Antimicrobial resistance characterization of Enterococcus faecium, Enterococcus faecalis and Enterococcus hirae isolated from marine coastal recreational waters in the State of São Paulo, Brazil

Low seawater quality is both an environmental concern and a public health problem. These waters are subject to several sources of pollution, including inadequate sewage disposal on beaches and rainwater runoffs (Adeniji et al. 2019; Müller et al. 2020; Reynolds et al. 2020).

There is substantial evidence supporting the correlation between low recreational coastal water quality and a range of diseases among visitors, including gastrointestinal, respiratory, skin, ear, and eye infections, posing a significant impact on the population's health (Kimiran-Erdem et al. 2007; Ahmad et al. 2013; WHO 2021; Adolf et al. 2023). This scenario is critical in countries like Brazil, where the extent of sewage collection is around 55.8%, from which only 51.2% is treated (SNIS 2021).

The genus Enterococcus spp. is a group of Gram-positive bacteria with 62 known species (https://lpsn.dsmz.de/genus/enterococcus), which are commensals in human and several other animals' gut. They are used by the World Health Organization (WHO 2003) and the United States Environmental Protection Agency (US EPA 2012) as an important indicator of fecal contamination in the monitoring of marine water microbiological quality. Although traditionally recognized as a commensal bacterium, Enterococcus has gained importance as a notable opportunistic pathogen, considered the third most common hospital-acquired pathogen in the United States between 2011 and 2014 (Fiore et al. 2019; Idris & Nadzir 2023) and included on the ESKAPE list (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa and Enterobacter spp.) of antibiotic-resistant priority pathogens, created by WHO in 2017 (Rice 2010; De Angelis et al. 2018; Pandey et al. 2021).

Enterococcus represents a good indicator of fecal pollution in natural aquatic ecosystems and a great interest to water quality assessment for several reasons: (1) These bacteria survive longer than coliforms, (2) they are closely related to health risks associated with bathing in the aquatic environment and (3) their multiplication is inhibited in polluted waters (Alibi et al. 2021; Adeniji et al. 2023).

The US EPA (2009) reports the importance of evaluating Enterococcus in samples of fresh or marine water and the direct relationship between the density of this microorganism and the risk of gastrointestinal diseases associated with the recreational use of these waters. In Brazil, the Federal Resolution of the National Council for the Environment (CONAMA) n° 274/2000 classifies the beaches in relation to bathing in two categories: proper and improper. This classification is established according to the densities of thermotolerant coliforms, E. coli and Enterococcus, measured in five consecutive weeks of sampling. The water is considered unsuitable for primary contact recreation when in the last sampling the values of thermotolerant coliforms are greater than 2,500/100 mL or >2,000/100 mL of Escherichia coli or >400 Enterococcus/100 mL.

Wastewater discharges into the seas pose a threat to public health because this sewage contains antibiotics and their by-products, which contribute to the proliferation and development of antibiotic-resistant bacteria in aquatic ecosystems, mainly due to the transfer of resistance genes (Zaheer et al. 2020; Alibi et al. 2021).

Evidence of antibiotic-resistant enterococci has been reported in recreational marine waters and beach sands, suggesting potential environmental contamination (Huijbers et al. 2015; Leonard et al. 2015). Although the specific health implications associated with resistance are still not fully understood, studies carried out in Fortaleza (Brazil) by Carvalho et al. (2014) and in Kidd's Beach (South Africa) by Adeniji et al. (2020) report the presence of Enterococcus species, particularly E. faecalis and E. faecium, with significant levels of antibiotic resistance. This fact emphasizes the need for further studies on the presence and impact of antibiotic-resistant Enterococci in coastal environments.

Therefore, this study aimed to characterize the phenotypic and genotypic profile of antimicrobial resistance in E. faecium, E. faecalis and E. hirae isolated from beach waters in the State of São Paulo, Brazil.

Four beaches in the State of São Paulo were selected for this study, based on their bathing suitability, which is periodically evaluated by the Environmental Company of the State of São Paulo (CETESB). The selection process involved several factors: the frequency of evaluation as unfit for bathing, the geometric mean of enterococci levels and the annual classification in the previous three years 2016, 2017 and 2018 (Table 1).

The selected cities present low rates of wastewater collection and treatment. In the State of São Paulo Coast, there are two main types of sewage treatment: sequencing batch reactor (SBR) and preconditioning treatment system (PTS). In the former, the treated sewage is released into rivers while the latter, after passing through a screening process, the sewage is discharged to the ocean throughout a submarine outfall, along with chlorination. Guarujá, with an urban population of 318,043 inhabitants, counts on one SBR and one PTS with a collection rate of sewage of 64.7% of which 6% is treated. São Sebastião, with an urban population of 86,606 inhabitants, possesses four SBR and two PTS, with sewage collection rate is 40.5% of which 54.8% is treated. São Vicente, with an urban population of 362,483 inhabitants, counts on two SBR to treat its sewage, being the collection rate of 72.6% of which 18% is treated. Ubatuba, with an urban population of 87,575 inhabitants, collects 39.1% of the sewage of which 99.6% is treated (CETESB 2019).

Enterococcus quantification for monitoring beaches' water quality was carried out by CETESB in accordance with the Standard Methods for the Examination of Water and Wastewater (Baird & Bridgewater 2017). Fifty-three Petri dishes showing typical colonies of enterococci in agar mEI (membrane-Enterococcus indoxyl-β-d-glucoside agar) were selected and sent to the laboratory within 24 h after incubation and colony counting. A total of up to five colonies were selected from each dish and individually cultured on TSA (Trypticase Soy Agar), at 37 °C/24 h. The isolates were then enriched in TSB medium (Trypticase Soy Broth) for additional 24 ± 2 h at 37 °C, and preserved in glycerol 30% at −80 °C.

The protein extraction procedure followed the protocol described by Christ et al. (2017). Briefly, each colony was diluted in 300 μL of ultrapure water, 900 μL of absolute ethanol were added to the suspension, and the mixture was centrifuged at 10,000 g for 2 min at room temperature. After careful removal of the ethanol, the resulting pellet was totally air-dried, and equal volumes (50 μL) of 70% formic acid solution and acetonitrile were added to the pellet. The suspension was centrifuged at 10,000 g for 2 min at room temperature, and the supernatant containing the protein extract was frozen at −20 °C for subsequent analysis using mass spectrometry.

For the MALDI-TOF/MS analysis, a volume of 1 μL of each protein extract was carefully applied to a ground steel plate and allowed to air-dry at room temperature. Then, 1 μL of the matrix solution (Sigma-Aldrich^®, Saint Louis, MO, USA) was added to each well, ensuring the coverage of the protein extract. The plate was dried in a dark environment at room temperature and transferred to a Bruker-Daltonics^® Microflex MALDI-TOF mass spectrometer. The FlexControl software (Bruker-Daltonics^®) was employed to program the method and associate the spectra IDs with the Biotyper^® Database. The obtained spectra were then compared with the manufacturer's library specific to each strain. The interpretation criteria provided by Bruker-Daltonics^® for the standards were as follows: scores below 1.7 indicated unidentified samples, scores ranging from 1.7 to 1.9 indicated identification at the genus level, and scores equal to or greater than 2.0 were deemed acceptable for species identification.

Genomic DNA extraction was conducted following a modified version of the method outlined by Costa et al. (2005), using E. faecalis ATCC 29212 as the positive control strain. To summarize, a 1,000 μL inoculum grown in TSI broth was centrifuged at 13,000 rpm for 5 min. The resulting pellet was resuspended in 0.5 mg/mL lysozyme and 25 μL of ultrapure water. This mixture was then incubated at 37 °C with agitation for 10 min. Following this, 50 μL of proteinase K (20 mg/mL) and 150 μL of Tris buffer (0.1 M, pH 7.5) were added to the sample, which was incubated at 37 °C in a shaking bath for 10 min. Subsequently, the sample was incubated at 95 °C in a water bath for 10 min and centrifuged at 13,000 rpm for 5 min, and the resulting supernatant was stored at −80 °C for future use.

Species confirmation was conducted using conventional PCR amplification of specific genes: ddl for E. faecalis and E. faecium and mur-2 for E. hirae (Table 2). Quality control strains for these assays were: E. faecalis ATCC 29212, E. faecium FSP 117/20 and E. hirae FSP 116/20.

The reaction mixtures, with a final volume of 25 μL, comprised 50 pmol of each primer, 1.5 mM MgCl₂, 0.25 mM dNTPs, 1X buffer and 1.25 U of GoTaq^® DNA Polymerase (Promega, EUA).

The identification of E. faecalis and E. faecium was adapted from the protocols of Kariyama et al. (2000) and Dutka-Malen et al. (1995), respectively. The cycling parameters included an initial denaturation step at 94 °C for 1 min, followed by 30 cycles of denaturation at 94 °C for 1 min, annealing at 56 °C for 1 min and extension at 72 °C for 2 min. A final extension step was performed at 72 °C for 5 min. The identification of E. hirae was performed using the method outlined by Arias et al. (2006). The cycling parameters included an initial denaturation step at 94 °C for 2 min, followed by 35 cycles of denaturation at 94 °C for 1 min, annealing at 55 °C for 2 min and extension at 72 °C for 3 min. A final extension step was performed at 72 °C for 10 min.

Amplification products were subjected to electrophoresis on a 1.5% agarose gel containing 0.5 μg/mL of SYBR^® Green, and the results were visualized under UV light.

The antimicrobial susceptibility of the isolates was evaluated using the disc diffusion method on agar, following the recommendations established by the Clinical Laboratory Standard Institute (CLSI 2019), including the quality control strains E. faecalis ATCC 29212 and Staphylococcus aureus ATCC 25923. A panel of 14 antimicrobial agents was selected: ampicillin (10 μg), ciprofloxacin (5 μg), chloramphenicol (30 μg), erythromycin (15 μg), fosfomycin (200 μg), linezolid (30 μg), norfloxacin (10 μg), penicillin (10 μg), rifampicin (5 μg), teicoplanin (30 μg), tetracycline (30 μg), vancomycin (30 μg) and discs with high concentrations of gentamicin (120 μg) and streptomycin (300 μg). Incubation was performed at 37 °C for 18–24 h.

For this investigation, we utilized the multidrug-resistance (MDR) classification system proposed by Magiorakos et al. (2012) and the multiple antibiotic resistance (MAR) index described by Krumperman (1983). Following the MDR criteria, isolates were categorized as MDR if they demonstrated resistance to a minimum of three antimicrobials from distinct classes. Isolates exhibiting non-susceptibility to at least one antimicrobial in all but two or fewer antimicrobial categories were designated as extensively drug-resistant (XDR). The pan-drug-resistant (PDR) classification was assigned to isolates displaying resistance to all evaluated antimicrobials across all categories. Additionally, the MAR index, calculated as a/b, where ‘a’ represents the number of antimicrobial classes to which the isolate exhibited resistance and ‘b’ is the number of antimicrobial classes to which the isolate was exposed, was employed as an additional measure of MAR. Isolates with a MAR index equal to or exceeding 0.2 were classified as multidrug-resistant.

The investigation of antimicrobial resistance-associated genes in the isolates was conducted using conventional PCR. The selected genes, known to confer resistance to various classes of antimicrobials, included vanA and vanB (vancomycin resistance), acc6′-aph2″ (gentamicin resistance), aph(3′)-llla (kanamycin resistance), ant(6)-la (streptomycin resistance), ermB (erythromycin resistance), tetM and tetL (tetracycline resistance), cat (chloramphenicol resistance), blaZ (penicillin resistance), fosB (fosfomycin resistance), optrA and poxtA (linezolid resistance) and the rifampicin resistance determining region of rpoB gene (RRDR-rpoB) (Table 3).

Reaction mixtures were the same as described above, and cycling parameters included an initial denaturation step at 94 °C for 2 min, followed by 30 cycles of denaturation at 94 °C for 1 min, annealing at 55 °C for 30 s and extension at 72 °C for 1 min. A final extension step was performed at 72 °C for 5 min. Amplification products were subjected to electrophoresis on a 1.5% agarose gel containing 0.5 μg/mL of SYBR^® Green, and the results were visualized under UV light. As a quality control measure, PCR products of each gene were randomly picked and submitted to sequencing, to confirm the amplicon specificity.

In total, 265 typical colonies were subjected to preliminary MALDI-TOF/MS identification, from which 256 (96.6%) belonged to the genus Enterococcus. The seven identified species and their corresponding percentages were as follows: E. faecium (77/256, 30.0%), E. faecalis (53/256, 20.7%), E. hirae (35/256, 13.7%), E. casseliflavus (33/256, 12.9%), E. durans (24/256, 9.4%), E. gallinarum (22/256, 8.6%) and E. mundtii (12/256, 4.7%). The three most frequent species (165/256, 64.5%), E. faecium, E. faecalis and E. hirae, were submitted to species-specific molecular identification by PCR. Conventional PCR results confirmed 130 isolates as E. faecium (62/130, 47.7%), E. hirae (36/130, 27.7%), or E. faecalis (32/130, 24.6%). These 130 isolates were then submitted to antimicrobial resistance (AMR) characterization. The distribution of species at each sampling site is shown in Table 4.

A total of 41.5% (54/130) of the isolates exhibited resistance to three or more antimicrobial agents from distinct classes (Table 5), indicating a MDR profile (Magiorakos et al. 2012), which was confirmed by the MAR > 0.2 index (Krumperman 1983). None of the isolates displayed an XDR or PDR profile.

Regarding the species E. faecalis, RRDR-rpoB was detected in all isolates, and the ermB and tetM genes were detected in 18.8% (6/32) of the species' isolates. The tetL gene was detected in 15.6% (5/32), while in lower frequencies genes that confer resistance to aminoglycosides were detected: aph(3′)-llla (9.38% 3/32), acc-6′-aph2″ (6.25% 2/32) and ant(6)-la (6.25% 2/32).

In all E. faecium isolates, the RRDR-rpoB gene was detected. Genes tetM and tetL were detected in 8.0% (5/62) of the species' isolates, and finally, ermB was identified in 1.6% (1/62).

Among E. hirae isolates, RRDR-rpoB was identified in 94.4% (34/36), followed by the tetM and tetL genes, which were detected in 22.2% (8/36) and 18.8% (6 /36) of the isolates, respectively. Finally, the aph(3′)-llla and ermB genes were both detected in 2.77% (1/36) of the isolates.

In this study, in total 256 Enterococcus sp. isolates were selected by the MALDI-TOF/MS method, with 7 identified species described in the results. Among these species, E. faecium, E. hirae and E. faecalis were the most frequent (64.5%) and were submitted to species-specific identification by PCR. From those, 35 were not confirmed, so 130 isolates were subjected to antimicrobial susceptibility characterization (Table 4).

Enterococci are generally present in aquatic and soil environments, in addition to being part of healthy humans and animals' intestinal bacterial flora (Adeniji et al. 2023). The presence of E. faecium, E. faecalis and E. hirae in environmental samples has been consistently documented in several studies including wastewaters, seawater, superficial water and sediment, with the aim to understand the origin, dissemination and persistence of enterococci resistance to some antibiotics (Ferguson et al. 2005; de Oliveira & Pinhata 2008; Prichula et al. 2016; Adeniji et al. 2019, 2023; Saingam et al. 2021; Machado et al. 2023).

E. faecalis and E. faecium are considered indicators of water fecal contamination (Maheux et al. 2011). In our study, E. faecium (47.7%) and E. faecalis (24.6%) were the most frequent species, a fact confirmed by Adeniji et al. (2023), that these species are associated with fecal pollution sources. A study carried out by Machado et al. (2023) assessed three different technologies used by wastewater treatment plants (WTTPs), in Brazil, aiming to evaluate their role in spreading AMR in the environment. The results reveal the Enterobacteriaceae family was dominant with high relative abundances of the genera Escherichia, Enterococcus, Shigella, Enterobacter, Klebsiella and Citrobacter among all WWTPs studied. MDR species of E. faecium and E. faecalis were present in both raw and treated wastewater in the WWTP that consisted of conventional activated sludge. Regarding E. hirae (27.7%), the species is considered related to E. faecalis (Lleò et al. 2005), which may explain their high frequency in the present study.

As recreational water activities increase, monitoring the quality of beach waters becomes necessary (Adeniji et al. 2019). Waters that receive domestic waste can be vehicles for the spread of bacteria that carry ARGs (Dada et al. 2013), and there are several reports of antimicrobial-resistant enterococci in the aquatic environment (Di Cesare et al. 2012; Adeniji et al. 2019, 2020; De Souza et al. 2023). This fact is worrying, as it adds to the already known pathogenicity of E. faecalis and E. faecium (Mancuso et al. 2021). Although E. hirae is considered an opportunistic pathogen with rare cases of human infections (Bourafa et al. 2015), the presence of any multidrug-resistant Enterococcus species in marine waters suggests the spread of AMR among environmental bacteria, which can raise both clinical and ecological concerns.

It is worth noting that the beaches of Camburi and Tombo, which have reasonable and good water quality indexes, respectively (Table 1), presented a lower frequency of resistant isolates (Table 5), perhaps indicating a reduced impact of sewage discharge. However, the detection of multidrug-resistant enterococci in these areas underscores the importance of continuous monitoring and treatment of potential sources of contamination. This observation can be noted at Tombo beach in Guarujá, as it was considered suitable for bathing during the study period, but carried isolates that showed resistance to at least 8 of the antibiotics analyzed (Figure 3). On the other hand, an association can be observed between fecal pollution and multidrug resistance to antibiotics at Itaguá beach in Ubatuba, where E. faecium showed resistance to almost all analyzed antibiotics (11 out of 14), including rifampicin (61.9%), tetracycline (33.0%) and erythromycin (14.3%) (Figure 3 and Table 5). The persistent poor water quality observed at Gonzaguinha and Itaguá beaches (Table 1), despite the existence of sewage collection and treatment systems, may be primarily attributed to ongoing sewage discharges originating from irregular settlements and the detrimental effects of diffuse pollution.

Among the enterococcal isolates analyzed in this study, the highest percentage of resistance was against rifampicin, expressed by 39.2% (51/130) of the isolates (Figure 2). Notably, all enterococcal isolates that displayed rifampicin resistance also tested positive for the mutated rpoB gene (RRDR-rpoB), which was the most frequently detected ARG in this study, with a prevalence of 98.4% (Figure 4), corroborating previous research (Enne et al. 2004; Urusova et al. 2022). Mutations in the rpoB gene may lead to changes in the target site of rifampicin, potentially affecting the antibiotic's ability to inhibit RNA synthesis effectively. Further sequencing analysis is needed to gain a deeper understanding of the relationship between rifampicin resistance and the presence of the mutated rpoB gene, allowing for a more detailed characterization of the molecular mechanisms involved in the resistance phenotype.

The study identified six other ARGs, encoding for resistance to tetracyclines (tetL, tetM), macrolides (ermB) and aminoglycosides (aac(6′), ant(6)-Ia, aph(3′)-IIIa) (Figure 4). The increasing emergence of multidrug-resistant pathogens, such as E. faecium and E. faecalis, is a concerning trend attributed to the genetic flexibility of their genomes and the widespread use of antibiotics (Zaheer et al. 2020). The capability of enterococci to acquire ARGs by horizontal gene transfer suggests other resistances may be present, and the persistent exposure to antibiotics and the acquisition of mobile genetic elements, such as plasmids and integrons, directly impact treatment strategies for infections caused by resistant bacteria enterococci (Uddin et al. 2021; Adeniji et al. 2023). The potential spread of ARGs among enterococci and other bacterial species is a significant concern that warrants attention (Martinez 2009). Moreover, the selective pressure exerted on commensal microorganisms in the gastrointestinal tract due to excessive antibiotic use contributes to the uncontrolled dissemination of resistance, leading to future patterns of multidrug resistance (Baquero 2001; Zaheer et al. 2020).

These findings reveal a significant level of antibiotic resistance among the Enterococcus spp. isolated from the studied beaches. Although the isolates did not reach XDR or PDR profiles, the presence of multidrug resistance raises concerns and underscores the necessity of implementing robust antibiotic stewardship programs and infection control measures in clinical settings, as well as policies to reduce antibiotic use in other settings and the discharge of sewage into environmental matrices. These measures are crucial to reduce the dissemination of resistant strains. Furthermore, the notable prevalence of genetic markers associated with resistance mechanisms highlights the urgency for additional research to elucidate the genetic determinants responsible for resistance in these isolates.

It is important to acknowledge that further investigation is necessary to ascertain the origin and dissemination of these resistant Enterococcus strains in beach environments. The monitoring and understanding of antibiotic resistance dynamics in environmental reservoirs, including beaches, play a pivotal role in assessing potential risks to public health and the environment.

In summary, this study provides significant insights into the identification and patterns of antibiotic resistance in Enterococcus isolates obtained from beaches in São Paulo, Brazil. The findings underscore the importance of continuous surveillance, further research endeavors, and the implementation of appropriate control measures to effectively mitigate the spread of antibiotic resistance in environmental domains.

The authors are thankful for the support by CETESB (Companhia Ambiental do Estado de São Paulo), especially to the laboratories of Cubatão (EDC) and Taubaté (EDT).

This work was partially supported by the National Council of Scientific and Technological Development (CNPq) for GSS (Number: 88887.334803/2019-00).

M.T.P.R. and G.S.S.: Conceptualization, writing and literature search; M.R.F.B. and V.T.M.G.: Analytical procedures for Maldi-Tof isolates identification, review and editing; S.M.-R., T.P.S. and M.D.: Analytical procedures for qPCR and antibiogram analysis; and S.M.-R. and M.D.: Writing, literature search and editing.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.