Correspondence: Ranadhir Chakraborty, Department of Biotechnology, University of North Bengal, PO NBU, Siliguri, 734 013 West Bengal, India. Tel.: +91353 2699127; fax: +91353 2699001; e-mail: firstname.lastname@example.org
Acinetobacter junii strain BB1A, a novel metal-tolerant bacterium, produced biofilm in the presence of added ions such as Ni2+, AsO2−, Cd2+ and Hg2+ on surfaces such as glass and polystyrene. Generation of a metal-sensitive and adhesion-deficient mutant by transposition of Tn5-mob in the A. junii genome has putatively confirmed the association of metal tolerance with the production of biofilm. The requirement of a critical cell density for biofilm formation and presence of acyl-homoserine lactone-like autoinducer molecules in the cell-free supernatant indicated the phenomenon of quorum sensing. Addition of a natural quorum-sensing inhibitor (garlic extract) or synthetic quorum-sensing inhibitor (4-nitro-pyridine oxide) significantly inhibited cell growth and biofilm formation in the presence of metal/metalloid ions.
Bacteria have a variety of heavy metal resistance mechanisms, including the formation and sequestration of heavy metals in complexes, reduction of a metal to a less toxic species and direct efflux of metal out of the cell (Nies & Silver, 1989, 1995; Silver, 1996; Silver & Phung, 2005). However, many bacteria in the environment exist in an aggregated state or as surface-attached communities termed biofilms; the ability of biofilms to withstand heavy metals and the mechanism by which this is achieved are little known. Biofilms are defined as assemblages of microorganisms attached to an abiotic or biotic surface through their associated extracellular products. Biofilms play a role in the biochemical cycling of minerals in the environment (Brown et al., 1999) and can survive in the presence of high concentrations of antimicrobials, including metals. It has been shown that biofilms of Pseudomonas aeruginosa were anywhere from two to 600 times more resistant to heavy metal stress than free swimming cells (Teitzel & Parsek, 2003). Several factors such as metabolic heterogeneity arising as a result of micronutrient and oxygen restriction or changes in physiology due to quorum-sensing-regulated gene expression may play a role in the metal tolerance of bacterial biofilms (Swift et al., 1996; McGowan et al., 1997; Pierson et al., 1999; Withers et al., 2001; Loh et al., 2002; Taga & Bassler, 2003; von Bodman et al., 2003). In gram-negative bacteria, quorum sensing is often achieved with the aid of acyl-homoserine lactone (AHL)-like signal molecules produced by the LuxI family of AHL synthases (Engebrecht et al., 1983; Cao & Meighen, 1989). These signal molecules differ in length, degree of substitution and saturation in their side chains (Parsek et al., 1999).
The presence of detectable AHL-like signal molecules was reported in one Acinetobacter strain (Gonzalez et al., 2001). It was also reported that strains of Acinetobacter form biofilms (Vidal et al., 1996; Mogilnaya et al., 2005; Kim & Wei, 2007; Shakeri et al., 2007). Acinetobacter junii BB1A, isolated from the waters of the River Torsa, India, was reported to demonstrate a novel induction phenomenon of nickel resistance and presumably to have a nickel resistance genetic system different from that previously characterized in other bacteria (Bhadra et al., 2006). The present study identifies that A. junii BB1A produced biofilm to counteract the presence of unfavourable concentrations of metal/metalloid ions and that quorum sensing plays a large role in the metal tolerance of strain BB1A.
Materials and methods
Reagents, chemicals and biosensor strains
Deionized double-distilled water and analytical-grade metal salts (NiCl2·6H2O, CdCl2·2H2O, HgCl2·2H2O and NaAsO2) were used to prepare 1 or 0.1 M stock solutions, which were filter-sterilized using filter paper (pore size 0.2 μm, marketed by Sartorius Ltd, Bangalore, India) before use. The chemical ingredients of Luria–Bertani medium (LB) and LB agar medium used in the experiments were purchased from HiMEDIA Chemicals, India. For growth experiments, the required quantity of metal salt solution was added singly to sterile medium prior to inoculation.
Preparation and use of 4-nitro-pyridine-N-oxide (4-NPO)
A 100 mM stock solution of the synthetic quorum-sensing inhibitor (QSI), 4-NPO (Aldrich) in sterile distilled water was prepared, filter sterilized and stored at 4 °C in the dark. From this stock, a calculated volume was added to the growth medium so that the required final concentration was attained.
Preparation and use of garlic extract
Garlic extract was prepared according to the method described by Rasmussen et al. (2005). The filter-sterilized aqueous phase was used as stock solution (100%) for QSI studies. A 1% (v/v) garlic extract in LB medium was used in general to study inhibition of biofilm.
AHL biosensor strains
All the AHL biosensor strains (kind gift from Professor Stephen K. Farrand, University of Illinois) were grown in AB minimal medium as described by Chilton et al. (1974) at a temperature not higher than 28 °C.
Determination of the maximum tolerable concentration (MTC) of Ni2+/AsO2–/Cd2+/Hg2+demonstrated by A. junii BB1A in liquid medium
For determination of MTC of the strain towards different metals, cells were inoculated from ‘preinduced’ culture (Bhadra et al., 2006) in a series of culture tubes containing LB supplemented with different concentrations of metal salt. MTCs of the strain were determined after incubation at 32 °C for 48 h.
Determination of critical cell number for initiation of biofilm in the presence of metal
To determine the critical cell density for the initiation of biofilm formation, aliquots of 0.2 mL of cultures grown overnight were separately inoculated into 10 mL LB medium containing metal/metalloid ions and allowed to grow at 32 °C in static condition. At 1h intervals, 100-μL volumes of the growing cultures were withdrawn, diluted and spread on LB agar plates and incubated at 32 °C for the appearance of visible colonies. Side-by-side, similar enumeration from culture grown in a medium without added metal was performed to compare and identify the plates where CFUs of metal-supplemented log-phase cultures have started departing from the usual exponential trend (as shown in culture without added metal). The identification of apparent stagnancy in cell count observed after a certain period of growth in metal-supplemented medium was corroborated by results obtained from turbidometric monitoring of growth.
Determination of the threshold concentration of metal ions required for biofilm formation
For determination of the threshold concentration of metal ions required for creating visible biofilm, overnight grown cells (2%, v/v) were inoculated in LB medium containing various concentrations of metal/metalloid ion(s) and allowed to grow for 24–48 h at 32 °C in static conditions. The concentration of added metal ions below which growth occurred without any development of biofilm is denoted as the ‘threshold concentration’. The highest concentration of added metal ions that demonstrated formation of biofilm, above which no growth occurred, is denoted as the ‘peak concentration’. Overnight grown cells (inoculum density identical to the test set) inoculated in LB without any addition of metal ions served as a positive control of growth as well as a negative control of biofilm formation.
Detection and quantification of biofilm formation
Biofilm formation was detected via the tissue culture plate (TCP) method and tube method (TM) as described earlier (Christensen et al., 1985) but with modifications, using LB medium in place of trypticase soy broth (TSB) and extending the duration of incubation to 48 h. ODs of stained adherent cells in different wells were determined with a micro-enzyme-linked immunosorbent assay (ELISA) autoreader (MERCK, Junior) at a wavelength of 600 nm. In the TM, medium with or without metal ions was inoculated with overnight grown culture and incubated for 48 h at 32 °C. A tube containing only sterile broth was used as control. The tubes were decanted, washed with phosphate-buffered saline (PBS; pH 7.3) and dried. They were stained with crystal violet (0.1%). Excess stain was removed and tubes were washed with deionized water, dried and observed for biofilm formation. Biofilm formation was considered positive when a visible film lined the wall and bottom of the tube. Both experiments were performed in triplicate and repeated at least three times; the data were then averaged and SD was calculated. Student's t-test was also done for statistical validation. To compensate for background absorbance, OD readings from sterile medium, fixative and dye were averaged and reduced from all test values.
Quantification of biofilm formation was performed as described by Sung et al. (2006). Cells were grown in polypropylene vials containing LB medium with or without metal/metalloid ions. After 48 h of growth, the growth medium containing the planktonic cells was decanted. The surface-attached cells were collected after adding 2 mL of fresh LB followed by vigorous shaking and repeated aspiration up and down using a micropipette. The biomasses of planktonic and surface-attached cells were estimated by measuring the OD540 nm. Three independent assays were performed and mean ODs were calculated. The adherence percentage was calculated using the following formula:
Cross-feeding assay for the presence of AHL molecules in cell-free supernatant
Cell-free filtrate (150 μL) from stationary phase cultures of strain BB1A grown in defined basal salt medium with or without added metal ions was coinoculated with 150 μL of stationary-phase cultures of Agrobacterium tumefaciens NTL4/pZLR4. The resulting mixtures were grown on plates containing AB minimal medium supplemented with 0.5% glucose. Then, 40 μL of 2% X-gal solution was spread on each plate before use. The presence of autoinducer or cross-feeding of the biosensor were determined by observing a change in colour to blue over the course of 2 days. Only sterile medium in place of cell-free filtrate of strain BB1A was used to rule out any nonspecific autoinduction. As positive control, spent culture supernatant of A. tumefaciens NTL4 (pTiC58ΔaccR) was used to cross-feed A. tumefaciens NTL4 (pZLR4).
The suicide vector pSUP5011 (Simon, 1984) used for transposon mutagenesis of A. junii BB1A was delivered by calcium chloride-mediated artificial transformation. Overnight grown BB1A cells were added to fresh LB medium containing 1.5 mM NiCl2 and allowed to grow in a shaker until an OD540 nm of 0.04. The cells were then harvested at 6400 g for 15 min at 4 °C, washed with sterile ice-cold water and finally suspended in 400 μL sterile ice-cold 0.1 M CaCl2. Transformation was carried out following standard protocols (Sambrook & Russell, 2001). The transformants were selected on LB plates containing 50 μg mL−1 kanamycin after 24 h of incubation at 30 °C. From the kanamycin-resistant (Kanr) colonies the nickel-sensitive mutant was selected by replica plating on LB plates containing different concentrations of NiCl2. To confirm the presence of Tn5-mob, genomic DNA was prepared (Bhadra et al., 2006) and PCR amplification was performed using primers F-5′-CCGACTGGGCTAAATCTGTG-3′ and R-5′-CTCGTCCTGCAGTTCATTCA-3′) (GenBank accession no. U00004 L19385).
For Southern hybridization the c. 4.7-kb BglII fragment of Tn5-mob from pSUP5011 was labelled and used as a probe against the Tn5-mob insertion mutants of BB1A. A digoxigenin-dUTP (DIG) high prime DNA labelling and detection kit (Roche Diognostics, Germany) was used as described by the manufacturer.
Growth characteristics of A. junii BB1A cells in liquid medium in the presence of metal/metalloid ions
When BB1A cells were grown in LB medium in the presence of metal ions such as Ni2+/AsO2−/Cd2+/Hg2+, they exhibited growth typical of biofilm structures. Such biofilm formation did not take place under similar growth conditions in cultures without added metal/metalloid ions. The ‘threshold’ concentration of supplemented metal ions that caused biofilm formation was 1.5 mM Ni2+ or 500 μM AsO2− or 150 μM Cd2+ or 10 μM Hg2+. The ‘peak’ concentration of supplemented metal/metalloid ions in LB medium at which the bacterium sustained and formed a biofilm was 4.5 mM Ni2+or 5 mM AsO2− or 450 μM Cd2+ or 12 μM Hg2+. In growth curves of strain BB1A in LB supplemented with or without metal/metalloid ion(s), an increase in the lag phase was noted in all the cultures where the medium was supplemented with metal/metalloid ions. The extent of the increase in the lag phase showed a relationship with the relative biological toxicity rendered by these metal ions. It was also observed that the length of the lag phase in nickel-supplemented medium increased with the increasing concentration of the metal.
Formation of biofilms of A. junii BB1A cells in liquid medium in the presence of added metal/metalloid ions
BB1A cells demonstrated surface-attached growth in polypropylene tubes or tissue culture plates with the TM or TCP method. In the presence of metal ions such as Ni2+/AsO2−/Cd2+ Hg2+, surface-attached growth was almost double (OD600 nm=0.12) that in absence of metal ions (OD600 nm=0.06). Quantification of biofilm production by measuring percentage adherence revealed that about 75% of total growth was surface attached when the cells were subjected to metal/metalloid challenge in comparison with about 53% surface attachment under normal growth conditions.
Presence of quorum sensing in A. junii BB1A
Viable cell counts of suspension cultures (CFU) with time revealed that when cells were grown in the presence of 1.5 mM nickel chloride or 500 μM sodium arsenite, quantification of viable cells became very inconsistent just after reaching a cell density of 109 CFU mL−1. This variability was concomitant with the initiation of biofilm formation. A similar observation was also noted during growth in the presence of mercury and cadmium. As in all the cases where the cells were grown in the presence of metal/metalloid ions, surface-attached growth was conspicuous after a particular cell number (in terms of viable cell number in suspension) was reached, i.e. quorum sensing. To confirm the phenomenon, a culture supernatant of strain BB1A was coinoculated with the biosensor A. tumefaciens NTL4 (pZLR4), which induced β-galactosidase activity of the 3-oxo and 3-hydroxy AHL derivatives. This result strongly suggested that strain BB1A produces a signalling molecule functionally similar to an AHL molecule having autoinducer activity.
Characterization of Tn5 mutant of A. junii BB1A
The Kanr Tn5 mutants were checked for the presence of vector and were found to be devoid of pSUP 5011. The presence of Tn5 in the genomic DNA of mutants was confirmed by generation of a 1.62-kb PCR product using primers as mentioned in the ‘Materials and methods’. No such product was found when genomic DNA of wild-type BB1A was used. Southern hybridization of the EcoRI-digested genomic DNA of Tn5-mob insertion mutants indicated the presence of a single copy of the transposon in the genome of BB1A (Fig. 1). No hybridization was detected when the wild-type DNA was probed with the c. 4.7-kb BglII fragment of Tn5-mob from pSUP5011 (data not shown). The mutants were grouped in two classes according to the size of the EcoRI fragment hybridized with the transposon: a 15.5-kb EcoRI fragment of one mutant SS01, and a 12-kb fragment of two mutants SS02 and SS03 contained Tn5-mob in the genome. The mutant SS01 was sensitive to Ni2+, AsO2−, Cd2+ or Hg2+ above 2.5, 0.5 mM, 50 and 8 μM, respectively, as compared with the MTC shown by wild-type BB1A (Table 1). The mutant is also capable of tolerating a higher concentration of QSI (4-NPO) compared with the wild-type BB1A (Table 1). Comparison of the growth curves of the wild-type with the metal-sensitive Tn5 mutant in LB showed greater growth yield of the mutant under identical conditions. The maximum OD540 nm attained by the mutant cells was 0.46 with to 0.30 obtained by the wild-type BB1A (Fig. 2a). Surface-attached growth of the mutant was found less than that of the wild-type in TM tests. In quantitative terms, the surface adherence of the mutant was significantly reduced in comparison with the wild-type (Table 1).
Table 1. Comparison of phenotypes in wild-type BB1A and its transposon mutant SS01 (BB1A:Tn5)
MTC, the highest concentration of metal ion/QSI that does not inhibit the formation of single colonies after 24–48 h of incubation at 32°C.
53.0 ± 2.02
25.0 ± 2.02
Effect of QSI on metal tolerance of strain BB1A
With natural or synthetic QSIs, garlic extract or 4-NPO, when added to the growth medium (LB with no supplementation of metal/metalloid ions), the growth or biofilm formation of strain BB1A cells was not affected. Cells of strain BB1A, which was otherwise seen to grow and form biofilm in the same medium supplemented with metal/metalloid ions, were significantly inhibited in growth with no visible formation of biofilm in the presence of garlic extract or 4-NPO: 1% (v/v) garlic extract or 100–150 μM 4-NPO was sufficient to prevent biofilm formation by strain BB1A in the presence of 2 mM nickel chloride or 250 μM cadmium chloride or 10 μM mercury chloride. When the growth curves(s) of BB1A cells in the presence of 1.5 mM nickel or 50 μM 4-NPO was compared, an increment in the lag phase was noticed in comparison with growth without any supplementation. However, when both 1.5 mM nickel and 50 μM 4-NPO were present, growth was significantly inhibited with a more pronounced lag phase (Fig. 2b).
Metal tolerance in A. junii BB1A is an inducible phenomenon. Southern hybridizations of strain BB1A genomic DNA with digoxigenin-dUTP-labelled DNA probes specific for the cnr, ncc and nre genes resulted no detectable signal, but the nir-specific probe yielded a weak hybridization signal with the restricted genome of strain BB1A (Bhadra et al. 2006). In the present study, the tolerance of this strain towards cadmium, mercury and arsenite ions was investigated. An observation common in all experimental growth flasks in the presence of metal/metalloid ion was the marked clumpy suspension or formation of biofilm in later phases of growth in shaking or still culture in contrast to evenly turbid growth in the absence of the same metal/metalloid ions irrespective of shaking or still conditions. The matrix-like structure and surface-attached growth led us to presume that strain BB1A produced biofilm to counteract the presence of unfavourable concentrations of metal/metalloid ions. Transpositional insertion mutagenesis has been conducted in strain BB1A, which identified a single insertional event in the genome responsible for disruption of metal resistance, namely biofilm formation (Fig. 1). The Tn5 insertion mutant, SS01, was found to be sensitive to the metals/metalloid tested with the loss of the ability to produce biofilm. Acinetobacter junii BB1A required a critical cell density for biofilm formation and also produced some form of quorum sensing molecule that was detected by biosensor strain of A. tumefaciens. To test whether quorum sensing has a role in the formation of biofilm, a natural or synthetic QSI was added to the metal-supplemented medium. Growth of strain BB1A leading to the formation of biofilm in metal-supplemented medium was significantly inhibited in the presence of natural or synthetic QSI (Fig. 2b). It was also observed that 4-NPO alone, at tolerable concentrations, does not prevent overall growth of wild-type cells in the absence of metal, but did extend the lag phase of growth. The metal-sensitive Tn5 mutant, lacking the inability to form biofilm, could withstand higher concentrations of QSI (4-NPO) compared with the wild-type and its growth in LB (in the absence of a metal salt) demonstrated higher planktonic cell growth, surpassing the maximum yield shown by the wild-type cells under similar conditions (Fig. 2a). Therefore, the apparent sterility attained by the mutant towards QSIs associated with the loss of biofilm-producing phenotype and attainment of greater sensitivity towards metal and metalloid ions revealed the role of quorum sensing in metal tolerance of strain BB1A through production of biofilm.
This work was supported by funding from the Department of Biotechnology, Ministry of Human Resource Development, Government of India. We acknowledge the input of Bhaskar Bhadra, Center for Cellular and Molecular Biology (CCMB), Hyderabad, India. We are grateful to Papai Singha for constant assistance.