Identification and functional analysis of a bacteriocin, pyocin S6, with ribonuclease activity from a Pseudomonas aeruginosa cystic fibrosis clinical isolate

Abstract

S-type pyocins are bacteriocins produced by Pseudomonas aeruginosa isolates to antagonize or kill other strains of the same species. They have a modular organization comprising a receptor-binding domain recognizing a surface constituent of the target bacterium, a domain for translocation through the periplasm, and a killing or toxic domain with DNase, tRNase, or pore-forming activity. Pyocins S2, S3, S4, and S5 recognize TonB-dependent ferri-siderophore receptors in the outer membrane. We here describe a new nuclease bacteriocin, pyocin S6, encoded in the genome of a P. aeruginosa cystic fibrosis (CF) clinical isolate, CF_PA39. Similarly to pyocins S1 and S2, the S6 toxin-immunity gene tandem was recruited to the genomic region encoding exotoxin A. The pyocin S6 receptor-binding and translocation domains are identical to those of pyocin S1, whereas the killing domain is similar to the 16S ribonuclease domain of Escherichia coli colicin E3. The cytotoxic activity was abolished in pyocin S6 forms with a mutation in the colicin E3-equivalent catalytic motif. The CF_PA39 S6 immunity gene displays a higher expression level than the gene encoding the killing protein, the latter being only detected when bacteria are grown under iron-limiting conditions. In the S1-pyocinogenic strain P. aeruginosa ATCC 25324 and pyocin S2 producer P. aeruginosa PAO1, a remnant of the pyocin S6 killing domain and an intact S6-type immunity gene are located downstream of their respective pyocin operons. Strain PAO1 is insensitive for pyocin S6, and its S6-type immunity gene provides protection against pyocin S6 activity. Purified pyocin S6 inhibits one-fifth of 110 P. aeruginosa CF clinical isolates tested, showing clearer inhibition zones when the target cells are grown under iron limitation. In this panel, about half of the CF clinical isolates were found to host the S6 genes. The pyocin S6 locus is also present in the genome of some non-CF clinical isolates.

Keywords: 16S ribonuclease; Pseudomonas aeruginosa; cystic fibrosis; iron; pyocins..

Figure 1

Genetic organization of the pyocin S6 locus in the genome of P. aeruginosa CF_PA39. Comparison between the pyocin S6 locus in P. aeruginosa CF_PA39 with the loci encoding pyocin S1 (P. aeruginosa ATCC 25324), pyocin S2 (P. aeruginosa PAO1), and putative pyocin S7 (P. aeruginosa BWH057). The equivalent immunity‐encoding genomic region of P. aeruginosa DK2 lacking a pyocin toxin gene is included. Synteny is visualized by sequence conservation (light gray shading) for the genes (represented by arrows) located between the orthologs of toxA (PA1148) and PAA1153 of P. aeruginosa PAO1 (nonannotated ORFs marked with an asterisk). The color code specifies the type of encoded nuclease and cognate or orphan immunity protein (Imm). Conserved receptor‐binding domains (RBD) are highlighted in the same color. Darker gray shading delineates the region encoding the translocation domain, conserved across the four pyocins. Dotted lines indicates the lack of an equivalent sequence. The same pyocin S6 gene context is found in P. aeruginosa strains AZPAE14840 and AZPAE14899. In several other S6 pyocinogenic strains the unknown gene upstream of CF_PA39 toxA, absent from strain PAO1, is equally lacking (AZPAE14862, AZPAE14931, AZPAE14934, AZPAE14937, AZPAE14951, AZPAE14976, 130_PAER, PA45) (Table S2). The P. aeruginosa PA45 locus further differs by the absence of the unknown gene downstream of the CF_PA39 PA1153 homologue.

Figure 2

Structural similarity of the killing domains of pyocin S6 and colicin E3. (A). 3D model of the KD of pyocin S6 (B). Cartoon representation of the rRNase KD of colicin E3. Side chains of conserved residues contributing to the activity in colicin E3 (B), and corresponding residues in pyocin S6 (A) are shown as sticks. Amino‐ and carboxy‐termini of the KDs are marked with N and C, respectively. (C). Sequence alignment of the KDs of PyS6 and ColE3. Gray shading reflects the degree of conservation. Residues involved in toxic activity of ColE3, and the corresponding residues in PyS6 are boxed in red. PyS6 residues that were mutated to alanine in this study are marked with an asterisk.

Figure 3

Effect of iron limitation on the expression of pys6 and imm6. Using semiquantitative reverse transcriptase (RT) RT‐PCR, the expression of pys6 (884 bp) and oprI (248 bp) (A) and imm6 (127 bp) and oprI (248 bp) (B) was determined for P. aeruginosa CF_PA17 and P. aeruginosa CF_PA39 grown in LB or casamino acids (CAA). +RT and −RT indicate that reverse transcriptase was added or not, respectively. Fifty nanograms of genomic DNA isolated from P. aeruginosa CF_PA39 was included as a positive control (lanes marked with a “+”). The molecular weight size marker is shown in the lanes marked with an “L”. The lower bands observed in the ‐RT reactions represent primer dimers and are not observed in the +RT reactions. pys6, pyocin S6 gene. imm6, pyocin S6 immunity gene.

Figure 4

Determination of pyocin S6 activity and MIC (minimum inhibitory concentration). Of the 110 tested strains, 22 were found to be sensitive to the purified protein. Clear zones of inhibition only appeared when grown on casamino acids rather than on LB medium, as demonstrated here for the strain PA109 in figures B and A, respectively. (C) The MIC was determined for CF_PA109 by a twofold serial dilution; number 1 represents the undiluted purified fraction (17 mg mL⁻¹). As dilutions increase, the killing zones become smaller and then become gradually more opaque due to incomplete inhibition.