Microbiological monitoring of coagulase-negative Staphylococcus in public drinking water fountains: Pathogenicity factors, antimicrobial resistance and potential health risks

One of the most concerning issues associated with drinking water is the occurrence of microbial contaminants, which impacts human health. Usually, the assessment of drinking water bacteriological quality is based on fecal indicator bacteria (FIB) such as thermotolerant coliforms or Escherichia coli. In current Brazilian legislation as well as in several other countries, FIB is used as a bacteriological quality indicator for drinking water in addition to heterotrophic plate count (HPC) and residual chlorine, according to the Ministry of Health of Brazil and the Northern Ireland Environment Agency (NIEA) (Ministério da Saúde Brasil 2011; NIEA 2016). FIB (E. coli) acts as an indicator of the potential presence of pathogens and they are effective to identify fragilities in drinking water distribution systems. Nevertheless, these indicators are not always sufficient to indicate the presence of opportunistic bacteria. Furthermore, the usage of indicators alone does not allow a broader view of the drinking water quality regarding the issue of the presence of resistant pathogens to antimicrobials. According to Getahun et al. (2020), antibiotic-resistant bacteria are part of a global crisis, requiring urgent action because untreatable drug-resistant infections and diseases pose the threat of a worldwide public health emergency.

The presence of opportunistic pathogens can lead to interspecies and other bacterial lineage interactions. These interactions favor exchanges of genetic elements that are responsible for antimicrobial resistance (AMR) characteristics. Moreover, the response of their metabolites can amplify virulence in other strains, impacting human health (Hartmann et al. 2018; Hu et al. 2021). Among the diversity of opportunistic bacteria, the group of microorganisms that is increasingly reported in bacterial infections is the coagulase-negative Staphylococcus (CoNS). CoNS have received a lot of attention because of the increasing number of cases of resistant infections in inpatients and individuals outside healthcare settings, according to the National Center for Emerging and Zoonotic Infectious Diseases (NCEZID) of Centers for Disease Control and Prevention (CDC) (NCEZID 2014). In Brazil, according to the Brazilian National Health Surveillance Agency (ANVISA), CoNS infections are the most reported agent in healthcare units and the most resistant agent to antimicrobials used in intensive care units (Ministério da Saúde Brasil 2017).

Although there is ample evidence that opportunistic bacteria exhibit multidrug resistance, their AMR profiles are poorly studied in drinking water (Santos et al. 2020; Hu et al. 2021). This shortage of data is a threat to human health, especially to individuals with compromised immunity (Sanganyado & Gwenzi 2019).

CoNS comprise 23 coagulase-negative staphylococcal species, all of which show a high frequency of AMR and a natural reservoir of genes associated with a virulence that notably favor characteristics for strains that turn out to be more infectious and more resistant to antibiotic treatments (Becker et al. 2014; Argemi et al. 2019; Heilmann et al. 2019). Considering the importance of CoNS in the potential human health impact from exposure to these opportunistic bacteria, this research evaluated the occurrence of them in drinking water distribution systems in urban parks, since the urban parks are public spaces with a high circulation of people (PMSP 2014). This unprecedented study aimed to detect, identify and characterize CoNS microorganisms in drinking water from public drinking fountains in the city of São Paulo, available in public parks.

From March 2017 to March 2018, drinking water samples (n = 468) from drinking water fountains were collected fortnightly from the four parks. Sample collection, storage and chilling for transportation were carried out in accordance with the recommendations of the Standard Methods for the Examination of Water and Wastewater (2017). Free residual chlorine concentration was performed by a colorimetric method using Free-chlorine Analyzer Policontrol^® (São Paulo, BR). E. coli quantification was carried out by the membrane filter method according to Standard Methods (2017) as well as the HPC.

A volume of 100 mL of the water samples was concentrated using a filter membrane (0.45 μm, 47 mm, Millipore^®, USA) and then transferred on a Petri dish containing Baird-Parker agar (BP) (Difco^®, MI, USA) (Standard Methods 2017) followed by incubation for 48 h at 35 ± 0.5 °C. After the incubation period, typical colonies were observed and then submitted to biochemical characterization for staphylococcal bacteria. The selected colonies were initially screened according to the recommendations of the Brazilian National Health Surveillance Agency (ANVISA 2004): Gram stain, catalase reaction (hydrogen peroxide solution 10 V, Laborclin^®, PR – BRA); tube coagulase reaction (Coagu-plasma, Laborclin^®, PR – BRA); DNAse agar test (Difco^®, MI, USA) and fermentation of Mannitol Salt agar (Difco^®, MI, USA).

After the screening step, the typical staphylococcal colonies were transferred to a BHI broth and incubated overnight at 35 ± 0.5 °C. From the bacterial growth, a volume of 1,000 μL was transferred to a microtube and centrifuged at 13,000 rpm for 10 min. The yielded supernatant was discarded, and the pellet was resuspended in 25 μL of lysostaphin enzyme (1 μg/mL) (Sigma^®, Missouri, USA) and 25 μL of ultrapure water and incubated for 10 min at 37° ± 0.2 °C with agitation. After this step, 50 μL of proteinase K (20 mg/mL in sterile deionized water) (Roche^®, California, USA) and 150 μL of Tris buffer (0.1 M, pH 7.5) (USB Corp.^®, Ohio, USA) were added and incubated in a shaking bath at 37 °C for 10 min, followed by incubation in a water bath at 95 °C for 10 min. Finally, the solution was centrifuged at 13,000 rpm for 10 min and the supernatant was stored. The primers used to track seven genes are shown in Table 1.

We used the Bruker MALDI Biotyper for the identification of the isolates. The identification was performed by matrix-assisted laser desorption/ionization mass spectrometry (MALDI-TOF MS) (Moreno et al. 2018). This technique is useful to analyze peptides and proteins in relatively complex samples, such as samples from environmental matrices. It allows us to differentiate species or subspecies, as well as new species and lineages. Mass spectra were obtained by Microflex™ mass spectrometer (Bruker Daltonik^®, Leipzig, Germany). The interpretative criteria were applied as follows: scores ≥2.0 were accepted for species assignment and for gender identification scores ranging from ≥1.7 to ≤2.0 according to the Bruker standard.

The susceptibility to antibiotics was determined by using the disk diffusion method according to the Clinical and Laboratory Standards Institute (CLSI 2020). CLSI has been developing international standards for the past 50 years which are used in more than 50 countries, and benefit safe and transparent laboratory practices (Weinstein & Lewis 2020). Antimicrobials (n = 11) tested were: cefoxitin (30 μg); ciprofloxacin (5 μg); clindamycin (2 μg); chloramphenicol (30 μg); erythromycin (15 μg); gentamicin (10 μg); penicillin g (10 μg); rifampicin (5 μg); sulfazotrin (25 μg); tetracycline (30 μg) and oxacillin (1 μg) (DME^®, SP, Brazil). Vancomycin susceptibility was also assessed by microdilution broth microplate technique for concentrations ranging from 0.5 to 32 μg/mL of Vancomycin hydrochloride (Sigma^®, Missouri, USA).

Out of 468 samples, 393 (83.9%) were in accordance with the standards set by the Brazilian legislation concerning E. coli and HPC, and free chlorine (Ministério da Saúde, Brasil 2011), while 75 (16%) of them did not meet the standard value for free chlorine. Staphylococcus genus was found in 104 (22.2%) samples within which 75 samples presented a lack of free residual chlorine. CoNS were found in 16% of the isolates in which the residual chlorine did not meet the established standard, as follows: 38.7% (29/75) in Aclimação Park; 22.7% (17/75) in Buenos Aires Park and Piqueri Park; and 16% (12/75) in Ibirapuera Park.

All of the 75 isolates were positive for the catalase test, and within them, 53.3% (40/75) were positive for mannitol fermentation and for the DNAse test. For the coagulase reaction 13.3% (10/75) of the isolates were positive as well as for gene coa presence. Genes sea, seg and luk-PVL were not detected; however, their genes recN and icaAB, commonly associated with CoNS, were detected in 4% (3/75) and in 6.7% (5/75), respectively.

A wide diversity of CoNS was identified in our samples (Table 2): 10 species and 25 phylogenetic variations were identified; 85.3% (64/75) of the isolates were identified at a taxonomic level showing logarithmic scores ≥2.0 while 12.3% with scores ≤2.0 were identified at the genus level. Within the 64 isolates, 40.6% (26/64) were identified as Staphylococcus epidermidis while the frequency of Staphylococcus sciuri and Staphylococcus warneri was 17.2% (11/64); 6.3% (4/64) for Staphylococcus saprophyticus and Staphylococcus condiment; 4.7% (3/64) for Staphylococcus haemolyticus; 3.1% (2/64) for Staphylococcus nepalensis; and 1.6% (1/64) for Staphylococcus cohnii, Staphylococcus gallinarum and Staphylococcus pscifermentans. A wide diversity of phylogenetic variation was identified. The greatest variation occurred for the following species: S. haemolyticus with 66.7% (2/3), S. warneri with 45.5% (5/11), S. sciuri with 27.2% (3/11) and S. epidermidis with 11.5% (3/26).

Concerning the antimicrobial susceptibility of the isolates, all of them were susceptible to vancomycin in the broth microdilution test. It was observed that 3.1% (2/64) of isolates were resistant to 63.6% (7/11) of the antibiotics tested (clindamycin, erythromycin, oxacillin, penicillin, tetracycline, cetoxitin and sulfazotrin). Table 2 shows the results obtained in this study, CoNS of isolates examined, organized by taxonomic group and the respective percentages of antibiotic resistance (or intermediary) phenotypes. The resistance to oxacillin was verified in 34.8% (22/64) of the isolates and resistance to cefoxitin in 9.3% (6/64) and for both was 9.3% (6/64). Regarding the detection of mecA gene, 15.6% (10/64) of the isolates carried this gene, which were: S. epidermidis (1/10), S. condimenti (1/10), S. haemolyticus (1/10), S. warneri (2/10) and S. sciuri (5/10). Within them, only S. haemolyticus expressed resistance against oxacillin.

The presence of opportunistic bacteria in various aquatic environments has been increasingly reported (Vikramjeet et al. 2019), including Staphylococcus aureus which has been detected in drinking water samples with the absence of free residual chlorine or below 0.1 mg/L as reported by Santos et al. (2020). The colorimetric method is one of the most used for field evaluation of free residual chlorine concentrations in drinking water (Badalyan et al. 2009; Standard Methods 2017). In the present study, it was shown that Staphylococcus that do not belong to the aureus species, including a variety of CoNS, are capable of surviving in drinking water even in the presence of chlorine. Staphylococcus infections cause several health concerns related to secondary bacterial infections in clinical outcomes and mortality in humans (Argemi et al. 2019). Therefore, evidence that CoNS are present and viable in drinking water is a significant worry for people who are exposed to those drinking water sources (Otzen & Nielsen 2008; Sanganyado & Gwenzi 2019). Taking into account the study carried out by Hu et al. (2021) which reported that there are several threatening environmental contaminants such as opportunistic pathogens compromising the safety of drinking water, raises the need to protect drinking water sources and protect human health.

Therefore, we investigated the occurrence of CoNS in drinking water supplied in public areas. Our study is unprecedented in focusing on the occurrence, taxonomy identification, and pathogenic potentials of negative staphylococci (CoNS) in drinking water distributed by public devices. The results reinforce the presence of these opportunistic bacteria in drinking water due to the absence of residual chlorine or its improper concentration. This is evidence for the poor sanitary conditions, which favor the survival of opportunistic pathogens and their proliferation in these environments.

Čuvalová et al. (2015), in Slovak Republic, reported that out of 10 coagulase-negative staphylococci species isolated 4 were from drinking water samples. The study carried out by Faria et al. (2009) reported the occurrence of coagulase-negative staphylococci in several sources of water, including drinking water from a distribution network. They found the prevalence of Staphylococcus pasteuri and S. epidermidis in these samples. In our study, 10 species were identified and 25 phylogenetic variations of CoNS, within which the most prevalent species were S. epidermidis, S. warneri and S. sciuri (Table 2).

CoNS has been increasingly reported in nosocomial infections worldwide (Table 3), but our study also shows that the presence of CoNS in drinking water is a risk factor for human health. S. warneri, detected in our samples, is considered an emerging pathogen whose pathogenesis and epidemiology have been little explored, but according to Espadinha et al. (2019), this species is frequent in infections of immunocompromised individuals and in sepsis of neonates. Another concerning species isolated from our samples is S. sciuri, which, according to Nemeghaire et al. (2014), is associated with polymicrobial infections and is considered a reservoir of virulence factors, playing a role in the horizontal transfer of genes to other staphylococcal species.

Although there are reports of the presence of genes associated with virulence factors in isolates of CoNS (Heilmann et al. 2019), we did not detect the genes sea, seg and luk-PVL. Only the virulence genes recN and icaAB were detected, which does not exclude that the isolates have other virulence characteristics. However, the identification of isolates using primers for specific recN and icaAB genes could be used to forecast virulent strains that possess the ability to initiate a lethal infection (Raheema et al. 2020).

Regarding antimicrobial susceptibility, the isolates of this study were resistant to penicillin, erythromycin, oxacillin, clindamycin, sulfazotrin and tetracycline, very similar to the resistance showed by clinical isolates (penicillin, erythromycin, oxacillin and gentamicin) (Duran et al. 2012; Čuvalová et al. 2015). Similar results were reported by Čuvalová et al. (2015) for AMR of coagulase-negative staphylococci isolated from drinking water and reported by Faria et al. (2009) as well.

We found the mecA gene in S. haemolyticus, S. epidermidis, S. warneri, Staphylococcus caprae and Staphylococcus capitis ssp. urealyticus, and the same results were found by Čuvalová et al. (2015). Among 10 CoNS isolates carrying mecA gene, only S. haemolyticus was simultaneously resistant to oxacillin and cefoxitin.

According to Shi et al. (2013), disinfection using chlorine can provide a favorable environment for antimicrobial-resistant bacteria (ARB), antimicrobial-resistant genes (ARGs) and mobile genetic elements (MGEs) dissemination. Moreover, infections caused by resistant CoNS are directly associated with worse clinical outcomes, with longer hospital stays, with an increment of mortality, and with an increasing burden and costs on the healthcare infrastructure (Gajdács et al. 2021; Kalantar-Neyestanaki et al. 2022). Currently, there are no data to estimate the risk due to exposure to AMR in the environment for establishing a safety level of risk (Wuijts et al. 2017); however, we have experienced that CoNS may colonize different types of water, including drinking water that fulfills bacteriological quality standards, which poses a risk for human health.

CoNS species and phylogenetic variations identified in drinking water samples from public fountains presented antibiotic-resistant profile, posing serious concerns for human health in spaces with a large circulation of people, especially children, newborns, immunocompromised people and elderly people, whose attendance at parks is high.

This verification is reinforced by the fact of detection of a strain of S. haemolyticus carrying mecA gene, and, moreover, to be resistant to oxacillin and cefoxitin (methicillin-resistant Staphylococcus haemolyticus, MRSH). These results bring us to the crux of the matter, the complexity of keeping drinking water safe, even meeting the established standards. Contamination of drinking water by opportunistic and antimicrobial-resistant pathogens is beyond monitoring classic microbiological indicators of drinking water quality, and, despite the increase of relevant studies related to this issue, it is still an ongoing discussion.

A better knowledge of pathogenic organisms toward survivability, antibiotic resistance profile and chlorine resistance in drinking water systems is needed to assess risks, and to better design sanitary barriers and contamination prevention measures in order to supply safe drinking water.

We thank the Environmental Company of the State of São Paulo – CETESB for all their support in identifying our isolates using the MALDI-TOF technique.

This work is based on research supported in part by the National Council for Scientific and Technological Development (CNPq) of (Contract No 134101/2017-0), Ministry of Science, Technology, Innovations and Communications to encourage research in Brazil.

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

The authors declare there is no conflict.