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

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.

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.

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).

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.

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).

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.