The conserved upstream region of lscB/C determines expression of different levansucrase genes in plant pathogen Pseudomonas syringae

Background Pseudomonas syringae pv. glycinea PG4180 is an opportunistic plant pathogen which causes bacterial blight of soybean plants. It produces the exopolysaccharide levan by the enzyme levansucrase. Levansucrase has three gene copies in PG4180, two of which, lscB and lscC, are expressed while the third, lscA, is cryptic. Previously, nucleotide sequence alignments of lscB/C variants in various P. syringae showed that a ~450-bp phage-associated promoter element (PAPE) including the first 48 nucleotides of the ORF is absent in lscA. Results Herein, we tested whether this upstream region is responsible for the expression of lscB/C and lscA. Initially, the transcriptional start site for lscB/C was determined. A fusion of the PAPE with the ORF of lscA (lscB UpN A) was generated and introduced to a levan-negative mutant of PG4180. Additionally, fusions comprising of the non-coding part of the upstream region of lscB with lscA (lscB Up A) or the upstream region of lscA with lscB (lscA Up B) were generated. Transformants harboring the lscB UpN A or the lscB Up A fusion, respectively, showed levan formation while the transformant carrying lscA Up B did not. qRT-PCR and Western blot analyses showed that lscB UpN A had an expression similar to lscB while lscB Up A had a lower expression. Accuracy of protein fusions was confirmed by MALDI-TOF peptide fingerprinting. Conclusions Our data suggested that the upstream sequence of lscB is essential for expression of levansucrase while the N-terminus of LscB mediates an enhanced expression. In contrast, the upstream region of lscA does not lead to expression of lscB. We propose that lscA might be an ancestral levansucrase variant upstream of which the PAPE got inserted by potentially phage-mediated transposition events leading to expression of levansucrase in P. syringae.

P. syringae pv. glycinea PG4180 causes bacterial blight of soybean plants. Like some other Pseudomonas species, this organism utilizes sucrose as a carbon source with the help of the enzyme levansucrase (EC 2.4.1.10, Lsc), in the process releasing glucose and forming the exopolysaccharide levan. PG4180 produces no alginate due to a native frameshift mutation in the algT gene and hence, the exopolysaccharide matrix of this strain is mainly composed of levan [11]. Additionally to several draft genome sequences [12][13][14][15][16][17][18], the complete genome sequences of three P. syringae pathovars are available, namely pv. tomato DC3000 [19], pv. phaseolicola 1448A [20] and pv. syringae B728a [21]. These strains serve as excellent model organisms to study plant-microbe interactions. Like in some other P. syringae pathovars, the PG4180 genome contains three copies of the lsc gene, of which two -lscA and lscCare chromosomally encoded while lscB is plasmidencoded. Of the three copies, only lscB and lscC have been shown to be expressed while no expression was observed for lscA under the tested growth conditions since a mutant, PG4180.M6, lacking lscB and lscC but containing lscA was levan-negative [10]. Interestingly, the ORF coding for LscA is fully functional since this gene from pv. glycinea, and its homologues from pv. phaseolicola and pv. tomato, could be expressed from recombinant promoters in Escherichia coli [9,22]. Even though LscB is predominantly extra-cellular and LscC is predominantly retained in the periplasm, the two enzymes are 98% identical at the amino acid level [23]. There are only five amino acid residues different, four of which are conserved changes. Since the enzymes are highly similar in their structure as well as function, all experiments in this study were done using lscB only.
As reported by Srivastava et al. [24], nucleotide sequence comparison of the lscA variants with those of lscB/C variants of P. syringae pathovars showed that the first 48-bp of the N-terminus of the ORF lscB/C were absent in lscA. In silico removal of this N-terminal region increased the identity from 87.5% to 93% at the amino acid residue sequence level between LscA and B/C variants. The comparison also showed that a~450-bp upstream region, which is highly conserved in all lscB/C variant loci, is missing upstream of lscA. This region spanning from −450-bp to +48-bp with respect to the translational start site of lscB/C was predicted to be a pro-phage borne DNA based on sequence similarities and hence was termed phage-associated promoter element (PAPE) [24]. P. syringae is the only Lsc-synthesizing organism having multiple gene copies coding for this enzyme. The rationale for the occurrence of multiple lsc gene copies, some of which carry upstream PAPEs, remained obscure and prompted the current study, during which the transcriptional start site of lscB/C was determined to be -339 bp upstream to the translational start codon. Subsequently, the PAPE with or without the N-terminal coding sequence was fused to lscA. Additionally, the upstream region of lscA was fused with the coding sequence of lscB while lscB and lscA with their native upstream sequences served as controls. All fusion constructs were expressed in the levan-negative mutant PG4180.M6 [10], and tested for their levan formation ability by zymographic detection followed by matrix-assisted laser desorption/ionization time of flight (MALDI-TOF) analysis as well as by Western blotting. Furthermore, the expression of the fusions at the mRNA level was checked by qRT-PCR analysis. In addition, a PCR approach with cDNA was undertaken to show that the expression of lscA is also cryptic in other P. syringae pathovars.

Results
Determination of the transcriptional start site of lscB The coding regions and upstream sequences of lscB/C are highly identical to each other (98.1% DNA identity for the coding sequences and 97.5% DNA identity for the 500-bp upstream sequences). As shown by Srivastava et al., a deletion construct ending at position −332-bp with respect to the lscB translational start codon does not lead to levan formation in levan negative mutant PG4180.M6 while the construct ending −440-bp leads to levan formation in the same mutant [24]. Consequently, primer extension experiments using total RNA from PG4180 cells and a set of reverse oligonucleotide primers were used to determine the transcriptional start site (TSS) of the lscB gene. Resolving the extension products on a polyacrylamide gel resulted in a clear signal at nucleotide position −339-bp upstream of the translational start codon of lscB ( Figure 1). The experiments were repeated for lscC giving identical results (Data not shown).

Qualitative analysis of lsc fusion proteins
The fusion constructs were introduced to the levannegative mutant PG4180.M6 and were first analyzed Figure 1 Determination of the transcriptional start site (TSS) of lscB in P. syringae pv. glycinea PG4180. The TSS was determined by electrophoresis of nucleotide sequencing reaction and primer extension product using primer pe.BC.PG~150 bp on 6% polyacrylamide gel. Nucleotide of the TSS (*) is shown at the right.
for their levan forming ability on sucrose supplemented mannitol-glutamate agar plates. Both, the PG4180.M6 mutant complemented with lscB UpN A and lscB Up A, showed levan formation indistinguishable from that of the PG4180. M6 mutant complemented with lscB ( Figure 2). In contrast, PG4180.M6 complemented with lscA Up B was levan negative, same as PG4180.M6 transformed with lscA, thus, suggesting that the upstream region of lscB mediates expression of downstream located genes while that of lscA does not.

Characterization of lsc fusion proteins
To verify the molecular sizes of Lsc encoded by the individual fusion constructs, a Western blot analysis using Lsc-specific antibodies was performed (Figure 3a). Under denaturing conditions, it was interesting to observe that LscB UpN A migrated at an intermediate rate i.e. faster than LscB but slower than LscB Up A. The signal for LscB Up A was weaker than those representing LscB or LscB UpN A suggesting that the N-terminus of LscB might contribute to the expression level or stability of Lsc. In contrast, protein samples of PG4180.M6 transformed with LscA or LscA Up B did not show any signal specific for Lsc at all thus confirming that lack of levan formation was due to lack of the corresponding protein.
To check for the enzymatic function of Lscs encoded by the individual fusion constructs, zymographic detection was done with non-denatured total protein samples of transformed mutants (Figure 3b). The above reported levan forming ability of transformants M6(lscB), M6(lscB UpN A) and M6(lscB Up A) could be attributed to the enzymatic functioning of proteins or fusion proteins. As expected, native protein samples derived from M6(lscA) or M6 (lscA Up B) did not exhibit any in-gel levan production ( Figure 3b). An interesting observation was the altered electrophoretic mobility of the enzymatically active proteins. The LscB UpN A migrated slower as compared to LscB even though the predicted molecular masses of both proteins were almost identical (~47.6 kDa) suggesting possible differences in the respective protein charges. In accordance with the Western blot results, LscB Up A seemed to be less expressed than LscB or LscB UpN A suggesting an important role of the N-terminus for transcriptional or translational processes.

MALDI-TOF analysis
The altered electrophoretic migration rate of LscB UpN A as compared to LscB during the native gel protein separation suggested that the two proteins were indeed different although their predicted protein sizes were almost identical. To demonstrate that LscB UpN A produced a unique and novel enzyme and to show that the other two transformants indeed also produced the intended Lsc proteins, we subjected the levan-forming fusion proteins to MALDI-TOF analysis. The peptides recovered in the MALDI-TOF analysis are shown in Figure 4. The recovered peptides gave rise to an overall good coverage in the protein sequences ( Table 1). Some of the peptides recovered were unique to each protein (Figure 4, underlined). E.g., peptides SFVQEVYDYGYIPAM from LscB UpN A and SFVQEEYDY-GYIPAM from LscB were located at the same position, namely 413-427, in the respective amino acid sequences

Analysis of lscA fusion protein expression by qRT-PCR
The difference in the amount of levan produced by LscB Up A as compared to LscB UpN A and LscB in the zymogram prompted us to check if this correlated at the RNA level. Samples were grown in HSC medium at 18°C and harvested at OD 600 of 0.5 since lsc transcription is maximum at this optical density [23]. The total RNA was extracted from the cells and the expression of lscB and lscA Up B was checked by lscB-specific primers while that of lscA, lscB UpN A and lscB Up A was checked by lscA-specific primers. The results showed that, considering the standard deviation obtained for the samples, the lscB UpN A had expression levels similar to lscB ( Figure 5) further supporting the results of the Western blot and zymogram. On the other hand lscB Up A had only 60% expression as compared to lscB. As was the trend seen in the Western blot and zymogram, lscA and lscA Up B had no expression. This indicated that even though the upstream region of lscB is sufficient to promote the expression of lsc, the expression level is enhanced by the presence 48-bp N-terminus of lscB.
Analysis of native gene expression of lscA in P. syringae pathovars Lack of expression of lscA had been shown before in P. syringae pv. glycinea PG4180 [10]. However, this has not been experimentally proven for other P. syringae pathovars. Consequently, possible expression patterns of lscA variants were also analyzed in the three P. syringae pathovars pv. phaseolicola 1448A, pv. syringae B728a and pv. tomato DC3000 using cDNA synthesis and PCR. No amplicon was detected in any of the four strains as shown in Figure 6 indicating that none of the lscA variants are expressed. The specificity of the primers was demonstrated by amplifying the lscA genes from corresponding genomic  DNA, all of which gave amplicons of the expected sizes. The accuracy of reverse transcription was checked by amplifying a cDNA of a PG4180.M6 transformant carrying a recombinant lscA gene under the control of P lac , where lscA is known to be expressed [10]. Successful cDNA synthesis of total mRNA was also demonstrated by PCR amplifying the cDNA derived from the mRNA of the hexR gene, a hexose metabolism regulator [25]. Gene hexR gave an amplicon of expected size ( Figure 6) indicating correct cDNA synthesis.

Discussion
Genomic co-existence of three highly conserved genes coding for levansucrase is a feature unique to the plant pathogen P. syringae despite the fact that numerous other bacterial species harbor just a single copy of this gene in their genomes. Artificial expression of lscA from P. syringae under the control of the P lac had been shown previously [10]. The same study also showed that lscA could not be expressed under its own promoter. Major differences between lscA and the natively expressed genes lscB and lscC are not found in the coding sequences but in their upstream DNA regions. The upstream regions of lscB and lscC represent a possible PAPE [24]. We previously hypothesized that this PAPE might harbor regulatory sites required for expression of levansucrase and general sugar metabolism in P. syringae. Herein, the PAPE of lscB was fused to the coding sequence of lscA and thus proven for its transcriptional activity in P. syringae. The nucleotide sequence of the predicted PAPE consists of two parts, the upstream region of lscB and the first 48-bp coding for the N-terminus of LscB. The importance of these 48-bp of the ORF for the expression was tested by generating fusion constructs of the upstream region and lscA with or without these coding nucleotides. Transformants carrying either of the two fusion constructs produced levan similar to the PG4180.M6 mutant complemented with lscB. Western blotting, zymographic detection, and qRT-PCR analyses confirmed these results but also allowed a more detailed view; native lscB and the lscB UpN A fusion had similar mRNA expression levels while that of the fusion lscB Up A, which lacked the 48-bp of N-terminal LscB-coding region, had less. Consequently, one might speculate that although the -450 bp upstream DNA region of lscB, which includes the TSS as determined in this study, is sufficient for expression of lscA, the first 48-bp of the lscB ORF increase the level of its expression. Since our respective results of Western blotting and zymographic detection of Lsc activity were indistinguishable from each other, it could be concluded that the N-terminus of LscB might not be involved in altering of enzymatic activities.
A peculiar observation was the electrophoretic migration of the individual proteins or fusion proteins in polyacrylamide gels. The observed faster migration of LscB UpN A Figure 5 Quantitative expression of different lsc genes and constructs in dependence of lscB. lscB UpN A shows similar levels of expression as lscB while lscB Up A, which does not contain the first 48 bp of lscB ORF, has lower expression. lscA and lscA Up B were not seen to be expressed. lscA, lscB UpN A and lscB Up A were detected using lscA primers ( 1 ) while the rest using lscB primers ( 2 ). The data represent the mean relative expression of 3 replicates ± standard deviations. Data were normalized to the highest expression value of lscB, which was set to 100%. as compared to LscB under denaturing conditions could potentially be attributed to the apparent mass shift for two proteins with nearly identical molecular masses as described earlier [26]. Interestingly, the migration of LscB UpN A was significantly slower than that of LscB under native conditions. This finding might demonstrate that modest changes in the protein's surface charge might result in significant alterations of electrophoretic mobility [22,27,28].
Although the different migration rates of the proteins or fusion proteins under native or denaturing conditions suggested that the synthesized proteins were indeed different from each other, a MALDI-TOF analysis of each of the proteins was conducted using protein samples from zymograms. The produced levan surrounding the proteins did not seem to impact mass spectrometric analysis. The MASCOT score for each of the identified proteins was above the significance threshold of 100. The sample from the PG4180.M6(lscB) sample gave LscB from P. syringae pv. phaseolicola 1448A as the first significant match which was in line with the high homology of the respective genes in the close relatives pv. glycinea and pv. phaseolicola [24]. The sample from PG4180.M6(lscB Up A) which should synthesize only LscA gave the first significant match as LscA from P. syringae pv. glycinea race 4 strain. This proved that the lscB Up A fusion actually synthesized an active LscA and confirmed earlier findings that artificial expression of LscA of PG4180 leads to levan formation [10]. Although the majority of obtained peptides for the sample representing LscB UpN A were LscA-borne as expected, the unique N-terminal 2,122-Da peptide NSPLASMSNINYAPTIWSR could be detected. This peptide is a consequence of the presence of the NheI restriction site coding for the amino acid residues alanine and serine. Oxidation of methionine, which was chosen as a variable modification parameter, added another 16 Da to the peptide mass which subsequently increased the mass of the NSPLASMSNINYAPTIWSR fragment to 2,138 Da. This mass was exactly the same as the mass of a recovered peptide which did not find a match during the NCBI search since the respective fusion peptide is not present in the database. Thus, the synthesis of the LscB UpN A fusion protein could also be proven.
The majority of previous LscA-related studies have been performed with P. syringae pv. glycinea PG4180 [9,10,23,24]. However, thus far, there was no evidence for a lack of lscA expression in other pathovars of P. syringae. Since the genomes of P. syringae pv. phaseolicola 1448A, pv. syringae B728a and pv. tomato DC3000 are fully sequenced [19][20][21], template-specific oligonucleotide primers for cDNA-based mRNA detection could be designed. Although mRNA samples were extracted during different growth stages, namely, early-logarithmic and late-logarithmic phase, no amplicons could be detected in any of the strains suggesting that lscA variants were not expressed. PCR amplification, using respective genomic DNA as template, proved that the primers were binding correctly. An independent gene, hexR, coding for a conserved hexose metabolism regulator protein HexR, was chosen to see if the total mRNA had been reverse transcribed correctly [25]. This PCR amplification gave correct sized amplicon of 880-bp for all the four strains demonstrating the accuracy of the used method. PCR amplification was also performed on the cDNA obtained from mRNA samples of PG4180.M6 containing lscA under the control of P lac . This experiment gave the same-sized amplicon as for genomic DNA again proving the accuracy of the method.
In summary, we propose that lscA could be an ancestral Lsc variant in P. syringae as suggested by Srivastava et al. [24]. During evolution, the inactive promoter perhaps did not allow expression of lscA after this gene had potentially been introduced to an ancestral P. syringae. An evolutionary gene duplication of lscA followed by an insertion of a prophage-borne PAPE might have led to a new lsc variant, i.e. lscB which in turn got duplicated yielding lscC or vice-versa. As a result of this evolutionary process, two functional and expressed lsc genes emerged in the plant pathogen, for which utilization of sucrose, and perhaps levan formation, might be particularly important. The advantage of an additional in planta fitness-increasing and possibly virulence-promoting factor [29] could have helped this organism to selectively establish itself as a potent plant pathogen. As a consequence of this hypothesis, one could speculate on a loss of the supposedly nonexpressed lscA during further evolutionary steps, a phenomenon also previously hypothesized by Smits et al. [30].

Conclusions
The differential expression of levansucrases in P. syringae was long known, but not tested. In this study, we have potentially solved the previously unexplainable phenomenon that P. syringae is the only organism possessing multiple levansucrase-encoding genes. We demonstrated the importance of the upstream region as well as the N-terminus of lscB/C required for the expression of Lsc in P. syringae. The upstream region of lscA does not seem to promote lsc expression. With careful controls, herein we also demonstrated that lscA is not expressed in other P. syringae pathovars.

Bacterial strains, plasmids and growth conditions
Bacterial strains, plasmids and oligonucleotides used in this study are listed in Tables 2 and 3. E. coli DH5α was used as the cloning host [31] [32] at 18°C. Bacterial growth in liquid media was monitored by measuring the optical density at 600 nm (OD 600 ) and harvested for (i) protein sampling at an OD 600 of 2.0 or (ii) RNA extraction and cDNA synthesis at an OD 600 of 0.5 and 2.0. Antibiotics were added to the media at the following concentrations (μg ml -1 ): ampicillin 50; tetracycline 25, and chloramphenicol 25.

Molecular genetic techniques
Plasmid isolation, restriction enzyme digests, agarose and polyacrylamide gel electrophoreses, electroporation, PCR, and other routine molecular methods were performed using standard protocols [31]. Nested deletion analysis of the upstream region of lscB in plasmid pRB7.2 [10] was conducted using the Erase-a-Base kit (Promega, Madison, USA). For analysis of the lsc upstream regions, PCR was used to generate products covering the respective regions (Table 3). PCR products of the lsc upstream regions were cloned in vectors pBBR1MCS or pBBR1MCS-3 [36].

Determination of transcriptional start site
Bacteria were incubated in HSC medium at 18°C to an OD 600 of 0.5 and harvested by mixing 15 ml of the culture with an equal volume of chilled killing buffer (20 mM Tris-HCl [pH7.5], 20 mM NaN 3 ). This mixture was centrifuged at 4°C for 15 min at 3,220 × g. Total RNA was isolated from the cell pellets by acid phenol/ chloroform extraction as described previously [37]. For primer extension analysis, 4 pmol of 32 P-labeled primer pe.BC.PG~150 bp (Table 3) were annealed with 10 μg of total RNA and reverse transcription was performed with M-MLV Reverse Transcriptase (Invitrogen, Karlsruhe, Germany). Nucleotide sequencing using 5 μg of plasmid pLB7.2 ( Table 2) and primer pe.BC.PG~150 bp was done with the Sequenase Version 2.0 DNA Sequencing Kit (USB, Cleveland, USA) according to the manufacturer's recommendation. The extension product and sequencing reaction were resolved on a 6% polyacrylamide sequencing gel. Signal detection was performed using a FLA-3000 phosphorimager (Raytest, Straubenhardt, Germany) according to the manufacturer's recommendations.

Generation of fusion constructs
All genes or DNA fragments were obtained by PCR amplification unless otherwise stated. All restriction enzymes used were obtained from Thermo Fisher Scientific Biosciences (St. Leon Rot, Germany). The nucleotide sequencing was done by Eurofins MWG Operon (Ebersberg, Germany). pBBR1MCS Cm r , broad-host-range cloning vector [36] pBBR1MCS-3 Tc r , broad-host-range cloning vector [36] pBBR3-500-lscB Tc r , lscB gene with −500-bp upstream sequence in pBBR1MCS-3 [24] pBBR3 ( Generation of lscB UpN A and lscB Up A: The sequences of the 518-bp PAPE and the 470-bp lscB upstream region without the 48-bp coding sequence, respectively, were ligated to the N-terminus of the 1,748-bp lscA fragment using T4 DNA Ligase (Thermo Fisher Scientific Biosciences) after treating the DNA with restriction enzyme NheI. The ligation products were then treated with HindIII, analysed by agarose gel electrophoresis, and the bands corresponding to the fusion products (2,284 and 2,224 bp, respectively) were purified from the gel using GeneJET Gel Extraction kit (Thermo Fisher Scientific Biosciences). The purified fusion products were ligated into pBluescript-KS(II) using HindIII in such a way that the fusion products were under control of the vector-borne lac promoter (P lac ). Formation of levan on LB agar containing 5% sucrose indicated a functional lscA gene driven by the P lac . The PAPE and lscB upstream regions were sequenced to exclude any possibility of mutations. The fusion products were then cloned into the broad host-range vector pBBR1MCS using HindII in order to ligate them in opposite orientation to the P lac and then cloned into pBBR1MCS-3 using restriction enzymes PstI and XhoI to keep the same opposite orientation with respect to P lac as in case of pBBR1MCS. The constructs were introduced into mutant PG4180.M6 via electroporation.
Generation of lscA Up B: A similar cloning strategy was used to generate the lscA Up B construct. The C-terminus of the 550-bp PCR-amplified lscA upstream region and the N-terminus of the 1,704-bp PCR-amplified ORF lscB were ligated using a combination of restriction enzymes XbaI and NheI which generate compatible DNA ends. This ligation product was treated with endonucleases BamHI and HindIII and subsequently ligated into pBluescript-SK (−). The constructs were cloned into pBBR1MCS using restriction enzymes BamHI and HindII in order to ligate them in opposite orientation to the P lac and then into pBBR1MCS-3 using restriction enzymes using XbaI and ApaI to keep the same opposite orientation with respect to P lac as in case of pBBR1MCS.

Immunological and enzymatic detection of Lsc
Total proteins from PG4180.M6 and PG4180.M6 transformants harboring the lsc fusion constructs were obtained as described previously [23]. For immunological detection of the Lsc enzyme, total proteins were separated by 10% SDS-PAGE and Western blot experiments were performed with total protein fractions using polyclonal antibodies raised against purified Lsc as reported earlier [10]. Zymographic detection of Lsc was done as described previously by separating the total proteins by 10% native-PAGE and incubating the gels in 5% sucrose solution [10]. Bacterial cells grown on mannitol-glutamate agar plates with 1.5% agar and 5% sucrose were used for the qualitative visualization of Lsc activity, which led to levan formation in form of a mucoid, dome-shaped colony morphology. Lsc activity was quantified by measuring the amount of glucose liberated during incubation with sucrose using the Gluco-quant Glucose/HK assay kit (Roche Diagnostics, Mannheim, Germany) at an absorbance of 340 nm. One unit of Lsc activity corresponded to the amount of enzyme which liberates 1 μmol glucose per minute from sucrose. The experiments were repeated threefold and mean values were expressed as the quantity of glucose release.

MALDI-TOF mass spectrometric analysis
Total proteins were separated using 10% native-PAGE and incubated in 5% sucrose solution overnight [10]. As soon as in-gel levan formation became apparent, the corresponding bands were cut out from the gel and subjected to an in-gel proteolytic cleavage using modified porcine trypsin (Promega, Madison, WI) as adapted from previous reports [38][39][40]. Trypsin digestion was carried out for 12-16 h at 37°C, and peptide samples were directly used for MALDI-TOF MS exposure using an Autoflex II TOF/TOF mass spectrometer (Bruker Daltonics, Bremen, Germany) equipped with a 337 nm nitrogen laser and operated with FlexControl 3.0 software. The matrix used was 1 mg ml −1 of a-cyano-4-hydroxycinnamic acid (HCCA; Bruker Daltonics) disolved in acetone and mixed with two volumes of ethanol. Peptide samples were acidified with 0.5% TFA in a ratio of 1:1 (v/v) and mixed with the HCCA solution in a ratio of 1:1 (v/v). Samples of 0.5 μL were spotted and air-dried on MTB AnchorChip targets with an anchor diameter of 600 μm (Bruker Daltonics). Spots were twice rinsed with 2 μL of 10 mM monobasic ammonium phosphate solution for~5 s, dried, and exposed to MALDI-TOF MS in positive-ion reflection mode with the laser offset set to 67% +/− 15% and an acquisition range of 800-4,000 Da. A signal-to-noise ratio of 6 was applied for peak identification using the Mascot search engine [41] from Biotools software 3.1. Mass lists were compared with NCBI databases and the Mascot score probability set for p <0.05. Peptide sequence analyses was done using the ExPASy bioinformatics resource portal [42].
Analysis of lsc gene expression by quantitative Reverse Transcriptase polymerase chain reaction (qRT-PCR) Total RNA was isolated by acid phenol/chloroform extraction as described previously [11]. The yield and the purity of RNA were determined by measuring absorption at 260 nm. Total mRNA samples were treated with TURBO DNA-free (Applied Biosystems, Darmstadt, Germany) to remove remaining traces of genomic DNA as described by the manufacturer's recommendation. SYBR-green based qRT-PCR was performed with 5 ng RNA template and 100 μM primer with QuantiTect SYBR Green one-step RT-PCR Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The thermocycler program comprised an initial step of 95°C for 15 min followed by 40 cycles of 95°C for 30 s, 58°C for 30 s, 72°C for 30 s. Reactions were performed with biological triplicates in a Mastercycler ep realplex2 real-time PCR system (Eppendorf, Hamburg, Germany) as described by the manufacturer using their universal program. Reactions with no addition of reverse transcriptase served as negative control and proved the absence of DNA contamination. Specificity of amplification was assessed by analyzing the melting curve of the amplification product. Primers to amplify lscB were used for constructs lscB and lscA Up B while primers to amplify lscA were used for constructs lscA, lscB UpN A and lscB Up A. All the results were normalized to amplification of the cDNA of gyrA (PSPPH3667) as described previously [43].

Analysis of lscA gene expression by Reverse-Transcriptase polymerase chain reaction (RT-PCR)
Template-specific primers were designed for the respective lscA variants of P. syringae pv. glycinea PG4180, pv. phaseolicola 1448A, pv. syringae B728a, and pv. tomato DC3000. Bacterial cells were grown in HSC medium and harvested at an OD 600 of 0.5 as well as 2.0. RNA was extracted by acid phenol/chloroform extraction method [11]. An RT-PCR was performed on total mRNA using RevertAid First Strand cDNA Synthesis Kit (Fermentas) as recommended by the manufacturer. The strain-specific lscA primers were used to check for presence of an lscA mRNA by PCR using cDNA as template. Regular PCR with the same primer-pairs and genomic DNA as template were used as controls. The thermocycler program was as follows: 1 cycle of 95°C for 90 s; 25 cycles of 95°C for 15 s, 66°C for 15 s, 72°C for 30 s; 1 cycle of 72°C for 5 min. The results were analyzed by 1% agarose gel electrophoresis.

Bioinformatics analyses
Vector NTI Advance 10.1.1 (Life Technologies, California, USA) was used for the nucleotide, amino acid sequence alignments, as well as for generating genetic maps. BLAST-N and BLAST-P programs were used for online sequence analyses [44]. The website www.pseudomonas.com was consulted for the determination of P. syringae gene orthologs and paralogs [45].