Limits...
Evidence for induction of integron-based antibiotic resistance by the SOS response in a clinical setting.

Hocquet D, Llanes C, Thouverez M, Kulasekara HD, Bertrand X, Plésiat P, Mazel D, Miller SI - PLoS Pathog. (2012)

Bottom Line: We found that the metronidazole, not an anti-pseudomonal antibiotic given to the first patient infected with S-Pae, triggered the SOS response that subsequently activated the integrase IntI1 expression.This resulted in the rearrangement of the integron gene cassette array, through excision of the gcuF1 cassette, and the full expression the β-lactamase in an isolate (R-Pae) highly resistant to ceftazidime, which further spread to other patients within our hospital.Our results demonstrate that in human hosts, the antibiotic-induced SOS response in pathogens could play a pivotal role in adaptation process of the bacteria.

View Article: PubMed Central - PubMed

Affiliation: Department of Immunology, Medicine and Microbiology, University of Washington, Seattle, Washington, United States of America.

ABSTRACT
Bacterial resistance to β-lactams may rely on acquired β-lactamases encoded by class 1 integron-borne genes. Rearrangement of integron cassette arrays is mediated by the integrase IntI1. It has been previously established that integrase expression can be activated by the SOS response in vitro, leading to speculation that this is an important clinical mechanism of acquiring resistance. Here we report the first in vivo evidence of the impact of SOS response activated by the antibiotic treatment given to a patient and its output in terms of resistance development. We identified a new mechanism of modulation of antibiotic resistance in integrons, based on the insertion of a genetic element, the gcuF1 cassette, upstream of the integron-borne cassette bla(OXA-28) encoding an extended spectrum β-lactamase. This insertion creates the fused protein GCUF1-OXA-28 and modulates the transcription, the translation, and the secretion of the β-lactamase in a Pseudomonas aeruginosa isolate (S-Pae) susceptible to the third generation cephalosporin ceftazidime. We found that the metronidazole, not an anti-pseudomonal antibiotic given to the first patient infected with S-Pae, triggered the SOS response that subsequently activated the integrase IntI1 expression. This resulted in the rearrangement of the integron gene cassette array, through excision of the gcuF1 cassette, and the full expression the β-lactamase in an isolate (R-Pae) highly resistant to ceftazidime, which further spread to other patients within our hospital. Our results demonstrate that in human hosts, the antibiotic-induced SOS response in pathogens could play a pivotal role in adaptation process of the bacteria.

Show MeSH

Related in: MedlinePlus

gcuF1 gene cassette exists in free circular form.(A) Position and direction of primers used for amplification (Table S2) are indicated by arrows. (B) Visualization of the PCR products on an agarose gel (resulted from the nested PCR with primers circ3 and circ4). PCRs were performed using total DNA from isolates R-Pae1 (without gcuF1, taken as a negative control) and S-Pae (with gcuF1) as templates. MW: Molecular weight.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3375312&req=5

ppat-1002778-g003: gcuF1 gene cassette exists in free circular form.(A) Position and direction of primers used for amplification (Table S2) are indicated by arrows. (B) Visualization of the PCR products on an agarose gel (resulted from the nested PCR with primers circ3 and circ4). PCRs were performed using total DNA from isolates R-Pae1 (without gcuF1, taken as a negative control) and S-Pae (with gcuF1) as templates. MW: Molecular weight.

Mentions: S-Pae differed from R-Pae1 by a 10-fold lower amount of blaOXA-28 transcripts (Figure 2B) and by the presence of a 319-bp cassette, gcuF1, inserted immediately upstream of blaOXA-28 (Figure 1B). Using nested-PCR, we demonstrated the presence of free circular cassettes of gcuF1 in S-Pae (Figure 3A, B), demonstrating that the recombination between its own attC site and the aacA4 attC site was occurring, though at extremely low level [2]. Computational analysis of the nucleotide sequence of the S-Pae integron predicted the translation of a new ORF, a fused protein consisting of gcuF1 and blaOXA-28 (GCUF1-OXA-28). We were able to show that these two cassettes (gcuF1 and blaOXA-28) could be transcribed in a single transcript. Hence, we could retrieve a specific amplicon after PCR amplification using cDNA prepared from S-Pae RNA as the matrix and with primers overlapping the junction between gcuF1 and blaOXA-28 (Figure S3). The GCUF1-OXA-28 peptide (368 residues) was predicted to have a molecular weight of 40.1 kDa, compared with the 29.3-kDa native OXA-28 (266 residues). To confirm these data, both ORFs were expressed in Escherichia coli BL21 from plasmid pET-28a which adds an N-terminal polyHis tag. After purification, we found that their molecular weights estimated by SDS-PAGE were in full agreement with our predictions (Figure 1D). This protein contained the original 19-residue long signal peptide now misplaced between the GCUF1 and the OXA-28 domains at position 103–121 of the GCUF1-OXA-28 protein (Figure 1B and 4A). Since β-lactamases are periplasmic proteins and are produced as preproteins with an N-terminal peptide signal [20], one would expect that the misplacement of the signal peptide in the GCUF1-OXA-28 protein will abolish the periplasmic process of the β-lactamase. However, cellular production of this altered protein conferred a residual resistance to ceftazidime (MIC of ceftazidime, 8 µg/ml; gcuF1-blaOXA-28 in Figure 1C), suggesting the presence of an active and processed OXA-28 in the periplasm of the GCUF1-OXA28-producing isolate. To clarify this point, we cloned the blaOXA-28 and gcuF1-blaOXA-28 sequences into the broad host range vector pBTK27 to encode C-terminal His-tagged polypeptides that were expressed in the reference strain P. aeruginosa PA14ΔampC. Western-blot analysis of periplasmic extracts of GCUF1-OXA28-producing bacteria revealed the presence of a reduced amount of processed periplasmic OXA-28 (Figure 1D), consistent with the lower resistance to ceftazidime when the gcuF1 cassette is inserted upstream of the blaOXA-28 (Figure 1C). We used a directed mutagenesis approach to clarify the origin of periplasmic OXA-28 in S-Pae and determine whether it is due to the export processing of the fusion protein or to an internal translational initiation at the original OXA-28 start codon (Figure 4A). In P. aeruginosa PA14ΔampC carrying a plasmid-borne gcuF1-blaOXA-28, the in-frame insertion of a stop codon just upstream of blaOXA-28 reduced the resistance level to ceftazidime down to 2 µg/ml. We also tested the effect of the in frame deletion of the blaOXA-28 ribosome binding site (GAAGGT), or its substitution by a sequence with no ribosome binding properties (CTCTCT). Finally, we tested the substitution of the ATG start codon with either a GTC or a GTG valine codons, which have no or weak translation initiation power [21]. None of these mutations led to a change in resistance level (Figure 1C). Hence, we confirmed that detected OXA-28 came entirely from the processing of the ORF2-OXA28 fusion protein, and that the inefficiency of its maturation was in part responsible for the low resistance level to ceftazidime. Additionally, the putative ribosome binding site of gcuF1 (TTAGG) is predicted to have a poor translation initiation efficiency [22], [23] (Figure 4A), likely leading to reduced translation of gcuF1-blaOXA-28.


Evidence for induction of integron-based antibiotic resistance by the SOS response in a clinical setting.

Hocquet D, Llanes C, Thouverez M, Kulasekara HD, Bertrand X, Plésiat P, Mazel D, Miller SI - PLoS Pathog. (2012)

gcuF1 gene cassette exists in free circular form.(A) Position and direction of primers used for amplification (Table S2) are indicated by arrows. (B) Visualization of the PCR products on an agarose gel (resulted from the nested PCR with primers circ3 and circ4). PCRs were performed using total DNA from isolates R-Pae1 (without gcuF1, taken as a negative control) and S-Pae (with gcuF1) as templates. MW: Molecular weight.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3375312&req=5

ppat-1002778-g003: gcuF1 gene cassette exists in free circular form.(A) Position and direction of primers used for amplification (Table S2) are indicated by arrows. (B) Visualization of the PCR products on an agarose gel (resulted from the nested PCR with primers circ3 and circ4). PCRs were performed using total DNA from isolates R-Pae1 (without gcuF1, taken as a negative control) and S-Pae (with gcuF1) as templates. MW: Molecular weight.
Mentions: S-Pae differed from R-Pae1 by a 10-fold lower amount of blaOXA-28 transcripts (Figure 2B) and by the presence of a 319-bp cassette, gcuF1, inserted immediately upstream of blaOXA-28 (Figure 1B). Using nested-PCR, we demonstrated the presence of free circular cassettes of gcuF1 in S-Pae (Figure 3A, B), demonstrating that the recombination between its own attC site and the aacA4 attC site was occurring, though at extremely low level [2]. Computational analysis of the nucleotide sequence of the S-Pae integron predicted the translation of a new ORF, a fused protein consisting of gcuF1 and blaOXA-28 (GCUF1-OXA-28). We were able to show that these two cassettes (gcuF1 and blaOXA-28) could be transcribed in a single transcript. Hence, we could retrieve a specific amplicon after PCR amplification using cDNA prepared from S-Pae RNA as the matrix and with primers overlapping the junction between gcuF1 and blaOXA-28 (Figure S3). The GCUF1-OXA-28 peptide (368 residues) was predicted to have a molecular weight of 40.1 kDa, compared with the 29.3-kDa native OXA-28 (266 residues). To confirm these data, both ORFs were expressed in Escherichia coli BL21 from plasmid pET-28a which adds an N-terminal polyHis tag. After purification, we found that their molecular weights estimated by SDS-PAGE were in full agreement with our predictions (Figure 1D). This protein contained the original 19-residue long signal peptide now misplaced between the GCUF1 and the OXA-28 domains at position 103–121 of the GCUF1-OXA-28 protein (Figure 1B and 4A). Since β-lactamases are periplasmic proteins and are produced as preproteins with an N-terminal peptide signal [20], one would expect that the misplacement of the signal peptide in the GCUF1-OXA-28 protein will abolish the periplasmic process of the β-lactamase. However, cellular production of this altered protein conferred a residual resistance to ceftazidime (MIC of ceftazidime, 8 µg/ml; gcuF1-blaOXA-28 in Figure 1C), suggesting the presence of an active and processed OXA-28 in the periplasm of the GCUF1-OXA28-producing isolate. To clarify this point, we cloned the blaOXA-28 and gcuF1-blaOXA-28 sequences into the broad host range vector pBTK27 to encode C-terminal His-tagged polypeptides that were expressed in the reference strain P. aeruginosa PA14ΔampC. Western-blot analysis of periplasmic extracts of GCUF1-OXA28-producing bacteria revealed the presence of a reduced amount of processed periplasmic OXA-28 (Figure 1D), consistent with the lower resistance to ceftazidime when the gcuF1 cassette is inserted upstream of the blaOXA-28 (Figure 1C). We used a directed mutagenesis approach to clarify the origin of periplasmic OXA-28 in S-Pae and determine whether it is due to the export processing of the fusion protein or to an internal translational initiation at the original OXA-28 start codon (Figure 4A). In P. aeruginosa PA14ΔampC carrying a plasmid-borne gcuF1-blaOXA-28, the in-frame insertion of a stop codon just upstream of blaOXA-28 reduced the resistance level to ceftazidime down to 2 µg/ml. We also tested the effect of the in frame deletion of the blaOXA-28 ribosome binding site (GAAGGT), or its substitution by a sequence with no ribosome binding properties (CTCTCT). Finally, we tested the substitution of the ATG start codon with either a GTC or a GTG valine codons, which have no or weak translation initiation power [21]. None of these mutations led to a change in resistance level (Figure 1C). Hence, we confirmed that detected OXA-28 came entirely from the processing of the ORF2-OXA28 fusion protein, and that the inefficiency of its maturation was in part responsible for the low resistance level to ceftazidime. Additionally, the putative ribosome binding site of gcuF1 (TTAGG) is predicted to have a poor translation initiation efficiency [22], [23] (Figure 4A), likely leading to reduced translation of gcuF1-blaOXA-28.

Bottom Line: We found that the metronidazole, not an anti-pseudomonal antibiotic given to the first patient infected with S-Pae, triggered the SOS response that subsequently activated the integrase IntI1 expression.This resulted in the rearrangement of the integron gene cassette array, through excision of the gcuF1 cassette, and the full expression the β-lactamase in an isolate (R-Pae) highly resistant to ceftazidime, which further spread to other patients within our hospital.Our results demonstrate that in human hosts, the antibiotic-induced SOS response in pathogens could play a pivotal role in adaptation process of the bacteria.

View Article: PubMed Central - PubMed

Affiliation: Department of Immunology, Medicine and Microbiology, University of Washington, Seattle, Washington, United States of America.

ABSTRACT
Bacterial resistance to β-lactams may rely on acquired β-lactamases encoded by class 1 integron-borne genes. Rearrangement of integron cassette arrays is mediated by the integrase IntI1. It has been previously established that integrase expression can be activated by the SOS response in vitro, leading to speculation that this is an important clinical mechanism of acquiring resistance. Here we report the first in vivo evidence of the impact of SOS response activated by the antibiotic treatment given to a patient and its output in terms of resistance development. We identified a new mechanism of modulation of antibiotic resistance in integrons, based on the insertion of a genetic element, the gcuF1 cassette, upstream of the integron-borne cassette bla(OXA-28) encoding an extended spectrum β-lactamase. This insertion creates the fused protein GCUF1-OXA-28 and modulates the transcription, the translation, and the secretion of the β-lactamase in a Pseudomonas aeruginosa isolate (S-Pae) susceptible to the third generation cephalosporin ceftazidime. We found that the metronidazole, not an anti-pseudomonal antibiotic given to the first patient infected with S-Pae, triggered the SOS response that subsequently activated the integrase IntI1 expression. This resulted in the rearrangement of the integron gene cassette array, through excision of the gcuF1 cassette, and the full expression the β-lactamase in an isolate (R-Pae) highly resistant to ceftazidime, which further spread to other patients within our hospital. Our results demonstrate that in human hosts, the antibiotic-induced SOS response in pathogens could play a pivotal role in adaptation process of the bacteria.

Show MeSH
Related in: MedlinePlus