and in root segments was confirmed using a R. solanacearum GMI1000 strain constitutively expressing the glucuronidase reporter gene (Figure 1B). Analysis of 158 plants indicated that nodules represent a frequent R. solanacearum-infected site. Indeed, the 936 analyzed nodules displayed identical infection frequencies (35%) as root tips (n = 1,685) that are considered as the preferential infection site in the classic bipartite pathosystem [18, 19]. However, in our culture system, plants displayed more root tips than nodules (Figures 1C and S1). Thus, the relative risk for plants of infection in a nodule was lower than in a root tip. Furthermore, R. solanacearum populations were found to be less dense in symbiotic organs (Figures 1D and S1). Various colonization patterns were observed for nodules, suggesting that the pathogen can enter in the symbiotic organs through multiple ways (Figure S2). For instance, continuous colonization of root segments and nodules suggested passage from the root to the nodule through the vasculature (Figure 1B). Among the 328 R. solanacearum-infected nodules, 39 were found on a root segment harboring colonized vasculature, Current Biology 30, 1–8, January 20, 2020 ª 2019 Elsevier Ltd. 1

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

among which, 28 and 11 were, respectively, totally colonized and colonized only in the nodule proximal zone. 183 infected nodules were found on root segments that did not harbor detectable glucuronidase activity, suggesting low or no vasculature colonization by the pathogen. Among these 183 nodules, 132 were colonized only at nodule apices and 23 in the proximal region and 28 were totally colonized (Figure S2). The high proportions of the three last categories suggest that nodules could be directly infected. Consistent with this, R. solanacearum cells were observed attached by one of their poles on the disorganized structure of the external nodule tissues (Figure 1E). Surprisingly, histological analysis indicated that this vascular pathogen was not restricted to the vasculature in nodules but infect all types of tissues (Figures 1 and S2). 14 dpi, the R. solanacearum cells essentially occupy the intercellular spaces of nodules. Later, the pathogenic bacteria colonize what appear to be cellular spaces in nodules, but we cannot currently differentiate whether these are living or dead plant cells (Figures 1 and S2). Remarkably, in agreement with the persistence of nitrogenase activity in R. solanacearum-colonized nodules (Figure 1A), fluorescence imaging indicated that pathogen and rhizobia can occupy the same nodule (Figure 1J). To evaluate nodule capacity to support R. solanacearum growth independently of any effect on the infection process, we directly inoculated the symbiotic organ by piercing. The results indicate that nodules support R. solanacearum colonization with efficiency similar to the reference organs (Figures 2A and S3). Furthermore, using natural diversity of Sinorhizobium as well as a S. medicae bacA mutant (impaired for nitrogen fixation), we observed that R. solanacearum growth in nodules is not limited by nitrogen fixation (Figures 2B and S3). Together, those data showed that M. truncatula nodules represent opportunities

Figure 1. R. solanacearum Impacts Nitrogen Fixation and Colonizes Nodules (A) 1 week after inoculation with R. solanacearum by flooding the nodulated root systems, nitrogen fixation activity is strongly reduced as compared to the control. However, high densities of pathogen in the nodules is not always correlated with a decreased nitrogenase activity, as all the nodules of a subset of plants able to fix nitrogen display high densities of R. solanacearum. Bars represent the medians. Stars illustrate the results of Mann and Whitney tests (a = 0.05). For acetylene reduction assay, three biological repetitions are represented by three colors, 72 plants analyzed, 36 inoculated with R. solanacearum, and 36 control plants. Dot line represents the detection threshold. For viability, two biological repetitions are represented by two colors and numbers of analyzed nodules are 23, 35, and 14 for control, fix
, and fix+ plants, respectively, N.D., not detected.

2 Current Biology 30, 1–8, January 20, 2020

(B) R. solanacearum colonizes all plant organs and is not restricted to vasculature in the nodules. Organs of nodulated M. truncatula A17 after inoculation of the all root system with R. solanacearum::uidA strain by flooding are shown. Scale bars: 1 mm for all pictures except for leaf: 5 mm. (C) Number of nodules (N), root tips (RT), and lateral root branching zones (LRBZ) per plant and their infection frequencies by R. solanacearum::uidA determined by visual inspection for blue staining (as illustrated in B). Data represent the mean values for 158 analyzed plants. Error bars represent the SEs. (D) R. solanacearum populations are lower in M. truncatula N than in RT 8 days after inoculation by immersion, Mann Whitney test indicates a p = 0.0076 (n = 21 for nodules; n = 22 for root tips). Dot line represents the detection threshold. (E) 3 days after inoculation, R. solanacearum is attached to poorly organized tissue of the nodule surface (scanning electron microscopy); scale bars: 500 mm for whole organ images; 20 mm for magnifications. (F–K) R. solanacearum colonizes inter- and intracellular nodule spaces, where it cohabits with S. medicae. (F–H) Sections of M. truncatula nodules infected by mCherry-tagged R. solanacearum. Green signal is autofluorescence that allows to distinguish nodule structure and cell walls. R. solanacearum intra- and intercellular bacteria are, respectively, visible in (F)–(H) and (J). Scale bars: (F) 1 mm and (G) and (H) 50 mm. (G) Magnification of the zone indicated in (F). (I and J) eGFP-tagged R. solanacearum was also observed intercellularly in nodule sections where mCherry-tagged S. medicae WSM419 are still visible 21 and 28 days post-inoculation with R. solanacearum (I and J, respectively). Scale bars: 50 mm. (K) Section of an M. truncatula nodule infected by R. solanacearum::uidA GMI1000 strain expressing glucuronidase; scale bar: 500 mm. See also Figures S1 and S2.

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

suppression of plant immunity [21], for nodule colonization. The results obtained with R. solanacearum hrcV T3SS mutant indicate that the T3SS does contribute to pathogenic bacterial colonization in nodules (Figures 2C and S3). This might be explained by the inability of the hrcV mutant to properly attach to nodules. Indeed, in contrast to the wild-type (WT), the hrcV mutant cannot attach in the typical polar fashion to host cells (Figure S3). Furthermore, following piercing inoculation procedure, the T3SS was also found to be required for optimal growth within nodules (Figure 2D). Together, these results indicate that the R. solanacearum T3SS contributes to R. solanacearum optimal growth in nodules.

Figure 2. Optimal Colonization of Nodules by R. solanacearum Requires the Type III Secretion System (T3SS), but Not Nitrogen Fixation (A) Growth of R. solanacearum in nodules, root tips, and root segments after inoculation by piercing. Dot lines represent detection thresholds: one bacterium per organ for 0, 12, and 24 h post-inoculation (hpi) and 20 bacteria per organ for 36 and 48 hpi. Bars represent the medians. For all time points, 9 nodules, 9 root tips, and 9 root segments were analyzed. Mann Whitney tests indicate no significant difference between organs. (B) R. solanacearum growth in nitrogen fixing (+, WT) and non-fixing (
, bacA) nodules (n = 77 and 63, respectively; bars represent the medians; samples with values below the detection threshold were not plotted but were considered for the median definition and for the Mann Whitney test that indicates no significant difference between fix+ and fix
nodules). (C and D) Importance of the T3SS for nodule colonization; dot lines represent detection thresholds, bars represent the medians, and stars illustrate the results of Mann and Whitney tests (a = 0.05). (C) Nodulated root systems were inoculated by immersion; at 14 dpi, 12 nodules inoculated with WT and 12 with the hrcV mutant were analyzed; at 21 dpi, 24 nodules inoculated with WT and 26 with the hrcV mutant were analyzed. (D) Nodules inoculated with R. solanacearum WT and the hrcV mutant strains by piercing; at 5 dpi, 81 nodules inoculated with WT and 79 with the hrcV mutant were analyzed; at 7 dpi, 65 nodules inoculated with WT and 59 with the hrcV mutant were analyzed. See also Figure S3.

for pathogen as they are as permissive as root tips for R. solanacearum infection and proliferation (Figures 1C and 2A) and this independently of nodule ability to fix nitrogen (Figures 2B and S3). R. solanacearum Optimal Colonization of Nodules Requires the Type III Secretion System Innate immunity is presumed to be suppressed, or partially suppressed, in nodules [6, 20], suggesting that virulence factors could be less important for pathogen growth in nodules as compared to other plant organs. We thus evaluated the role of the bacterial type III secretion system (T3SS), a key actor in the

Nodules and Roots Respond to R. solanacearum Infection by Activating Defense Genes The fact that R. solanacearum T3SS strongly contributed to the infection and development of R. solanacearum in symbiotic organs (Figure 2) suggests that innate immunity might not be totally switched off in nodules. In order to characterize the defenses of the symbiotic organ, RNA sequencing (RNA-seq) analysis was performed on nodules and roots of the same plants (Figure 3A). Nodulated plants were treated with water (as controls) or inoculated by flooding the root system with R. solanacearum. Roots (including root tips and root sections) and nodules were collected for analysis 24 h, 48 h, and 7 dpi, and the expression of M. truncatula genes was characterized. The six ‘‘mock’’ versus ‘‘inoculated’’ comparisons for roots and nodules (24 h, 48 h, and 7 dpi) revealed 739 genes whose expression varied significantly (adjusted p % 0.05; Figure 3B; Tables S1 and S2). Most of these genes (617/739) were upregulated upon inoculation with R. solanacearum. The response started earlier in nodules than in roots but concerns fewer (293) induced genes in nodules (versus 486 in roots). The highest response, in terms of upregulated genes, was observed 48 h after inoculation for both organs. R. solanacearum-induced genes showing the same expression kinetics in roots and nodules (102 genes) include typical defense-related genes, among which are 23 genes predicted to encode proteins involved in the biosynthesis of flavonoids, isoflavonoids, and terpenoids; five Toll/interleukin1 (IL-1) receptors (TIRs); two regulators of the WRKY family; and the respiratory burst oxydase homolog C (Table S1). Thus, our data indicate that nodules are able to activate defense reactions upon infection with R. solanacearum. Nodules Display a Specific Defense Program Strikingly, nodules and roots displayed distinct responses with only a limited number of genes commonly activated upon pathogen infection in the two organs (163/739), among which 37% (60 genes) are more rapidly induced in the nodules (Figure 3B; Table S1). One hundred and thirty genes were found specifically induced in R. solanacearum-infected nodules and not in roots (Figures 3B and 3D; Table S1). These notably include six genes of the flavonoid/isoflavonoid biosynthetic pathway, ten TIRs, three WRKY transcription factors, two thaumatins, two genes involved in the biosynthesis of ethylene, two chitinases, an endo-glucanase, two members of the RPW8 family, two late embryogenesis abundant protein homologs, and a pathogenesis-related protein 4. Remarkably, 10 out of the 22 receptor-like kinases (RLKs) Current Biology 30, 1–8, January 20, 2020 3

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

specifically induced in nodules harbor the non-arginine aspartate motif that is typical of pattern recognition receptors crucial for pathogen-associated-molecular-pattern-triggered immunity [22] (Figure 3D; Table S1). Among these ten RLKs, nine harbor an ectodomain similar to the pathogenesis-related protein 5. The specific induction of these defense-related genes in the nodules and not in the roots clearly indicates that nodules develop defense reactions distinct from roots as a whole. Despite Significant Upregulation, the Expression of Defense-Related Genes Remains Low in Nodules upon R. solanacearum Inoculation A striking feature emerging from this transcriptomic dataset is that essentially all the genes upregulated upon R. solanacearum colonization (Figure 3B) had a higher expression level in roots than in symbiotic organs independently of the plant inoculation by R. solanacearum (Figure 3B). For instance, out of the 289 genes that were upregulated by R. solanacearum in nodules 48 h after inoculation, 263 have levels in control nodules that are more than two times lower than in control roots. This is notably true for all the genes that are significantly induced by R. solanacearum and that harbor the ‘‘defense response’’ GO (gene ontology) term (Figure 3C). Thus, in nodule, because of a very low expression level and despite significant activation, defenses do not reach high intensity upon R. solanacearum attack. Perhaps because of this, the nodule defense response is not intense enough to systematically drastically reduce nitrogen fixation of infected nodules (Figure 1A) either to compromise radically rhizobia viability (Figure 1) or R. solanacearum growth. In agreement, nodulin expression was not significantly altered upon R. solanacearum infection, no signs of nodule senescence were visible, and no induction of the expression of senescence markers was detected (Figure 3C). Interestingly, despite an apparent low intensity of nodule defense response, the R. solanacearum T3SS strongly contributed to the infection and development of R. solanacearum in symbiotic organs (Figure 2), confirming that innate immunity is not totally switched off in nodules. Importantly, our transcriptomic data were produced with whole nodules that are constituted of various tissues (meristematic zone, uninfected cortex, infection zone, nitrogen fixation zone, etc.). Further characterization will be required to determine

Figure 3. Nodules Develop Weak and Partially Specific Defense Responses upon R. solanacearum Infection Responses of M. truncatula root and nodules to R. solanacearum were studied by RNA-seq. (A) Nodule and root RNA of the same plants, inoculated or not with R. solanacearum by flooding, were collected separately 24 h, 48 h, and 7 days

4 Current Biology 30, 1–8, January 20, 2020

after inoculation for sequence analysis. Three independent biological repeats were analyzed. (B) Hierarchical clustering of the 739 M. truncatula genes, in which expression varies significantly in at least one of the ‘‘mock’’ versus ‘‘R. solanacearum’’ comparisons (left). The color code represents the log of fold change. The red rectangle indicates the 130 genes that are specifically upregulated upon R. solanacearum infection in nodules and not in root. For these 739 genes, the nodules versus roots comparisons (expressed as log of fold change, right) illustrate that essentially all the genes differentially expressed upon R. solanacearum infection in roots and in nodules are downregulated in nodules as compared to roots independently of the infection. (C) Expression of 17 nodulin genes, six nodule senescence markers, as well as of the 66 genes harboring the ‘‘defense response’’ GO term (GO:006952) that are differentially expressed in one of the mock versus R. solanacearum comparisons presented in (B). (D) Venn diagram indicating the overlap and specificity of upregulated genes in nodules and in roots upon R. solanacearum infection. Examples of defenserelated genes specifically induced in nodules and not in root are indicated. See also Tables S1 and S2.

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

Figure 4. R. solanacearum Dissemination from a Nodule to Non-symbiotic Organs Is Delayed Independently of Nodule Functionality (A) Experimental setup: each plant was inoculated by piercing a root segment and a nodule. The two inoculation sites are equidistant from the aerial parts. mCherry and eGFP-tagged R. solanacearum were used to distinguish the origin of bacteria (nodule or root). All experiments included swapping of tagged strains (A). (B–D) Data from the swapped experiments were pooled. The presented results only refer to the type of inoculated organs (N and RS for nodule and root segment, respectively) and not to the type of the fluorescent tag. The experiments were performed three times independently, and results were similar. Asymmetric flows between nodules and non-symbiotic organs were observed during the race experiments. Inoculations by piercing were performed in (B) nitrogen fixing nodules and in (C) non-fixing nodules induced by the S. medicae bacA mutant. In (D), an inoculation by scalping the nodule apex with a blade soaked in R. solanacearum suspension was performed to improve accessibility to nodule vasculature. R. solanacearum densities in the indicated organs (the aerial parts, the nodules, and the roots) 14 days after inoculations and the origin of these populations are shown. In the aerial parts of the plants, in all cases, we observed significantly more bacteria originating from the root segments than from the nodule. Upon the three conditions, high densities of bacteria originating from root segments are observed in the nodules. In root segments, 14 days after inoculations, pathogens originating from nodules were not detected in most cases, in contrast to bacteria inoculated in root segments. ****p < 0.0001 after Mann Whitney tests. Dot lines represent the detection thresholds (20 bacteria/analyzed organ). See also Figure S4.

whether the different nodule zones display similar or different responsiveness to pathogen. Similarly, the reference roots were treated as a whole (including root tips and root sections). However, they are heterogeneous organs in which tips display a particular physiology. Like nodules, root tips harbor a relatively high content of young meristematic tissue. Further studies will be required to determine whether the specificity of the nodule response is due to their high content of young tissues and whether this can explain the R. solanacearum colonization patterns observed in the symbiotic organ. R. solanacearum Dissemination from Nodules to Nonsymbiotic Organs Is Hampered The R. solanacearum colonization patterns often display a sharp delimitation between totally infected nodules and the rest of the root system (Figure 1B), suggesting the existence of restrictions delaying or preventing R. solanacearum spreading out of nodules. In order to evaluate the dispersion capacity of R. solanacearum from the symbiotic organs, a race experiment was set up. Two R. solanacearum strains were used: one producing the enhanced green fluorescent protein (eGFP) and one accumulating the red fluorescent mCherry. Each plant root system was inoculated twice with R. solanacearum using needles. The two inoculation sites, a nodule and a root segment, were equidistant to the upper root branching (Figures S4 and 4A). 14 days after R. solanacearum inoculations, the aerial parts of

the plants were analyzed for the presence of green and red fluorescent bacteria to determine the origin of R. solanacearum reaching the leaves (Figure S4). The aerial parts were essentially colonized by the root segment inoculated R. solanacearum (61 plants out of 77), and in most cases (58 out of 77), R. solanacearum coming from the nodules was not detected in leaves (Figure 4). Swapping experiments indicated that fluorescent markers had no effect on the colonization pattern (Figure 4A). Like the aerial part, the root-branching zones located at equal distance of the two inoculation sites display more R. solanacearum coming from the root segments than from the nodules (Figure S4). As expected, the inoculated nodules and root segments both contained high densities of the R. solanacearum strain that were introduced in these organs (Figure 4B). In addition, nodules frequently contained R. solanacearum originating from root segments (63 out of 77). In contrast, R. solanacearum coming from the nodules were only detected in 16 root segments out of the 77 analyzed (Figure 4B). In addition, to be less frequently detected away from their site of inoculation, R. solanacearum coming from nodules only formed small populations outside of symbiotic organs as compared to R. solanacearum that originated from root segments (Figure 4B). Similar results were obtained with the bacA mutant, suggesting that nitrogen fixation has no influence on these asymmetric movements (Figure 4C). Furthermore, when nodule vasculatures were made directly Current Biology 30, 1–8, January 20, 2020 5

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

accessible for pathogen by cutting the symbiotic organs transversally using a scalpel soaked in R. solanacearum cell suspension (Figure S4), results were comparable to those obtained by piercing (Figure 4B and 4D), indicating that access to nodule vasculature is not the main factor limiting pathogen dissemination in the non-symbiotic organs. Thus, although symbiotic organs, like root tips, represent points of weakness with respect to pathogen entry within the plant root system and a proliferation site with all tissues supporting R. solanacearum growth, the dispersion of R. solanacearum from nodules to the rest of the plant appears hampered (Figure 4). The reason for the delayed dissemination from nodules remains to be clarified and might be explained by a tropism for nodules and/or an asymmetric diffusion barrier. Interestingly, during nodule life (including senescence), the rhizobial populations also do not spread in the plant. The reason is not clear, but in M. truncatula, compact tissue without meatuses between cells is observed at the level of nodule basis, which might contribute to hamper bacteria dissemination through apoplast. Furthermore, in plants of this clade, intracellular rhizobia are controlled by plant-produced peptides [23], among which some were proposed to interfere with rhizobial motility [24]. One cannot exclude that prevention of rhizobia and pathogen dissemination to non-symbiotic organs rely on the same mechanisms. To conclude, here, we described the setup and characterization of a tripartite interaction system that allowed us to highlight the weak defense capacity of nodules but also the existence of a mechanism that tends to compensate this deficiency by preventing dissemination of the pathogen beyond nodules. The plant benefit of the rhizobia-legume relationship is frequently highlighted for its agronomical and environmental interests, and this justifies current efforts to transfer the symbiotic capacity to non-legume plants [25–27]. However, recently, phylogenomic analysis showed that strong selective pressure acts against nodulation [28], suggesting that tradeoffs are associated with the symbiotic association. The experimental tool presented herein will be useful, in the near future, to investigate the causes and consequences of legume tolerance for rhizobia. STAR+METHODS Detailed methods are provided in the online version of this paper and include the following: d d d d

d d

KEY RESOURCES TABLE LEAD CONTACT AND MATERIALS AVAILABILITY EXPERIMENTAL MODEL AND SUBJECT DETAILS METHOD DETAILS B Nodulation B R. solanacearum inoculations and quantification of bacterial populations B Bacterial strain constructions B Histological analysis B Acetylene reduction assays (ARA) B Transcriptomic analysis QUANTIFICATION AND STATISTICAL ANALYSIS DATA AND CODE AVAILABILITY

6 Current Biology 30, 1–8, January 20, 2020

SUPPLEMENTAL INFORMATION Supplemental Information can be found online at https://doi.org/10.1016/j. cub.2019.11.066. ACKNOWLEDGMENTS bastien Carre`re, LIPM, who mapped the RNA-seq data We are grateful to Se onto the M. truncatula genome; to Laurent Sauviac, LIPM, for constant help cile Pouzet, Alain Jauneau, Marie Christine Auriac, and fruitful discussion; to Ce and Yves Martinez for help with histology and imaging technics; and to Clare Gough, LIPM, for critical reading of the manuscript. We express our gratitude to Gabriella Endre, Biological Research Centre of HAS, Szeged, Hungary, who kindly provided the S. medicae strain WSM419 carrying the plasmid pBHRmRFP. We also thank Anthony Perrier for his help with bacterial strain construction and his expertise on R. solanacearum. B.G. benefited from grants from the Agence Nationale de la Recherche (JCJC program) and from grants from the Labexes TULIP and SPS (ANR-10-LABX-41, ANR-11-IDEX-000202, ANR-17-CE20-0013, ANR-10-LABX-0040-SPS, and ANR-11-IDEX-0003partement Sante  des Plantes et Environne02) as well as support from the De ment of the INRA. AUTHOR CONTRIBUTIONS Conceptualization, C.B. and B.G.; Methodology, C.B. and B.G.; Validation, C.B., M.-F.J., and B.G.; Formal Analysis, C.B., M.-F.J., and B.G.; Investigation, C.B., F.B., A.L.S., M.M., G.J., and B.G.; Writing – Original Draft Preparation, C.B. and B.G.; Writing – Review and Editing, C.B., M.-F.J., P.R., F.V., and B.G.; Supervision, F.V. and B.G.; Project Administration and Funding Acquisition, B.G. DECLARATION OF INTERESTS The authors declare no competing interests. Received: May 29, 2019 Revised: August 2, 2019 Accepted: November 21, 2019 Published: January 2, 2020 REFERENCES 1. Sinharoy, S., Torres-Jerez, I., Bandyopadhyay, K., Kereszt, A., Pislariu, C.I., Nakashima, J., Benedito, V.A., Kondorosi, E., and Udvardi, M.K. (2013). The C2H2 transcription factor regulator of symbiosome differentiation represses transcription of the secretory pathway gene VAMP721a and promotes symbiosome development in Medicago truncatula. Plant Cell 25, 3584–3601. 2. Berrabah, F., Bourcy, M., Cayrel, A., Eschstruth, A., Mondy, S., Ratet, P., and Gourion, B. (2014). Growth conditions determine the DNF2 requirement for symbiosis. PLoS ONE 9, e91866. 3. Bourcy, M., Brocard, L., Pislariu, C.I., Cosson, V., Mergaert, P., Tadege, M., Mysore, K.S., Udvardi, M.K., Gourion, B., and Ratet, P. (2013). Medicago truncatula DNF2 is a PI-PLC-XD-containing protein required for bacteroid persistence and prevention of nodule early senescence and defense-like reactions. New Phytol. 197, 1250–1261. 4. Berrabah, F., Bourcy, M., Eschstruth, A., Cayrel, A., Guefrachi, I., Mergaert, P., Wen, J., Jean, V., Mysore, K.S., Gourion, B., and Ratet, P. (2014). A nonRD receptor-like kinase prevents nodule early senescence and defense-like reactions during symbiosis. New Phytol. 203, 1305– 1314. 5. Berrabah, F., Ratet, P., and Gourion, B. (2015). Multiple steps control immunity during the intracellular accommodation of rhizobia. J. Exp. Bot. 66, 1977–1985. 6. Gourion, B., Berrabah, F., Ratet, P., and Stacey, G. (2015). Rhizobiumlegume symbioses: the crucial role of plant immunity. Trends Plant Sci. 20, 186–194.

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

7. Wang, C., Yu, H., Luo, L., Duan, L., Cai, L., He, X., Wen, J., Mysore, K.S., Li, G., Xiao, A., et al. (2016). NODULES WITH ACTIVATED DEFENSE 1 is required for maintenance of rhizobial endosymbiosis in Medicago truncatula. New Phytol. 212, 176–191.

24. Penterman, J., Abo, R.P., De Nisco, N.J., Arnold, M.F., Longhi, R., Zanda, M., and Walker, G.C. (2014). Host plant peptides elicit a transcriptional response to control the Sinorhizobium meliloti cell cycle during symbiosis. Proc. Natl. Acad. Sci. USA 111, 3561–3566.

8. Domonkos, A´., Kova´cs, S., Gomba´r, A., Kiss, E., Horva´th, B., Kova´ts, G.Z., Farkas, A., To´th, M.T., Ayaydin, F., Bo´ka, K., et al. (2017). NAD1 controls defense-like responses in Medicago truncatula symbiotic nitrogen fixing nodules following rhizobial colonization in a BacA-independent manner. Genes (Basel) 8, 187.

25. Charpentier, M., and Oldroyd, G. (2010). How close are we to nitrogenfixing cereals? Curr. Opin. Plant Biol. 13, 556–564.

9. Yamaya-Ito, H., Shimoda, Y., Hakoyama, T., Sato, S., Kaneko, T., Hossain, M.S., Shibata, S., Kawaguchi, M., Hayashi, M., Kouchi, H., and Umehara, Y. (2018). Loss-of-function of ASPARTIC PEPTIDASE NODULE-INDUCED 1 (APN1) in Lotus japonicus restricts efficient nitrogen-fixing symbiosis with specific Mesorhizobium loti strains. Plant J. 93, 5–16. 10. Gourion, B., and Alunni, B. (2018). Strain-specific symbiotic genes: a new level of control in the intracellular accommodation of rhizobia within legume nodule cells. Mol. Plant Microbe Interact. 31, 287–288. 11. Pecrix, Y., Staton, S.E., Sallet, E., Lelandais-Brie`re, C., Moreau, S., Carre`re, S., Blein, T., Jardinaud, M.F., Latrasse, D., Zouine, M., et al. (2018). Whole-genome landscape of Medicago truncatula symbiotic genes. Nat. Plants 4, 1017–1025. , F., Oldroyd, G.E., Geurts, R., Cannon, S.B., Udvardi, 12. Young, N.D., Debelle M.K., Benedito, V.A., Mayer, K.F., Gouzy, J., Schoof, H., et al. (2011). The Medicago genome provides insight into the evolution of rhizobial symbioses. Nature 480, 520–524. 13. Jones, K.M., Kobayashi, H., Davies, B.W., Taga, M.E., and Walker, G.C. (2007). How rhizobial symbionts invade plants: the SinorhizobiumMedicago model. Nat. Rev. Microbiol. 5, 619–633. 14. Sugawara, M., Epstein, B., Badgley, B.D., Unno, T., Xu, L., Reese, J., Gyaneshwar, P., Denny, R., Mudge, J., Bharti, A.K., et al. (2013). Comparative genomics of the core and accessory genomes of 48 Sinorhizobium strains comprising five genospecies. Genome Biol. 14, R17. 15. Mansfield, J., Genin, S., Magori, S., Citovsky, V., Sriariyanum, M., Ronald, P., Dow, M., Verdier, V., Beer, S.V., Machado, M.A., et al. (2012). Top 10 plant pathogenic bacteria in molecular plant pathology. Mol. Plant Pathol. 13, 614–629. 16. Peeters, N., Guidot, A., Vailleau, F., and Valls, M. (2013). Ralstonia solanacearum, a widespread bacterial plant pathogen in the post-genomic era. Mol. Plant Pathol. 14, 651–662. 17. Deslandes, L., and Genin, S. (2014). Opening the Ralstonia solanacearum type III effector tool box: insights into host cell subversion mechanisms. Curr. Opin. Plant Biol. 20, 110–117. 18. Vailleau, F., Sartorel, E., Jardinaud, M.F., Chardon, F., Genin, S., Huguet, T., Gentzbittel, L., and Petitprez, M. (2007). Characterization of the interaction between the bacterial wilt pathogen Ralstonia solanacearum and the model legume plant Medicago truncatula. Mol. Plant Microbe Interact. 20, 159–167. 19. Turner, M., Jauneau, A., Genin, S., Tavella, M.J., Vailleau, F., Gentzbittel, L., and Jardinaud, M.F. (2009). Dissection of bacterial Wilt on Medicago truncatula revealed two type III secretion system effectors acting on root infection process and disease development. Plant Physiol. 150, 1713– 1722. 20. Berrabah, F., Ratet, P., and Gourion, B. (2019). Legume nodules: massive infection in the absence of defense induction. Mol. Plant Microbe Interact. 32, 35–44.

26. Mus, F., Crook, M.B., Garcia, K., Garcia Costas, A., Geddes, B.A., Kouri, E.D., Paramasivan, P., Ryu, M.H., Oldroyd, G.E.D., Poole, P.S., et al. (2016). Symbiotic nitrogen fixation and the challenges to its extension to nonlegumes. Appl. Environ. Microbiol. 82, 3698–3710. 27. Beatty, P.H., and Good, A.G. (2011). Plant science. Future prospects for cereals that fix nitrogen. Science 333, 416–417. 28. Griesmann, M., Chang, Y., Liu, X., Song, Y., Haberer, G., Crook, M.B., Billault-Penneteau, B., Lauressergues, D., Keller, J., Imanishi, L., et al. (2018). Phylogenomics reveals multiple losses of nitrogen-fixing root nodule symbiosis. Science 361, eaat1743. 29. Howieson, J.G., and Ewing, M.A. (1986). Acid tolerance in the Rhizobium meliloti-Medicago symbiosis. Aust. J. Agric. Res. 37, 55–64. na, G. 30. Rangin, C., Brunel, B., Cleyet-Marel, J.C., Perrineau, M.M., and Be (2008). Effects of Medicago truncatula genetic diversity, rhizobial competition, and strain effectiveness on the diversity of a natural sinorhizobium species community. Appl. Environ. Microbiol. 74, 5653–5661. 31. Sallet, E., Roux, B., Sauviac, L., Jardinaud, M.F., Carre`re, S., Faraut, T., de Carvalho-Niebel, F., Gouzy, J., Gamas, P., Capela, D., et al. (2013). Nextgeneration annotation of prokaryotic genomes with EuGene-P: application to Sinorhizobium meliloti 2011. DNA Res. 20, 339–354. 32. Galardini, M., Mengoni, A., Brilli, M., Pini, F., Fioravanti, A., Lucas, S., Lapidus, A., Cheng, J.F., Goodwin, L., Pitluck, S., et al. (2011). Exploring the symbiotic pangenome of the nitrogen-fixing bacterium Sinorhizobium meliloti. BMC Genomics 12, 235. 33. Galardini, M., Bazzicalupo, M., Biondi, E., Brambilla, E., Brilli, M., Bruce, D., Chain, P., Chen, A., Daligault, H., Davenport, K.W., et al. (2013). Permanent draft genome sequences of the symbiotic nitrogen fixing Ensifer meliloti strains BO21CC and AK58. Stand. Genomic Sci. 9, 325–333. 34. Boucher, C.A., Barberis, P.A., and Demery, D.A. (1985). Transposon mutagenesis of Pseudomonas solanacearum: isolation of Tn5-induced avirulent mutants. Microbiology 131, 2449–2457. 35. Salanoubat, M., Genin, S., Artiguenave, F., Gouzy, J., Mangenot, S., Arlat, M., Billault, A., Brottier, P., Camus, J.C., Cattolico, L., et al. (2002). Genome sequence of the plant pathogen Ralstonia solanacearum. Nature 415, 497–502. 36. Etchebar, C., Trigalet-Demery, D., van Gijsegem, F., Vasse, J., and Trigalet, A. (1998). Xylem colonization by an HrcV mutant of Ralstonia solanacearum is a key factor for the efficient biological control of tomato bacterial wilt. Mol. Plant Microbe Interact. 11, 869–877. 37. Perrier, A., Barlet, X., Rengel, D., Prior, P., Poussier, S., Genin, S., and Guidot, A. (2019). Spontaneous mutations in a regulatory gene induce phenotypic heterogeneity and adaptation of Ralstonia solanacearum to changing environments. Environ. Microbiol. 21, 3140–3152. 38. Cruz, A.P., Ferreira, V., Pianzzola, M.J., Siri, M.I., Coll, N.S., and Valls, M. (2014). A novel, sensitive method to evaluate potato germplasm for bacterial wilt resistance using a luminescent Ralstonia solanacearum reporter strain. Mol. Plant Microbe Interact. 27, 277–285.

21. Jones, J.D., and Dangl, J.L. (2006). The plant immune system. Nature 444, 323–329.

39. Cunnac, S., Occhialini, A., Barberis, P., Boucher, C., and Genin, S. (2004). Inventory and functional analysis of the large Hrp regulon in Ralstonia solanacearum: identification of novel effector proteins translocated to plant host cells through the type III secretion system. Mol. Microbiol. 53, 115–128.

22. Dardick, C., and Ronald, P. (2006). Plant and animal pathogen recognition receptors signal through non-RD kinases. PLoS Pathog. 2, e2.

40. R Core Team (2017). R: A language and environment for statistical computing (R Foundation for Statistical Computing).

23. Alunni, B., and Gourion, B. (2016). Terminal bacteroid differentiation in the legume-rhizobium symbiosis: nodule-specific cysteine-rich peptides and beyond. New Phytol. 211, 411–417.

41. RStudio Team (2015). RStudio: Integrated development for R (RStudio). 42. Dray, S., and Dufour, A.-B. (2007). The ade4 Package: implementing the duality diagram for ecologists. J. Stat. Softw. 22, 1–20.

Current Biology 30, 1–8, January 20, 2020 7

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

43. Kolde, R. (2018). pheatmap: pretty heatmaps. R package version 1.0.10 (CRAN). https://CRAN.R-project.org/package=pheatmap. 44. Neuwirth, E. (2014). RColorBrewer: ColorBrewer palettes. R package version 1.1-2. https://cran.r-project.org/web/packages/RColorBrewer/ index.html. 45. Warnes, G.W., Bolker, B., Bonebakker, L., Gentleman, R., Huber, W., Liaw, A., et al. (2016). gplots: various R programming tools for plotting data. R package version 3.0.1 (CRAN). https://CRAN.R-project.org/ package=gplots. 46. Hansen, K.D., Gentry, J., Long, L., Gentleman, R., Falcon, S., Hahne, F., et al. (2017). Rgraphviz: Provides plotting capabilities for R graph objects. R package version 2.22.0 (CRAN). https://CRAN.R-project.org/ package=Rgraphviz. 47. Robinson, M.D., McCarthy, D.J., and Smyth, G.K. (2010). edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26, 139–140. 48. Fahraeus, G. (1957). The infection of clover root hairs by nodule bacteria studied by a simple glass slide technique. J. Gen. Microbiol. 16, 374–381. 49. Beringer, J.E. (1974). R factor transfer in Rhizobium leguminosarum. J. Gen. Microbiol. 84, 188–198. , F., Gough, C., Bisseling, T., 50. Smit, P., Raedts, J., Portyanko, V., Debelle and Geurts, R. (2005). NSP1 of the GRAS protein family is essential for rhizobial Nod factor-induced transcription. Science 308, 1789–1791.

8 Current Biology 30, 1–8, January 20, 2020

51. Poueymiro, M., Cunnac, S., Barberis, P., Deslandes, L., Peeters, N., Cazale-Noel, A.C., Boucher, C., and Genin, S. (2009). Two type III secretion system effectors from Ralstonia solanacearum GMI1000 determine host-range specificity on tobacco. Mol. Plant Microbe Interact. 22, 538–550. 52. Sauviac, L., and Bruand, C. (2014). A putative bifunctional histidine kinase/ phosphatase of the HWE family exerts positive and negative control on the Sinorhizobium meliloti general stress response. J. Bacteriol. 196, 2526– 2535. 53. Glazebrook, J., Ichige, A., and Walker, G.C. (1993). A Rhizobium meliloti homolog of the Escherichia coli peptide-antibiotic transport protein SbmA is essential for bacteroid development. Genes Dev. 7, 1485–1497. 54. Jefferson, R.A., Kavanagh, T.A., and Bevan, M.W. (1987). GUS fusions: beta-glucuronidase as a sensitive and versatile gene fusion marker in higher plants. EMBO J. 6, 3901–3907. 55. Cerutti, A., Jauneau, A., Auriac, M.C., Lauber, E., Martinez, Y., Chiarenza, , R., and Noe¨l, L.D. (2017). Immunity at cauliS., Leonhardt, N., Berthome flower hydathodes controls systemic infection by Xanthomonas campestris pv campestris. Plant Physiol. 174, 700–716. 56. Roux, B., Rodde, N., Moreau, S., Jardinaud, M.F., and Gamas, P. (2018). Laser capture micro-dissection coupled to RNA sequencing: a powerful approach applied to the model legume Medicago truncatula in interaction with Sinorhizobium meliloti. Methods Mol. Biol. 1830, 191–224.

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

STAR+METHODS KEY RESOURCES TABLE

REAGENT or RESOURCE

SOURCE

IDENTIFIER

Sinorhizobium medicae WSM419

[29]

N/A

Sinorhizobium medicae WSM419 pBHR-mRFP

Gabriella Endre

N/A

Sinorhizobium medicae WSM419 DbacA

This paper

N/A

Sinorhizobium medicae MD4

[30]

N/A

Sinorhizobium meliloti 2011

[31]

N/A

Sinorhizobium meliloti BL225C

[32]

N/A

Sinorhizobium meliloti AK83

[32]

N/A

Sinorhizobium meliloti BO21CC

[33]

N/A

Sinorhizobium meliloti AK58

[33]

N/A

Ralstonia solanacearum GMI1000

[34, 35]

N/A

Ralstonia solanacearum GMI1559

[36]

N/A

Ralstonia solanacearum GMI1000 eGFP

[37]

N/A

Ralstonia solanacearum GMI1000 mCherry

[38]

N/A

Ralstonia solanacearum GMI1000 hrcV

[39]

N/A

miRNeasy Mini Kit

QIAGEN

217004

TruSeq stranded mRNA Library Prep

Illumina

20020595

HS NGS Fragment Kit (1-6000bp), 500

Agilent Technologies

DNF-474-0500

HiSeq 3000/4000 PE Cluster kit

Illumina

PE-410-1001

HiSeq 3000/4000 SBS Kit (300 cycles)

Illumina

FC-410-1003

Kapa Library Quantification Kit

Roche

07960140001

AMPure

Beckman-Coulter

A63882

This paper

NCBI SRA SRP229031

[11]

N/A

OCB1706 50 -GTCGACGATGATCACGGCCTACGGC-30

This paper

N/A

OCB1707 50 -GGATCCGGGACGGCACTCTCGTTTC-30

This paper

N/A

OCB1708 50 -GGATCCCGGCAGGCATCAGAAGGCGG-30

This paper

N/A

OCB1709 50 -GAGCTCGGCTGGCCCAGCTTCGCA-30

This paper

N/A

0

This paper

N/A

OCB1790 50 -TGTCTCTCAGCGACGGCCT-30

This paper

N/A

This paper

N/A

GraphPad Prism 7

GraphPad Software

https://www.graphpad.com/

Imagej

N/A

https://imagej.nih.gov/ij/

R version 3.4.3

[40]

https://www.R-project.org/.

RStudio Version 1.0.143

[41]

http://www.rstudio.com/.

Ade4 package version 1.7-11

[42]

https://CRAN.R-project.org/package=ade4

Pheatmap version 1.0.10

[43]

https://CRAN.R-project.org/package=pheatmap

RColorBrewer version 1.0-2

[44]

https://CRAN.R-project.org/package=RColorBrewer

Bacterial Strains

Critical Commercial Assays

Deposited Data Raw and analyzed data Experimental Models: Organisms/Strains Medicago truncatula A17 Oligonucleotides

0

OCB1789 5 -GGCCTGCTGGTCGAGGCG-3 Recombinant DNA pCB1 Software and Algorithms

(Continued on next page)

Current Biology 30, 1–8.e1–e4, January 20, 2020 e1

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

Continued REAGENT or RESOURCE

SOURCE

IDENTIFIER

gplots version 3.0.1

[45]

https://CRAN.R-project.org/package=gplots

Rgraphviz version 2.22.0

[46]

https://CRAN.R-project.org/package=Rgraphviz

EdgeR package version 3.20.9

[47]

https://CRAN.R-project.org/package=EdgeR

LEAD CONTACT AND MATERIALS AVAILABILITY Further information and requests for the Sinorhizobium medicae DbacA mutant generated during this study should be directed to and will be fulfilled by the Lead Contact, Benjamin Gourion ([email protected]). EXPERIMENTAL MODEL AND SUBJECT DETAILS M. truncatula A17 seeds were surface sterilized by ten minutes incubation in sulfuric acid. Surface sterilized seeds were then abundantly washed with distilled sterile water and laid onto 1% agar plate prior to incubation for 48 h at 4 C in the darkness. Then seeds were incubated for 48 h at room temperature for germination. Eight seedlings were transferred under sterile conditions onto paper covered 12 cm side square plates containing Fahraeus medium (0.132 g/L CaCl2, 0.12 g/L MgSO4.7H2O, 0.1 g/L KH2PO4, 0.075 g/L Na2HPO4.2H2O, 5 mg/L Fe-citrate, and 0.07 mg/L each of MnCl2.4H2O, CuSO4.5H2O, ZnCl2, H3BO3, and Na2MoO4.2H2O, adjusted to pH 7.5 before autoclaving) [48]. For nodulation, plants were then cultivated for fourteen days (except when indicated in the Figure legends) in those Fahraeus containing plates in a growth chamber at 25 C with a 16/8 h light/dark photoperiod. Sinorhizobium meliloti strains 2011 [31], BL225C [32], AK83 [32], BO21CC [33], AK58 [33] and Sinorhizobium medicae strains MD4 [30] and WSM419 as well as the bacA mutant [29] were cultivated in TY medium (5g/L bacto tryptone, 3g/L yeast extract) [49] supplemented with 6 mM CaCl2 at 30 C. S. medicae strain WSM419 carrying the plasmid pBHR-mRFP [50] was cultivated in TY medium [49] supplemented with 6 mM CaCl2 with tetracycline. R. solanacearum strain GMI1000 [34, 35], GMI1559 (uidA labeled GMI1000) [36], eGFP [37] and mCherry tagged GMI1000 [38] as well as R. solanacearum strain GMI1000 altered in hrcV [39] were cultivated in complete B medium (10 g/L bacto peptone, 1g/L casamino acids, 1g/L yeast extract) [51] at 28 C. METHOD DETAILS Nodulation For nodulation, overnight cultures of rhizobia were pelleted before being suspended in sterile water to reach an OD600nm of 0.1. One milliliter of bacterial suspension was used per plate to homogeneously inoculate the root of eight seedlings immediately after transfer on Fahraeus plates [48]. Plants were cultivated in a growth chamber at 25 C with a 16/8 h light/dark photoperiod. Nodule bumps could be observed from 7 dpi and nodules were mature at 14 dpi. R. solanacearum inoculations and quantification of bacterial populations To inoculate plants, Ralstonia was cultivated in B medium and overnight cultures were washed and re-suspended in distilled sterile water. The ODs were adjusted for inoculation. Depending on the aim of the experiment, nodulated roots of M. truncatula were inoculated using different methods that are indicated in the figure legends and in the Results and Discussion section. For all the experiments described in this study, plants were cultivated and nodulated in Fahraeus containing plates before inoculation with Ralstonia. For flooding inoculation, the nodulated root systems of eight plants cultivated in one plate were flooded with 5 mL of R. solanacearum suspension (OD600nm = 0.01). For spotting inoculation, 5 mL of OD600nm = 0.01 R. solanacearum cell suspension was dropped per root apex. For piercing inoculation, a sterile needle soaked in the inoculum (R. solanacearum suspended in sterile water, OD600nm = 0.01) was used to pierce the nodule. Root tips and root segments of nodulated plants were inoculated using the same procedure and analyzed as reference organs in which R. solanacearum is known to multiply. Alternatively, when indicated for race experiments, a scalpel soaked in the R. solanacearum suspension was used in order to make nodule vasculature directly accessible for pathogen by cutting the symbiotic organs transversally with the soaked scalpel. To determine R. solanacearum populations in the plant material, the different organs, nodules, root tips, root segments, aerial parts or root branching zones were individually placed in tubes and grinded. The material was resuspended in sterile water and serial dilution were plated on B medium. Colonies were counted after incubation at 28 C for 48 h. When indicated, eGFP and mCherry colonies were counted using a fluorescence stereo zoom microscope (Axiozoom V16, Zeiss). Bacterial strain constructions S. medicae bacA deletion mutant was constructed by double recombination following the method described in [52]. Briefly, approximately 400 base pair fragments corresponding to the upstream and the downstream regions of the WSM419 bacA gene e2 Current Biology 30, 1–8.e1–e4, January 20, 2020

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

(Smed_4510) were amplified by PCR using primers OCB1706 50 -GTCGACGATGATCACGGCCTACGGC-30 , OCB1707 50 -GGATCCGGGACGGCACTCTCGTTTC-30 , for the upstream region and OCB1708, 50 -GGATCCCGGCAGGCATCAGAAGGCGG-30 , OCB1709 50 -GAGCTCGGCTGGCCCAGCTTCGCA-30 for the downstream region. The PCR products were cloned in a pJQ200mp19 suicide vector to generate the pCB1 and this plasmid was introduced by triparental mating in strain WSM419. Integrant were selected using the gentamicin resistance gene carried by the pJQ200mp19. Vector excision was then selected on 5% sucrose plate and deletion of the bacA gene was confirmed by PCR using the primers OCB1789, 50 -GGCCTGCTGGTCGAGGCG-30 and OCB1790, 50 -TGTCTCTCAGCGACGGCCT-30 and by sequencing. Acetylene reduction assay (see below for method) confirmed the inability of WSM419 bacA mutant to fix nitrogen as observed for the bacA mutant of S. meliloti strain 1021 [53]. Histological analysis GUS staining of infected nodulated plants was performed as follow: nodulated plants were inoculated with R. solanacearum GMI1559 [36] by flooding as described in the ‘‘R. solanacearum inoculations and quantification of bacterial populations’’ paragraph. This strain is a GMI1000 derivative, not altered in pathogenicity and carrying the uidA reporter gene inserted in its genome [36]. Plants were harvested and fixed in 0.1% paraformaldehyde in a 0.1 M phosphate buffer under vacuum. Then samples were washed three times in 0.1 M phosphate buffer. b-glucuronidase activity was detected using the X-gluc (5-bromo-4-chloro-3-indolyl-b-D-glucuronide, Clonetech Laboratories) as a substrate [54]. Samples were incubated 48 h at 28 C under stirring in 1 mg/mL X-gluc, 0.1 M NaH2PO4 (pH 7.0), 5 mM K3Fe(CN)6, 5 mM K4Fe(CN)6, and 10 mM EDTA. Then samples were fixed in 2.5% glutaraldehyde and washed in 0.1 M phosphate buffer (pH 7.0) before dehydration and depigmentation in 70% ethanol. Semi-thin sections of nodules were prepared after embedding material in Technovit resin. 10 mm sections were prepared using a microtome. Sections were stained with 1% ruthenium red. Stained sections were observed using an AxioPlan Imaging microscope equipped with a color CCD camera (Zeiss). Alternatively, 70 mm thickness sections were prepared with a vibratome (VT1000S; Leica). Fresh nodule sections were mounted in water on a glass slide and covered with a coverslip. Confocal images were acquired with a laser scanning confocal microscope (Leica SP8). A 488nm ray line of an argon laser was used for the excitation of eGFP, and the emitted fluorescence was collected between 500 and 550nm. A 561nm ray line of an argon laser was used for the excitation of RFP and the emitted fluorescence was collected between 570 and 650nm. For scanning electron microscopy, samples were fixed in 2.5% glutaraldehyde, 5 mM sodium cacodylate (pH 7.2), progressively dehydrated in ethanol, and then critical-point dried with liquid CO2, grounded with conductive silver paint on the observation plate, and sputter coated with platinum. Images were acquired with a scanning electron microscope (SEM. Quanta 250 FEG FEI) at 5 kV, spot size 3 with a working distance of 1 cm. Observations of vascular bundles in clarified nodules were performed after nodules incubation for three weeks in an aqueous solution of chlorate hydrate (45 g of chloral hydrate and 9.3 mL of a solution 60% glycerol in 7.6 mL water). Samples were mounted in the same solution and images were acquired in Nomarsky (DIC) using an inverted microscope (DMIRBE; Leica) equipped with a color CCD camera (DFC300 FX; Leica) as described previously [55]. Acetylene reduction assays (ARA) Assays were performed on single plants as previously described [2]. Briefly, single nodulated plants were placed into 10 mL glass vials containing 200 mL of sterile water and closed with septa. 250 mL of acetylene were injected per vial. Plants were incubated in growth chambers (25 C with light) for one to three hours (linear phase of ethylene production). Then, 400 mL of gas samples were analyzed on an Agilent 7020 equipped with a flame ionization detector. Activity was normalized with the incubation time. Transcriptomic analysis Fourteen days after inoculation with S. medicae, nodulated plants were inoculated with R. solanacearum by flooding root systems. Roots and nodules from eight to sixteen plants were collected at 24 h, 48 h and seven days after inoculation with R. solanacearum and immediately frozen in liquid nitrogen. Three independent biological repeats were performed. Material was grounded and RNA extracted using Tissue Lyser II (QIAGEN) for one minute on dry samples and for one minute in a 1% b-Mercaptoethanol RLT solution of the miRNeasy Mini Kit. Then, samples were centrifugated and the supernatant was cleaned by a second centrifugation step to remove debris. DNA was precipitated using absolute ethanol. RNeasy Mini Spin Columns were used to bind total RNA, columns were washed with RW1 solution and DNase treatment was performed directly on the column membrane. Then columns were washed with RW1 and RPE solutions. RNAs were eluted in 0.1% DEPC water. RNaseq was performed at the GeT-PlaGe core facility, INRA Toulouse. RNA-seq libraries were prepared according to Illumina’s protocols using the Illumina TruSeq Stranded mRNA sample prep kit. Briefly, mRNAs were selected using poly-T beads. Then, RNA were fragmented to generate double stranded cDNA and adaptors were ligated to be sequenced. Eleven cycles of PCR were applied to amplify libraries. Library quality was assessed using a Fragment Analyzer and libraries were quantified by QPCR using the Kapa Library Quantification Kit. RNA-seq experiments have been performed on an Illumina HiSeq3000 using a paired-end read length of 2x150 bp with the Illumina HiSeq3000 sequencing kits. Mapping of RNA sequencing reads on the latest M. truncatula genomic sequence (v5) [11] was done using the Glint software (Faraut T. and Courcelle E.; http://lipm-bioinfo.toulouse.inra.fr/download/glint/) with the following parameters–mmis 5–matedist 10000 –lmin 80. When several reads mapped to two different positions with different scores only the reads with the best score

Current Biology 30, 1–8.e1–e4, January 20, 2020 e3

Please cite this article in press as: Benezech et al., Medicago-Sinorhizobium-Ralstonia Co-infection Reveals Legume Nodules as Pathogen Confined Infection Sites Developing Weak Defenses, Current Biology (2019), https://doi.org/10.1016/j.cub.2019.11.066

were kept. Multi-mapped reads were discarded. We used the latest annotation release (version 1.6) to count the number of completely overlapping hits onto gene loci. A raw count was thus obtained for each gene. QUANTIFICATION AND STATISTICAL ANALYSIS For all results except the RNaseq analysis, the statistical analyses were performed using GraphPad Prism 7 (GraphPad Software, La Jolla, CA). The statistical tests used, definition of center, dispersion as well as the exact values of n and what n represents are indicated in the Figure legends, except for Figure 4 where sampling sizes are indicated directly on the Figure. For transcriptomics, we performed statistical analysis as previously described [56] using R software. We removed genes without at least 1 read/count per million. Using the EdgeR package, we normalized the counts using the Trimmed Mean Method. Reproducibilty was assessed using principal component analysis (Ade4 package [42]) and heatmaps illustrating Euclidean distances between biological replicates (pheatmap and RColorBrewer packages). Using Genewise Negative Binomial Generalized Linear Models, we set the design matrix taking into account treatment, time, and organ factors using biological replicates as the bloking factor and conducted likelihood ratio tests as decribed in the EdgeR manual. After selecting differentially expressed genes, we conducted hierarchical clustering (Ward.D2 method) on Euclidean distances and represented then with heatmaps (gplots package). DATA AND CODE AVAILABILITY The transcriptomic datasets generated during this study are available at https://www.ncbi.nlm.nih.gov/sra under the reference SRP229031.

e4 Current Biology 30, 1–8.e1–e4, January 20, 2020