A novel small RNA is important for biofilm formation and pathogenicity in Pseudomonas aeruginosa

The regulation of biofilm development requires multiple mechanisms and pathways, but it is not fully understood how these are integrated. Small RNA post-transcriptional regulators are a strong candidate as a regulatory mechanism of biofilm formation. More than 200 small RNAs in the P. aeruginosa genome have been characterized in the literature to date; however, little is known about their biological roles in the cell. Here we describe the identification of the novel regulatory small RNA, SrbA. This locus was up-regulated 45-fold in P. aeruginosa strain PA14 biofilm cultures. Loss of SrbA expression in a deletion strain resulted in a 66% reduction in biofilm mass. Furthermore, the mortality rate over 72 hours in C. elegans infections was reduced to 39% when infected with the srbA deletion strain compared to 78% mortality when infected with the parental wild-type P. aeruginosa strain. There was no significant effect on culture growth or adherence to surfaces with loss of SrbA expression. Also loss of SrbA expression had no effect on antibiotic resistance to ciprofloxacin, gentamicin, and tobramycin. We conclude that SrbA is important for biofilm formation and full pathogenicity of P. aeruginosa.


Introduction
Bacterial biofilms are aggregated communities of cells that are embedded within a self-produced extracellular matrix [1,2]. The matrix can contain various biopolymers including polysaccharides, DNA, and protein [3][4][5][6][7]; it enables structured association of cells within the biofilm, mediates tight adhesion to surfaces, and promotes the mechanical stability of biofilms. The matrix also helps to maintain an internal environment and entrap extracellular degradative enzymes [8]. While biofilm colonies undergo dispersal to spread cells into the environment, they are intrinsically resilient and difficult to disrupt [9,10]. The resilience and resistance to treatment of biofilms poses a continual challenge in clinical settings when treating bacterial infections and decontaminating equipment [11]. Biofilm infections are significantly more PLOS  adaptively resistant to antibiotics due primarily to their altered gene expression patterns and in part due to the protection provided by the extracellular matrix [12][13][14]. In addition, biofilms contaminating surfaces are difficult to fully sterilize or remove mechanically, creating reservoirs of pathogenic bacteria in hospitals [15,16]. P. aeruginosa is a Gram-negative γ-Proteobacteria and is a significant opportunistic pathogen in individuals with compromised immune systems and natural barriers. Individuals with the genetic condition cystic fibrosis are highly susceptible to having their airways colonized by environmental sources of P. aeruginosa [17]. P. aeruginosa is highly adaptable to the host environment and is capable of altering regulatory networks to enable survival during chronic infections.
It has become increasingly appreciated that complex regulatory mechanisms govern biofilm development and enable both major and subtle responses that are dependent on cues from the external environment. Recent studies on pathogenic bacteria have demonstrated that non-coding, small RNA (sRNA) transcripts have important effects on biofilm formation and virulence in a host [18][19][20][21]. Recent studies into the transcriptome of P. aeruginosa uncovered novel sRNAs that are expressed under conditions that replicate aspects of the host environment during infections and biofilm development [22][23][24]. Many studies have catalogued novel sRNAs, however, there have been few studies that characterize their biological roles and determine the importance of sRNA-mediated regulation in complex adaptive modes of growth such as biofilm formation.
A particular category of sRNAs, called trans-sRNAs, are encoded as independent genes and usually do not form part of an operon [25]. These are often around 50-400 nucleotides in length [26,27] and act as post-transcriptional regulators of protein synthesis acting through short stretches (5-7 nucleotides) of base-pairing complementarity with target mRNAs to either promote or inhibit translation [27]. Trans-sRNAs are of particular interest when studying regulation of complex activities like biofilm formation, because such trans-sRNAs can have a high number of mRNA interaction targets throughout the genome, leading to broad post-transcriptional regulation. Typically these sRNAs interact at or near the ribosomal binding site (RBS) of an mRNA transcript [27]. Through interaction with the target mRNA, an sRNA may have a negative regulatory effect by blocking the ribosome or a positive regulatory effect by altering secondary structures through binding to the target mRNA and making the RBS available [25,27]. Trans-sRNAs can also exert their regulation through affecting mRNA stability. An sRNA can bind to its target and recruit RNases that will degrade an mRNA target [25]. Regulation by sRNAs can affect highly complex and diverse expression networks as well as providing cross-talk between signalling networks.
Having a better understanding of the biofilm lifestyle and regulation is of significant importance to developing new treatments for bacterial biofilm infections that comprise two thirds of all infections. Here we describe the novel sRNA regulator of biofilms A, SrbA that is important for biofilm formation and pathogenesis in P. aeruginosa. Biofilms grown under laboratory conditions were significantly diminished in an srbA deletion strain. Using a Caenorhabditis elegans model of infection, it was also found that the srbA deletion strain displayed a significant reduction in pathogenicity.

Bacterial strain generation and growth conditions
Strains and plasmids used in this study are listed in Table 1. All primers used in this study are listed in S1 Table. In the wild-type strain of P. aeruginosa UCBPP-PA14, a chromosomal deletion mutant of the sRNA gene locus was generated by allelic exchange [28]. Two 1 kb fragments flanking srbA were amplified using PCR. The two flanking fragments were digested with BamHI and ligated. The ligated 2 kbp deletion fragment and pEX18Gm suicide vector were digested with EcoRI and SalI before being ligated together to generate the pEXΔsrbA construct. The deletion construct was first introduced into Escherichia coli by heat-shock. P. aeruginosa UCBPP-PA14 was transformed with the deletion construct through incubation with the conjugative transfer strain S17-1. Transconjugants were isolated by growth on LB agar plates containing 15 μg/ml gentamicin to select for PA14 cells carrying the pEXΔsrbA plasmid and 30 μg/ml nalidixic acid to eliminate E. coli S17-1 cells. Strains containing the chromosomal deletion were confirmed by PCR as well as by sequencing at the StemCore facility of the Ottawa Hospital Research Institute.
To enable complementation by srbA expression, the entire 239 bp region was cloned into the expression plasmid pUCP18. PA14 was transformed by electroporation as described previously using 5 ms pulses at 2.5 kV in 0.2 cm electroporation cuvettes [36]. PA14 transformants carrying the pUCsrbA construct were selected on LB agar with 100 μg/ml carbenicillin. Successful transformation of the plasmid was confirmed by restriction enzyme digestion and visualization on an agarose gel. For most experiments, cultures were first grown overnight in LB medium at 37˚C. For growth assays (described below) 1% tryptone and M63 minimal media were used in addition to LB. When used, M63 basal medium at pH 7 consisted of final concentrations of 1x M63 salts (22 mM KH 2 PO 4 , 40 mM K 2 HPO 4 , 15 mM (NH 4 ) 2 SO 4 ), 0.4% (w/v) Larginine, and 1 mM MgSO 4 .

RNA isolation and quantitative PCR
Whole cell RNA was isolated from biofilm and planktonic cultures. Colony biofilms were grown by spotting multiples of 5 μl of overnight cultures onto M63 agar plates. Inoculated plates were incubated at 37˚C for 24 h plus an additional 24 h at room temperature before harvesting. Planktonic cultures were grown by inoculating a 1/100 dilution of overnight cultures into 3 ml M63 medium and incubating at 37˚C with shaking for 4 h or until an OD 600 of between 0.3-0.5 was reached. RNA isolation was performed by pelleting re-suspended biofilm colonies or planktonic cultures and incubating cells in 1 ml of TRIzol1 from Thermo Fisher Scientific, Inc. for 5 min with regular pipetting to homogenize the samples. RNA was then purified using the PureLink1 RNA Minikit according to instructions from Thermo Fisher Scientific, Inc. with an additional DNase digestion step before the final isolation of RNA. RNA was tested for DNA contamination by PCR. cDNA was generated using the iScript™ kit from Bio-Rad Laboratories, Inc. For each sample, 0.7 μg of RNA was used to synthesis cDNA. Quantitative PCR (RT-qPCR) was performed using the MyIQ™ system from Bio-Rad Laboratories, Inc. and fold-changes in expression were calculated by the ΔΔCt method using rpoD as a reference gene. Primers used in this study are listed in S1 Table. Crystal violet staining assays Static biofilms were grown according to previously established protocols in 96-well microtitre plates [37]. Overnight cultures were diluted 1/100 into fresh LB containing a final concentration of 100 μg/ml carbenicillin and 100 μl aliquoted into each well with several aliquots tested for each strain and biological replicate. Static biofilms were grown for 24 h at 37˚C, after which the microtitre plates were washed twice and each well was loaded with 100 μl of 0.1% (w/v) crystal violet. Plates were incubated at room temperature for 20 minutes before washing twice. Biofilm formation was quantified by solubilizing the crystal violet stain in 110 μl per well of 70% (v/v) ethanol. Absorbance was read at 595 nm to measure relative differences in biofilm biomass.

Growth curves
Growth studies were performed by inoculating 100 ml of fresh medium in an Erlenmeyer flask with overnight cultures at a 1/100 dilution. Inoculated flasks were then placed on a shaker at 140 rpm. Samples were taken immediately after inoculation and every 30 min to assess optical density at 600 nm (OD 600 ).

Minimal inhibitory concentrations (MICs) and minimal bactericidal concentrations (MBC) for planktonic cultures
Assaying for MICs was performed by serial dilution in a microtitre plate as described previously [38]. Two-fold serial dilutions of ciprofloxacin, gentamicin, and tobramycin were prepared and innoculated with overnight cultures of P. aeruginosa strains and MICs were taken after incubating at 37˚C for 18 hours. The MIC was taken as the first well in the microtitre plate to have no observable growth. The MBCs for planktonic cultures-were assayed by spotting 5 μl of culture from MIC assay microtitre plates starting at the MIC followed by increasing concentrations onto LB agar plates with no antibiotic added. The MBC was taken as the first spotted culture to have no colonies growing.

Bioinformatic analyses
Sequence complementarity searches and alignments were performed using the BLASTn and TargetRNA2 servers [41][42][43]. The SrbA transcript sequence was queried limiting the searches to the UCBPP-PA14 genome. Alignments returned were then manually checked to determine if the SrbA transcript does have a complementary alignment to the expressed mRNA transcript of the potential interaction sequence. Regulator motif searches and database comparisons were carried out using the PRODORIC database [44]. Searches for protein coding regions were performed using ExPASy [45]. Multiple sequence alignments were performed using Clustal Omega [46].

Results
The sRNA SrbA was highly up-regulated in P. aeruginosa biofilms SrbA was previously identified in independent published studies characterizing the P. aeruginosa transcriptome using second-generation sequencing methods [22][23][24]. These previous studies referred to SrbA as PA2633.1 [22], pant235 [23], and PA14sr_067 [24]. Using transcriptomic data available through the Pseudomonas Genome Database [37] and the Pseudomonas Browser [24], we determined that the gene srbA was encoded in an intergenic region on the reverse strand of the UCBPP-PA14 genome. srbA is 239 bp in length and encoded from nucleotides 2,604,298 through to 2,604,536 in an intergenic region with no presence of operons between isocitrate lyase aceA and an uncharacterized gene [22][23][24]46] (Fig 1A). There are no rho-independent terminators present in srbA. The Pseudomonas Genome Database, in addition to BLASTn searches, indicated that the sequence of srbA was conserved in all sequenced strains of P. aeruginosa [42,47]. Expression of SrbA in biofilm and planktonic cultures was assessed by RT-qPCR. SrbA was found to be up-regulated by 45-fold during biofilm growth compared to planktonic exponential growth phase cultures ( Fig 1B). Furthermore, SrbA was 4-fold up-regulated in planktonic stationary phase growth cultures relative to exponential phase growth cultures (Fig 1B).
Given the possibility that sRNAs might express short peptides that are biologically active, it was important to check if there was any recognizable peptide coding sequence within SrbA. The gene lacked a recognizable Shine-Delgarno sequence upstream of any potential start codon found in the sequence and indeed in silico translation of the SrbA transcript did not reveal any prospective expressed peptide sequences in any reading frame on either strand [24,45].
The srbA deletion strain had a significantly reduced ability to form biofilms A srbA deletion mutant strain was constructed in order to carry out phenotypic analysis of the mutant. Deletion of srbA was confirmed through sequencing, PCR amplification of the gene locus on the chromosome (S1 Fig), and loss of expression of SrbA through use of RT-qPCR (S2 Fig). It was also determined that there were no polar effects on expression of aceA, the gene immediately downstream of srbA (S3 Fig). Since SrbA was highly up-regulated under biofilm conditions, the ability of a strain carrying a deletion of srbA to form biofilms was tested by using a static biofilm formation model. The deletion mutant strain produced only 34% the amount of biofilm (p < 0.05) compared to the wild-type strain (Fig 2A).
Rapid attachment assays were performed to determine if there was a deficiency in the ability of ΔsrbA to adhere to surfaces [48][49][50]. In contrast to its deficiency in biofilm formation, ΔsrbA demonstrated no significant reduction in its ability to adhere to surfaces (Fig 2B).
Complementing the mutant by expressing SrbA from the pUCsrbA plasmid restored wildtype levels of biofilm formation (Fig 2A) and had no added effect on rapid attachment ( Fig  2B). Additionally, the wild-type strain overexpressing SrbA from the pUCsrbA plasmid displayed no alteration in biofilm phenotype.
Deletion of srbA had no impact on growth or antibiotic resistance in P. aeruginosa To determine if the biofilm deficiency was due to generally depressed cell health as a result of mutagenesis, growth and antibiotic resistance of the srbA deletion mutant were assayed. ΔsrbA had no growth deficiency in defined minimal medium and rich medium when compared to the wild-type strain (Fig 3). Furthermore, the susceptibility phenotype to three clinically relevant antibiotics (ciprofloxacin, gentamicin, and tobramycin) was tested. The srbA deletion 1% crystal violet and absorbance at 595 nm was taken. A student's t test was performed to determine significance. * represents p < 0.05. B) An assessment of rapid attachment for early biofilm formation was performed by incubating midlog phase cultures for 30 min at room temperature in a microtitre plate before staining with crystal violet for cells attached to wells. WT and ΔsrbA were transformed with empty pUCP18 plasmid and all strains were grown in 100 μg/ml carbenicillin. A one-way ANOVA was performed to determine no significance (ns). Both graphs represent the results of 4 biological replicates and error bars represent the standard error.
https://doi.org/10.1371/journal.pone.0182582.g002 sRNA involved in biofilm formation and Pathogenicity in Pseudomonas aeruginosa strain showed no significant change in resistance to any of these antibiotics compared to the wild-type strain ( Table 2).
The srbA deletion strain was attenuated in a C. elegans slow killing model The ability to form biofilms can contribute to the ability of a pathogen to persist within its host. Thus, it was hypothesized that the biofilm deficiency observed in the srbA deletion strain would result in reduced virulence and persistence in an animal host. We used a slow-killing model system to assay the ability of P. aeruginosa to persist in C. elegans that is considered a biofilm infection model of Pseudomonas [32,50,51]. After 72 hours, C. elegans infected with the srbA deletion strain displayed a 39% mortality rate that was significantly reduced compared to the 78% mortality rate observed in worms infected with the wild-type strain (Fig 4).  The srbA has complementarity with sixty-one putative mRNA targets The reduced biofilm phenotype of the srbA deletion strain could be due to multiple factors being affected by loss of srbA. Trans-sRNAs often have large numbers of targets on which they exert their effects [26,27]. sRNAs typically act on their targets through short stretches of complementarity to mRNA transcripts either resulting in an inhibition or enhancement of translation through mechanisms including altering the availability of the RBS or affecting mRNA stability via recruitment of RNase E [25,26,27,52]. A search was performed using the entire SrbA transcript sequence to query against the entire P. aeruginosa PA14 genome using both TargetRNA2 and BLASTn. TargetRNA2 specifically searches for complementarity of the 5' UTR of a putative target. Putative targets manually selected using BLAST searches were chosen on the basis that interactions with targets can occur outside of the 5' UTR for sRNA mechanisms affecting mRNA stability. A multiple sequence alignment of these putative targets was also generated (Fig 5). To determine whether the loss of srbA had any effect, the expression of the 61 putative mRNA targets was assessed by RT-qPCR. It was found that the transcript levels of 26 of the putative mRNA targets were increased (2 by !4-fold, 7 more by !2-fold) or decreased (12 by !4-fold and 5 more by 2-fold) by greater than two-fold in the srbA deletion strain compared to the parental PA14 wildtype grown as biofilms (Table 3).

Discussion
This study aimed to characterize the biological role of the novel sRNA SrbA. This sRNA was found to be highly upregulated relative to other sRNAs under biofilm conditions in P. aeruginosa (Fig 1B). It was determined that ΔsrbA was highly reduced in its ability to develop as a biofilm compared to the wild-type strain and restoring expression of SrbA from a plasmid was sufficient to restore wild-type levels of biofilm formation (Fig 2A). It was also found that this biofilm deficiency is not due to the srbA deletion strain having an inability to adhere to surfaces and establish biofilms (Fig 2B) or any growth deficiency of the mutant strain (Fig 3). Additionally, no effect on antibiotic resistance was observed in the srbA deletion strain ( Table 2). Taken together, these data demonstrate that the SrbA has an important role for the biofilm mode of growth beyond the attachment step.
Use of the C. elegans slow-killing model demonstrated that there is greater survival of C. elegans when infected with the srbA deletion strain compared to the wild-type strain. During an infection with P. aeruginosa, the host mounts an innate immune response through neutrophil activity that is of particular interest due to the tissue damage that occurs through inflammation [53]. C. elegans possesses an ancestral analogue of our innate immune system and is an applicable model to study host-pathogen interactions in P. aeruginosa infections [32]. This slow killing model assesses the ability of P. aeruginosa to develop and persist as a biofilm being the major contributor to pathogenesis [32,51,54]. It is therefore reasonable to conclude that the attenuated phenotype of the srbA deletion strain is due to its reduced ability to develop a biofilm.
Through use of TargetRNA2 and BLAST searches, 61 putative targets in the genome were identified that were within open reading frames and where the sequence complementarity was evident between the sRNA and mRNA transcripts ( Table 3). The alignment of the complementary regions of putative targets with SrbA demonstrates that there is a region of SrbA within the span of nucleotides 120-150 where there is greater complementarity (24 of 61 putative targets) than in any other region of SrbA (Fig 5). Such a primary seed region for complementarity is common for sRNAs. However, the putative targets with complementarity in this region did   Of the 61 putative mRNA targets of SrbA, there were 26 putative targets that displayed changes greater than 2-fold in transcript levels (Table 3). While sRNAs tend to affect their target gene expression at the level of protein stability, they are also known to exert effects on mRNA transcript stability [25,27]. Binding of an sRNA to its mRNA target can act to promote stability of a transcript or it may encourage degradation through recruitment of RNase E. Therefore, it is likely that the significant changes observed in 26 of the putative mRNA targets of SrbA in the srbA deletion strain are due to a loss of SrbA affecting regulation of stability and degradation. Additionally, of these 26 putative targets there are 9 genes (PA14_03560, hpaA, PA14_26810, PA14_32750, ilvA2, PA14_48010, nemO, PA14_59030, and adhA) that have complementarity in the region of nucleotides 120-150 where there was found to be a greater concentration of alignment of targets with SrbA (Fig 5). Future work with purified transcripts and RNase E could be performed to validate this.
The remaining 35 putative mRNA targets that did not display any significant change in transcript levels are likely under SrbA regulation through another mechanism such as affecting availability of the RBS. Future work could investigate the involvement of sRNAs like SrbA in regulating biofilm components such as polysaccharide secretion, regulation of pili and flagella, as well as regulatory effects on metabolic pathways represented in the list of putative mRNA targets. sRNA regulation might have a significant role in these complex responses that are sRNA involved in biofilm formation and Pathogenicity in Pseudomonas aeruginosa regulated by subtle changes in environmental conditions. Indeed, sRNA regulation has already been shown to be involved in pathways important for regulating biofilms in P. aeruginosa [55,56,57]. Taken together these putative target searches assist in guiding future work to investigate the specific regulatory involvement this sRNA has and how that contributes to the phenotypes observed in the srbA deletion.

Conclusions
Transcriptomic data and deep-sequencing have provided a vastly greater resolution of expression profiles in pathogens. These technologies have also provided new perspectives on previously underappreciated regulatory mechanisms such as sRNAs. However, determining how sRNAs fit into regulatory networks and what roles they have in the cell is still poorly understood. In this work, we demonstrated that the SrbA is important for biofilm growth in P. aeruginosa. We also determined that the expression of SrbA has a role in P. aeruginosa having full pathogenicity when infecting C. elegans. It is possible that SrbA is interacting with multiple targets that result in the phenotypic effects observed based on the 61 putative mRNA targets found here and that trans-sRNAs are characterized by having a large number of diverse mRNA targets [25,26]. It is important that continued work builds on recent transcriptomic studies to characterize the functional roles of novel, regulatory sRNAs found in P. aeruginosa. This will help us to not only gain a better understanding of basic biology of bacteria but also infectious states of pathogenic bacteria where nuances of regulation contribute to continued difficulty in treating infections due to biofilm adaptation to the stress of the host environment.