Prevalence of Potentially Pathogenic Antibiotic-Resistant Aeromonas spp. in Treated Urban Wastewater Effluents versus Recipient Riverine Populations: a 3-Year Comparative Study

Aeromonads are Gram-negative, asporogenous rod-shaped bacteria that are autochthonous in fresh and brackish waters. Their pathogenic nature in poikilotherms and mammals, including humans, pose serious environmental and public health concerns especially with rising levels of antibiotic resistance. Wastewater treatment facilities serve as major reservoirs for the dissemination of antibiotic resistance genes (ARGs) and resistant bacterial populations and are, thus, a potential major contributor to resistant populations in aquatic ecosystems. However, few longitudinal studies exist analyzing resistance among human wastewater effluents and their recipient aquatic environments. In this study, considering their ubiquitous nature in aquatic environments, we used Aeromonas spp. as bacterial indicators of environmental antimicrobial resistance, comparing it to that in postchlorinated wastewater effluents over 3 years. Furthermore, we assessed the potential of these resistant populations to be pathogenic, thus elaborating on their potential public health threat.

ABSTRACT

Antibiotic resistance continues to be an emerging threat both in clinical and environmental settings. Among the many causes, the impact of postchlorinated human wastewater on antibiotic resistance has not been well studied. Our study compared antibiotic susceptibility among Aeromonas spp. in postchlorinated effluents to that of the recipient riverine populations for three consecutive years against 12 antibiotics. Aeromonas veronii and Aeromonas hydrophila predominated among both aquatic environments, although greater species diversity was evident in treated wastewater. Overall, treated wastewater contained a higher prevalence of nalidixic acid-, trimethoprim-sulfamethoxazole (SXT)-, and tetracycline-resistant isolates, as well as multidrug-resistant (MDR) isolates compared to upstream surface water. After selecting for tetracycline-resistant strains, 34.8% of wastewater isolates compared to 8.3% of surface water isolates were multidrug resistant, with nalidixic acid, streptomycin, and SXT being the most common. Among tetracycline-resistant isolates, efflux pump genes tetE and tetA were the most prevalent, though stronger resistance correlated with tetA. Over 50% of river and treated wastewater isolates exhibited cytotoxicity that was significantly correlated with serine protease activity, suggesting many MDR strains from effluent have the potential to be pathogenic. These findings highlight that conventionally treated wastewater remains a reservoir of resistant, potentially pathogenic bacterial populations being introduced into aquatic systems that could pose a threat to both the environment and public health.

IMPORTANCE Aeromonads are Gram-negative, asporogenous rod-shaped bacteria that are autochthonous in fresh and brackish waters. Their pathogenic nature in poikilotherms and mammals, including humans, pose serious environmental and public health concerns especially with rising levels of antibiotic resistance. Wastewater treatment facilities serve as major reservoirs for the dissemination of antibiotic resistance genes (ARGs) and resistant bacterial populations and are, thus, a potential major contributor to resistant populations in aquatic ecosystems. However, few longitudinal studies exist analyzing resistance among human wastewater effluents and their recipient aquatic environments. In this study, considering their ubiquitous nature in aquatic environments, we used Aeromonas spp. as bacterial indicators of environmental antimicrobial resistance, comparing it to that in postchlorinated wastewater effluents over 3 years. Furthermore, we assessed the potential of these resistant populations to be pathogenic, thus elaborating on their potential public health threat.

INTRODUCTION

A global health crisis is arising due to the lack of new antibiotic development combined with increasing antibiotic resistance. Economically, each case of resistance costs more than 20,000 dollars, impacting the global gross domestic product with a loss of over 3 trillion dollars (1). In less than 100 years since the discovery of antibiotics, the modern antibiotic era is quickly fading due to overuse and misuse in patient and veterinary clinics, as well as in agriculture and aquaculture. One of the main reservoirs of antibiotic-resistant bacteria (ARB) and antibiotic resistance genes (ARGs) is wastewater treatment plants (WWTPs) (2). The high concentrations of bacteria coexisting with subinhibitory concentrations of antibiotics throughout the wastewater treatment process (2) provide optimal conditions enabling the development of resistance as well as persister cells, thus serving as vehicles of ARG dissemination in the environment (3, 4). Although urban WWTPs provide several log reductions in bacterial populations in comparison to influents, treated effluents remain rich sources of multidrug-resistant (MDR) bacterial populations entering the recipient aquatic ecosystem. One predominant bacterial genus in treated wastewater is Aeromonas (5), which exhibits moderate resilience to chlorine treatment (6).

The genus Aeromonas is comprised of Gram-negative, oxidase-positive, glucose-fermenting, facultative anaerobic bacilli. They are widespread in nature, populating fresh and brackish water habitats as well as wastewater, sewage, and drinking and agricultural waters (7). Aeromonas spp. cause a wide spectrum of diseases in cold- and warm-blooded animals ranging from fish to humans (8) with varying severities. The majority of Aeromonas-associated human illnesses involve gastroenteritis as well as soft tissue and wound infections, with immunocompromised and younger populations being the most affected and having increased disease severity (9). Of the over 30 species known (10), human diseases are primarily caused by four species, Aeromonas hydrophila, Aeromonas veronii, Aeromonas caviae, and Aeromonas dhakensis (11). Due to their high presence in aquatic environments and various antibiotic-resistant mechanisms, this genus could identify as indicator bacteria for aquatic environmental antibiotic-resistant dissemination (12). Health concerns related to drug-resistant aeromonad populations are 2-fold: Aeromonas has a pathogenic nature makes it more difficult to treat clinically (11, 13–15), and it has the potential to act as a reservoir of ARGs capable of transferring to other pathogenic bacteria through horizontal gene transfer (16, 17).

Khajanchi et al. linked biochemical analysis, pulse-field gel electrophoresis XbaI molecular fingerprints, and prevalence of virulence factors between environmental and clinical isolates supporting water-to-human transmission (18). This finding stresses the urgent need for surveillance studies to monitor antibiotic resistance among Aeromonas spp. and their potential to be pathogenic within natural aquatic reservoirs, as well as their anthropogenic contaminating sources. However, longitudinal studies are essential for surveillance studies to implicate resistance trends and assess the need for remedial countermeasures (19). To our knowledge, no longitudinal studies exist using Aeromonas spp. as indicator bacteria of environmental resistance in surface water. Thus, in this study, we (i) provided a longitudinal study using Aeromonas spp. as indicator bacteria assessing antibiotic susceptibility from postchlorinated wastewater effluents to recipient surface water over three consecutive summers, (ii) assessed their potential to be pathogenic, (iii) identified the predominant Aeromonas spp. in each aquatic reservoir and their correlation with resistance and cytotoxicity to further assess health risk, (iv) identified ARGs associated with resistant phenotypes, and (v) determined association of serine protease with cytotoxicity.

RESULTS

Prevalence of antibiotic resistance within surface river water and treated wastewater effluents.

Aeromonas spp. are ubiquitous in stormwater and sewage (20) as well as lakes (21) and rivers (22); therefore, their impact on the aquatic environmental antibiotic resistome is a serious concern. Our study analyzed bacterial populations in the Rock River upstream of an urban wastewater treatment facility. The Rock River is 285 mi long initiating in Wisconsin and flowing through Illinois where it becomes a major tributary to the Mississippi River. With 85.9% of the 16,786 square kilometers of the Rock River basin being used for agriculture (23) and 165 million liters/day of human wastewater discharge into the Rock River in Illinois alone (Illinois Environmental Protection Agency), it remains a source of human and agricultural contamination. In our studies over three summers, from June to August 2012 through 2014, we identified Aeromonas populations at similar concentrations in postchlorinated effluents (POC) (22,492 ± 40,088 CFU/100 ml) compared to those in surface water upstream on the Rock River, IL, USA (44,006 ± 29,105 CFU/100 ml). Aeromonas populations (2.5 × 10³ to 1.0 × 10⁵ CFU/100 ml) consistently outnumbered coliforms (2 to 148 CFU/100 ml) by 1 to 3 logs within postchlorinated wastewater. To determine a potential health risk associated with this aeromonad population, antibiotic susceptibility profiles using CLSI clinical resistance breakpoints were characterized over 3 years (2012 through 2014) in treated effluent and surface river water. Among these summers, the prevalence of resistance to nalidixic acid (NA) showed that this was the only antibiotic exhibiting an annual statistical difference throughout the study, with increased levels among postchlorinated effluent isolates (2012 through 2014, 75% to 95%) compared to complete susceptibility in surface water isolates (Table 1) (P < 0.01). Statistical differences were found during the second summer (2013) and overall (2012 through 2014) where the prevalence of postchlorinated effluent isolates were more resistant against tetracycline (TET) (Table 1) (2013, 18.8% versus 1.8%; 2012 through 2014, 10.4% versus 2.1%) and trimethoprim-sulfamethoxazole (SXT) (Table 1) (2013, 15.6% versus 0%; 2012 through 2014, 10.4% versus 0%) compared to surface water. Among 209 isolates, all isolates were susceptible to ceftazidime, ceftriaxone, meropenem, and gentamicin (GEN), with only one effluent isolate in 2014 lacking susceptibility to the third-generation cephalosporin cefotaxime (CTX) and one in 2013 resistant to ciprofloxacin (CIP) (Table 1). An isolate was considered multidrug resistant (MDR) if it was resistant against three or more categories of antibiotics. This did not include the intrinsic resistance to ampicillin and vancomycin, which is typically evident among Aeromonas spp. MDR isolates from treated effluents (Table 1) (2013, 12.5%) were significantly more prevalent in summer of 2013 than surface water isolates (Table 1) (2013, 0%). Over the 3 years studied, the multiple antibiotic resistance (MAR) index was statistically different between treated effluent and surface water (Table 1) (P < 0.05) and nearly doubled every year compared to that of the previous year (2012, 3.15; 2013, 5.47; 2014, 10.82).

TABLE 1.

Longitudinal analysis of antibiotic resistance (%) from chlorinated wastewater effluents compared to that of upstream surface river water

Drug target and antibiotic	% (no.) of isolates with antibiotic resistance
	Summer 1b		Summer 2b		Summer 3b		All summersb
	RS (n = 40)	POC (n = 12)	RS (n = 57)	POC (n = 32)	RS (n = 47)	POC (n = 21)	RS (n = 144)	POC (n = 65)
Cell wall
Cefotaxime	0.0	0.0	0.0	0.0	0.0	4.5 (1)	0.0	1.5 (1)
Ceftazidime	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
Ceftriaxone	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
Meropenem	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
Protein synthesis
Chloramphenicol	0.0	0.0	1.8 (1)	0.0	2.1 (1)	9.1 (2)	1.4 (2)	3.0 (2)
Gentamicin	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
Kanamycin	0.0	0.0	0.0	3.1 (1)	0.0	4.5 (1)	0.0	3.0 (2)
Streptomycin	20.0 (8)	0.0	3.5 (2)	3.1 (1)	8.5 (4)	13.6 (3)	9.7 (14)	6.0 (4)
Tetracycline	3.0 (1)	0.0	1.8 (1)	18.8 (6)**	2.1 (1)	4.5 (1)	2.1 (3)	10.4 (7)**
DNA synthesis
Ciprofloxacin	0.0	0.0	0.0	3.1 (1)	0.0	0.0	0.0	1.5 (1)
Nalidixic acid	0.0	75 (9)**	0.0	81 (26)**	0.0	95 (20)**	0.0	82 (55)
Folic acid synthesis
SXT	0.0	0.0	0.0	15.6 (5)**	0.0	9.5 (2)	0.0	10.4 (7)**
MDR	0.0	0.0	0.0	12.5 (4)*	2.1 (1)	9.5 (2)	0.7 (1)	9.2 (6)**
MAR indexa	0.020	0.063*	0.019	0.104**	0.011	0.119**	0.017	0.101**

^aRepresents the median multiple antibiotic resistance (MAR) index among isolates within each column.

^bRS, upstream surface river samples; POC, postchlorinated effluents; MDR, multidrug resistant; n, number of isolates in each category. Statistical analysis was performed between RS and POC populations within each summer and cumulatively; *, P < 0.05; **, P < 0.01.

Diversity of Aeromonas spp. in water sources.

To further assess the potential risk associated with resistant isolates, a random subset of each population from the surface water and treated effluents underwent gyrB sequencing to determine species level. In both water samples, the A. veronii group was the predominant species comprising 95.5% in the river and 54.5% in the effluent (Fig. 1 and 2). Between the two, a more diverse representation of species existed in treated effluents (n = 33) where A. veronii group (51.5%) > A. hydrophila (18.2%) > Aeromonas media (15.2%) > A. caviae/punctata (9.1%) > Aeromonas jandaei (3%). Of the 44 species determined from river isolates, only the A. veronii group and A. hydrophila were evident (Fig. 1).

FIG 1.

Speciation of aeromonads comprising water samples. A species representation from river water (n = 44) upstream of the effluent and within the postchlorinated wastewater (n = 33).

FIG 2.

Dendrograms of a subpopulation of Aeromonas isolates. Neighbor-joining method was utilized to infer evolutionary history from a 480-nucleotide sequence of gyrB. Branches were determined using the maximum composite likelihood method comparing a subpopulation of Aeromonas isolates (n = 129) from this study with known Aeromonas spp. sequences obtained from BLAST. Distances were determined based on the number of base substitutions per site.

When determining the prevalence of resistance within species, the MAR index was highest for A. media (n = 4) with 0.21 (Table 2), indicating isolates resistant to the most antibiotics, including the only species exhibiting resistance to ciprofloxacin and cefotaxime. The cefotaxime-resistant isolate also lacked susceptibility to four other antibiotics (i.e., streptomycin [STR], SXT, chloramphenicol [CHL], and nalidixic acid [NA]). However, the low prevalence of these isolates makes further conclusions difficult. The two most prevalent, A. veronii group and A. hydrophila, exhibited the lowest MAR indices, 0.04 and 0.07, respectively.

TABLE 2.

Resistance profile of Aeromonas populations

Species	No. of isolates	% (no.) of isolates with resistance toa :										MAR indexb
		CHL (≤12)	CIP (≤15)	CTX (≤14)	GEN (≤12)	KAN (≤13)	NA (≤13)	STR (≤11)	SXT (≤10)	TET (≤14)	MDR
A. caviae/punctata	3	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	33 (1)	0 (0)	33 (1)	33 (1)	33 (1)	0.11
A. hydrophila	7	0 (0)	0 (0)	0 (0)	0 (0)	14 (1)	43 (3)	0 (0)	13 (1)	13 (1)	0 (0)	0.07
A. jandaei	1	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	100 (1)	0 (0)	0 (0)	0 (0)	0 (0)	0.00
A. media	4	25 (1)	25 (1)	25 (1)	0 (0)	0 (0)	50 (2)	25 (1)	50 (2)	50 (2)	50 (2)	0.21
A. veronii group	59	0 (0)	0 (0)	0 (0)	0 (0)	0 (0)	29 (17)	12 (7)	2 (1)	2 (1)	0 (0)	0.04

^aAntibiotics not included in this table were drugs to which aeromonads exhibited complete susceptibility. CHL, chloramphenicol; CIP, ciprofloxacin; CTX, cefotaxime; GEN, gentamicin; STR, streptomycin; SXT, trimethoprim-sulfamethoxazole; TET, tetracycline; MDR, multidrug resistant. The numbers in parentheses after the antibiotic names in the heading are zones of inhibition (in mm).

^bRepresents the median multiple antibiotic resistance (MAR) index among species.

Characterization of TET-resistant population.

Our study identified TET resistance within 6 of 12 strains resistant to two or more antibiotics; therefore, it was selected to identify and characterize more MDR phenotypes. Isolates on ampicillin dextrin agar with vancomycin and triclosan (Irgasan [BASF]) (ADA-VI) with TET were further screened for a ≤14-mm zone of inhibition (Table 3). Amongst resistance profiles, MDR phenotypes from TET-resistant isolates were more common among postchlorinated effluents (34.8%) than among river samples (8.3%). Resistance to STR and SXT was paired together with TET resistance at 22.0% and 15.3%, respectively (Table 3). A. veronii group isolates exhibited the strongest multidrug resistance with an isolate lacking susceptibility to CIP, STR, SXT, TET, NA, and kanamycin (KAN) (Table 3 and data not shown).

TABLE 3.

Antibiotic resistance profiles of tetracycline-resistant populations

Sample type or species	No. of isolates	% (no.) of isolates resistant to:								MAR index
		CHL	CIP	GEN	KAN	NA	STR	SXT	MDR
River	36	0 (0)	0 (0)	0 (0)	0 (0)	13.9 (5)	13.9 (5)	11.1 (4)	8.3 (3)	0.10
Postchlorinated	23	4.3 (1)	4.3 (1)	0 (0)	8.7 (2)	95.6 (22)	34.7 (8)	21.7 (5)	34.8 (8)a	0.22
A. caviae	1	0 (0)	0 (0)	0 (0)	0 (0)	100 (1)	0 (0)	100 (1)	100 (1)	0.25
A. hydrophila	8	0 (0)	0 (0)	0 (0)	12.5 (1)	75 (6)	37.5 (3)	16.0 (2)	25.0 (2)	0.21
A. media	1	0 (0)	0 (0)	0 (0)	0 (0)	100 (1)	100 (1)	100 (1)	100 (1)	0.25
A. veronii group	23	4.3 (1)	4.3 (1)	0 (0)	4.3 (1)	39 (9)	39.0 (9)	8.6 (2)	13.0 (3)	0.17

^aP < 0.05 comparing the prevalence of MDR in postchlorinated isolates to that in river isolates.

ARGs associated with tetracycline and cefotaxime resistance.

Considering TET resistance was strongly evident among our aeromonad population, molecular mechanisms of resistance were further analyzed. Clinically, drug efflux pump and ribosomal protection proteins appear to be the most prevalent (24). Thus, in this study, we focused on ARGs associated with drug efflux pumps (tetA, tetB, tetC, tetD, and tetE) as well as ribosomal proteins (tetM and tetO). Among a subpopulation of 59 TET-resistant Aeromonas spp. isolates from effluent and surface river water, tetB, tetC, tetD, tetM, and tetO genes were not present (data not shown). Only tetA and tetE genes were detected, with tetE being significantly more prevalent (88%) than tetA (17%) (Table 4) (P < 0.005), though both were evident in surface water and effluent isolates. Out of 59 isolates, only 4 coharbored tetA and tetE. We next compared the presence of tetA and tetE to the degree of TET resistance. Isolates with tetA were significantly more resistant to ≥64 μg/ml of TET compared to tetA-negative populations (Fig. 3) (P < 0.05).

TABLE 4.

Prevalence of tet genes among Aeromonas isolates

Sample type or species	No. of isolates	% (no.) of isolates with tetracycline gene(s)
		tetA	tetEa	tetA and tetE
River	35	11.4 (4)	91.4 (32)	5.7 (2)
Postchlorinated	24	25.0 (6)	83.3 (20)	8.3 (2)
A. caviae	1	0 (0)	100 (1)	0 (0)
A. media	1	0 (0)	100 (1)	0 (0)
A. hydrophila	8	12.5 (1)	87.5 (7)	0 (0)
A. veronii group	25	24 (6)	76 (19)	8.0 (2)
Total	59	16.9 (10)	88.1 (52)	6.8 (4)

^aStatistical analysis comparing prevalence of tetE to that of tetA indicated significant differences (P < 0.005).

FIG 3.

Distribution of MICs among tetracycline genotypes. tetA and tetE genotypes were categorized among Aeromonas isolates selected from tryptic soy agar (TSA) with tetracycline plates. The prevalence of each MIC was determined for the following genotypes: tetA-negative tetE⁺ (n = 49), tetA⁺ tetE-negative (n = 6), and tetA⁺ tetE⁺ (n = 4). Statistical differences existed between tetA⁺ genotypes compared to tetA-negative for MICs of ≥64 mg/liter. *, P < 0.005.

The only cefotaxime-resistant isolate was an A. media isolate, although it was susceptible in the presence of the extended-spectrum β-lactamase (ESBL) inhibitor clavulanic acid (data not shown). This isolate carried both TEM and CTX-M ESBL genes (data not shown).

Pathogenic phenotypes.

To determine the potential of Aeromonas isolates to cause disease, filtered supernatants were analyzed for their ability to lyse human cancer epithelial cells (Fig. 4A and B). Minimal differences in cytotoxicity were evident between Aeromonas populations in the river (64%) and those of postchlorinated effluents (55%) (Fig. 4C). The majority of A. hydrophila (50%) and A. veronii group isolates (70.2%) secreted lethal toxins, although these were absent in A. caviae/punctata and A. media isolates (Fig. 4C). When comparing MDR isolates with cytotoxicity, A. caviae/punctata and A. media exhibited the highest prevalence of MDR, although there was an absence of lethal exotoxins (Fig. 4). Among MDR strains within the A. veronii group and A. hydrophila, only a few exhibited pathogenic potential (4.5% and 0%, respectively), with one A. veronii group isolate exhibiting resistance to 6 of 12 antibiotics (STR, KAN, TET, NA, CIP, SXT) (Table 2).

FIG 4.

Prevalence of cytotoxic phenotypes among Aeromonas isolates. Filtered supernatants from 24-h bacterial cultures of a subpopulation of Aeromonas isolates were tested for cytotoxicity against a human epithelial cancer cell line using an MTT viability assay to determine pathogenic potential. Supernatants with absorbances of ≤50% compared to TSB alone were considered cytotoxic. Medium alone (A) and A. hydrophila ATCC 7966-filtered supernatants (B) were used in each experiment as a negative and positive control, respectively. (C) The number of samples tested for each group was reported above each bar. The percentage of cytotoxicity was determined by dividing the number of isolates exhibiting cytotoxicity by the total number of isolates within that group. POC, postchlorinated effluents.

Serine protease activity correlated with cytotoxicity.

Serine protease is a putative virulence factor due to its involvement in cleaving, and thus activating, precursor virulence factors like glycerophospholipid cholesterol acyltransferase (GCAT) (25) and due to its potential involvement in sepsis and shock (26, 27). In this study, only 5% of noncytotoxic isolates exhibited serine protease activity, whereas 26 of 48 cytotoxic isolates contained protease activity (Table 5) (P < 0.005). A. hydrophila and the A. veronii group isolates exhibited the highest prevalence of protease activity (40% and 32.6%, respectively).

TABLE 5.

Correlation of serine protease activity and cytotoxicity

Species	No. of isolates	No. with protease activity (%)	P valuea
A. caviae
Cytotoxic	0	0	n.s.
Noncytotoxic	4	0 (0)
A. hydrophila
Cytotoxic	5	2 (40)	n.s.
Noncytotoxic	7	0 (0)
A. media
Cytotoxic	0	0	n.s.
Noncytotoxic	6	0 (0)
A. veronii group
Cytotoxic	43	14 (32.6)	0.037
Noncytotoxic	23	2 (8.7)
Total
Cytotoxic	48	16 (33.3)	<0.005
Noncytotoxic	40	2 (5)

^an.s., not significant.

DISCUSSION

Positive correlations between Aeromonas spp. and indicator bacteria (28) support their use in assessing health risks with aquatic environments. Due to the ubiquitous nature of Aeromonas spp. in water environments along with their presence in antibiotic-rich environments (i.e., humans, fish farms, and agriculture), they can serve as major contributors to horizontal gene transfer, both as donors and recipients. Their numbers have commonly been 1 or 2 logs higher than fecal coliforms in both surface water (28) as well as wastewater (29). Though Aeromonas spp. have been suggested as a bacterial indicator to monitor aquatic environmental antibiotic resistance (12, 30), to our knowledge, no study has used this genus to track resistance in a longitudinal study of surface water. This presents a challenge when attempting to assess the long-term impact of agricultural and anthropogenic pollution on environmental antibiotic resistance levels. Additionally, very few studies have further characterized antibiotic-resistant populations with their potential to cause disease, thereby providing incomplete information for public health risk consideration.

Throughout the wastewater treatment process, Aeromonas spp. predominate in influents (29) as well as untreated (5) and treated effluents (5, 31). Our study identified A. veronii group and A. hydrophila isolates comprising 70% of the aeromonad population in postchlorinated wastewater effluents. In contrast, Figueira et al. analyzed raw and treated wastewater, finding A. media and A. punctata comprising more than 65% of each water source (32). However, their use of selective glutamate starch phenol-red agar (GSP) with 30°C incubation for Aeromonas isolation implies that techniques may be the cause of variability.

With detectable levels of antibiotics, like fluoroquinolones (54 to 2,200 ng/liter) and SXT (2.64 to 1,600 ng/liter), in wastewater (33, 34), along with high bacterial burdens, and stressful conditions favoring hypermutation rates due to elevated SOS levels (35), wastewater treatment sites provide ideal environments for acquiring resistance (36). In this study, postchlorinated wastewater effluents exhibited resistance to all antibiotics except ceftazidime, ceftriaxone, meropenem, and gentamicin. This stressful environment was further enriched with MDR strains by more than 13-fold compared to those in upstream surface water with the MAR index increasing every year in effluents. Others have also found a higher MDR prevalence in treated wastewater than that upstream (32, 37). Resistance levels of Aeromonas spp. within treated wastewater in this study were similar to previous studies exhibiting a high susceptibility (>95%) to CHL, CIP, gentamicin, meropenem, and ceftazidime compared to that of TET, STR, and SXT (<85%) (32, 38, 39). Considering the presence of fluoroquinolones in wastewater, both influents and effluents (34), it was surprising that only a few postchlorinated isolates in this study demonstrated resistance. Varela et al. characterized NA-resistant Aeromonas populations from urban wastewater and identified 37% susceptibility to CIP, supporting varying mechanisms of resistance between the two fluoroquinolones (30). Although chromosomal mutations in gyrA and parC are principle mechanisms of fluoroquinolone resistance clinically (40), the majority of wastewater Aeromonas isolates had aac(6′)-lb-cr (59%) and plasmid-carried qnrS2 (51%), suggesting their involvement in stabilizing resident wastewater populations (30). Considering the high susceptibility to CIP in our study, chromosomal ARGs and mutations are a more probable source. Reduced susceptibility to TET, STR, and SXT might be attributed to the Aeromonas mobilome. Previously, tetA, aadA, and sul1 and sul2 have all been found on the same R plasmids, though of various sizes and incompatibility groups (41). Additionally, these ARGs have been identified on other mobile genetic elements, including transposons and class I integrons (41).

Among Aeromonas populations from surface and wastewater, quinolone and ceftazidime resistance is more common with A. media (32). Our data also confirmed A. media isolates to have the highest MAR index (0.21), including an isolate resistant to ciprofloxacin as well as cefotaxime. The cefotaxime-resistant isolate encoded two ESBLs, TEM and CTX-M, which have also been identified among A. media isolates in river sediment in the United Kingdom (42). Overall, ESBL-producing aeromonads are becoming more evident in wastewater (38) and rivers in the United States (43) and worldwide (12, 44). Within the wastewater reclamation process, chlorinated effluents are rich with Aeromonas species, possibly due to their moderate resilience to chlorine (1.2 mg/liter) (6). Considering the high prevalence of resistance within this genus, it suggests other methods of treatment like UV radiation (45) or ultrasound (46) might serve as better alternatives in wastewater management.

Over 3 years, this study identified similarly high levels of Aeromonas spp. in the summer months in surface water from the Rock River and treated wastewater effluents, with the river predominantly populated with A. veronii group and A. hydrophila. Tacao et al. analyzed 12 different rivers and identified only three Aeromonas spp., with A. hydrophila and A. veronii also being the most prominent species when cultured at 37°C comprising 17% of all culturable bacteria (47). This lack of species diversity could be due to seasonal fluctuations in aeromonad populations, where the spring harbors more variety (48). However, incubation temperature is important when comparing studies. Baron et al. identified Aeromonas bestiarum as the most common species and A. veronii as the fifth most common (22) when cultured at 22°C using GSP agar.

Among surface river water isolates, resistance was relatively rare with TET and STR being the most common. Both antibiotics are common in veterinary medicine and agriculture. The Rock River basin north of the sampling site is primarily agricultural land cover (62%) contributing over 157,000 tons per year of suspended solids (49). Fang et al. identified a higher antibiotic resistance index among bacteria associated with suspended particles compared to that of free-floating particles (50), with Aeromonas populations preferring suspended particle colonization (51). We and others (32) have demonstrated a high prevalence of resistance to NA (≥75%) in treated wastewater, though low to complete susceptibility in surface water. In our study, the nearest wastewater sites upstream of our surface water collection site are 43 and 55 km, where each release 2.5 and 11.4 million liters per day (MLD), respectively. The complete susceptibility of surface water Aeromonas spp. to NA in our study suggests that there is minimal impact of treated wastewater effluents on surface water resistance levels in the Rock River. This is in contrast to many other studies where downstream of wastewater effluents exhibited elevated ARBs and ARGs in surface water (37, 52) as well as sediment. However, these studies have primarily focused on shorter distances (<8 km) downstream of effluents. As spatial distances increase, others have also demonstrated a low impact of wastewater effluent on environmental antibiotic resistance and ARG levels (53–55). This is most likely due to the immense breadth of the Mississippi River flow rate. Together, it suggests the surface water resistant isolates in this study are products of agricultural pollution. Though, in the absence of environmental stresses, these bacterial populations might dispense their ARGs, making their origins unidentifiable.

Over 60 different TET resistance genes have been reported, encoding proteins associated with drug efflux pumps, ribosomal protection proteins, drug inactivating enzymes, and undefined (56). Within both water sources, the drug efflux pump gene tetE was the most frequently detected ARG (>80%) among TET-resistant Aeromonas isolates. Other studies have also identified tetE as the most common ARG from surface river waters (12, 57), with elevated tetE-carrying Aeromonas isolates from anthropogenically impacted waters compared to that from unpolluted waters (58). The tetE gene has been reported chromosomally as well as in large 150-kbp plasmids, and Agerso et al. found it in 90% of TET-resistant Aeromonas spp. (59). The continual use of TET in agriculture (60) and fish farms (61) has retained multiple tet genes within manure bacterial populations (62, 63). The only other ARG detected among this resistant population was tetA, where it was twice as common in postchlorinated effluents as surface water. The 12-transmembrane H⁺ antiporter efflux pump, Tet(A), is among the most common efflux pumps within Gram-negative clinical samples (64), where its presence alone has increased TET resistance over 50-fold (65). Others have found similar findings in surface water and river sediment samples with or without human fecal pollution where tetE was the predominant gene compared to tetA (58, 66). Subtherapeutic antibiotic levels throughout the wastewater treatment process and dense bacterial communities propose that horizontal gene transfer leads to the elevated presence of ARGs in effluent, including tetA and tetE (67). Both genes can be found within chromosomes or plasmids (68), even the same plasmid (69). In this study, the presence of both tetA and tetE correlated with increased resistance to TET. Aeromonas populations harboring tetA and tetE have been detected in meat-processing plants (70), indicating agriculture as a reservoir source of ARGs. In the Rock River basin, agriculture comprises 27 concentrated animal feeding operations as well as 61 municipal wastewater treatment facilities (WWTF), together providing 91% of the total suspended solids in the Rock River basin (49) and proposing a feedback loop between the environment, humans, and agriculture.

Cytotoxicity induced by A. veronii and A. hydrophila populations is common (21) and correlates with disease isolates (71). Our findings identified cytotoxic activity among over 50% of A. veronii group and A. hydrophila isolates as well as 30% of MDR isolates, thus posing a potential health risk. With hypervirulent strains of A. hydrophila jeopardizing aquaculture (72) and ESBL-producing strains associating with bacteremia (73, 74), it is critical that we continue to assess antibiotic resistance with pathogenic phenotypes to elaborate on the potential health risk. Previously, our group and others have identified a correlation with Aeromonas strains encoding serine protease and aerolysin genes with cytotoxic phenotypes (21, 75) and disease (11, 76). This study further confirmed these findings by demonstrating a significant positive correlation of serine protease activity with cytotoxicity. The role of this virulence factor in sepsis and shock (77) is mediated through early induction of inflammation resulting in neutrophil migration and vascular leakage, which appears kinin- (26) and complement-mediated through cleavage of C5 to C5a (78). To combat its role in disease, host defense systems have evolved to inhibit serine protease activity through plasma alpha-2-macroglobulin, thus limiting its activity in vivo (79).

In conclusion, wastewater treatment plants provide ideal environments for the enrichment and sustainability of antibiotic resistance. Our longitudinal findings identify annual increases of MAR indices from chlorinated wastewater effluents among aeromonad populations. These resistant isolates existed in both cytotoxic and nonlethal strains proposing a public health concern. Together, our findings reemphasize the urgent need for more longitudinal studies determining if environmental resistance levels are being impacted by anthropogenic and agricultural pollutants. Given that urban wastewater treatment plants release millions of liters of treated wastewater per day into aquatic reservoirs, more research needs to focus on the human health impact considering the ubiquitous nature of these potentially harmful bacteria.

MATERIALS AND METHODS

Sample collection.

Postchlorinated effluent samples were collected at the Rock River Water Reclamation District (Rockford, IL, USA) by plant technicians from a pump connected directly to the effluent water flow just before it enters the Rock River. This treatment facility serves 188,954 inhabitants with a 151 MLD average outflow. Effluents from June to August in 2012 through 2014 were treated with sodium hypochlorite solution (Vertex, MA) to achieve chlorine levels of 4.5 to 7.8 mg/liter, followed by sodium bisulfite (Viking Chemical, Rockford, IL, USA) to remove residual chlorine. River samples were obtained 5.9 km upstream of effluent from the Rock River in Rockford, IL. Surface water collection was performed on days ranging between 3 and 13 days postrainfall, with an average of 7.6 days. Temperatures ranged from 18.6 to 27.2°C with an average of 22.4°C. The mean gage of the Rock River ranged from 0.686 to 0.878 m with an average of 0.790 and contained an average discharge rate of 120.8 m³/s (range of 45 to 227.7 m³/s). All samples were collected in sterile 500-ml collection bottles and immediately placed on ice until processed, within 2 h.

Determining viable bacterial counts and isolation of aeromonads.

Fecal coliforms were quantified from postchlorinated effluents using a standard fecal coliform membrane filter procedure (80). Briefly, different volumes of water were filtered through a sterile 0.45-μm membrane filter (Fisher Scientific, Hampton, NH) and placed on a sterile absorbent pad (EMD Millipore, Burlington, MA, USA) saturated with m-FC broth (BD Biosciences, San Jose, CA, USA). Plates were incubated at 44.5 ± 0.2°C for 24 ± 2 h with blue colonies identified as fecal coliforms. Aeromonads were quantified by passing sample water over 10-μm filters (Thermo Fisher Scientific, Waltham, MA) to remove suspended solids with subsequent passage over 0.45-μm filters (Thermo Fisher Scientific). Filters were then placed onto 60-mm by 15-mm ampicillin dextrin agar (Hardy Diagnostic, CA, USA) with vancomycin (2 μg/ml) and triclosan (10 μg/ml) (ADA-VI) plates to select for aeromonads (81) according to the U.S. Environmental Protection Agency methodology (82). To further select for tetracycline-resistant strains, filters were placed on ADA-VI + tetracycline (16 μg/ml) plates and further incubated at 37°C for 18 to 24 h, and presumptive Aeromonas colonies were identified by the presence of a yellow halo. To ensure isolation, random presumptive isolated colonies were subsequently streaked using the four-quadrant method over ADA-VI medium. After 18 to 24 h, an isolated colony was inoculated into 5 ml of tryptic soy broth (TSB; BD Biosciences). Glycerol (15%) frozen stocks were made from overnight cultures of each sample and stored at −80°C for further study.

Identifying antibiotic susceptibility.

Isolates were tested for antibiotic resistance using the Kirby-Bauer disk diffusion assay. Briefly, overnight cultures were diluted in sterile 0.85% NaCl (pH 7.4) to obtain a 0.5 McFarland turbidity standard (BD Biosciences). Isolates were cultured onto Mueller-Hinton plates, which were stamped with the following antibiotics: cefotaxime (30 μg), ceftazidime (30 μg), ceftriaxone (30 μg), ceftazidime-clavulanic acid (30 μg/10 μg), meropenem (10 μg), chloramphenicol (30 μg), gentamicin (10 μg), kanamycin (30 μg), streptomycin (10 μg), tetracycline (30 μg), ciprofloxacin (5 μg), nalidixic acid (30 μg), and sulfamethoxazole (23.75 μg) with trimethoprim (1.25 μg) (BD Biosciences). Plates were incubated at 37°C for 18 to 24 h, and zones of inhibition were recorded. Determination of antibiotic resistance or susceptibility was based on the Clinical and Laboratory Standards Institute (CLSI) M45-A for Aeromonas hydrophila complex; however, in the absence of specific limits for aeromonads, Enterobacteriaceae family standards were used for streptomycin, nalidixic acid, and kanamycin resistance. The MAR index was calculated by dividing the number of antibiotic-resistant phenotypes by the total number of antibiotics (n = 12) evaluated for each isolate (83). MAR index values of ≥0.2 suggest that the isolate originated from a high-risk source of contamination (83). Determination of MIC levels among tetracycline-resistant isolates was performed as previously described (22). Briefly, overnight bacterial cultures were diluted to a 0.5 MacFarland standard with subsequent 1:100 dilution in Mueller-Hinton broth. In 96-well microplates, serial 2-fold dilutions (8 to 512 mg/liter) in sterile water were performed leaving 50 μl per well. Diluted bacterial cultures (50 μl) were then combined with tetracycline solutions to obtain final concentrations (4 to 256 mg/liter). MIC values were determined by identifying the lowest concentration without turbidity after 24 h ± 2 h at 35°C.

Isolation of genomic DNA and species identification.

To determine the species diversity from postchlorinated effluents and river samples, 129 random isolates from plates, with or without tetracycline, were further analyzed. Genomic DNA was isolated from overnight stocks using the alkaline lysis method (84). Briefly, 100-μl overnight cultures were centrifuged and washed with phosphate-buffered saline (PBS). Pellets were resuspended in 20 μl of 0.1 M KOH and placed in a 100°C water bath for 15 min before cooling and neutralizing with 0.1 M HEPES. Genomic DNA was stored at −20°C until further use. To determine Aeromonas species, the housekeeping gene gyrB was amplified by PCR (85) using GoTaq master mix (Promega, Madison, WI, USA) with 0.02 μM gyrB primers (Table 6) according to the following protocol: initial denaturation at 94°C for 5 min, followed by 35 cycles of 30 s at 94°C, 30 s at 60°C, and 2 min at 72°C, with a final extension at 72°C for 7 min. A sample of amplicons was analyzed using a 1.5% agarose gel and stained with GelRed (Biotium, CA, USA) with subsequent visualization via UV transilluminator (UVP, CA, USA). Samples with a strong 669-bp band were further purified using a MinElute PCR purification kit (Qiagen, Hilden, Germany) and sequenced (Eurofins Genomics, Louisville, KY, USA). Chromatograms were analyzed with Chromas 2.6.5 software with subsequent sequence identity determined using the National Center for Biotechnology Information BLAST software. The percent identities of Aeromonas sobria, Aeromonas allosaccharophila, and some A. veronii were typically within 0.5% identity and therefore grouped together as the A. veronii group. Additionally, one isolate shared 100% identity to both A. punctata and A. caviae, and therefore these two species were paired together in the analysis. Evolutionary analysis was conducted with MEGA X (86) using the neighbor-joining method (87) followed by the maximum composite likelihood method to determine evolutionary distances (88) on 480-bp positions within the gyrB amplicons (Fig. 2).

TABLE 6.

Primer sequences

Gene	Primer	Sequence (5′ to 3′)	Size (bp)	Reference
gyrB	Forward	GGGGTCTACTGCTTCACCAA	669	85
	Reverse	CTTGTCCGGGTTGTACTCGT
tetA	Forward	TTGGCATTCTGCATTCACTC	494	90
	Reverse	GTATAGCTTGCCGGAAGTCG
tetB	Forward	CAGTGCTGTTGTTGTCATTAA	571	90
	Reverse	GCTTGGAATACTGAGTGTAA
tetC	Forward	CTGCTCGCTTCGCTACTTG	897	89
	Reverse	GCCTACAATCCATGCCAACC
tetD	Forward	TGTGCTGTGGATGTTGTATCTC	844	89
	Reverse	CAGTGCCGTGCCAATCAG
tetE	Forward	TATTAACGGGCTGGCATTTC	544	90
	Reverse	AGCTGTCAGGTGGGTCAAAC
tetM	Forward	ACACGCCAGGACATATGGAT	536	90
	Reverse	ATTTCCGCAAAGTTCAGACG
tetO	Forward	GCGGTAATTATGGGAAACGA	550	90
	Reverse	TTTCCCGCTGTTCAGATTTC
CTX-M	Forward	TTTGCGATGTGCAGTACCAGTAA	478	43
	Reverse	TCCGCTGCCGGTTTTATC
TEM-1	Forward	CCGTGTCGCCCTTATTCC	800	43
	Reverse	AGGCACCTATCTCAGCGA

Determination of antibiotic resistance genes.

Prevalence of tet genes was determined following previous methods (89, 90). Briefly, DNA was amplified using 2× GoTaq green PCR master mix (Promega) and 0.2 μM tet primers (Table 6). tetA and tetB PCRs were amplified using the following protocol: initial denaturation at 95°C for 60 s followed by 30 cycles of 30 s at 96°C, 30 s at 60°C, and 30 s at 72°C, and final elongation at 72°C for 10 min. tetC, tetD, tetE, tetM, and tetO followed similar cycling steps except annealing temperature was adjusted to 55°C for 30 s. The presence of extended-spectrum beta-lactamases in the cefotaxime-resistant isolate was determined using the primers in Table 6. Agarose gel electrophoresis was used to visualize PCR amplicons for correct sizes (Table 6) using GelRed nucleic acid stain and viewed with a UV transilluminator (UVP, CA, USA). The CTX-M amplicon was sequenced to further confirm identity (see “Data availability” below).

Cytotoxicity of aeromonad supernatants.

Cytotoxicity of human cancer epithelial cells was assessed using a previously described protocol (21). Briefly, isolates were incubated in TSB for 24 h at 37°C. Aliquots of each isolate were centrifuged at 12,000 × g for 20 min, and supernatants were passed through a 0.22-μm syringe filter. Filtrate was diluted 1:5 in Dulbecco’s Minimal Essential Medium (Sigma, St. Louis, MO, USA) with 5% fetal bovine serum and added to HeLa-229 (ATCC CCL2) cells in a sterile 96-well plate containing 1 × 10⁴ cells/well. The cytotoxicity was measured at 18 to 24 h postinoculation using an MTT [3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide] assay (Sigma) following the manufacturer’s protocol. Briefly, 5 μl of MTT (5 mg/ml) solution diluted in phosphate-buffered saline (PBS) was added per well and incubated for 4 h at 37°C. Medium was aspirated and MTT crystals were solubilized in MTT lysis buffer (10% Triton X-100 in 0.1 N HCl diluted in isopropanol). Cytotoxicity was determined by measuring the absorbance at 570 nm minus 630 nm. The negative control for this assay was TSB alone, and the positive control was A. hydrophila ATCC 7966 reference strain filtrate. All isolates were analyzed in triplicates and cytotoxicity determined by demonstrating 50% or less the absorbance values of TSB alone (≥50% cytotoxicity).

Quantification of serine protease activity.

Serine protease activity was determined following a slight modification of previous methods (91, 92). Filtered bacterial supernatants were mixed with 0.4% azocasein (Sigma) in 25 mM Tris-HCl buffer at a pH of 7.5. The mixed solution was then incubated at 37°C for 1 h with subsequent incubation in 10% trichloroacetic acid (TCA) solution to inhibit the proteolytic activity. After a 30-min incubation period, samples were centrifuged at 12,000 × g for 10 min. Reaction mixtures were transferred to a 96-well plate where they were diluted 1:2 with 1 N NaOH before recording absorbance at 450 nm minus the 570-nm background. All samples were analyzed in duplicate.

Statistics.

Significant differences between the prevalences of antibiotic susceptibility among river and postchlorinated wastewater isolates were analyzed with a two-tailed Fisher’s exact test. Logistic regression was used to determine if the presence of tetA and tetE was associated with the zone of inhibition from tetracycline. All statistical analyses were performed with Stata/IC 12.1 software (College Station, TX, USA).

Data availability.

The Sanger-based sequence of CTX-M from A. media isolate APOC93 demonstrating cefotaxime resistance in Table 2 was deposited in NCBI under GenBank accession number MN657314.