Functional Characterization of a Global Virulence Regulator Hfq and Identification of Hfq-Dependent sRNAs in the Plant Pathogen Pantoea ananatis

Abstract

To successfully infect plant hosts, the collective regulation of virulence factors in a bacterial pathogen is crucial. Hfq is an RNA chaperone protein that facilitates the small RNA (sRNA) regulation of global gene expression at the post-transcriptional level. In this study, the functional role of Hfq in a broad host range phytopathogen Pantoea ananatis was determined. Inactivation of the hfq gene in P. ananatis LMG 2665^T resulted in the loss of pathogenicity and motility. In addition, there was a significant reduction of quorum sensing signal molecule acyl-homoserine lactone (AHL) production and biofilm formation. Differential sRNA expression analysis between the hfq mutant and wild-type strains of P. ananatis revealed 276 sRNAs affected in their abundance by the loss of hfq at low (OD₆₀₀ = 0.2) and high cell (OD₆₀₀ = 0.6) densities. Further analysis identified 25 Hfq-dependent sRNAs, all showing a predicted Rho-independent terminator of transcription and mapping within intergenic regions of the P. ananatis genome. These included known sRNAs such as ArcZ, FnrS, GlmZ, RprA, RyeB, RyhB, RyhB2, Spot42, and SsrA, and 16 novel P. ananatis sRNAs. The current study demonstrated that Hfq is an important component of the collective regulation of virulence factors and sets a foundation for understanding Hfq-sRNA mediated regulation in the phytopathogen P. ananatis.

Introduction

Pantoea ananatis, formerly described as the pineapple pathogen Erwinia ananas (Serrano, 1928), is a Gram-negative bacterium belonging to the family Enterobacteriaceae. To date, the occurrence of P. ananatis has been reported from various ecological niches spanning both the aquatic and terrestrial environments, including fresh (Morohoshi et al., 2007) and marine water (Jatt et al., 2014) as well as the rhizosphere of crop plants (Oliveira et al., 2008; Marquez-Santacruz et al., 2010). The bacterium exhibits ecologically diverse roles in association with its environment. For example, P. ananatis can be found as an epiphyte of crop and weed plants (Gitaitis et al., 2002) or as an endophyte in maize kernels (Rijavec et al., 2007) and rice seeds (Okunishi et al., 2005). Moreover, the ability of P. ananatis to solubilize phosphate, and produce indole-acetic acid and siderophores, makes the bacterium an ideal plant growth-promoting agent in the production of pepper (Kang et al., 2007), soybean (Kuklinsky-Sobral et al., 2004), and sugarcane (da Silva et al., 2015).

Pantoea ananatis is better known as a phytopathogen affecting the yield of many economically important plant species that causes blight and dieback of Eucalyptus (Coutinho et al., 2002), maize leaf spot disease and brown stalk rot (Goszczynska et al., 2006; Pérez-y-Terrón et al., 2009; Alippi and López, 2010; Krawczyk et al., 2010), leaf blight and bulb rot of onion (Gitaitis and Gay, 1997; Schwartz and Otto, 2000; Goszczynska et al., 2007), palea browning and stem necrosis of rice (Azegami et al., 1983; Cother et al., 2004; Cortesi and Pizzatti, 2007), and fruit rot of netted melon (Kido et al., 2008). P. ananatis has also been considered an emerging plant pathogen due to increasing reports of disease outbreaks in the previously undescribed host and geographical regions (Coutinho and Venter, 2009). This emergence is likely to have resulted from the persistent nature of P. ananatis in diverse environments through its association with a wide range of non-host plant and even insect vectors (Gitaitis et al., 2003; Dutta et al., 2014).

The virulence factors that have been identified as necessary for pathogenesis of P. ananatis in onion are motility for attachment (Weller-Stuart et al., 2016) and quorum sensing (QS) for production of biofilm and exopolysaccharide (EPS) (Morohoshi et al., 2007). In addition, genomic regions named “HiVir” (Asselin et al., 2018) and “Onion Virulence Region” (Stice et al., 2018), encoding enzymes catalyzing phosphonate biosynthetic pathway and cell wall degradation, respectively, have been characterized in the onion pathogenic strains of P. ananatis. For successful infection by P. ananatis, a rapid and collective expression of these virulence genes in response to the surrounding environment is critical as it results in the modulation of cellular pathways that predispose the pathogen for infection, pathogenesis, and survival in the host.

Hfq is an RNA-binding protein that constitutes a key component of post-transcriptional gene regulation exhibited by small non-coding regulatory RNAs (sRNAs) (Vogel and Luisi, 2011). Hfq is a ring-like homohexameric protein that was initially identified as a host factor needed for the replication of RNA bacteriophage Qβ (Franze de Fernandez et al., 1968). It is now known that the chaperone Hfq is essential for the structural stabilization of the class of trans-acting sRNAs whose regulatory mechanisms are dependent on Hfq (Updegrove et al., 2016). The chaperone facilitates imperfect base-pairing between the sRNA and its cognate messenger RNA (mRNA), forming an Hfq–sRNA–mRNA complex that determines the fate of target mRNA translation (Gottesman and Storz, 2011; Storz et al., 2011). Suppression of the protein synthesis is achieved by the formation of a sRNA–mRNA duplex at the 5′-untranslated region (UTR) of the transcript by occlusion of ribosome binding and/or by recruiting ribonucleases for mRNA degradation (De Lay et al., 2013). Conversely, translation of the mRNA is enhanced by Hfq–sRNA complexes that alter the 5′-secondary inhibitory structure of an mRNA, making it more accessible for initiation of translation.

Hfq-dependent sRNAs are typically 50–300 nucleotides in length and are trans-encoded from their cognate mRNAs. They are mostly found in, but not limited to, the intergenic regions of bacterial chromosomes (Argaman et al., 2001; Chao et al., 2012; Guo et al., 2014), and are characterized by often possessing a Rho-independent terminator at the 3′-end, resulting in a poly-uridine tail of sRNA that are recognized by Hfq (Otaka et al., 2011). The cellular functions modulated by Hfq–sRNAs are diverse, ranging from cell membrane integrity, acquisition, and metabolism of nutrients, motility, secretion systems, stress response, and virulence (Chao and Vogel, 2010). Their role in virulence regulation has been extensively studied in bacterial pathogens of animals. For example, in Salmonella typhimurium, motility and expression of the T3SS encoded within Salmonellapathogenicity island SPI-1 and SPI-2 are dependent on Hfq and contribute significantly to the adhesion and invasion of Salmonella into the host cells (Sittka et al., 2007, 2008) whereas in Vibrio cholera, Hfq and its trans acting sRNAs Qrr 1–4 regulate cholera toxin (CT) biosynthesis (Bardill and Hammer, 2012) and QS as an ultrasensitive switch to transition V. cholerae from low to high cell density mode for colonization and disease development (Lenz et al., 2004).

Despite the growing evidence of Hfq and Hfq-dependent sRNAs as a global post-transcriptional gene regulatory complex, the functionality of Hfq and its trans acting sRNAs in plant pathogenic bacteria has only been investigated in a few bacterial species to date, namely in Agrobacterium tumefaciens (Wilms et al., 2012a, b), Burkholderia glumae (Kim et al., 2018), Dickeya dadantii (Yuan et al., 2019), Erwinia amylovora (Zeng et al., 2013; Zeng and Sundin, 2014), Pectobacterium carotovorum (Wang et al., 2018), and Xanthomonas spp. (Schmidtke et al., 2013). Consequently, the functional role of Hfq and the diversity of Hfq-dependent sRNAs in phytopathogens remain largely elusive. We hypothesized that Hfq and Hfq-dependent sRNAs would play a critical role in P. ananatis pathogenesis, through direct regulation of specific virulence traits and through regulation of QS system. In this study, we functionally characterized the role of Hfq as a regulator in the production of acyl-homoserine lactones (AHLs), biofilm development, motility, and virulence, and identified the Hfq-dependent sRNAs that are potentially implicated in the regulation of the virulence traits of the ubiquitous plant pathogen P. ananatis.

Materials and Methods

Bacterial Strains and Growth Conditions

The bacterial strains and plasmids used in this study are listed in Table 1. P. ananatis LMG2665^T and Escherichia coli DH5α strains were cultured in Luria-Bertani (LB) broth [1% (w/v) NaCl, 1% (w/v) tryptone, and 0.5% (w/v) yeast extract; pH 7.2] or on LB agar plates [LB broth amended with 1.5% (w/v) agar; pH 7.2] at 28 and 37°C, respectively. The growth medium was supplemented with either ampicillin (100 μg/ml), chloramphenicol (50 μg/ml), gentamicin (20 μg/ml), or kanamycin (50 μg/ml) for plasmid DNA selection and maintenance.

Generation of a P. ananatis hfq Mutant and Complemented Strains

A mutant strain with chromosomal deletion of a single copy gene hfq (locus tag: PANA_RS17940) was constructed as previously described (Katashkina et al., 2009; Shyntum et al., 2015). The modification was made in the preparation of the knockout cassette which was amplified from the pKD13 plasmid using the Kan-F and Kan-R primers (Table 2) consisting of 50 bp homologous sequences of hfq flanking regions and 20 bp of kanamycin resistance gene priming sequences. Insertion of the kanamycin resistance gene was verified by Southern blotting, PCR amplification, and sequencing of the hfq region.

The promoter sequence of hfq determined in E. coli K12 MG1655 by Kim et al. (2012) was searched against the upstream sequence of hfq start codon in P. ananatis LMG2665^T. An amplicon (1038 bp) containing the hfq gene (315 bp), its native promoter (58 bp), and flanking sequences (662 bp) was cloned into a pBBR1MCS-5_START vector (Obranić et al., 2013) restricted with SmaI and BamHI enzymes. Electrocompetent hfq deletion mutant P. ananatis was transformed with hfq complementing plasmid, pBBR1MCS::hfq and the resulting transformants were selected on the gentamicin amended LB agar. The integrity of the hfq complementation was determined by plasmid extraction, PCR, and sequencing using Test-F and Test-R primers (Table 2).

In vitro and in planta Growth Assay

The growth of wild-type P. ananatis with an empty pBBR1MCS-5_START vector (WT), hfq deletion mutant with an empty pBBR1MCS-5_START vector (Δhfq), and hfq complementing pBBR1MCS-5_START::hfq (pBBR1MCS::hfq) strains of P. ananatis was monitored both in vitro and in planta conditions.

Pantoea ananatis strains grown overnight in LB broth were normalized to an OD_600nm reading of 0.5. For in vitro growth assay, the normalized cultures were diluted 100-fold in fresh LB medium and incubated with shaking at 200 rpm. The absorbency of each culture was periodically measured. There were three replicates for each culture and the experiment was repeated twice.

The previously described red onion scale assay (Stice et al., 2018) was adapted for quantifying in planta growth of P. ananatis between the WT, Δhfq, and hfq complementing strains. In summary, sliced red onion (Allium cepa L) scales of approximately 9 cm² in area were surface sterilized in 3% bleach solution for 1 min and were rinsed twice in distilled water. Each scale was inoculated with 1 μl of bacterial cells (1 × 10⁷ CFU/ml) suspended in 1× PBS [0.8% (w/v) NaCl, 0.02% (w/v) KCl, 0.144% (w/v) NaHPO₄, 0.024% KH₂PO₄; pH 7.4] using a sterile pipette tip. Inoculated scales were placed on moistened paper towels in a surface sterilized container and were incubated at room temperature for 5 days. To quantify growth, three onion scales per strain were harvested at 24 h intervals. Each scale was macerated in 1 ml of 1× PBS and the extract was serially diluted and cells were enumerated on LB supplemented with gentamicin. Experiments were repeated in triplicate, and the results were presented as CFU/g of onion tissue. Sterile water was used as a negative control.

Virulence Assay

Virulence assay was performed as previously described for in planta growth assay. The vertical diameter of the water-soaked lesion on onion scales inoculated with WT, Δhfq, and hfq complementing P. ananatis strains was measured at 3 days post inoculation (dpi). The virulence assay was repeated twice, and there were three technical replicates for each P. ananatis strains.

Motility Assay

Overnight cultures of P. ananatis strains (WT, Δhfq, and pBBR1MCS::hfq) were normalized to OD_600nm = 0.5 and 1 μl of each culture was inoculated in the center of the soft agar [0.5% (w/v) NaCl, 1% (w/v) tyrptone, and 0.3% (w/v) agar; pH 7.2]. The inoculated plates were incubated at 28°C, and swimming motility was determined after 24 h. Negative control plates were inoculated with sterile water. The swimming motility experiment was repeated three times with three biological replicates in each experiment.

Bioassay Detection of Acyl-Homoserine Lactones

Formation of AHL by WT, Δhfq, and hfq complementing strains of P. ananatis was determined using experimental procedures adapted from McClean et al. (1997). An aliquot (0.5 ml) of AHL reporter strain Chromobacterium violaceum (C. violaceum 026) grown in LB overnight was spread plated on LB agar plates and air-dried. Thereafter, three wells were (three replicates) created on each plate by puncturing the agar with a sterile cork-borer and inoculated with 100 μl of cell-free filtrate of P. ananatis WT, Δhfq, and hfq complementing strains overnight cultures. The inoculated plates were incubated at 28°C for 48 h. The formation of violacein (purple halo) by CV026, around the inoculated wells were indicative of AHL production. The assay was repeated twice and carried out in three technical replicates.

Biofilm Quantification

The biofilm of WT, Δhfq, and hfq complementing strains of P. ananatis was quantified as previously described by Santander and Biosca (2017) with slight modifications. An aliquot of 160 μl broth culture diluted to an OD_600nm of 0.5 in half-strength LB [0.5% (w/v) NaCl, 0.5% (w/v) tryptone, and 0.25% (w/v) yeast extract; pH 7.2] was made into each well of a polystyrene 96-well microplate (Nunc^TM MicroWell^TM, Thermo Scientific, Waltham, MA, United States) and incubated for 24 h under static conditions. Eight replicates per P. ananatis strain were included in each experiment with sterile half-strength LB broth serving as a negative control. Thereafter, the inoculated 96-well plates were inverted to remove the excess LB broth, air-dried, and incubated at 60°C for 40 min to heat-fix the biofilms. The biofilms were stained with 1% crystal violet (220 μl) for 15 min before being rinsed with distilled water. After rinsing and invert-air-drying the microplate, 220 μl of ethanol:acetone in 8:2 ratio was added to the wells to solubilize the crystal violet dye for 20 min at room temperature. The solubilized biofilm was measured at OD₆₀₀ using Safire Microplate Reader (Tecan, Research Triangle Park, NC, United States), and this assay was repeated three times.

RNA Extraction and Transcriptomic Analysis

Total RNA of P. ananatis WT and Δhfq strains grown in LB broth was extracted at OD_600nm readings of 0.2 (T1 = low cell density) and 0.6 (T2 = high cell density) using the miRNeasy Mini kit (Qiagen, Hilden, Germany). Genomic DNA was removed by including an on-column DNase digestion step during the RNA extraction. The purity (A260/A280) of extracted RNA was measured by Nanodrop2000 (Thermo Scientific, Sugarland, TX, United States) and RNA integrity was determined by Agilent2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, United States). Illumina Truseq Small RNA Library (Illumina, San Diego, CA, United States) preparation was performed on the RNA samples, and deep sequencing of the library was conducted on Illumina HiSeq2500 platform (single-end, 1 × 50 bp) by Macrogen (South Korea).

Bioinformatic Analysis and sRNA Identification

Raw sequencing reads (BioProject accession number: PRJNA550544) were stringently trimmed and filtered using Trimmomatic (Bolger et al., 2014) to remove adapter sequences and low quality reads. Following adapter trimming and filtering, quality was verified using FastQC (Andrews, 2010) and reads were mapped to the P. ananatis LMG20103 genome (De Maayer et al., 2010) using Bowtie2 (Langmead and Salzberg, 2012), as the genome of LMG 20103 was the only P. ananatis genome with a complete annotation at the time of analysis. For sRNA identification, a custom python script (Supplementary Data Sheet S1, see the section “genic_filter.py” in the Supplementary Material) was compiled to remove reads that mapped to coding sequences, ribosomal RNA, and transfer RNA, or within 120 bases upstream or downstream of these features from the resulting sequence alignment map (SAM) files. The purpose of the 120 base buffer was to reduce the number of sRNAs identified that originated from extended 5′- or 3′-UTR regions. All wild-type sequencing replicates from the same sampling time point were merged into a single gene-filtered SAM file for sRNA identification.

To identify putative sRNAs from gene-filtered SAM files, a custom python script (Supplementary Data Sheet S1, see the section “peak_ID.py” in the Supplementary Material) was used to calculate per base depth relative to the genome-wide per-base sequencing depth by replicate, which was also normalized to library size. A threshold of 10-fold increased abundance above background with a minimum length of 10 nucleotides was chosen for sRNA identification. Using the script, putative sRNAs at the low cell density and high cell density sampling time points were identified and the lists of sRNAs were merged using a custom python script (Supplementary Data Sheet S1, see the section “mergeList.py” in the Supplementary Material), combining any overlapping identified sRNAs into a single sRNA to generate a single list of putative P. ananatis sRNAs (pPARs sRNA).

Computational Prediction of Rho-Independent Terminators

Following established criteria (Zeng and Sundin, 2014), Rho-independent terminators were searched in the P. ananatis LMG20103 genome using a custom python script (Supplementary Data Sheet S1, see the section “RI_term.py” in the Supplementary Material). Briefly, the search was conducted in an effort to detect poly-T regions with at least six continuous Ts and for those that had at least four GC base pairs in the last six bases before the poly-T stretch. Of these, those that had at least 50% GC content in the last 25 bases before the poly-T were considered to be putative Rho-independent terminators.

Differential sRNA Expression in P. ananatis LMG 2665 WT vs. δhfq

Using the genomic coordinates from the BLAST + search of the pPAR sRNAs against the P. ananatis LMG2665 genome (Adam et al., 2014), a gene format file (.gff) for all the pPAR sRNAs was generated. The sRNA sequencing reads that had been trimmed and filtered were mapped to the LMG2665 genome using Bowtie2 (Langmead and Salzberg, 2012). The mapped reads were sorted using SAMtools (Li et al., 2009) and the number of reads mapping to pPAR sRNAs in the LMG2665 genome was counted using HTSeq (Anders et al., 2015). Read counts tables were analyzed for statistically significant differential expression of pPAR sRNAs between WT- and δhfq-mutant samples at corresponding time points using the DESeq R package which utilizes a negative binomial distribution model (Anders and Huber, 2010; R Core Team, 2013). Resulting genes with a false-discovery rate of 0.05 were considered differentially expressed.

sRNA Conservation Analysis

The bacterial genomes were downloaded from NCBI and searched using BLAST+ (Camacho et al., 2009) with all pPAR sRNA sequences as queries. Because BLAST uses local alignment, the global percent identity to pPAR sRNAs was calculated by multiplying the percent identity by the length of the BLAST alignment and dividing by the length of the pPAR sRNA. Heatmaps showing percent identity of sRNA by genome were generated using ClustVis (Metsalu and Vilo, 2015).

qRT-PCR Validation of sRNA Expression

To validate the expression of putative sRNAs identified, a quantitative RT-PCR was conducted on a subset of sRNAs. The 2 μg of total RNA extracted from the two time points (low and high cell density) was converted to cDNA using random primers using the High-Capacity cDNA Synthesis Kit (Applied Biosystems, Carlsbad, CA, United States). Subsequently, PowerUP^TM SYBR^TM Green Master Mix (Applied Biosystems, Carlsbad, CA, United States) was used to quantify expression levels of the selected sRNAs real time in QuantStudio 12K Flex Real-Time PCR System (Applied Biosystems, Carlsbad, CA, United States). The list of primers used for qRT-PCR is found in Table 2. The relative expression of sRNA was calculated using 2^–ΔΔCT method (Livak and Schmittgen, 2001) with the gene ffh encoding a signal recognition particle protein, serving as an endogenous mRNA control (Takle et al., 2007; Sibanda et al., 2018).

5′-Rapid Amplification cDNA Ends Analysis

The 5′-Rapid Amplification cDNA Ends (RACE) analysis was conducted on the selected putative sRNAs to capture their transcription start sites (TSS). Total RNA (up to 15 μg) of P. ananatis strains grown to high density (OD₆₀₀ = 0.6) was extracted as above mentioned (see the section “RNA Extraction and Transcriptomic Analysis”). The resulting RNA was ligated to 300 pmol of RNA linker: GACGAGCACGAGGACACUGACAUGGAGGAGGGAGUAG AAA in the presence of RNA 5′-pyrophosphohydrolase (RppH) (New England BioLabs, Ipswich, MA, United States) and T4 RNA ligase (New England BioLabs, Ipswich, MA, United States) at 37°C for 4 h. The linker-ligated RNA was purified using Trizol-chloroform (2:1) extraction method, as described by Rio et al. (2010). The resulting RNA was ethanol precipitated and suspended in 10 μl of RNase-free water. The cDNA of linker-ligated RNA was synthesized as previously described (see the section “qRT-PCR Validation of sRNA Expression”) and Gene Specific PCR (GSP) was performed using nested linker and sRNA-specific PCR primers (Table 2). The GSP using genomic DNA was used as a control and resulting bands from 5′-RACE were gel-purified and cloned into pJET1.2 blunt (Thermo Fischer Scientific Baltics UAB, Vilnius, Lithuania) prior to sequencing.

Secondary Structure and mRNA Target Prediction

The secondary structures of sRNAs, of which their TSS have been determined by 5′-RACE analysis, were predicted in silico using RNAfold web server¹ (Hofacker, 2003). The putative target mRNAs of novel sRNAs pPAR237, pPAR238, and pPAR395 and their putative interacting domains were computationally predicted using CopraRNA and IntaRNA² (Wright et al., 2014). The above information is presented in Supplementary Figure S6 and Supplementary Table S4, respectively.

Image and Statistical Analysis

Images resulting from motility, AHL detection, and virulence assays were analyzed in ImageJ (Schneider et al., 2012) for measurement of halos and lesion diameter. Statistical analyses are performed with R 3.2.6 (R Core Team, 2013) and significance of the data (P < 0.05) were determined by analysis of variance (ANOVA) and Tukey’s honestly significantly difference (HSD) tests. Except where otherwise mentioned, all data shown in this study represent mean values and error bars represent standard error (SE) of the samples.

Results

hfq Mutation Negatively Affects Growth

To investigate the functional role of Hfq in the pathogenesis of P. ananatis, an hfq deletion mutant (Δhfq) was constructed by replacing the hfq gene with a kanamycin resistance marker (the section “Materials and Methods”). Southern blotting (Supplementary Figure S1) and PCR amplification of the hfq region (Supplementary Figure S2) verified a single insertion of the antibiotic marker in the hfq mutant strain. For the construction of the hfq-complementing plasmid, hfq promoter sequence of E. coli K12 was used to search for hfq promoter in P. ananatis and a highly conserved hfq promoter sequence (93% nucleotide identity) of P. ananatis compared to that of E. coli K12 was found overlapping in the coding region of the adjacent gene miaA.

In vitro growth analyses of P. ananatis WT, Δhfq, and hfq complementing strains cultured in LB medium showed that the hfq mutation affected the growth of P. ananatis. The hfq mutant exhibited a slower growth rate relative to the WT and hfq-complementing strains, but similar cell density was reached at stationary phase as the both strains (Supplementary Figure S3). Similarly, in planta growth curves at 12 h showed that WT, Δhfq, and hfq complementing strains of P. ananatis exhibited comparable cell densities to one another (Supplementary Figure S4B) which were at sufficient levels for the onset of symptoms by 3 dpi (Supplementary Figure S4A).

Loss of Hfq Attenuates Virulence in P. ananatis

Virulence assay on red onion scales demonstrated clearing of the red pigment and formation of a water-soaked lesion in the onion scales inoculated with WT P. ananatis while no disease symptoms were observed on the scales infected with the P. ananatis Δhfq mutant (Figure 1). The impaired virulence of the P. ananatis Δhfq mutant was restored to the wild-type levels by trans expression of hfq gene on the plasmid pBBR1MCS-5_START::hfq. The finding that the P. ananatis Δhfq-mutant strain was able to attain in planta population densities equivalent to WT (Supplementary Figure S4) suggests that the lack of disease symptoms is not associated with a growth defect, and that hfq is required for virulence of this strain when inoculated into red onion.

FIGURE 1

Hfq Regulates Motility, AHL Production, and Biofilm Formation

To determine whether the P. ananatis Hfq regulates virulence traits, swimming motility, production of AHL molecules, and biofilm were quantified in WT, Δhfq, and hfq-complementing strains of P. ananatis. The results show that P. ananatis hfq mutant was impaired in swimming motility relative to the wild-type strain, as determined by the size of the halo that formed on the soft agar (Figure 2). In addition, AHL production, as determined by the production of the purple pigment violacein by the C. violaceum 026 biosensor demonstrated a statistically significant reduction in the size of the purple halo formed by the hfq-mutant strain relative to the wild type, indicating a significant reduction in AHL production by the mutant strain (Figure 3). Furthermore, a threefold reduction (P < 0.05) in the biofilms formed by the hfq-mutant strain relative to the WT P. ananatis was also observed (Figure 4). These findings are consistent with previous studies that showed that AHL molecules are needed as a signal for QS to regulate biofilm formation in P. ananatis (Morohoshi et al., 2007; Sibanda et al., 2016). The phenotypic defects resulting from loss of hfq, which were restored to wild-type levels by trans-complementation of hfq, suggest that Hfq regulates the production of multiple virulence traits in P. ananatis.

FIGURE 2

FIGURE 3

FIGURE 4

Identification of Putative sRNAs

Due to impaired motility, AHL production, biofilm formation, and virulence caused by the loss of Hfq in P. ananatis LMG 2665, a sRNA sequencing analysis was conducted to identify the regulatory sRNAs that are dependent on Hfq for stability and function. Deep sequencing of the sRNA transcriptomes of WT and Δhfq-mutant P. ananatis strains at low (OD₆₀₀ = 0.2) and high cell density (OD₆₀₀ = 0.6) time points resulted in a total of 172.03 million reads. Following trimming of adapters and filtering for high-quality reads (Phred score =30), 66.74 million reads were retained, of which 83.2% mapped uniquely to the P. ananatis LMG20103 genome. Following removal of the reads that mapped to protein coding genes, rRNAs, or tRNAs, 9.72 million reads remained for the sRNA identification and analysis. The distribution of reads across the WT and Δhfq-mutant P. ananatis strains, each with three technical replicates at low and high cell density time points, are included in Supplementary Table S1.

For identification of sRNAs in the transcriptome dataset, the WT sequencing data were utilized and calculated for the per-base depth across the genome relative to the genome-wide average per-base depth. To select a threshold that would allow for sensitive detection of sRNAs while also filtering out noise in the sequencing data, the number of putative sRNAs identified across a broad range of signal-to-noise thresholds was calculated. A strong linear relationship (R² = 0.9981) between Log₁₀ (Threshold) and Log₁₀ (# sRNAs identified) was found (Supplementary Figure S5), and a signal-to-noise threshold ratio of 10 was selected for calling of putative sRNAs from the sequencing data. Using this threshold, a total of 615 pPARs sRNAs was identified. Of these, 425 pPARs were identified in both time points, 90 were identified only in the low cell density time point, and 100 only in the high cell density time point in P. ananatis LMG2665.

Characterization of pPAR sRNAs

The 615 identified pPARs were further classified as intergenic, antisense, or overlapping. The classification resulted in 249 intergenic pPAR sRNAs, 302 antisense pPAR sRNAs, and 64 overlapping pPAR sRNAs (Figure 5A). The mean length of pPAR sRNAs was 66.4 bases with a median of 42 bases (Figure 5B) and mean GC content of pPAR SRNAs was 52.3% with a median of 52.2% (Figure 5C). Both of these are quite close to the genome average of 53.7% GC bases (De Maayer et al., 2010). Of note, seven pPAR sRNAs had GC content below 30%, and 14 pPAR sRNAs had GC content above 70%, suggesting the potential horizontal acquisition of the genomic regions containing these sRNAs.

FIGURE 5

In addition, we performed a genome-wide computational search for putative Rho-independent terminators that are associated with the transcription termination of Hfq-dependent sRNAs (Otaka et al., 2011). The results revealed that there were 5,002 poly-T stretches with at least 6 continuous Ts and 2,437 of these had four or more GC base pairs in the last 6 bases before the poly-T. A total of 1,842 of poly-T stretches had approximately 50% GC content in the final 25 bases before the poly-T, meeting the established criteria of Rho-independent terminators (Zeng and Sundin, 2014). Based on these criteria, only 569 were associated with protein-coding genes and 69 were associated with the identified pPAR sRNAs. The key features of select pPAR sRNAs are presented in Table 3 and for full data, including genomic coordinates and sRNA sequences, refer to Supplementary Table S2.

Identification of Hfq-Dependent sRNAs

Because trans encoded sRNAs have been shown to depend on RNA chaperone proteins such as Hfq for stability and activity (Vogel and Luisi, 2011), an hfq mutant was included in the sRNA sequencing experiment in order to determine pPAR sRNAs that are dependent on or influenced by the loss of hfq. The analysis of hfq to WT samples from both low cell density and high cell density samples identified a total of 276 pPAR sRNAs affected in abundance by Hfq. Sixty-four pPAR sRNAs were affected in abundance by loss of hfq in both cell density samples, 58 pPAR sRNAs in low cell density samples, and 154 pPAR sRNAs only in high cell density samples (Figure 6). Of all the Hfq-dependent pPAR sRNAs, 145 had decreased abundance and 131 had increased abundance in the hfq mutant relative to wild type. Overall, results indicate that Hfq affects the abundances of numerous pPAR sRNAs either positively and/or negatively. Supplementary Table S3 lists all pPAR sRNAs affected in abundance by loss of hfq as well as corresponding fold changes in both low and high cell densities.

FIGURE 6

Of the pPAR sRNAs affected by the loss of hfq, 41 have predicted Rho-independent terminators. Of these, 25 are intergenic and 16 are antisense, consistent with the classical model that Hfq-dependent sRNAs are frequently intergenic (Vogel and Luisi, 2011). Among the sRNAs detected in intergenic regions and Hfq-dependent with Rho-independent terminator, 9 known sRNAs and 16 novel sRNAs were identified. The known sRNAs included ArcZ, FnrS, GlmZ, RprA, RyeB/SdsR, RyhB, RyhB2, Spot42, and SsrA. The depth plots for a number of selected known and novel pPAR sRNAs of interest were generated, showing per-base sequencing depth across the length of the sRNA (Figure 7). Several pPAR sRNAs have certain regions

FIGURE 7

with far greater sequencing depth than the rest of the sRNA which suggests that mechanisms such as post-transcriptional processing (Davis and Waldor, 2007; Chao et al., 2017) might be active in P. ananatis, playing a role in sRNA maturation and/or processing.

Conservation of Identified sRNAs

To visualize the degree of conservation of the identified pPAR sRNAs, a genome-wide analysis was performed to identify sequences similar to those of the pPAR sRNAs for several bacterial species both within and outside the genus Pantoea were conducted. Nearly all of the pPAR sRNAs are highly conserved within the P. ananatis strains, and a large portion of sRNAs was also well conserved within the genus Pantoea (Figure 8A). As some P. ananatis genomes are in draft form, it is possible that some sRNAs have not yet been assembled, accounting for low levels of conservation within the species. However, it will be interesting to determine experimentally if some sRNAs are specific within P. ananatis or within the genus Pantoea as far fewer pPAR sRNAs are conserved across different enterobacterial species (Figure 8B).

FIGURE 8

Experimental Validation and Characterization of Individual sRNAs

Expression of the arcZ, fnrS, glmZ, rprA, ryeB, ryhB2, pPAR237, pPAR238, and pPAR395 sRNA genes was quantified in the P. ananatis hfq-mutant strain relative to WT using qRT-PCR (Figure 9). The resulting expression profile of aforementioned sRNA transcript levels (except glmZ and ryhB2) was decreased in the absence of hfq, which was in agreement with the depth plots analysis (Figure 7). In WT P. ananatis, glmZ expression is likely repressed at low cell density (OD₆₀₀ = 0.2) and increased at high cell density (OD₆₀₀ = 0.6) in a Hfq-dependent manner as the opposite expression levels were observed in hfq-mutant P. ananatis where abundances of glmZ transcripts were detected at low cell density but not at high cell density condition. Similarly, Hfq may negatively affect ryhB2 expression, as the abundances of ryhB2 transcript in WT at low and high cell density conditions were both low relative to the hfq-mutant ryhB2 levels. The TSS of FnrS, GlmZ, pPAR237, pPAR238, and pPAR395 was determined by 5′-RACE analysis. Their predicted structures, sequence, and targets are reported in Supplementary Figure S6 and Supplementary Table S4.

FIGURE 9

Discussion

In the present study, we investigated the functional role of Hfq in the pathogenesis of the Gram-negative phytopathogen P. ananatis, and demonstrated that Hfq is important for motility, AHL and biofilm formation, and virulence of the pathogen. We also identified several putative sRNAs, which include known and novel sRNAs that are Hfq-dependent for their abundances in P. ananatis. The pleiotropic phenotypes caused by hfq mutation is due to global post-transcriptional gene regulation operated by Hfq and Hfq-dependent sRNAs that modulate stress response and virulence of numerous bacterial pathogens (Chao and Vogel, 2010). The ability of P. ananatis to survive in diverse ecological niches and to successfully infect susceptible plant hosts requires a timely and collective regulation of cellular functions in response to environmental conditions.

Inactivation of hfq in bacteria generally results in pleiotropic effects, of which growth retardation is common. Decreased growth rate has been reported in hfq-attenuated bacteria such as Acinetobacter baumannii (Kuo et al., 2017), Haemophilus influenzae (Hempel et al., 2013), Yersinia enterocolitica (Kakoschke et al., 2016), and the plant pathogens A. tumefaciens (Wilms et al., 2012a) and P. carotovorum (Wang et al., 2018). This phenotype was consistent with the P. ananatis hfq deletion mutant growing in vitro; however, this alteration did not prevent P. ananatis from entering logarithmic growth phase and eventually reaching the wild-type cell density at a stationary phase which was also observed in planta (Supplementary Figure S4). Unlike the E. amylovora hfq mutant (Zeng et al., 2013), which exhibited reduced growth in an immature pear fruit infection model, the P. ananatis hfq mutant strain was able to reach a population density comparable to that of the wild-type strain when inoculated into onion, indicating that the abolishment of virulence in the hfq-mutant P. ananatis was not due to a growth defect.

The loss of hfq gives rise to impairment of important virulence determinants such as motility in bacterial pathogens. Impaired motility affects the overall fitness of a bacterium as a pathogen as it disables attachment and dispersal of the pathogen in the host. This, in turn, results in the diminished invasion, colonization, and hence virulence (Sittka et al., 2007; Kulesus et al., 2008). In enterobacterial pathogens, Hfq and Hfq-dependent sRNAs control flagellar-based motility. For example, in E. coli, multiple sRNAs including ArcZ, OmrAB, OxyS, and RyeB/SdsR have been shown to modulate the expression and/or translation of flhDC, the master regulator of flagellar biosynthesis (De Lay and Gottesman, 2012). In the phytopathogen, E. amylovora, the sRNAs ArcZ, OmrAB, and RmaA have been found to regulate flhDC at both transcriptional and post-transcriptional levels (Schachterle and Sundin, 2019; Schachterle et al., 2019). In this way, the integration of different environmental cues is achieved through several sRNAs, allowing fine-tuning of flagellar expression and production.

The lack of swimming motility displayed by the P. ananatis hfq mutant clearly demonstrates the role of Hfq in regulating flagellar motility. In a previous study by Weller-Stuart et al. (2016), a P. ananatis flgK mutant deficient in flagellar assembly enzyme, FlgK, was abolished in swimming motility and pathogenicity in onion seedlings. Together with the current study, these findings suggest that flagellar motility is required for the virulence of P. ananatis, and this trait is regulated by functional Hfq. Similarly, in P. carotovorum (Wang et al., 2018) and Serratia sp. ATCC 39006 (Wilf et al., 2013; Hampton et al., 2016), attenuation of flagellar motility was observed in hfq-deletion mutant strains, as an expression of the flhDC genes was dependent on Hfq. Given that sRNAs namely, ArcZ, OmrAB, and RyeB/SdsR, were also identified in our sRNA sequencing data (Supplementary Table S2), and that their wild-type transcript levels are dependent on the functional copy of hfq (Figure 9), we hypothesize that Hfq, in conjunction with the identified sRNAs, may regulate the flagellar motility of P. ananatis in a similar manner as the other enterobacterial species.

In addition to impaired motility, disruption of hfq in Gram-negative pathogens often results in reduced biofilm formation (Kulesus et al., 2008; Monteiro et al., 2012; Zeng et al., 2013; Wang et al., 2018). One possible explanation for this phenotype is an effect on QS-mediated regulation of motility and biofilm formation. As biofilm formation is a developmental and co-operative process, the process necessitates cell to cell communication that enables perception of the signals generated from the community. The signal or information is packaged in the form of autoinducer molecules, acylated homoserine lactones (AHLs), or may be communicated to the QS circuit by secondary signaling molecule such as cyclic dimeric guanosine monophosphate (c-di-GMP) (Castiblanco and Sundin, 2016). Through protein phospho-relay, signals resulting from high cell density reach Hfq-dependent sRNAs which initiate the expression of genes required for the extracellular matrix synthesis and maturation of biofilm (Lenz et al., 2004; Kay et al., 2006; Tu and Bassler, 2007).

Consistent with the findings that in some bacteria, Hfq positively regulates biofilm production, a significantly reduced amount of biofilm was formed by the P. ananatis strain lacking an hfq gene. Interestingly, a decreased level of extracellular AHLs diffused in the supernatant of an hfq mutant overnight culture was detected by the CV026 biosensor, suggesting a potential role of Hfq in the positive regulation of AHL synthesis. According to Pomini et al. (2006), the main P. ananatis-derived AHL molecule is N-hexaonoyl-L-homoserine lactone (C6-HSL) and is synthesized by the LuxI homolog EanI, while LuxR homolog EanR is a transcriptional regulator that down-regulates the expression of eanIR operon in the absence of AHL (Morohoshi et al., 2007). Functional characterization further demonstrated that eanI or AHL synthase was required for the formation of biofilm, EPS, and pathogenicity of P. ananatis (Morohoshi et al., 2007; Sibanda et al., 2016). Thus, decreased production of AHL in the hfq-mutant strain of P. ananatis would mean down-regulation of eanI, virulence-related genes, as well as the genes within the QS regulon (Sibanda et al., 2018).

In the present study, two putative novel Hfq-dependent sRNAs were identified in the vicinity of the eanIR genes namely, pPAR237 and pPAR238 (Supplementary Figure S7A). These sRNAs are partially antisense to each other and are abundantly present in the low cell density condition but are drastically reduced in abundance at high cell density (Figure 9). In contrast to the wild-type expression levels, expression of pPAR237 and pPAR238 was almost non-existent in the two cell density conditions in the P. ananatis hfq-mutant strain (Figure 9). The decreased transcript levels of pPAR237 and pPAR238 in hfq-mutant relative to WT P. ananatis were validated experimentally by qRT-PCR, reinforcing the idea that the expression of these sRNAs is dependent on cell density and Hfq. Potential pairing sites of pPAR237 and pPAR238 to eanI and eanR were predicted in silico using IntaRNA (Supplementary Figures S7B–D). Further experimental confirmation of their interaction will indicate the role of pPAR237 and pPAR238 in QS through modulation of AHL synthesis in P. ananatis. Moreover, it will be also interesting to determine whether there are other upstream and downstream transcriptional or translational regulators of putative sRNAs pPAR237 and pPAR238. This is the case in V. cholera and Pseudomonas aeruginosa whose Hfq-dependent sRNA Qrr 1,2,3, and 4 and RsmY are transcriptionally activated by LuxO and GacA, respectively, and are used to repress transcription of hapR or sequester translational regulator RsmA (Lenz et al., 2004; Kay et al., 2006; Tu and Bassler, 2007; Brencic et al., 2009).

To date, factors that contribute to the pathogenicity of P. ananatis have been characterized, resulting in an expansion in our understanding of virulence mechanisms of this pathogen. A collective regulation of all virulence traits seems likely for the success and persistence of P. ananatis in hostile environments, and this can be achieved through Hfq and its global regulatory networks constituted by Hfq-dependent sRNAs. Overall, this study provided valuable insights into the essential role of Hfq in regulating different virulence traits of P. ananatis. A total of 276 sRNAs were identified that are affected in abundance by Hfq at low and high cell density conditions. These sRNAs include those that are well characterized as well as novel putative sRNAs that may possess novel function involved in the QS of P. ananatis.

References

• 1

  AdamZ.TambongJ. T.LewisC. T.LévesqueC. A.ChenW.BromfieldE. S.et al (2014). Draft genome sequence of Pantoea ananatis strain LMG 2665T, a bacterial pathogen of pineapple fruitlets.Genome Announc.2:e489-14. 10.1128/genomeA.00489-14

  • CrossRef
  • Google Scholar

• 2

  AlippiA. M.LópezA. C. (2010). First report of leaf spot disease of maize caused by Pantoea ananatis in Argentina.Plant Dis.94:487. 10.1094/PDIS-94-4-0487A

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 3

  AndersS.HuberW. (2010). Differential expression analysis for sequence count data.Genome Biol.11:R106. 10.1186/gb-2010-11-10-r106

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4

  AndersS.PylP. T.HuberW. (2015). HTSeq—a Python framework to work with high-throughput sequencing data.Bioinformatics31166–169. 10.1093/bioinformatics/btu638

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5

  AndrewsS. (2010). FastQC: A Quality Control Tool for High Throughput Sequence Data. Available at: http://www.bioinformatics.babraham.ac.uk/projects/fastqc(accessed May 21, 2018).

  • Google Scholar

• 6

  ArgamanL.HershbergR.VogelJ.BejeranoG.WagnerE. G.MargalitH.et al (2001). Novel small RNA-encoding genes in the intergenic regions of Escherichia coli.Curr. Biol.11941–950. 10.1016/S0960-9822(01)00270-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 7

  AsselinJ. E.BonaseraJ. M.BeerS. V. (2018). Center rot of onion (Allium cepa) caused by Pantoea ananatis requires pepM, a predicted phosphonate-related gene.Mol. Plant Microbe Interact.311291–1300. 10.1094/MPMI-04-18-0077-R

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 8

  AzegamiK.OzakiK.MatsudaA.OhataK. (1983). Bacterial palea browning, a new disease of rice caused by Erwinia herbicola (in Japanese with English summary).Bull. Natl. Inst. Agric. Sci. Ser. C371–12.

  • Google Scholar

• 9

  BardillJ. P.HammerB. K. (2012). Non-coding sRNAs regulate virulence in the bacterial pathogen Vibrio cholerae.RNA Biol.9392–401. 10.4161/rna.19975

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10

  BolgerA. M.LohseM.UsadelB. (2014). Trimmomatic: a flexible trimmer for Illumina sequence data.Bioinformatics302114–2120. 10.1093/bioinformatics/btu170

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 11

  BrencicA.McFarlandK. A.McManusH. R.CastangS.MognoI.DoveS. L.et al (2009). The GacS/GacA signal transduction system of Pseudomonas aeruginosa acts exclusively through its control over the transcription of the RsmY and RsmZ regulatory small RNAs.Mol. Microbiol.73434–445. 10.1111/j.1365-2958.2009.06782.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 12

  CamachoC.CoulourisG.AvagyanV.MaN.PapadopoulosJ.BealerK.et al (2009). BLAST+: architecture and applications.BMC Bioinformatics10:421. 10.1186/1471-2105-10-421

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 13

  CastiblancoL. F.SundinG. W. (2016). New insights on molecular regulation of biofilm formation in plant-associated bacteria.J. Integr. Plant Biol.58362–372. 10.1111/jipb.12428

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 14

  ChaoY.LiL.GirodatD.FörstnerK. U.SaidN.CorcoranC.et al (2017). In vivo cleavage map illuminates the central role of RNase E in coding and non-coding RNA pathways.Mol. Cell6539–51. 10.1016/j.molcel.2016.11.002

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 15

  ChaoY.PapenfortK.ReinhardtR.ScharmaC. M.VogelJ. (2012). An atlas of Hfq-bound transcripts reveals 3’ UTRs as a genomic reservoir of regulatory small RNAs.EMBO J.314005–4019. 10.1038/emboj.2012.229

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 16

  ChaoY.VogelJ. (2010). The role of Hfq in bacterial pathogens.Curr. Opin. Microbiol.1324–33. 10.1016/j.mib.2010.01.001

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 17

  CortesiP.PizzattiC. (2007). Palea browning, a new disease of rice in Italy caused by Pantoea ananatis.J. Plant Pathol.89:S76.

  • Google Scholar

• 18

  CotherE. J.ReinkeR.McKenzieC.LanoiseletV. M.NobleD. H. (2004). An unusual stem necrosis of rice caused by Pantoea ananas and the first record of this pathogen on rice in Australia.Aust. Plant Pathol.33495–503. 10.1071/AP04053

  • CrossRef
  • Google Scholar

• 19

  CoutinhoT. A.PreisigO.MergaertJ.CnockaertM. C.RiedelK.-H.SwingsJ.et al (2002). Bacterial blight and die-back of Eucalyptus species, hybrids and clones in South Africa.Plant Dis.8620–25. 10.1094/PDIS.2002.86.1.20

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 20

  CoutinhoT. A.VenterS. N. (2009). Pantoea ananatis: an unconventional plant pathogen.Mol. Plant Pathol.10325–335. 10.1111/j.1364-3703.2009.00542.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 21

  da SilvaJ. F.BarbosaR. R.de SouzaA. N.da MottaO. V.TeixeiraG. N.CarvalhoV. S.et al (2015). Isolation of Pantoea ananatis from sugarcane and characterization of its potential for plant growth promotion.Genet. Mol. Res.1415301–15311. 10.4238/2015.November.30.6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 22

  DatsenkoK. A.WannerB. L. (2000). One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products.Proc. Natl. Acad. Sci. U.S.A976640–6645. 10.1073/pnas.120163297

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 23

  DavisB. M.WaldorM. K. (2007). RNase E-dependent processing stabilizes MicX, a Vibrio cholerae sRNA.Mol. Microbiol.65373–385. 10.1111/j.1365-2958.2007.05796.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 24

  De LayN.GottesmanS. (2012). A complex network of small non-coding RNAs regulate motility in Escherichia coli.Mol. Microbiol.86524–538. 10.1111/j.1365-2958.2012.08209.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 25

  De LayN.SchuD. J.GottesmanS. (2013). Bacterial small RNA-based negative regulation: Hfq and its accomplices.J. Biol. Chem.2887996–8003. 10.1074/jbc.R112.441386

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 26

  De MaayerP.ChanW. Y.VenterS. N.TothI. K.BirchP. R.JoubertF.et al (2010). Genome sequence of Pantoea ananatis LMG20103, the causative agent of Eucalyptus blight and dieback.J. Bacteriol.1922936–2937. 10.1128/JB.00060-10

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27

  DuttaB.BarmanA. K.SrinivasanR.AvciU.UllmanD. E.LangstonD. B.et al (2014). Transmission of Pantoea ananatis and P. agglomerans, causal agents of center rot of onion (Allium cepa), by onion thrips (Thrips tabaci) through feces.Phytopathology104812–819. 10.1094/PHYTO-07-13-0199R

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 28

  Franze de FernandezM. T.EoyangL.AugustJ. T. (1968). Factor fraction required for the synthesis of bacteriophage Qβ-RNA.Nature219588–590. 10.1038/219588a0

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29

  GitaitisR.WalcottR.CulpepperS.SandersH.ZolobowskaL.LangstonD. (2002). Recovery of Pantoea annatis, causal agent of center rot of onion, from weeds and crops in Georgia, USA.Crop Prot.21983–989. 10.1016/S0261-2194(02)00078-9

  • CrossRef
  • Google Scholar

• 30

  GitaitisR. D.GayJ. D. (1997). First report of leaf blight, seed stalk rot, and bulb decay of onion by Pantoea ananas in Georgia.Plant Dis.81:1096. 10.1094/PDIS.1997.81.9.1096C

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31

  GitaitisR. D.WalcottR. R.WellsM. L.Diaz PerezJ. C.SandersF. H. (2003). Transmission of Pantoea ananatis, causal agent of center rot of onion, by tobacco thrips, Frankliniella fusca.Plant Dis.87675–678. 10.1094/PDIS.2003.87.6.675

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32

  GoszczynskaT.MolotoV. M.VenterS. N.CoutinhoT. A. (2006). Isolation and identification of Pantoea ananatis from onion seed in South Africa.Seed. Sci. Technol.34655–668. 10.15258/sst.2006.34.3.12

  • CrossRef
  • Google Scholar

• 33

  GoszczynskaT.VenterS. N.CoutinhoT. A. (2007). Isolation and identification of the causal agent of brown stalk rot, a new disease of corn in South Africa.Plant Dis.91711–718. 10.1094/PDIS-91-6-0711

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34

  GottesmanS.StorzG. (2011). Bacterial small RNA regulators: versatile roles and rapidly evolving variations.RNA21511–512. 10.1261/rna.050047.115

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 35

  GuoM. S.UpdegroveT. B.GogolE. B.ShabalinaS. A.GrossC. A.StorzG. (2014). MicL, a new σE –dependent sRNA, combats envelope stress by repressing synthesis of Lpp, the major outer membrane lipoprotein.Genes Dev.281620–1634. 10.1101/gad.243485.114

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 36

  HamptonH. G.McNeilM. B.PatersonT. J.NeyB.WilliamsonN. R.EasingwoodR. A.et al (2016). CRISPR-Cas gene-editing reveals RsmA and RsmC act through FlhDC to repress the SdhE flavinylation factor and control motility and prodigiosin production in Serratia.Microbiology1621047–1058. 10.1099/mic.0.000283

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 37

  HempelR. J.MortonD. J.SealeT. W.WhitbyP. W.StullT. L. (2013). The role of the RNA chaperone Hfq in Haemophilus influenzae pathogenesis.BMC Microbiol.13:134. 10.1186/1471-2180-13-134

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 38

  HofackerI. L. (2003). Vienna RNA secondary structure server.Nucleic Acids. Res.313429–3431. 10.1093/nar/gkg599

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 39

  JattA. B.TangK.LiuJ.ZhangZ.ZhangX.-H. (2014). Quorum sensing in marine snow and its possible influence on production of extracellular hydrolytic enzymes in marine snow bacterium Pantoea ananatis B9.FEMS Microbiol. Ecol.911–13. 10.1093/femsec/fiu030

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 40

  KakoschkeT. K.KakoschkeS. C.ZeuzemC.BouabeH.AdlerK.HeesemannK.et al (2016). The RNA chaperone Hfq is essential for virulence and modulates the expression of four adhesins in Yersinia enterocolitica.Sci. Rep.6:29275. 10.1038/srep29275

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 41

  KangS. H.ChoH.-S.CheongH.RyuC.-M.KimJ. F.ParkS.-H. (2007). Two bacterial endophytes eliciting both plant growth promotion and plant defense on pepper (Capsicum annuum L.).J. Microbiol. Biotechnol.1796–103.

  • Pubmed Abstract
  • Google Scholar

• 42

  KatashkinaJ. I.HaraY.GolubevaL. I.AndreevaI. G.KuvaevaT. M.MashkoS. V. (2009). Use of the λ Red-recombineering method for genetic engineering of Pantoea ananatis.BMC Mol. Biol.10:34. 10.1186/1471-2199-10-34

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 43

  KayE.HumairB.DenervaudB.RiedelK.SpahrS.EberlL.et al (2006). Two GacA-dependent small RNAs modulate the quorum-sensing response in Pseudomonas aeruginosa.J. Bacteriol.1886026–6033. 10.1128/JB.00409-06

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 44

  KidoK.AdachiR.HasegawaM.YanoK.HikichiY.TakeuchiS.et al (2008). Internal fruit rot of netted melon caused by Pantoea ananatis (=Erwinia ananas) in Japan.J. Gen. Plant Pathol.74302–312. 10.1007/s10327-008-0107-3

  • CrossRef
  • Google Scholar

• 45

  KimD.HongJ. S.QuiY.NagarajanH.SeoJ.-H.ChoB.-K.et al (2012). Comparative analysis of regulatory elements between Escherichia coli and Klebsiella pneumoniae by genome-wide transcription start site profiling.PLoS Genet.8:e1002867. 10.1371/journal.pgen.1002867

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 46

  KimJ.MannaaM.KimN.LeeC.KimJ.ParkJ.et al (2018). The roles of two hfq genes in the virulence and stress resistance of Burkholderia glumae.Plant Pathol. J.34412–425. 10.5423/PPJ.OA.06.2018.0097

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 47

  KrawczykK.KamasaJ.ZwolinskaA.PospiesznyH. (2010). First report of Pantoea ananatis associated with white leaf spot disease of maize in Poland.J. Plant Pathol.92807–811.

  • Google Scholar

• 48

  Kuklinsky-SobralJ.AraújoW. L.MendesR.GeraldiI. O.Pizzirani-KlelnerA. A.AzevedoJ. L. (2004). Isolation and characterization of soybean-associated bacteria and their potential for plant growth promotion.Environ. Microbiol.61244–1251. 10.1111/j.1462-2920.2004.00658.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 49

  KulesusR. R.Diaz-PerezK.SlechtaE. S.EtoD. S.MulveyM. A. (2008). Impact of the RNA chaperone Hfq on the fitness and virulence potential of uropathogenic Escherichia coli.Infect. Immun.763019–3026. 10.1128/IAI.00022-08

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 50

  KuoH.-Y.ChaoH.-H.LiaoP.-C.HsuL.ChangK.-C.TungC.-H.et al (2017). Functional characterization of Acinetobacter baumannii lacking the RNA chaperone Hfq.Front. Microbiol.8:2068. 10.3389/fmicb.2017.02068

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 51

  LangmeadB.SalzbergS. L. (2012). Fast gapped-read alignment with Bowtie 2.Nat. Methods9:357. 10.1038/nmeth.1923

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 52

  LenzD. H.MokK. C.LilleyB. N.KulkarniR. V.WingreenN. S.BasslerB. L. (2004). The small RNA chaperone Hfq and multiple small RNAs control quorum sensing in Vibrio harveyi and Vibrio cholerae.Cell11869–82. 10.1016/j.cell.2004.06.009

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 53

  LiH.HandsakerB.WysokerA.FennellT.RuanJ.HomerN.et al (2009). The sequence alignment/map format and SAMtools.Bioinformatics252078–2079. 10.1093/bioinformatics/btp352

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 54

  LivakK. J.SchmittgenT. D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2-ΔΔCT method.Methods25402–408. 10.1006/meth.2001.1262

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 55

  Marquez-SantacruzH. A.Hernandez-LeonR.Orozco-MosquedaM. C.Velazquez-SepulvedalI.SantoyoG. (2010). Diversity of bacterial endophytes in roots of Mexican husk tomato plants (Physalis ixocarpa) and their detection in the rhizosphere.Genet. Mol. Res.92372–2390. 10.4238/vol9-4gmr921

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 56

  McCleanK. H.WinsonM. K.FishL.TaylorA.ChhabraS. R.CamaraM.et al (1997). Quorum sensing and Chromobacterium violaceum: exploitation of violacein production and inhibition for the detection of N-acylhomoserine lactones.Microbiology1433703–3711. 10.1099/00221287-143-12-3703

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 57

  MetsaluT.ViloJ. (2015). ClustVis: a web tool for visualizing clustering of multivariate data using principal component analysis and heatmap.Nucleic Acids Res.43W566–W570. 10.1093/nar/gkv468

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 58

  MonteiroC.PapenfortK.HentrichK.AhmadI.Le GuyonS.ReimannR.et al (2012). Hfq and Hfq-dependent small RNAs are major contributors to multicellular development in Salmonella enterica Serovar typhimurium.RNA Biol.9489–502. 10.4161/rna.19682

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 59

  MorohoshiT.NakamuraY.YamazakiG.IshidaA.KatoN.IkedaT. (2007). The plant pathogen Pantoea ananatis produces N-acylhomoserine lactone and causes center rot disease of onion by quorum sensing.J. Bacteriol.1898333–8338. 10.1128/JB.01054-07

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 60

  ObranićS.BabićF.Maravić-VlahovičekG. (2013). Improvement of pBBR1MCS plasmids, a very useful series of broad-host-range cloning vectors.Plasmid70263–267. 10.1016/j.plasmid.2013.04.001

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 61

  OkunishiS.SakoK.ManoH.ImamuraA.MorisakiH. (2005). Bacterial flora of endophytes in the maturing seed of cultivated rice (Oryza sativa).Microbes Environ.20168–177. 10.1264/jsme2.20.168

  • CrossRef
  • Google Scholar

• 62

  OliveiraC. A.AlvesV. M. C.MarrielI. E.GomesE. A.ScotiiM. R.CarneiroN. P.et al (2008). Phosphate solubilizing microorganisms isolated from rhizosphere of maize cultivated in an oxisol of the Brazilian Cerrado Biome.Soil Biol. Biochem.411782–1787. 10.1016/j.soilbio.2008.01.012

  • CrossRef
  • Google Scholar

• 63

  OtakaH.IshikawaH.MoritaT.AibaH. (2011). PolyU tail of rho-independent terminator of bacterial small RNAs is essential for Hfq action.Proc. Natl. Acad. Sci. U.S.A.10813059–13064. 10.1073/pnas.1107050108

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64

  Pérez-y-TerrónR.VillegasM. C.CuellarA.Muñoz-RojasJ.Castañeda-LucioM.Hernández-LucasI.et al (2009). Detection of Pantoea ananatis, causal agent of leaf spot disease of maize, in Mexico.Aust. Plant Dis. Notes496–99.

  • Google Scholar

• 65

  PominiA. M.AraújoW. L.MarsaioliA. J. (2006). Structural elucidation and biological activity of acyl-homoserine lactones from the phytopathogen Pantoea ananatis Serrano, 1928.J. Chem. Ecol.321769–1778. 10.1007/s10886-006-9108-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 66

  R Core Team (2013). R: A Language and Environment for Statistical Computing.Vienna: R Foundation for Statistical Computing.

  • Google Scholar

• 67

  RijavecT.LapanjeA.DermastiaM.RupnikM. (2007). Isolation of bacterial endophytes from germinated maize kernels.Can. J. Microbiol.53802–808. 10.1139/W07-048

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 68

  RioD. C.AresM.Jr.HannonG. J.NilsenT. W. (2010). Purification of RNA using TRIzol (TRI reagent).Cold Spring Harb. Protoc.6:pdb.prot5439. 10.1101/pdb.prot5439

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 69

  SantanderR. D.BioscaE. G. (2017). Erwinia amylovora psychrotrophic adaptations: evidence of pathogenic potential and survival at temperate and low environmental temperatures.PeerJ5:e3931. 10.7717/peerj.3931

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 70

  SchachterleJ. K.SundinG. W. (2019). The leucine-responsive regulatory protein participates in virulence regulation downstream of small RNA ArcZ in Erwinia amylovora.mBio10:e0757-19. 10.1128/mBio.00757-19

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 71

  SchachterleJ. K.ZengQ.SundinG. W. (2019). Three Hfq-dependent small RNAs regulate flagellar motility in the fire blight pathogen Erwinia amylovora.Mol. Microbiol.1111476–1492. 10.1111/mmi.14232

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 72

  SchmidtkeC.AbendrothU.BrockJ.SerraniaJ.BeckerA.BonasU. (2013). Small RNA sX13: a multifaceted regulator of virulence in the plant pathogen Xanthomonas.PLoS Pathog.9:e1003626. 10.1371/journal.ppat.1003626

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 73

  SchneiderC. A.RasbandW. S.EliceiriK. W. (2012). NIH Image to ImageJ: 25 years of image analysis.Nat. Methods9671–675. 10.1038/nmeth.2089

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 74

  SchwartzH. F.OttoK. (2000). First report of a leaf blight and bulb decay of onion by Pantoea ananatis in Colorado.Plant Dis.84:808. 10.1094/PDIS.2000.84.7.808A

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 75

  SerranoF. B. (1928). Bacterial fruitlet brown-rot of pineapple in the Philippines.Philipp. J. Sci.36271–324.

  • Google Scholar

• 76

  ShyntumD. Y.TheronJ.VenterS. N.MolelekiL. N.TothI. K.CoutinhoT. A. (2015). Pantoea ananatis utilizes a type VI secretion system for pathogenesis and bacterial competition.Mol. Plant Microbe Interact.28420–431. 10.1094/MPMI-07-14-0219-R

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 77

  SibandaS.KwendaS.TanuiC. K.ShyntumD. Y.CoutinhoT. A.MolelekiL. N. (2018). Transcriptome profiling reveals the EanI/R quorum sensing regulon in Pantoea ananatis LMG 2665T.Genes91–17. 10.3390/genes9030148

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 78

  SibandaS.TheronJ.ShyntumD. Y.MolelekiL. N.CoutinhoT. A. (2016). Characterization of two LuxI/R homologs in Pantoea ananatis LMG 2665T.Can. J. Microbiol.62893–903. 10.1139/cjm-2016-0143

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 79

  SittkaA.LucchiniS.PapenfortK.SharmaC. M.RolleK.BinnewiesT. T.et al (2008). Deep sequencing analysis of small noncoding RNA and mRNA targets of the global post-transcriptional regulator, Hfq.PLoS Genet.4:e1000163. 10.1371/journal.pgen.1000163

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 80

  SittkaA.PfeifferV.TedinK.VogelJ. (2007). The RNA chaperone Hfq is essential for the virulence of Salmonella typhimurium.Mol. Microbiol.63193–217. 10.1111/j.1365-2958.2006.05489.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 81

  SticeS. P.StumpfD. S.GitaitisR. D.KvitkoB. H.DuttaB. (2018). Pantoea ananatis genetic diversity analysis reveals limited genomic diversity as well as accessory genes correlated with onion pathogenicity.Front. Microbiol.9:184. 10.3389/fmicb.2018.00184

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 82

  StorzG.VogelJ.WassarmanK. M. (2011). Regulation by small RNAs in bacteria. Expanding frontiers.Mol. Cell43880–891. 10.1016/j.molcel.2011.08.022

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 83

  TakleG. W.TothI. K.BrurbergM. B. (2007). Evaluation of reference genes for real-time RT-PCR expression studies in the plant pathogen Pectobacterium atrosepticum.BMC Plant Biol.7:50. 10.1186/1471-2229-7-50

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 84

  TuK. C.BasslerB. L. (2007). Multiple small RNAs act additively to integrate sensory information and control quorum sensing in Vibrio harveyi.Genes Dev.21221–233. 10.1101/gad.1502407

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 85

  UpdegroveT. B.ZhangA.StorzG. (2016). Hfq: the flexible RNA matchmaker.Curr. Opin. Microbiol.30133–138. 10.1016/j.mib.2016.02.003

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 86

  VogelJ.LuisiB. F. (2011). Hfq and its constellation of RNA.Nat. Rev. Microbiol.9578–589. 10.1038/nrmicro2615

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 87

  WangC.PuT.LouW.WangY.GaoZ.HuB.et al (2018). Hfq, a RNA chaperone, contributes to virulence by regulating plant cell wall–degrading enzyme production, type VI secretion system expression, bacterial competition, and suppressing host defense response in Pectobacterium carotovorum.Mol. Plant Microbe Interact.311166–1178. 10.1094/MPMI-12-17-0303-R

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 88

  Weller-StuartT.TothI.De MaayerP.CoutinhoT. A. (2016). Swimming and twitching motility are essential for attachment and virulence of Pantoea ananatis in onion seedlings.Mol. Plant Pathol.18734–745. 10.1111/mpp.12432

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 89

  WilfN. M.ReidA. J.RamsayJ. P.WilliamsonN. R.CroucherN. J.GattoL.et al (2013). RNA-seq reveals the RNA binding proteins, Hfq and RsmA, play various roles in virulence, antibiotic production and genomic flux in Serratia sp. ATCC 39006.BMC Genomics14:822. 10.1186/1471-2164-14-822

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 90

  WilmsI.MöllerP.StockA.GurskiR.LaiE.NaberhausF. (2012a). Hfq influences multiple transport systems and virulence in the plant pathogen Agrobacterium tumefaciens.J. Bacteriol.1945209–5217. 10.1128/JB.00510-12

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 91

  WilmsI.OverlöperA.NowrousianM.SharmaC. M.NarberhausF. (2012b). Deep sequencing uncovers numerous small RNAs on all four replicons of the plant pathogen Agrobacterium tumefaciens.RNA Biol.9446–457. 10.4161/rna.17212

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 92

  WrightP. R.GeorgJ.MannM.SorescuD. A.RichterA. S.LottS.et al (2014). CopraRNA and IntaRNA: predicting small RNA targets, networks and interaction domains.Nucleic Acids Res.42119–123. 10.1093/nar/gku359

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 93

  YuanX.ZengQ.KhokhaniD.TianF.SeverinG. B.WatersC. M.et al (2019). A feed-forward signalling circuit controls bacterial virulence through linking cyclic-di-GMP and two mechanistically distinct sRNAs; ArcZ and RsmB.Environ. Microbiol.212755–2771. 10.1111/1462-2920.14603

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 94

  ZengQ.McNallyR. R.SundinG. W. (2013). Global small RNA chaperone Hfq and regulatory small RNAs are important virulence regulators in Erwinia amylovora.J. Bacteriol.1951706–1717. 10.1128/JB.02056-12

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 95

  ZengQ.SundinG. W. (2014). Genome-wide identification of Hfq-regulated small RNAs in the fire blight pathogen Erwinia amylovora discovered small RNAs with virulence regulatory function.BMC Genomics15:414. 10.1186/1471-2164-15-414

  • Pubmed Abstract
  • CrossRef
  • Google Scholar