Prevalence and virulence potential of Aeromonas spp. isolated from human diarrheal samples in North East Italy

ABSTRACT Aeromonas spp. are emerging human pathogens causing intestinal and extra-intestinal infections. Since their relevance in Western Europe as gastrointestinal pathogens is not well established, we investigated Aeromonas spp. prevalence in diarrheal fecal samples in an Italian University Hospital and characterized the virulence mechanisms of the isolates. Aeromonas spp. isolated from diarrheic stools using standard culture methods were identified by molecular techniques. Antimicrobial resistance was assessed by the micro broth dilution. Toxins, flagella, and type III secretion system genes were evaluated by polymerase chain reaction. Biofilm was quantified by crystal-violet staining. Interaction with human intestinal epithelial cells (Caco-2) was assessed by quantifying adhesion, interleukin (IL)-8 secretion, and epithelial barrier integrity. Aeromonas spp. represented 20.6% of bacterial pathogens isolated from diarrheic feces, the second most common enteropathogen. A. cavieae constituted 75% of the identified species, showing a relatively low clustering value. About 52% of Aeromonas isolates showed resistance to amikacin, whereas only 7.5% showed multiple drug resistance. Four or more virulence genes were identified in 66.7% of A. cavieae isolates and 100% of A. dakensis. Aeromonas isolates (82.5%) showed moderate or important biofilm-producing ability. Adhesion to Caco-2 cells correlated to fla+ gene, whereas ascV+ and aexU+ strains significantly induced IL-8 release from Caco-2. Aeromonas aer+ strains caused ZO1 and occluding redistribution and a significative reduction in trans-epithelial resistance. Aeromonas spp. emerge as relevant human intestinal pathogens with a disparate arsenal of pathogenicity factors causing diarrhea through different mechanisms. IMPORTANCE In this work, we demonstrate the epidemiologic relevance of the Aeromonas genus as the cause of infective diarrhea in North East Italy, both in children and adult subjects, with the significative presence of highly pathogenic strains. Aeromonas strains possess a heterogeneous armamentarium of pathogenicity factors that allows the microbe to affect a wide range of human intestinal epithelial cell processes that justify the ability to induce diarrhea through different mechanisms and cause diseases of variable severity, as observed for other gastrointestinal pathogens. However, it remains to be determined whether specific genotype(s) are associated with clinical pictures of different severity to implement the diagnostic and therapeutic approaches for this relevant enteric pathogen.

have been isolated from drinking water supplies and food items such as meat, milk, dairy products, and vegetables (2,3).Granting that Aeromonads are well-recognized disease-causing pathogens in fish and other cold-blooded organisms, these microbes have emerged as relevant human pathogens in the past few years (4).Clinical and experimental studies have demonstrated that Aeromonads, coming into the human body through contact or ingestion of polluted water or food, cause various infections in both immunocompetent and immunocompromised subjects (4,5).Aeromonas spp.have been implicated in extra-intestinal diseases, such as septicemia, soft-tissue wounds, eyes, and respiratory tract infections.However, the most common human infections involve the gastrointestinal tract (5,6).Consistent with the ability of Aeromonads to colonize water supply systems and foods, gastrointestinal diseases caused by these pathogens are relevant in developing countries.Nevertheless, Aeromonas spp.infections in industrial ized countries have been reported too, but the real relevance is not well documented (7,8).
Numerous Aeromonas species can be identified from the environment and foods.However, only a few are considered pathogenic in humans, namely A. hydrophila, A. caviae, and A. veronii bv.sobria (9).Symptoms of Aeromonas-associated gastrointestinal infection vary from diarrhea with loose stools to severe watery or bloody diarrhea with fever and abdominal pain lasting up to 4 weeks (10).The variability in Aeromonas-associ ated gastrointestinal symptoms has been correlated, at least partly, to the substantial variability in putative virulence-protein-encoding genes carried by different species and strains (9,10).These organisms express an assortment of virulence factors, which allow them to colonize, invade, and infect numerous hosts.For instance, in addition to various biologically active surface structures and a repertoire of exoenzymes that digests cellular components, Aeromonas spp.produce several toxins targeting the intestinal mucosa (11).Aeromonas spp.secrete cytotoxic (the heat-labile Act) and cytotonic (the heat-labile Alt and the heat-stable Ast) enterotoxins and various hemolysins (including AerA and HlyA) (11).Indeed, genes encoding different toxins can be concomitantly found, in variable combinations, in the same strain (12,13).The relationship between the occurrence of toxins genes and strains' pathogenicity in animals and humans is unclear since toxigenic strains have been identified in the environment and asymptomatic subjects (9,10).Thus, the pathogenic potential of Aeromonas spp.seems multifactorial and complex and may result from products of different genes acting individually or collectively.
An additional point of concern is the growing incidence of resistance to diverse groups of antibiotics reported in Aeromonads.Several reports, mainly on Aeromonas spp.isolate from fish farms and the environment, have demonstrated a worrying incidence of drug-resistant strains correlated to the extensive use of antibiotics and other chemotherapeutics (14,15).For instance, isolates of Aeromonas have shown relatively high resistance to β-lactam antibiotics, usually correlating with naturally occurring phenotypes of β-lactamases production (16).Furthermore, strains containing multiple antibiotic resistance (MAR) have been isolated from patients' stools and the environment.However, antimicrobial resistance seems to differ between strains isolated from different geographic environments but also compared to those from clinical sources.
In this study, we aimed to determine the incidence of Aeromonas spp. in human diarrheic stool specimens collected at the Microbiology Unit in North Eastern Italy.We characterized the isolated strains regarding their antimicrobial resistance, enterotoxin armamentarium, and pathogenicity mechanisms.

Prevalence of Aeromonas spp. in diarrheal stools
A total of 6,570 consecutive stool specimens collected from patients suffering from diarrhea were examined at the Microbiology Laboratory of Padua University Hospital in 2021.One hundred sixty-two fecal samples resulted positive for bacterial pathogens (Table 1), corresponding to 2.5% of the examined samples.Aeromonas spp.were identified as the causative agent of diarrhea in 40 patients, corresponding to 20.6% of subjects with a diagnosis of bacterial-mediated diarrhea, resulting as the second most common enteropathogen in our series (Table 2).As expected, Campylobacter sp. was the most common enteropathogen identified in patients with diarrhea, whereas Salmonella enterica and Yersinia spp.were less frequent (Table 2).The incidence of stool specimens positive for Aeromonas from patients suffering from diarrhea was comparable between 2021 and 2020, suggesting that the prevalence of this enteropathogen in our area is relatively stable (Tables 2 and 3).
The gender ratio (male:female) was 1.1 (21/19) among patients presenting with diarrhea caused by Aeromonas spp.(Table 3).Aeromonas infections are known to occur in all age groups; however, we observed a peak incidence in children younger than 15 years and adults of 45 years and older (Table 3).
Finally, since the conventional biochemical method (VITEK2 Compact) appeared to provide unsatisfactory identification at the species level, we performed housekeeping gene sequencing with phylogenetic analysis (17,18).Among the isolates recovered from feces of diarrheic patients in 2021, A. caviae was the most prevalent (75%), A. dhakensis represented about 12.5% of isolates, A. media 7.5%, whereas other species such as A. veronii and A. hydrophyla were rarely isolated (Fig. 1).

Phylogenetic analysis of Aeromonas isolates
A multiple sequence alignment was built with the sequences obtained from the PCR of the rpoB housekeeping gene, followed by a phylogenetic analysis (Fig. 2).One reference sequence extracted from GenBank was included for each detected species (accession numbers OQ330882-OQ330921).The tree shows a good clustering among the Aeromo nas spp., although with different levels of bootstrap consistency.A. caviae, the species most frequently detected, shows a relatively low clustering value (56%) (SI Fig. 1), which is affected by the high heterogeneity among sequences.The other Aeromonas spp.show a lower degree of intraspecies diversity, but the small sample sizes of these groups limit further considerations.

Susceptibility to antimicrobials of Aeromonas clinical isolates
Resistance profiles of the Aeromonas isolate against eight antimicrobial agents are shown in Fig. 3.All the bacterial isolates were susceptible to cefepime, cefotaxime, and ceftazidime; however, the susceptibility to other antimicrobial agents varied and 65% of Aeromonas strains showed resistance to one or more antibiotics.Resistance was most prevalent for amikacin, given that 52.5% of strains were unresponsive.Only 5% and 12.5% of Aeromonas were resistant to ciprofloxacin or trimethoprim-sulfamethoxazole, respectively.Resistance to meropenem, a carbapenem antibiotic, was barely detectable.MAR patterns were reported in 7.5% of the strains.
To evaluate the presence of the type III secretion system (T3SS) such as AexU and AexT (13), which injects toxins into target cells, we investigated the presence of the ascV gene that encodes for an inner membrane component of the T3SS apparatus.AscV gene was detected in 47.5% (19/40) of Aeromonas clinical isolates (Fig. 3B).
Since the flagella in Aeromonas spp.are considered necessary to adhere to biotic or abiotic substrates to form a biofilm, we investigated the presence of the laf and fla genes encoding lateral and polar flagella, respectively (19,20).As reported in Table 5, 13/40 strains carried laf gene, whereas 21/40 possessed the fla gene (Fig. 3B).

Biofilm formation by Aeromonas clinical isolates
Aeromonas spp.can adhere to biotic or abiotic surfaces and form biofilms, a key virulence factor (21).Therefore, we determined the ability of Aeromonas clinical isolates to produce biofilm in static conditions.Although all strains could form biofilms, they had distinct productivities (Fig. 5A).The strains were classified as weak, moderate, or strong biofilm producers (Fig. 5A).Of the Aeromonas strains isolated from diarrheic feces, only 4 (10%) were strong biofilm producers, whereas 29/40 (72.5%) resulted in moderate biofilm producers (Table 6).Interestingly, the presence of fla gene improved biofilm formation (Fig. 5B).

Attachment and internalization of Aeromonas clinical isolates to Caco-2 cells
To evaluate how Aeromonas strains interact with human intestinal epithelial cells, we first evaluated attachment and internalization to Caco-2 monolayers.Among the tested Aeromonas strains, only one presented an excellent attachment ability since almost 10 CFU/cell adhered to Caco-2 cells; 18/40 strains presented an adhesion ability lower than 1 CFU/cell (Fig. 6A). A. caviae strains exhibit an average adhesion of 1.5 CFU/cell.Nine isolates showed a high adhesion efficiency (>2 CFU/cell), whereas 13 strains showed low adhesion abilities (<1.0 CFU/cell) (Fig. 6A).Other Aeromonas spp.exhibited a similar  ability to adhere to Caco-2 (Fig. 6A).Interestingly, we observed that the adhesion ability to human intestinal epithelial cells increases with the presence of fla gene that encodes for polar flagella (Fig. 6C); the presence or absence of gene coding lateral flagella (laf) does not influence the ability of clinical isolates to adhere to Caco-2 cells (Fig. 6D).
To assess the internalization ability of Aeromonas spp.into intestinal epithelial cells, we incubated Caco-2 cells exposed for 2 h to the bacteria with gentamicin to kill bacteria attached to the cells.The results summarized in Fig. 6B indicate that Aeromonas strains barely invade epithelial cells, suggesting that the adhesion to intestinal epithelium is the dominant phenomenon for this pathogen.

Aeromonas clinical isolates stimulate inflammatory cytokines release from Caco-2
To determine whether the interaction of Aeromonas with intestinal epithelial cells triggers an inflammatory phenotype, we quantified IL-8 released by Caco-2 cells in response to Aeromonas strains isolated from patients with diarrhea.Among A. caviae, we again observed a dichotomic aptitude since 7 strains strongly induced IL-8 secretion, but 19 strains did not induce a pronounced cytokine release, in line with other Aeromonas spp.(Fig. 7A).
Aeromonas spp.-inducedIL-8 release directly correlated with adhesion ability, suggesting that the direct interaction with intestinal epithelial cells is required to trigger inflammatory cytokines release (Fig. 7B).Moreover, the presence of the genes ascV and aexU stimulates Caco-2 cells in producing IL-8 (Fig. 7C and D).

Aeromonas clinical isolates exert cytotoxic effects on Vero cells in vitro
To measure the ability of Aeromonas strains to damage the epithelial cells directly, we determined the release of cytotoxic products.To this goal, we added Aeromonas cultureconditioned media to Vero cell monolayers.A. caviae strains reported a low cytotoxic effect since 21/30 isolates showed less than 10% cytotoxicity, whereas 9/30 exhibited a moderate cytotoxic effect ranging from 10 to 30% (Fig. 8).The other Aeromonas spp.showed an average cytotoxicity activity higher than the A. caviae group, even if the cytotoxic activity of the isolates was highly variable (Fig. 8).

AerA+ Aeromonas strains damage the epithelial barrier
To assess the impact of Aeromonas strains on the integrity of the intestinal epithelial barrier, we measured transepithelial electrical resistance (TEER) in polarized Caco-2 monolayers grown in transwell apparatus (22).Polarized Caco-2 cells were apically exposed to different clinical isolates, either aerA+ or aerA−, or to enteropathogens known to disrupt epithelial barrier integrity, such as enteropathogenic Escherichia coli or S. enterica.In these experiments, we tested only aerA− A. caviae since all A. dhakensis, A. hydrophila, and A. veronii resulted aerA+ (Table 4).All Aeromonas aerA+ caused a significant decrease in TEER 5 h post-exposure that persisted up to 24 h (Fig. 9), compara ble to enteroinvasive E. coli and S. enterica.Indeed, A. dhakensis induced a more pro nounced decrease in TEER than A. caviae aerA+ (Fig. 6).On the contrary, monolayers apically exposed to Aeromonas aerA− showed a modest transient decrease in TEER following 5 h incubation and exhibited resistances comparable to controls following 24 h (Fig. 9A and B).Therefore, Aeromonas spp.-induced damage of epithelial monolayer strongly increases with the presence of aerA gene, suggesting that the toxin aerolysin A damages the inter-epithelial cell junctions adhesion apparatus (Fig. 9C).
To investigate whether Aeromonas-induced intestinal barrier disruption is associated with changes in TJ protein expression or subcellular distribution, we performed an immunofluorescence assay evaluating ZO-1 and occludin in Caco-2 monolayers exposed to aerA− or aerA+ strains.Following exposure to Aer+ strains, we observed substantial disorganization in ZO-1 and occludin that appeared redistributed in microdomains in Caco-2 cells (Fig. 9D).On the contrary, we did not observe any change in ZO-1 and occludin immunoreactivity in Caco-2 monolayers exposed to A.caviae22, an Aer− strain (Fig. 9D).

DISCUSSION
In diagnostic procedures, many coprocultures performed on diarrheic stool samples fail to identify a specific pathogen causing diarrhea (23).Even more sensitive techniques, such as antigen or nucleic acid detections in addition to coproculture, do not significantly improve the failure rate to discover a specific pathogen (23).The negative results can have various explanations, such as not infectious diarrhea and the inadequate handling and storage of samples.However, the causative agent is usually not investigated as in the cases of viral infection or not yet identified enteropathogens.Therefore, in the United States, most diarrhea cases remain with unknown etiology (24,25).Actually, several new enteric pathogens have been proposed in the past few years and Aeromonas spp. is among the newly identified human pathogens since this bacterial genus has been widely accepted as the etiological agent of infectious diarrhea (26,27).
We undertook a prospective study at the Microbiology Unit of Padua University Hospital in Northeast Italy to define the frequency of identification of Aeromonas spp. in diarrheal disorders in our region and characterize the phenotype of the isolates.The factual incidence of Aeromonas in human gastrointestinal infections worldwide is unknown.However, it is well-accepted that it varies by geographical location and is higher in regions with low hygiene standards (10,26,27).In several studies worldwide, Aeromonas spp.have been isolated at a rate of 0.6-7.2% in patients with diarrhea, predominantly in infants and children (28,29).In European countries, North America, and Israel, the incidence is estimated at 2% in patients with traveler's diarrhea (28,30,31).
In contrast, Aeromonas spp.infection is the third cause of bacterial gastroenteritis in Spanish children, accounting for 2.5-4% of diagnoses (32)(33)(34).Similarly, in the USA, the incidence has been reported in 2.5% of children with diarrhea (34).In this study, 0.61% of diarrheic stools of patients were positive for Aeromonas spp., in agreement with previous reports and with the well-known notion that in industrialized countries, Campylobacter spp.(1.32% in our series) are the prevalent bacterial pathogens (35).Interestingly, in our series Aeromonas spp.were the second most common enteropathogen isolated from feces, being more numerous than S. enterica (0.46%).Therefore, Aeromonas spp.need to be considered as relevant human enteropathogens, like Campylobacter and Salmonella spp.The detection based on classical coproculture methods without selective medium could underestimate the actual incidence of this pathogen, even in "industrialized" countries (36).In our study, the Aeromonas spp.infection was monomicrobial in most cases (93%).In contrast, in previous studies evaluating traveler's or children's diarrhea, multiple pathogens were detected in a higher percentage of cases (31,32,37), probably due to differences in the study population.Our survey is free of the bias of patient recruitment and reveals a bimodal case distribution, with a peak in children younger than 15 years and a second large peak in subjects older than 45.According to previous reports, our study reported that 30% of the patients were infants and children (31).
In the current study, Aeromonas spp.were identified using the rpoB gene sequencing method, a procedure considered much more accurate than biochemical identification methods or 16S rRNA gene sequencing (17,37,38).In agreement with recent studies, A. caviae was recognized as humans' predominating species associated with diarrhea (32,39).In contrast with previous reports, reporting A. veronii or A. hydrophyla as the second largest detected group (9,31,39); in our study, A. dhakensis represented the second most frequent group of Aeromonas strains isolated from diarrheic stool specimens.Overall, our data support the need for molecular tests to identify Aeromonas species to perform accurate epidemiological studies.The low clustering value observed among A. caviae isolates suggests different sources of infection (i.e., contaminated food) for each patient and seems to exclude clusters of infection or direct inter-human transmission.
Aeromonas spp.are intrinsically susceptible to most antibiotics active against non-fastidious Gram-negative bacilli, except for many beta-lactams due to chromoso mally encoded β-lactamases (31).In line with previous studies, all strains were suscep tible to third-generation cephalosporin antibiotics (cefotaxime and ceftazidime), and only two strains (5%) were resistant to the second-generation fluoroquinolone antibi otic, ciprofloxacin (31,40,41).Surprisingly, the majority of strains (52.5%) were resist ant to aminoglycoside antibiotics (i.e., amikacin), in contrast to the high susceptibility previously reported for human isolates in Israel, China, and Malaysia (32,42,43).Since aminoglycoside-resistant strains have been reported mainly from aquatic sources, our data suggest that patients in our area might be exposed principally to Aeromonas strains originating from fish of aquaculture (42,44).Recently, several reports described carbapenemase genes in this group of bacteria (45,46).Carbapenemase-producing A. hydrophila has been identified by routine perirectal surveillance culture and in clinical isolates from various sources, including stools and polluted water (42,43).However, we have not identified carbapenemase-producing Aeromonas, suggesting that these pathogens are not contributing to the spread of these genes in our area.
The mechanism of Aeromonas pathogenesis in the gut is complex and no single putative virulence-associated factor can be unequivocally pinpointed as responsible for GI diseases (44).Most studies characterizing virulence factors associated with Aeromonas pathogenicity have been performed in strains isolated from environmental sources such as polluted and drinking water (44).However, regardless of their origin, Aeromonas strains possess an impressive variety of virulence factors (44).In addition to structural components of the bacterial cell (i.e., LPS, pili, and flagella), Aeromonas strains can produce a variety of extracellular enzymes and toxins; as a result, the toxigenic pro file is extremely mutable (1,47,48).In agreement with previous studies, also in the human isolates retrieved in North Eastern Italy, we observed high heterogeneity in the distribution of toxin genes among the isolates (31,44,49).Thus, our clinical isolates harboured at least one of the putative virulence genes ast, alt, aer, act, hlyA, ela, and aexU.The percentages of positivity for virulence genes obtained in our study were in line with food and environmental strains, as well as clinical isolates previously described (31,44,47).Interestingly, we observed a strong presence of virulence genes in A. dhakensis according to a previous study which pointed out the importance of this species due to a large number of virulence genes but also to its high rates of drug resistance and ability to cause both intestinal and extra-intestinal infections (39).
Aeromonas is characterized by the ability to produce biofilms on the biotic or abiotic surfaces that plays a crucial role in the colonization process and amplify antibiotic resistance.In agreement with previous reports, the present study demonstrates that clinical isolates had different abilities to form biofilms, even under the same experimental conditions (19).Moreover, biofilm formation depends on the specific strain and is not a general characteristic of a bacterial species or serotype (50).Moreover, by comparing biofilm biomass by diverse Aeromonas spp.for 48 h, our data support the view that functional polar flagella seem required for optimal adherence and biofilm formation, in line with most of the published studies (21,51).
To study the interaction of Aeromonas clinical isolates with gut epithelial cells, we used the Caco-2 cells model, widely used to investigate pathogens' effects on GI mucosa, including Aeromonas spp.(9,52,53).Adhesion and invasion of bacteria to mucosal surfaces are a critical step in the pathogenesis of most GI infections.The adherence capacity of Aeromonas spp. to several human cell lines has been previously reported (54,55).Compared to previous studies reporting adherence between 3.2 and 6.5 bacteria per epithelial cell of strains isolated from the polluted waterways, the Aeromonas isolates described in this study and obtained from diarrheic feces showed a slightly less adhering ability (8,19,52).In contrast with a previous study, we observed that the expression of the gene encoding for polar flagellum (fla), rather than for the gene encoding for lateral flagella (laf), correlates with the adhesion and colonization ability (19).Our clinical isolates, in agreement with previous reports on strains isolated from the environment and food, showed a low index of invasion (56,57).The limited invasive ability of Aeromonas strains observed in this and previous studies suggests that the pathogenetic mechanism of Aeromonas spp.differs from that of "classical" invasive enteropathogens such as E. coli, Shigella, or Salmonella spp (53).
Although tissue-dependent responses have been reported, it is well accepted that Aeromonas spp.enhance several proinflammatory cytokine overexpression in animal models (9).In line with the reported ability of Aeromonas spp. to promote the activation of transcription factors and secretion of proinflammatory cytokines, our clinical isolates stimulate IL-8 release from Caco-2 cells (9).Intriguingly, Aeromonas-induced cytokine release required direct epithelial cell-pathogen contact, directly correlating to the extent of adhesion to epithelial cells.
Since our results point out that Aeromonas spp.inducing IL-8 release require a type III secretion system or the association with toxin AexU, they support the view that Aeromonas strains modulate inflammatory responses in intestinal epithelial cells by directly inoculating specific toxins (31,58).Moreover, the amplitude of cytokine release was strain-dependent suggesting that the variable clinical presentation of Aeromonas GI infection might be determined by to different extent of proinflammatory cytokines secreted by the epithelial cells that amplify mucosal damage, such as in other infective diseases (53,59,60).However, further studies are required to correlate the severity of the clinical picture and the virulent armamentarium of Aeromonas spp.and to dissect the innate immune mechanisms engaged by the gastrointestinal system to detect and clear invading Aeromonas.
Impairment of TJs function, caused by gastrointestinal pathogens, is implicated in the pathology of gastrointestinal infections (61).In agreement with previous studies, we demonstrated that the aerolysin is the main effector of Aeromonas-induced barrier impairment, as bacteria strains lacking this gene failed to affect TJs function and integrity (62).Direct exposure of Caco-2 monolayers to Aer+ strains induced a fast drop in TEER along with TJ proteins redistribution in raffles caused by differential claudin/ZO-1 interactions (63).Indeed, aerolysin induces TJ protein redistribution via Ca 2 + signaling thus producing actomyosin contraction, which in turn causes retraction and redistribution of TJ proteins forming membrane ruffles (62). A. dhakensis strains caused a more prominent and persistent decrease in transepithelial resistance than A. caviae Aer+.Intriguingly, A. dhakensis strains carried, in addition to Aer gene, the hlyA gene coding a toxin that, in inflammatory or inflammation-prone conditions, potentiates the leaky gut phenomenon, showing barrier-breaking effects (64).Thus, we support the view that various toxins contribute to damage to the epithelial barrier, eventually increasing the severity of clinical pictures (65).
In conclusion, this work demonstrates the epidemiologic relevance of the Aeromo nas genus as the cause of infective diarrhea in North East Italy, both in children and adult subjects, with the significative presence of highly pathogenic strains.Phylogenetic analysis and antibiotic resistance pattern suggest that the major source of Aeromonas strains causing diarrhea in North Eastern Italy might derive from aquaculture fish, posing the need for more strict surveillance to improve food safety standards.Aeromonas strains possess a heterogeneous armamentarium of pathogenicity factors that allows the microbe to affect a wide range of human intestinal epithelial cells processes that justify the possibility of inducing diarrhea through different mechanisms and cause diseases of variable severity, as observed for other gastrointestinal pathogens (66).However, it remains to be determined whether specific genotype(s) are associated with clinical pictures of different severity (i.e., bloody or watery diarrhea) to implement the diagnostic approaches (i.e., search of specific genes in isolates) and therapeutic approaches for this relevant enteric pathogen.

Aeromonas prevalence in diarrheal fecal samples
The isolation and identification of Aeromonas spp.were performed using standard microbiological procedures in diarrheic stool specimens submitted to the Microbiology Laboratory of Padua University Hospital, which provides services to a vast population in the Padua metropolitan area (North Eastern Italy) (67).Samples were collected between January 2021 and December 2021.All specimens were examined routinely for conven tional enteropathogens, namely Shigella, Salmonella, Yersinia, and Campylobacter spp., identified by established methods (67).For the isolation of Aeromonas spp., the fecal specimens were diluted (1:10), and a loop of material was streaked on selective Shigella Aeromonas agar.The agar plates were incubated for 48 h at 25°C.Colonies morpholog ically suspected as Aeromonas were identified at the genus level using an automatic bacteriologic analyzer (VITEK2 Compact, BioMerieuX, France).Table 3 summarizes the gender and age of the patients and the month in which each strain was isolated.Strains were stored in Luria-Bertani (LB, Becton Dickinson) broth and glycerol mixture (70:30) at −80°C until genotypic and phenotypic tests were performed.For subsequent tests, Aeromonas strains were grown in LB broth at 37°C.
Aeromonas species were determined by amplifying and sequencing the RNA polymerase β subunit gene (rpoB), a housekeeping gene, following a previously published method (38).To amplify the sequence, we used the following primers Fw-G CAGTGAAAGARTTCTTTGGTTC and Rv-GTTGCATGTTNGNACCCAT.The PCR products were sequenced by Eurofins Genomics (Germany).Newly obtained sequences were compared to those available in the GenBank database, using the standard nucleotide-nucleo tide BLAST program (BLASTN; http://www.ncbi.nlm.nih.gov) to establish their closest relatives.The sequences were submitted to the GenBank database under accession numbers OQ330882-OQ330921.A phylogenetic tree was generated using the Maximum Likelihood method with MEGA 11 (https://doi.org/10.1093/molbev/msab120)based on alignments from CLUSTAL W (10.1093/nar/22.22.4673).

DNA extraction and PCR amplification
Strains kept at −80°C were streaked on selective Shigella Aeromonas agar, and then an isolated colony was inoculated in 10 mL of LB and grown overnight at 37°C with agitation (150 rpm/min).The culture was centrifuged (3,800 rpm for 10 min), and bacteria were incubated in 500 µL of Lysis Buffer (NaCl 10 mM, MgCl 2 3 mM, Tris-HCl 20 mM pH 7.4, 0.3% Nonidet P40, and 1.25% sucrose), 62.5 µL of SDS 10% and 20 µL of proteinase K 200 µg/mL for 1 h at 56°C with frequent vortexing.Total chromosomal DNA from Aeromonas spp. was purified using standard phenol/chloroform methods.The DNA was suspended in DNAse and RNAse-free water, quantified and subjected to PCR using primers and conditions described in Table 7. PCR products were electrophoresed in 2% wt/vol agarose gel and visualized using Gel Doc EZ System Bio-Rad.

Biofilm quantification
Aeromonas strains were grown overnight in LB at 37°C with agitation (150 rpm/min), and then cultures were diluted 1:100 in fresh LB medium and incubated for additional 2 h at 37°C.The cultures were centrifuged and suspended in Brain Heart Infusion broth (BHI) at 10 8 CFU/mL.About 200 µL of bacterial suspensions was placed into 96-well polystyrene microtiter plates and incubated at 37°C in an aerobic environment under static conditions.The biofilm quantification was conducted utilizing the crystal violet method with some modifications (69).Following 48 h of incubation, the wells were emptied and washed three times with sterile phosphate-buffered saline (PBS) to remove planktonic cells.Adhering cells were fixed with ethanol, and the biofilm was incubated with crystal violet 0.1% wt/vol in the dark at 37°C.After 30 min, the excess of crystal violet was removed, and the plate was left to dry for 24 h at 37°C.Finally, 100 µL of acetic acid 30% vol/vol was added to solubilize the dye.Optical density (OD) at 570 nm was recorded as a measure of biofilm biomass using a microplate reader (Perkin Elmer Victor X2 Multilabel Microplate Reader).

Adhesion and invasion of Caco-2 monolayers
The adhesion of Aeromonas strains to Caco-2 monolayers was assessed as previously described (52).Confluent Caco-2 monolayers in 24 wells/plate were washed two times with warm HBSS and then incubated with 500 µL of DMEM w/o antibiotics.Aeromo nas strains cultured in LB-broth overnight at 37°C with agitation (150 rpm/min) were collected by centrifugation and adjusted to a concentration of 10 8 CFU/L in DMEM w/o antibiotics.Then bacteria were added to Caco-2 monolayers (MOI 1:20) and incubated at 37°C in a humidified atmosphere containing 5% CO 2 , with slight shaking to promote cell/bacteria interaction.After 1 h, non-adhering Aeromonas were removed by extensive washes with warm HBSS.Epithelial cells were then detached, collected, and lysed to quantify the bacteria adhering to Caco-2 monolayers.Living bacteria were enumerated by quantitative bacterial vital count assay.Serial dilutions of lysed epithelial cells were seeded on LB agar plates and incubated at 37°C.After 16 h, the CFU of bacteria per cell was calculated.To quantify the bacteria invading Caco-2 monolayers, following extensive washes with HBSS, the cells were incubated in DMEM containing gentamycin.After 2 h, living bacteria associated with Caco-2 cells were enumerated by quantitative bacterial vital count assay as described above.

Cytokine release
Aeromonas strains-induced cytokine release from Caco-2 monolayers was performed as previously described for enteropathogens (70).Aeromonas strains cultured in LB-broth overnight at 37°C with agitation (150 rpm/min) were centrifuged, adjusted to a concentration of 10 8 CFU/mL and added (MOI 1:20) to confluent Caco-2 monolay ers.Monolayers and bacteria were co-incubated at 37°C in a humidified atmosphere containing 5% CO 2 .After 2 h, the culture media was removed, monolayers were washed two times with warm HBSS, and then cells were incubated with DMEM containing gentamycin (50 µg/mL).Conditioned media were collected after additional 16 h and stored at −20°C until IL-8 was measured by a commercially available ELISA kit (eBioscien zeTM).

Effect of Aeromonas clinical isolates on transepithelial electrical resistance
Caco-2 cells were seeded on Transwell polyester membrane cell culture inserts (trans parent PET membrane: 1.0 cm 2 growth surface area, 0.4 µm pore size; BD Falcon) in 24-well plates and incubated with DMEM medium w/o antibiotics.The culture medium was replaced every 48 h until the confluence.As previously described, the TEER values were measured daily in HBSS (22).Monolayers were used when TEER values remained stable for three consecutive days.Then, Aeromonas clinical isolates cultured overnight at 37°C in LB were placed (MOI 1:20) in the upper chamber of Transwell inserts and left in contact with monolayers.After 2 h, the culture medium was replaced with fresh DMEM containing gentamycin (50 µg/mL), and Transwells were incubated at 37°C.TEER values were measured after 5 and 24 h.TEER values of Caco-2 monolayers exposed to bacteria were expressed as a percentage of monolayer resistance not exposed to bacteria.

Evaluation of Aeromonas spp. effects on tight junction proteins
Caco-2 cells were seeded on glass coverslips in six-well plates and kept in culture for 7-10 days after reaching confluence.Then, cells were washed and incubated with Aeromonas strains (MOI 1:20) in DMEM without antibiotics.The culture medium was removed after 2 h and substituted with fresh DMEM supplemented with FBS and gentamycin to eliminate residual Aeromonas.After 5 or 24 h, epithelial cells were fixed in 4% wt/vol paraformaldehyde (PFA) for 15 min.Cells were washed three times with PBS 1× and incubated with anti-zonula occludens (ZO)-1 or anti-occludin polyclonal antibody (Invitrogen) in PBS/0.2%Triton X100/2% BSA (PBS-T).After 1 h, cells were washed three times in PBS-T (10 min each), and monolayers were incubated with proper secondary antibody FITC labeled.After 30 min, following extensive washing, in PBS/0.2%Triton X-100 (3 × 10 min).Samples were visualized with a Nikon A1RSi laser scanning inverted confocal microscope equipped with NIS-elements advanced research software using 40× ocular objectives.Microscope settings were established to collect images below saturation and were kept constant for all image acquisition.

Cytotoxic effect of Aeromonas clinical isolates on Vero cells
Aeromonas isolates (10 8 CFU/mL) were grown overnight in LB-broth at 37°C with agitation (150 rpm/min).The culture was centrifuged and supernatant sterile filtered using a 0.22-µm syringe filter.Cell-free supernatant was added to confluent Vero monolayers at the ratio of 1:10 vol/vol.Cell cytotoxicity was determined by crystal violet staining following previously reported methods (71).After 16 h of incubation with the supernatants, cells were fixed in PFA 4% for at least 20 min.Cells were then washed with PBS 1× three times.Plates were stained with a 0.4% crystal violet solution in methanol for 30 min.After three additional washing, stained cells were solubilized with 30% acetic acid.Absorbance at 590 nm was measured using plate reader (Varioskan Lux Reader, Thermo Fisher Scientific).No treated wells were arbitrarily assigned 0% of cytotoxicity.All tests were performed in triplicate.

Statistical analysis
The data were analyzed using one-way analysis of variance, followed by Bonferroni multiple comparisons using GraphPad Prism (version 8.0) in Fig. 9A-B.The correlation diagram and regression line were conducted using GraphPad by calculating Pearson R 2 in Fig. 7C.In Fig. 5B-6C/D-7B/D-9C, were used t test analysis to compare two different groups.P values of ≤0.05 were taken as statistically significant.

FIG 4
FIG 4 Distribution of virulence genes by species in Aeromonas strains isolated from patients with diarrhea.DNA purified from each strain was subjected to PCR for the specified genes; amplicons were visualized using Gel Doc EZ System.(A) Enumeration of presence or absence in each Aeromonas isolates of genes involved in the production of cytolytic toxins-act, aer, hlyA, ela, cytotonic toxins-ast, alt, ADP-ribosylating toxin-aexU.(B) Evaluation of presence/absence of genes coding for a component of type III secretion systems-ascV, lateral flagella-laf, and polar flagella-fla.(C) Illustration of numbers of toxins genes in each isolate and their representation of distributions among different species.

FIG 6
FIG 6 Adherence and invasion on Caco-2 cells of Aeromonas strains isolated from patients with diarrhea.Caco-2 monolayers were incubated with Aeromonas isolates (MOI 1:20) for 2 h at 37°C.Bacteria non-adhering to Caco-2 were removed by extensive washing, and (A) cells were lysed; aliquots were seeded on chocolate agar plates; adhering microbes were quantified by quantitative bacterial vital count assay.(B) Cells cultured for additional 2 h in complete medium containing gentamycin were lysed and invading microbes were quantified by quantitative bacterial vital count assay.Adhering or invading microbes were expressed as CFU/cell.All tests were performed in triplicate.(C and D) The influence of fla (C) or laf (D) genes on adhesion ability of Aeromonas isolates.*P value < 0.05.

FIG 7
FIG 7 IL-8 in Caco-2 cells cultured with or without Aeromonas strains isolated from patients with diarrhea.(A) Caco-2 monolayers were incubated with Aeromonas isolates (MOI 1:20).After 2 h at 37°C culture medium was removed, the cells were washed and cultured in complete medium.Culture media were collected after 24 h.Levels of interleukin-8 (IL-8) were measured using commercially available enzyme-linked immunosorbent assay (ELISA) kits and expressed as pg/mL.All tests were performed in triplicate.(B) Correlation analysis between adhesion to Caco-2 cells and IL-8 release for each Aeromonas isolate R 2 = 0.1804; **P < 0.01.(C and D) Effect of ascV (C) or aexU (D) genes presence (ascV/U+) or absence (ascV/U−) on IL-8 induced production in Caco-2.*P < 0.05.

FIG 8
FIG 8Cytotoxic effects of Aeromonas strains isolated from patients with diarrhea.Vero cells monolayers were incubated for 16 h with cell-free supernatant in a ratio of 1:10 vol/vol.Cell cytotoxicity was determined by crystal violet staining and measuring absorbance at 590 nm.The 0% of cytotoxicity was arbitrarily assigned to not-treated wells.All tests were performed in triplicate.

TABLE 1
Overall picture of coproculture at Microbiology Laboratory of Padua University Hospital

TABLE 2
Enteropathogens isolated in diarrheal fecal samples

TABLE 3
Characteristics of patients with Aeromonas-positive coproculture

TABLE 4
Virulence genes in Aeromonas clinical isolates

TABLE 5
Distribution of virulence genes among Aeromonas spp

TABLE 6
Biofilm production by different Aeromonas isolates

TABLE 7
Primers and amplification conditions for PCR