SEARCH

SEARCH BY CITATION

Keywords:

  • antibiotic;
  • interspecific signal;
  • quorum sensing;
  • violacein;
  • bacterial virulence

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Increasing evidence has shown that antibiotics function as intermicrobial signaling molecules instead of killing weapons. However, mechanisms and key factors that are involved in such functions remain poorly understood. Earlier findings have associated antibiotic signaling with quorum sensing (QS); however, results varied among experiments, antibiotics, and bacterial strains. In this study, we found that antibiotics at subinhibitory concentrations improved the violacein-producing ability of Chromobacterium violaceum ATCC 12472. Quantitative real-time polymerase chain reaction of QS-associated gene transcripts and bioassay of violacein production in a QS mutant strain demonstrated that antibiotics enhanced the production of N-acyl-l-homoserine lactones (AHLs; QS signaling molecules) and increased AHL-inducing QS-mediated virulence, including chitinase production and biofilm formation. Moreover, a positive flagellar activity and an increased bacterial clustering ability were found, which are related to the antibiotic-induced biofilm formation. Our findings suggested that antibiotic-mediated interspecific signaling also occurs in C. violaceum, thereby expanding the knowledge and language of cell-to-cell communication.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Quorum sensing (QS) is a classical population density-dependent signaling mechanism that allows bacteria to behave coordinately based on the size of their population (DeAngelis et al., 2008; Defoirdt et al., 2008). In Gram-negative bacteria, QS is mediated by N-acyl-l-homoserine lactones (AHLs), intraspecific signal molecules, which are known as autoinducers (AIs). Among the diverse species, the AHL side chain may vary in length (C4 to C16) and in the substitution at C3 (oxo or hydroxyl functions as well as no substitution) (Schaefer et al., 2000; Marketon et al., 2002). The bacterial QS system controls a wide range of prokaryotic phenotypes, including enzyme secretion, virulence factor production, bioluminescence, and biofilm development by monitoring fluctuations in AI concentrations (Niu et al., 2006; Babić et al., 2010). Moreover, the bacterial QS system is important in the response mechanism against environmental deterioration. For instance, in pathogenic bacteria, QS enables organisms to fight against attacks and express virulence factors to overwhelm the host's immune response (Moine & Abraham, 2004).

Most antibiotics are derived from naturally occurring microbial products. In general, environmentally existing antibiotics are ecologically used to fight against competitors (Hoffman et al., 2005). However, the antibiotic concentration is significantly lower than the minimum inhibitory concentration (MIC) in environments or in postantibiotic phases of clinical therapy. An increasing number of findings have revealed that antibiotics function as interspecific signals (Davies et al., 1998, 2006; Yim et al., 2007; Skindersoe et al., 2008).

Given that antibiotic signaling and QS are global regulatory mechanisms in bacteria, few studies have focused on the possible link that exists between antibiotic signaling and QS. For example, Garske et al. (2004) found that both ceftazidime and tobramycin reduce C12-homoserine lactone (C12-HSL) and C4-homoserine lactone (C4-HSL) in Pseudomonas aeruginosa. Shen et al. (2008) found that vancomycin, tetracycline, ampicillin, and azithromycin activate the expression of QS-related virulence factors in a QS-independent manner (Garske et al., 2004; Shen et al., 2008). These findings indicated that antibiotic signaling is diverse among bacteria.

In this study, we collected direct evidence that antibiotics at subinhibitory concentrations (SICs) enhanced the QS-controlled behaviors, including violacein production, chitinolytic activity, and biofilm formation in Chromobacterium violaceum.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Materials

All of the chemicals used were of analytical grade or higher unless specifically indicated. Antibiotics were purchased from Sigma-Aldrich (USA). Furanone compound 30 (C30; a QS inhibitor) and homoserine lactone (C10-HSL; a QS signal molecule) were synthesized at the Synthetic Medicinal Chemistry Laboratory of School of Medicine and Pharmacy, Ocean University of China.

Bacterial strains and growth media

Two related strains of C. violaceum were used in this study: the wild-type C. violaceum ATCC 12472 and a mutant that is affected in QS regulation. CV12472 is a Gram-negative bacterium that produces a natural antibiotic (violacein) under the control of QS (Chernin et al., 1998). For CV12472 strain, the QS signal molecule AHL is synthesized by AI synthase CviI and recognized by the transcriptional activator protein CviR. Its cviI deletion mutant VIR24 is defective in AHL production and thus nonpigmented. Exogenous AHLs can retrieve violacein synthesis in this mutant (Someya et al., 2009). Both strains were grown in Luria–Bertani (LB) broth at 30 °C with shaking (150 r.p.m.).

Determination of MICs

MICs of antibiotics for CV12472 were determined in LB broth after the strains were incubated at 30 °C for 24 h according to the dilution method described in a previous study (Maravic Vlahovicek et al., 2008). In brief, CV12472 was cultured overnight, then diluted to a density of 0.002 (OD600), mixed with an equal volume of LB broth that contains 2–256 μg mL−1 of antibiotics, and cultured in 96-well flat-bottom plastic microplates at 30 °C for 24 h without shaking. MIC is determined as the minimum concentration of antibiotics that prevents the visible growth of bacteria (OD600 nm ≤ 0.01).

Determination of crude violacein concentration

After the cells were incubated in glass tubes at 30 °C for 16 h, 1 mL of the cells was centrifuged at 14 000 g for 20 min and resuspended in 1 mL of absolute ethanol to dissolve the crude violacein (violacein and deoxyviolacein). The supernatant was collected after the cells were centrifuged at 12 000 g for 10 min. The amount of crude violacein was determined according to the absorbance at 575 nm (Blosser & Gray, 2000).

Determination of AI levels

The overnight culture of CV12472 was diluted with 50 mL of fresh LB broth and grown to a density of 2.0 (OD600 nm). AI molecules were extracted from the used medium with an equal volume of ethyl acetate that was acidified with 0.1% acetic acid. The extract was dried and dissolved in 50 μL of sterile methyl hydrate as described previously (Steindler et al., 2009).

VIR24 strain was used for AI quantification. The overnight culture (4 mL) was mixed with molten solid LB agar and quickly plated. The wells were punched on the agar plate using a sterile cork borer (6 mm in diameter). AI extract (10 μL) was added into the wells before the plate was dried. The plate was then incubated at 30 °C for at least 16 h (Someya et al., 2009).

Quantitative analysis of cviI and vioB

Transcript abundance of cviI and vioB was determined by quantitative real-time reverse-transcription polymerase chain reaction (RT-PCR). The primers used are listed in Table 1. The abundance was normalized by referring to the housekeeping 16S ribosomal RNA gene (Pfaffl, 2001).

Table 1. The quantitative real-time RT-PCR primers used in this study
GeneForward primersReverse primers
cvi I 5′CTGAAACTAAGCTGCGACAGTTG3′5′GAAACCGTCCTCGCATAAGG3′
vio B 5′ACGGTGGAGGAGGTCGATTA 3′5′CGCACATCTGCCACATCAG 3′
16S 5′ GCGCAACCCTTGTCCTTAGTT 3′5′TGTCACCGGCAGTCTCCTTAG 3′

Determination of chitinolytic activity on agar plates

Cells were evenly inoculated on a 1.5% agar plate that contained 0.4% K2HPO4, 0.2% KH2PO4, MgSO4•7H2O, 0.1% NaCl, 0.1% yeast extract, and 1.0% chitin colloid. Unless otherwise indicated, the antibiotics (1/6 MIC of kanamycin), C30 (20 μM), or their combination were added. The plate was incubated at 30 °C for 48–72 h until transparent zones appeared around the colonies (Chernin et al., 1998; Folders et al., 2001).

Biofilm formation assay

Biofilm was prepared as described previously (O'Toole et al., 1999; Carter et al., 2003). In brief, cells (107 CFU mL−1) were cultured in glass tubes at 30 °C for 24 h with shaking (150 rpm). Biofilm developed was routinely quantified by reading absorbance of crystal violet-stained adherent cells at 590 nm. Each assay was performed in triplicate and repeated thrice.

Motility assay

Swimming ability was assayed on semi-flowing agar plates as described previously (Deziel et al., 2001; Fonseca et al., 2004). The medium contained 1% tryptone, 0.5% NaCl, and 0.3% agar. Overnight culture (1 μL) was spotted onto plates with or without antibiotics. Motility was assessed according to the diameter of the turbid zone.

Adherence assay

Overnight culture of CV12472 was diluted 100-fold with LB medium and cultured for 6 h. The cells were collected by centrifugation, washed thrice with PBS, and resuspended in PBS with the density adjusted to 0.01 (OD600 nm). The diluents were inoculated into wells of 24-well flat-bottom plastic microplates and cultured at 30 °C for 1 or 5 h without shaking. The cells that adhered to the walls of the wells were stained with crystal violet for 10 min, washed five times with PBS, and observed under an inverted microscope (Olympus IX70)(Letourneau et al., 2011).

Agar plating assay

Agar plating assay was performed according to the method described previously (Allison et al., 2011). Well-developed biofilm was obtained and washed with sterile PBS. To dislodge biofilm cells, biofilm was resuspended in PBS, sonicated in water bath for 30 min at 40 kHz, and serially diluted. Each diluent was spot-plated onto LB agar plates, and CFUs were determined after the colonies were cultured for 16 h. Only those plates with 10–100 colonies were used in our assay.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Effect of antibiotics (at SICs) on violacein production

An enhanced violacein production was observed at the edge of the kanamycin inhibition zone when the CV12472 strain was grown on the plate (Fig. 1a). The antibiotics diffused from the split-spot to the edge, which revealed that such an enhancement was caused by the antibiotics in a concentration-dependent manner. To verify our hypothesis, the strain was incubated with serial concentrations of kanamycin. After the strain was incubated for 16 h, crude violacein was extracted and quantified. Violacein production was promoted (Fig. 1b) when the antibiotic concentration was 1/4, 1/6, and 1/8 of MICs (MIC of kanamycin, 8 μg mL−1). A peak induction occurred and the amount of violacein increased by 25% at 1/6 MIC of kanamycin. To detect whether or not this feature is specific to kanamycin, we screened several antibiotics other than the aminoglycoside. Similar results were observed in 1/4 MIC of amikacin, 1/2 MIC of gentamycin, 1/16 MIC of tetracycline, and 1/8 MIC of erythromycin (Fig. 1c). The MICs of these antibiotics were 16, 8, 2, and 4 μg mL−1 of amikacin, gentamycin, tetracycline, and erythromycin, respectively, under the growth conditions used. Bacterial growth was not inhibited by the SICs of antibiotics based on CFUs (data not shown).

image

Figure 1. Effect of antibiotics at SICs on violacein production of CV12472 bacterial strain. (a) Assay of the effect of kanamycin at SICs on violacein production on agar plate. Violacein production was enhanced (arrow pointing to the zone). (b) Assay of the effect of kanamycin at SICs on violacein production in glass tubes. CV12472 strain was treated with 1/2, 1/4, 1/6, and 1/8 MICs of kanamycin. (c) Assay of other antibiotics at SICs on violacein production in glass tubes. CV12472 strain was treated with 1/4 MIC of amikacin, 1/2 MIC of gentamycin, 1/16 MIC of tetracycline, and 1/8 MIC of erythromycin. Data are shown as means ± SEM and reflect the combined results of three independent experiments. Ami, amikacin; Gen, gentamycin; Tet, tetracycline; Ery, erythromycin.

Download figure to PowerPoint

Effect of antibiotics (at SICs) on AI production

Fig. 2(a) shows that VIR24 strain almost lost the ability to produce violacein because AI synthetase was deleted, indicating that QS is important for the violacein production in C. violaceum. Such an association was further verified by adding C30, which suppresses QS by accelerating the turnover of AI receptors (Manefield et al., 2002; Wu et al., 2004). When 20 μM C30 was added in the wild-strain culture, violacein production was inhibited approximately 50% (Fig. 2a). To determine the function of QS in antibiotic-induced violacein production, we added 1/6 MIC of kanamycin to the CV12472 culture (with 20 μM C30) and the VIR24 culture (with 10 μM C10-HSL). At 10 μM C10-HSL, the violacein-producing ability of VIR24 was partly retrieved. We also found that C30 nearly inhibited the antibiotic effect in the wild-strain culture, and the antibiotics no longer increased the violacein production in VIR24 with C10-HSL (Fig. 2a). All of these findings revealed that antibiotics promote violacein production in a QS-dependent manner.

image

Figure 2. Association of QS with antibiotic-induced violacein production. (a) Violacein production assay of CV12472 and VIR24 (CV12472 ΔcviI) bacterial strains. CV12472 strain was treated with 1/6 MIC of kanamycin, 20 μM furanone compound 30 (C30), or their combination. VIR24 strain was treated with either 1/6 MIC of kanamycin or a combination of 1/6 MIC of kanamycin and 10 μM C10-homoserine lactone. (b) Bioassay of autoinducers using VIR24. a) Control group; b) 1/2, c) 1/4, d) 1/6, and e) 1/8 MICs of kanamycin. (c) Relative abundance of vioB and cviI transcripts detected with quantitative real-time reverse-transcription polymerase chain reaction. Data are shown as means ± SEM and reflect the combined results of three independent experiments. Kan, kanamycin.

Download figure to PowerPoint

Previous studies demonstrated that violacein production is proportional to AI concentration (McClean et al., 1997). To detect the influence of antibiotics on AI production, we extracted AI from the CV12472 culture with kanamycin and assayed the amount of AI with VIR24. We found that the violacein-producing ability of VIR24 was recovered by the extracted AI, and the highest retrieving power was achieved from the LB medium that contains 1/6 MIC of kanamycin (Fig. 2b). Moreover, quantitative real-time RT-PCR assay showed that AI synthetase gene (cviI) and one of the violacein biosynthetic genes (vioB) were upregulated (Fig. 2c).

Effect of antibiotics (at SICs) on chitinase production

Chitinase production is another QS-controlled trait of CV12472 (Chernin et al., 1998). The chitinase activity was determined by inoculating the bacterial cells on the agar–colloidal chitin plate that was supplemented with either antibiotics or C30. After the cells were cultured at 30 °C for 48–72 h, a clearing zone around the growing bacteria was observed. We found that the diameter of the hydrolyzing zone of the antibiotic plates was significantly larger than that of the control plate, whereas the zone diameter on the plate that contains both antibiotics and 20 μM C30 was similar to that of the control (Fig. 3).

image

Figure 3. Effect of kanamycin at SICs on the chitinolytic activity of CV12472 bacterial strain. CV12472 strain was treated with 1/6 MIC of kanamycin, 20 μM furanone compound 30 (C30), or their combination. Chitinolytic activity was assayed on a plate supplemented with colloidal chitin based on the clearing zone diameter. After the strain was cultured for 48 h, the diameter was measured: 11 mm (control plate), 13 mm (1/6 MIC of kanamycin), 7 mm (20 μM C30), and 7.3 mm (20 μM C30 + 1/6 MIC of kanamycin). Kan, kanamycin.

Download figure to PowerPoint

Effect of antibiotics (at SICs) on biofilm formation

Regulation of biofilm formation is extremely complex, which is attributed to an array of coordinated regulatory mechanisms (Davies et al., 1998). The effect of QS on biofilm formation depends on strains and testing conditions. In this study, we initially determined the association of QS with biofilm formation in C. violaceum. Wild strain CV12472 was cultured in glass tubes, and the formed biofilm was quantified at A590 nm after crystal violet staining was performed. After 24 h of incubation, the biomass in the biofilm was significantly reduced when 20 μM C30 was added, indicating that QS is necessary in biofilm formation. Further verification was performed in the QS-deficient strain VIR24, which lost most of the biofilm formation ability (data not shown). Furthermore, we examined the role of antibiotics at SICs on biofilm formation. When 1/6 MIC of kanamycin was added, biofilm formation was increased by approximately 34%, whereas such increase was blocked by C30 (Fig. 4).

image

Figure 4. Effect of kanamycin at SICs on biofilm formation of CV12472 bacterial strain. CV12472 strain was treated with 1/6 MIC of kanamycin, 20 μM furanone compound 30 (C30), or their combination. The biofilm that formed was quantified at an absorbance of 590 nm of the crystal violet-stained adherent bacteria. Data are shown as means ± SEM and reflect the combined results of three independent experiments. Kan, kanamycin.

Download figure to PowerPoint

To determine the mechanism involved in the antibiotic-induced biofilm formation, we investigated the swimming ability of CV12472, which has an important effect on early biofilm formation or biofilm structure maintenance (Deziel et al., 2001; Atkinson et al., 2006; Kozlova et al., 2011). Fig. 5(a) shows that the swimming ability was activated when bacteria were inoculated onto a soft agar plate with antibiotics, whereas initial adherence difference was not observed under the microscope. Aggregation ability was enhanced and a higher number of microcolonies appeared when the strain was treated with antibiotics for 5 h (Fig. 5b). The number of bacteria in antibiotic-treated biofilm was about twofold of that in the control plate (Fig. 5c).

image

Figure 5. Factors that affect antibiotic-induced biofilm formation. (a) Swimming motility assay. CV12472 bacterial strain was inoculated on 0.3% Luria–Bertani agar plates with or without antibiotics. (b) Bacterial attachment and clustering ability were observed under an inverted microscope. (c) Agar plating assay was used to determine the number of bacteria in the biofilm. Results are averages of 5 replicates ± SEM. < 0.05 as compared with the same strain exposed to no antibiotics. Kan, kanamycin.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In microbiology, MIC is used to express the minimum concentration of antibiotics that inhibit the visible growth of clinically significant numbers of bacteria in vitro (Babić et al., 2010). However, in the postantibiotic treatment period in clinical therapy or in the environment, the antibiotic concentration is always lower than MIC. Such concentrations are defined as SICs, which are one of the major causes of drug resistance, although bacterial growth is not affected by antibiotics at SICs (Taylor et al., 2005; Martinez et al., 2009). In this study, we found that the purple area in CV12472 was darker at the edge of the kanamycin inhibition zone, which prompted us to determine the mechanism involved in the antibiotic-induced violacein production. We also found that antibiotics at some SICs enhanced the violacein production without affecting the bacterial growth. Different antibiotics function similarly, indicating that a common mechanism is likely involved in the antibiotic-enhanced violacein production. We focused on QS, particularly in bacterial QS, which is a density-dependent population behavior. QS directly regulates violacein production in C. violaceum (Chernin et al., 1998; Cegelski et al., 2008; Boyer & Wisniewski-Dye, 2009). Our findings on AI production and AI synthetase gene transcript indicated that the antibiotic-induced violacein production was mediated by QS. Because of the diverse structure of antibiotics, the doubt that antibiotics acted as AHL analogues was precluded. Earlier studies found that some antibiotics at SICs function as signal molecules, which affected 5–10% of all transcripts including some QS-related regulators like rsmA in bacteria (Goh et al., 2002; Linares et al., 2006; Yim et al., 2007). Those results may be helpful to determine the molecular mechanisms in antibiotic signaling and identify the primary response components that could trigger the regulatory cascade in future studies.

Chitin is the major structural component of various organisms, including fungi, insects, and crustaceans (Chernin et al., 1998). Chitinase is an enzyme that functions as a protective component of some bacteria in fighting against fungal or other pathogens with chitinous exoskeleton (Neiendam Nielsen & Sorensen, 1999; Nguyen et al., 2008). In C. violaceum, chitinase production was regulated by QS and upregulated by antibiotics, indicating that the antibiotics at SICs contribute to the success of C. violaceum in its competition with other inhabitants by improving QS-regulated behaviors.

Bacterial biofilm is a dense microorganism community with low sensitivity to antibiotics and high resistance to environmental stress (Davies et al., 1998). Antibiotics at SICs induce biofilm formation in P. aeruginosa (Hoffman et al., 2005). In this study, we demonstrated that QS was necessary for the biofilm formation in C. violaceum. Biofilm was induced when C. violaceum was cultured with kanamycin. The antibiotic-induced biofilm formation was related to the increase in the swimming motility and bacterial aggregation.

In summary, we found that the antibiotic-mediated interspecific signal at SIC can also function intraspecifically, which provides additional information in understanding the ecological function of antibiotics in nature. At high concentrations, antibiotics are bacterial killers, whereas at low concentrations, antibiotics are beneficial for the survival of susceptible bacteria in natural environments. The antibiotic-induced virulence and biofilm formation might help guide antibiotic therapy and lead to the development of novel co-therapeutic drugs, which could suppress the QS-related response. However, further studies on the correlation of antibiotics and QS at molecular level were appreciated.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We would like to thank Dr. R.J.C. Mc Lean (Texas State University, USA) and Dr. Tomohiro Morohoshi (Utusnomiya University, Japan) for providing CV12472 and VIR24 strains, respectively. This work was supported by the National Natural Science Foundation of China (Grant nos. 31070712 and 81102368), the Special Fund for Marine Scientific Research in the Public Interest (Grant no. 201105027), and the National High-technology Research and Development (863 Program; Grant no. 2011AA09070304).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References