Integrated downstream regulation by the quorum-sensing controlled transcription factors LrhA and RcsA impacts phenotypic outputs associated with virulence in the phytopathogen Pantoea stewartii subsp. stewartii

Strains and growth conditions

Table 1 lists strains and plasmids used in this study. E. coli strains were grown in Luria-Bertani (LB) (10 g/l tryptone, 5 g/l yeast extract, and 5 g/l NaCl) broth or plates with 1.5% agar at 37 °C. P. stewartii strains were grown in either LB or Rich Minimal (RM) medium (1 × M9 salts, 2% casamino acids, 1 mM MgCl₂, and 0.4% glucose) at 30 °C. Growth medium was supplemented with the following antibiotics: ampicillin (Ap, 100 µg/ml), chloramphenicol (Cm, 35 µg/ml), kanamycin (Kn, 50 µg/ml), nalidixic acid (Nal, 30 µg/ml), or streptomycin (Str, 100 µg/ml) as required (see Table 1).

Green fluorescent protein fusion (GFP) construction and testing

A transcriptional fusion between the lrhA promoter (903 bp) and the gene for GFP was created through traditional molecular techniques as described previously (Kernell Burke et al., 2015). PCR primers (Table S1) with the restriction sites EcoRI and KpnI added to the 5′ and 3′ ends of the promoter sequence, respectively, were used to facilitate subcloning into the pPROBE′-GFP [tagless] vector (Miller, Leveau & Lindow, 2000). E. coli S17-1 was transformed with this plasmid construct containing P_lrhA-gfp, which was then moved into the wild-type P. stewartii DC283, ΔlrhA and ΔlrhA/lrhA⁺ strains (Table 1) via conjugation. The transconjugates were grown in RM medium supplemented with Kn and Nal overnight, diluted in fresh medium to an OD₆₀₀ of 0.05 at 30 °C with shaking at 250 rpm to an OD₆₀₀ of 0.2–0.5, diluted a second time in fresh medium to an OD₆₀₀ of 0.025 and grown to an OD₆₀₀ of 0.5. GFP measurements were done as previously described (Kernell Burke et al., 2015) with average relative fluorescence/OD₆₀₀ from three experiments of triplicate samples, standard errors, and two-tailed homoscedastic Student’s t-test values calculated for each strain.

Overexpression of LrhA

The lrhA coding sequence was amplified using primers with BamHI and HindIII sites (Table S1), cloned into pGEM-T (Promega, Madison, WI, USA), and sequenced. After double digestion with BamHI and HindIII, the construct was ligated into pET28a (Novagen, Madison, WI, USA) and transformed into E. coli (BL21-DE3) (Studier & Moffatt, 1986) to express LrhA with a His₆ tag at the N-terminus (37 kDa). Induction of protein expression with 0.1 M isopropyl β-D-1-thiogalactopyranoside (IPTG) was performed at an OD₆₀₀ of 0.5–0.8, 19 °C, overnight, shaking at 250 rpm. Cells were pelleted by centrifugation at 5,000 rpm in a JA-10 rotor (Beckman Coulter, Brea, CA, USA) for 20 min at 4 °C, snap-frozen with liquid nitrogen and stored at −75 °C. The cell pellet was then resuspended in Ni-NTA wash buffer (50 mM Tris–HCl, 300 mM NaCl, 50 mM imidazole) and sonicated to release proteins. Ultracentrifugation at 40,000 rpm in a Beckman Ti70 rotor for 1 h at 4 °C was used to subsequently remove the cell debris. The protein was purified using a Ni-NTA column (HisTrap HP, GE Healthcare) with Ni-NTA elution buffer (50 mM Tris–HCl, 300 mM NaCl, 500 mM imidazole). The protein purity was observed through standard SDS-PAGE electrophoresis.

Electrophoretic mobility shift assays (EMSA)

Promoter regions of genes of interest were amplified with FAM-labeled primers (Table S1) and extracted from a 1% agarose gel to examine the specific binding with purified His₆-LrhA over a range of concentrations. Twenty µl reactions with purified His₆-LrhA, 5 nM FAM-DNA in 1X EMSA buffer (10% glycerol, 1 mM MgCl₂, 0.5 mM EDTA, 0.5 mM DTT, 50 mM NaCl, 10 mM Tris–HCl, 50 µg/ml poly (dI-dC) and 150 µg/ml BSA) were incubated at room temperature for 1 h before loading on to 1 × TBE (10.8 g/l Tris–HCl, 5.5 g/l boric acid, 2 mM EDTA, pH 8.0) 4%, 5%, or 6% acrylamide native gels followed by electrophoresis at 80 V for 2–3 h. Images were visualized on a Typhoon Trio Scanner (GE Healthcare). Experiments were done in duplicate.

Construction of unmarked deletion mutant strains

Chromosomal deletions of CKS_0458/CKS_0459, CKS_5208, CKS_5211, and CKS_5211/CKS_5208 were constructed based on the Gateway system (Life Technologies) and suicide vectors as described previously (Kernell Burke et al., 2015), but with primers listed in Table S1. In addition, another chromosomal deletion of rcsA was constructed using the same approach as described in a previous study (Kernell Burke et al., 2015), due to a deletion of ∼66-kilobases (kb) in the chromosome of the original construct.

Construction of chromosomal complementation strains

Complementation strains were constructed by generating a chromosomal insertion of the promoter and coding regions of the target gene into the neutral region downstream of glmS on the P. stewartii chromosome using the pUC18R6K-mini-Tn7-cat vector system (Choi et al., 2005) as previously described (Kernell Burke et al., 2015), but with primers listed in Table S1.

Phenotypic surface motility assay

Swarming motility for the wild-type, deletion and complementation strains was investigated under strict conditions to ensure a reproducible phenotype as previously described (Kernell Burke et al., 2015). Briefly, five µl of cell culture at an OD₆₀₀ of 0.5 were spotted directly on the agar surface of LB 0.4% agar quadrant plates supplemented with 0.4% glucose (Herrera et al., 2008). Plates were incubated in a closed plastic box at 30 °C for 2 days prior to observation.

Phenotypic capsule production assay

Bacterial strains were grown overnight in LB broth supplemented with the appropriate antibiotics at 30 °C with shaking. The overnight cultures were subcultured in fresh LB medium to an OD₆₀₀ of 0.05 and grown to an OD₆₀₀ of 0.5 at 30 °C with shaking. The strains were then cross-streaked with sterilized wooden sticks on 1.5% agar plates containing 0.1% casamino acids, 1% peptone and 1% glucose (CPG) (Kernell Burke et al., 2015; Von Bodman, Majerczak & Coplin, 1998). Plates were incubated at 30 °C, lid-up for 2 days to observe the capsule production and visualized using the Bio-Rad Gel Doc imager system.

Plant virulence assay

Virulence assays with P. stewartii strains in Zea mays seedlings were adapted from established methods (Von Bodman, Majerczak & Coplin, 1998; Kernell Burke et al., 2015) with some modifications. In this study, Zea mays cv. Jubilee, 2B seeds were planted in Sunshine mix #1 or Promix soil for seven or six days, respectively, in a 28 °C growth chamber with ∼100–200 µE m⁻² s⁻¹ light intensity, 16 h light/8 h dark and ∼80% relative humidity (Percival Scientific, Inc.). Fifteen seedlings between 6 and 10 cm of height with two separated leaves were inoculated with five µl (∼3 × 10⁵ CFU) of bacterial culture grown to an OD₆₀₀ of 0.2 in LB broth (∼6 × 10⁷ CFU/ml). Prior to plant inoculation, the bacterial cells were washed and resuspended in phosphate buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄ and 2 mM KH₂PO₄, pH 7.4). Wild-type strain and PBS controls were included in each trial, then, accumulated numbers of control-inoculated plants across all experiments were analyzed. A sterile needle (26G 5/8, 15.9 mm, SUB-Q Becton, Dickinson and Company) was used to make an ∼1 cm incision in the stem ∼1 cm above the soil line and the bacteria were added to the plant by slowly pipetting the inoculum while moving across the wound five times. The plants were observed on day 12 post-infection to assess the virulence by two independent observers. Symptom severity was scored based on a five-point scale with 0 = no symptoms; 1 = few scattered lesions; 2 = scattered water soaking symptoms; 3 = numerous lesions and slight wilting; 4 = moderately severe wilt; 5 = death. The data for each treatment were averaged together and used to calculate the mean and standard error. A Student’s t-test was used to calculate the p-value for experimental treatments compared to the wild-type treatment.

LrhA autorepresses its own gene expression in P. stewartii

A GFP reporter was used to measure levels of transcription from the lrhA promoter in the wild-type, ΔlrhA and ΔlrhA/lrhA⁺ strains of P. stewartii DC283 (Table 1). Expression levels of GFP in the ΔlrhA strain were significantly higher than the wild-type strain (Fig. 1, p = 0.00001) indicating that LrhA normally represses its own expression in the wild-type strain. The expression level of the lrhA promoter in the complementation ΔlrhA/lrhA⁺ strain was restored to levels closer to those of the wild-type strain, and was also significantly different than the deletion strain (Fig. 1, p = 0.00002).

Identification of LrhA direct targets through EMSAs

To determine if the observed lrhA autorepression occurred directly or indirectly, electrophoretic mobility shift assays (EMSAs) were performed. First, direct binding of LrhA to the promoter of its own gene was demonstrated by EMSA analysis (Fig. 2A). Next, the ability of LrhA to directly regulate additional gene targets was explored, using the lrhA promoter as a positive control for the His₆-LrhA activity and unlabeled P_lrhA DNA to prove the specificity of the binding. In E. coli, LrhA is known as a repressor of motility by direct interaction with the promoter region of flhD/flhC, whose products promote the expression of flagellar gene synthesis (Lehnen et al., 2002). However, RNA-Seq data of expression levels of flhD/flhC in P. stewartii showed less than a two-fold difference between wild-type and ΔlrhA strains (Kernell Burke et al., 2015) suggesting a lack of transcriptional regulation. Here, EMSA analysis showed that His₆-LrhA does not bind to the promoter of flhD/flhC (Fig. 2B), explaining the observed lack of transcriptional regulation.

Additional analysis of the LrhA-regulated transcriptome in P. stewartii revealed that LrhA repressed the expression level of several more downstream targets, including rcsA, CKS_0458, CKS_5208 and CKS_5211 (Kernell Burke et al., 2015). RcsA activates capsule production, a known virulence factor in P. stewartii (Kernell Burke et al., 2015; Minogue et al., 2005; Poetter & Coplin, 1991; Wehland et al., 1999). The putative roles of genes for fimbria encoded by CKS_0458, annotated as a putative fimbrial subunit, (and CKS_0459 located downstream in an operon) and for surfactant expression encoded by CKS_5208 and CKS_5211, annotated as a rhamnosyltransferase I subunit B and a putative alpha/beta superfamily hydrolase/acyltransferase, respectively, in plant colonization and/or virulence had not been established. However, it seemed plausible that they might also play roles in host association as they were some of the most highly LrhA-repressed genes, four-fold or greater (Kernell Burke et al., 2015). Therefore, the binding of LrhA to the promoters of these genes was also examined via EMSA. The direct binding of His₆-LrhA to P_rcsA, P_{CKS_0458} and P_{CKS_5211} (Figs. 2C–2E), was demonstrated via EMSAs while P_{CKS_5208} did not interact with His₆-LrhA in vitro (Fig. 2F). Collectively, these findings identified four directly controlled gene targets in the LrhA regulon. The lack of LrhA regulation of FlhD₂C₂, the master regulator of flagellar-based motility and chemotaxis in E. coli, indicates a different role for LrhA in controlling P. stewartii motility. The direct binding of LrhA to the promoter of rcsA further links LrhA to P. stewartii pathogenesis. The role of the two other LrhA direct targets CKS_0458 and CKS_5211 remained to be established.

Examining the role of putative fimbrial and surfactant production genes in the surface motility and virulence of P. stewartii

To further investigate the role of the downstream targets of LrhA putatively involved in production of fimbriae and surfactant, a reverse genetic approach was utilized. Markerless deletions of CKS_0458/CKS_0459, CKS_5208, CKS_5211 and CKS_5211/CKS_5208 were successfully generated. Corresponding chromosomal complementation strains were also generated with the exception of a double deletion mutant of CKS_5211/CKS_5208 complementation strain, due to the length constraint of the DNA fragment containing the adjacent genes. In surface motility assays, the P. stewartii wild-type strain showed either uni-directional (Fig. 3A and Fig. S1A) or omni-directional (Fig. 3B and Fig. S1B) expansion from the inoculum sites as had been previously observed (Herrera et al., 2008; Kernell Burke et al., 2015). In comparison to the wild type, there is no obvious difference between the various deletion and complementation strains; they all possessed similar level of expansion on the agar surface (Figs. S1C –S1J). Therefore, these genes do not appear to play any detectable role in surface motility via this assay.

The same deletion and complementation strains of the genes putatively involved in fimbriae and surfactant production were also tested for virulence via in planta xylem infection assays. A lrhA deletion strain caused an intermediate level of disease severity in corn seedlings during xylem-infection assays (Kernell Burke et al., 2015). However, similar to the surface motility assays, no significant impacts on the virulence of P. stewartii were observed in the strains with deletions in either the fimbriae or surfactant synthesis genes (Fig. S2). Hence, the contribution of these genes individually to the virulence of the phytopathogen could not be measured.

Re-examining the role of RcsA in the capsule production, surface motility and virulence of P. stewartii

The important finding that LrhA directly binds to the promoter of rcsA, led to a reexamination of the previous findings about the physiological role of RcsA in P. stewartii. In prior work, rcsA deletion and complementation strains of DC283 had been constructed (ΔrcsA-2015 and ΔrcsA/rcsA⁺-2015) (Kernell Burke et al., 2015). However, complete assembly of the genome of P. stewartii DC283 (Duong, Stevens & Jensen, 2017) revealed that there is a large deletion, ∼66 kb containing 68 genes (Table S2), in the ΔrcsA-2015 and ΔrcsA/rcsA⁺-2015 strains. This deletion was not obvious using the incomplete genome sequence (NCBI GenBank accession no. AHIE00000000.1), but was found during a re-analysis of previously generated RNA-Seq data (Kernell Burke et al., 2015) using the new genome sequence (NCBI GenBank accession no. CP017581).

Therefore, a new set of rcsA deletion and complementation strains was re-constructed (ΔrcsA-2017 and ΔrcsA/rcsA⁺-2017) and shown to include the 66-kb region using PCR (data not shown). These new strains then were subjected to three assays to establish the true phenotypes of the rcsA deletion strain. First, capsule production assays have re-confirmed that RcsA regulates EPS production, as was shown for the 2015 strains (Kernell Burke et al., 2015). Both the ΔrcsA-2015 (Fig. 4D) and ΔrcsA-2017 strains (Fig. 4F) are not as mucoid as the parental wild-type strain (Fig. 4A) or the pair of lrhA deletion and complementation strains (Figs. 4B and 4C) as assessed by visual observation. The chromosomal complementation strains ΔrcsA/rcsA⁺-2015 (Fig. 4E) and ΔrcsA/rcsA⁺-2017 (Fig. 4G) had mucoid levels as high or higher than those seen in the wild type (Fig. 4A).

Second, surface motility assays were performed. These original strains ΔrcsA-2015 and ΔrcsA/rcsA⁺-2015 strains had not previously been examined for surface motility (Kernell Burke et al., 2015), but both were surprisingly defective for this phenotype (Figs. 3E and 3F). The fact that surface motility was not complemented by addition of rcsA back into the chromosome provided further evidence for the importance of the 66-kb region deletion that had been discovered initially through bioinformatics analysis. The new ΔrcsA-2017 also has severely reduced surface movement (Fig. 3G) while its complementation strain (Fig. 3H) restored motility levels similar to the wild-type strain (Figs. 3A and 3B). Thus, it has been demonstrated that RcsA plays a previously unappreciated role in the surface motility of P. stewartii. The defect in surface motility associated with the deletion of rcsA (Fig. 3G) appears to be greater than the defect in ΔlrhA (Fig. 3C). The ΔlrhA/lrhA⁺ strain has restored levels of surface motility (Fig. 3D), similar to the wild type (Figs. 3A and 3B), as previously reported (Kernell Burke et al., 2015).

Finally, the xylem infection assays for the newly constructed ΔrcsA-2017 strain and its complement, with the inclusion of the wild-type strain and PBS as controls, indicated that the absence of rcsA significantly (p < 0.05) reduces the severity of the disease compared to the wild-type and complementation strains (Fig. 5) (p < 0.05). These results have similar trends with those reported for the 2015 strains (Kernell Burke et al., 2015) which confirms the role of rcsA in virulence of this phytopathogen. However, strains from the 2015 study that were missing the 66-kb region were reduced in their average disease severity (score ∼0 and ∼1.5 for the deletion and complementation strains, respectively) in comparison to the new 2017 strains (score ∼1.5 and ∼3.5 for the deletion and complementation strains, respectively) while the wild-type control had similar levels in both studies, implicating a role for the 66-kb region in virulence as well as surface motility.