Quorum  sensing N-acylhomoserine lactone signals affect nitrogen fixation in the cyanobacterium Anabaena sp. PCC7120

Authors

  • Manuel Romero,

    1. Departamento de Microbiología y Parasitología, Facultad de Biología-CIBUS, Universidad de Santiago de Compostela, Santiago de Compostela, Spain
    Search for more papers by this author
  • Alicia M. Muro-Pastor,

    1. Instituto de Bioquímica Vegetal y Fotosíntesis, Consejo Superior de Investigaciones Científicas and Universidad de Sevilla, Seville, Spain
    Search for more papers by this author
  • Ana Otero

    1. Departamento de Microbiología y Parasitología, Facultad de Biología-CIBUS, Universidad de Santiago de Compostela, Santiago de Compostela, Spain
    Search for more papers by this author

  • Editor: Karl Forchhammer

Correspondence: Ana Otero, Departamento de Microbiología y Parasitología, Facultad de Biología-CIBUS, Universidad de Santiago de Compostela, 15782 Santiago de Compostela, Spain. Tel.: +34 981 563 100, ext. 16913; fax: +34 981 528 006; e-mail: anamaria.otero@usc.es

Abstract

Bacteria secrete small signal molecules into the environment that induce self and neighbour gene expression. This phenomenon, termed quorum sensing, allows cooperative behaviours that increase the fitness of the group. The best-studied signal molecules are the N-acylhomoserine lactones (AHLs), secreted by a growing number of bacterial species including important pathogen species such as Pseudomonas aeruginosa. These molecules have recently been proposed to have properties other than those of signalling, functioning as iron quelants or antibiotics. As the presence of an acylase capable of inactivating long-chain AHLs in Anabaena sp. PCC7120 could constitute a defence mechanism against these molecules, in this work we analyse the effects of different AHLs varying in length and substitutions on the growth and nitrogen metabolism of the cyanobacterium Anabaena sp. PCC7120. All the AHLs tested strongly inhibited nitrogen fixation. The inhibition seems to take place at post-transcriptional level, as no effect on heterocyst differentiation or on the expression of nitrogenase was observed. Moreover, N-(3-oxodecanoyl)-l-homoserine lactone (OC10-HSL) showed a specific cytotoxic effect on this cyanobacterium in the presence of a combined nitrogen source, but the mechanism involved seems to be different from that described so far for tetramic acid derivatives of oxo-substituted AHLs. These results suggest a variety of new unexpected activities for AHLs, at least on cyanobacterial populations.

Introduction

The term ‘quorum sensing’ (QS) (Fuqua et al., 1994) describes a phenomenon of bacterial communication that confers on these organisms the ability to perceive and respond to the community density through coordinated regulation of gene expression, thus being able to adopt an advantageous social behaviour. Bacteria communicate their presence to others by secreting small chemical signals called autoinducers, allowing the individuals to distinguish between high and low population densities.

By means of QS, bacterial populations can coordinate important biological functions including motility, swarming, aggregation, plasmid conjugal transfer, luminescence, antibiotic biosynthesis, virulence, symbiosis and biofilm maintenance and differentiation (Williams et al., 2007). Several chemically distinct families of QS signal molecules have now been described, but the most studied QS signalling system involves N-acylhomoserine lactones (AHLs) employed by diverse Gram-negative bacteria. AHLs differ in the acyl side chain, which is usually 4–18 carbons in length, with or without saturation or C3 hydroxy- or oxo-substitutions (Whitehead et al., 2001). AHLs have been initially described as being exclusively produced by a relatively small number of Alpha-, Beta- and Gammaproteobacteria (Williams et al., 2007), but recently the production of these signals has also been reported for the colonial cyanobacterium Gloeothece (Sharif et al., 2008) and different marine Bacteroidetes (Huang et al., 2008; Romero et al., 2010), which might indicate a significant role for QS systems in natural populations/environment.

Besides acting as quorum signals, some AHLs have been proposed to have other possible biological functions, for example acting as iron quelants and antibiotics (Kaufmann et al., 2005; Schertzer et al., 2009). A naturally occurring degradation product of N-(3-oxododecanoyl)-l-homoserine lactone (OC12-HSL), one of the AHL signals produced by Pseudomonas aeruginosa, is the tetramic acid 3-(1-hydroxydecylidene)-5-(2-hydroxyethyl)pyrrolidine-2,4-dione, which exhibits iron-binding ability. This AHL derivative is able to bind iron in a 3 : 1 complex with an affinity comparable to that exhibited by standard quelators and siderophores (Schertzer et al., 2009). In addition, antibiotic properties of the tetramic acid derivative of OC12-HSL have been described, through the disruption of membrane potential and proton gradient of bacteria, thus eliminating the proton-motive force and leading to bacterial death (Lowery et al., 2009).

The existence of QS blockage systems adopted by competitors to destroy or inhibit the functions of AHLs also indicates the ecological importance of these molecules. The different mechanisms of interference with QS communication systems have been generally termed ‘quorum quenching’ (QQ) (Dong et al., 2001). An example of QQ is the enzymatic inactivation of AHLs, with two groups of AHL-degrading enzymes identified so far. The lactonases hydrolyse the homoserine lactone (HSL) ring of the AHL molecule to produce acyl homoserines (Dong et al., 2007), whereas the acylases cleave the AHL amide bond, generating the corresponding fatty acid and HSL ring (Dong et al., 2007). Enzymatic QQ activity has been described in Gram-positive and -negative bacteria and more recently in the cyanobacterium Anabaena sp. PCC7120 (Romero et al., 2008).

Anabaena sp. PCC7120 is a filamentous cyanobacterium simultaneously able to perform photosynthesis and dinitrogen fixation under aerobic conditions. In the presence of a source of combined nitrogen, filaments grow as undifferentiated chains of vegetative cells. In contrast, when Anabaena sp. PCC7120 is deprived of combined nitrogen, approximately 10% of the cells differentiate into morphologically distinct heterocysts that supply the rest of the filament with fixed nitrogen and in return receive carbohydrate from vegetative cells (Wolk et al., 1994). In the absence of combined nitrogen the heterocysts are spaced along the filament in a semi-regular pattern that is controlled by a regulatory loop established between two master regulators, NtcA and HetR (Muro-Pastor et al., 2002).

Because AHLs have been described in natural environments where cyanobacteria are prevalent, such as microbial mats and algal blooms (McLean et al., 1997; Bachofen & Schenk, 1998), the acylase-type QQ activity found in Anabaena sp. PCC7120 (Romero et al., 2008) could serve either to mitigate possible negative effects of AHLs themselves and/or their tetramic acid derivatives (Kaufmann et al., 2005; Schertzer et al., 2009) or to confer a competitive advantage against AHL-producing competitors through the disruption of their communication system.

In this work, we study the effects of exogenous AHL addition to cultures of the filamentous heterocyst-forming cyanobacterium Anabaena sp. PCC7120 to assess the possible physiological role of the AHL-acylase present in this cyanobacterium.

Materials and methods

Growth conditions

Stock cultures of Anabaena sp. PCC7120 were maintained photoautotrophically at 30 °C with a continuous irradiance of 75 μE m−2 s−1. Cultures were aerated by connecting each culture unit to an aeration system with a continuous filtered (0.45 μm) air flow or carbon dioxide (CO2)-enriched air (1% v/v).

Diazotrophic cultures were carried out in BG110C medium [BG11 medium (Rippka et al., 1979) without NaNO3 and supplemented with 0.84 g L−1 of NaHCO3 (C)].

Nondiazotrophic cultures of Anabaena sp. PCC7120 were established in BG110C supplemented with either 17 mM NaNO3 (BG11C) or 6 mM NH4Cl and 12 mM of N-Tris(hydroxymethyl)methyl-2-aminoethanesulphonic acid-NaOH buffer pH 7.5 (BG110C+NH4+). To study the effect of AHL addition on the process of heterocyst differentiation, the biomass of nondiazotrophic cultures was collected by filtration (0.45 μm), washed and resuspended in fresh BG110C (nitrogen step-down procedure).

Solid media plates were prepared mixing equal volumes of double-concentrated sterilized BG110 or BG110+NH4+ and agar 10 g L−1. Plates inoculated with Anabaena sp. PCC7120 were incubated at 30 °C with light.

Addition of synthetic AHLs to cultures

AHLs were first assayed in solid media to check a possible antibiotic effect (Lowery et al., 2009). Cells from a liquid exponentially growing culture of Anabaena sp. PCC7120 in BG110C+NH4+ were harvested by filtration, washed and resuspended in BG110C at a concentration of 5 μg chlorophyll a (Chl a) mL−1 and 100 μL of the suspension was spread on top of BG110+NH4+ or BG110 plates. Small holes were made in the centre of each plate and filled with 100 μL of 100 μM AHL or acetonitrile (as control). Growth was checked after 7 days of incubation at 30 °C with light.

Synthetic AHLs were also added to liquid cultures of Anabaena sp. PCC7120 both under nondiazotrophic conditions (BG110C+NH4+ medium) and during nitrogen step-down. Anabaena sp. PCC7120 was grown to exponential phase in BG110C+NH4+ [cultures with about 5 μg Chl a mL−1; Chl a levels were determined in methanolic extracts (Mackinney, 1941)]. The cells were filtered, washed with BG110C, inoculated in fresh BG110C+NH4++AHL (100 μM) or BG110C+AHL (100 μM) and bubbled with air or CO2-enriched air with a final Chl a concentration of 4 μg mL−1. The AHLs used were: N-butyryl-homoserine lactone (C4-HSL), N-(3-oxobutyryl)-l-homoserine (OC4-HSL), N-(3-hydroxybutyryl)-l-homoserine (OHC4-HSL), N-decanoyl-l-homoserine (C10-HSL) N-(3-oxodecanoyl)-l-homoserine lactone (OC10-HSL), N-(3-hydroxydecanoyl)-l-homoserine (OHC10-HSL), N-dodecanoyl-l-homoserine (C12-HSL) OC12-HSL and N-(3-hydroxydodecanoyl)-l-homoserine (OHC12-HSL) (unsubstituted AHLs were purchased from Sigma-Aldrich, all other AHLs were kindly provided by Prof. Miguel Cámara from the University of Nottingham). AHL stock solutions of 1 mg mL−1 were prepared in acetonitrile. Parallel control assays were carried out using equal amounts of acetonitrile (AHL solvent). In nitrogen step-down cultures, the differentiation of heterocysts was monitored by Alcian blue staining of polysaccharides in the heterocyst envelope (Olmedo-Verd et al., 2006).

To further evaluate the lethal effect observed for OC10-HSL in ammonium-grown nondiazotrophic cultures of Anabaena sp. PCC7120 (BG110C+NH4+), different concentrations of this signal (0.01, 0.1, 1, 10, 25, 50, 75 and 100 μM) as well as OC12-tetramic acid (100 μM) were also assayed. The effect of OC10-HSL (100 μM) was also tested in cultures with nitrate as combined nitrogen source (BG11C). OD600 nm of the cultures was measured at different time points after treatment (Kuznetsova et al., 2008).

Nitrogenase activity measurement

Biomass (200 mL, 2–3 μg mL−1 Chl a) from BG110C+NH4+ aerated cultures of Anabaena sp. PCC7120 was harvested, washed and resuspended in fresh BG110C at a Chl a concentration of 2 μg mL−1 to induce the differentiation of heterocysts. Cultures of 20 mL were established in flasks supplemented with AHLs (100 μM) or acetonitrile as control. After 20 h of incubation at 30 °C, 120 r.p.m. and light, the nitrogenase activity was measured as follows: cells were concentrated to 4 mL by removing part of the supernatant after centrifugation, and they were then divided in two 17-mL flasks sealed with silicon caps (2 mL each, 10 μg Chl a). For each AHL, one flask was incubated under standard aerobic conditions. Another flask was incubated with an anaerobic atmosphere by injecting argon for 3 min and adding 10 μM 3-(3,4dichlorophenyl)-1,1-dimethylurea (DCMU) to inhibit photosynthesis and therefore oxygen (O2) production (Rippka & Stanier, 1978) to avoid a possible inhibition of nitrogenase activity derived from the formation of abnormal heterocyst cell walls during maturation or the damage from other mechanisms responsible for maintaining low O2 concentration within the heterocysts.

After 1-h incubation at 30 °C, 2 mL of acetylene was injected. Samples of 1 mL from the air in the sealed flask were taken at different times during 20 h starting 15 min after acetylene injection to determine the concentration of the ethylene produced using a GC-MS (HP 5890 series II) equipped with injector, column (Porapak Q) and flame ionization detector (kept at 100, 80 and 150 °C, respectively). The detected signals were processed with the computing integrator PYE Unicam DP88. The equipment was calibrated with known concentrations of ethylene.

To determine the nitrogenase activity of the cultures per unit Chl a, the following formula was used: nitrogenase activity=nmol ethylene in sample × 14 mL/2 ×μg Chl a mL−1; where 14 was the atmosphere volume in 17-mL flasks and 2 the volume of culture in the flask.

C10-HSL was also added to BG110C cultures of Anabaena sp. PCC7120 with mature heterocysts (24 h after nitrogen step-down) and the nitrogenase activity then measured as described before.

RNA isolation and analysis

To assess a possible effect of AHLs on the expression of genes involved in nitrogen fixation, Northern hybridization was carried out with probes for the nifH and fdxH genes. Samples of 50 mL were taken at 0, 3, 6, 20 and 24 h after nitrogen step-down. Cells were filtered, washed and resuspended in 1 mL of Tris 50 mM/EDTA 100 mM, centrifuged and the pellet was frozen in liquid nitrogen before RNA extraction. RNA from whole filaments was extracted in the presence of ribonucleoside–vanadyl complex as described previously (Muro-Pastor et al., 2002).

For Northern analysis, 30 μg of RNA was loaded per lane and electrophoresed in 1% agarose denaturing formaldehyde gels. Transfer and fixation to Hybond-N+ membranes (Amersham Biosciences) were carried out using 0.1 M NaOH. Hybridization was performed at 65 °C according to the recommendations of the manufacturer of the membranes. The nifH and fdxH probes were fragments of these genes amplified by PCR. The nifH probe was amplified using plasmid pCSAV60 (containing the nifH gene cloned in pGEM-T vector) as a template and oligonucleotides NH-1 (corresponding to positions −334 to −314 with respect to the translation start of nifH) and NH-4 (complementary to nucleotides +884 to +863 with respect to the translation start of nifH) (Valladares et al., 2007). The fdxH probe was amplified using plasmid pCSAV164 (containing the fdxH gene cloned in pGEM-T vector) as a template and oligonucleotides FH-1 (corresponding to nucleotides+3 to +20 with respect to the translation start of fdxH) and FH-2 (complementary to nucleotides +297 to +269 with respect to the translation start of fdxH) (Valladares et al., 2007). rnpB, encoding the RNA subunit of RNase P (Vioque, 1997), was used as a loading and transfer control. All probes were 32P-labeled with a Ready-to-Go DNA labeling kit (Amersham Biosciences) using [α-32P]dCTP. Images of radioactive filters and gels were obtained and quantified with a Cyclone storage phosphor system and optiquant image analysis software (Packard).

Results and discussion

Effect of synthetic AHLs addition

AHLs were added to Anabaena sp. PCC7120 cultures to evaluate possible effects on growth and nitrogen metabolism of the cyanobacterial filaments both in solid and liquid media. We selected saturated and substituted representatives of short- (C4, OC4 and OHC4-HSL), middle- (C10, OC10 and OHC10-HSL) and long-chain AHLs (C12, OC12 and OHC12-HSL). A first experiment was carried out in solid media, as described in Materials and methods. Growth inhibition halos surrounding the wells were observed after 7 days for OC10-HSL and OC12-HSL in cultures subjected to nitrogen step-down (transferred to nitrogen-free BG110 medium) (Fig. 1). OC10-HSL also inhibited growth in the presence of combined nitrogen (BG110+NH4+, data not shown). These observations suggested that at least these two AHLs could have an effect on heterocyst differentiation or maturation, which was further investigated.

Figure 1.

Anabaena sp. PCC7120 growth inhibition halos surrounding wells filled with 100 μL of OC10-HSL and OC12-HSL (100 μM) in comparison with normal growth (control with acetonitrile) in BG110 plates. Plates were incubated for 7 days with continuous light at 30°C.

AHLs were also added to liquid cultures under nondiazotrophic conditions (BG110C+NH4+) and to cultures subjected to nitrogen step-down to study the effect on growth and heterocyst differentiation. None of the tested AHLs showed cytotoxic effects in liquid cultures subjected to step-down after 20 h of exposure. Moreover, no effect on heterocyst differentiation and distribution pattern was found in step-down cultures for any of the tested AHLs after Alcian blue staining and microscope observation (data not shown). The discrepancy between the inhibitory effects obtained for OC10 and OC12-HSL in solid plates (Fig. 1) and in liquid cultures could be derived from the longest period of incubation of solid plates or could also be due to the higher initial cell concentration in the liquid cultures compared with plates resulting in a higher AHL-acylase activity (Romero et al., 2008) that would diminish the effect of initial AHL concentration.

Possible effects of AHLs on heterocyst differentiation were also tested with Anabaena sp. PCC7120 strain CSEL4a (Olmedo-Verd et al., 2006). This strain expresses gfp gene under the control of ntcA promoter, the master regulator of nitrogen assimilation, which also controls the early phases of heterocyst differentiation (Herrero et al., 2004). Expression of gfp in this strain is induced in specific cells upon nitrogen step-down, indicating the induction of ntcA during heterocyst differentiation (Olmedo-Verd et al., 2006). To test for possible effects of AHLs, cells of strain CSEL4a grown in the presence of BG110C+NH4+ were transferred to BG110C in the presence of AHLs (100 μM). Induction of the expression of gfp from the ntcA promoter proceeded in the same way both in the presence and in the absence of AHLs, indicating that the AHLs were not affecting the process of heterocyst differentiation (data not shown).

In contrast, and consistent with the results obtained in solid plates, a strong cytotoxic effect was observed after only 5 h for OC10-HSL (100 μM) in BG110C+NH4+ in liquid media (Fig. 2a). The same effect could also be observed in cultures with nitrate as nitrogen source (BG11C) supplemented with OC10-HSL at the same concentration (data not shown). This effect could not be observed for any of the other AHLs tested. To determine the OC10-HSL minimal lethal concentration, the assay was repeated using: 0.01, 0.1, 1, 10, 25, 50, 75 and 100 μM of OC10-HSL in BG110C+NH4+ cultures. Concentrations >25 μM were lethal (Fig. 2a and b) and the filaments appeared completely lysed under the microscope after 5 h of culture. Cells incubated in the presence of 25 μM of OC10-HSL showed black dots, resembling cyanophycin granules, in the inner side of the cell walls (data not shown). No lethal effect on Anabaena sp. PCC7120 was observed in cultures supplemented with 100 μM OC12-HSL or OC12-tetramic acid (data not shown). The half maximal effective concentration (EC50) observed for other bacteria is between 8 and 55 μM for the OC12-HSL-derived tetramic acid and between 22.1 and 100 μM for OC12-HSL itself, depending on the bacterial strain (Kaufmann et al., 2005). These ranges match the lethal concentration observed for OC10-HSL in BG110C+NH4+ cultures of Anabaena sp. PCC7120, but it should be noted that this activity was described only for Gram-positive bacteria, as the outer Gram-negative membrane seems represent a permeability barrier for tetramic acids (Lowery et al., 2009). Nevertheless, the antibiotic effect observed for OC10-HSL under nondiazotrophic conditions seems to be highly specific and different from the antibiotic effect described so far for tetramic acids, as no cytotoxic effect of OC12-HSL or its tetramic acid derivative could be observed. It has been reported that a degradation product of oxo-substituted AHLs such as OC12-HSL is a tetramic acid with a high affinity for iron, comparable to standard quelants and siderophores (Kaufmann et al., 2005; Schertzer et al., 2009), therefore the cytotoxic effect of OC10-HSL could be related to iron quelant properties, but this could not explain the dramatic lethal effect observed, with total lysis of the filaments already after 5 h of the addition of OC10-HSL to nondiazotrophic cultures. Moreover, it is highly improbable that OC10-HSL is acting through the disruption of membrane potential, as already described for OC12-HSL or its tetramic acid derivative (Lowery et al., 2009), because no effect was recorded for these two compounds, which are expected to be more active than OC10-HSL in this respect (Schertzer et al., 2009). Therefore, the observation that OC10-HSL is lethal only in the presence of combined nitrogen in liquid media could be the result of a specific inhibitory effect of this molecule on the metabolism of combined nitrogen. Alternatively, OC10-HSL signal might lead to the activation of the wrong pathways. For instance, overactivation of arginine biosynthesis in the presence of combined nitrogen could lead to cyanophycin accumulation (dense, presumptive cyanophycin granules are observed in the damaged filaments), blocking the entire nitrogen metabolism and resulting in cell death.

Figure 2.

 (a) Antibiotic effect of different concentrations of OC10-HSL (25–100 μM) on Anabaena sp. PCC7120 cultures in BG110C+NH4+. The photo was taken 7 h after AHL addition. C, control culture containing acetonitrile. (b) Evolution of OD600 nm of Anabaena sp. PCC7120 cultures in BG110C+NH4+ with different concentrations of OC10-HSL (□, 25 μM; ▴, 50 μM; ○, 75 μM; and ◆, 100 μM) and acetonitrile (◊) as control. Time 0, addition of OC10-HSL.

Nitrogenase activity

Although no macroscopic effect of AHLs on survival and heterocyst differentiation was recorded in diazotrophic cultures in short-time experiments, the effect of the signals on the nitrogenase activity was evaluated in BG110C+NH4+ cultures transferred to BG110C for the induction of heterocyst formation and nitrogen fixation in the presence of the AHLs. Nitrogenase measurements were carried out 20 h after the nitrogen step-down treatment to allow formation of mature heterocysts. A strong inhibition of the nitrogenase activity was recorded for all AHLs tested (Fig. 3). The lower ethylene production in AHL-treated cultures was already evident 5 min after acetylene addition. The inhibition was specially marked in cultures treated with OC10 and OC12-HSL, in which none or residual nitrogenase activity could be detected (Fig. 3). This result is consistent with the inhibition of growth observed in the cyanobacterium, with these two AHLs in solid BG110 media (Fig. 1).

Figure 3.

Anabaena sp. PCC7120 nitrogenase activities under aerobic (black bars) and anaerobic (grey bars) conditions in BG110C supplemented with the AHLs: C4, OC4, OHC4, C10, OC10, OHC10, C12, OC12 and OHC12-HSL (100 μM). Control culture was set with acetonitrile (C). Nitrogenase activities are expressed as percentages of the value for control culture, which corresponds to 2.04 (aerobic) and 6.5 (anaerobic) nmol ethylene per μg Chl−1h−1.

To evaluate whether the inhibition of nitrogenase activity was due to defects in heterocyst wall formation or defects in any of the other mechanisms driving the creation of a microoxic environment inside the heterocysts, nitrogenase activity was also measured under anaerobic atmosphere (Fig. 3). Air inside the flasks was substituted by argon and DCMU was added to the cultures to inhibit PSII-dependent O2 production. As expected, slightly higher nitrogenase activity was observed in anaerobic conditions than in aerobic ones (Valladares et al., 2007), but the effect of AHL addition was still observed (Fig. 3). This indicates that the lower nitrogenase activity observed in the presence of AHLs was not due to alterations in the microoxic environment of the heterocysts and confirms that they have no effect on heterocyst differentiation as observed in AHL-supplemented cultures described before. As observed under aerobic conditions, the OC10 and OC12-HSL signals had the strongest inhibitory effect on nitrogenase activity (Fig. 3). Twenty hours after the addition of acetylene still no recovery of normal levels of nitrogenase activity of the cultures was observed either in aerobic or anaerobic conditions (data not shown).

To determine whether the inhibitory effect of the AHLs on nitrogen fixation took place only in developing heterocysts, nitrogenase activity was also measured in diazotrophic cultures in which C10-HSL (100 μM) was added 24 h after nitrogen step-down, when mature heterocysts are already present. The amount of ethylene produced in early samples was similar in cultures with or without C10-HSL but, interestingly, a progressively decreased ethylene production was observed in the C10-HSL-treated culture, resulting in a 30% decrease of nitrogenase activity (data not shown). The progressive increase of the inhibitory effect of AHLs in acclimated cultures could perhaps be caused by the entry of the AHLs in the new generations of heterocysts, as the impermeability of the wall of mature heterocysts could prevent the penetration of the AHLs. Nonetheless it cannot be excluded that although AHLs could enter through vegetative walls and spread along the filaments by the periplasmic space (Flores et al., 2006; Mariscal et al., 2007), entering in both mature and forming heterocysts, these molecules could only act at the molecular level in newly formed heterocysts. In that case the results observed would suggest a nonreversible inhibition of nitrogenase in very early stages at the level of either gene expression or its enzymatic activity.

Effect on the expression of nitrogen fixation-related genes

Because all tested AHLs showed inhibitory activity on nitrogen fixation mostly in newly formed heterocysts, to study possible effects at the level of expression of nitrogen metabolism genes, Northern blots were carried out to detect changes in expression of the dinitrogenase reductase subunit gene (nifH) and fdxH, encoding a heterocyst-specific ferrodoxin that is a likely electron donor to dinitrogenase reductase (Razquin et al., 1995).

No significant differences in the expression of either gene could be detected at 20 and 24 h after nitrogen step-down (no expression of nifH and fdxH was detected at 0, 3 or 6 h) in total RNA extracted from C10-HSL-treated cultures when compared with control samples (Fig. 4). This indicates that the process of heterocyst differentiation proceeds normally in the presence of AHLs and therefore AHL inhibition could be affecting either the expression of other genes related to nitrogen fixation or be acting on nitrogenase-related genes at a post-transcriptional level.

Figure 4.

 Effect of C10-HSL addition on heterocyst differentiation upon nitrogen step-down. RNA (30 μg) was isolated from samples taken at 0 (NH4+), 3, 6, 20 and 24 h in the presence or absence of C10-HSL. Cultures contained 100 μM C10-HSL or acetonitrile as control. Hybridizations were carried out with a probe for the nifH or fdxH gene or for the rnpB gene (Vioque, 1997), which was used as a loading and transfer control.

Finally, the strong inhibition of nitrogenase demonstrated for all the AHLs tested and the cytotoxic effect of OC10-HSL in the presence of combined nitrogen represent novel biological activities of these signal molecules. The observation that antibiotics cannot easily reach the lethal concentrations in natural environments has led to a questioning of whether these molecules could act, in subinhibitory concentrations, as signal molecules (Davies, 2006; Linares et al., 2006). Low concentrations of several antibiotics can alter expression patterns of bacteria without any effect on growth rate (Davies et al., 2006), which resembles the mode of action of QS signals. Thus one possibility is that the AHL signals have inhibitory effects when added at higher concentrations than those found in natural environments. In fact, the concentrations reported in the literature for AHLs in the culture media of the model microorganism Vibrio fischeri usually range between 0.4 and 400 nM (Kaplan & Greenberg, 1985; Schaefer et al., 2002; Burton et al., 2005), significantly lower than the concentrations exhibiting inhibitory activity against Anabaena sp. PCC7120.

In conclusion, AHLs strongly inhibit nitrogen fixation in Anabaena sp. PCC7120, although they do not affect the process of heterocyst differentiation because no changes were observed in the frequency, pattern of differentiation, permeability of the heterocyst cell wall or expression of regulatory genes whose products are involved in differentiation (ntcA). The strong inhibition of nitrogenase activity observed could be related to nitrogen fixation blockage at a post-transcriptional level, mainly on newly formed heterocysts. Moreover, a possible new activity of AHL signals was found for OC10-HSL in the presence of combined nitrogen, differing from those activities described for oxo-substituted and AHL tetramic acid derivatives. The presence of acylase activity against long-chain AHLs described in the biomass of Anabaena sp. PCC7120 (Romero et al., 2008) could be related to the negative effects of AHLs in this cyanobacterium. This AHL-degradation mechanism would protect the filaments, at normal environmental concentrations, from exogenous signals with potential cytotoxic and inhibitory activities on the cyanobacterium.

Acknowledgements

This work was financed by a grant from Consellería de Innovación e Industria, Xunta de Galicia PGIDIT06PXIB200045PR. M.R. was supported by an FPU fellowship from the Spanish Ministry of Education and Science and a predoctoral fellowship from Diputación de A Coruña. We would like to thank Prof. Kim D. Janda and Dr Gunnar F. Kaufmann for kindly providing us with OC12-tetramic acid. We also would like to thank Prof. Miguel Cámara for providing us with synthetic AHLs.

Ancillary