Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

An upstream sequence modulates phenazine production at the level of transcription and translation in the biological control strain Pseudomonas chlororaphis 30-84

  • Jun Myoung Yu,

    Roles Conceptualization, Data curation, Formal analysis, Methodology, Validation, Writing – original draft, Writing – review & editing

    Affiliations Department of Horticultural Sciences, Texas A&M University, College Station, TX, United States of America, Department of Plant Pathology and Microbiology, Texas A&M University, College Station, TX, United States of America

  • Dongping Wang,

    Roles Data curation, Formal analysis, Writing – original draft, Writing – review & editing

    Current address: Oil-Dri Corporation of America, Vernon Hills, IL, United States of America.

    Affiliation Department of Plant Pathology and Microbiology, Texas A&M University, College Station, TX, United States of America

  • Tessa R. Ries,

    Roles Data curation, Formal analysis

    Affiliation Department of Plant Pathology and Microbiology, Texas A&M University, College Station, TX, United States of America

  • Leland S. Pierson III,

    Roles Conceptualization, Funding acquisition, Investigation, Project administration, Supervision, Writing – original draft, Writing – review & editing

    Affiliation Department of Plant Pathology and Microbiology, Texas A&M University, College Station, TX, United States of America

  • Elizabeth A. Pierson

    Roles Conceptualization, Data curation, Funding acquisition, Investigation, Project administration, Supervision, Writing – original draft, Writing – review & editing

    eapierson@tamu.edu.

    Affiliations Department of Horticultural Sciences, Texas A&M University, College Station, TX, United States of America, Department of Plant Pathology and Microbiology, Texas A&M University, College Station, TX, United States of America

Abstract

Phenazines are bacterial secondary metabolites and play important roles in the antagonistic activity of the biological control strain P. chlororaphis 30–84 against take-all disease of wheat. The expression of the P. chlororaphis 30–84 phenazine biosynthetic operon (phzXYFABCD) is dependent on the PhzR/PhzI quorum sensing system located immediately upstream of the biosynthetic operon as well as other regulatory systems including Gac/Rsm. Bioinformatic analysis of the sequence between the divergently oriented phzR and phzX promoters identified features within the 5’-untranslated region (5’-UTR) of phzX that are conserved only among 2OHPCA producing Pseudomonas. The conserved sequence features are potentially capable of producing secondary structures that negatively modulate one or both promoters. Transcriptional and translational fusion assays revealed that deletion of 90-bp of sequence at the 5’-UTR of phzX led to up to 4-fold greater expression of the reporters with the deletion compared to the controls, which indicated this sequence negatively modulates phenazine gene expression both transcriptionally and translationally. This 90-bp sequence was deleted from the P. chlororaphis 30–84 chromosome, resulting in 30-84Enh, which produces significantly more phenazine than the wild-type while retaining quorum sensing control. The transcriptional expression of phzR/phzI and amount of AHL signal produced by 30-84Enh also were significantly greater than for the wild-type, suggesting this 90-bp sequence also negatively affects expression of the quorum sensing genes. In addition, deletion of the 90-bp partially relieved RsmE-mediated translational repression, indicating a role for Gac/RsmE interaction. Compared to the wild-type, enhanced phenazine production by 30-84Enh resulted in improvement in fungal inhibition, biofilm formation, extracellular DNA release and suppression of take-all disease of wheat in soil without negative consequences on growth or rhizosphere persistence. This work provides greater insight into the regulation of phenazine biosynthesis with potential applications for improved biological control.

Introduction

Phenazines are bacterial secondary metabolites produced by a diversity of plant-associated Pseudomonas species and that contribute to their ability to promote plant health [18]. Pseudomonas chlororaphis 30–84 is a rhizosphere-colonizing bacterial species capable of producing an array of secondary metabolites with beneficial agronomic applications. P. chlororaphis 30–84 was isolated as a biological control agent for take-all disease of wheat caused by the fungal pathogen Gaeumannomyces graminis var. tritici (Ggt). Phenazines are the principal antifungal secondary metabolites produced by P. chlororaphis 30–84 and several other well-studied biological control agents [17]. Phenazines are redox active molecules that are involved in diverse biological functions [8, 9]. P. chlororaphis 30–84 produces three phenazine derivatives, phenazine-1-carboxylic acid (PCA), 2-hydroxy-phenazine-1-carboxylic acid (2OHPCA) and a small amount of 2-hydroxy-phenazine (2OHPZ). PCA is produced via expression of the phenazine biosynthetic operon phzXYFABCD, and the phenazine modifying gene phzO, encoding an aromatic monooxygenase, is responsible for the conversion of PCA to the 2-hydroxy derivatives [1, 8, 10].

The ecological benefits of phenazine production have been well documented (reviewed in [8, 9, 11, 12]). These include their antibiotic characteristics, which facilitate survival in competition with other microorganisms [8, 9, 13, 14]. Phenazine production by P. chlororaphis 30–84 is primarily responsible for the inhibition of Ggt in vitro and in situ on roots as well as for its persistence on roots in competition with other rhizosphere microorganisms [1, 13]. Phenazines also contribute to biofilm formation [1519]. In P. chlororaphis 30–84 this was demonstrated using a phenazine biosynthetic mutant of P. chlororaphis 30–84, 30-84ZN (phzB::lacZ), which was defective in cell attachment and biofilm development [17]. Subsequently, using isogenic derivatives of P. chlororaphis 30–84 producing only PCA or overproducing 2OHPCA, the roles of different phenazines in specific aspects of biofilm formation and architecture were demonstrated [16]. More recently, differences in the ecological roles and transcriptional influence of each phenazine derivative produced by P. chlororaphis 30–84 were reported [19], including that 2OHPCA production more readily promotes extracellular DNA release, which results in a greater structured biofilm matrix. RNA-seq analysis revealed that phenazine production has broad impacts on gene expression patterns, including genes involved in the biosynthesis of exoenzymes, secondary metabolites, and other genes important for survival. Similar results have been shown for other phenazine producing species [20].

In most phenazine-producing bacteria including the plant growth promoting rhizosphere colonists P. synxantha (formerly P. fluorescens) 2–79, P. chlororaphis PCL1391 and P. chlororaphis 30–84, phenazine biosynthesis is controlled directly by the PhzR/PhzI quorum sensing system [2125]. Typically, phzR and phzI are located immediately upstream of the phenazine biosynthetic operon. The gene phzI encodes an N-acyl homoserine lactone (AHL) synthase and phzR encodes a transcriptional regulator of the phz biosynthetic operon [22, 24, 25]. Once AHL signals reach a threshold level, they interact with PhzR forming an active complex that binds to the specific sequence motif known as a “phz box” within the phenazine biosynthetic promoter resulting in the activation of the expression of the phenazine biosynthetic genes. The activated PhzR-AHL complex also binds to a phz box in the promoter region of phzI to enhance phzI expression resulting in increased AHL signal production.

The GacS/GacA two-component global regulatory system (TCS) is essential for biosynthesis of bacterial secondary metabolites including phenazines, AHL signals, exoprotease, lipase, gelatinase and HCN in P. chlororaphis 30–84 [26, 27] and other Pseudomonas species [2830]. In Pseudomonas, the Gac/Rsm system is comprised of three small non-coding RNAs (ncRNA, e.g. rsmX, rsmY and rsmZ) and two RNA binding repressor proteins (RsmA and RsmE) [27, 28, 3133]. The RsmA/E proteins function as posttranscriptional repressors by binding to a specific sequence motif (e.g. -GGA- or ribosome binding site) in the mRNA and blocking translation initiation and/or targeting mRNA degradation [32, 34, 35]. The binding of the ncRNAs to RsmA and RsmE results in sequestration of these repressor proteins and alleviation of their translational regulation [27, 3638]. For example in P. chlororaphis 30–84, the Gac system controls the expression of rsmZ, and in turns activates the expression of phz genes by titrating the translational repressor RsmE [27]. In addition to quorum sensing and the Gac/Rsm network, other regulatory genes are involved in the regulation of phenazine biosynthesis in P. chlororaphis 30–84 including, sigma factor RpoS, the two component system RpeA/RpeB, and the transcription regulator Pip [26, 27, 39, 40].

Predictably, much attention related to the regulation of phenazine biosynthesis has focused on the integration of these additional circuits as a regulatory network with quorum sensing to control phenazine production. In the present study, bioinformatics analysis was used to compare the 430-bp intergenic region between the divergently oriented phzR and phzX promoters in P. chlororaphis 30–84 (Fig 1A) and other closely related phenazine producing strains. This region includes the previously defined 18-bp palindromic phz box (and -10 hexamer sequence), which is highly conserved with previously identified phz box sequences in the phenazine operon promoters of biological control agents P. synxantha 2–79 and P. chlororaphis PCL1391 [22, 23, 41]. The P. chlororaphis 30–84 sequence motifs located within the 5’-UTR between the phz box and the ATG start codon of the first gene of the phenazine operon (phzX) differ from the two strains that do not produce 2OHPCA, P. synxantha 2–79 and P. chlororaphis PCL1391, but are similar to other strains that do produce 2OHPCA. These sequences are predicted to generate significant secondary structure. We examined the role of this region in phenazine regulation by creating a series of transcriptional and translation fusion plasmid derivatives with deletions or alterations aimed at disrupting the predicted secondary structures. We discovered a sequence within the 5’-UTR of P. chlororaphis 30–84, which is conserved among 2OHPCA producers, that modulates phenazine gene expression via transcriptional and translational regulation. The effects of the deletion of this sequence on P. chlororaphis 30–84 phenazine gene expression and on biological control efficacy were investigated.

thumbnail
Fig 1. Comparison of the nucleotide sequences of the phenazine biosynthetic promoter regions.

(A) Quorum sensing genes phzR and phzI are located immediately upstream of the phenazine biosynthetic operon and arrows indicate divergent transcription of phzR and phzX (B) The boxed region indicates the putative phz box sequences for the phenazine biosynthetic promoter of the six different phenazine-producing strains. Restriction enzyme sites are underlined. The putative -10 sequences, transcription start site (+1), ribosome binding site sequences (RBS), and ATG of PhzX are bolded. The hollow arrows indicate the direct repeat sequences (CACCCCCAA). Solid arrows indicate the four palindromic sequences. The 90-bp of 5’-UTR of phenazine biosynthetic operon is grey highlighted. The asterisks (*) indicate fully conserved residues, and gaps introduced for alignment are indicated by dashes (-). DNA sequences were obtained from National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov/) and aligned using Clustal Omega (http://www.ebi.ac.uk/Tools/msa/clustalo/). (C) Maximum-Likelihood (ML) tree based on a 250 bp region upstream from the translation start site of the phenazine biosynthetic operon from 27 different phenazine-producing pseudomonads. Sequences were retrieved from the Pseudomonas Genome database (www.pseudomonas.com) and NCBI. The tree with the highest log likelihood (-1149.3348) is shown, and only ML bootstrap values ≥ 50% are shown at nodes.

https://doi.org/10.1371/journal.pone.0193063.g001

Results

Analysis of the sequence between phzR and the phenazine biosynthetic operon in P. chlororaphis 30–84 and comparison to other phenazine producing strains

Sequence analysis of the promoter region and 5’-UTR of the phenazine biosynthetic operon in P. chlororaphis 30–84 revealed that there are four palindromic sequences: one centered on the PstI site (5’-GCGGGCTGCAGCCAGCCCGC-3’, Fig 1B: -87 to -67), a second centered on the BglII site (5’-CACTACAAGATCTGGTAGT-3’, Fig 1B: -54 to -35), a third on the first SspI site (5’-GCTGTAAATATTCACC-3’, Fig 1B: +35 to +52), and the fourth centered on the second SspI site (5’-AACTAAATATTTACTT-3’, Fig 1B: +61 to +77). The sequence surrounding the BglII site is a perfect match with the phz box found in other phenazine producing Pseudomonas strains that contain a single phenazine biosynthetic operon in their genome [23, 42]. The -10 site is located 9 bp upstream of transcription start site (TSS), which was previously identified by RNA-seq analysis [27] with a good sequence match to consensus sequences. Located between the phz box and the translation start site (ATG) of phzX are two identical repeats (5’-CACCCCCAA-3’, Fig 1B: -33 to -24 and +87 to +96) flanking a 112 bp sequence that has 38.4% GC content (as compared to the 51% GC content throughout the rest of the region and 62.9% GC overall content in the P. chlororaphis 30–84 genome) [43].

We compared sequence in this region (Fig 1B: -138 to the ATG) to sequence in the same region in other phenazine producing strains that have a single phz operon in their genome including P. chlororaphis O6, PA23 and GP72 (producing PCA, 2OHPCA and 2OHPZ), P. chlororaphis PCL1391 [producing PCA and phenazine-1-carboxamide (PCN)] and P. synxantha 2–79 (producing PCA only). The sequence surrounding the promoter region is highly conserved in all six strains, e.g., the sequence of the phz box and -35 and -10 regions match perfectly (Fig 1B). However, the palindromic sequence surrounding the PstI site is highly conserved only in the other 2OHPCA producing strains P. chlororaphis O6, PA23 and GP72, is less conserved in P. chlororaphis PCL1391 and is absent in P. synxantha 2–79 (Fig 1B). The 5’-UTR also is highly conserved only among the 2OHPCA producing strains with similar GC content (38.7% GC in P chlororaphis O6, PA23 and GP72) and all contain the direct repeats and repetitive sequence motifs (SspI sites) (Fig 1B). Although a similar region with low GC content is present in the 5’-UTR of phzA in P. synxantha 2–79 (40% GC) and P. chlororaphis PCL1391 (38% GC), both strains lack these direct repeats. The third palindromic sequence surrounding the first SspI site is identical to the sequence of 2OHPCA producing strains P. chlororaphis O6, PA23 and GP72, but is less conserved in P. synxantha 2–79 and P. chlororaphis PCL1391. The last palindromic sequence surrounding the second SspI site is highly conserved even though the SspI site is not present in P. synxantha 2–79 or P. chlororaphis PCL1391.

To examine whether similarities in this region are related to the types of phenazines each strain produces, we expanded the comparison to include twenty seven other phenazine producing pseudomonads that have a single phenazine operon in their chromosome. Based on Maximum Likelihood analysis, all the phylograms sorted into 4 major groups (Fig 1C). The 2OHPCA producing strains (with phzO located downstream of the phz operon) including P. chlororaphis 30–84, O6, PA23 and GP72 grouped together and formed a 2OHPCA producer clade with moderate bootstrap support. The second clade contained PCN producing strains (with phzH located downstream of the phz operon) including P. chlororaphis PCL1391, and PCA only producing strains (lacking any phenazine modifying gene) including P. synxantha 2–79 grouped together and formed a PCA producer clade with moderate bootstrap support. The forth clade included the two 2OHPCA producing strains without the quorum sensing system upstream of phz operon.

Role of conserved sequence motifs on phenazine gene expression in P. chlororaphis 30–84

To investigate the potential role of the sequence motifs in P. chlororaphis 30–84 on phenazine gene expression, a series of transcriptional fusion plasmid derivatives with deletions in the different features were cloned into the promoter trap vectors (pGT2-lacZ or pKT2-lacZ), resulting in the plasmids pJMYX1 (control) and pJMYX2 (containing a deletion of 90-bp in the 5’-UTR of phenazine biosynthetic operon) (Table 1 and Fig 2) and pJMYX3 (control), pJMYX4 (disruption of the first palindrome) and pJMYX5 (disruption of phz box) (Table 1 and S1 Fig). Each plasmid was introduced separately into 30-84Ice (phzB::inaZ), a phenazine non-producing mutant employed so that phenazine production did not interfere with the β-galactosidase assays.

thumbnail
Fig 2. Effect of 90-bp deletion of the 5’-UTR in phzX expression using a transcriptional fusion to the lacZ reporter gene.

The region of the P. chlororaphis 30–84 chromosome that contains the promoter of the phenazine biosynthetic operon (phzXYFABCD), four palindromic sequences (solid arrowheads) and two direct repeats (open arrowheads). Construction map for transcriptional fusion reporter plasmids pJMYX1 (control) and pJMYX2 (90-bp deletion): the rectangles with the diagonal lines indicate the sequence included in each derivative, the hollow rectangles indicate the 90-bp region not included in pJMYX2, the solid black rectangle represents the lacZ reporter gene sequence, and the black triangles under the lacZ reporter represent the ribosome binding site of lacZ. The transcriptional expressions of each reporter plasmids in 30-84Ice were determined via the β-galactosidase activities and presented as Miller Units. Data represent the average of eight replicates with standard errors. Asterisks indicate significant differences as determined by unpaired t-test (P < 0.05).

https://doi.org/10.1371/journal.pone.0193063.g002

thumbnail
Table 1. Bacterial strains and plasmids used in this study.

https://doi.org/10.1371/journal.pone.0193063.t001

The greatest increase in β-galactosidase activity relative to the appropriate control was observed for pJMYX2, which had 4-fold higher expression than the control plasmid pJMYX1 (Fig 1). Plasmid pJMYX2 has a 90-bp deletion, resulting in the removal of both SspI sites and the second direct repeat (+8 from the TSS to the ribosome binding site). This result indicates that the 5’-UTR of the phenazine biosynthetic operon negatively modulates phenazine expression.

The secondary structure predictive program Mfold [49] identified the second direct repeat (5’-CACCCCCAA-3’) in the 5’-UTR of phenazine biosynthetic operon as potentially capable of providing significant secondary structure in mRNA, which may contribute to the observed reduced expression (S2A Fig). To determine whether this secondary structure affected phenazine expression, plasmid derivatives were constructed with nucleotide alterations in the second direct repeat or both direct repeats (5’-CACCCCCAA-3’ to 5’-CACTATCAT-3’, S2B Fig). However, there were no significant differences in β-galactosidase expression with modifications to the direct repeats. The results indicated that small nucleotide sequence modifications were insufficient to relieve reduced expression. Predictions of potential secondary structure by Mfold were consistent with these derivatives altering but not eliminating possible secondary structures in this region.

Interestingly, β-galactosidase activity for pJMYX4 (missing half of the first palindromic sequence) was ~2-fold higher expression than for the control plasmid pJMYX3 (containing all palindromic sequences), suggesting that the first palindrome also may negatively affect phenazine expression to lesser extent under the conditions used (S1 Fig). Plasmid pJMYX5 (missing half of the phz box) had no β-galactosidase activity confirming that an intact phz box is required for phenazine promoter function (S1 Fig).

Deletion of 90-bp of the 5’-UTR of phenazine biosynthetic operon resulted in enhanced gene expression and phenazine production

Chromosomal deletions of the 90-bp sequence in the 5’-UTR of the phenazine biosynthetic operon were constructed via homologous recombination in the wild-type and the phenazine mutant 30-84ZN, resulting in derivatives 30-84Enh and 30-84ZN-Enh, respectively (Table 1). Compared to the wild-type, strain 30-84Enh produced significantly more phenazine in different types of media (i.e. 3.1-fold, 1.9-fold, and 2.4-fold in AB, LB and PPMD, respectively, Fig 3A and S3A Fig). Also β-galactosidase activity of 30-84ZN-Enh was 3.2-fold higher than that of 30-84ZN, indicating this was a result of enhanced expression of the phenazine biosynthetic genes (S3B and S3C Fig). Transcript abundances of phzX, phzB, and the phenazine modifying gene phzO measured via quantitative PCR (qPCR) were also greater in 30-84Enh relative to the wild-type (Fig 3B). These results suggest that the 90-bp of the 5’-UTR negatively modulates transcript abundance of the phenazine biosynthetic operon.

thumbnail
Fig 3. Phenazine production and gene expression patterns.

(A) Phenazine production by the 30–84 wild-type (WT) and 30-84Enh in different media (AB minimal, LB and PPMD). Data points represent means of three replicates ± standard error. Asterisks indicate significant differences as determined by unpaired t-test (P < 0.05). Experiments were repeated twice. (B) Expression of the phenazine regulatory genes in 30-84WT and 30-84Enh. The relative expression of selected phz operon (phzX, phzB and phzO) in 30-84WT and 30-84Enh were determined by qPCR using the16s rDNA gene as the reference. (C) Time course of phenazine production by 30-84WT and 30-84Enh in AB-C medium. During the growth, samples were taken periodically and from them total phenazines were extracted. Data points represent means of three replicates ± standard error. In some cases, error bars do not exceed the size of the symbol. Experiments were repeated twice. (D) AHL production by 30-84WT and 30-84Enh. AHLs obtained from 30-84WT and 30-84Enh were quantified using the AHL-specific reporter strain 30-84I/Z (phzI -, phzB::lacZ). AHLs were quantified based on β-galactosidase activity and reported in Miller Units (MU). Data are the means and standard errors of 8 replicates. Asterisks indicate significant differences as determined by unpaired t-test (P < 0.05). (A) and (C) Phenazines were quantified by UV-visible light spectroscopy at absorbance of 367 nm.

https://doi.org/10.1371/journal.pone.0193063.g003

In order to determine whether deletion of the 90-bp of the 5’-UTR of phenazine biosynthetic operon affected quorum sensing control of phenazine gene regulation, the relationship between the cell growth and the production of phenazine was followed over time. The wild-type and 30-84Enh had similar growth rates, indicating that alteration of the 5’-UTR of phenazine biosynthetic operon and enhanced phenazine production did not alter growth rates in vitro (Fig 3C). Neither strain produced detectable amounts of phenazines below OD620 = 0.8, indicating both strains lack phenazine production at low cell density. Phenazine production by 30-84Enh was significantly greater than the wild-type after OD620 = 1.8, and by late stationary phase the phenazine concentration from strain 30-84Enh was 5.8-fold higher than the wild-type (Fig 3C). Importantly, the requirement for quorum sensing was further verified by introducing plasmids pJMYX1 and pJMYX2 into the AHL synthase defective mutant 30-84I. No β-galactosidase activity was detected from 30-84I harboring either plasmid. These results indicate that loss of the 90-bp region did not relieve the requirement for quorum sensing activation as evidenced by the kinetics of phenazine production and the requirement for a functional phzI for gene expression.

Deletion of the 90-bp of the 5’-UTR of phenazine biosynthetic operon results in enhanced expression of quorum sensing

Given that phzR is located immediately upstream of phzX and divergently transcribed (Fig 1A), alterations in adjacent regulatory elements may influence the transcription of phzR, which may in turn lead to changes in the transcription of phzI, since phzI also has a phz box associated with its promoter region. To measure phzR expression, phzR transcriptional fusion vectors pJMYR1 (control, with the 90-bp region) and pJMYR2 (deletion of the 90-bp region) were constructed (Table 1. and S4 Fig). The β-galactosidase assays showed that lacZ gene expression was slightly, but statistically higher (p < 0.005, n = 8) in pJMYR2 (205 ± 4 MU) compared to the control, pJMYR1 (134 ± 4 MU). In addition, quantification of total AHL produced by the wild-type and 30-84Enh in different medium revealed that 30-84Enh produced significantly more AHL than the wild-type (Fig 3D). The transcript abundances of phzR and phzI were also greater in 30-84Enh compared to the wild-type as verified by qPCR (Fig 3B). These data indicate that the 90-bp region also has a negative influence on phzR expression which in turn reduces phzI expression, and that removal of the region enhanced the expression of the quorum sensing system.

Interaction between Gac/RsmE system and the 5’-UTR of phenazine biosynthetic operon

To determine whether the Gac/RsmE system negatively influences transcriptional expression via interaction with the 5’-UTR of phenazine biosynthetic operon, plasmids pJMYX1 (control) and pJMYX2 (90-bp deletion) were introduced separately into the wild-type, 30-84W (a spontaneous gacA mutant) and 30-84RsmE (rsmE::KmR). The β-galactosidase activities of both plasmids in 30-84W were below the detectable amount, whereas β-galactosidase activity of pJMYX2 was higher than pJMYX1 in the wild-type (consistent with the previous experiments) and in the RsmE mutant (Fig 4A). The results suggest that pJMYX2 requires a functional Gac system to activate the phzX expression and also indicate that loss of rsmE has no effect on either pJMYX1 or pJMYX2 transcription. Since RsmE is a post-transcriptional regulator, we constructed the translational fusion reporter plasmids pJMYX6 (control) and pJMYX7 (90-bp deletion) (Table 1 and Fig 4B). Sequence spanning -124 to +175 relative to the TSS of phzX (includes first 20 codons of PhzX) with and without with the 90-bp sequence in the 5’-UTR of phenazine biosynthetic operon were fused in frame with the 8th codon of lacZ to create the translational fusions pJMYX6 (control) and pJMYX7 (90-bp deletion), respectively, in pME6015 [34, 45]. The β-galactosidase activity (translational expression) of pJMYX7 (1743 ± 46 MU) was 3.2-fold higher than pJMYX6 (540 ± 63 MU) in the wild-type background (Fig 4C), which is similar to the relative fold change in transcriptional expression between pJMYX1 and pJMYX2 in the wild-type (Fig 4A). The β-galactosidase activity of pJMYX7 (90-bp deletion) was similar in both wild-type and the rsmE mutant backgrounds, indicating that loss of rsmE has no effect on pJMYX7 translation (Fig 4C). Importantly, the β-galactosidase activity of JMYX6 was significantly greater in the rsmE mutant compared to the wild-type, but was still significantly less than the activity of pJMYX7 in the rsmE mutant, indicating disruption of rsmE relieves, but only partially, repression of pJMYX6. Together the results using the transcriptional and translational fusions indicate that the 90-bp sequence is involved in RsmE-mediated translational regulation, but this does not explain entirely the effect of this region on phenazine expression.

thumbnail
Fig 4. Interaction between the 5’-UTR of phenazine biosynthetic operon and post-transcriptional repressor RsmE.

(A) The β-galactosidase activity of pJMYX1 and pJMYX2 in 30–84 wild-type (WT) and 30-84RsmE. Transcriptional activities of each reporter are expressed in Miller Units as the average of 8 replicates with standard error. Values with the same letter do not differ significantly as determined by a Fishers protected Least Significantly Difference (LSD) test (P ≥ 0.05). (B) Construction map for translational fusion reporter plasmids: the rectangles with the diagonal lines indicate the sequence included in each derivative, the hollow rectangle indicates the 90-bp regions not included in pJMYX7, the solid grey rectangle represents 20 codons of PhzX and the solid black rectangle represents the lacZ reporter gene sequence. (C) The β-galactosidase activity of pJMYX6 and pJMYX7 in the 30-84WT and 30-84RsmE. Translational activities of each reporter are expressed in Miller Units as the average of 8 replicates with standard error. Values with the same letter do not differ significantly as determined by a Fishers protected Least Significantly Difference (LSD) test (P ≥ 0.05).

https://doi.org/10.1371/journal.pone.0193063.g004

Ecological role of 90-bp deletion of the 5’-UTR of phenazine biosynthetic operon

Our previous research showed that altering the ratio of phenazines produced by P. chlororaphis 30–84 affected the attachment, density and structure of biofilms formed on solid surfaces as well as matrix production in floating biofilms [16, 17, 19]. To examine the effects of enhanced phenazine production on biofilm formation, the wild-type and 30-84Enh were grown in static culture. Strain 30-84Enh formed significantly more biofilm than the wild-type in all medium tested (Fig 5A). Since eDNA release is a key component of biofilm matrix [18, 19, 50], we compared eDNA release by the wild-type and 30-84Enh. Strain 30-84Enh produced significantly more eDNA than the wild-type over time (Fig 5B).

thumbnail
Fig 5. Effect of enhanced phenazine production on biofilm formation and eDNA release.

(A) Biofilm formation by the wild-type (WT) and 30-84Enh. Bacteria were grown in AB-C, LB and PPMD in static plates for 48 h. Attached cells were stained with crystal violet and quantified by OD540. Relative biofilm was calculated by standardizing to the wild-type (assigned a value of 1). Data represent average of six replicates from two separate experiments with standard errors. Asterisks indicate significant differences as determined by unpaired t-test (P < 0.05). (B) Release of eDNA by the 30-84WT and 30-84Enh. Cultures were grown in AB minimal for 120 h with rapid agitation. Samples were taken periodically and the concentration of eDNA was quantified using a Quibit (Invitrogen) fluorometer. In some cases, error bars do not exceed the size of the symbol. These experiments were repeated with similar results.

https://doi.org/10.1371/journal.pone.0193063.g005

We also tested whether wheat root colonization by the 30-84Enh strain differed from the wild-type using two inoculation methods: root-dip (3-day old seedlings) and soil inoculation (inoculum mixed with soil). Bacterial populations of root-dip inoculated wild-type and 30-84Enh on roots were similar (9.60 ± 0.2 and 9.48 ± 0.1 log CFU/g of roots, respectively, n = 8, p>0.05) after 30 days of growth. Wheat roots also recruited similar populations of wild-type and 30-84Enh from the soil after 24 days (7.72 ± 0.12 and 7.89 ± 0.06 log CFU/g of roots, respectively, n = 10, p > 0.05). These results suggest that enhanced phenazine production did not alter P. chlororaphis 30–84 colonization or persistence on roots.

Improved biological control ability to inhibit fungal growth and take-all suppression

As expected, enhanced phenazine production resulted in greater in vitro fungal inhibition of Ggt. Zones of fungal growth inhibition were significantly greater for 30-84Enh than the wild-type (8.2 ± 0.4 mm and 6.5 ± 0.3 mm, respectively, p<0.005, n = 9). To determine whether the increased phenazine production also results in enhanced disease suppression, the wild-type and 30-84Enh were seed inoculated and planted into a soil mix infested with Ggt as described previously [1, 51]. Strain 30-84Enh inoculated seedlings had significantly improved take-all suppression (reduced lesions on roots and disease symptoms on leaves) compared to the wild-type or the negative control (no bacterial inoculant) (Table 2). These results confirmed that enhanced phenazine production resulted in greater pathogen inhibition and disease suppression.

thumbnail
Table 2. Bacterial populations and take-all disease symptoms on wheat roots inoculated with P. chlororaphis 30–84 wild-type or 30-84Enh.

https://doi.org/10.1371/journal.pone.0193063.t002

Discussion

In P. chlororaphis 30–84, phenazines are essential for rhizosphere colonization, biofilm formation, pathogen inhibition and disease suppression [1, 16, 17, 39]. Given the importance of the bacterial functions for which phenazine production is vital, much attention has focused on phenazine regulation, including the influence of quorum sensing and other phenazine regulatory systems (e.g. transcriptional/translational regulatory proteins or ncRNAs) [22, 23, 27, 35, 40, 42, 52]. Analysis of the sequence between the divergently oriented phzR and phzX promoters identified multiple repetitive sequences and other features within the 5’-UTR of phzX and phzR capable of producing secondary structures (S2A Fig), which may alter gene expression by one or both promoters. Transcriptional fusion of phzX derivatives with deletions were analysed for their effect on gene expression. Deletion of the 90-bp sequence containing the third and the fourth palindrome sequences (SspI repeat sites) and the second direct repeat sequence resulted in 4-fold greater expression by the transcriptional reporter compared to the reporter with the wild-type sequence (Fig 2). These results indicate that this region is involved in negatively modulating phenazine gene expression. However, modifications of specific sequence motifs did not alleviate reduced expression significantly (S1B Fig), likely because the AT-rich nature of this region created other secondary structures. This high level of overall secondary structure may explain why deletion of this entire region was necessary to alleviate reduced expression. Deletion of this region from the P. chlororaphis 30–84 chromosome resulted in the construction of strain 30-84Enh that produced significantly more phenazine than wild-type (Fig 3A and 3C). Importantly, phenazine production by 30-84Enh still required activation via quorum sensing and still required a functional AHL synthase. Deletion of the 90-bp sequence also significantly improved AHL production and phzR and phzI transcript abundances (Fig 3B and 3D), which probably contributes to the increased phenazine production by 30-84Enh. Our evidence suggests that deletion of this sequence facilitated higher expression of the phenazine biosynthetic operon as well as the divergently transcribed phzR, the latter leading to increasing PhzR levels and enhancing interaction with the AHL signal. Since the promoters of both phzI and phzX have phz boxes, this would lead to higher phzI and phenazine gene expression.

In many plant-beneficial Pseudomonas species, Gac/Rsm-mediated regulation plays a crucial role in the control of secondary metabolites [27, 28, 33, 35, 36, 38, 53, 54]. However, the mechanisms by which RsmA/E mediate regulation of phenazine biosynthesis are somewhat species specific and can be either positive or negative. For example, RsmA differentially regulates the expression of the two phenazine biosynthetic operons in P. aeruginosa M18 [35]. RsmA binds upstream of the RBS of the P. aeruginosa M18 phz2 operon, resulting in enhancement of translation via destabilization of the stem loop structure. In contrast, RsmA binds motifs near the RBS of the P. aeruginosa M18 phz1 operon and negatively regulates its expression, likely due to RsmA blocking the RBS and/or targeting mRNA degradation as observed in other studies [28, 29, 5557]. Similar to RsmA regulation of the P. aeruginosa M18 phz1 operon, translational fusion assays suggest that RsmE targets the stem-loop structure in the P. chlororaphis 30–84 5’-UTR of phenazine biosynthetic operon, and deletion of the 90-bp enables 30-84Enh to avoid RsmE-mediated translational repression. However, we observed that in an rsmE mutant the translational activity of the wild-type reporter was still less than that of the reporter with the 90-bp deletion, indicating that disruption of rsmE only partially relieves repression. This finding reaffirms that the 90-bp sequence is also involved in interactions with transcriptional regulatory mechanisms that contribute to the reduced expression of the phenazine biosynthetic operon. The specific point of interaction between RsmE/rsmZ and the 5’-UTR of phenazine biosynthetic operon is currently being investigated.

Why do only 2OHPCA producers contain the conserved sequence that negatively modulates phenazine production?

One interesting finding of this study was that the multiple repetitive sequences we identified within the 5’-UTR of phenazine biosynthetic operon in P. chlororaphis 30–84 are also highly conserved among other 2OHPCA producing strains (e.g. having a single phenazine operon with phzO downstream of the phz operon), but are absent or differ from strains that produce other types of phenazines (Fig 1A and 1B). This led us to speculate why strains that produce 2OHPCA such as P. chlororaphis 30–84 and require quorum sensing for phenazine production modulate expression of the biosynthetic genes via the conserved presence of this upstream sequence. One hypothesis is that the cluster of phenazine regulatory (phzI/phzR) and biosynthetic genes (including phzO) share a common inheritance among 2OHPCA producers, with selection pressure for maintaining the conserved sequence. It was noted previously that despite the benefits of phenazine production to the ecological fitness of phenazine producers [1, 13, 16, 17, 39], phenazine production is stressful for the producing strain [45, 58]. It is interesting to speculate that some phenazine derivatives are more stressful to the producer than others. For example in P. aeruginosa, phenazine production, especially pyocyanin, promotes eDNA release presumably via the generation of reactive oxygen species (ROS) leading to cell lysis [59]. Wang et al. [19] demonstrated that cell lysis and release of eDNA by P. chlororaphis 30–84 was greater among derivatives that produced 2OHPCA compared to those that produced only PCA or do not produce phenazines. Additionally, cell lysis and eDNA release increased in derivatives producing more 2OHPCA. Consistent with the hypothesis that 2OHPCA production might be stressful for the cell, transcriptomic analysis revealed that as compared to derivatives that produce only PCA or do not produce phenazine, production of 2OHPCA by wild-type and the 2OHPCA overproducer significantly increased expression of genes involved in oxidative stress response and management, including ROS detoxifying enzymes, efflux pumps, DNA repair/modification enzymes, as well as a gene cluster encoding a bacteriophage-derived pyocin under the control of a stress-inducible promoter [19]. Similarly, 30-84Enh released up to 3-fold more eDNA than the wild-type (Fig 5B). Because production of 2OHPCA may be stressful in certain environments, 2OHPCA producers may have evolved or maintained the unique 5’-UTR sequence for tighter regulation of phenazine production as compared to strains that do not produce 2OHPCA. Interestingly, 30-84Enh is not altered in growth in shaking culture or in survival on wheat roots in growth chamber studies compared to the wild-type, suggesting that overproduction of phenazines under quorum sensing control is not detrimental in these environments.

P. chlororaphis 30-84Enh improves biological control

Phenazines have been studied extensively for their broad spectrum antibiotic activity against a diversity of soil-borne plant pathogens. We hypothesized that increased phenazine production should result in enhanced biological control. Consistent with this hypothesis, 30-84Enh has greater ability to inhibit fungal growth of the Ggt in vitro as well as take-all disease suppression than the wild-type (Table 2). Encouragingly for biological control applications, creation of 30-84Enh via the deletion of 90-bp sequence resulted in a strain with improved ability to form biofilms, produce eDNA, inhibit target pathogens and suppress disease, without deficiencies in growth rate or the ability to colonize wheat roots under our test conditions. The latter may be due to control of phenazine production by quorum sensing regulation, which determines the timing of phenazine production. Thus rather than achieving enhanced phenazine production via constitutive expression of genes in the chromosome or via the insertion of foreign genes in trans [60, 61], deletion of the modulating sequence may be both ecologically germane and environmentally safer for a biological control agent. The benefits of this approach include reducing potential interference in the establishment of a beneficial rhizosphere microbiome via constant phenazine antibiotic production and minimizing the possibility for horizontal gene transfer [62]. Future studies will consider whether these capabilities result in better suppression of other plant diseases or more reliable interactions with host plants and their rhizosphere microbiomes, contributing to enhanced plant health.

Materials and methods

Bacterial strains, plasmids and growth conditions

Bacterial strains and plasmids used in this study are described in Table 1, and primers used in this study are listed in S1 Table. A spontaneous rifampicin-resistant strain of P. chlororaphis 30–84 and its derivatives were grown at 28°C in Luria-Bertani medium (LB) (5 g of NaCl per liter), AB minimal media (AB), AB amended with 2% casamino acids (AB-C) (Difco, Franklin Lakes, NJ), or pigment production medium-D (PPM-D) [16, 17, 40, 44]. Where applicable, antibiotics were used in the following concentrations: gentamicin (Gn; 50 μg/ml), kanamycin (Km; 50 μg/ml), rifampicin (Rif; 100 μg/ml), tetracycline (Tc; 50 μg/ml) and piperacillin (Pip; 50 μg/ml) for P. chlororaphis 30–84, and ampicillin Ap (100 μg/ml), Gn (15 μg/ml), Tc (25 μg/ml) for E. coli.

DNA manipulations, sequence analysis, and PCR

Standard methods were utilized for plasmid DNA isolation, restriction enzyme digestion, ligation, transformation, and agarose gel electrophoresis [63]. Plasmids were introduced into P. chlororaphis 30–84 and its derivatives using either triparental matings or electroporation using methods described previously [1, 40]. Standard polymerase chain reaction (PCR) were carried out using FideliTaq DNA polymerase (Affymetrix, Santa Clara, CA) as described previously [45]. DNA sequencing was performed by the Laboratory for Genome Technology within Institute for Plant Genomics and Biotechnology, Texas A&M University using an ABI 3130xl Genetic Analyzer.

Phylogenetic analysis

To obtain the nucleotide sequence upstream of phenazine biosynthetic operon, the 250 bp nucleotide sequence upstream of the translation start site of phzX were blasted against the Pseudomonas Genome Database (www.pseudomonas.com) and National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih.gov/). A total of 27 phenazine producing strains were retrieved and analyzed for the types of a phenazine modifying enzyme. The nucleotide sequences (250 bp flanking sequences from the translation start site of each strain’s phenazine biosynthetic operon) were aligned and edited using MUSCLE (MEGA7). MEGA7 was used to build a, Maximum Likelihood (ML) trees based on the Tamura-Nei Model [64] [65]. ML bootstrapping was performed with 1,000 replicates to assess the relative stability of the branches.

Construction of phzX transcriptional fusion derivatives

To determine the function of the 5’-UTR of phenazine biosynthetic operon in phenazine gene expression, a 1.5-Kb DNA fragment contains a deletion of 90-bp in the 5’-UTR of phenazine biosynthetic operon was synthesized (GeneScript, Piscataway, NJ) and obtained as pUC57-Enh (Table 1). This fragment lacks 90-bp of sequence (including both SspI sites and the second direct repeat) starting at the +8 bp (from the TSS) to the RBS of phzX, but maintaining the endogenous RBS and start codon of PhzX. PCR fragments of 271 bp and 181 bp containing sequence from -124 to +147 of phzX were amplified from genomic DNA of P. chlororaphis 30–84 and pUC57-Enh using the primers, phzXF1/phzXR1 (S1 Table) and cloned into the promoter trap vector pGT2-lacZ resulting plasmids pJMYX1 and pJMYX2, respectively (Fig 2 and Table 1). To determine the function of upstream sequence of phenazine biosynthetic promoter region, PCR fragments of 354bp (from -124 to +230), 310bp (from -80 to +230), and 274bp (from -44 to +230) containing the promoter of phenazine biosynthetic operon, 5’-UTR and 115bp of phzX coding sequence were amplified from P. chlororaphis 30–84 genomic DNA using the following primer sets, phzXF1/phzXR2, phzXF2/phzXR2 and phzXF3/phzXR2, respectively (S1 Table) These fragments were cloned into the promoter trap vectors pKT2-lacZ, creating the plasmids pJMYX3, pJMYX4 and pJMYX5, respectively (Table 1 and Fig 2). The plasmids pKT2-lacZ and pGT2-lacZ contain a promoterless lacZ gene with its own ribosome binding site (RBS) located downstream of multiple cloning locus, which enables the study of transcriptional activities.

All transcriptional fusion plasmids were separately introduced into 30-84Ice via triparental mating and the transcriptional activity was determined by β-galactosidase activity [66] in 30-84Ice after 24 h growth in LB with rapid agitation. Strain 30-84Ice was used for the transcriptional fusion assays because it contains a phzB::inaZ insertion (Table 1) and as a consequence does not produce phenazine, which interferes with the β-galactosidase assay.

Generation of a phenazine enhanced mutant

In order to generate the phenazine enhanced mutant strain, the 1.5 Kb sequence from pUC57-Enh was cloned into the vector, pLAFR3, generating pLAF-phzEnh (Table 1). This 1.5 Kb fragment contains flanking regions upstream from the EcoRV site in the first half of phzR and downstream to the BamHI site at the 3’-end of phzY (the second gene of phz operon) to facilitate homologous recombination. The pLAF-phzEnh plasmid was introduced into P. chlororaphis 30–84 strains containing pUCP18-RedS via triparental mating, and the chromosomal 90-bp deletion of 5’-UTR of phenazine biosynthetic operon was obtained with the support of λ phage recombinases [40, 48]. A dark orange colony (for 30-84Enh) from the PPMD plate or a dark blue colony (for 30-84ZN-Enh) from PPMD plate supplemented with 2% X-gal were chosen, and pUCP18-RedS plasmid (containing sacB) was cured by counter selection on LB plates supplemented with 5% sucrose. A Tcs, Pips, and SucroseR colony was chosen and mutation was verified by PCR and sequencing.

Construction of an rsmE deletion mutant

To inactivate rsmE gene in P. chlororaphis 30–84, fragments of upstream (611 bp) and downstream (706 bp) of the rsmE open reading frame (ORF) were amplified from P. chlororaphis 30–84 genomic DNA using primers RsmEUPF/RsmEUPR and RsmEDWF/RsmEDWR (S1 Table). These fragments were designed to carry a KpnI site that permitted the insertion of a kanamycin resistance cassette at the 3’ end of upstream fragment and 5’ end of downstream fragment with 183 bp deletion of rsmE ORF. Each fragment was simultaneously cloned into the pEX18Ap (Table 1). A 961 bp fragment containing kanamycin resistance cassette was PCR amplified from pUC4K (Table 1) using the primers, KmKpnF/KmKpnR, and inserted into KpnI site between upstream and downstream fragments in pEX18Ap. The resulting plasmid, pEX18Ap-rsmEKO (Table 1), was electroporated into P. chlororaphis 30–84, and mutant was selected for by amending LB plates with the appropriate antibiotics. A KmR, Pips, and SucroseR colony was chosen and the rsmE mutation was verified by PCR and sequencing.

Construction of translational fusion vector with 90-bp deletion of the 5’-UTR of phenazine biosynthetic operon

To determine the function of the 5’-UTR on translation of the phenazine biosynthesis genes, translation fusion vectors were created containing the sequence from the phenazine biosynthetic operon promoter to the 20th codon of PhzX with or without 90-bp sequence of 5’-UTR of phenazine biosynthetic operon (Fig 4B). The fragments with the 90-bp sequence (299 bp) and without the 90-bp sequence (209 bp) were PCR amplified from genomic DNA of P. chlororaphis 30–84 and pUC57-Enh, respectively, using primer set phzXF1/phzXR3 (S1 Table). The products were cloned, in frame, with the 8th codon of lacZ in the translational fusion vector, pME6015 [34, 45], resulting pJMYX7 and pJMYX8 respectively (Fig 4B and Table 1). These plasmids were introduced in the wild-type, 30-84W and 30-84RsmE via electroporation, and translational activities were determined by β-galactosidase activity.

RNA preparation for quantitative PCR

To isolate RNA from the wild-type and 30-84Enh, single colonies of each strain were grown with rapid agitation at 28°C in 3 ml of AB-C broth. When cell density reached OD620 = 1.8, 1 ml aliquots of each sample were mixed with 2 ml of Qiagen RNA Protect reagent (Qiagen, Hilden, Germany) to stabilize bacterial RNA, and cells were harvested by centrifugation for 10 min at 2400 x g. Total RNA was extracted using a Qiagen RNeasy Mini Kit (Qiagen) according to the manufacturer’s recommended protocol. The genomic DNA was removed using on-column DNase-I digestion (Qiagen). Five micrograms of total RNA were reverse-transcribed using random primers (Invitrogen Life Technologies, Carlsbad, CA) and Superscript III (Invitrogen) at 50°C for 1 h and inactivated at 75°C for 15 min. For the negative control, the same reaction was performed using sterilized water instead of reverse transcriptase.

qPCR methods and analysis

SYBR Green reactions were performed using the ABI 7900 HT Fast System (Applied Biosystems, Foster City, CA) in 384 well optical reaction plates. Quantitative PCR (qPCR) assays were performed to measure the expression levels of the target genes as previously described [45]. Briefly, aliquots (1 μl) of cDNA (2 ng/reaction) or negative controls were used as template for qPCR reactions with Fast SYBR Green PCR Master Mix (Applied Biosystems) and primers (500 nM final concentration). qPCR amplifications were carried out at 50°C for 2 min, 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min, and a final dissociation curve analysis step from 65°C to 95°C. Technical replicate experiments were performed for each of the biological samples, in triplicate. Amplification specificity for each reaction was confirmed by the dissociation curve analysis. The Ct values were used for further ΔΔCt analysis. The 16S rDNA was used as a reference gene to normalize samples. A relative quantification value was calculated for each gene with the control group as a reference [27, 40, 45].

Quantification of phenazine and AHL production

P. chlororaphis 30–84 strains were grown at 28°C in AB minimal, LB and PPMD broth with rapid agitation. Phenazines were extracted from cell-free supernatants as described previously [1, 16, 45]. Phenazines concentration was calculated via serial dilution of the extract at absorbance of 367 nm. Total AHLs were extracted as described previously [39, 45] from 5 ml cultures, which were grown at 28°C with shaking in AB minimal, LB and PPMD broth. AHL production was quantified by inoculating the extracted AHLs with the AHL-specific reporter strain 30-84I/Z (phzI-, phzB::lacZ). 30-84I/Z is deficient in AHL production due to mutation of AHL synthase gene phzI, but responds to AHL by expressing the reporter gene lacZ. β-galactosidase activity was determined on cultures grown at 28°C for 24 h with rapid agitation.

Construction of phzR transcriptional reporters

In order to determine whether the 90-bp of the 5’-UTR of phenazine biosynthetic operon also negatively influence phzR expression, phzR transcriptional fusions were constructed. PCR fragments containing a sequence from -359 bp to +155 of phzR was amplified from genomic DNA of the wild-type and 30-84Enh using the primers, phzRF-phzRR (S1 Table). The 514 bp and 424 bp PCR amplicons containing phzR promoter region with or without the 90-bp of the 5’-UTR of phzX, respectively, were ligated into promoter trap vector pGT2-lacZ to make the phzR transcriptional fusion vectors, pJMYR1 and pJMYR2 (Table 1 and S4 Fig). These reporters were separately introduced into 30-84Ice by triparental mating, and transcriptional activities were determined by β-galactosidase activity.

Microtiter plate biofilm assay and eDNA quantification

To measure the ability of strains 30–84 wild-type and 30-84Enh to form a biofilm, static biofilm assays were conducted in 24-well polystyrene microtiter plates. Briefly, overnight cultures grown in 3 different media (AB-C, LB and PPMD) were adjusted to an OD620 of 0.8 with fresh medium. The adjusted cultures were diluted 1:100 into the appropriate media, and 1.5 ml of the dilution inoculate was transferred into 24-well plates. Plates were incubated at 28°C without shaking. After 48 h, the adherent cell population was quantified by crystal violet staining as described previously [17, 67].

The concentration of eDNA was determined quantitatively using Qubit 2.0 Fulorometer (Invitrogen), as described previously with few modification [19]. Briefly, overnight cultures grown in AB-C broth at 28°C with agitation were adjusted to an OD620 of 0.8. The adjusted cultures were re-inoculated at a 1:100 dilution into 20 ml AB-C broth. Cultures were grown at 28°C with rapid agitation and sampled every 8–12 h. Cell-free supernatant by centrifugation and filtration were mixed with double-strand DNA fluorescent dyes (dsDNA BR) from Qubit (Invitrogen Life Technologies, Carlsbad, CA), and the concentration of eDNA was quantified using Qubit 2.0 Fluorometer (Invitrogen Life Technologies). The amount of eDNA was reported as μg/ml.

Root colonization assay

To determine the ability of colonizing to the host plant, the wild-type and 30-84Enh were inoculated to wheat (cultivar TAM112) by two different methods. For the root-dip inoculation methods, bacterial cultures were grown in 10 ml KMB broth for 24 h, and inoculum was standardized to OD620 = 0.8 (ca. 2 x 109) in sterile 1X PBS. Seeds were surface sterilized as previously described [44], and surface sterilized seeds were pregerminated on sterilized germination paper for two days. Seedling roots were dipped into the bacterial solution for 10 min, and sown into 25 × 200 mm cone-tainers that contain a natural wheat rhizosphere (Uvalde, TX) soil mix (soil: sand, 2:1, v:v). Plants were grown for 30 days before the entire root system was processed for the CFU calculation, as described previously [44]. For the soil inoculation method, bacterial cultures were prepared as described above. Bacteria cultures were washed 3 times with sterilized water, and inoculum were thoroughly mixed with natural wheat rhizosphere soil mix (soil: sand, 2:1, v:v) for final bacterial population of 106 CFU/g of soil. Pregerminated wheat seedlings (two days) were sown into the bacterial amended soil mix and were grown for 24 days. Bacterial populations were collected from the entire root system and quantified by CFU. Total populations were determined by serial dilution on LB agar amended with rifampicin.

Fungal inhibition and take-all suppression assay

To quantify the ability of strains 30–84 wild-type and 30-84Enh to inhibit the take-all causal agent Ggt strain ARS-A1, an in vitro dual culture assay was conducted as described previously [45]. After 7 days of co-culture on potato dextrose agar plates, the zone of inhibition was measured as the distance between edge of the bacterial colony and the fungal mycelium.

Assays to determine the ability of strains 30–84 wild-type and 30-84Enh to suppress take-all disease on wheat seedlings were conducted as described previously [1, 51]. Briefly, bacteria-coated seeds or control seeds (coated with methyl cellulose) were sown in tubes (25 × 200 mm) filled with 5g of sterilized vermiculate layer overlaid with 20 g of a natural wheat rhizosphere soil mix (soil: sand, 2:1, v:v) amended with Ggt colonized oat kernels fragments (0.85%, w/w). For the control, sterilized oat kernels were ground and amended to the soil mix with same amount (0.85%, w/w). Seeds (cv. TAM112) were covered with 1cm of sterilized vermiculate and incubated for 3 days at room temperature to facilitate germination. Seedlings were arranged in a complete randomized block design, and transferred to a growth chamber (16°C, 12 h dark-light cycle). After 20 days, root disease was evaluated on a scale of 0–5, where 0 = no disease and 5 = nearly dead. as described previously [1].

Statistical analysis

All data presented are mean ± the standard error of the mean (STE) from at least two experiments. Data were analyzed using ANOVA and Fisher’s protected Least Significant Difference (LSD) test (P<0.05) or unpaired t-test. Data were processed with GraphPad Prism (GraphPad Software, San Diego, CA).

Supporting information

S1 Table. Oligonucleotides used for gene cloning and qPCR.

https://doi.org/10.1371/journal.pone.0193063.s001

(PDF)

S1 Fig. Analysis of phzX expression in P. chlororaphis 30–84 using transcriptional fusion to the lacZ reporter gene.

The region of the P. chlororaphis 30–84 chromosome that contains the promoter of the phenazine biosynthetic operon (phzXYFABCD) includes four palindromic sequences (solid arrowheads) and two direct repeats (open arrowheads). For each plasmid derivative, the rectangles with the diagonal lines indicate the sequence included in each derivative, the solid black rectangle represents the lacZ reporter gene sequence, and the black triangles under the lacZ reporter represent the ribosome binding site of lacZ. The transcriptional expressions of the phz promoter derivatives in 30-84Ice were determined via the β-galactosidase activities and presented as Miller Units. Data represent the average of eight replicates with standard errors. The different letters indicate significant differences by Fisher’s protected Least Significantly Difference (LSD) test (P < 0.05). Note: Expression of control plasmids pJMYX1 (Fig 2) and pJMYX3 were not significantly different.

https://doi.org/10.1371/journal.pone.0193063.s002

(PDF)

S2 Fig. Analysis of secondary structure prediction and the activity of promoter derivatives.

(A) Potential secondary structure predicted from the RNA sequence spanning the transcription start site (+1) site to the SalI site of phzX. The two SspI sites are marked with blue highlight, the second direct repeats are marked with red highlight, and the putative phzX RBS and start codon are marked with green highlight. (B) The pJMYX4 and its derivatives with specific sequence alterations fused to lacZ. The black solid arrows represent the two direct repeats (5’-CACCCCCAA-3’), DR1 and DR2. The black rectangles represent the RBS of phzX. The hollow solid arrows represent modified sequence motifs. The orange rectangles represent partial ORF of phzX. The hatched rectangles represent lacZ and its RBS (triangle); dotted lines indicate restriction enzyme sites. (C) The β-galactosidase activity of pJMYX4 promoter (control) and each derivative in 30-84Ice. Promoter activity is expressed in Miller Units as the average of 8 replicates with standard error. Values with the same letter do not differ significantly as determined by a Fishers protected Least Significantly Difference (LSD) test (P ≥ 0.05).

https://doi.org/10.1371/journal.pone.0193063.s003

(PDF)

S3 Fig. Characterizing phenazine production by 30-84Enh and 30-84ZN-Enh.

(A) Phenazine production by P. chlororaphis 30–84 wild-type and 30-84Enh after 48 h on PPMD agar plates. (B) Overnight culture of 30-84ZN and 30-84ZN-Enh on LB broth supplemented with 2% X-Gal. (C) Expression of the phz biosynthetic operon in 30-84ZN and 30-84ZN-Enh. Expression of phz operon was quantified by β-galactosidase assay. Each bar represent means ± SE of six replicates from two independent experiments.

https://doi.org/10.1371/journal.pone.0193063.s004

(PDF)

S4 Fig. Map of the promoter sequences used to construct the PhzR transcriptional fusion reporters.

Construction map for transcriptional fusion reporter plasmids, pJMYR1(control) and pJMYR2 (90bp deletion). These reporters containing flanking sequence from -359 to +155 including phzR promoter (-35 and -10), ribosome binding site, transcription and translation start site and the first 84 bp of the phzR gene. The rectangles with the diagonal lines indicate the sequence included in each derivative, the hollow rectangles indicate the 90-bp region not included in pJMYR2, the solid grey rectangle represents partial ORF of phzR, the solid black rectangle represents the lacZ reporter gene sequence, and the black triangles under the lacZ reporter represent the ribosome binding site of lacZ. Divergently transcribed phzX promoter and the relative nucleotide based on +1 of phzX are presented as grey.

https://doi.org/10.1371/journal.pone.0193063.s005

(PDF)

Acknowledgments

We thank Dr. David Weller (U.S. Department of Agriculture-Agricultural Research Service, USA) for generously providing the fungal pathogen Gaeumannomyces graminis var. tritici and Dr. Panatda Saenkham (Texas A&M University, USA) for valuable technical support and helpful discussions. We thank Dr. Jayasimha Rao (Virginia Tech Carilion School of Medicine, USA) for generously contributing a translational fusion vector pME6015. We also acknowledge the contribution of Tim Malan, Julien Levy, and Gerardo Puopolo (former Pierson Lab members) for preliminary data analysis.

References

  1. 1. Pierson LS III, Thomashow LS. (1992) Cloning and heterologous expression of the phenazine biosynthetic locus from Pseudomonas aureofaciens 30–84. Mol Plant-Microbe Interact.5:330–9. pmid:1325219
  2. 2. Thomashow LS, Weller DM, Bonsall RF, Pierson LS III. (1990) Production of the antibiotic phenazine-1-carboxylic acid by fluorescent Pseudomonas species in the rhizosphere of wheat. Appl Environ Microbiol.56:908–12. pmid:16348176
  3. 3. Chin-A-Woeng TF, Bloemberg GV, van der Bij AJ, van der Drift KM, Schripsema J, Kroon B, et al. (1998) Biocontrol by phenazine-1-carboxamide-producing Pseudomonas chlororaphis PCL1391 of tomato root rot caused by Fusarium oxysporum f. sp. radicis-lycopersici. Mol Plant-Microbe Interact.11:1069–77.
  4. 4. Powell J, Vargas J Jr, Nair M, Detweiler A, Chandra A. (2000) Management of dollar spot on creeping bentgrass with metabolites of Pseudomonas aureofaciens (TX-1). Plant Dis.84:19–24.
  5. 5. Tambong JT, Höfte M. (2001) Phenazines are involved in biocontrol of Pythium myriotylum on cocoyam by Pseudomonas aeruginosa PNA1. Eur J Plant Pathol.107:511–21.
  6. 6. Hu HB, Xu YQ, Chen F, Zhang XH, Hur BK. (2005) Isolation and characterization of a new fluorescent Pseudomonas strain that produces both phenazine-1-carboxylic acid and pyoluteorin. J Microbiol Biotech.15:86–90.
  7. 7. Thomashow LS, Essar DW, Fujimoto DK, Pierson LS, Thrane C, Weller DM. Genetic and Biochemical Determinants of Phenazine Antibiotic Production in Fluorescent Pseudomonads that Suppress Take-All Disease of Wheat. In: Nester EW, Verma DPS, editors. Advances in Molecular Genetics of Plant-Microbe Interactions, Vol 2: Proceedings of the 6th International Symposium on Molecular Plant-Microbe Interactions, Seattle, Washington, USA, July 1992. Dordrecht: Springer Netherlands; 1993. p. 535–41.
  8. 8. Mavrodi DV, Blankenfeldt W, Thomashow LS. (2006) Phenazine compounds in fluorescent Pseudomonas spp. biosynthesis and regulation. Annu Rev Phytopathol.44:417–45. pmid:16719720
  9. 9. Pierson LS III, Pierson EA. (2010) Metabolism and function of phenazines in bacteria: impacts on the behavior of bacteria in the environment and biotechnological processes. Applied microbiol Biotechnol.86:1659–70.
  10. 10. Delaney SM, Mavrodi DV, Bonsall RF, Thomashow LS. (2001) phzO, a gene for biosynthesis of 2-hydroxylated phenazine compounds in Pseudomonas aureofaciens 30–84. J Bacteriol.183:318–27. pmid:11114932
  11. 11. Chin-A-Woeng TF, Bloemberg GV, Lugtenberg BJ. (2003) Phenazines and their role in biocontrol by Pseudomonas bacteria. New phytologist.157:503–23.
  12. 12. Mavrodi DV, Parejko JA, Mavrodi OV, Kwak YS, Weller DM, Blankenfeldt W, et al. (2013) Recent insights into the diversity, frequency and ecological roles of phenazines in fluorescent Pseudomonas spp. Envrion Microbiol.15:675–86.
  13. 13. Mazzola M, Cook RJ, Thomashow LS, Weller DM, Pierson LS (1992) Contribution of phenazine antibiotic biosynthesis to the ecological competence of fluorescent pseudomonads in soil habitats. Appl Environ Microbiol.58:2616–24. pmid:1514808
  14. 14. Dwivedi D, Johri BN, Ineichen K, Wray V, Wiemken A. (2009) Impact of antifungals producing rhizobacteria on the performance of Vigna radiata in the presence of arbuscular mycorrhizal fungi. Mycorrhiza.19:559–70. pmid:19458967
  15. 15. Ramos I, Dietrich LE, Price-Whelan A, Newman DK. (2010) Phenazines affect biofilm formation by Pseudomonas aeruginosa in similar ways at various scales. Res Microbiol.161:187–91. pmid:20123017
  16. 16. Maddula VS, Pierson EA, Pierson LS III. (2008) Altering the ratio of phenazines in Pseudomonas chlororaphis (aureofaciens) strain 30–84: effects on biofilm formation and pathogen inhibition. J Bacteriol.190:2759–66. pmid:18263718
  17. 17. Maddula VS, Zhang Z, Pierson EA, Pierson LS III. (2006) Quorum sensing and phenazines are involved in biofilm formation by Pseudomonas chlororaphis (aureofaciens) strain 30–84. Microb Ecol.52:289–301. pmid:16897305
  18. 18. Das T, Kutty SK, Kumar N, Manefield M. (2013) Pyocyanin facilitates extracellular DNA binding to Pseudomonas aeruginosa influencing cell surface properties and aggregation. PLoS One.8:e58299. pmid:23505483
  19. 19. Wang D, Yu JM, Dorosky RJ, Pierson LS III, Pierson EA. (2016) The phenazine 2-hydroxy-phenazine-1-carboxylic acid promotes extracellular DNA release and has broad transcriptomic consequences in Pseudomonas chlororaphis 30–84. PLoS One.11:e0148003. pmid:26812402
  20. 20. Dietrich LE, Price-Whelan A, Petersen A, Whiteley M, Newman DK. (2006) The phenazine pyocyanin is a terminal signalling factor in the quorum sensing network of Pseudomonas aeruginosa. Mol Microbiol.61:1308–21. pmid:16879411
  21. 21. Chin-A-Woeng TF, Thomas-Oates JE, Lugtenberg BJ, Bloemberg GV. (2001) Introduction of the phzH gene of Pseudomonas chlororaphis PCL1391 extends the range of biocontrol ability of phenazine-1-carboxylic acid-producing Pseudomonas spp. strains. Mol Plant-Microbe Interact.14:1006–15. pmid:11497461
  22. 22. Khan SR, Herman J, Krank J, Serkova NJ, Churchill ME, Suga H, et al. (2007) N-(3-hydroxyhexanoyl)-L-homoserine lactone is the biologically relevant quormone that regulates the phz operon of Pseudomonas chlororaphis strain 30–84. Appl Environ Microbiol.73:7443–55. pmid:17921283
  23. 23. Khan SR, Mavrodi DV, Jog GJ, Suga H, Thomashow LS, Farrand SK. (2005) Activation of the phz operon of Pseudomonas fluorescens 2–79 requires the LuxR homolog PhzR, N-(3-OH-Hexanoyl)-L-homoserine lactone produced by the LuxI homolog PhzI, and a cis-acting phz box. J Bacteriol.187:6517–27. pmid:16159785
  24. 24. Pierson LS III, Keppenne VD, Wood DW. (1994) Phenazine antibiotic biosynthesis in Pseudomonas aureofaciens 30–84 is regulated by PhzR in response to cell density. J Bacteriol.176:3966–74. pmid:8021179
  25. 25. Wood DW, Pierson LS III. (1996) The phzI gene of Pseudomonas aureofaciens 30–84 is responsible for the production of a diffusible signal required for phenazine antibiotic production. Gene.168:49–53. pmid:8626064
  26. 26. Chancey ST, Wood DW, Pierson LS III. (1999) Two-component transcriptional regulation of N-acyl-homoserine lactone production in Pseudomonas aureofaciens. Appl Environ Microbiol.65:2294–9. pmid:10347004
  27. 27. Wang D, Lee SH, Seeve C, Yu JM, Pierson LS III, Pierson EA. (2013) Roles of the Gac-Rsm pathway in the regulation of phenazine biosynthesis in Pseudomonas chlororaphis 30–84. MicrobiologyOpen.2:505–24. pmid:23606419
  28. 28. Heeb S, Blumer C, Haas D. (2002) Regulatory RNA as mediator in GacA/RsmA-dependent global control of exoproduct formation in Pseudomonas fluorescens CHA0. J Bacteriol.184:1046–56. pmid:11807065
  29. 29. Selin C, Manuel J, Fernando WD, de Kievit T. (2014) Expression of the Pseudomonas chlororaphis strain PA23 Rsm system is under control of GacA, RpoS, PsrA, quorum sensing and the stringent response. Biol Control.69:24–33.
  30. 30. Heeb S, Haas D. (2001) Regulatory roles of the GacS/GacA two-component system in plant-associated and other gram-negative bacteria. Mol Plant-Microbe Interact.14:1351–63. pmid:11768529
  31. 31. Kay E, Dubuis C, Haas D. (2005) Three small RNAs jointly ensure secondary metabolism and biocontrol in Pseudomonas fluorescens CHA0. Proc Natl Acad Sci USA.102:17136–41. pmid:16286659
  32. 32. Kay E, Humair B, Dénervaud V, Riedel K, Spahr S, Eberl L, et al. (2006) Two GacA-dependent small RNAs modulate the quorum-sensing response in Pseudomonas aeruginosa. J Bacteriol.188:6026–33. pmid:16885472
  33. 33. Lapouge K, Schubert M, Allain FH-T, Haas D. (2008) Gac/Rsm signal transduction pathway of γ‐proteobacteria: from RNA recognition to regulation of social behaviour. Mol Microbiol.67:241–53. pmid:18047567
  34. 34. Blumer C, Heeb S, Pessi G, Haas D. (1999) Global GacA-steered control of cyanide and exoprotease production in Pseudomonas fluorescens involves specific ribosome binding sites. Proc Natl Acad Sci USA.96:14073–8. pmid:10570200
  35. 35. Ren B, Shen H, Lu ZJ, Liu H, Xu Y. (2014) The phzA2-G2 transcript exhibits direct RsmA-mediated activation in Pseudomonas aeruginosa M18. PLoS One.9:e89653. pmid:24586939
  36. 36. Reimmann C, Valverde C, Kay E, Haas D. (2005) Posttranscriptional repression of GacS/GacA-controlled genes by the RNA-binding protein RsmE acting together with RsmA in the biocontrol strain Pseudomonas fluorescens CHA0. J Bacteriol.187:276–85. pmid:15601712
  37. 37. Babitzke P, Romeo T. (2007) CsrB sRNA family: sequestration of RNA-binding regulatory proteins. Curr Opin Microbiol.10:156–63. pmid:17383221
  38. 38. Duss O, Michel E, Yulikov M, Schubert M, Jeschke G, Allain FH-T. (2014) Structural basis of the non-coding RNA RsmZ acting as a protein sponge. Nature.509:588–92. pmid:24828038
  39. 39. Whistler CA, Pierson LS III. (2003) Repression of phenazine antibiotic production in Pseudomonas aureofaciens strain 30–84 by RpeA. J Bacteriol.185:3718–25. pmid:12813064
  40. 40. Wang D, Yu JM, Pierson LS III, Pierson EA. (2012) Differential regulation of phenazine biosynthesis by RpeA and RpeB in Pseudomonas chlororaphis 30–84. Microbiol.158:1745–57.
  41. 41. Chin-A-Woeng TF, van den Broek D, de Voer G, van der Drift KM, Tuinman S, Thomas-Oates JE, et al. (2001) Phenazine-1-carboxamide production in the biocontrol strain Pseudomonas chlororaphis PCL1391 is regulated by multiple factors secreted into the growth medium. Mol Plant-Microbe Interact.14:969–79. pmid:11497469
  42. 42. Selin C, Fernando WD, de Kievit T. (2012) The PhzI/PhzR quorum-sensing system is required for pyrrolnitrin and phenazine production, and exhibits cross-regulation with RpoS in Pseudomonas chlororaphis PA23. Microbiol.158:896–907.
  43. 43. Chen Y, Shen X, Peng H, Hu H, Wang W, Zhang X. (2015) Comparative genomic analysis and phenazine production of Pseudomonas chlororaphis, a plant growth-promoting rhizobacterium. Genomics data.4:33–42. pmid:26484173
  44. 44. Wood DW, Gong F, Daykin MM, Williams P, Pierson LS III. (1997) N-acyl-homoserine lactone-mediated regulation of phenazine gene expression by Pseudomonas aureofaciens 30–84 in the wheat rhizosphere. J Bacteriol.179:7663–70. pmid:9401023
  45. 45. Yu JM, Wang D, Pierson LS, Pierson EA. (2017) Disruption of MiaA provides insights into the regulation of phenazine biosynthesis under suboptimal growth conditions in Pseudomonas chlororaphis 30–84. Microbiol.163:94–108.
  46. 46. Staskawicz B, Dahlbeck D, Keen N, Napoli C. (1987) Molecular characterization of cloned avirulence genes from race 0 and race 1 of Pseudomonas syringae pv. glycinea. J Bacteriol.169:5789–94. pmid:2824447
  47. 47. Hoang TT, Karkhoff-Schweizer RR, Kutchma AJ, Schweizer HP. (1998) A broad-host-range Flp-FRT recombination system for site-specific excision of chromosomally-located DNA sequences: application for isolation of unmarked Pseudomonas aeruginosa mutants. Gene.212:77–86. pmid:9661666
  48. 48. Lesic B, Rahme LG. (2008) Use of the lambda Red recombinase system to rapidly generate mutants in Pseudomonas aeruginosa. BMC Mol Biol.9:20. pmid:18248677
  49. 49. Zuker M. (2003) Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res.31:3406–15. pmid:12824337
  50. 50. Allesen-Holm M, Barken KB, Yang L, Klausen M, Webb JS, Kjelleberg S, et al. (2006) A characterization of DNA release in Pseudomonas aeruginosa cultures and biofilms. Mol Microbiol.59:1114–28. pmid:16430688
  51. 51. Wilkinson H, Cook R, Alldredge J. (1985) Relation of inoculum size and concentration to infection of wheat roots by Gaeumannomyces graminis var. tritici. Phytopathol.75:98–103.
  52. 52. Li Y, Du X, Lu ZJ, Wu D, Zhao Y, Ren B, et al. (2011) Regulatory Feedback Loop of Two phz Gene Clusters through 5′-Untranslated Regions in Pseudomonas sp. M18. PLoS One.6:e19413. pmid:21559370
  53. 53. Humair B, Wackwitz B, Haas D. (2010) GacA-controlled activation of promoters for small RNA genes in Pseudomonas fluorescens. Appl Environ Microbiol.76:1497–506. pmid:20048056
  54. 54. Zhang X, Wang S, Geng H, Ge Y, Huang X, Hu H, et al. (2005) Differential regulation of rsmA gene on biosynthesis of pyoluteorin and phenazine-1-carboxylic acid in Pseudomonas sp. M18. World J Micriobiol Biotech.21:883–9.
  55. 55. Brencic A, Lory S. (2009) Determination of the regulon and identification of novel mRNA targets of Pseudomonas aeruginosa RsmA. Mol Microbiol.72:612–32. pmid:19426209
  56. 56. Kulkarni PR, Jia T, Kuehne SA, Kerkering TM, Morris ER, Searle MS, et al. (2014) A sequence-based approach for prediction of CsrA/RsmA targets in bacteria with experimental validation in Pseudomonas aeruginosa. Nucleic Acids Res.42:6811–25. pmid:24782516
  57. 57. Mercante J, Edwards AN, Dubey AK, Babitzke P, Romeo T. (2009) Molecular geometry of CsrA (RsmA) binding to RNA and its implications for regulated expression. J Mole Biol.392:511–28.
  58. 58. Girard G, Rigali S. (2011) Role of the phenazine-inducing protein Pip in stress resistance of Pseudomonas chlororaphis. Microbiol.157:398–407.
  59. 59. Das T, Manefield M. (2012) Pyocyanin promotes extracellular DNA release in Pseudomonas aeruginosa. PLoS One.7:e46718. pmid:23056420
  60. 60. Huang Z, Bonsall RF, Mavrodi DV, Weller DM, Thomashow LS. (2004) Transformation of Pseudomonas fluorescens with genes for biosynthesis of phenazine-1-carboxylic acid improves biocontrol of rhizoctonia root rot and in situ antibiotic production. FEMS Microbiol Ecol.49:243–51. pmid:19712418
  61. 61. Timms-Wilson T, Ellis R, Renwick A, Rhodes D, Mavrodi D, Weller D, et al. (2000) Chromosomal insertion of phenazine-1-carboxylic acid biosynthetic pathway enhances efficacy of damping-off disease control by Pseudomonas fluorescens. Mol Plant-Microbe Interact.13:1293–300. pmid:11106021
  62. 62. Bailey M, Lilley A, Thompson I, Rainey P, Ellis R. (1995) Site directed chromosomal marking of a fluorescent pseudomonad isolated from the phytosphere of sugar beet; stability and potential for marker gene transfer. Molecular Ecology.4:755–64. pmid:8564013
  63. 63. Sambrook J, Russell DW. Molecular cloning: a laboratory manual. 3rd ed. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory Press; 2001.
  64. 64. Tamura K, Nei M. (1993) Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees. Mol Biol Evol.10:512–26. pmid:8336541
  65. 65. IKumar S, Stecher G, Tamura K. (2016) MEGA7: Molecular Evolutionary Genetics Analysis version 7.0 for bigger datasets. Mol Biol Evol.33:1870–4. pmid:27004904
  66. 66. Miller JH. Experiments in molecular genetics. NY: Cold Spring Harbor Laboratory; 1972. 352–5 p.
  67. 67. O'Toole GA, Kolter R. (1998) Initiation of biofilm formation in Pseudomonas fluorescens WCS365 proceeds via multiple, convergent signalling pathways: a genetic analysis. Mol Microbiol.28:449–61. pmid:9632250