This first step, attachment, is pivotal for biofilm formation (Figure 1). Aeromonads are able to colonize both biotic surfaces in plants and animals (Mizan et al., 2015), and abiotic surfaces, notably sediment, steel, glass, and polyvinyl chloride (Zalmum et al., 1998; Béchet and Blondeau, 2003; Bomo et al., 2004; Doğruöz et al., 2009; Balasubramanian et al., 2012). The substratum properties, chemical components, and nutrient availability are critical conditions influencing bacterial attachment. For instance, Jahid et al. (2013, 2015) have shown that low salinity (0.25% wt./vol.) enhances biofilm formation by Aeromonas hydrophila, whereas glucose concentration above 0.05% (wt./vol.) impairs its formation. In addition, Aeromonas spp. harbor several structures and/or mechanisms, including flagella and chemotaxis, lipopolysaccharides (LPS), and other surface polysaccharides (α-glucan), Mg²⁺ transporters and cytoskeletons that are actively involved in the first steps of biofilm formation (Figure 1).

Motility is decisive for attachment, and any system that promotes motility may stimulate attachment. Among these systems, the constitutive polar flagellum of Aeromonas spp., responsible for swimming in liquid, plays a critical role in biofilm formation and contributes to colonization of surfaces, as demonstrated for Aeromonas caviae strain Sch3 and A. hydrophila {A. piscicola} strain AH-3 (Kirov et al., 2004; Merino et al., 2014). In A. hydrophila, the O-glycosylation of polar flagella seems to be a prerequisite for adhesion and biofilm formation because mutants with reduced flagella glycosylation are unable to form biofilms (Merino et al., 2014; Fulton et al., 2015). In addition to polar flagella, members of Aeromonas spp. display inducible lateral flagella distributed randomly on the cell surface (Kirov et al., 2002). These lateral flagella are responsible for the swarming motility, enabling bacteria to migrate over surfaces by rotative movements and the formation of side-by-side cell groups called rafts (Gavín et al., 2002; Kirov et al., 2002). They also contribute to biofilm formation for Aeromonads (Gavín et al., 2002, 2003). Similarly, swimming, swarming, and twitching motility are known to be pivotal for P. aeruginosa biofilm formation (Barken et al., 2008), but Aeromonas strains do not develop any detectable twitching motility (Kirov et al., 1999). Chemotaxis systems mediated by the histidine kinase CheA allow bacterial cells to navigate in chemical gradients by regulating bacterial flagellar motility and are necessary for swimming, swarming, and for biofilm formation (Porter et al., 2011). Consistent with this, a chemotactic mutant strain ΔcheA of A. caviae is unable to swim or swarm in agar assays, and its biofilm formation ability is decreased by more than 80% (Kirov et al., 2004). Ten clusters of chemotaxis genes have been described in the genome of A. hydrophila ATCC 7966^T, including two gene-system homologs to the gene clusters I and V of P. aeruginosa PAO1 (Wuichet and Zhulin, 2010) shown to be essential for chemotactic motility (Masduki et al., 1995). The VgrG proteins corresponding to both components and effectors of the type VI secretion system (T6SS) of the A. hydrophila {A. dhakensis} strain SSU promote biofilm formation. Their function is not fully characterized but may occur at the attachment step because VgrG3 also enhances swimming motility (Sha et al., 2013).

The O-antigen is the most surface-exposed moiety of LPS, which acts as an attachment factor and enhances the formation of biofilm in A. hydrophila strains (Merino et al., 2014; Fulton et al., 2015). The cell surface hydrophobicity and charge conferred by LPS were involved in P. aeruginosa biofilm formation (Ruhal et al., 2015), but their exact roles in the Aeromonas biofilm development are not fully understood. The A. hydrophila surface α-glucan independent of the LPS also improves biofilm formation (Merino et al., 2012). In addition, Mg²⁺ is suspected to contribute to the integrity and stability of the outer membrane by preventing electrostatic repulsion between LPS molecules (Hancock, 1984). This divalent cation and its transporter MgtE are directly involved in adherence to epithelial cells, in swarming and in biofilm formation of A. hydrophila {A. piscicola} AH-3 (Merino et al., 2001). Genomic data confirm the links between MgtE and cell motility because the gene mgtE is adjacent to the polar flagellar operon flg in the strain A. hydrophila ATCC 7966^T (Seshadri et al., 2006). However, the level of evidence is still low for understanding how MgtE enhances biofilm formation. Finally, the protein MinD, a cytoskeletal ATPase involved in the septal placement of bacteria and plastid division sites in many bacteria (Shih and Rothfield, 2006), is also markedly involved in adherence, bacterial motility, and formation of biofilm in the A. hydrophila strain W (Huang et al., 2015).

Once attachment is completed, cell division processes maintain bacterium to bacterium bonds and lead to microcolony formation (Figure 1; Lynch et al., 2002). Type IV pili and fimbriae participate in the formation of microcolonies and biofilm for strains of Aeromonas (Béchet and Blondeau, 2003; Kozlova et al., 2008). Three families of type IV pili structures (Bfp, Flp, and Tap) have been involved in microcolony and biofilm formation for several species, e.g., Aggregibacter actinomycetemcomitans and Vibrio vulnificus (Paranjpye and Strom, 2005; Perez-Cheeks et al., 2012). Bacteria of the genus Aeromonas also harbor the three families of type IV pili structures (Kirov et al., 1999; Boyd et al., 2008). Bfp has been shown to be critical for biofilm formation in Aeromonas veronii (Hadi et al., 2012), but the involvement of Flp and Tap has not yet been demonstrated in aeromonads. Further evaluation deserves to be conducted to more precisely specify the role of each type of pili in early steps of Aeromonas biofilm formation.

From microcolonies, colonies grow and the biofilm matures. After 48 h in a stainless steel flow-through model, Lynch et al. (2002) observed mushroom-like “large microcolonies” of A. hydrophila, characteristic of mature biofilms (Figure 1). Aeromonads live embedded in a self-produced matrix of extracellular polymeric substances (EPS), mainly composed of polysaccharides, proteins, nucleic acids, and lipids (Andersson et al., 2011). The resulting structure provides the mechanical stability of biofilms (Peterson et al., 2015). When forming biofilms, A. hydrophila produces more capsular and colloidal EPS compared to planktonic cells (Castro et al., 2014). As in other bacteria including P. aeruginosa (Rasamiravaka et al., 2015), the persistence of a mature biofilm in Aeromonas is largely controlled by signaling systems triggered by high bacterial density (Lynch et al., 2002; Khajanchi et al., 2009). These aspects are developed in the section “How are biofilm and quorum sensing interconnected?”

Detachment corresponds to passive escape from a biofilm, occurring under the influence of external factors such as shear stress or degradation of extracellular polymeric matrix by enzymes, chelating agents, or detergents. In contrast, dispersion is an active mechanism of biofilm escape depending on biofilm growth, cell density, and related factors (Figure 1; Petrova and Sauer, 2016). Very few data on Aeromonas biofilm detachment and dispersion are available, and most knowledge relies on data from P. aeruginosa biofilm models. Dispersion is triggered by exogenous factors such as nutrient availability and toxic compounds, and by internal regulatory systems including quorum sensing systems (Kim and Lee, 2016). To enhance its dispersion, it is suspected that aeromonads degrade some compounds of their own extracellular polymeric matrix or other bacteria. Consistent with this, Bansal et al. (2015) showed that the depolymerase produced by an Aeromonas punctata strain is able to degrade the capsular polysaccharides of Klebsiella pneumoniae within a biofilm.

Under environmental conditions, biofilms often mix several bacterial species including Aeromonas sp., as reviewed in Table 1. Aeromonads have been co-isolated with one or more representatives of other genera from natural freshwater and seawater biofilms (Stine et al., 2003; September et al., 2007; Balasubramanian et al., 2012; Farkas et al., 2012; Marti et al., 2014). Aeromonas spp. are also observed in mixed-species biofilms inside dental care plastic lines (Tall et al., 1995) or the surfaces of food plants (Gunduz and Tuncel, 2006).

Bacterial biofilms harbor some inherent degree of structural heterogeneity in response to chemical gradients and adaptation to local microenvironments. Heterogeneity also concerns the mixture of bacterial species in a biofilm; juxtaposition of bacteria occurs such that mutualistic interactions are facilitated (Stewart and Franklin, 2008). Experimentally, the A. hydrophila strain AH-1 and Flavobacterium sp. form a mixed biofilm on chitin-containing particles. The A. hydrophila strain AH-1 is able to degrade chitin due to extracellular chitinases, and Flavobacterium sp. acts as a cheater that uses chitin degradation products as nutrients (Jagmann et al., 2012). The biofilm formation of Aeromonas was enhanced when co-cultured with a Bacillus cereus strain that improves aggregation between bacteria (Cheng et al., 2014). In addition to mutualism, competitive behavior was shown between aquatic bacteria organized in biofilms. For instance, A. hydrophila exhibits antagonism against L. pneumophila through the production of bacteriocine-like substances. This interfering effect enhanced the detachment of Legionella from biofilms, contributing to its dissemination (Guerrieri et al., 2008).

The medicinal leech Hirudo verbana is the natural host of the two symbiotic bacterial species A. veronii and Mucinovorans hirudinis (Graf, 1999; Worthen et al., 2006; Nelson et al., 2015). Synergy occurs between the two bacterial endosymbionts because A. veronii associated with M. hirudinis forms larger mixed microcolonies than each species alone (Kikuchi and Graf, 2007). The mechanisms involved in this synergy are not yet known, and it is unknown whether it could be controlled by QS signaling and whether bacterial cooperation could initiate or influence leech gut colonization in the two symbionts.

The AI-1 system was discovered within the LuxRI bioluminescent system in Vibrio fisheri (Engebrecht and Silverman, 1984) and is widespread among gram-negative bacteria including P. aeruginosa (Parsek et al., 1999). AI-1 signals are small molecules with a chemical structure based on N-acyl homoserine lactone (AHL), which is derived from common components of the bacterial metabolism, i.e., S-adenosyl methionine and acyl-acyl carrier proteins derived from fatty acid biosynthesis (Parsek et al., 1999). Depending on the species, the acyl chain length of AHLs varies from C4 to C18 and can be modified by unsaturation, methyl branches, and oxo- or hydroxyl substituents (Churchill and Chen, 2011).

The AI-2 QS system was initially described in Vibrio harveyi to control the expression of bioluminescence in response to fluctuation of bacterial population density (Bassler et al., 1994; Surette et al., 1999). A putative schematic model of Aeromonas AI-2 quorum sensing system is presented in Figure 3. AI-2 molecules are produced and detected by many gram-positive and gram-negative bacteria, including Aeromonas spp. and are considered a “universal signal autoinducer” with functions in interspecies cell-to-cell communication (Figure 3; Fong et al., 2001; Miller et al., 2004; Federle, 2009). AI-2 molecules are by-products of S-adenosyl-methionine (as AI-1) and correspond to a furanosyl borate diester in V. harveyi (Chen et al., 2002) or a variant lacking borate in Salmonella enterica (Figure 3; Miller et al., 2004). These molecules are synthesized by the enzyme LuxS (Xavier and Bassler, 2003) and freely diffuse across bacterial membranes (Figure 3). Under conditions of high cell density in V. cholerae, AI-2 molecules bind to a periplasmic receptor and lead indirectly to the derepression of the transcriptional regulator HapR (Figure 3; Henke and Bassler, 2004; Xavier and Bassler, 2005).

The genomes of A. hydrophila ATCC 7966^T and A. hydrophila {A. dhakensis} SSU contain homologs for AI-2 synthase LuxS and enzymes involved in signal transformation (AI-2 sensor kinase/phosphatase LuxQ, phosphorelay protein LuxU, regulatory protein LuxO) and a LitR encoding-gene, a homolog for the transcriptional regulator HapR of V. cholerae, and Lit-R-regulated genes (Figure 3; Kozlova et al., 2011). The transcription of luxS is negatively regulated by the expression of the locus ahyRI (Kozlova et al., 2011). The functions of AI-2 in the A. hydrophila {A. dhakensis} strain SSU have been studied by constructing ΔluxS deletion mutants. AI-2 is involved in the up-regulation of swimming motility and the down-regulation of biofilm formation and bacterial virulence in a murine model (Kozlova et al., 2008).

The type 3 autoinducer (AI-3) system is a hormone-like signal transduced by the two-component QseBC system in which QseC is the sensor kinase and QseB the response regulator (Figure 4). A putative schematic model of the Aeromonas AI-3 quorum sensing system is presented in Figure 4. AI-3 is suspected to behave similar to eukaryotic hormones because QseC is also a bacterial adrenergic receptor for the eukaryotic host hormones epinephrine and norepinephrine and is thus involved in interkingdom cross-signaling (Figure 4; Sperandio et al., 2003). AI-3 molecules are usually produced by gastrointestinal microbiota, e.g., in the context of symbiotic relationships between microbiota and host (Clarke et al., 2006). The periplasmic sensing domain of QseC is conserved among several gram-negative bacterial species (e.g., E. coli, S. enterica; Clarke et al., 2006).

Khajanchi et al. (2012) have identified and characterized the QseBC QS system in A. dhakensis that encodes a functional homolog for E. coli QseBC. In Aeromonas, AI-3 enhances swarming and swimming motility and virulence (e.g., hemolytic activity, protease and cytotoxic enterotoxin production, toxicity in murine models) and negative regulation of biofilm formation, as shown in the ΔqseB deleted mutant of the A. hydrophila {A. dhakensis} strain SSU (Khajanchi et al., 2012). The transcription of qseB and qseC genes is negatively regulated by the AI-1 QS system (Kozlova et al., 2012; Figure 4).

Bacteria commonly form dense communities in biofilms (Nadell et al., 2009). Consequently, biofilm lifestyle has emerged as a suitable model to study cell-to-cell signaling mechanisms. The three QS systems described in Aeromonas spp. are able to coordinate and influence biofilm formation, maintenance, and possibly dispersion. In addition, the major intracellular second messenger c-di-GMP has an impact on biofilm formation of Aeromonas spp. through QS systems, which is similar to the role observed in P. aeruginosa (Harmsen et al., 2010).

After the initial attachment phase, the AHL synthase AhyI and AHLs (C4-HSL and C6-HSL) are produced within the biofilm during microcolony formation and biofilm maturation (Lynch et al., 2002), as observed in other gram-negative bacteria and in particuliar P. aeruginosa (Davies et al., 1998; Huber et al., 2001; Labbate et al., 2004). Evidence of AI-1 effects on biofilms in aeromonads is provided by mutagenesis experiments and the effect of specific inhibitors.

Mutagenesis experiments clearly demonstrated that AI-1 influence depends on the biofilm development stage. The early attachment phase was not altered in ΔahyI (AI-1 synthase) or ΔahyR (AI-1 sensor/regulator) mutants of the A. hydrophila strain AH-1N (Lynch et al., 2002). Consistent with this, both swimming and swarming motilities were able to influence biofilm formation and were conserved in a double ΔahyRI mutant of A. hydrophila {A. dhakensis} strain SSU (Khajanchi et al., 2009). However, the later phases of biofilm maturation were affected by AI-1; the AH-1N ΔahyI (AI-1 synthase) mutant developed a biofilm less differentiated than that of the wild type (WT; Lynch et al., 2002). In addition, the number of viable cells was reduced in ΔahyI mutant biofilms compared to WT, but there was no difference in the viability of planktonic cells between ΔahyI and WT (Lynch et al., 2002). These defects in biofilm formation were partially restored by addition of exogenous AI-1, C4-HSL (Lynch et al., 2002). Confirming these observations, an A. hydrophila {A. dhakensis} SSU ΔahyRI mutant was defective and unable to form a well-structured biofilm, with alterations in filamentation, strain aggregation, and the (Khajanchi et al., 2009). In contrast to the ΔahyI (AI-1 synthase) mutant, the ΔahyR (AI-1 sensor/regulator) mutant of A. hydrophila strain AH-1N retains the ability to form mature biofilm, suggesting that the role of AHLs in the formation of large mushroom-like microcolonies does not depend on AhyR (Lynch et al., 2002) but presumably depends on transcriptional regulators other than AhyR. For instance, Kozlova et al. (2011) reported the presence of two loci that encode AhyR homologs in the strain A. hydrophila {A. dhakensis} SSU. In summary, these mutagenesis experiments showed that AI-1 molecules and AI-1 synthase enhanced biofilm formation.

Evidence for the role of AI-1 on biofilm formation is reinforced by studies focusing on AI-1 inhibitors, i.e., compounds that bear AI-1 quorum-quenching properties. For instance, chestnut honey decreases the level of C4-HSL molecules produced by the A. hydrophila strain CECT 839^T and leads to significant inhibition of biofilm formation (Truchado et al., 2009). Synthetic 2(5H)-furanone, derived from natural furanones produced by the marine algae D. pulchra, inhibits in vitro AHLs, including those usually produced by aeromonads (C4-HSL and C6-HSL), and results in biofilm inhibition in a strain of A. hydrophila (Ponnusamy et al., 2010). Vanillin exhibited quorum-quenching activity against aeromonad AHLs and reduced A. hydrophila biofilm by altering its architecture and decreasing surface cell-density and protein content (Ponnusamy et al., 2013). Mentha piperata essential oil and menthol have a quorum-quenching effect against C6-HSL. Sub-minimal inhibitory concentrations of mint essential oil and menthol decrease extracellular polymeric substance production and biofilm formation in the A. hydrophila strain WAF-28 (Husain et al., 2015). Examples of inhibition on biofilm formation displayed by several AI-1 inhibitors confirm dependence of biofilm formation to AI-1 QS.

Overall, the planktonic form is promoted by AI-2 and AI-3 because these two QS systems operate in negative feedback regulation in the biofilm formation of Aeromonas (Kozlova et al., 2008, 2012; Khajanchi et al., 2012).

A ΔluxS mutant of A. hydrophila {A. dhakensis} strain SSU defective in AI-2 showed enhanced amount of biofilm (Kozlova et al., 2008). Meanwhile, swimming motility, was reduced when luxS was deleted (Kozlova et al., 2008), suggesting that LuxS up-regulates swimming motility. This result should not be viewed as conflicting because it favors a switch to a planktonic lifestyle and is compatible with down-regulation of biofilm formation exhibited by the AI-2 QS system. Indeed, polar flagella are involved in biofilm formation at the initial stage, but this appendage is still fully expressed in the planktonic life of aeromonads (Canals et al., 2006). AI-3 signaling seems to exert similar effects to AI-2 on the down-regulation of biofilm formation and up-regulation of swimming motility as observed in the strain A. hydrophila {A. dhakensis} SSU and also up-regulates swarming motility (Khajanchi et al., 2012; Kozlova et al., 2012).

Cyclic di-guanosine monophosphate (c-di-GMP) is a secondary messenger that is mainly involved in the transition from a planktonic to sessile lifestyle in the domain Bacteria and was well-studied in P. aeruginosa or V. cholerae (Simm et al., 2004; Tischler and Camilli, 2005; Kuchma et al., 2007). For a comprehensive review, see Römling et al. (2013). Synthesis and degradation of c-di-GMP are under control of peptide domains widespread and universally conserved in bacteria: GGDEF (di-guanylate synthase activity) and EAL (phosphodiesterase activity), respectively. The A. hydrophila ATCC 7966^T genome contains 32 proteins with GGDEF, 9 proteins with EAL and 13 proteins with both (Rahman et al., 2007). In Aeromonas, the overexpression of the GGDEF domain enhances biofilm formation and inhibits swimming (polar flagellum dependent) and swarming motility (lateral flagella dependent) whereas, overexpression of the EAL domain leads to opposite effects. Thus, it is thought that c-di-GMP positively regulates biofilm maturation (sessile life) and negatively affects bacterial motility for aeromonads (planktonic life; Rahman et al., 2007; Kozlova et al., 2008, 2011, 2012; Khajanchi et al., 2012).

AI-2 and AI-3 QS are involved in the control of biofilm by c-di-GMP. For instance, one gene coding for a GGDEF domain protein, namely AHA0701, is located immediately downstream of the LuxS encoding-gene (AI-2 synthase gene) in the genomes of A. hydrophila ATCC 7966^T and A. hydrophila {A. dhakensis} SSU strains (Seshadri et al., 2006; Kozlova et al., 2011). Overexpression of AHA0701 increased the number of surface-attached cells and enhanced biofilm formation (Kozlova et al., 2011) and putatively led to extensive EPS production (Kozlova et al., 2012) by down-regulation of AI-2 synthase transcription. Corroborating the involvement of AI-2 and AI-3 QS in the c-di-GMP dependent regulation of sessile lifestyle in Aeromonas, the phenotype of strains that overproduced GGDEF domains varied in the same direction as ΔluxS (AI-2) and ΔqseB (AI-3) mutants vs. WT, i.e., increasing the maturation of biofilm and decreasing bacterial motility (Kozlova et al., 2008, 2011, 2012; Khajanchi et al., 2012). The regulation of sessile life through c-di-GMP also involves the AI-1 system in Aeromonas, as proved by the increase in C4-HSL production and enhancement of ahyI and ahyR transcription when c-di-GMP increases. In conditions of c-di-GMP overproduction, the transcript levels of luxS, qseB, and qseC genes also increased, but this influence is dependent on the AI-1 system (Kozlova et al., 2011). Finally, the ahyRI locus is required to promote c-di-GMP-mediated biofilm maturation (Rahman et al., 2007; Kozlova et al., 2011). Moreover, several homologs for genes encoding effectors of c-di-GMP signaling involved in biofilm formation of V. cholerae or P. aeruginosa have been found in Aeromonas spp. genomes. Among them, two transcriptional regulators containing the DNA-binding domains VpsR/FleQ and VpsT/CsgAB, and one master regulator for biofilm formation FleN were identified (Kozlova et al., 2011). In A. dhakensis, the transcription of effectors VpsR/FleQ, VpsT/CsgAB, and FleN is enhanced by c-di-GMP and is also modulated by AI-1, AI-2, and AI-3 systems, with opposite effects (Kozlova et al., 2011, 2012).

In summary, there is cross-talk between the three QS systems AI-1, AI-2, and AI-3 in the c-di-GMP regulation of the sessile life in Aeromonas (Kozlova et al., 2011, 2012).

A large amount of knowledge is available for understanding biofilm formation, QS systems, and interconnections between biofilms and QS in aeromonads. However, some knowledge is still missing for a comprehensive understanding of aeromonad social life. The type 1 autoinducer (AI-1) system is well-described, but further investigations are needed to specify the AHL profiles actually produced according to species, growth phase and origin of isolates. The AI-2 and AI-3 QS systems are poorly described and deserve to be more deeply investigated. Although, the AI-2 system has proved to be involved in Aeromonas biology thanks to mutagenesis evidence, the system could be better characterized. More specifically, homologs for the AI-2 receptor (LuxP in V. harveyi) have not been detected in the genome of A. hydrophila {A. dhakensis} strain SSU or A. hydrophila ATCC 7966^T (Kozlova et al., 2011), and the chemical nature of AI-2 produced from aeromonads should be clarified. Further studies are also needed to explore the AI-3 system and characterize the AI-3 signal(s) at the level of bacterial clone and within host interacting microbiota. Further studies designed to better understand this regulation system in Aeromonas spp. in high-density cell and biofilm situations are needed. Above all, QS and biofilm are the beginning of the understanding of complex and fine systems involved in aeromonad social life. Once the Pandora's box opened, additional exciting, unstudied questions will arise. Mixed biofilm structures and lifestyles likely result from a balance between cooperation and competition that we will have to unravel. Changes in gene expression during different stages of biofilms are another field to explore (Nadell et al., 2009). In addition, packed communities secrete into the local environment and share beneficial metabolites that are costly to produce, with benefits for individuals (Dumas and Kümmerli, 2012). These “public goods” may be exploited by non-producing mutants, i.e., cheaters (Heilmann et al., 2015). How biofilm structure, population dynamics within biofilm is impacted by cheaters is unknown. Understanding these aspects will help us to better understand the social life of aeromonads.

Activated sludge is a typical process for treating sewage and industrial wastewater taking advantage of the sociobiology of aeromonads, i.e., biofilm formation, QS and synergistic behavior. Members of the genus Aeromonas have been estimated to account for ~2% of the total biomass of activated sludge (Kampfer et al., 1996). Aeromonads organized in biofilm and aggregates have the biotechnological interest in the removal of organic pollutants (Chong et al., 2012), nitrogen (Chen et al., 2014), and phosphorus (Andersson et al., 2011) and are resistant to extreme conditions of pH, salinity, heavy metals, and temperature (Chen et al., 2014; Mohd Yasin et al., 2014).

Flocs in activated sludge are microbial aggregates enchased in a matrix of EPS composed of carbohydrates, proteins, lipids, and nucleic acids. These polymers come from bacterial cell metabolism, autolysis, and exogenous wastewater particles (Frølund et al., 1996). An analogy between biofilms and flocs is generally drawn because these bioaggregates correspond to a microbial surface producing EPS that improve the cohesion of the supracellular structure. The ability to flocculate was shown in vitro with a strain of Aeromonas sp. isolated from activated sludge (Chong et al., 2012).

To reduce hydraulic retention times, biofilm techniques are increasingly used to increase the surface of biodegradation. In a model of mixed biofilms composed of strains regularly present in activated sludge, Andersson et al. (2011) reported that mixed species biofilms that included A. hydrophila were associated with synergistic effects on denitrification and phosphorus removal. A shift in the composition of the produced EPS was also highlighted, with a significant rise in the carbohydrate fraction in the Brachymonas denitrificans model mixed with A. hydrophila, compared with single species biofilms. New polysaccharides were also detected but only from a mixed culture of A. hydrophila and Comamonas denitrificans or Acinetobacter calcoaceticus (Andersson et al., 2011). Another multi-species biofilm containing Comamonas testosteroni and A. hydrophila strains display synergistic biofilm formation, enabling better resistance of the biofilm to the repetitive physical and chemical shocks occurring in wastewater (Li et al., 2008). These interactive adaptive behaviors within the biofilm have great potential to improve wastewater treatment efficiency.

From typical activated sludge flocs containing Alpha-, Beta-, and Gammaproteobacteria, as well as members of Bacteroidetes and Actinobacteria, Li and Zhu (2014) showed by a quorum quenching method that AI-1 QS is necessary for the aerobic granulation of flocs and qualitatively and quantitatively regulates the content of exopolymeric substances. Aeromonas sp. in activated sludge are reported to produce AHLs (Chong et al., 2012; Ochiai et al., 2013). Interestingly, AHLs were localized in flocs community but absent from the aqueous phase (Chong et al., 2012). Moreover, the exogenous addition of AHLs led to increase the chitinase activity of the aeromonads isolated from activated sludge and of the whole activated sludge (Chong et al., 2012). Furthermore, QS signaling within activated sludge can also be modulated by bacteria degrading AHLs, such as Acinetobacter (Ochiai et al., 2013). All these data support that Aeromonas in flocs display QS-dependent social behavior improving water treatment.

Biofilm formation is involved in the pathogenesis of numerous human colonizations and/or infections (e.g., healthcare-associated infections) and persistent infections (e.g., chronic infection during cystic fibrosis, endocarditis, biliary tract infections, periodontitis, otitis media, wounds; Lebeaux et al., 2013). Aeromonads are occasionally involved in biofilm-associated colonization of medical devices, leading to infection such as lenses-associated keratitis (Willcox et al., 2001; Pinna et al., 2004) and central venous catheter-associated bloodstream infection (Andreoli-Pinto and Graziano, 1999; Tang et al., 2014). These reports suggest that biofilm acts as a virulence or patho-adaptive factor in Aeromonas infections. Consequently, biofilms are constantly evaluated in virulence studies. However, the correlation between biofilm producer and aeromonad pathogenesis is still weak. Presumably, the ability to produce biofilm in vitro may explain long-term digestive epithelium colonization by some strains whether they are associated with an infection (Janda and Abbott, 2010). Bfp type IV pili and polar/lateral flagella were suspected to be factors of intestinal colonization (Kirov et al., 1999, 2004). However, the production of lateral flagella is unlikely to be compulsory in digestive colonization because it is inconstantly expressed in Aeromonas spp. clinical isolates recovered from diarrheal feces (Kirov et al., 2002). The level of evidence for the exact contribution of biofilm to Aeromonas pathogenicity needs to be improved, and further studies are welcomed.

Because it was shown that several virulence determinants (e.g., alpha-hemolysin, cholesterol acetyltransferase, lipase, and serine protease) were overproduced at high cell density in late exponential/stationary phase, cell density could be a prerequisite for aeromonad pathogenic behavior and virulence expression (MacIntyre and Buckley, 1978; Whitby et al., 1992; Anguita et al., 1993). This suggests a pivotal role for QS in the course of infection. Both AI-1 and AI-3 signaling enhance the expression of some virulence determinants of aeromonads (Table 2; Figures 2, 4) and increase lethality in a murine model, as proved by the virulence-reduced phenotypes exhibited by ΔahyRI and ΔqseB mutants vs. WT (Khajanchi et al., 2009, 2012). In contrast, AI-2 signaling is associated with a down-regulation of bacterial virulence in a murine model (Kozlova et al., 2008), although molecular targets have yet to be identified. The pathogenesis of Aeromonas infections remain to be better understood, but QS-based cross-talk is indubitably involved in the expression of virulence from a threshold of cell density.

As described in P. aeruginosa, AHL molecules (e.g., 3-oxo-C12-HSL) are not only involved in bacterial virulence regulation but also interact with several eukaryotic cells and play a role in the immuno-modulation of the host response (Liu et al., 2015). Khajanchi et al. (2011) have reported a similar immune-modulatory effect with C4-HSL, C6-HSL, and N-3-oxo-C12-HSL in septicemic mice infected with A. hydrophila {A. dhakensis}. Indeed, AHL-based pretreatment was associated with reduced levels of cytokines/chemokines in tissues, increased neutrophil recruitment from blood (C6-HSL), enhanced bacterial clearance, limitation of clinical symptoms, and increase in survivability for infected mice. In addition, the AHL-based pretreatment increased in vitro bacterial phagocytosis by murine macrophages. Thus, it seems that exogenous AHLs act as a signal for the activation of host response to infection (Khajanchi et al., 2011). Both promising QS inhibitors that interfere with aeromonad QS, e.g., menthol (Husain et al., 2015), and some purified QS molecules (Khajanchi et al., 2011), represent new therapeutic perspectives to fight against aeromonosis threatening multi-drug antibiotic resistance.

There is increasing evidence that suggests the influence of multicellular microbial interaction during the course of infection, in the case of mixed infections driven by cooperation and/or competition (Trejo-Hernández et al., 2014; Armbruster et al., 2016). Microbial interactions may be particularly critical in the Aeromonas genus. Indeed, infections caused by Aeromonas are frequently polymicrobial, i.e., caused by Aeromonas and other bacterial genera (30–80% of the cases), (Lamy et al., 2009; Lay et al., 2010; Figueras and Beaz-Hidalgo, 2015); aeromonads are mainly associated with enterobacteria, Staphylococcus aureus and anaerobes. In 5–10% of human aeromonosis, infections are caused by heterogeneous populations associating several clones of Aeromonas spp. (Lamy et al., 2009). For some pairs of aeromonads co-isolated from clinical samples, the virulence during infection in Caenorhabditis elegans was higher for pairs than for each individual strain (Mosser et al., 2015). The exact mechanism of pathogenicity and aeromonads interactions during mixed infections remains unsolved, but intercellular communication likely plays a critical role in cooperation and/or competition behaviors. Ponnusamy et al. (2016) have also reported complex cross-talk in which two different clones of A. hydrophila affect each other, leading to a changing course of infection in a murine model. Indeed, an antagonistic effect through direct and host-mediated elimination was shown. In parallel, authors have observed a synergistic effect on the dissemination via local tissue barrier damage (Ponnusamy et al., 2016).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.