For determination of AHL-like molecules with short acyl chains, the biosensor C. violaceum CV026 was utilized. This strain is a mini Tn5 double mutant defective in the synthesis of violacein pigment. The production of this pigment in vitro is activated by AHLs with short acyl chains. These autoinducers in peanut-nodulating strains were detected by the method of McClean et al. [26].

A. tumefaciens NTL4 (pZLR4) was utilized to detect AHLs with long acyl chains. This strain carries the plasmid pZLR4, which contains a traG::lacZ fusion and traR. In the presence of AHLs with long acyl chains the TraR protein is activated, transcription of the traG::lacZ fusion is turned on, and LacZ (β-galactosidase) activity can be used as a reporter of traG transcription [30]. AHL-like molecules with long acyl chains in peanut-nodulating strains were detected by the method of Farrand et al. [31].

Extraction of AHLs was performed as described by Shaw et al. [32]. Peanut-nodulating bacteria were grown to the early stationary phase, and cells were removed from 25 mL growth medium by centrifugation at 12,000 g for 15 min at 4 °C. AHLs were then extracted from culture supernatants with three equal volumes of ethyl acetate, and the extracts were dried and resuspended in 500 μL ethyl actetate.

Aliquots (100 μL) of AHL extracts as above were added to 10 mL cultures of A. tumefaciens NTL4 (pZLR4) grown to OD₆₀₀ 0.5. The cultures were incubated for 6–8 h until reaching OD₆₀₀ 1.0. β-galactosidase activity was determined in Miller units as described by Miller [33]. For each strain extract, the values presented are means of four repeated experiments.

Aliquots (10 μL) of AHL extracts were analyzed by reverse-phase C18-thin layer chromatography (RP-C18 TLC) using methanol/water (60:40 v/v) as the mobile phase. The plates were air-dried, overlaid with AB medium, 0.7% agar containing X-Gal 40 μg mL⁻¹, and the biosensor A. tumefaciens NTL4 (pZLR4), and incubated overnight at 30 °C [32].

AHL extracts were resuspended in 100 μL methanol (100%) for identification and quantification in LC-MS/MS, using a Restek C₁₈ (Restek, USA) column (2.1 × 100 mm, 5 μm) with injection volume 10 μL. AHLs were separated at 28 °C using a gradient solvent system with increasing methanol concentration, constant glacial acetic acid concentration 0.2% (v/v) in water, and initial flow rate 0.2 mL min⁻¹. The gradient was increased linearly from 40% (v/v) methanol/ 60% (v/v) water–acetic acid to 80% (v/v) methanol/ 20% (v/v) water–acetic acid over 25 min.

Mass spectrometry was performed on a quadrupole tandem mass spectrometer (MS–MS, Quattro Ultima™ PT; Micromass, Manchester, UK) equipped with an electrospray ion source (ESI).

AHLs were identified by comparison of retention times and m/z transitions with those of the following standards: N-hexanoyl-dl-homoserine lactone (C₆), N-(3-oxodecanoyl)-l-homoserine lactone (3OC₁₀), N-(3-oxododecanoyl)-l-homoserine lactone (3OC₁₂), and N-(3-oxotetradecanoyl)-l-homoserine lactone (3OC₁₄).

Bacterial motility was determined in plates containing TY medium and reduced 1/10 TY medium with a final concentration of 0.3% agar. The 3OC₁₀, 3OC₁₂, and 3OC₁₄ AHLs were added at various concentrations. Strains were inoculated by puncture in the center of the plate and incubated for 8 days at 28 °C. Halo diameters as indicators of motility were measured in cm.

Biofilm formation capacity was determined macroscopically by the method of O'Toole and Kolter [34], with some modifications. Glass tubes were inoculated with 800 μL bacterial culture (OD₆₀₀ 0.5) and incubated with agitation for 72 h at 30 °C. Planktonic cells were removed, and each tube was washed three times with saline solution, emptied, stained with crystal violet 0.1% for 15 min, and rinsed three times with distilled water to remove excess crystal violet. Biofilms formed were quantified by adding 1 mL 95% ethanol to the stained tube. The absorbance of solubilized crystal violet was determined by spectrophotometry at 570 nm.

Each peanut-nodulating strain was grown for 5 days at 28 °C in 25 mL TY medium supplemented with the appropriate AHLs. The bacterial suspension (5 mL) was transferred to a glass tube (10 × 70 mm) and left to settle for 24 h at 4 °C. An aliquot (0.2 mL) of the upper portion of the suspension was carefully transferred to a microtiter plate, and OD₆₀₀ was measured (OD_final). A control tube was vortexed for 30 s and OD₆₀₀ was determined (OD_initial). The autoaggregation percentage was calculated as 100[1 – (OD_final/OD_initial)] [35].

Microorganisms with the ability to move have an ecological advantage in that they are able to find more favorable or less hazardous niches for colonizing and persisting in a given environment. In the case of rhizobia, motility has been shown not to be essential for nodulation, but allows the bacteria to find their specific host legume and establish symbiosis. Bacterial motility is an energetically costly process, and therefore must be finely regulated [40]. Several studies using S. marcescens, V. cholerae, E. coli, and Y. pseudotuberculosis have shown that QS systems are involved in the control of bacterial motility [41–44]. We evaluated for the first time the effect of QS signals on the swimming motility of peanut-nodulating bacterial strains. The optimal experimental condition for evaluating the motility of these strains was found to be limited nutrient availability (reduced 1/10 TY medium), whereby their motility was two-fold higher than in full growth medium (entire TY medium) (data not shown). We therefore added the AHLs to reduced 1/10 TY medium and measured the swimming halo of peanut-nodulating strains.

In general, the motility of 62B, P8A, and P5 was increased as a result of exposure to various types and concentrations of AHLs (Figure 3). In the case of 62B and P8A, a low autoinducer concentration (5 μM) resulted in a significant increase in motility, and this effect was maintained or increased in the presence of higher autoinducer concentrations (10 and 20 μM) (Figure 3(A,B,C)). P5 also responded positively to the presence of autoinducers, but to a lesser degree than 62B or P8A. We concluded that, regardless of the type and concentration of autoinducer, these rhizobial strains show increased motility in the presence of exogenous signaling molecules. Interestingly, USDA 4438, a reference strain that did not induce blue haloes in biosensor strains, displayed a more complex swimming behavior when various exogenous AHLs were added. Low concentrations of 3OC₁₀ AHL (5 or 10 μM) significantly decreased the motility of USDA 4438, whereas the maximal autoinducer concentration tested (20 μM) significantly increased its swimming ability in comparison to controls without addition of exogenous AHLs (Figure 3(A)). The addition to USDA 4438 of 5 or 10 μM 3OC₁₂ AHL caused a slight but significant increase in motility (Figure 3(B)), and the addition of 10 μM 3OC₁₄ AHL also caused a slight increase (Figure 3(C)).

Bacterial motility probably occurs as a dispersal strategy in nutrient-limited environments, or to searching for a new host [41–43]. Hoang et al. [45] showed that motility of the symbiotic S. meliloti is negatively regulated by a QS system, ExpR/SinI. When the bacterial population of S. meliloti grows, AHL production increases and motility is repressed. This phenomenon plays an important role in the interaction of S. meliloti with its host (alfalfa); i.e., the accumulation of bacteria around the host root and the scope of a quorum could provide an elegant strategy to suppress motility and simultaneously activate the invasion of the root.

There is no previous evidence for a connection between a QS system and motility control in peanut-nodulating Bradyhizobium sp. strains. We show here for the first time that the motility process in these bacteria is affected by the presence of QS signaling molecules (AHLs). It is clear that a finely regulated mechanism of cell signaling via AHLs coordinates the process of motility in these bacteria. It remains to be determined whether peanut-nodulating strains employ a motility mechanism for dispersion, colonization of new niches, invasion of the legume host, or other activities.

Bacteria are social organisms that live in terrestrial environments and form highly complex and highly organized communities. Such multicellular structures develop on various surfaces and facilitate survival and optimized resource utilization in hostile environments. This structure and associated lifestyle are termed a “biofilm” [6,46,47]. It has been estimated that over 99% of all bacterial activity in natural ecosystems is associated with bacteria organized in biofilms [48]. Thus, understanding the biology of biofilms is crucial in studies of ecosystems, including those in the soil. QS is a key mechanism that regulates several aspects of biofilm development, including adhesion, motility, maturation, and dispersal [49–52]. The present study is the first to examine the relationship between AHL production and biofilm development in peanut-nodulating bacteria.

Regardless of the presence of exogenously-added AHLs, all rhizobial strains tested were able to develop sessile biomass on a glass surface, with OD₅₇₀ values ranging from 0.4 to ∼5.0 (Figure 4). A high concentration (20 μM) of 3OC₁₀ AHL caused a significant (3–5 fold) increase in the biofilm formation ability of all strains, relative to strains grown without added AHL (Figure 4(A)). With lower concentrations (5, 10 μM) of added 3OC₁₀ AHL, there was either no difference in biofilm formation or a slight inhibitory effect (Figure 4(A)).

In comparison to the results with 3OC₁₀ and 3OC₁₄, all concentrations of 3OC₁₂ AHL tested caused a larger increase in biofilm formation (Figure 4(B)). Similarly to results with 3OC₁₀ AHL, a high concentration (20 μM) of 3OC₁₂ AHL produced the greatest increase in biofilm formation relative to controls. Such an increase was observed even at low concentrations (5, 10 μM) (Figure 4(B)), indicating that, in contrast to results with other signaling molecules, a significant increase in biofilm formation occurred in the presence of 3OC₁₂ AHL regardless of its concentration.

The effect of various concentrations of exogenously added 3OC₁₄ AHL was strain-dependent (Figure 4(C)). This molecule only slightly modified biofilm formation by P5 and USDA 4438. Surprisingly, an intermediate concentration (10 μM) of C₁₄ AHL greatly increased biofilm formation by 62B strain, whereas all C₁₄ AHL concentrations tested reduced biofilm formation by P8A relative to control (Figure 4(C)).

The ability of rhizobia to establish a biofilm can be used as a strategy for survival or for colonization and/or invasion [53,54]. In general, peanut-nodulating strains show increased biofilm formation ability in the presence of added AHLs with long acyl chains, independently of their capacity to synthesize AHLs. However, some strains showed different behaviors depending on the type and concentration of AHL. Since rhizobia are exposed to various and fluctuating environmental conditions, this range of behaviors is presumably employed depending on the prevailing environmental and biological conditions in order to increase rhizosphere fitness and to adapt and survive in a hostile soil ecosystem. The decision to change rhizobial lifestyle from planktonic cell to biofilm cell probably depends on the detection of certain environmental factors that direct the bacteria to activation of cell communication mechanisms that regulate the conduct of the entire population to the new state [6,47,54].

Several recent studies have addressed the regulatory effect of cell signaling mechanisms mediated by AHL on the process of biofilm formation in various Gram-negative bacteria, including Pseudomonas aeruginosa [55,56], Pseudomonas putida [57], Serratia liquefaciens [58], Aeromonas hydrophila [59], and Burkholderia cepacia [60]. Although there is no direct evidence for a mechanistic relationship between QS and biofilm formation in peanut-nodulating strains, AHLs clearly have an effect on biofilm formation. The ability of some AHL-non-producing strains to form biofilms may be due to the presence of signaling molecules (AHLs or other types) that we were not able to detect. However, there was a clear response to added AHLs by both AHL-producing and AHL-non-producing strains in this study. While there is no previous evidence that a sensing system mediated by AHLs is involved in the mechanism of biofilm formation in peanut-nodulating strains, the present results suggest that these signaling molecules play an important role in the regulation of this process. Such QS regulation may be exerted on surface components such as EPS, LPS, flagella, and pili, which are essential for different steps of biofilm development [54].

It is interesting that putative AHL-non-producing strains are able to respond to the exogenous addition of these signaling molecules. This finding is probably attributable to the ability of strains that are incapable of synthesizing AHLs to recognize them. This could be due to the presence of LuxR-like regulator molecules capable of binding the signaling molecules produced by other bacteria within a given population or community. We hypothesize that this mechanism is used by AHL-non-producing strains and confers an adaptive advantage for regulating certain activities, depending on the bacterial signaling molecules produced by other community members. In nature, biofilms are composed of multiple bacterial species and many other types of organisms, resulting in assemblies that are taxonomically and functionally complex and diverse [61]. Perhaps it is a common occurrence that QS signals produced by certain members regulate the activities of the community as a whole, in ways advantageous for bacterial survival [9].

Cell-cell interactions in bacterial cultures (planktonic cells) are manifested as the autoaggregative phenomenon [62]. Cell aggregation allows the establishment of microcolonies as well as the development of mixed microbial communities in natural habitats. Autoaggregation has been proposed as the initial step in biofilm formation; the cell aggregates formed could then attach to surfaces and develop a sessile population, conferring a competitive advantage and promoting bacterial survival in hostile environments [63,64].

Based on this process of physical cell-cell interaction, there may be mechanisms of cellular communication that mediate the autoaggregative phenomenon. No previous studies have addressed this process or its possible relationship with QS systems in Bradyrhizobium sp. We therefore evaluated the effect of AHLs on cell aggregation of peanut-nodulating strains.

Under our experimental conditions, the peanut-nodulating Bradyrhizobium strains studied showed a low capacity for cell aggregation, with values ranging from ∼5–20%. In comparison to controls, the addition of exogenous AHLs caused decreased autoaggregative ability in all strains except P5, which showed a slight increase (Figure 5(A,B,C)). All concentration of 3OC₁₀ AHL produced the greatest increases in autoaggregation of P5, with values ∼10-fold higher than those of controls (Figure 5(A)). 3OC₁₂ (Figure 5(B)) and 3OC₁₄ AHLs (Figure 5(C)) showed effects that were similar to but smaller than that of 3OC₁₀ AHL. Autoaggregation of USDA 4438 was not affected by addition of 3OC₁₀ AHL, but autoaggregation of 62B and P8A decreased in response to increasing concentrations of this autoinducer (Figure 5(A)). A negative dose-dependent effect by 3OC₁₂ AHL was observed on autoaggregation of USDA 4438, and a negative dose-independent effect by this autoinducer was observed on autoaggregation of 62B and P8A (Figure 5(B)). Exposure to any concentration of 3OC₁₄ AHL significantly reduced autoaggregation of USDA 4438 and 62B, and a high concentration (20 μM) of this autoinducer reduced autoaggregation of P8A (Figure 5(C)).

Previous studies have shown that QS regulatory systems are involved in control of bacterial autoaggregation, e.g., in Rhodobacter sphaeroides [65], Burkholderia thailandensis [66], Yersinia pseudotuberculosis [67], and Pseudomonas aureofaciens [68]. Our results indicate that the process of autoaggregation of peanut-nodulating strains is reduced or increased by QS signaling molecules. We can therefore speculate regarding the controlling effect of this process on the behavior of bacterial populations in nature. For some bacterial strains, QS signals may induce the dispersal of aggregates, leading individual bacteria to colonize new microniches in the soil. For other strains, QS mechanisms may positively regulate the autoaggregative process, enhancing the survival of the microorganisms in the soil.

In order to provide a more integrated picture of the impact of AHLs on autoaggregation, biofilm formation, and motility phenotypes of peanut-nodulating strains, we carried out a statistical multivariate PCA. This analysis allowed us to project the data set and variables in a graph in which they could be more easily visualized. The type of AHLs (depending on acyl chain length) associated with the different concentrations tested were the observations (or cases), while the variables were the motility, autoaggregation, and biofilm formation ability of each peanut-nodulating Bradyrhizobium strain.

According to the PCA, the analysis of all data in three dimensions (PC1, PC2, and PC3) explained ∼80% of the total variability in the study. The graph generated according to PC1 and PC2 (which explained ∼60% of the variability) (Figure 6) shows that motility and biofilm formation were processes regulated by the influence of 3OC₁₂ AHL. Similarly (although not conclusively), it appears that 3OC₁₀ and 3OC₁₄ AHLs affect the autoaggregative process.

Regarding the correlation between variables, there is a strong positive linkage between the variables of motility and biofilm formation for each strain (acute angle, positive correlation). This is consistent with previous findings that the different mechanisms of bacterial motility play important roles in the colonization of surfaces, which is the first step in the formation of biofilms [69].

Less strongly than for the correlation between motility and biofilm formation, we observed certain linkages between the processes of motility and autoaggregation (acute angles more open). This may reflect the logical necessity of bacterial motility to allow physical contact between the microorganisms prior to cellular aggregation.

In agreement with previous reports [69], motility appears to be a key mechanism for peanut-nodulating bacteria to establish various processes of cell interaction, such as cell aggregation and formation, remodeling, and disassembly of biofilms.

Surprisingly, there were weak or no linkages between biofilm formation and autoaggregation (right or obtuse angles), suggesting that, at least for peanut-nodulating Bradyrhizobium strains, these physiological processes may be regulated differently and probably used differently for the bacteria according to their exposure to environmental signals. Thus, the behavior of the bacteria could be driven to aggregation-disaggregation or biofilm formation-biofilm dispersal depending on the type of signal (physical, chemical, etc.), the presence-absence of a surface (biotic or abiotic), and other factors. The high biofilm formation capacity (Figure 4) compared to the low autoaggregation percentage (Figure 5) determined for these bacteria suggests that the former mechanism is more important in terms of the environmental benefits that these physiological processes confer to these peanut-symbiotic microorganisms.