The Knockout of Enterobactin-Related Gene in Pectobacterium atrosepticum Results in Reduced Stress Resistance and Virulence towards the Primed Plants

Siderophores produced by microorganisms to scavenge iron from the environment have been shown to contribute to virulence and/or stress resistance of some plant pathogenic bacteria. Phytopathogenic bacteria of Pectobacterium genus possess genes for the synthesis of siderophore enterobactin, which role in plant-pathogen interactions has not been elucidated. In the present study we characterized the phenotype of the mutant strain of Pba deficient for the enterobactin-biosynthetic gene entA. We showed that enterobactin may be considered as a conditionally beneficial virulence factor of Pba. The entA knockout did not reduce Pba virulence on non-primed plants; however, salicylic acid-primed plants were more resistant to ΔentA mutant than to the wild type Pba. The reduced virulence of ΔentA mutant towards the primed plants is likely explained by its compromised resistance to oxidative stress.

Siderophores are small metabolites (usually less than 1 kDa) that scavenge iron from mineral and organic substrates due to high iron-affinity and deliver it to cells via specific receptors [13,14]. Iron is an essential nutrient for fundamental metabolic processes in living organisms. Owing to its capacity to undergo reversible changes in its oxidation state, iron is a privileged cofactor for many proteins mediating electron transfer and redox reactions. This metal is required for pathogens to manifest their virulence potential. Even though iron is the fourth most abundant element in the Earth's crust, it is poorly bioavailable at alkaline or neutral pH because of the low solubility of ferric hydroxides prevailing in aerobic soils [15,16]. Siderophores serve as iron carriers in bacteria and other organisms. Iron-free siderophores are exported from the microbial cell, while the ferric-siderophore complexes can be taken up via specific outer-membrane transporters associated with the TonB complex. Inside the microbial cell, iron within siderophore is The presented values are means ± SD of three biological replicates. The activity levels of the wild type were equated to one. No significant differences in the assessed parameters were revealed between the wild type and ΔentA mutant (Mann-Whitney two-sided test, p < 0.05).

Virulence of ΔentA Mutant
To check if entA gene knockout affected Pba virulence, tobacco plants were infected with the wild type or mutant strain. Almost 90% of plants infected by the wild type displayed maceration symptoms expressed in tissue softening and disintegration 3 days post inoculation. ΔentA mutant-infected plants showed similar symptoms, and only 10% less plants had macerated zones compared to plants infected by the wild type ( Figure 2A). Similar results were obtained on the infected potato plants (data not shown). The titer of bacterial cells in the diseased tobacco plants infected with either the wild type or ΔentA mutant differed only around two times that was not statistically significant ( Figure 2B). Since the marker of plant susceptibility to Pba is the expression level of jasmonic acid-regulated genes [10,35], the levels of gene transcripts of lipoxygenase (LOX2) and allene oxide cyclase (AOC) was compared in tobacco plants infected with either the wild type or ΔentA mutant, on which symptoms were equally manifested. LOX2 and AOC genes were induced in plants infected with the wild type as well as mutant strain compared to control non-infected plants ( Figure 2C). However, the level of induction was lower in the plants infected by the ΔentA mutant. Thus, the knockout of entA gene did not make significant contribution to virulence properties of Pba. However, the expression level of jasmonic acid-regulated genes, the markers of Pba-caused disease, was decreased in ΔentA mutant-infected plants compared to the wild type-infected ones. The enzymatic activities were determined in the culture supernatants after one day of bacteria cultivation. The presented values are means ± SD of three biological replicates. The activity levels of the wild type were equated to one. No significant differences in the assessed parameters were revealed between the wild type and ∆entA mutant (Mann-Whitney two-sided test, p < 0.05).

Virulence of ∆entA Mutant
To check if entA gene knockout affected Pba virulence, tobacco plants were infected with the wild type or mutant strain. Almost 90% of plants infected by the wild type displayed maceration symptoms expressed in tissue softening and disintegration 3 days post inoculation. ∆entA mutant-infected plants showed similar symptoms, and only 10% less plants had macerated zones compared to plants infected by the wild type ( Figure 2A). Similar results were obtained on the infected potato plants (data not shown). The titer of bacterial cells in the diseased tobacco plants infected with either the wild type or ∆entA mutant differed only around two times that was not statistically significant ( Figure 2B). Since the marker of plant susceptibility to Pba is the expression level of jasmonic acidregulated genes [10,35], the levels of gene transcripts of lipoxygenase (LOX2) and allene oxide cyclase (AOC) was compared in tobacco plants infected with either the wild type or ∆entA mutant, on which symptoms were equally manifested. LOX2 and AOC genes were induced in plants infected with the wild type as well as mutant strain compared to control non-infected plants ( Figure 2C). However, the level of induction was lower in the plants infected by the ∆entA mutant. Thus, the knockout of entA gene did not make significant contribution to virulence properties of Pba. However, the expression level of jasmonic acid-regulated genes, the markers of Pba-caused disease, was decreased in ∆entA mutant-infected plants compared to the wild type-infected ones. Bacterial colony forming unit (CFU) titer in the tobacco plants infected with the wild type (dark grey) or ΔentA mutant (light grey) of P. atrosepticum. (C) Expression levels of genes LOX (lipoxygenase) and AOC (allen oxide cyclase) in non-infected plants (hashed column) and infected with the wild type (dark grey) or ΔentA mutant (light grey) of P. atrosepticum. The expression levels and CFU titers were determined one day post inoculation. The presented values in B and C are means ± SD of five biological replicates. Asterisks (*) show the significance of difference (Mann-Whitney two-sided test, p < 0.05).

Stress Resistance of ΔentA Mutant
Since siderophores were shown to contribute not only to virulence, but also to resistance to some stressors (foremost oxidative stress and heavy metals) [14,34,[36][37][38][39], we compared the resistance of the wild type and ΔentA mutant as well as the complementation mutant carrying the lost entA gene within a recombinant plasmid. Herewith, the ability of bacterial strains to grow in iron-depleted medium (in the presence of 20 µM Na-EDTA) or in presence of hydrogen peroxide or CuSO4 was assessed. In contrast to the wild type and complementation mutant, ΔentA mutant was unable to grow in the presence of 20 µM Na-EDTA ( Figure 3A).
During 24 h of cultivation at 3 mM concentration of H2O2, the cell titer of the wild type and complementation mutant increased 7-8-fold, while that of the mutant strain decreased by more than 40-fold ( Figure 3B). At 1.5 mM of H2O2, all three strains were able to grow, but herewith, the cell titers in the wild type and complementation mutant cultures were 3-4-fold higher than that in the cultures of ΔentA mutant. At lower H2O2 concentrations, the differences between the wild type and mutant strains were insignificant. The CFU number decreased at 5 µM of CuSO4 in the cultures of all three strains compared to the inoculation titer, however, much greater reduction was observed for the ΔentA mutant strain compared to the wild type and complementation mutant ( Figure 3C). At 2.5 µM CuSO4, the CFU titer of the wild type and complementation mutant increased slightly compared to the inoculation titer, while the CFU titer of the mutant strain decreased more than 10-times. At lower CuSO4 concentrations, the differences between the wild type and mutant strains were insignificant. Thus, the knockout of entA gene results in the decreased stress resistance of Pba.

Stress Resistance of ∆entA Mutant
Since siderophores were shown to contribute not only to virulence, but also to resistance to some stressors (foremost oxidative stress and heavy metals) [14,34,[36][37][38][39], we compared the resistance of the wild type and ∆entA mutant as well as the complementation mutant carrying the lost entA gene within a recombinant plasmid. Herewith, the ability of bacterial strains to grow in iron-depleted medium (in the presence of 20 µM Na-EDTA) or in presence of hydrogen peroxide or CuSO 4 was assessed. In contrast to the wild type and complementation mutant, ∆entA mutant was unable to grow in the presence of 20 µM Na-EDTA ( Figure 3A).
During 24 h of cultivation at 3 mM concentration of H 2 O 2 , the cell titer of the wild type and complementation mutant increased 7-8-fold, while that of the mutant strain decreased by more than 40-fold ( Figure 3B). At 1.5 mM of H 2 O 2 , all three strains were able to grow, but herewith, the cell titers in the wild type and complementation mutant cultures were 3-4-fold higher than that in the cultures of ∆entA mutant. At lower H 2 O 2 concentrations, the differences between the wild type and mutant strains were insignificant. The CFU number decreased at 5 µM of CuSO 4 in the cultures of all three strains compared to the inoculation titer, however, much greater reduction was observed for the ∆entA mutant strain compared to the wild type and complementation mutant ( Figure 3C). At 2.5 µM CuSO 4 , the CFU titer of the wild type and complementation mutant increased slightly compared to the inoculation titer, while the CFU titer of the mutant strain decreased more than 10-times. At lower CuSO 4 concentrations, the differences between the wild type and mutant strains were insignificant. Thus, the knockout of entA gene results in the decreased stress resistance of Pba.

Virulence of ΔentA Mutant towards Salicylic Acid-Primed Plants
Even though the ΔentA mutant strain did not display reduced virulence compared to the wild type in the experiments described above, its increased stress susceptibility might impede the disease manifestation in the primed plants. To test this hypothesis, tobacco plants were pretreated with 0.2 mM salicylic acid (SA). This concentration was shown to be below the one (1 mM) that led to a significant reduction of the disease development caused by the wild type Pba (data not shown). Tobacco plants were mock-or SA-treated 24 h before the infection with either the wild type or ΔentA mutant. The number of plants with maceration symptoms was scored 1, 2, 4 and 6 days post inoculation. The number of the diseased plants in the mock-treated wild type-infected plant group (assign as a control group) was equated to 100% at each time point. The disease incidence rate in the SA-treated wild type-infected group and mock-treated ΔentA mutant-infected group did not differ significantly from the control group, but herewith, the disease incidence rate in the SA-treated ΔentA mutant-infected group was 2-3-times lower ( Figure 4). Thus, enterobactin deficiency impeded Pba to cause the disease on the SA-primed plants, but not on non-primed plants.

Virulence of ∆entA Mutant towards Salicylic Acid-Primed Plants
Even though the ∆entA mutant strain did not display reduced virulence compared to the wild type in the experiments described above, its increased stress susceptibility might impede the disease manifestation in the primed plants. To test this hypothesis, tobacco plants were pretreated with 0.2 mM salicylic acid (SA). This concentration was shown to be below the one (1 mM) that led to a significant reduction of the disease development caused by the wild type Pba (data not shown). Tobacco plants were mock-or SA-treated 24 h before the infection with either the wild type or ∆entA mutant. The number of plants with maceration symptoms was scored 1, 2, 4 and 6 days post inoculation. The number of the diseased plants in the mock-treated wild type-infected plant group (assign as a control group) was equated to 100% at each time point. The disease incidence rate in the SA-treated wild type-infected group and mock-treated ∆entA mutant-infected group did not differ significantly from the control group, but herewith, the disease incidence rate in the SA-treated ∆entA mutant-infected group was 2-3-times lower ( Figure 4). Thus, enterobactin deficiency impeded Pba to cause the disease on the SA-primed plants, but not on non-primed plants. The LOX2 gene expression (the marker of plant susceptibility to Pba) was induced in the wild type Pba-and ΔentA mutant-infected plants during the disease progression irrespective of whether the plants were treated with SA or not ( Figure 5). However, the level of induction was slightly (non-significantly) lower in the SA-treated plants compared to non-treated ones. Herewith, the LOX2 gene expression level was greater in the non-treated wild type Pba-infected plants compared to that in the non-treated ΔentA mutant-infected plants (as also described above). However, in the SA-treated plants, the expression level of this gene did not differ significantly between the plants infected with the wild type and ΔentA mutant ( Figure 5). The bacterial cell titer in planta also did not differ in the non-treated and SA-treated plants infected by the wild type Pba and ΔentA mutant (data not shown).  The LOX2 gene expression (the marker of plant susceptibility to Pba) was induced in the wild type Pbaand ∆entA mutant-infected plants during the disease progression irrespective of whether the plants were treated with SA or not ( Figure 5). However, the level of induction was slightly (non-significantly) lower in the SA-treated plants compared to non-treated ones. Herewith, the LOX2 gene expression level was greater in the non-treated wild type Pba-infected plants compared to that in the non-treated ∆entA mutant-infected plants (as also described above). However, in the SA-treated plants, the expression level of this gene did not differ significantly between the plants infected with the wild type and ∆entA mutant ( Figure 5). The bacterial cell titer in planta also did not differ in the non-treated and SA-treated plants infected by the wild type Pba and ∆entA mutant (data not shown).  The LOX2 gene expression (the marker of plant susceptibility to Pba) was induced in the wild type Pba-and ΔentA mutant-infected plants during the disease progression irrespective of whether the plants were treated with SA or not ( Figure 5). However, the level of induction was slightly (non-significantly) lower in the SA-treated plants compared to non-treated ones. Herewith, the LOX2 gene expression level was greater in the non-treated wild type Pba-infected plants compared to that in the non-treated ΔentA mutant-infected plants (as also described above). However, in the SA-treated plants, the expression level of this gene did not differ significantly between the plants infected with the wild type and ΔentA mutant ( Figure 5). The bacterial cell titer in planta also did not differ in the non-treated and SA-treated plants infected by the wild type Pba and ΔentA mutant (data not shown).  Then we analyzed whether the complementation of ∆entA mutation restored the virulence towards the SA-primed plants. In this case, 69 and 81% of non-primed plants displayed disease symptoms three days after inoculation with ∆entA mutant and the complementation mutant, respectively. Herewith, in 23% and 69% of the SA-primed plants, the disease symptoms were developed after infection with ∆entA mutant and the complementation mutant, respectively ( Figure 6). This means that the complementation of ∆entA mutation restored the virulence towards the SA-primed plants. Then we analyzed whether the complementation of ΔentA mutation restored the virulence towards the SA-primed plants. In this case, 69 and 81% of non-primed plants displayed disease symptoms three days after inoculation with ΔentA mutant and the complementation mutant, respectively. Herewith, in 23% and 69% of the SA-primed plants, the disease symptoms were developed after infection with ΔentA mutant and the complementation mutant, respectively ( Figure 6). This means that the complementation of ΔentA mutation restored the virulence towards the SA-primed plants.

H2O2-Level in Infected and Non-Infected Tobacco Plants Pretreated or Not with Salicylic Acid
To check whether the SA-priming of plants (resulting in the increased plant resistance to ΔentA mutant but not to the wild type Pba) influenced the level of ROS and whether the ROS accumulation was different in plants (SA-treated and non-treated) infected by the wild type Pba, or ΔentA mutant, or complementation mutant, the level of hydrogen peroxide was measured. Herewith, among the plants infected with each of the strains, plants that displayed disease symptoms and plants that remained symptomless after the infection were analyzed separately.
The SA-treatment resulted in the increased H2O2-level in control non-infected plants compared that in non-treated plants (Figure 7). Symptomatic infections in both the SA-treated and non-treated plants, irrespective of the strain that caused them (wild type Pba, or ΔentA mutant, or complementation mutant), were associated with the increased H2O2-level compared to that in non-treated non-infected plants; and herewith, the SA-treatment did not increase the H2O2-level in infected plants compared to that in non-treated infected plants.
In symptomless plants infected by the wild type Pba, the level of H2O2 was also increased compared to control; herewith, the H2O2 level did not differ in symptomless plants infected by different strains. During symptomatic infection caused by ΔentA mutant, the level of H2O2 was higher than that in symptomless plants infected by the same strain. Mean values of H2O2-level during symptomatic infections caused by two other strains (wild type Pba and complementation mutant) were also higher than during the asymptomatic ones although these differences were not statistically significant (Figure 7). During symptomatic infections caused by all three strains, the H2O2 levels did not differ in SA-primed and non-primed plants. However, in symptomless plants infected by the wild type Pba or ΔentA mutant, the levels of H2O2 were greater if the plants were SA-primed.

H 2 O 2 -Level in Infected and Non-Infected Tobacco Plants Pretreated or Not with Salicylic Acid
To check whether the SA-priming of plants (resulting in the increased plant resistance to ∆entA mutant but not to the wild type Pba) influenced the level of ROS and whether the ROS accumulation was different in plants (SA-treated and non-treated) infected by the wild type Pba, or ∆entA mutant, or complementation mutant, the level of hydrogen peroxide was measured. Herewith, among the plants infected with each of the strains, plants that displayed disease symptoms and plants that remained symptomless after the infection were analyzed separately.
The SA-treatment resulted in the increased H 2 O 2 -level in control non-infected plants compared that in non-treated plants (Figure 7). Symptomatic infections in both the SAtreated and non-treated plants, irrespective of the strain that caused them (wild type Pba, or ∆entA mutant, or complementation mutant), were associated with the increased H 2 O 2 -level compared to that in non-treated non-infected plants; and herewith, the SAtreatment did not increase the H 2 O 2 -level in infected plants compared to that in non-treated infected plants.
In symptomless plants infected by the wild type Pba, the level of H 2 O 2 was also increased compared to control; herewith, the H 2 O 2 level did not differ in symptomless plants infected by different strains. During symptomatic infection caused by ∆entA mutant, the level of H 2 O 2 was higher than that in symptomless plants infected by the same strain. Mean values of H 2 O 2 -level during symptomatic infections caused by two other strains (wild type Pba and complementation mutant) were also higher than during the asymptomatic ones although these differences were not statistically significant (Figure 7). During symptomatic infections caused by all three strains, the H 2 O 2 levels did not differ in SA-primed and non-primed plants. However, in symptomless plants infected by the wild type Pba or ∆entA mutant, the levels of H 2 O 2 were greater if the plants were SA-primed.
The level of H 2 O 2 did not differ in non-primed plants infected by different strains during symptomatic infections, and the same was true for the non-primed symptomless plants infected by different strains. As for the SA-primed plants, their H 2 O 2 levels did not differ during both symptomatic and asymptomatic infections caused by the wild type and ∆entA mutant. However, in the SA-primed symptomless plants infected by the complementation mutant, the level of H 2 O 2 was lower than that in the SA-primed symptomless plants infected by the wild type and ∆entA mutant; a similar tendency was observed for symptomatic infection, however the differences were statistically insignificant (Figure 7).
The level of H2O2 did not differ in non-primed plants infected by different strains during symptomatic infections, and the same was true for the non-primed symptomless plants infected by different strains. As for the SA-primed plants, their H2O2 levels did not differ during both symptomatic and asymptomatic infections caused by the wild type and ΔentA mutant. However, in the SA-primed symptomless plants infected by the complementation mutant, the level of H2O2 was lower than that in the SA-primed symptomless plants infected by the wild type and ΔentA mutant; a similar tendency was observed for symptomatic infection, however the differences were statistically insignificant (Figure 7). Thus, the SA-priming leads to the increase in ROS level in tobacco plants. In turn, the increased ROS level within the SA-primed plants can prevent the manifestation of full virulence of ΔentA mutant due to its high ROS-vulnerability. In addition, given the decreased H2O2-level in the complementation mutant-infected SA-primed plants (compared to the wild type Pba-and ΔentA mutant-infected SA-primed plants) (Figure 7) as well as the increased expression level of entA gene within the recombinant plasmid (in the complementation strain) compared to its expression within the bacterial chromosome (in the wild type strain) (data not shown) it may be presumed that enterobactin can be involved in the detoxification of ROS within the infected plants. Thus, the SA-priming leads to the increase in ROS level in tobacco plants. In turn, the increased ROS level within the SA-primed plants can prevent the manifestation of full virulence of ∆entA mutant due to its high ROS-vulnerability. In addition, given the decreased H 2 O 2 -level in the complementation mutant-infected SA-primed plants (compared to the wild type Pbaand ∆entA mutant-infected SA-primed plants) (Figure 7) as well as the increased expression level of entA gene within the recombinant plasmid (in the complementation strain) compared to its expression within the bacterial chromosome (in the wild type strain) (data not shown) it may be presumed that enterobactin can be involved in the detoxification of ROS within the infected plants.

Discussion
One of the genera of soft rot Pectobacteriaceae, Dickeya, was unequivocally shown to use siderophores (chrysobactin and achromobactin) to cause disease in host plants. Herewith, Dickeya siderophores act as both iron carriers and inducers of host reactions related to iron assimilation, which is necessary for the disease development [12,26,28,29,33]. As for the second genus of soft rot Pectobacteriaceae, Pectobacterium, siderophores have not been characterized. Pectobacterium species do not have genes related to chrysobactin and achromobactin synthesis but possess a cluster of the enterobactin-biosynthetic genes, which expression is highly induced during plant colonization [9,10]. In the present study we characterized the phenotype of the mutant strain of Pba deficient for one of the enterobactinbiosynthetic genes entA. We have shown that the knockout of entA gene causes only tiny, if any, effect on the disease development. The mutant produced as much extracellular enzymes (pectate lyase, polygalacturonase, cellulase, protease) as the wild type did and the symptoms of disease did not differ on plants infected with the wild type strain and ∆entA mutant. Only low non-significant reduction in the number of the diseased plants was observed for the ∆entA mutant-infected plants compared to the wild type-infected plants.
Herewith, plant genes of the jasmonate-related pathway, which expression may serve as a marker of Pba-caused disease, were upregulated in the ∆entA mutant-infected tobacco plants significantly less than in the wild type-infected plants. Considering these results, it can be preliminary concluded that enterobactin does not make a significant contribution to Pba virulence. This is consistent with the fact that siderophore-deficient mutants of some phytopathogenic bacteria fully retained virulence [19,[40][41][42].
Differences in the consequences of siderophore-deficiency for closely related phytopathogens, Pectobacterium and Dickeya, can be explained by the following. First, D. dadantii, produces a specific pectate lyase, named PelN, that requires iron as a co-factor for its enzymatic activity in contrasts to iron-independent pectate lyases of pectobacteria [43]. Therefore, Pectobacterium and Dickeya may require different levels of iron to maintain their major virulence factors in active state. Second, siderophores are not the only way of iron acquisition [2] and alternative iron-acquisition systems may compensate siderophore deficiency in Pectobacterium species better than in Dickeya species. For example, Pectobacterium species have a transport system (FecABCD) for ferric citrate, which is formed in plant tissues and provides a long-distance transport of iron [15].
In turn, the enterobactin-deficiency caused a significant reduction in Pba stress resistance, including oxidative stress. It is known, that siderophores, including enterobactin, act as antioxidants and contribute to bacterial resistance to ROS as well as heavy metals [14,34,[36][37][38][39]. Taking this into account, we have hypothesized that we did not observe the reduction of ∆entA mutant virulence compared to the wild type because the virulence test was carried out under infection-promoting conditions. We presumed that the ∆entA mutant virulence could be impaired if tested on the primed plants. The priming state is characterized by preinduced immunity, which enables to respond more rapidly and effectively to pathogen invasion. To test this hypothesis, we used SA. This phytohormone is known to contribute to oxidative stress-related plant defenses [44] and is shown to confer resistance to SRP [45,46]. We used 0.2 mM SA, a concentration that did not reduce significantly the disease development caused by the wild type. However, this SA concentration was sufficient to significantly reduce disease incidents caused by ∆entA mutant strain. The SA-priming of plants led to an increase in their H 2 O 2 -level, to which ∆entA mutant displayed compromised resistance in vitro compared to the wild type. Thus, enterobactin is likely to contribute to the resistance of Pba to the host plant-produced ROS. However, the increased level of plant-produced H 2 O 2 did not influence the growth of the mutant strain since the CFU titers of the ∆entA mutant and the wild type did not differ significantly in the SA-primed plants. This means that the reduced virulence of ∆entA mutant towards the SA-primed plants was not related to its reduced growth in planta. The observed influence of the target mutation on Pba stress resistance and virulence towards the SA-primed plants was not a result of a polar effect of the mutation. The complementation of this mutation restored the wild type phenotype in ∆entA mutant carrying the entA gene within a recombinant plasmid. In addition, the gene located downstream of the entA gene (ECA0482, encoding non-ribosomal peptide synthetase) was expressed at similar level in the wild type and ∆entA mutant (data not shown).
Our results show that in addition to its role in stress resistance, enterobactin may be considered as a conditionally beneficial virulence factor of Pba. Its effect on virulence is undetectable if non-primed plants are tested. However, in the SA-primed plants, enterobactin significantly contributes to Pba virulence. Given that under natural conditions plants are continuously exposed to stress factors and thus are in primed state, the significant upregulation of enterobactin related genes [10] evidently gives strong benefit to Pba in plant colonization. Interestingly, another siderophore, DFO, can also be regarded as a conditionally beneficial virulence factor of E. amylovora. DFO appeared to be unnecessary for the bacteria to cause the disease on apple seedlings but was explicitly required for the colonization of apple flowers and ROS resistance [19]. This means that the conclusions that siderophores do not make contribution to the virulence of some plant pathogenic bacteria [19,[40][41][42] may be premature and a number of experimental conditions should be tested to judge about the role of a gene or metabolite in virulence.
Taken together, enterobactin contributes to Pba stress resistance. Enterobatin is also necessary for Pba to cause disease in that plant, whose immune system is in the primed state. Herewith, if the plant is non-primed, the enterobactin does not make significant contribution to Pba virulence.

Bacterial Strains, Media and Culture Conditions
Pectobacterium atrosepticum SCRI1043 (Pba) (ATCC BAA-672) was grown in Luria-Bertani (LB) medium on a rotary shaker (180 rpm) at 28 • C. The ∆entA mutant strain was grown in the presence of kanamycin (20 µg/mL) and the complementation mutant was grown in the presence of kanamycin (20 µg/mL) and ampicillin (200 µg/mL). The CFU titer was determined by the plating of serial 10-fold dilutions of the cultures onto 1.5% LB agar. For stress tolerance assay, minimal medium D5 (0.1 mM Na-K phosphate buffer (pH 7.5), 1.0 g L −1 NH 4 Cl, 0.3 g L −1 MgSO 4 ·H 2 O, 2.0 g L −1 sucrose) was used. For the extracellular enzymatic activity assays, minimal medium D5 was supplemented by 2.0 g L −1 pectin (FLUKA) instead of sucrose.

Construction of entA Deletion Mutant
The entA deletion mutant (∆entA) was constructed by the method described by Kaniga et al. (1991). The target gene entA (ECA0481 locus) together with the adjacent regions (approximately 1000 bp up-and downstream of entA ORF) were amplified by PCR with primers upentAF and dnentAR (Table S1; Figure S1) using Q5 high-fidelity DNA polymerase (NEB, Ipswich, MA, USA). The amplified PCR fragment was cloned into the bacterial cloning vector system pGEM-T Easy (Promega, Madison, WI, USA). The obtained plasmid (pGEM:entA) was introduced into E. coli NovaBlue by chemical transformation. Transformants carrying the recombinant plasmid were screened by ampicillin resistance and further verified by PCR using plasmid specific primers for the T7 and SP6 polymerase promoters, which flank the multiple cloning regions of pGEM-T Easy.
To replace entA ORF with the Km R cassette, a part of pGEM:entA plasmid (including 1000 bp regions up-and downstream of entA ORF but not entA ORF itself) was amplified with primers dnentAKmF and upentAKmR (Table S1; Figure S1), whose 5 -ends were complementary to the end regions of Km R cassette. The amplified PCR fragment was treated with restriction endonuclease DpnI to remove the original methylated plasmid and then purified using DNA cleanup kit (NEB, USA). Km R cassette was amplified from pKD4 plasmid with primers KmentAF and KmentAR, whose 5 -ends were complementary to Pba DNA regions adjacent to entA ORF. Two obtained PCR fragments (corresponding to pGEM plasmid with~1000 bp regions up-and downstream of entA ORF and to Km R cassette) were joined by circular polymerase extension cloning method [47] ( Figure S1). The obtained plasmid (pGEM:∆entA;Km R ) was introduced into E. coli NovaBlue by chemical transformation. The mutant locus was confirmed by DNA sequencing.
The mutant locus (containing the Km R cassette and~1000 bp regions up-and downstream of entA ORF) was amplified with primers upentAF and dnentAR (Table S1; Figure S1) and ligated (T4 ligase, NEB, USA) into the SmaI-digested (NEB, USA) suicide vector pKNG101 to generate the recombinant plasmid containing the allelic exchange cassette for the target locus. The obtained plasmid (pKNG101:∆entA;Km R ) was introduced into E. coli cc118 by electroporation. The transfer of pKNG101:∆entA;Km R plasmid from E. coli cc118 into Pba was achieved by triparental mating using E. coli HH26 as a helper strain. The clones, in which pKNG101:∆entA;Km R plasmid was integrated into the chromosome by a single crossover event, were selected by streptomycin and kanamycin resistance. The clones, in which the second crossover event led to the replacement of the target locus with the mutant one and the donor plasmid was eliminated, were selected on M9 agar medium containing 10% sucrose. Then, the clones were tested for the sensitivity to streptomycin. Clones without streptomycin resistance were analyzed by PCR with primers CheсkentAF and CheckentAR to identify ∆entA mutants.

Construction of the Complementation ∆entA Mutant Strain Carring the entA Gene within the Plasmid
The entA is the fifth gene within the operon that contains seven genes (entCEBFA and two non-ribosomal peptide synthetases). To construct the complementation plasmid, the target gene entA as well as 343 bp promoter region of the operon and 153 bp terminator region were amplified by PCR with primers promentAF and entAtermR, promF and promentAR, entAtermF and termR, respectively (Table S1), using Q5 high-fidelity DNA polymerase (NEB, USA). The amplified PCR fragments were assembled by overlapping PCR. The assembled fragment was cloned into the bacterial cloning vector system pGEM-T Easy (Promega, USA). The obtained plasmid (pGEM:entA; complementation construct) was introduced into E. coli NovaBlue by chemical transformation. Transformants carrying the recombinant plasmid were screened by ampicillin resistance and further verified by PCR using plasmid specific primers for the T7 and SP6 polymerase promoters, which flank the multiple cloning regions of pGEM-T Easy. The correct assembly was confirmed by DNA sequencing.
The obtained plasmid (pGEM:entA; complementation construct; Amp R ) was introduced into Pba ∆entA (Km+) strain by electroporation. Clones with ampicillin and kanamycin resistance were analyzed by PCR with primers KmentAF/KmentAR and entAF/entAR to identify ∆entA mutants with the complementation construct.

Plant Cultivation and Infection
Nicotiana tabacum cv. Petit Havana SR1 plants were grown axenically in test tubes placed in a growth chamber with a 16-h light/8-h dark cycle photoperiod. Seeds were surface-sterilized using diluted bleach (0.8% of active chlorine) and 1% sodium dodecyl sulfate for 30 min, washed seven times with sterile distilled water, then transferred to Murashige and Skoog medium (MS) in Petri dishes. In this case, 10-day-old seedlings were transferred to individual flasks containing MS. Four to five weeks after planting, plants were infected with Pba, or ∆entA mutant, complementation mutant. For plant inoculation, bacteria were grown until the early stationary phase (~2 × 10 9 colony-forming units, CFU mL −1 ), then washed with sterile 10 mM MgSO 4 and resuspended in the same solution up to a density of~2 × 10 7 CFU mL −1 . Sterile 10 mM MgSO 4 or bacterial suspensions containing~2 × 10 5 cells were placed as 10 µL drops into the bosoms of the leaves in the middle part of the stems using sterile pipette tips and slight scratches were made simultaneously. In the experiments with plant priming, the plants were treated by pulverization with~350 µL per plant of a solution of 0.2 mM salicylic acid or water 24 h before inoculation.
Virus-free potato (Solanum tuberosum cv. Condor) plants were vegetatively propagated under sterile conditions. Apical parts (2 cm length) of sterile plants were rooted aseptically in vermiculite. Then, the plants were grown in 300 mL plastic pots on commercial soil (Peter Peat, Dzerzhinsky, Russia) with a 16-h light/8-h dark cycle photoperiod at 20 • C.
Five weeks after planting into the soil, plants were infected by Pba or ∆entA mutant by injecting 20 µL of the bacterial suspension (~2 × 10 7 CFU mL −1 ) into the middle part of a stem.

Stress Tolerance Assay
To compare stress resistance of the wild type Pba, ∆entA mutant and complementation mutant, bacteria were cultured in D5 medium supplemented with 2.0 g L −1 sucrose in the absence or in the presence of Na-EDTA (20 µM), H 2 O 2 (0.7, 1.5 or 3.0 mM) or CuSO 4 (1.25, 2.5, 5 µM). After 24 h of cultivation, suspensions were plated onto 1.5% LB agar as serial 10-fold dilutions. The plates were incubated at 28 • C for 2 days before the CFUs were counted.

Enzymatic Activity Assays
Pectate lyase activity was determined by measuring the degradation of polygalacturonic acid (PGA) into unsaturated products [48]. First, 435 µL of 0.25% PGA (Sigma, Ronkonkoma, NY, USA) in 50 mM Tris-HCl buffer (pH 8.5) was mixed with 50 µL of 10 mM CaCl 2 and 50 µL of the cultural supernatant at 37 • C. The accumulation of the unsaturated products was measured at 234 nm. One unit of pectate lyase activity was defined as the amount of enzyme releasing 1 µmol of unsaturated products/min per 10 9 bacterial cells.
Cellulase and polygalacturonase activities were determined by measuring the reducing sugars released after the enzymatic hydrolysis of the corresponding substrates. The reducing sugars were measured using 3,5-dinitrosalicylic acid (DNS reagent) (Sigma, USA) at 540 nm [49]. Cellulase (endoglucanase) activity was determined using carboxymethyl cellulose as a substrate (Sigma, USA). Herewith, 250 µL of the cultural supernatant were mixed with 250 µL of 2% carboxymethyl cellulose in 100 mM citrate buffer (pH 5.5) and incubated 30 min at 50 • C [50]. For the determination of polygalacturonase activity, 180 µL of PGA (5 mg/mL) in 50 mM sodium acetate buffer pH 5.0 were mixed with 20 µL of bacterial supernatant and incubated 60 min at 37 • C [51]. The reactions were stopped by heating at 100 • C, 5 min before the analysis of products by DNS reagent. One unit (U) of activities (cellulase and polygalacturonase) was defined as the amount of enzyme releasing 1 µmol of reducing sugars (glucose)/min per 10 9 bacterial cells.
For protease activity assay, 500 µL of 1% azocasein in 100 mM Tris HCl (pH 7.5) and 100 µL of culture supernatant were mixed and incubated 60 min at 37 • C and then 100 µL of 10% trichloroacetic acid were added to the reaction mixture. The sediment was removed by filtration and 500 µL of the supernatant were incubated with 166 µL of 1M NaOH 10 min at 25 • C and the absorbance was measured at 440 nm [52]. One unit (U) of the protease activity was defined as the amount of enzyme required to produce an absorbance change of 1.0/min per 10 9 bacterial cells. All enzymatic activities were measured in three biological replicates using a РB2201B spectrophotometer (SOLAR, Belarus).

Gene Expression Analysis
Plant leaves (one day after plant inoculation) were ground in liquid nitrogen in mortars.
be stable under the experimental conditions (data not shown), were used for normalization of the target gene expression. Relative expression levels were determined as the ratios between the quantities of cDNA corresponding to the target genes and values of the normalization factor, which was calculated for each sample using geNorm software based on transcript levels of reference genes. The presented data were obtained by the analysis of four-five biological replicates.

Determination of the Hydrogen Peroxide Level
H 2 O 2 -level in the tobacco leaves was measured 1 day after mock or bacterial inoculation. One day before the inoculation, plants were treated with water or salicylic acid as described above. H 2 O 2 -level was determined by the method based on the peroxidemediated oxidation of Fe 2+ followed by the reaction of Fe 3+ with xylenol orange (Sigma, USA) [53]. Leaves (100 mg) were ground in 1 mL of cold 50 mM borate buffer, pH 8.4 in mortars. The homogenates were centrifuged (7000× g, 10 min) and 100 µL of the supernatants were added to 500 µL of the assay reagent (500 mM ammonium ferrous sulfate, 50 mM H 2 SO 4 , 200 mM xylenol orange and 200 mM sorbitol). The absorbance of the Fe 3+ -xylenol orange complex (A560) was detected after 45 min. Standard curves of H 2 O 2 were obtained for each independent experiment by adding variable amounts of H 2 O 2 to 100 mL of borate buffer mixed to 500 mL of assay reagent. Data were normalized and expressed as µmol H 2 O 2 per gram of fresh weight. The presented data are means ± SD of three biological replicates.