Introduction

Aeromonas species (spp.) are ubiquitous in nature predominately found in freshwater habitats and estuarine ecosystems. The organism commonly infects fish, amphibians, and reptiles, wreaking havoc on the fish-farming industry. The first documented case of a human Aeromonas infection was recorded in 1951 when the organism was cultured from cerebral spinal fluid during a patient’s autopsy in Jamaica (Caselitz 1996). Since this landmark case, 36 spp. have been added to the genus, and at least 19 of them have been classified as emerging human pathogens (Fernández-Bravo and Figueras 2020). Unlike many other human pathogens, Aeromonas spp. are unique in their ability to inhabit an enormous range of hosts. In addition to humans, they have been isolated from leeches, insects, mollusks, birds, livestock, fresh produce, preserved food, domestic animals, drinking water, and wastewater sludge (Didugu et al. 2015; Govender et al. 2021; Janda and Abbott 2010; McMahon and Wilson 2001; Wang et al. 2011; Wu et al. 2019). The organism’s ability to grow at refrigeration temperatures is an added concern in the food industry (Hoel et al. 2019). Alarmingly, human Aeromonas infections are not associated with just one predictable tissue type or set of symptoms but rather have been implicated in an impressive array of clinical syndromes including wound/soft tissue infections, septicemia/bacteremia, gastroenteritis, colitis, intraabdominal infections/peritonitis, urinary tract infections, pneumonia, and even dreaded necrotizing fasciitis (Janda and Abbott 2010). Historically, symptomatic infections have mostly been associated with immuno-compromised patients, and a number of comorbidities are commonly correlated with severe infection, mostly liver disease (Clark and Chenoweth 2003; Valcarcel et al. 2021; Xu et al. 2022). While the majority of Aeromonas research is focused on aquaculture and relieving the economic burden of fish disease, there is a pressing need to understand this pathogen in a human-disease context and to innovate and develop novel, clinically relevant treatments as antimicrobial resistance (AMR) continues to spread in this pathogen.

Overuse/misuse of antibiotics, both clinically and commercially, has accelerated AMR development world-wide in a large number of human bacterial pathogens. This has resulted in bacterial infections that are increasingly difficult to treat, leading to higher mortality rates and longer hospital stays (Orosz et al. 2022; Wagenlehner and Dittmar 2022). The seemingly unchecked emergence of AMR pathogens poses a significant global threat to public health, prompting the need for a better understanding of the molecular mechanisms of resistance and the development of novel countermeasures. Aeromonas spp. are no exception to this emerging trend and have demonstrated their ability to rapidly acquire and share new AMR genes. Identifying/developing solutions to this problem requires a multifaceted approach including achieving a better understanding of Aeromonas spp. acquisition and retention of AMR genes, as well as identifying and characterizing virulence factors/mechanisms as potential drug targets. One promising antibiotic-alternative drug target is quorum sensing (QS), an essential virulence mechanism for Aeromonas spp. during human infections.

Antibiotic resistance in Aeromonas: an ongoing problem

Considering Aeromonas’ pervasiveness in both ecological and clinical environments, promiscuity, and ability to cope with (and endure) environmental stressors, it should not be overlooked as a potentially significant source and/or reservoir of clinically relevant AMR genes. AMR mechanisms, particularly the presence of ß-lactams, including carbapenemases (Hayes et al. 1994, 1996), have been well characterized in Aeromonas spp. over the years. Chromosomally encoded ß-lactams were among the first to be detected and genetically characterized in Aeromonas (Iaconis and Sanders 1990; Ko et al. 1998; Walsh et al. 1997). Today, AMR genes that encode penicillinases, cephalosporinases, and metallo-ß-lactamases are of common occurrence in many Aeromonas spp. (Nwaiwu and Aduba 2020; Pourmohsen et al. 2023; Hilt et al. 2020). AMR genes do not observably accumulate in one species of Aeromonas more than another (Bertran et al. 2021). This is, in part, due to the fact that accurate identification of Aeromonas on a species level is historically problematic due to standard diagnostic techniques that are not tailored to Aeromonas spp. and shifting taxonomic definitions (Zhang et al. 2023; Fernández-Bravo and Figueras 2020). For example, A. dhakensis is being reevaluated in recent years for clinical significance, especially as a culprit in monomicrobial, systemic infections because of frequent misidentification as A. hydrophila (Wu et al. 2015, Puah et al. 2022). These factors combine to make studying species-specific AMR trends difficult.

Importantly, clinical AMR Aeromonas strains have been isolated across the globe; 3rd-generation cephalosporin-resistant Aeromonas has been isolated in Southern India (Bhaskar et al. 2015), and 3rd-generation cephalosporin and carbapenem-resistant isolates have been found in Croatia (Drk et al. 2023). Two broad-spectrum carbapenemase KPC-24-producing A. veronii strains were isolated recently from hospital sewage in China (Yang et al. 2022). Indeed, AMR profiles have been characterized in Aeromonas spp. isolated from India (Indra et al. 2015), including North Bengal (Dey Bhowmick and Bhattacharjee 2017), Tunisia (Bargui et al. 2023), Egypt (El-Hossary et al. 2023), Thailand (Hatrongjit et al. 2020), European countries, and Brazil, just to name a few (Table 1). Sequence analysis of nine independent A. veronii isolates from fish, humans, and Brazilian environments found all isolates to be remarkably similar to the uploaded Aeromonas genomes found in NCBI, demonstrating that widespread distribution of AMR genes in Aeromonas can originate from vastly different sources and geographic locations (Maia et al. 2023). Environmental isolates unassociated with human/clinical disease also demonstrate an alarmingly high rate of AMR. This suggests a possible ecological reservoir of AMR genes with the potential to be transferred to human-associated strains (Canellas et al. 2023; Goñi-Urriza et al. 2000; Igbinosa et al. 2015). It is clear that Aeromonas spp. are formidable enemies capable of horizontally spreading AMR genes, and it is important to understand how and why such propensity for AMR acquisition and retention exists.

Table 1 A table summarizing representative studies characterizing AMR in Aeromonas over the years
Full size table

Sources and selective pressures for AMR acquisition

As alluded to earlier, one commonly studied mechanism of AMR acquisition in Aeromonas is horizontal gene transfer from other bacteria. A. caviae, for example, has been shown to be naturally competent and readily acquires DNA from its environment (Sayeed et al. 1996). In one study, 73% of environmental Aeromonas isolates were able to serve as recipients of donor DNA, while 100% of tested isolates were able to act as donors to at least some other aeromonads under optimal laboratory conditions (Tris buffer with magnesium or calcium, pH 5–8, and a saturating concentration of 0.5 μg of DNA per assay, at 30 °C; sodium was also required) (Huddleston et al. 2013). On account of this naturally transformative state, it is no surprise that so many AMR genes have been found in Aeromonas spp. that are derived from other common but unrelated pathogens. For example, a Verona integron-encoded family metallo-β-lactamase (VIM) producing A. hydrophila strain carrying a VIM-4 gene was described in a case report from Budapest in 2008. Sequencing showed an identical match to a previously characterized integron in Pseudomonas aeruginosa from southern Hungary, suggesting DNA transfer between the two (Libisch et al. 2008). Temoniera-24 (TEM-24), a prominent extended-spectrum β-lactamase gene variant, was observed in a clinical isolate of Aeromonas for the first time in 2003. TEM-24 is typically isolated across Western Europe from Enterobacter spp. and Pseudomonas spp., and since this particular Aeromonas spp. was isolated alongside Enterobacter aerogenes, this also suggests cross-species horizontal AMR gene acquisition (Marchandin et al. 2003).

Another mechanism that could be promoting the alarming rate of AMR acquisition is the presence of AMR genes in the context of mobile genetic elements (MGE) such as plasmids and integrons. Aeromonas spp. are known to possess a collection of plasmids constituting its plasmidome. The Aeromonas plasmidome is of particular interest in the context of AMR genes and other virulence factors (Vincent et al. 2021). Virulence-related plasmids released by bacterial cells have been shown to persist in harsh environments such as treated wastewater and can readily be acquired by nearby pathogens (Drk et al. 2023). Additionally, integrons, small sections of chromosome that can capture gene cassettes from the environment and incorporate them into the genome via integrase genes and site-specific recombination, have been found to play an important role in the acquisition and spread of antibiotic resistance genes (Fluit and Schmitz 1999). Such integrons have been found within multiple species of the Aeromonas genus. Characterization of 133 Aeromonas spp. isolates (50 A. caviae, 45 A. hydrophila, 31 A. sobria, 6 A. encheleia, and 1 A. veronii) revealed the presence of several different class I integrons, including 10 different gene cassettes, encoding resistance to a variety of antibiotics including trimethoprim, aminoglycosides, ß-lactams, and phenicol. As to be expected, antibiotic resistance rates were different between integron-positive and integron-negative strains. Specifically, resistance to trimethoprim and trimethoprim–sulphamethoxazole was more commonly associated with integron-positive isolates, and all integron-positive isolates were resistant to more than 3 antibiotics. In fact, resistance to as many as 10 antimicrobial chemotherapeutics was observed in some integron-positive strains (Chang et al. 2007). Additionally, a global study of 38 A. salmonicida isolates revealed that 21/38 isolates contained a class I integron with all gene cassettes described in the study being associated with human bacterial infections (L'Abée-Lund and Sørum 2001).

Non-pathogenic, environmental/aquatic Aeromonas spp. have also proven to be significant sources/reservoirs of AMR gene acquisition of clinically relevant strains. In one such example, the transfer of oxytetracycline-resistant plasmids between Aeromonas spp. found in fish hatcheries and Aeromonas spp. recovered from hospital effluent was observed in A. hydrophila (8 isolates), A. sobria (6 isolates), and A. caviae (1 isolate) (Rhodes et al. 2000). Additionally, A. veronii isolated from catfish ponds in the South-Eastern USA was found to harbor tetracycline resistance gene on a MGE, similar to ones found in Vibrio parahaemolyticus and other Aeromonas spp. isolated from human stool (Dubey et al. 2023). When 66 Aeromonas isolates [A. caviae (58%), A. hydrophila (17%), A. media (11%), and A. veronii (11%)] from both untreated hospital wastewater and treated municipal water were examined, almost all of them (65/66) demonstrated multidrug-resistant phenotypes. Prevalent carbapenem genes found among the isolates included blaKPC-2, blaVIM-2, blaOXA-48, and blaIMP-13, with the latter three being described for the first time in Aeromonas. This same study demonstrated the ability of some of these Aeromonas isolates to transfer these resistance phenotypes to susceptible recipients (Escherichia coli), suggesting conventionally treated municipal and untreated hospital wastewater may be a reservoir for AMR, and that Aeromonas spp. could be mediating the spread of AMR to other pathogens in that environment (Drk et al. 2023). By demonstrating overlap between aquatic and clinical Aeromonas spp., these findings warn of the perils of compartmentalizing human and agricultural/environmental niches and conversely suggest that they should be considered as one combined environment since the transfer of genetic information can occur between them (Jones et al. 2023). Roh and Kannimuthu (2023) in a recent genomic analysis of the resistomes of 400 Aeromonas aquaculture strains found resistance against carbapenem, fluoroquinolone, cephalosporin, elfamycin, aminoglycoside, and tetracycline was “more or less evenly distributed across all species, while resistance against the other classes varied between species” (Roh and Kannimuthu 2023). This distribution highlights the genetic promiscuity displayed by Aeromonas across species and underscores its potentially global relevance as an indicator organism of the spread of antibiotic resistance (Usui et al. 2016).

Mechanisms of Aeromonas-associated drug resistance

Aeromonas employs a multivariate platform of strategies that confer antimicrobial resistance. One such strategy is through the exploitation of escape mutations in genes encoding the protein targets of the antibiotics. For example, quinolone resistance observed in an A. caviae human isolate was a result of an accumulation of point mutations in the type II topoisomerase genes gyrA and parC allowing for the continued function of the enzymes while reducing antibiotic binding affinity (Sinha et al. 2004). Another AMR mechanism employed by Aeromonas is the use of substrate-specific antibiotic degrading enzymes such as ß-lactamases (Majiduddin et al. 2002; Rasmussen and Bush 1997). For example, MOX-9, a class C enzyme belonging to a novel sub-lineage of MOX ß-lactamases, was found to be encoded by a chromosomal transposon in A. media. Biochemical characterization of this MOX-9 gene revealed a strong binding preference for cephalosporins and cephamycins. By comparing MOX-9 binding affinity and its hydrolysis activity to other more common MOX-type enzymes, this study not only demonstrated the variations that exist within this family of resistance genes but also provided a genetic context by which resistance genes can be easily mobilized onto transmissible plasmids and horizontally shared among other organisms (Piccirilli et al. 2022).

A third strategy is the use of broad-spectrum, non-specific AMR techniques such as drug uptake resistance, efflux pumps, and/or enhanced biofilm production. In one demonstration, when the ompR gene encoding an outer membrane protein (OMP) was deleted in an A. veronii isolate, increased sensitivity of the mutant culture to both ceftriaxone and neomycin, two different classes of drugs, was observed. Interestingly, the ΔompR mutant was shown to exhibit reduced biofilm production as well. It was speculated that this increased antibiotic susceptibility may be due, in part, to the reduction in biofilm formation since biofilm typically impedes an antibiotic’s access to the bacteria (Wang et al. 2023b). Indeed, when 29 OMP knockout strains were created in A. hydrophila, 22 gene deletions affected susceptibility levels to at least one of the 20 antibiotics tested. That being said, no OMP mutant exhibited consistent responses to all the tested antibiotics, eluding to more complicated downstream signaling/regulatory mechanisms underlying OMP-related drug uptake (Li et al. 2019). Additionally, when the porin protein Aha1 was mutated in A. hydrophila at its lysine-acetylation sites, increased resistance to tetracyclines and ß-lactams was observed, presumably due to decreased drug uptake (Zhang et al. 2022d). Recently, in China, a resistance-nodulation-division (RND)-type efflux pump gene cluster named tmexCD1-toprJ1636 was discovered in a related Gram-negative enteric pathogen Klebsiella pnuemoniae which confers resistance to different classes of antibiotics including tetracyclines, cephalosporins, aminoglycosides, phenicol, quinolones, and the last-resort antibiotic tigecycline (Lv et al. 2020). Such an efflux pump has since been shown to play a role in Aeromonas spp. drug resistance as well; in an environmental study, 36 of the 636 Aeromonas spp. isolated from livestock, meat, water, and humans [A. caviae ( 22), A. hydrophila (5), A. salmonicida (1), and A. veronii ( 8)] were positive for the multidrug-resistant gene cluster mentioned above, either encoded chromosomally or on a plasmid. Importantly, the characterized tmexCD-toprJ genes were associated with different Aeromonas spp., phylogenetic lineages, environments, and genetic locations and were surrounded by varying MGEs, demonstrating alarming diversity (Wu et al. 2023).

In reality, AMR in Aeromonas likely results from many different complicated factors, all playing simultaneous and even interactive roles. Employing a proteomic approach, a quinolone, norfloxacin (NOR), stress response study in A. hydrophila revealed 186 downregulated proteins and 220 upregulated proteins following exposure. Interestingly, many of the differentially expressed proteins were involved in sulfur metabolism and homologous recombination. Seven of these differentially expressed proteins were chosen as targets for site-directed mutagenesis in their encoding genes. Some mutants exhibited increased sensitivity to NOR such as ΔAHA_0904 (an uncharacterized protein) and ΔcirA (colicin I receptor), whereas the ΔhlyD (in the secretion family) mutant significantly increased NOR resistance. Other mutants, ΔAHA_4275 (a ferrichrome receptor), Δicd (isocitrate dehydrogenase [NADP]), ΔcheV (chemotaxis coupling protein), and ΔppsA (phosphoenolpyruvate synthase) displayed no differences compared to the parental A. hydrophila strain (Liu et al. 2023). This suggests that genes with no apparent ties to AMR can play an important role in an organism’s resistance/susceptibility to antibiotic stress.

Transcriptional regulators have also been shown to influence Aeromonas spp. drug resistance. More specifically, the transcriptional regulator AhslyA was shown to play a role in fluoroquinolone resistance. In an A. hydrophila ΔahslyA mutant, increased fluoroquinolone Enoxacin (ENX) sensitivity was observed. Proteomic analysis revealed differentially produced proteins involved in DNA metabolism, the SOS response, and cell communication following ENX treatment. Site-specific mutations were then engineered in several targets’ encoding genes, three genes related to decreasing protein abundance (AHA_0655, AHA_1195, and AHA_3721), and three genes related to increasing protein abundance (AHA_1239, AHA_2114, and narQ). The ΔAHA_2114 and ΔnarQ mutants had slightly decreased resistance to ENX at 0.01 μg/mL, and mutants ΔAHA_1239 and ΔAHA_3721 demonstrated an increase in resistance to ENX at 0.01 μg/mL. This further demonstrates the genetic diversity of expression/regulation involved in conferring drug resistance (Li et al. 2021). Collectively, these studies underscore the complex network of overlapping known pathways and mechanisms involved in AMR. When considering the potential contributions of yet unknown pathways, the network becomes even more complex.

Quorum sensing

Quorum sensing (QS), broadly, is a sophisticated mechanism of communication utilized by bacteria to coordinate behavior in a population. There are three major types of quorum sensing systems in Aeromonas known as autoinducers 1, 2, and 3. Autoinducer 1 (AI-1) QS is a system found exclusively in Gram-negative bacteria and is thought to detect and respond to the population density of members of the same species in an environment (Vanetti et al. 2020). Autoinducer 2 (AI-2) QS is a mechanism thought to mediate cross-species communication, given the machinery for this system is found in both Gram-positive and Gram-negative bacteria (Zhao et al. 2018). Autoinducer 3 (AI-3) QS is a two-component response system found in bacteria that responds to signals produced by members of the eukaryotic kingdom, demonstrating its use as an inter-kingdom mode of communication (Fan et al. 2022) (Fig. 1).

Fig. 1
figure 1

A schematic demonstrating the basic mechanisms that govern the three QS systems: AI-1 QS (far left) is only found in Gram-negative bacteria and is thought to be the mode of intra-species communication. Bacteria in a community simultaneously produce AI-1 signal via an AI-1 synthase (AhyI) where the signal is then sensed and responded to via the response regulator AhyR. AI-2 QS (center) is found in both Gram-positive and Gram-negative bacteria and is thought to be the mechanism of communication between different bacterial species. AI-2 signal is produced by all members of the bacterial community via LuxS. LuxR is responsible for sensing and responding to AI-2 signals. AI-3 QS (far right) is a two-component phosphorylative response system thought to be a mode of communication between prokaryotes and eukaryotes (inter-kingdom). Signals produced by a eukaryotic host collectively known as AI-3 cause a conformational change in membrane-bound sensor kinase (QseC), allowing for phosphorylation of the cytoplasmic response regulator (QseB), which activates it. (Image produced in BioRender)

Full size image

In many pathogenic bacteria, including Aeromonas spp., QS has been shown to globally regulate virulence gene expression and/or disease-causing mechanisms. Some of these virulence factors/mechanisms, utilized by Aeromonas and regulated by QS, include biofilm formation, motility, and effector protein secretion through various (e.g., types 2, 3, and 6) secretion systems (Table 2).

Table 2 Well-characterized Aeromonas virulence factors and their respective functions including helpful references on the topic
Full size table

Autoinducer 1 quorum sensing: AhyRI

A QS system homologous to the LuxRI system in V. fischeri (Kempner and Hanson 1968) was first described in Aeromonas in 1997 (Swift et al. 1997). The system produces acyl-homoserine lactones (AHLs), molecular signals collectively known as AI-1, which are synthesized by the AHL synthase AhyI. A corresponding response regulator, AhyR, is then modulated by this signal (Chu et al. 2013; Swift et al. 1999; Van Houdt et al. 2007) to alter downstream gene expression. By far the most studied of the 3 Aeromonas-associated QS systems, AI-1 QS is ubiquitous across Aeromonas spp. (Jangid et al. 2007) and has been shown to influence the development of biofilm (Lynch et al. 2002), exo-proteases production (Khajanchi et al. 2009; Swift et al. 1999), outer membrane protein profiles, S (surface)-layer thickness (Bi et al. 2007), and type 6 secretion system (T6SS) effector secretion (Khajanchi et al. 2009) (Table 2). It has also been shown that mutations in this QS system result in decreased virulence potential of Aeromonas. More specifically, the virulence of a ΔahyRΔahyI double mutant was reduced by 50% when compared to its parental strain A. hydrophila SSU [since re-classified as A. dhakensis (Grim et al. 2014)] in a murine model of infection (Khajanchi et al. 2009). In a fish infection model using a challenge dose of 109 colony forming units (CFU)/ml, A. hydrophila J-1 mutant ΔahyR was rendered avirulent, as evidenced by the 100% survival of challenged fish. In contrast, 100% fish mortality was observed when challenged with the parental strain at the same dose (Bi et al. 2007). To better understand the AI-1 QS pathway, an AHL lactonase was used to block AI-1 signaling. Further evaluation of the data revealed differential expression of genes, post AHL lactonase treatment, that were involved in a myriad of metabolic pathways including metabolite transport, amino acid metabolism, central metabolism, and respiration, suggesting universal metabolic regulation by AI-1 QS (Gui et al. 2017).

While it is well established that AI-1 QS plays a crucial role in the virulence of Aeromonas (Bi et al. 2007; Khajanchi et al. 2009), little is known about the specifics of AHL synthesis or substrate specificity. It was historically thought Aeromonas only had the ability to synthesize two AHLs, N-butanoyl-L-homoserine lactone (C4-HSL) and N-hexanoyl-L-homoserine lactone (C6-HSL) (Kirke et al. 2004; Swift et al. 1997, 1999). To expand that knowledge, one study successfully purified 6 unique AHLs from A. hydrophila, although C4-HSL and C6-HSL continued to be the most abundant signals. The mechanism by which AhyI is able to catalyze the formation of the various AHLs was proposed to likely employ small molecules S-adenosyl-L-methionine (SAM) and butyryl-acyl carrier protein (ACP) as facilitators. If indeed SAM and ACP are involved in AHL synthesis, then AHL synthesis utilizes an acyl-ACP-derived fatty-acyl substrate and not acyl-CoA, as previously thought (Jin et al. 2020).

Autoinducer 2 quorum sensing: LuxS

The LuxS universal QS system mediated by AI-2 has also been described in the Aeromonas genus as early as 2008 (Kozlova et al. 2008). Unlike AI-1, this QS system is found in both Gram-positive and Gram-negative bacteria and is thought to be the means of cross-species communication (Xavier and Bassler 2003). Since its discovery, other publications have corroborated the existence of AI-2 systems in Aeromonas (Zhao et al. 2015); however, less research has focused on this QS system in Aeromonas spp. compared with AI-1 QS. The phenomenon generally observed has been an overall increase in virulence when AI-2 (luxS gene) is deleted. This is in sharp contrast to the deletion of the AI-1 system components. More specifically, an A. dhakensis SSU ΔluxS mutant exhibited decreased motility, increased virulence (as observed by increased lethality in a murine model), and altered biofilm structure. Surprisingly, the increased virulence in a septic mouse model of infection was not due to alterations/enhancements in hemolytic activity, AexU (a type 3 section system effector) translocation, or T6SS effector translocation (Kozlova et al. 2008) (Table 2). Furthermore, LuxS deficiency negatively affected expression levels of the A-layer gene encoding VapA, potentially reducing survivability in host macrophages (Meng et al. 2017). In an effort to uncover the mechanism(s) for these phenotypes, the DNA adenine-methyltransferase (Dam) encoding gene was overexpressed in both the parental and the ΔluxS mutant. The overexpression of dam caused the ΔluxS mutant to become hyper-motile and demonstrated increased hemolytic activity as compared to the isogenic dam-overexpressing parental strain. However, the overexpression of dam did not alter the virulence potential of the ΔluxS mutant in vivo. Taken together, these results suggest that the methylation of LuxS may play a role in the regulation of the AI-2 QS system (Kozlova et al. 2008) and needs further investigation.

To gain more insight into the signaling pathway downstream of AI-2 QS, the LuxS-regulated gene B protein (LsrB) was investigated in A. veronii. This protein belongs to the high-affinity substrate-binding protein family and is one of the two D-type receptors (LuxP and LsrB) of the AI-2 molecule in the AI-2 QS system. The major role of LsrB is to internalize extracellular AI-2 (Reading and Sperandio 2006). When this receptor was deleted in A. veronii, there was no apparent impact on growth, hemolytic activity, or antibiotic sensitivity. Motility was slightly decreased, likely on account of reduced flagellar gene expression, and a significant reduction in biofilm formation was also observed. Interestingly, the subsequent interruption of the AI-2 signaling pathway following LsrB deletion resulted in an unexpected decrease in virulence in a zebrafish model (as measured by an increased LD50) (Gao et al. 2023) contradicting the previous murine study (Kozlova et al. 2008). Additionally, AI-2 QS signaling seems to be affected by post-translation modifications (PTMs). By mapping out lysine-acetylation and lysine-succinylation sites in A. hydrophila, the sites were found to be largely overlapping. One such overlap was in the amino acid K165 in the luxS gene. Acetylated LuxS was found to negatively regulate LuxS enzymatic activity in A. hydrophila, while conversely, succinylated LuxS (at the same residue) positively regulated enzymatic activity. Interestingly, two distinct PTMs of LuxS on a specific residue oppositely influenced bacterial AI-2 QS activity (Sun et al. 2019), suggesting that the role LuxS plays in Aeromonas’ biological functions may be partially dependent on PTM status. This aspect may potentially contribute to the difference in phenotypic virulence in murine versus Zebra fish models and requires further investigation.

Autoinducer 3 quorum sensing: QseB/QseC

A third QS system mediated by two-component regulatory proteins QseB and QseC, which respond to AI-3 molecules, was identified in A. dhakensis SSU in 2012 (Khajanchi et al. 2012). Since then, 15 environmental Aeromonas isolates from China have been found to possess qseBC genes, demonstrating the widespread nature of this system within the genus (Sarkodie et al. 2019). Of the three QS systems identified in Aeromonas spp., the QseB/QseC system is the most poorly understood. Upon discovery, a ΔqseB mutant was constructed in A. dhakensis SSU, and the mutant exhibited diminished swarming and swimming motility, increased biofilm density, reduced protease production, and a slightly decreased virulence with 30% lower mortality over a test period of 16 days in an in vivo murine model of septicemic infection compared to the parental strain (Khajanchi et al. 2012). In contradiction to this study, a recent 2023 study reported that AI-3 QS component deletions in A. hydrophila did not affect motility, decreased biofilm production, and promoted increased virulence in an in vivo fish model (Qin et al. 2023a). Given the contrary nature of these two reports in different animal models, clearly, more studies are needed to better understand this complicated QS system.

One interesting finding in a fish model of infection study (Qin et al. 2023a) was its use of the host-derived stress hormone norepinephrine (NE). QseBC has previously been shown to enable many entero-bacteria to sense and interact with the host-derived environment (Lustri et al. 2017; Moreira and Sperandio 2016). This may also be true for Aeromonas spp. since the addition of NE to the medium of A. hydrophila increased its growth rate and dramatically increased biofilm production. However, ΔqseB and ΔqseC mutants did not display the aforementioned NE-mediated responses, suggesting that host signaling molecules are, in some way, associated with Aeromonas behavior, likely through the AI-3 QS (Qin et al. 2023a).

Interactions between QS systems

To complicate matters further, the three QS systems described above could possibly interact with one another, creating a complicated network of QS pathways replete with crosstalk and overlap. In an attempt to elucidate the ambiguity, one study systematically compared QS-related gene expression in mutants of all three QS systems. It was found that individual component deletions resulted in altered expression levels of the other QS system genes. In the ΔahyRI mutant, qseB, qseC, and luxS genes were all upregulated. In the ΔqseB mutant, ahyR and ahyI gene expression levels were downregulated; however, no changes were observed in luxS expression. Finally, in the ΔluxS mutant, no changes were observed in qseB and qseC expression levels. Taken together, these findings demonstrate that crosstalk and/or compensatory interactions between/among the various Aeromonas spp. QS systems occur (Kozlova et al. 2012).

Role of C-di-GMP in QS system interactions

C-di-GMP is a small signaling molecule that plays a crucial role in the regulation of bacterial behavior and physiology including all three Aeromonas spp. QS systems. An initial report demonstrated a link between c-di-GMP and AI-1 QS in A. sobria (Rahman et al. 2007). C-di-GMP overexpression in A. hydrophila was shown to enhance biofilm formation and reduce motility in the ΔluxS mutant and its parental strain. In contrast, the ΔahyRI mutant only showed a marginal increase in biofilm formation with no effect on motility when c-di-GMP was overexpressed (Kozlova et al. 2011). Overexpression of c-di-GMP reduced protease activity in the ΔqseB mutant when compared to the isogenic parental strain, and no changes in protease activity in the ΔahyRI mutant were observed. Furthermore, increased c-di-GMP expression in parental A. dhakensis SSU produced denser biofilms while increased c-di-GMP in the ΔqseB mutant decreased biofilm density (Kozlova et al. 2012). Collectively, the varying regulations each QS system exerts on one another, either positively or negatively, may be mediated by this small signaling molecule that has the demonstrable ability to communicate with all three.

Investigating the role of c-di-GMP in QS regulation led to the discovery that both AI-1 and AI-2 QS systems in Aeromonas affect expression levels of the transcriptional regulator LitR (Kozlova et al. 2011). The LitR homolog, HapR, has been shown to universally regulate virulence factors in V. cholerae (Kovacikova and Skorupski 2002). LitR has since been shown to bind to the promoter regions of the hemolysin and serine protease genes, as well as T6SS effector protein VrgG in A. hydrophila. LitR was also found to positively regulate hemolytic and extracellular protease activities (Zhao et al. 2023). This establishes LitR as a master transcriptional regulator used to control the expression of many essential virulence factors, and the expression level of LitR is regulated by both AI-1 and AI-2 QS systems, further demonstrating the overlap between the systems.

QS inhibition: an alternative therapeutic to antibiotics

Understanding the Aeromonas spp. QS systems and their crosstalk could enable exploitations that may result in promising alternative therapeutics for Aeromonas infections and beyond. The need for alternative therapeutics is especially apparent in Aeromonas spp. on account of their being both the cause of severe infections in humans and reservoirs of AMR genes. In that vein, the first use of a QS inhibitor in Aeromonas appears in 2009 when Truchado et al. (2009) found culturing A. hydrophila with chestnut honey resulted in the degradation of AHLs and decreased biofilm production. Since then, many natural and synthetic compounds have been shown to decrease QS-mediated virulence factors including biofilm, motility, protease production, and hemolysis to great effect via QS inhibition in Aeromonas (Table 3). AI-1 QS inhibitor cinnamaldehyde was shown to significantly decrease virulence phenotypes of A. hydrophila (Li et al. 2023). The plant-derived citrus flavonoid, hesperidin methyl chalcone (HMC), was found to not only downregulate the QS gene ahyR but also reduce the overall virulence potential of A. hydrophila in an in vivo fish model (Roshni et al. 2023). Tannic acid has been proven to be an effective QS inhibitor in A. hydrophila with demonstrably lower expression levels of ahyI and ahyR post-treatment and reduced hemolysis, motility, and biofilm formation. Tannic acid treatment also resulted in decreased virulence potential in an in vivo fish model (Patel et al. 2017). Genistein caused the downregulation of ahyRI expression levels, decreased virulence factors like biofilm and aerolysin production, and increased survival in an in vivo fish model (Dong et al. 2021). Another compound, carvacrol, a naturally derived monoterpenoid present in many herbs, was found to decrease the virulence potential of A. hydrophila by inducing decreased biofilm formation, protease production, hemolytic activity, and AHL production. The transcriptional analysis uncovered the downregulation of ahyR with carvacrol treatment in two separate studies, suggesting the involvement of AI-1 QS inhibition (Wang et al. 2022; Lu et al. 2023).

Table 3 A table summarizing studies that have been performed on QS inhibition in Aeromonas to date
Full size table

High-throughput screening for QS inhibitory molecules is becoming increasingly common to discover novel QS inhibitors (Zhang et al. 2022c). In silico methodologies can be used to do this via predicted 3-dimensional structures of the proteins involved. In that vein, the protein structure of AhyI was predicted and functionally characterized. Following that, the AI-1 synthase inhibitor N-cis-octadec-9Z-enoyl-L-homoserine lactone was then identified using high-throughput virtual screening. When tested, this molecule was found to effectively inhibit AI-1 activity at a concentration of 40 mM (Ali et al. 2022). Work has also been carried out to uncover novel AI-2 QS inhibiting compounds. In silico modeling of the AI-2 QS LuxS protein structure facilitated the prediction of putative binders and inhibitors of LuxS. From those predictions, a compound named ( −)-dimethyl 2,3-O-isopropylidene-l-tartrate was chosen for downstream testing, and it was shown to be an effective AI-2 QS inhibitor also at a concentration of 40 μM. Furthermore, A. hydrophila growth was significantly reduced when AI-2 QS inhibitor was added in conjunction with 1 mg/ml of oxytetracycline treatment (Ali et al. 2018). The use of in silico predictive models can more efficiently inform the discovery/design of novel drug candidates, especially when used synergistically with sub-lethal concentrations of bonafide antibiotics.

Alternatively, some research efforts have focused on exploring the role of commensal bacteria in pathogenic QS degradation. For example, one study reported that co-culturing A. hydrophila with three separate fish-gut-derived probiotic bacteria decreased AHL production by A. hydrophila and increased survival in an in vivo tilapia model when challenged (Omar et al. 2023). Similar results are found when a Streptomyces commensal (Liang et al. 2022) and a Bacillus commensal (Chen et al. 2020) were used in a zebrafish model and challenged with A. hydrophila. Because these studies have all focused on aquaculture and fish models, the efficacy of this technique in a mammalian model and the use of human commensals remain unexplored.

The emerging body of literature strongly suggests that blocking QS can be an effective way to reduce Aeromonas-related disease burden in aquaculture. Unfortunately, its potential in humans is left almost entirely unexplored. Given that all in vivo QS inhibition studies to date have been performed in a fish model of infection, a more clinically relevant understanding of many of these QS inhibitors needs to be established. Toward that end, some QS inhibitors have been tested in mammalian cell lines. Resveratrol, while effective in fish, demonstrated cytotoxicity in the murine macrophage cell line J774A.1 at higher concentrations (Qin et al. 2023b). One group observed that the plant extract, sanguinarine, was successful at reducing QS-regulated virulence factors like biofilm production and hemolysis at concentrations of 4 mg/ml. Sanguinarine was found to provide significant protection to human A549 cells from aerolysin-induced cell injury at this same concentration (Zhang et al. 2022b). In fact, thymol, genistein, and cinnamaldehyde have all demonstrated anti-QS activity and reduced cytotoxicity in human A549 cells (Dong et al. 2020a, b; Dong et al. 2021; Li et al. 2023). So, what little evidence we do have of these inhibitors in mammalian cell lines is, at best, varied. To further complicate the situation in a mammalian model, it has been shown that pre-treating mice with QS AI-1 signaling molecule AHL before challenging with A. dhakensis prevents clinical sequelae and produces increased survival in a septicemic model of infection (Khajanchi et al. 2011). While QS inhibition is a demonstrably effective way to reduce pathogenicity in a fish model, more research needs to be performed to ascertain its effectiveness in humans. Furthermore, the vast majority of QS inhibitory studies have been conducted using A. hydrophila as the model pathogen. Studies on this topic need to shift from the discovery of new compounds with anti-QS activity to fully characterizing the known compounds in different Aeromonas spp. and infection models.

Conclusion

Aeromonas spp. are well-established aquatic fish and emerging human pathogens (Fernández-Bravo and Figueras 2020; Hayatgheib et al. 2020). Control of these aquatic pathogens is critical to both protecting aquaculture and its associated economy, as well as to prevent potential human disease. When considering treatment, antibiotic resistance is a major global threat in all bacterial pathogens, and Aeromonas infections are no exception. In fact, antibiotic resistance of various types has been globally documented in Aeromonas spp. (Bargui et al. 2023; Bhaskar et al. 2015; Hayes et al. 1994). Unfortunately, Aeromonas spp. have been shown to acquire resistance from other pathogens as well as readily share resistance with other species/strains (Canellas et al. 2023; Goñi-Urriza et al. 2000; Igbinosa et al. 2015). On account of this, Aeromonas, being aquatic by nature, has rendered aquatic environments including treated waters, significant reservoirs for AMR acquisition and retention (Drk et al. 2023; Rhodes et al. 2000). This is particularly challenging given the potential economic impact of disrupting the aquaculture industry. Furthermore, in serving as potential reservoirs for AMR, these hardy aquatic pathogens can not only cause drug-resistant human diseases but also facilitate the spread of AMR to other unrelated bacterial pathogens. Viewed in this light, AMR Aeromonas could become a major contributor to the problem, setting the stage for nightmarish scenarios associated with the post-antibiotic era.

When facing antimicrobial challenges, Aeromonas spp. employ numerous AMR strategies of evasion including mutations of the drug targets themselves and, in some cases, expression of specific antibiotic degrading enzymes (if their encoding genes are present) (Piccirilli et al. 2022; Rasmussen and Bush 1997; Sinha et al. 2004). Aeromonas spp. also employ a number of highly effective broad-spectrum AMR strategies such as efflux pumps and drug uptake inhibition (Wu et al. 2023; Zhang et al. 2022d). When attempting to acquire a clear picture of AMR pathways involved in antimicrobial evasion, some ambiguity is found on account of pathway crosstalk, synergy, and even, at times, antagonism. Indeed, proteomic analysis reveals the truly complicated nature of Aeromonas’ phenotypic resistance to antibiotic-associated stress which further complicates the issue (Li et al. 2021; Liu et al. 2023). Because there are so many different strategies employed to this effect, a complete understanding of antibiotic resistance in Aeromonas spp. cannot be reached by examining each strategy/pathway independently. As a result, this remains a fruitful area of study.

Aeromonas spp. have also been shown to harbor AI-1, AI-2, and AI-3 QS systems (Khajanchi et al. 2012; Kozlova et al. 2008; Swift et al. 1997). All three systems are essential for the regulation of each other and overall virulence potential (Talagrand-Reboul et al. 2017). Without them, infection cannot be established (Natrah et al. 2012). Because of its essential and ubiquitous nature, QS may be a point of vulnerability to be exploited from a therapeutic perspective. In that vein, many pharmaceutical/natural alternatives to conventional antibiotics/antimicrobials therapies have been proposed specifically targeting the QS systems in Aeromonas spp. to great effect (Li et al. 2023; Patel et al. 2017; Qin et al. 2023b; Roshni et al. 2023; Tan et al. 2019; Wang et al. 2022). The limitation of these therapeutic studies has been their primary focus on aquaculture and fish health. In large part, these studies have employed in vivo fish models with the ecological and economic health of fish culturing in mind, thereby limiting data available on efficacy for use in a human context. Unfortunately, Aeromonas spp. infections continue to pose a challenge to human health, and when considering multiple-drug-resistant Aeromonas spp. infections, treatment alternatives become critical. In fact, the post-antibiotic era has created a pressing need for the development of alternative therapeutics for bacterial infections extending well beyond those caused by Aeromonas spp. alone. Ultimately, additional investigations on the use of these alternative therapeutics in a clinically relevant context are warranted for Aeromonas spp. infections and beyond.

Funding

This work was supported by the National Institutes of Health (NIH) R21 AI135453 grant and the pilot grants from the Institute for Human Infections and Immunity, UTMB, as well as the John S. Dunn chair in Global health Endowment to AKC, and the National Science Foundation (NSF) HRD-1345173 (JAR), HRD-1400962 (JAR), and HRD-1622993 (JAR) awards. The authors declare there are no conflicts of interest.