Pedro-Jové et al. BMC Genomics (2021) 22:170 https://doi.org/10.1186/s12864-021-07457-w RESEARCH ARTICLE Open Access Dynamic expression of Ralstonia solanacearum virulence factors and metabolism-controlling genes during plant infection R. de Pedro-Jové1,2†, M. Puigvert1,2†, P. Sebastià2†, A. P. Macho3, J. S. Monteiro4, N. S. Coll2, J. C. Setúbal4 and M. Valls1,2* Abstract Background: Ralstonia solanacearum is the causal agent of bacterial wilt, a devastating plant disease responsible for serious economic losses especially on potato, tomato, and other solanaceous plant species in temperate countries. In R. solanacearum, gene expression analysis has been key to unravel many virulence determinants as well as their regulatory networks. However, most of these assays have been performed using either bacteria grown in minimal medium or in planta, after symptom onset, which occurs at late stages of colonization. Thus, little is known about the genetic program that coordinates virulence gene expression and metabolic adaptation along the different stages of plant infection by R. solanacearum. Results: We performed an RNA-sequencing analysis of the transcriptome of bacteria recovered from potato apoplast and from the xylem of asymptomatic or wilted potato plants, which correspond to three different conditions (Apoplast, Early and Late xylem). Our results show dynamic expression of metabolism-controlling genes and virulence factors during parasitic growth inside the plant. Flagellar motility genes were especially up-regulated in the apoplast and twitching motility genes showed a more sustained expression in planta regardless of the condition. Xylem-induced genes included virulence genes, such as the type III secretion system (T3SS) and most of its related effectors and nitrogen utilisation genes. The upstream regulators of the T3SS were exclusively up- regulated in the apoplast, preceding the induction of their downstream targets. Finally, a large subset of genes involved in central metabolism was exclusively down-regulated in the xylem at late infection stages. (Continued on next page) * Correspondence: marcvalls@ub.edu † R. de Pedro-Jové, M. Puigvert and P. Sebastià contributed equally to this work. 1 Department of Genetics, University of Barcelona, Barcelona, Catalonia, Spain 2 Centre for Research in Agricultural Genomics (CSIC-IRTA-UAB-UB), Bellaterra, Catalonia, Spain Full list of author information is available at the end of the article © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 2 of 18 (Continued from previous page) Conclusions: This is the first report describing R. solanacearum dynamic transcriptional changes within the plant during infection. Our data define four main genetic programmes that define gene pathogen physiology during plant colonisation. The described expression of virulence genes, which might reflect bacterial states in different infection stages, provides key information on the R. solanacearum potato infection process. Keywords: Ralstonia solanacearum, Bacterial wilt, RNAseq, Virulence factors, Dynamic gene expression, Metabolism, T3SS, Effectors, Xylem, Apoplast Background colonisation, since multiple deletion of the egl, pehA/ Brown rot or bacterial wilt of potato is a vascular disease B/C, pme and cbhA genes compromised pathogenicity caused by the bacterial phytopathogen Ralstonia solana- [25]. Bacterial motility also plays important roles dur- cearum. This gram-negative β-proteobacterium is among ing parasitic life in planta. For instance, R. solana- the most threatening bacterial phytopathogens world- cearum flagellar components were shown to be wide, as it can infect over 200 different plant species, in- essential at early stages of infection [26] and mutants cluding many important crops such as potato, tomato, in the main twitching gene pilA were less pathogenic peanut, eggplant and banana [1–3]. Although R. solana- [27]. On the other hand, the R. solanacearum genome cearum is endemic of tropical and sub-tropical regions, encodes the necessary enzymes to use nitrate as an phylotype IIB-1 strains such as UY031 are acclimated to energy source (i.e. dissimilatory nitrate reduction), to lower temperatures and have caused important out- incorporate nitrate as a molecular building block (i.e. breaks in temperate areas [4–6]. assimilatory nitrate reduction) [28] and to detoxify R. solanacearum has a complex life cycle. The patho- reactive nitrogen species (i.e. denitrification) [29]. The gen survives in soil and water for long periods of time ability to use nitrate as terminal electron acceptor has [7]. When R. solanacearum senses the roots of natural been proposed to sustain rapid bacterial growth in hosts by plant exudates [8], it penetrates the host the xylem, a hypoxic environment that is nonetheless through the root elongation zone, root wounds or sec- rich in nitrate [29, 30]. ondary root emerging points [9]. The root intercellular Gene regulation analyses are essential to decipher how spaces (the apoplast) constitutes a front line in the arms R. solanacearum finely tunes its pathogenicity. For race in plant-pathogen interactions and it is thus a hos- instance, transcription of the hrp genes -encoding the tile environment to phytopathogens [10]. Therefore, col- T3SS- and its related effectors was found to be con- onisation of the apoplast is key for R. solanacearum trolled by the HrpB transcriptional activator. HrpB lies pathogenicity [11–13]. Successful infections involve downstream of a regulatory cascade induced by bacterial entry into the vascular cylinder and extensive colonisa- contact with the plant cell wall [31] . The cascade tion of the xylem vessels [14, 15]. Occlusion of the vas- includes the membrane receptor PrhA, the signal trans- culature due to massive exopolysaccharide (EPS) ducer PrhI and the transcriptional regulators PrhJ and production and bacterial multiplication ultimately cause HrpG, the latter directly activating hrpB transcription wilting symptoms and plant death [9, 16]. [32, 33]. Gene expression studies demonstrated that the To progress across the different plant tissues, R. R. solanacearum hrp genes and T3SS effectors were solanacearum uses a panoply of virulence determi- transcribed in planta at late infection stages [34]. Based nants [17, 18]. The main virulence factor in this and on these results, it was speculated that R. solanacearum many other pathogenic bacteria is the Type III Secre- could inject T3Es to the xylem parenchyma cells in tion System (T3SS) [19, 20], which delivers effector order to hijack plant defences and manipulate the host proteins inside the plant cells, hijacking the cellular metabolism [34]. These findings were later confirmed in machinery for bacterial benefit [21]. Another key viru- gene expression studies using bacteria extracted from in- lence determinant is EPS. EPS leads to the clogging fected tomato and banana plants [14, 35] or bacterial of the xylem vessels and plant symptom appearance, transcripts isolated from infected potato roots [36]. Simi- and it can also bind to the cell wall and protect the lar to the T3SS, EPS production is also stringently con- bacterium from plant defences [22, 23]. In addition, trolled through the expression of the eps operon, which the general secretion system (type II) secretes import- encodes all EPS biosynthesis genes [37]. The eps operon ant virulence factors into the apoplast, including cell promoter is dependent on the global regulator PhcA, wall degrading enzymes [24]. These enzymes are whose production is induced at bacterial densities above collectively important for R. solanacearum plant 107 CFU/ml [37–39] . Finally, it has been described that Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 3 of 18 some crosstalk exists between the eps and the hrp gene Results regulation, since hrpG is negatively regulated by phcA R. solanacearum transcriptomes reflect four main genetic [32, 40]. programmes inside the plant Bacterial interactions in plant hosts do not consist To elucidate the genes deployed by R. solanacearum on one static phase, but rather in a dynamic inter- throughout infection, we profiled the gene expression of action during disease development. However, all R. strain UY031 in its natural susceptible potato host. We solanacearum in planta transcriptomic studies have collected bacterial samples from the apoplast —a condi- focused so far on a specific stage of the infection tion mimicking early root infection, when the bacterium process: xylem colonisation at the onset of disease traverses and multiplies in this compartment [47]— and symptoms [14, 35, 41, 42], with the exception of a from the xylem of infected plants at six and ten days single study indirectly analysing bacterial reads from post-inoculation, which correspond to the onset of the infected roots [36]. Among the differentially expressed disease (early xylem) or to the final stages when plants (DE) genes identified in these previous studies, the are completely wilted (late xylem) (Additional File 1 B T3SS, T3Es, motility genes, ROS detoxifying enzymes and 2A). R. solanacearum plant infection through roots and cell wall degrading enzymes were found up- is highly variable due to stochastic changes in the regulated in most cases. Dynamic transcriptomic stud- physiological state of the plant, the initial inoculum and ies of the model plant pathogen Pseudomonas syrin- available root entry sites. To overcome this problem, we gae analysing different moments of the disease took advantage of a luminescent R. solanacearum development have recently revealed a changing bac- reporter strain previously developed in our group to terial behaviour. For example, flagellar motility and measure bacterial colonisation and we normalized values chemotaxis-related genes were transcribed in the epi- for tissues containing comparable bacterial loads at dif- phytic phase, while genes controlling metabolism were ferent times of infection [48]. The in planta transcrip- expressed in the apoplast [43]. In another study, gene tomes were compared with that obtained from bacteria expression of virulent and avirulent P. syringae strains grown in liquid rich B medium, a reference condition was studied at different time points after inoculation known to repress many of the pathogen’s virulence of various Arabidopsis thaliana defence-related mu- determinants [49]. Principal component analysis (PCA) tants. This work identified an iron response regulator of the transcripts from each sample showed a clear clus- that was induced at early infection stages, counteract- tering of the biological replicates and a clear differenti- ing plant immunity [44]. Other time course transcrip- ation of the xylem samples from the reference and tomes in P. syringae have described an up-regulation apoplast samples (PC1, explaining 65% of the variation) of flagellar, chemotaxis and two-component system (Additional File 2 C). Comparison of the in planta tran- genes and a down-regulation of bacterial secretion scriptomes with that obtained in axenic growth in rich systems and general metabolism at late infection medium identified 418 differentially expressed genes stages in bacteria recovered from plants with pre- (DEGs) in the apoplast, 531 in the early xylem and 922 induced immunity compared to naïve plants [45]. To- in the late xylem (log2 fold change ≥ |1.5| and adjusted gether, these studies have started revealing the com- p-value ≤ 0.01). Of these genes, 226 and 192 were up- plex landscape of transcriptomic changes occurring and down-regulated, respectively, in the apoplast, 290 over time during the course of a bacterial infection. and 241 in the early xylem, and 378 and 544 in the late Due to the various environments it encounters xylem (Fig. 1a and Additional File 3). along the infection process and because of its eco- Comparison of the DEGs in each in planta condition nomic relevance, R. solanacearum is an excellent is in agreement with the previously published R. solana- model to analyse gene expression in different plant cearum in planta transcriptomic studies (Additional tissues, which correspond to distinct phases of the in- File 4 A). DE transcripts from the same UY031 strain fection process. Here, we have analysed the transcrip- retrieved from total RNAs of infected wild potato roots tome of the cold adapted R. solanacearum UY031 at [36] showed up to 17–18% overlap with the apoplast three different conditions. We have used the econom- condition and lower overlap with the other conditions ically important crop potato plant where the R. sola- assayed in the present study, and the gene expression nacearum UY031 was naturally identified a decade values showed a high correlation (Additional File 4 B). ago [46]. Our data clearly shows that R. solanacearum This is logical, since the transcriptome previously ob- genes behave dynamically inside the plant during the tained from roots of asymptomatic plants corresponds course of infection. We have identified condition- to an early time of infection where most bacteria grow specific expression of virulence and metabolic genes, apoplastically and only a small proportion of bacteria providing a new dynamic perspective of the R. solana- have already reached the xylem. The highest overlap cearum infection process. (34% overlap in up- and 36% in down-regulated genes, Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 4 of 18 Fig. 1 Transcriptomic profile of R. solanacearum UY031 in planta. a. Shared and unique DE genes across the three in planta conditions for the up-regulated (left) and down-regulated (right) genes. Each vertical bar plot represents the number of shared DE between the conditions indicated by the lines and dots in the schematic below. The horizontal bar plots on the right indicate the total DE genes per in planta condition compared to rich medium. b. For the intersection of Apoplast, Early and Late (in planta genetic programme), Early and Late (Xylem genetic programme), Apoplast and Late xylem alone, the list of genes was extracted and surveyed for enriched KEGG pathways. Dot plots of the enriched KEGG pathways for the up- (left) and down-regulated (right) genes in each environment are shown below. DE genes were identified with DEseq2 (p-adj > 0.01, log2 FC ± 1.5) and plotted using the R package UpsetR respectively) was found between our early xylem condi- shared in all in planta conditions: apoplast, early and tions and the microarray transcriptome of the phylo- late xylem), the xylem (genes shared in early and late genetically close strain UW551 isolated from tomato xylem), the apoplast, and the late xylem. Similarly, DE in plants at a comparable infection time (onset of wilting all in planta conditions were 104 up- and 81 down- symptoms) [14], which further validates our results regulated genes. The differentially expressed genes in the (Additional File 4). The overlap is obviously lower with xylem genetic programme (both time points analysed) comparable transcriptomes obtained using the distantly included a total of 162 and 156 up- and down-regulated related GMI1000 strain. genes. Finally, 100 and 80 genes were, respectively, up- To discover the DEGs common or unique to the dif- or down-regulated solely in the plant apoplast and 96 ferent plant environments, we analysed the shared genes and 278 only in the late xylem condition, when plants among the different conditions studied. As can be are mostly dead. The remaining conditions or overlaps observed in Fig. 1a, two intersections (i.e. in planta and between conditions included fewer than 30 specifically xylem) and two conditions (i.e. apoplast and late xylem) DEGs (Fig. 1a) and we did not consider them a proper that correspond to bacterial growth in precise environ- “genetic programme”. Overall, as hinted by the PCA ments included most of the DEGs. On this basis, we analysis, the apoplast showed the most divergent tran- defined four genetic programmes where R. solana- scriptome of the in planta conditions, whereas the sam- cearum expresses exclusive gene sets: in planta (genes ples extracted from the xylem (early and late) were the Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 5 of 18 most similar. However, a substantial fraction of genes containing both flagellar and type IV pili. Similarly, the was only differentially expressed in the late xylem (40% polygalacturonase gene pglA, encoding one out of the of those DE in this condition). six cell-wall degrading enzymes in the genome was also up-regulated in the plant. Other virulence genes up- R. solanacearum upregulates a variety of virulence factors regulated in bacteria growing in any of the studied in in planta planta conditions included efe, responsible for ethylene Functional enrichment of gene annotations is a powerful formation, the reactive oxygen species (ROS) scavenging tool to evaluate the genes involved in similar roles or superoxide dismutase sodC, and epsR, encoding the exo- pathways in each experimental condition. Thus, we polysaccharide (EPS) repressor. Finally, only the EPS investigated the enrichment of KEGG pathways and GO subcategory was under-represented in planta (p-value = terms in the genes that appeared DE in all in planta 1.25·10− 2), which can be explained by the high expres- conditions. Since the KEGG database contains metabolic sion of the exopolysaccharide synthesis operon in the pathways and terms specifically for prokaryotes, we reference rich medium [38]. ocused on its categories for enrichment analysis. Among the genes up-regulated in all in planta conditions, only Flagellar genes and the upstream regulators of the T3SS the KEGG flagellar assembly pathway was enriched are exclusively up-regulated in the apoplast (Fig. 1b). This result was confirmed by the GO Once R. solanacearum has infected the roots of a sus- enrichment analysis, where the bacterial flagellum- ceptible host plant it must cross the root cortex through dependent cell motility term was similarly over- the apoplast. The KEGG flagellar assembly pathway was represented, together with transposase activity and enriched in the genes exclusively up-regulated in the DNA-mediated transposition (Additional File 5). On apoplast (Fig. 1b). Similarly, the four GO terms referring the other hand, the enriched KEGG terms amongst to the flagellum (bacterial-type flagellum-dependent cell the genes down-regulated in all in planta conditions motility, bacterial-type flagellum basal body, bacterial- were all related to metabolism: inositol phosphate type flagellum and bacterial-type flagellum assembly) metabolism, and porphyrin and chlorophyll metabol- and phosphopantetheine binding were also enriched in ism (Fig. 1b), and the GO term cobalamin biosyn- this genetic programme (Additional File 5). A closer thetic process (Additional File 5). perusal of the list of up-regulated genes in the apoplast Manual curation of gene annotations enabled us to genetic programme also revealed that the “virulence and pinpoint a high number of pathogenicity-related func- parasitic fitness” category was enriched (p-value = tions up-regulated in all in planta conditions. These 4.2·10− 15). PrhJ and hrpG, key upstream regulators of genes had been overlooked by the global enrichment the T3SS activation cascade [31], were up-regulated in analysis because virulence genes are not in a KEGG this genetic programme. On the other hand, none of the pathway and pathogenicity-related terms in GO are too downstream T3SS transcriptional activators and only general and have not been widely used. Thus, we used two of 60 T3E genes (ripE2 and ripAD) were exclusively genomic and bibliographic information to create the up-regulated in this genetic programme. None of the gene category “virulence and parasitic fitness” for the KEGG pathways nor GO terms were enriched amongst UY031 strain and calculated its enrichment in all condi- the genes down-regulated in the apoplast. tions or genetic programmes analysed in this work (see Methods). The new category included all genes encoding R. solanacearum adapts to the xylem environment by the type III secretion system (T3SS) and its associated inducing virulence, chemotaxis and nitrogen metabolism effectors (T3Es), genes involved in motility, EPS and genes phytohormone biosynthesis, ROS scavenging, cell-wall After travelling through the root apoplast, R. solana- degrading enzymes, and nitrogen metabolism cearum crosses the Casparian strip, reaching the plant (Additional File 5). As expected, the created “virulence vasculature and heavily colonising the xylem vessels. As and parasitic fitness” category was clearly enriched in mentioned before, a substantial number of R. solana- the up-regulated genes in the in planta genetic cearum genes was DE in the xylem genetic programme, programme (p-value = 1.4·10− 14). Detailed analysis of the both at early and late conditions (Fig. 1a). Almost one subcategories included in “virulence and parasitic fit- third (12 out of 38) of the genes with associated KEGG ness” indicated that T3SS and T3Es (p-value = 2.4·10− 12) pathways differentially up-regulated in the xylem irre- and motility (p-value = 5.7·10− 5) were also significantly spective of the condition belonged to the enriched cat- enriched among the up-regulated genes. For instance, egory two-component system (Fig. 1b). This includes 20% (12 out of 60) of the genes annotated as T3Es were genes that participate in chemotaxis signal transduction, overexpressed in all in planta conditions. The enriched nitrate reduction, and oxidative phosphorylation. Three motility subcategory included a total of 11 genes, other categories were enriched in the genes up-regulated Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 6 of 18 in the xylem: oxidative phosphorylation (six genes), bac- categories translation, ribosome, structural constituent terial chemotaxis (five genes) and nitrogen metabolism of ribosome, RNA binding and tricarboxylic acid cycle (five genes). The up-regulated nitrogen metabolism (Additional File 5). In sum, R. solanacearum exclusively genes included nitrate transporters (nark1/2), enzymes downregulates at late infection stages in the xylem a involved in the denitrification pathway (aniA, norB) and large subset of genes involved in the central metabolism in the dissimilatory nitrate reduction pathway (narG/H/ and its derived metabolic pathways. I, nirB/D) as well as in reactive nitrogen species detoxifi- cation (hmpX). The enriched term bacterial chemotaxis Expression profiles reinforce the existence of specific included genes involved in different steps of swimming genetic programmes in the apoplast and the xylem motility, including membrane chemosensors, signal The findings described so far strongly suggest that R. transduction components (i.e. cheZ1, cheA, cheR) and solanacearum expresses specific sets of genes at each flagellar motor genes (i.e. motB). The “virulence and step of the infection process. To better understand this parasitic fitness” category was also enriched in the xylem dynamic process, we obtained the expression profiles of genetic programme up-regulated genes (p-value = the R. solanacearum UY031 genes in the three in planta 8.8·10− 5). Amongst these genes were 9 out of 60 T3Es conditions: apoplast, early and late xylem. To this end, annotated in strain UY031 genome (ripAE, ripY, ripAN, fold-change values of DE genes in each condition in re- ripC1, ripN, ripAP, ripF2, ripBH, and ripS5), and one lation to growth in rich culture medium were used as in- out of six cell wall degrading enzymes (pme). Other put to the Mfuzz clustering package. Six different gene overexpressed genes in the category included 10 motility expression profile clusters were identified according to genes and the cytokinin biosynthesis gene tzs. Finally, the condition or temporal progression, considering that amongst the 102 KEGG tagged down-regulated genes in the apoplast is the earliest stage during infection, the xylem, the enriched categories were: ribosome, oxi- followed by early and late xylem (Fig. 2, Additional File 7). dative phosphorylation and citrate (TCA) cycle (Fig. 1b). According to this, the profile named “specific apoplast” GO enrichment in down-regulated genes similarly contained 807 genes up-regulated in the apoplast but showed the over-represented categories translation, ribo- down-regulated in early and late xylem (Fig. 2a), and the some, structural constituent of ribosome, RNA binding, profile “specific xylem” contained 1286 genes down- rRNA binding (Additional File 5). In summary, a large regulated in the apoplast but up-regulated in the other set of R. solanacearum genes was found DE in the xylem conditions (Fig. 2b). We identified two additional pro- throughout infection, including up-regulation of nitro- files, including genes that continuously decreased (561 gen utilisation and virulence genes, such as T3Es and genes up-regulated in the apoplast with transcripts grad- down-regulation of genes encoding the citrate cycle en- ually decreasing in xylem) (Fig. 2c) or increased (334 zymes and the electron transport chain. genes, opposite profile) (Fig. 2d) their expression over the infection period. Finally, the genes specifically R. solanacearum inhibits a large number of metabolic repressed (Fig. 2e) or induced (Fig. 2f) in the early xylem pathways at late infection stages that showed the opposite trend in the apoplast and late Besides the DE genes in the xylem throughout infection, xylem were 105 and 107, respectively. a large set of R. solanacearum genes was exclusively DE To unveil the biological functions behind each expres- in the Late xylem genetic programme, at late stages of sion profile, we performed enrichment analyses. infection when plants are already wilted (Fig. 1a). Sur- Enriched KEGG pathways in the “specific apoplast” prisingly, no KEGG category was enriched in this abun- expression profile included various biosynthetic pro- dant set of up-regulated genes, but our “virulence and cesses, especially biosynthesis of secondary metabolites parasitic fitness” category was enriched in the up- (99 out of 308 tagged genes) and related pathways such regulated genes (p-value = 5·10− 3). Within this category, as biosynthesis of amino acids (53 genes) and flagellar two subcategories were also enriched: T3SS & T3Es, in- assembly (25 genes) (Fig. 2a). Our manually-defined mo- cluding six effectors, three of the GALA family (ripG3, tility subcategory was enriched in this expression profile ripG4 and ripG6) (p-value = 8.5·10− 3), and motility, with (p-value = 1.78·10− 2). In the “specific xylem” profile, the six involved in chemosensing and signal transduction (p- KEGG enrichment analysis yielded terms related with value = 3.68·10− 2). In the genes differentially down- metabolism adaptation such as microbial metabolism in regulated in the late xylem condition, five KEGG cat- diverse environments (106 out of 411 tagged genes), egories were enriched: carbon metabolism (18 out of ABC transporters (63 genes), and nitrogen metabolism 108 genes tagged), ribosome (17 genes), TCA cycle (9 (19 genes) among others (Fig. 2b). Our manually-defined genes), RNA degradation (six genes) and protein export subcategories T3SS & T3Es (p-value = 5.2·10− 3), phyto- (six genes) (Fig. 1b). GO enrichment analysis also hormones (p-value = 2.5·10− 3) and nitrogen metabolism showed similar results with the overrepresented (p-value = 2·10− 6) were also significantly enriched in this Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 7 of 18 Fig. 2 Gene expression dynamics of R. solanacearum throughout infection. Six clusters were obtained through Mfuzz clustering of log2-fold- change data of the apoplast, early and late xylem conditions normalised to the reference rich liquid media. Clusters include the genes with a membership higher than 70% and consistently associated to the same cluster on at least 30 out of 40 iterations. Number of genes indicated above each graph. The list of genes associated to each cluster was extracted and surveyed for enriched KEGG pathways. Dot plots of the enriched KEGG pathways in each cluster are shown next to the cluster profile. KEGG enriched terms within the continuous normalised read counts in all assayed conditions, includ- decrease profile were linked to transcription and ing the reference condition in rich medium. This pro- carbohydrate metabolism such as ribosome (43 out of vided an unbiased view on the gene expression data 191 tagged genes) and carbon metabolism (24 genes) avoiding the effect of the reference condition in the (Fig. 2c). Finally, the profile containing genes with spe- DESeq analysis. Detailed observation of gene expression cific up-regulation in the early xylem, was enriched in values in heatmap representations for the T3SS (hrp and the ubiquinone and other terpenoid-quinone biosyn- hrc genes) and T3E (rip genes) reinforced the above- thesis pathway (3 out of 22 tagged genes). The subcat- described enrichment in various genetic programmes or egory T3SS & T3Es was significantly enriched in this conditions (Fig. 3, Additional File 9). Both the rip T3Es expression profile as well (p-value = 1.34·10− 2), contain- and the hrp/hrc genes displayed a very homogeneous ex- ing genes such as the master regulator hrpB, and three pression pattern with high expression levels in the xylem T3 effectors (Fig. 2f). GO enrichment analysis confirmed genetic programme (early and late) and low expression these results, showing over-represented categories with levels in the apoplast. The only exceptions among the ef- similar biological functions (Additional File 8). fectors were the two ripI genes, with low expression levels in all studied conditions, ripE2, with higher ex- R. solanacearum specifically activates different sets of pression in the apoplast, and a cluster of effector genes virulence factors in different plant environments (i.e. ripAD and ripD), showing high transcript levels in As described above, key virulence activities were induced all conditions (Fig. 3). Heatmap visualisation of the nor- in specific plant environments or at specific disease malised transcriptomic data also indicated that flagellar stages. To analyse in further detail the genes in this genes —essential for swimming motility— were highly “virulence and parasitic fitness” (Additional File 6) and expressed in all in planta conditions, but to a higher ex- its subcategories we graphically represented their tent in the apoplast (Fig. 4 top panel). This is in Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 8 of 18 Fig. 3 T3E gene expression profile. Heatmaps showing the normalised transcripts per million (TPM) of the genes coding the 60 T3E described in R. solanacearum UY031 in the reference condition and in planta apoplast, early and late conditions. Only putatively functional T3E genes are included according to Peeters et al., 2013 accordance with the enrichment of this category in bottom panel), suggesting that the bacterium is using planta and in the late xylem genetic programmes up- the pilus appendix in all assayed plant environments. Ex- regulated genes, as well as in the specific apoplast pro- ceptions to this trend were the flagellar genes (i.e. fliM, file. The pil twitching motility genes encoding type IV fliS, fliD, fliT, motA, motB, fliC, fliO) and the type IV pi- pili followed a similar trend, although their expression lus genes (e.g. pilE1, pilY1, pilW, pilV, pilX), which were was more similar in the apoplast and the xylem (Fig. 4 down-regulated in the apoplast compared to the xylem Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 9 of 18 Fig. 4 Motility gene expression profile. Heatmaps showing the normalised transcripts per million (TPM) of the genes involved in motility in the reference (rich Medium), apoplast, early and late xylem conditions. Two heatmaps are shown to differentiate the swimming (top panel) and twitching (bottom panel) Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 10 of 18 genetic programme. The genes encoding chemotactic the infection. It should be noted that our gene expres- sensors and chemotaxis signal transduction proteins sion experiments in planta were all performed at com- showed low expression levels in the apoplast and parable bacterial loads. The reason for this is that R. progressive induction in the early and late xylem condi- solanacearum forms microcolonies and biofilms at early tions (Additional File 10) in accordance with the enrich- infection stages in the apoplast [50], so that neither the ment of these specific genes in the late xylem genetic effective local bacterial concentrations nor if density- programme. Finally, all the UY031 genes that synthetize dependent regulatory circuits are already induced at the plant hormones ethylene (efe), cytokinin (tzs), and these early stages are known. Consequently, although auxin (RSUY_RS1835 to RSUY_RS18970) [33] were our results perfectly reflect R. solanacearum adaptation highly expressed in the xylem genetic programme and to different plant environments, the influence that bac- to a lower extent in the apoplast, efe and tsz display- terial cell densities have on their gene expression during ing a more sustained expression in the apoplast. This disease progression is not reflected in our results. stable expression in all in planta conditions was also observed in the differential expression analysis T3Es expression is prevalent throughout the in planta (Additional File 12). infection process, especially in the xylem Here, we carefully investigated the expression pattern of Discussion the most important virulence factors to elucidate the R. solanacearum gene expression displays a behavioural bacterial strategies used to rewire the plant environ- differentiation into four plant genetic programmes that ments to its own benefit. The T3SS is the main patho- develop over time during in planta infection genicity determinant in R. solanacearum, as hrp mutants Previous R. solanacearum transcriptomic studies com- are completely avirulent [51]. The T3SS is tightly regu- pared gene expression profiles obtained using a specific lated by a transcriptional regulatory cascade that con- in planta condition, such as root apoplast [36] or early tains the constitutive receptor and transducer elements xylem colonisation [14], to reference bacteria grown in PrhA and PrhR and the transcriptional regulators PrhJ, rich medium. In our study, we analysed the whole infec- HrpG and HrpB [31]. Interestingly, in this work we tion process, including three different in planta condi- found that this cascade appears sequentially induced tions: apoplast, early and late xylem, which typify during infection. As depicted in Fig. 5, prhI and prhJ are paradigmatic stages of infection. Intersection of the exclusively induced in the apoplast, hrpG expression also DEGs of each of the three in planta experimental condi- peaks in this environment but is sustained at lower levels tions showed that most of the DEGs of R. solanacearum in the early xylem and hrpB is expressed in the apoplast during the infection are grouped in four biologically but highly induced in the early xylem, preceding the ex- relevant genetic programmes: genes commonly DE in all pression of the T3SS and most T3E, which is maximal at in planta conditions, genes exclusively DE in the apo- all xylem stages (Additional File 9). plast, genes expressed in the xylem at any stage of the Our gene expression dataset also shows that most of disease and genes exclusively DE in the xylem when the 60 T3Es are highly induced in the xylem genetic plants are already wilted (Fig. 1). One of the previous programme, confirming our previous results [34] that transcriptomic studies sampled bacteria from plants 5 challenged the view of T3Es as key only early after infec- days post-inoculation [14], similar to our early xylem tion [32, 40]. In agreement our finding that almost all condition. With the addition of our novel late xylem T3Es are simultaneously expressed in the xylem, a re- condition 10 days after inoculation, and the apoplast cently published study showed that deletion of 42 R. condition, we provide a more detailed expression land- solanacearum T3E genes was required to compromise scape of R. solanacearum, encompassing important dif- virulence of the bacteria on tobacco and eggplant and ferent stages of the infection process. To study the proliferation inside the xylem [52]. transcriptomic data from a tissue-specific perspective, Interestingly, all R. solanacearum T3Es belonging to we clustered the DEGs based on their expression profile gene families (PopA/B/C, AWR2/3/4/5_1/5_2, SKWP1/ across the three in planta experimental conditions (i.e. 2/3/5/7, HLK1/2/3 and GALA2/3/4/5/6/7) [53] were apoplast, early and late xylem) (Fig. 2). Reinforcing the clearly induced in the xylem throughout infection concept of a specific behaviour of R. solanacearum in (Fig. 3). The GALA effectors (e.g. ripG2 to ripG7) and different genetic programmes, the largest number of the AWR effectors (e.g. ripA2 to RipA5) have both DEGs appeared exclusively up-regulated in the xylem been shown to be collectively required for full bacter- genetic programme (1286 genes) and the apoplast (807 ial virulence and to target the proteasome or inhibit genes). This finding confirms that R. solanacearum has the Target Of Rapamycin (TOR) pathway, respectively different sets of genes that are deployed to infect the [54–56]. Their expression pattern suggests that these plant and adapt to the environments encountered along biochemical activities are likely carried out in the Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 11 of 18 Fig. 5 T3SS regulatory cascade. Log2 fold-change expression profile in the Apoplast, Early and Late Xylem of the genes involved in the T3SS regulatory cascade and downstream activated genes. Log2 fold changes in transcript levels with respect to the control condition (axenic growth in rich medium) are indicated in the boxes (left to right: Apoplast, Early and Late Xylem conditions) in colour gradients according to the Colour Key xylem. Similarly, The T3E ripAY, which has been GABA in plant cells in the absence of sufficient bac- proven to impair the redox status of the plant cell teria to consume it has been shown to induce cell degrading glutathione through its gamma-glutamyl death [62]. Therefore, we hypothesize that RipI ex- cyclotransferase activity [57–60], was clearly induced pression inside the plant must be tightly regulated to in the xylem (Fig. 3). This points to the xylem as a induce the production of nutrients without triggering stressing redox environment R. solanacearum must plant stress signals. cope with. In contrast, a few T3Es showed alternative induc- R. solanacearum modulates twitching and swimming in tion patterns to the one described above. For in- different plant environments stance, ripE2 can be clearly classified as an “early R. solanacearum uses two types of motility during the effector” since it was highly induced in the apoplast colonisation of plant tissues: swimming [26] and twitch- compared to the other conditions, while ripD and ing [27]. Swimming motility is an individual bacterial ripAD were highly induced in all in planta conditions movement through liquid environments in which flagella (Fig. 3). RipD localizes in vesicle-like structures and rotate by a proton-driven motor that is directed by che- blocks the flg22-induced ROS response in Nicotiana mosensor proteins [63]. Previous research showed that benthamiana [61]. This fact, linked with its high ex- both flagella (fliC) and chemosensor (cheA and cheW) pression in the apoplast and the activation of flagellar mutants were less virulent than the wild-type R. solana- genes in this condition, suggests that R. solanacearum cearum, demonstrating that not just the flagellar move- counteracts flg22 plant defence responses from the ment but also the ability to direct it are essential for full first stages of infection onwards. On the contrary, virulence in planta [8]. Interestingly, full virulence was ripI, which was shown to enhance plant production restored when the chemotactic mutants were directly in- of gamma-aminobutyric acid (GABA), was lowly oculated in the plant stem, indicating that swimming expressed in all in planta conditions (Fig. 3) [62]. Al- motility is of crucial importance at the very early stages though GABA catabolization by R. solanacearum en- of infection [8]. In our data (Fig. 4), most of the hances its infection capacity, the overproduction of flagellar-encoding genes were highly induced in the Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 12 of 18 apoplast and, to a lower extent, in the early and late R. solanacearum specifically activates different nitrogen xylem, supporting the previously mentioned hypothesis. metabolism genes to thrive in the xylem A small subset of flagellar genes including the motor R. solanacearum encounters a hypoxic environment in (motA, motB), the flagellin subunit (fliC) and the fila- the plant xylem, which could limit its growth as the bac- ment cap (fliD) among others showed low expression in terium usually uses oxygen as the main terminal electron the apoplast, for which we have no plausible acceptor. However, the xylem contains an optimal con- explanation. centration of nitrate that R. solanacearum can use as ter- R. solanacearum displays twitching motility, which in- minal electron acceptor to maintain its growth rates in volves the extension and retraction of type IV pili to move this environment [29]. Our gene expression dataset on solid or viscous surfaces [64]. This motility is involved shows a faint induction of the nitrogen metabolism in in natural transformation, biofilm formation and virulence the apoplast, reaching its expression peak in the xylem [27]. Inactivation of the genes encoding the pilin protein (Fig. 6). When nitrate is available in the extracellular (pilA), the secretin involved in the pilus extrusion (pilQ) space, it diffuses the outer membrane and is imported to or the protein required for pilus retraction (pilT) reduced the cytoplasm by NarK1/2. Once nitrate enters the cyto- R. solanacearum virulence [65]. In our transcriptomic plasm, the nitrate reductase (NarG/H/I) converts it to data, twitching motility genes showed a similar expression nitrite and then to ammonia through nitrite reductase pattern than swimming motility, but they were less in- (NirB/D). We found both the transporters- and duced in the apoplast and their expression was often reductase-encoding genes induced in the xylem (Fig. 6), maintained in early and late xylem (Fig. 4). This empha- suggesting that both import and dissimilatory nitrate re- sizes the importance of twitching motility throughout duction are active in this compartment (Fig. 6). the plant infection process, as showed by the effect on Nitrite diffusing back to the periplasm allows R. solana- virulence of pil deletion mutants [27, 65, 66]. Finally, cearum to perform denitrification, first by reducing nitrite pilI, which encodes the type IV pili chemosensor protein, to nitric oxide via the nitrite reductase AniA and finally by was especially induced in the apoplast (Fig. 4 bottom reducing nitric oxide to nitrous oxide via the nitric oxide panel), in agreement with our recent findings that it reductase NorB. Expression of these denitrification path- is involved in virulence especially during the early way genes is also induced in the xylem (Fig. 6), suggesting infection stages [66]. that R. solanacearum has the ability to detoxify the Fig. 6 Nitrogen metabolism expression profile. Log2 fold-change expression profile in the Apoplast, Early and Late Xylem of the genes involved in the denitrification (aniA, norB), assimilatory (nasA), dissimilatory (narK1/2, narG, nirB/D) and nitrite detoxification (hmpX) pathways of nitrogen in R. solanacearum UY031. Log2 fold changes in transcript levels with respect to the control condition (axenic growth in rich medium) are indicated in the boxes (left to right: Apoplast, Early and Late Xylem conditions) in colour gradients according to the Colour Key. O. M = Outer membrane, I.M. = Inner membrane Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 13 of 18 reactive nitrogen species produced during nitrate dissimi- solanacearum has the highest expression of motility latory pathway in the anaerobic xylem vessels [29]. genes in the apoplast, while the majority of T3Es and Moreover, R. solanacearum can also incorporate nitro- nitrogen metabolism genes are highly induced in the gen to its central metabolism through the assimilatory xylem environment. This study provides for the first nitrate reduction. The nitrate present in the cytoplasm is time a dynamic gene expression landscape of the reduced to nitrite by NasA/B. A previous study showed bacterial plant pathogen R. solanacearum and is a that nitrate assimilation was essential for initial root first step towards the transcriptomic characterisation attachment but was dispensable for growth, virulence, of its complete infection cycle. and competitive fitness [28]. The fact that nasA is induced in the xylem and not in the apoplast is in dis- Methods agreement with these results and may indicate strain- or Bacterial strains and plant growth conditions condition-specific roles of N genes in R. solanacearum. The highly aggressive Ralstonia solanacearum strain Finally, the nitric oxide anion in the cytoplasm can be UY031 (phylotype IIB, sequevar 1) isolated from potato detoxified using HmpX, whose expression is also highly tubers in Uruguay [46] carrying the synthetic luxCDABE induced in the xylem genetic programme (Fig. 6), an operon under the control of the psbA promoter was indicator of a highly active N metabolism in this plant used in this study [34]. The luminescence allowed indir- environment. ect but precise quantification of bacteria and to track bacterial proliferation in planta [48]. Bacteria were rou- Phytohormone and ROS scavenging enzymes are tinely grown at 30 °C in rich B medium supplemented expressed along the infection with 0.5% glucose [34]. R. solanacearum genome codes for phytohormone bio- Solanum tuberosum cv. Desirée potato plants were synthetic genes that drive the production of auxin [33], propagated in vitro [36] and 2-week old apexes were cytokinin [67] and ethylene [68]. Interestingly, transferred to a soil:silica sand mixture in a 1:1 ratio for bacterially-produced auxin was described to block plant RNA-seq sampling or moved to a substrate:perlite:ver- defences against the plant pathogen Pseudomonas syrin- miculite mixture in a 30:1:1 ratio for in planta visualisa- gae pv savastanoi [69] and ethylene was involved in tion. Plants were grown at 22 °C under long day (16 h / wilting development in the pathosystem A. thaliana-R. 8 h light/dark) conditions for 3 weeks. solanacearum [70]. In this study, we observed induction of the cytokinin (tzs) and the ethylene (efe) biosynthetic Bacterial sampling genes as well as the auxin operon in the xylem For liquid medium samples, bacterial cultures were set (Additional File 11). Apoplastic induction of the master to an starting OD600 = 0.1 (108 CFU/ml) and grown for regulator hrpG, which also controls auxin and ethylene 5 h in rich B medium (10 g/L bacteriological peptone, 1 synthesis genes [33], precedes the xylematic expression g/L yeast extract, 1 g/L casamino acids), until they of phytohormone biosynthesis genes as was observed for reached exponential growth phase (OD600 ~ 0.4–0.5). the T3SS (Fig. 5). Bacteria were then centrifuged at 4 °C for 2 min at After pathogen infection, plant cells respond with ROS maximum speed and the pellet was immediately frozen production to create a hostile environment against the bac- in liquid nitrogen. terium [71]. Interestingly, R. solanacearum contains several To assess bacterial colonisation levels, especially in genes that code for ROS scavenging enzymes, helping the asymptomatic plants, stems were placed under a lumin- bacterium survive in the plant apoplast and xylem [72]. ometer to visualize bacterial densities within the vascular Amongst them, alkyl hydroperoxide reductase genes system, and only plants showing luminescence were used. (ahpC1/C2/D/F) were mostly induced in the xylem (Add- To avoid bias of quorum sensing signals in the xylem itional File 12). Several studies have linked the induction of stages and not in the apoplast, similar bacterial yields were ahp genes in biofilm-forming cells in different bacterial path- infiltrated in potato leaves for the initial stage. Finally, to ogens, contributing to protection against oxidative stress, epi- identify the best time point at which bacterial colonisation phytic survival and attachment in the intercellular spaces or within xylem vessels of almost asymptomatic plants was to the xylem vessels [73–77]. most similar to that in dead plants, we monitored bacterial growth, luminescence and disease symptoms over time Conclusion (Additional File 1 A). As shown in Additional File 2 A, In summary, we performed a transcriptomic analysis of bacterial densities recovered from the three in planta R. solanacearum at different conditions in potato plants. conditions were in the same order of magnitude (be- DEG analysis revealed that R. solanacearum deploys in- tween 107 and 108 CFUs/ml). The in vitro reference side the plant host four different genetic programmes. condition corresponding to bacteria grown in liquid Functional enrichment analysis showed that R. rich medium, was also obtained to better define R. Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 14 of 18 solanacearum gene expression. We ensured that the RNA extraction, sequencing and library preparation difference of the final bacterial yields from the Total RNA was extracted using the SV Total RNA different conditions was not higher than one log Isolation System kit (Promega) following manufacturer’s (Additional File 2 A). These conditions allowed us to instructions for Gram-negative Bacteria. RNA concen- obtain enough R. solanacearum RNA-seq reads to tration was measured with a ND-8000 Nanodrop and have a robust representation of the whole genome RNA integrity was validated for all samples using the (Additional File 13). Principal component analysis Agilent 2100 Bioanalyzer. For rRNA depletion, 2.5 μg of revealed that these conditions are consistent among total RNA were treated with the Ribo-zero (TM) mag- biological replicates and sensitive enough to detect netic kit for bacteria (Epicentre). Three biological repli- biological differences between conditions (Additional cates per condition were subjected to sequencing on a File 2 C). HiSeq2000 Illumina System apparatus using multiplex- To obtain more reproducible samples, leaf apoplast ing and kits specially adapted to obtain 100 bp paired- was used as a mimic condition of root apoplast, since it end reads in stranded libraries. Rich media reference has been reported that R. solanacearum behaves simi- samples were sequenced by Macrogen Inc. In all other larly in these two apoplastic spaces [47]. cases, RNA-sequencing was performed in the Shanghai To obtain leaf apoplast samples, bacterial cells from an PSC Genomics facility. Raw sequencing data will be overnight culture were washed with water and resus- available upon publication in the Sequence Read Archive pended to a final concentration of 5 × 108 CFU/ml. The under Bio Project: PRJNA660623 (accession codes aerial part of the plants was vacuum-infiltrated for 30 s SAMN15955133 to SAMN15955144). to 1 min and the leaves were dried in paper towel before incubating the plants in the inoculation chamber (27 °C, Read alignment, mapping and differential gene 12 h / 12 h). After 6 h, leaves were vacuum-infiltrated expression analysis with sterile distilled water, dried in paper towel, rolled in RNA-seq raw data quality was evaluated using FAST a cut tip and centrifuged inside a 50 ml tube at 4 °C for QC (version 0.11.4, [78]). R. solanacearum reads were 5 min at 2000 rpm. Apoplast fluid extract was pooled mapped using Bowtie2 (version 2.3.3, [79]) with strin- (each pool representing approximately 15 plants) and gent parameters [36] using as reference the com- centrifuged at 4 °C at maximum speed for 2 min. Bacter- pletely sequenced genome of UY031 strain [80]. ial pellets were frozen in liquid nitrogen. Alignment files were quantified with HTSeq-count For early and late xylem samples, potato roots were (version 0.11.3, [81]) using NCBI’s RefSeq sequences injured with a 1 ml tip before inoculation. A total of 40 NZ_CP012687.1 (chromosome) and NZ_CP012688.1 ml of a 108 CFU/ml R. solanacearum suspension was (megaplasmid). The DESeq2 package (version 1.28.1, used to soil-inoculate each plant. After inoculation, [82]) in R ( [83], ver. 3.6.3) was employed to perform plants were kept inside the inoculation chamber (27 °C, differential expression (DE) analysis of high quality 12 h / 12 h) for 6 days (mean disease index = 0–1) for RNAseq reads. Genes with |log2(fold-change) | > 1.5 early xylem condition, or 10 days (disease index = 4 in all and adjusted p-value < 0.01 were considered as DE in the plants) for late xylem condition. Plants were photo- planta when compared to bacteria grown on liquid graphed in a Fuji Film LAS4000 light imager system to rich medium as reference condition (Additional File 2 check individual infection levels and only plants showing C). The results of the DeSeq2 analysis is shown in luminescence were used. Stem pieces of 2 cm were cut Additional File 3. The UpSetR [84] R package was from each plant, placed in a 1.5 ml tube containing used to visualise the intersection of DE genes in the 500 μl of sterile distilled water and centrifuged 2 min at different in planta conditions. For gene expression maximum speed at 4 °C to release bacteria from the comparison, gene counts were also normalised to xylem vessels. In all cases, bacterial densities were mea- transcripts per million (TPM) (Additional File 14). sured by luminescence before freezing and dilutions were plated to measure CFUs before addition of 5% of Gene expression pattern clustering and enrichment an ice-cold transcriptional stop solution (5% [vol/vol] analysis water saturated phenol in ethanol). This enabled nor- To obtain expression profiles of R. solanacearum UY031 malisation of early or late xylem samples for bacterial genes, a soft clustering analysis was performed using concentrations comparable to those of apoplast and ref- Mfuzz package (version 2.48, [85]) in R. Input data cor- erence medium samples. Bacterial pellets were pooled responds to the DE fold-change values yielded by together for each biological replicate and frozen in liquid DESeq2 of apoplast, early and late xylem samples nor- nitrogen. Approximately 30 plants were used for each malised to the reference liquid rich medium. The cluster early xylem replicate and 7 plants for every late xylem number was manually set at c = 6. To be more stringent, replicate (Additional File 2 A). a gene was considered to belong to a specific cluster if Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 15 of 18 the gene was allocated in the same cluster in 30 out of Supplementary Information 40 iterations with the membership value set to μ ≥ 0.75. The online version contains supplementary material available at https://doi. org/10.1186/s12864-021-07457-w. To further characterise the genes differentially expressed or belonging to any of the clusters, we looked Additional file 1: R. solanacearum reporter strains and bacterial growth for enriched Gene Ontology (GO) terms or Kyoto En- show equivalent infection rates. (A) Luminescence levels or bacterial cyclopaedia of Genes and Genomes (KEGG) pathways growth (bar plot) and symptom development (line plot) in potato plants were monitored over time to detect the precise time points at which among our genes. Since no GO terms had been previ- similar bacterial yields but different symptoms could be detected. The ously associated to UY031 strain genes, we used Blas- disease index scale (DI) ranges from 0 to 4 being 0 symptomless plants t2GO [86] software to annotate the UY031 genome. For and 4 plants completely wilted. Luminescence measurement were conducted on stem sections of infected plants. (B) GFP-labelled bacteria the KEGG and GO enrichment analysis, we used the en- were monitored at the sampled time points in potato plants. RLU = Rela- richer function of the ClusterProfiler package [87] in R tive light units. having previously created the TERM2GENE and Additional file 2: RNAseq experimental set-up and bioinformatic pipe- TERM2NAME lists to do the hypergeometric test. line. (A) Experimental set-up for the three in planta conditions, corre- sponding to an early (leaf apoplast), mid (xylem from asymptomatic Because KEGG enrichment analysis is limited to a plants) and late stages (xylem from dead plants) of the disease. As refer- number of pre-established pathways or terms that do ence condition, bacteria grown in rich liquid media were used. The aver- not include important virulence categories, and be- age of bacterial yields recovered in each condition are indicated as CFU/ ml. The grey background section of the figure contains the representa- cause pathogenicity-related terms in GO are too gen- tion of how bacteria was enriched in each condition (see M&Ms). (B) eral and have not been widely used, we decided to Transcriptomic analysis pipeline. (C) Two-dimensional Principal Compo- create a manually curated category that we defined as nent Analysis representation of the expression data of the conditions’ biological replicates used in the study. “virulence and parasitic fitness (Additional File 6). Additional file 3: DEGs in the three in planta conditions. Differentially This category included the T3SS and type III effec- expressed genes of R. solanacearum in apoplast, early and late xylem tors, motility genes, exopolysaccharides secretion, phy- compared to liquid rich medium obtained with DESeq2 (p-adj > 0.01, log2 tohormone biosynthesis, ROS scavenging, nitrogen FC ± 1.5). metabolism and cell-wall degrading enzymes. After Additional file 4: Overlap of DEGs in apoplast, early and late xylem compared to previous gene expression analysis. (A) Percentage of defining the genes included in this category, we con- common DE genes in each in planta condition (versus rich medium) ducted a hypergeometric test using the R stats pack- compared to previous in planta gene expression analyses (− Puigvert age on the differentially expressed genes or the gene et al. 2017; −Jacobs et al. 2012; −Khokhani et al. 2017). Fractions represent the overlapping genes from the total of DEGs in each of our clusters to find out whether the “virulence and para- conditions compared to a given previous gene expression analysis. sitic fitness” or any of the subcategories was Colors were plotted using the Conditional Formatting tool in Microsoft overrepresented. Excel. (B) Expression correlation of the DE data of the common genes between our Apoplast data and the RNAseq data from the potato root (Puigvert et al. 20,107). In planta visualisation of R. solanacearum Additional file 5: Transcriptomic profile of R. solanacearum in in planta To visualize R. solanacearum bacterial cells in early genetic programmes. Up-regulated (left) and down-regulated (right) (6 days post-inoculation, d.p.i) and late (10 d.p.i.) genes shared and unique across the three in planta conditions. Each ver- tical bar plot represents the number of shared DE between the condi- xylem stages, UY031 with the psbA constitutive pro- tions indicated by the lines and dots in the schematic below. The moter was fused to the GFP gene. This reporter horizontal bar plots on the right indicate the total DE genes per in planta strain was soil-inoculated with root wounding at condition compared to rich medium. For the intersection of Apoplast, Early and Late (in planta environment), Early and Late (Xylem environ- OD600 = 0.1 (108 CFUs/ml) in 3 week-old potato ment), Apoplast and Late xylem alone, the list of genes was extracted plants. Potato stem slices from the first node of in- and surveyed for enriched GO terms. Dot plots of the enriched GO terms fected plants with GFP-containing bacteria were ob- for the up- (left) and down-regulated (right) genes in each environment is shown below. DE genes were identified with DEseq2 (p-adj > 0.01, log2 served in the SZX16 stereomicroscope equipped with FC ± 1.5) and plotted using the R package UpsetR. a DP71 camera system (Olympus). Pictures were ob- Additional file 6: “Virulence and parasitic fitness” manually defined tained using the following settings: GFP filter, 10 s category. Genes belonging to specific virulence categories (T3SS & T3Es, exposure time, ISO 1/800. Control plants were soil- Motility, ROS scavenging enzymes, phytohormone biosynthesis, EPS, nitrogen metabolism, cell wall degrading enzymes) of R. solanacearum inoculated with water (Additional File 1 B). are listed showing information related to: UY031 NCBI locus tag (first column), gene name (second column), gene description (third column), Abbreviations category, (forth column), reference (fifth column). CFU: Colony forming units; Cv: Cultivar; d.p.i.: Days post inoculation; DEG: Differentially Expressed Genes; DI: Disease index; Additional file 7: List of genes included in each of the six expression DNA: desoxyribonucleic acid; EPS: exopolysaccharide; GABA: gamma- profiles. aminobutyric acid; GFP: Green Fluorescent Protein; GO: Gene Ontology; Additional file 8: Gene expression dynamics of R. solanacearum Hrp: Hypersensitive response and pathogenicity; IVET: In vivo expression throughout infection. Six clusters were obtained through Mfuzz clustering technology; KEGG: Kyoto Encyclopedia of Genes and Genomes; OD: Optical of log2- fold-change data of the apoplast, early and late xylem conditions Density; PCA: Principal component analysis; RLU: Relative Light Unit; normalised to the reference rich liquid media. Clusters include the genes RNA: Ribonucleic acid; ROS: Reactive Oxygen Species; rRNA: Ribosomal (number indicated above each graph) with a membership higher than ribonucleic acid; T3E: Type 3 Effector; T3SS: Type III Secretion System; 70% and consistently associated to the same cluster on at least 30 out of TCA: Tricarboxylic acid Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 16 of 18 40 iterations. The list of genes associated to each cluster was extracted datasets are available in Sequence Read Archive repository under Bio Project: and surveyed for enriched GO terms. Dot plots of the enriched GO terms PRJNA660623 (accession codes SAMN15955133 to SAMN15955144) in in each cluster is shown next to the cluster. https://www.ncbi.nlm.nih.gov/sra. Additional file 9: T3SS regulatory cascade and apparatus gene expression profile. Heatmap showing the normalised transcripts per Declarations million (TPM) of the genes involved in the T3SS regulatory cascade and Ethics approval and consent to participate the T3SS apparatus in the reference and in the in planta conditions. Not applicable. Additional file 10: hemosensors and signal transduction gene expression profile. Heatmap showing the normalised transcripts per Consent for publication million (TPM) of the genes involved in chemosensing and signal Not applicable. transduction in the reference and in the in planta conditions. Additional file 11: Phytohormones biosynthesis gene expression profile. Competing interests Heatmap showing the normalised transcripts per million (TPM) of the The authors declare that they have no competing interests. genes involved in phytohormones biosynthesis in the reference and in the in planta conditions. Author details 1 Additional file 12: ROS scavenging enzymes gene expression profile. Department of Genetics, University of Barcelona, Barcelona, Catalonia, Spain. 2 Heatmap showing the normalised transcripts per million (TPM) of the Centre for Research in Agricultural Genomics (CSIC-IRTA-UAB-UB), Bellaterra, genes coding for ROS scavenging enzymes in the reference condition Catalonia, Spain. 3Shanghai Center for Plant Stress Biology, CAS Center for and in planta apoplast, early and late condition. Excellence in Molecular Plant Sciences, Chinese Academy of Sciences, Shanghai 201602, China. 4Departamento de Bioquímica, Universidade de São Additional file 13: Proportion of reads aligned to R. solanacearum Paulo, São Paulo, Brazil. UY031 genome. Total number of reads obtained in each biological replicate for each condition (first column). Total number of reads aligned Received: 21 September 2020 Accepted: 19 February 2021 to R. solanacearum UY031 genome (second column). Proportion of reads aligned to R. solanacearum genome expressed as percentage (third column). References Additional file 14: Transcripts Per Million of each gene in rich medium, 1. Coll NS, Valls M. Current knowledge on the Ralstonia solanacearum type III apoplast, early and late xylem. Reads normalized per Transcripts Per secretion system. Microb Biotechnol [Internet]. 2013 Nov [cited 2020 Sep Million for each R. solanacearum gene in every condition: rich medium 18];6(6):614–20. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/ (philiq1, 2, 3), apoplast (Apo.10, .7, .9), early xylem (Early.D, .E, .G) and late PMC3815929/ xylem (Xylem.E, .O, Fresh.xylem). (CSV 986 kb) 2. Peeters N, Guidot A, Vailleau F, Valls M. Ralstonia solanacearum, a widespread bacterial plant pathogen in the post-genomic era. Mol Plant Pathol [Internet]. 2013 Sep 1 [cited 2020 Sep 18];14(7):651–62. Available Acknowledgements from: https://bsppjournals.onlinelibrary.wiley.com/doi/full/10.1111/mpp.1203 We thank C. Balsalobre and C. Madrid, from the Microbiology Section of the 8 University of Barcelona for their assistance in interpreting the RNAseq data 3. Hayward AC. Bacterial wilt caused by pseudomonas solanacearum. Annu and their interesting suggestions. Also, we would like to thank L. Tondo and Rev Phytopathol. 1991;29:65–87. E. Orellano from CONACYT Rosario for their useful advice, and CM Piroupo 4. Ciampi L, Sequeira L, French ER. Latent infection of potato tubers by for help in submitting the sequence datasets to the Short Read Archive. Pseudomonas solanacearum. Am Potato J. 1980;57(8):377–86. 5. Janse JD, Van Den Beld HE, Elphinstone J, Simpkins S, Tjou-Tam-Sin NNA, Authors’ contributions Van Vaerenbergh J. Introduction to Europe of Ralstonia solanacearum biovar MV and JCS designed the study. APM sequenced the bacterial samples. 2, race 3 in Pelargonium zonale cuttings. J Plant Pathol. 2004;86(2):147–55. RdeP, PS and MP performed the experimental work. RdeP, PS, MP, NSC, JSM, 6. Champoiseau PG, Jones JB, Allen C. Ralstonia solanacearum race 3 Biovar 2 JCS, APM and MV analysed the data. RdeP, PS, MP, and MV wrote the paper. causes tropical losses and temperate anxieties. Plant Heal Prog. 2009; All authors have read and approved the manuscript. 10(1):35. 7. Álvarez B, López MM, Biosca EG. Influence of native microbiota on survival Funding of Ralstonia solanacearum phylotype II in river water microcosms. Appl This work was funded by project PID2019-108595RB-I00 from the Spanish Environ Microbiol. 2007;73(22):7210–7. Ministry of Economy and Competitiveness. We also acknowledge financial 8. Yao J, Allen C. Chemotaxis is required for virulence and competitive fitness support from the “Severo Ochoa Program for Centres of Excellence in R&D” of the bacterial wilt pathogen Ralstonia solanacearum. J Bacteriol. 2006; 2016–2019 (SEV-2015-0533) and the CERCA Program of the Catalan Govern- 188(10):3697–708. ment (Generalitat de Catalunya). R. de Pedro-Jové received FI and FPU fel- 9. Vasse J, Frey P, Trigalet A. Microscopic studies of intercellular infection and lowships from Generalitat de Catalunya and the Spanish Ministry of protoxylem invasion of tomato roots by Pseudomonas solanacearum. Mol Economy and Competitiveness. P. Sebastià received the support of a fellow- Plant-Microbe Interact. 1995;8(2):241–51. ship (code is LCF/BQ/IN17/11620004) from la Caixa Foundation (identifier [ID] 10. Du Y, Stegmann M, Misas-Villamil JC. Meetings the apoplast as battleground 100010434). This project has received funding from the European Union’s for plant – microbe interactions; 2015. Horizon 2020 research and innovation program under Marie Skłodowska- 11. Hikichi Y, Yoshimochi T, Tsujimoto S, Shinohara R, Nakaho K, Kanda A, et al. Curie grant agreement no. 713673. M. Puigvert received an APIF doctoral fel- Global regulation of pathogenicity mechanism of Ralstonia solanacearum. lowship from Universitat de Barcelona and travel fellowship funded by Fun- Plant Biotechnol. 2007;24(1):149–54. dació Montcelimar and Universitat de Barcelona to carry out a short stay in 12. Planas-Marquès M, Bernardo-Faura M, Paulus J, Kaschani F, Kaiser M, Valls M, J.C. Setubal’s lab. J.C. Setubal holds a CNPq senior researcher fellowship. J.S. et al. Protease Activities Triggered by Ralstonia solanacearum Infection in Monteiro is funded by a FAPESP doctoral fellowship under process number Susceptible and Tolerant Tomato Lines. Mol Cell Proteomics [Internet]. 2018 2019/05287–2. A.P. Macho is funded by the Chinese Academy of Sciences Jun 1 [cited 2020 Sep 18];17(6):1112–25. Available from: https://pubmed. and the Chinese 1000 Talents Program. The funder had no role in study de- ncbi.nlm.nih.gov/29523767/ sign, data collection and analysis, decision to publish, or preparation of the 13. Zuluaga AP, Puigvert M, Valls M. Novel plant inputs influencing Ralstonia manuscript. solanacearum during infection. Front Microbiol [Internet]. 2013 Nov 20 [cited 2020 Sep 18];4(NOV):349. Available from: http://journal.frontiersin. Availability of data and materials org/article/10.3389/fmicb.2013.00349/abstract All data generated or analysed during this study are included in this 14. Jacobs JM, Babujee L, Meng F, Physiological C, Strategies V, Wilt B, et al. The published article and its supplementary information files and the supporting In Planta Transcriptome of Ralstonia solanacearum. Conserved. 2012;3(4):1–11. Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 17 of 18 15. Planas-Marquès M, Kressin JP, Kashyap A, Panthee DR, Louws FJ, Coll NS, 38. Clough SJ, Flavier AB, Schell MA, Denny TP. Differential expression of et al. Four bottlenecks restrict colonization and invasion by the pathogen virulence genes and motility in Ralstonia (Pseudomonas) solanacearum Ralstonia solanacearum in resistant tomato. J Exp Bot [Internet]. 2020 Mar 1 during exponential growth. Appl Environ Microbiol. 1997;63(3):844–50. [cited 2020 Sep 18];71(6):2157–71. Available from: https://academic.oup. 39. Huang J, Carney BF, Denny TP, Weissinger AK, Schell MA. A complex com/jxb/article/71/6/2157/5686149 network regulates expression of eps and other virulence genes of 16. Genin S. Molecular traits controlling host range and adaptation to plants in Pseudomonas solanacearum. J Bacteriol. 1995;177(5):1259–67. Ralstonia solanacearum. New Phytol. 2010;187(4):920–8. 40. Yoshimochi T, Hikichi Y, Kiba A, Ohnishi K. The global virulence regulator 17. Vasse J, Genin S, Frey P, Boucher C, Brito B. The hrpB and hrpG regulatory PhcA negatively controls the Ralstonia solanacearum hrp regulatory cascade genes of Ralstonia solanacearum are required for different stages of the by repressing expression of the PrhIR signaling proteins. J Bacteriol. 2009; tomato root infection process. Mol Plant-Microbe Interact. 2000; 191(10):3424–8. 18. Lu H, Lema SA, Planas-Marquès M, Alonso-Díaz A, Valls M, Coll NS. Type III 41. Meng F, Babujee L, Jacobs JM, Allen C. Comparative transcriptome analysis secretion-dependent and-independent phenotypes caused by Ralstonia reveals cool virulence factors of Ralstonia solanacearum race 3 biovar 2. solanacearum in Arabidopsis roots. Mol Plant-Microbe Interact. 2018;31(1): PLoS One. 2015;10:10. 175–84. 42. Khokhani D, Lowe-Power TM, Tran TM, Allen C. A single regulator mediates 19. Büttner D, He SY. Type III protein secretion in plant pathogenic bacteria. strategic switching between attachment/spread and growth/virulence in Plant Physiol. 2009;150(4):1656–64. the plant pathogen Ralstonia solanacearum. MBio. 2017;8:5. 20. Coburn B, Sekirov I, Finlay BB. Type III secretion systems and disease. Clin 43. Yu X, Lund SP, Scott RA, Greenwald JW, Records AH, Nettleton D, et al. Microbiol Rev. 2007;20(4):535–49. Transcriptional responses of Pseudomonas syringae to growth in epiphytic 21. Peeters N, Carrère S, Anisimova M, Plener L, Cazalé AC, Genin S. Repertoire, versus apoplastic leaf sites. Proc Natl Acad Sci U S A. 2013;110:5. unified nomenclature and evolution of the Type III effector gene set in the 44. Nobori T, Velásquez AC, Wu J, Kvitko BH, Kremer JM, Wang Y, et al. Ralstonia solanacearum species complex. BMC Genomics. 2013;14:1. Transcriptome landscape of a bacterial pathogen under plant immunity. 22. Prior P, et al. 1998. Molecular and Ecological Aspects: Bacterial Wilt Disease; Proc Natl Acad Sci U S A. 2018;115(13):E3055–64. 2013. 449 p. 45. Lovelace AH, Smith A, Kvitko BH. Pattern-triggered immunity alters the 23. Huang J, Denny TP, Schell MA. VsrB, a regulator of virulence genes of transcriptional regulation of virulence-associated genes and induces the Pseudomonas solanacearum, is homologous to sensors of the two- sulfur starvation response in pseudomonas syringae pv. Tomato DC3000. component regulator family. J Bacteriol. 1993;175(19):6169–78. Mol Plant-Microbe Interact. 2018;31(7):750–65. 24. Yaowei K, Jianzhong H, Mao G, Yuan HL, Schell MA. Dramatically reduced 46. Siri MI, Sanabria A, Pianzzola MJ. Genetic diversity and aggressiveness of virulence of mutants of Pseudomonas solanacearum defective in export of Ralstonia solanacearum strains causing bacterial wilt of potato in Uruguay. extracellular proteins across the outer membrane. Mol Plant-Microbe Plant Dis. 2011;95(10):1292–301. Interact. 1994;7(3):370–7. 47. Hikichi Y. Interactions between plant pathogenic bacteria and host plants 25. Liu H, Zhang S, Schell MA, Denny TP. Pyramiding unmarked deletions in during the establishment of susceptibility [Internet]. Vol. 82, Journal of General Ralstonia solanacearum shows that secreted proteins in addition to plant Plant Pathology. Springer Tokyo; 2016 [cited 2020 Aug 18]. p. 326–31. Available cell-wall-degrading enzymes contribute to virulence. Mol Plant-Microbe from: https://link.springer.com/article/10.1007/s10327-016-0680-9 Interact. 2005;18(12):1296–305. 48. Cruz APZ, Ferreira V, Pianzzola MJ, Siri MI, Coll NS, Valls M. A novel, sensitive 26. Tans-Kersten J, Brown D, Allen C. Swimming motility, a virulence trait of method to evaluate potato germplasm for bacterial wilt resistance using a Ralstonia solanacearum, is regulated by FlhDC and the plant host luminescent ralstonia solanacearum reporter strain. Mol Plant-Microbe environment. Mol Plant-Microbe Interact. 2004 Feb;17(6):686–95. Interact. 2014;27(3):277–85. 27. Kang Y, Liu H, Genin S, Schell MA, Denny TP. Ralstonia solanacearum 49. Arlat M, Gough CL, Zischek C, Barberis PA, Trigalet A, Boucher CA. Transcriptional requires type 4 pili to adhere to multiple surfaces and for natural organization and expression of the large hrp gene cluster of Pseudomonas transformation and virulence. Mol Microbiol. 2002 Oct;46(2):427–37. solanacearum. Molecular plant-microbe interactions : MPMI. 1992;5:187–93. 28. Dalsing BL, Allen C. Nitrate assimilation contributes to Ralstonia 50. Mori Y, Inoue K, Ikeda K, Nakayashiki H, Higashimoto C, Ohnishi K, et al. The solanacearum root attachment, stem colonization, and virulence. J Bacteriol. vascular plant-pathogenic bacterium Ralstonia solanacearum produces 2014;196(5):949–60. biofilms required for its virulence on the surfaces of tomato cells adjacent 29. Dalsing BL, Truchon AN, Gonzalez-Orta ET, Milling AS, Allen C. Ralstonia to intercellular spaces. Mol Plant Pathol. 2016. solanacearum uses inorganic nitrogen metabolism for virulence, ATP 51. Boucher CA, Barberis PA, Trigalet AP, Demery DA. Transposon mutagenesis production, and detoxification in the oxygen-limited host xylem of Pseudomonas solanacearum: isolation of Tn5-induced avirulent mutants. environment. MBio. 2015;6(2):1–13. J Gen Microbiol. 1985;131(9):2449–57. 30. Pegg GF. Life in a black hole — the micro-environment of the vascular 52. Lei N, Chen L, Kiba A, Hikichi Y, Zhang Y, Ohnishi K, et al. Super-Multiple pathogen. Trans Br Mycol Soc. 1985;85(1):IN1–20. Deletion Analysis of Type III Effectors in Ralstonia solanacearum OE1–1 for 31. Brito B, Marenda M, Barberis P, Boucher C, Genin S. PrhJ and hrpG, two new Full Virulence Toward Host Plants Bacterial Strains and Culture Conditions. components of the plant signal-dependent regulatory cascade controlled 2020;11(July):1–10. by PrhA in Ralstonia solanacearum. Mol Microbiol. 1999;31(1):237–51. 53. Poueymiro M, Genin S. Secreted proteins from Ralstonia solanacearum: a 32. Genin S, Brito B, Denny TP, Boucher C. Control of the Ralstonia hundred tricks to kill a plant. Curr Opin Microbiol. 2009;12(1):44–52. solanacearum type III secretion system (Hrp) genes by the global virulence 54. Remigi P, Anisimova M, Guidot A, Genin S, Peeters N. Functional regulator PhcA. FEBS Lett. 2005;579(10):2077–81. diversification of the GALA type III effector family contributes to Ralstonia 33. Valls M, Genin S, Boucher C. Integrated regulation of the type III secretion solanacearum adaptation on different plant hosts. New Phytol. 2011;192(4): system and other virulence determinants in Ralstonia solanacearum. PLoS 976–87. Pathog. 2006;2(8):0798–807. 55. Solé M, Popa C, Mith O, Sohn KH, Jones JDG, Deslandes L, et al. Type III 34. Monteiro F, Genin S, van Dijk I, Valls M. A luminescent reporter evidences effectors displaying virulence and Avirulence activities. Mol Plant-Microbe active expression of Ralstonia solanacearum type III secretion system genes Interact. 2012;25(7):941–53. throughout plant infection. Microbiol (United Kingdom). 2012;158(8):2107–16. 56. Popa C, Li L, Gil S, Tatjer L, Hashii K, Tabuchi M, et al. The effector AWR5 35. Ailloud F, Lowe TM, Robène I, Cruveiller S, Allen C, Prior P. In planta from the plant pathogen Ralstonia solanacearum is an inhibitor of the TOR comparative transcriptomics of host-adapted strains of Ralstonia signalling pathway. Sci Rep. 2016. solanacearum. Peer J. 2016;2016:1. 57. Sang Y, Wang Y, Ni H, Cazalé AC, She YM, Peeters N, et al. The ralstonia 36. Puigvert M, Guarischi-Sousa R, Zuluaga P, Coll NS, Macho AP, Setubal JC, solanacearum type III effector ripay targets plant redox regulators to et al. Transcriptomes of ralstonia solanacearum during root colonization of suppress immune responses. Mol Plant Pathol. 2018;19(1):129–42. solanum commersonii. Front Plant Sci. 2017;8. 58. Wei Y, Sang Y, Macho AP. The ralstonia solanacearum type III effector RipAY 37. Garg RP, Huang J, Yindeeyoungyeon W, Denny TP, Schell MA. is phosphorylated in plant cells to modulate its enzymatic activity. Front Multicomponent transcriptional regulation at the complex promoter of the Plant Sci. 2017;8(November):1–7. exopolysaccharide I biosynthetic operon of Ralstonia solanacearum. J 59. Mukaihara T, Hatanaka T, Nakano M, Oda K. Ralstonia solanacearum type III Bacteriol. 2000;182(23):6659–66. effector RipAY is a glutathione-degrading enzyme that is activated by plant Pedro-Jové et al. BMC Genomics (2021) 22:170 Page 18 of 18 cytosolic thioredoxins and suppresses plant immunity. MBio [Internet]. 2016 81. Anders S, Pyl PT, Huber W. HTSeq-A Python framework to work with high- Apr 12 [cited 2020 Aug 18];7(2). Available from: https://pubmed.ncbi.nlm. throughput sequencing data. Bioinformatics. 2015;31(2):166–9. nih.gov/27073091/ 82. Love MI, Huber W, Anders S. Moderated estimation of fold change and 60. Fujiwara S, Kawazoe T, Ohnishi K, Kitagawa T, Popa C, Valls M, et al. RipAY, a dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):1–21. Plant Pathogen Effector Protein, Exhibits Robust γ-Glutamyl Cyclotransferase 83. R Team C. R Core Team (2017). R: a language and environment for statistical Activity When Stimulated by Eukaryotic Thioredoxins. J Biol Chem [Internet]. computing. R Found Stat Comput Vienna, Austria URL http//www R-project 2016 Mar 25 [cited 2020 Aug 18];291(13):6813–30. Available from: https:// org/, page R Found Stat Comput 2017; pubmed.ncbi.nlm.nih.gov/26823466/ 84. Conway JR, Lex A, Gehlenborg N. UpSetR: an R package for the visualization 61. Jeon H, Kim W, Kim B, Lee S, Jayaraman J, Jung G, et al. Erratum: Ralstonia of intersecting sets and their properties. Bioinformatics. 2017;33(18):2938–40. solanacearum Type III effectors with predicted nuclear localization signal 85. Kumar L, Futschik ME. Mfuzz: A software package for soft clustering of localize to various cell compartments and modulate immune responses in microarray data. Bioinformation. 2007. Nicotiana spp. (Plant Pathol. J., (2020) 36(1), (43–53), https://doi.org/10.5423/ 86. Götz S, García-Gómez JM, Terol J, Williams TD, Nagaraj SH, Nueda MJ, et al. PPJ.OA.08.2019.0227). Plant Pathol J. 2020;36(3):303. High-throughput functional annotation and data mining with the Blast2GO 62. Xian L, Yu G, Wei Y, Rufian JS, Li Y, Zhuang H, et al. A Bacterial Effector suite. Nucleic Acids Res. 2008;36(10):3420–35. Protein Hijacks Plant Metabolism to Support Pathogen Nutrition. Cell Host 87. Yu G, Wang LG, Han Y, He QY. ClusterProfiler: an R package for Microbe [Internet]. 2020 Jul 30 [cited 2020 Aug 18];1–10. Available from: comparing biological themes among gene clusters. Omi A J Integr Biol. https://doi.org/10.1016/j.chom.2020.07.003. 2012;16(5):284–7. 63. Sampedro I, Parales RE, Krell T, Hill JE. Pseudomonas chemotaxis. FEMS Microbiol Rev. 2015;39(1):17–46. 64. Mattick JS. Type IV pili and twitching motility. Annu Rev Microbiol 2002;56: Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in 289–314. published maps and institutional affiliations. 65. Liu H, Kang Y, Genin S, Schell MA, Denny TP. Twitching motility of Ralstonia solanacearum requires a type IV pilus system. Microbiology. 2001;147(12): 3215–29. 66. Corral J, Sebastià P, Coll NS, Barbé J, Aranda J, Valls M. Twitching and Swimming Motility Play a Role in Ralstonia solanacearum Pathogenicity. mSphere. 2020;5(2):1–16. 67. Genin S, Denny TP. Pathogenomics of the ralstonia solanacearum species complex. Annu Rev Phytopathol. 2012;50:67–89. 68. Freebairn HT, Buddenhagen IW. Ethylene production by Pseudomonas solanacearum. Nature. 1964;313(314). 69. Robinette D, Matthysse AG. Inhibition by agrobacterium tumefaciens and Pseudomonas savastanoi of development of the hypersensitive response elicited by Pseudomonas syringae pv. Phaseolicola. J Bacteriol. 1990;172(10): 5742–9. 70. Hirsch J, Deslandes L, Feng DX, Balagué C, Marco Y. Delayed symptom development in ein2-1, an Arabidopsis ethylene-insensitive mutant, in response to bacterial wilt caused by Ralstonia solanacearum. Phytopathology. 2002;92(10):1142–8. 71. Bolwell GP. The apoplastic oxidative burst in response to biotic stress in plants: a three-component system. J Exp Bot. 2002;53(372):1367–76. 72. Flores-Cruz Z, Allen C. Ralstonia solanacearum encounters an oxidative environment during tomato infection. Mol Plant-Microbe Interact. 2009; 22(7):773–82. 73. Büttner D, Bonas U. Regulation and secretion of Xanthomonas virulence factors. FEMS Microbiol Rev. 2010;34(2):107–33. 74. Jang IA, Kim J, Park W. Endogenous hydrogen peroxide increases biofilm formation by inducing exopolysaccharide production in Acinetobacter oleivorans DR1. Sci Rep. 2016;6(February):1–12. 75. Oh E, Jeon B. Role of alkyl hydroperoxide reductase (AhpC) in the biofilm formation of Campylobacter jejuni. PLoS One. 2014;9:1. 76. Panmanee W, Hassett DJ. Differential roles of OxyR-controlled antioxidant enzymes alkyl hydroperoxide reductase (AhpCF) and catalase (KatB) in the protection of Pseudomonas aeruginosa against hydrogen peroxide in biofilm vs. planktonic culture. FEMS Microbiol Lett. 2009;295(2):238–44. 77. Wasim M, Bible AN, Xie Z, Alexandre G. Alkyl hydroperoxide reductase has a role in oxidative stress resistance and in modulating changes in cell- surface properties in Azospirillum brasilense Sp245. Microbiology. 2009; 155(4):1192–202. 78. Andrews S, Krueger F, Seconds-Pichon A, Biggins F, Wingett S. FastQC. A quality control tool for high throughput sequence data. Babraham Bioinformatics [Internet]. Vol. 1, Babraham Institute. 2015. p. 1. Available from: https://www.bioinformatics.babraham.ac.uk/projects/fastqc/, http:// www.bioinformatics.bbsrc.ac.uk/projects/fastqc/ 79. Langmead B, Bowtie SS. Nat Methods [Internet]. 2013;9(4):357–9 Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3322381/pdf/nihms-3 66740.pdf. 80. Guarischi-Sousa R, Puigvert M, Coll NS, Siri MI, Pianzzola MJ, Valls M, et al. Complete genome sequence of the potato pathogen Ralstonia solanacearum UY031. Stand Genomic Sci 2016;11(1):1–8.