https://orcid.org/0000-0001-9825-5762
Abstract
In recent decades, antibiotic-resistant bacteria (ARB) emerged and spread among humans and animals worldwide. In this study, we evaluated the presence of ARB and antibiotic resistance genes (ARGs) in the raw sewage of two hospitals in Brazil. Sewage aliquots were inoculated in a selective medium with antibiotics. Bacterial identification was performed by MALDI-TOF and ARGs were assessed by polymerase chain reaction (PCR). A total of 208 strains from both hospitals were isolated (H1 = 117; H2 = 91). A wide variety of Enterobacterales and non-Enterobacterales species were isolated and most of them were Enterobacter spp. (13.0%), Proteus mirabilis (10.1%), and Klebsiella pneumoniae (9.6%). blaTEM and blaKPC were the most frequent β-lactamase-encoding genes and the predominant macrolide resistance genes were mph(A) and mel. Many species had the three tetracycline resistance genes (tetD, tetM, tetA) and strB was the prevalent aminoglycoside resistance gene. Two Staphylococcus haemolyticus strains had the mecA gene. Quinolone, colistin, and vancomycin resistance genes were not found. This study showed that hospital raw sewage is a great ARB and ARG disseminator. Strict monitoring of hospital sewage treatment is needed to avoid the spread of these genes among bacteria in the environment.
HIGHLIGHTS
Diverse Enterobacterales and non-Enterobacterales species have antibiotic resistance genes, including traditionally environmental bacteria.
ESBL-encoding (blaTEM) and carbapenemase genes (blaKPC) were found in hospital raw sewage.
Multiple macrolide and tetracycline resistance genes were reported.
Hospitals contribute to ARB and ARG dissemination to the municipal sewage system.
INTRODUCTION
Antimicrobial resistance is recognized as a global challenge (WHO 2014). In recent decades, antibiotic-resistant bacteria (ARB) emerged and spread among humans and animals worldwide due to the selective pressure caused by the intensive and inappropriate use of antibiotics. In this context, cephalosporin, fluoroquinolone, and carbapenem resistance in Enterobacterales, carbapenem resistance in Pseudomonas aeruginosa and Acinetobacter baumannii, methicillin resistance in Staphylococcus spp., and vancomycin resistance in Enterococcus spp. stand out (CDC 2019). This growing threat requires a global response and the development of strategies to inhibit its spread in hospitals and the environment. Besides these hospital pathogens, several traditionally environmental bacterial species have shown resistance to antibiotics, such as Alcaligenes faecalis, Ochrobactrum spp., and Achromobacter spp. (Furlan & Stehling 2017; Ngbede et al. 2020). By the presence of antibiotic resistance genes (ARGs), these bacteria are able to adapt to new environments and it shows the high gene transmission rate between hospital and environmental bacteria (Picão et al. 2013; Haller et al. 2018). ARG dissemination to environmental and zoonotic bacteria can be a potential risk to human health (Cacace et al. 2019; Ngbede et al. 2020). Ngbede et al. (2020) found mobile colistin resistance genes in Enterobacterales and A. faecalis of animal and human origin in Nigeria. Huang (2020) observed emergent extensively drug-resistant A. faecalis causing human infections. In Brazil, Furlan & Stehling (2017) reported a high level of resistance to β-lactam antibiotics and different β-lactamase-encoding genes in Ochrobactrum and Achromobacter bacteria isolated from soil.
Enterobacterales is the main group that links the hospital and the aquatic and soil environments (Picão et al. 2013; Haller et al. 2018). β-lactamase production is responsible for the most important resistance mechanisms among them and they are traditionally present in health care strains. Extended-spectrum β-lactamases (ESBLs) and carbapenemase-producing bacteria are global problems and strains with β-lactamases in environmental matrices are increasing (Woodford et al. 2014; White et al. 2016). In Brazil, Enterobacterales carrying ESBL genes, such as blaSHV, blaTEM, and blaCTX-M, and carbapenemases, mainly blaKPC, emerged from the soil and different aquatic environments (Picão et al. 2013; Furlan & Stehling 2017).
In hospitals, the use of antibiotics is more intensive and hospital effluents can receive antimicrobial residues and ARB (Harris et al. 2013; Varela & Manaia 2013). These effluents are discharged to municipal wastewater treatment plants (WWTP), which may also serve as important reservoirs for the transfer of ARGs by mobile genetic elements (MGEs) (Rizzo et al. 2013; Varela et al. 2014). The presence of resistance genes in the final WWTP effluents shows the inefficiency of the water treatment process and can contribute to ARB and ARG dissemination (Schwartz et al. 2003; Li et al. 2016).
This study aimed to monitor ARB and ARG in two hospital raw sewage systems from Espírito Santo, Southeastern Brazil, to subsidize strict legislation for environmental control of antibiotic resistance dissemination.
METHODS
Sample collection and hospital characteristics
This study was performed in two hospital raw sewage systems (H1 and H2) in the city of Vitória, Southeastern Brazil. Five samples were collected monthly from October 2020 to February 2021 in each hospital. H1 is a reference teaching institution for the treatment of tuberculosis and HIV/AIDS, which provides from low-complexity to highly specialized health services and has 295 beds and about 2,108 employees. H2 is a reference philanthropic institution for the treatment of various types of cancer and has 265 beds, around 1,500 employees, and over 400 doctors in the clinical staff. The distance between both hospitals is 100 m. At the end of the hospital raw sewage system, before its content was discharged into the municipal sewage system, the samples were collected at a depth of about 0.5 m from the surface. Then, 500 mL sterile plastic bottles were submerged in the effluent, immediately transported in a portable icebox to the laboratory, and processed during the following 6 h.
Sample processing and screening in selective medium
For primary growth, 100 μL of the raw sewage aliquot and its serial dilutions (10−1 and 10−2) were inoculated in triplicate in plates with Mueller-Hinton agar (Oxoid, England) medium supplemented with antibiotics. Amikacin (16 μg/mL), cephalexin (32 μg/mL), erythromycin (2 μg/mL), colistin (2 μg/mL), levofloxacin (2 μg/mL), and tetracycline (4 μg/mL) were the antibiotics and concentrations used. Mannitol salt agar (Oxoid, England) with oxacillin (4 μg/mL) and bile-esculin agar (Himedia, India) with vancomycin (8 μg/mL) were also used to screen for methicillin-resistant Staphylococcus spp. and vancomycin-resistant Enterococcus spp., respectively.
After primary growth, a secondary isolation of colonies was performed on nutrient agar medium (Oxoid, England). For this isolation, the size, pigmentation, opacity, and edges of the colonies on the primary plate were analyzed. One colony of each profile and only one among all colonies with identical profiles were selected for isolation. All strains were stored in tryptone soy broth (TSB) (Himedia, India) with 20% glycerol at −20 °C.
Bacterial identification
Bacterial identification was performed by the matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) technique, using a Bruker Microflex™ MALDI-TOF/MS instrument (Billerica, MA, USA).
DNA extraction and ARG detection in ARB
For DNA extraction, the Wizard™ Genomic DNA kit (Promega, USA) was used, according to the manufacturer's technical specifications.
The presence of resistance genes in ARB was assessed by single or multiplex PCR, according to the antibiotic present in the culture medium for primary isolation (Table 1).
Genes used in PCR to detect antibiotic resistance genes
| Antibiotic used in the screening medium (class) | Resistance genesa | References |
|---|---|---|
| Amikacin (Aminoglycosides) | aacA, aacA1, aac(6’)-Im, aacC1, aadA4, aph, aphA, aph2, strA, strB | Szczepanowski et al. (2009) |
| Erythromycin (Macrolides) | mph(A), mph(B), ermA, ermB, mefA, mel | |
| Levofloxacin (Quinolones) | qnrA3, qnrB1, qnr | |
| Tetracycline (Tetracyclines) | tetA, tetD, tetM | |
| Cephalexin (β-lactams) | blaKPC1-5, blaIMP, blaVIM (M-PCR 1) | Dallenne et al. (2010) |
| blaSHV, blaTEM, blaOXA1-like (M-PCR 2) | ||
| blaOXA48-like | ||
| blaCTX-Mgp1, blaCTX-Mgp2,blaCTX-Mgp9 (M-PCR 3) | ||
| blaOXA23, blaOXA24-like, blaOXA51, blaOXA58 (M-PCR 4) | Woodford et al. (2006) | |
| blaNDM | Poirel et al. (2011) | |
| Colistin (Polimixins) | mcr1, mcr2, mcr3, mcr4, mcr5 (M-PCR 5) | Rebelo et al. (2018) |
| Oxacillinb (β-lactams) | mecA | Milheiriço et al. (2007) |
| Vancomycinb (Glycopeptides) | vanA, vanB (M-PCR 6) | Depardieu et al. (2004) |
| Antibiotic used in the screening medium (class) | Resistance genesa | References |
|---|---|---|
| Amikacin (Aminoglycosides) | aacA, aacA1, aac(6’)-Im, aacC1, aadA4, aph, aphA, aph2, strA, strB | Szczepanowski et al. (2009) |
| Erythromycin (Macrolides) | mph(A), mph(B), ermA, ermB, mefA, mel | |
| Levofloxacin (Quinolones) | qnrA3, qnrB1, qnr | |
| Tetracycline (Tetracyclines) | tetA, tetD, tetM | |
| Cephalexin (β-lactams) | blaKPC1-5, blaIMP, blaVIM (M-PCR 1) | Dallenne et al. (2010) |
| blaSHV, blaTEM, blaOXA1-like (M-PCR 2) | ||
| blaOXA48-like | ||
| blaCTX-Mgp1, blaCTX-Mgp2,blaCTX-Mgp9 (M-PCR 3) | ||
| blaOXA23, blaOXA24-like, blaOXA51, blaOXA58 (M-PCR 4) | Woodford et al. (2006) | |
| blaNDM | Poirel et al. (2011) | |
| Colistin (Polimixins) | mcr1, mcr2, mcr3, mcr4, mcr5 (M-PCR 5) | Rebelo et al. (2018) |
| Oxacillinb (β-lactams) | mecA | Milheiriço et al. (2007) |
| Vancomycinb (Glycopeptides) | vanA, vanB (M-PCR 6) | Depardieu et al. (2004) |
aM-PCR (multiplex PCR). All other genes were detected by single PCR.
bused specifically to detect methicillin-resistant Staphylococcus spp. and vancomycin-resistant Enterococcus spp., respectively.
All PCRs conditions used in this study were validated in laboratory (according to conditions established in the references) and PCR amplicons were sequenced and DNA sequences analysis was performed using BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi).
Statistical analysis
Chi-square or Fisher's exact test was used to evaluate the difference in ARB and ARGs from the hospitals (H1 and H2). Statistical analyses were performed using GraphPad Prism (version 7.04) and p < 0.05 was considered significant.
RESULTS
Bacterial isolated from selective screening medium
Bacterial species isolated from two hospital raw sewage systems (H1 and H2) (n = 208).
Bacterial species isolated from two hospital raw sewage systems (H1 and H2) (n = 208).
The prevalent species in H1 and H2 were significantly different. We found P. mirabilis only in H1 (p < 0.0001) and Enterobacter spp. mainly in H2 (25 × 2; p < 0.0001). P. aeruginosa, K. pneumoniae, and E. coli presented no significant differences between them (p > 0.05). In general, various species existed in only one hospital (Figure 1).
ARG detection in ARB
A total of 36 widely varied strains, including traditionally environmental bacteria (Alcaligenes faecalis and Kluyvera spp.), were isolated from the selective medium with cephalexin. Among them, 31 strains (86.1%) had at least one β-lactam resistance gene. blaTEM (n = 22), blaKPC (n = 17), and blaOXA-1-like (n = 9) were the most frequent genes (Table 2). We did not find blaIMP, blaVIM, blaNDM, blaOXA-24-like, blaOXA-23, blaOXA-48, blaOXA-51, and blaOXA-58. We found blaTEM in 10 different species, mainly Klebsiella pneumoniae and Enterobacter spp., and blaKPC in seven, mainly Enterobacter spp. (Table 2).
Presence of β-lactamase resistance genes in 36 strains isolated from the selective medium with cephalexin
| Species (number of strains H1/H2) | Resistance genesa (number of strains) | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| blaKPC1-5b | blaCTX-M1 group | blaCTX-M2 group | blaCTX-M 9 group | blaTEM | blaSHV | blaOXA1like | ||||||||
| H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | |
| Aeromonas spp. (1/1) | – | 1 | – | – | – | – | – | – | 1 | – | – | – | – | – |
| Alcaligenes faecalis (0/1) | – | – | – | – | – | – | – | – | – | 1 | – | – | – | – |
| Burkholderia cepacia (0/1) | – | 1 | – | – | – | – | – | – | – | – | – | – | – | 1 |
| Enterobacter spp. (2/4) | 2 | 3 | – | – | – | – | – | – | 2 | 2 | 1 | 1 | 1 | 1 |
| Escherichia coli (2/0) | – | – | 2 | – | – | – | – | – | – | – | – | – | 2 | – |
| Klebsiella pneumoniae (3/2) | – | 2 | 1 | – | – | – | – | – | 2 | 1 | – | – | 1 | – |
| Klebsiella oxytoca (0/1) | – | 1 | – | – | – | 1 | – | 1 | – | 1 | – | – | – | – |
| Kluyvera spp. (0/2) | – | 2 | – | – | – | – | – | 2 | – | 1 | – | 1 | – | 1 |
| Ochrobactrum spp (1/0) | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Proteus mirabilis (2/0) | 1 | – | – | – | – | – | – | – | 1 | – | – | – | – | – |
| Pseudomonas stutzeri (1/0) | – | – | – | – | – | – | – | – | 1 | – | – | – | – | – |
| Pseudomonas aeruginosa (3/0) | – | – | – | – | – | – | – | – | 3 | – | 1 | – | 2 | – |
| Pseudomonas spp. (1/0) | – | – | – | – | – | – | – | – | 1 | – | – | – | – | – |
| Serratia spp. (1/0) | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Unidentified (5/2) | 3 | 1 | – | – | – | – | – | – | 4 | 1 | 1 | 1 | – | – |
| Total (22/14) | 6 | 11 | 3 | – | – | 1 | – | 3 | 15 | 7 | 3 | 3 | 6 | 3 |
| Species (number of strains H1/H2) | Resistance genesa (number of strains) | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| blaKPC1-5b | blaCTX-M1 group | blaCTX-M2 group | blaCTX-M 9 group | blaTEM | blaSHV | blaOXA1like | ||||||||
| H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | |
| Aeromonas spp. (1/1) | – | 1 | – | – | – | – | – | – | 1 | – | – | – | – | – |
| Alcaligenes faecalis (0/1) | – | – | – | – | – | – | – | – | – | 1 | – | – | – | – |
| Burkholderia cepacia (0/1) | – | 1 | – | – | – | – | – | – | – | – | – | – | – | 1 |
| Enterobacter spp. (2/4) | 2 | 3 | – | – | – | – | – | – | 2 | 2 | 1 | 1 | 1 | 1 |
| Escherichia coli (2/0) | – | – | 2 | – | – | – | – | – | – | – | – | – | 2 | – |
| Klebsiella pneumoniae (3/2) | – | 2 | 1 | – | – | – | – | – | 2 | 1 | – | – | 1 | – |
| Klebsiella oxytoca (0/1) | – | 1 | – | – | – | 1 | – | 1 | – | 1 | – | – | – | – |
| Kluyvera spp. (0/2) | – | 2 | – | – | – | – | – | 2 | – | 1 | – | 1 | – | 1 |
| Ochrobactrum spp (1/0) | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Proteus mirabilis (2/0) | 1 | – | – | – | – | – | – | – | 1 | – | – | – | – | – |
| Pseudomonas stutzeri (1/0) | – | – | – | – | – | – | – | – | 1 | – | – | – | – | – |
| Pseudomonas aeruginosa (3/0) | – | – | – | – | – | – | – | – | 3 | – | 1 | – | 2 | – |
| Pseudomonas spp. (1/0) | – | – | – | – | – | – | – | – | 1 | – | – | – | – | – |
| Serratia spp. (1/0) | – | – | – | – | – | – | – | – | – | – | – | – | – | – |
| Unidentified (5/2) | 3 | 1 | – | – | – | – | – | – | 4 | 1 | 1 | 1 | – | – |
| Total (22/14) | 6 | 11 | 3 | – | – | 1 | – | 3 | 15 | 7 | 3 | 3 | 6 | 3 |
a‘–’: absence. We did not find blaIMP, blaVIM, blaOXA-24 like, blaOXA23, blaOXA-51, blaOXA-58, blaNDM1, or blaOXA48 in any strain.
b: statistical difference in gene detection between H1 and H2.
Only blaKPC showed a significant difference (p = 0.0054) in gene detection between H1 and H2.
Among the 28 strains isolated from the medium with erythromycin, mph(A) (n = 23) was the prevalent gene. We found mel and ermB in 18 and 16 strains, respectively, and did not find mph(B), ermA, and mefA. Different species present the coexistence of distinct macrolide resistance genes. We found mph(A) in all identified species, except for Klebsiella oxytoca, which had no genes (Table 3).
Presence of macrolide resistance genes in 28 strains isolated from the selective medium with erythromycin
| Species (number of strains H1/H2) | Resistance genesa (number of strains) | |||||
|---|---|---|---|---|---|---|
| mph(A) | ermB | mel | ||||
| H1 | H2 | H1 | H2 | H1 | H2 | |
| Citrobacter koseri (1/0) | 1 | – | 1 | – | 1 | – |
| Citrobacter spp. (0/1) | – | 1 | – | – | – | – |
| Enterobacter spp. (0/4) | – | 3 | – | 1 | – | 2 |
| Escherichia coli (3/0) | 2 | – | 2 | – | 1 | – |
| Klebsiella oxytoca (0/1) | – | – | – | – | – | – |
| Klebsiella pneumoniae (1/1) | 1 | 1 | – | 1 | – | 1 |
| Klebsiella spp. (1/1) | – | 1 | – | – | – | – |
| Klebsiella variicola (1/0) | 1 | – | 1 | – | 1 | – |
| Proteus mirabilis (3/0) | 3 | – | 2 | – | 3 | – |
| Pseudomonas aeruginosa (3/0) | 3 | – | 2 | – | 3 | – |
| Pseudomonas spp. (1/0) | 1 | – | 1 | – | 1 | – |
| Unidentified (3/3) | 3 | 2 | 3 | 2 | 3 | 2 |
| Total (17/11) | 15 | 8 | 12 | 4 | 13 | 5 |
| Species (number of strains H1/H2) | Resistance genesa (number of strains) | |||||
|---|---|---|---|---|---|---|
| mph(A) | ermB | mel | ||||
| H1 | H2 | H1 | H2 | H1 | H2 | |
| Citrobacter koseri (1/0) | 1 | – | 1 | – | 1 | – |
| Citrobacter spp. (0/1) | – | 1 | – | – | – | – |
| Enterobacter spp. (0/4) | – | 3 | – | 1 | – | 2 |
| Escherichia coli (3/0) | 2 | – | 2 | – | 1 | – |
| Klebsiella oxytoca (0/1) | – | – | – | – | – | – |
| Klebsiella pneumoniae (1/1) | 1 | 1 | – | 1 | – | 1 |
| Klebsiella spp. (1/1) | – | 1 | – | – | – | – |
| Klebsiella variicola (1/0) | 1 | – | 1 | – | 1 | – |
| Proteus mirabilis (3/0) | 3 | – | 2 | – | 3 | – |
| Pseudomonas aeruginosa (3/0) | 3 | – | 2 | – | 3 | – |
| Pseudomonas spp. (1/0) | 1 | – | 1 | – | 1 | – |
| Unidentified (3/3) | 3 | 2 | 3 | 2 | 3 | 2 |
| Total (17/11) | 15 | 8 | 12 | 4 | 13 | 5 |
a‘–’ absence. We did not find mph(B), ermA, and mefA.
The genes found in H1 and H2 presented no significant difference between them (p > 0.05).
All 31 strains isolated from the selective medium with tetracycline had at least one of the three genes: we found tetA, tetD, and tetM in 23 (74.2%), 27 (87.1%), and 26 (83.9%) strains, respectively. We found these genes in 10 different species, mainly E. coli (n = 8) and K. pneumoniae (n = 6) (Table 4).
Presence of tetracycline resistance genes in 31 strains isolated from the selective medium with tetracycline
| Species (number of strains H1/H2) | Resistance genes (number of strains) | |||||
|---|---|---|---|---|---|---|
| tetA | tetD | tetM | ||||
| H1 | H2 | H1 | H2 | H1 | H2 | |
| Acinetobacter spp. (1/0) | 1 | – | 1 | – | 1 | – |
| Aeromonas spp. (1/2) | 1 | 1 | 1 | 2 | 1 | 2 |
| Enterobacter spp. (0/3) | – | 1 | – | 3 | – | 1 |
| Escherichia coli (3/5) | 3 | 3 | 3 | 5 | 3 | 3 |
| Klebsiella oxytoca (0/1) | – | – | – | 1 | – | 1 |
| Klebsiella pneumoniae (4/2) | 4 | 2 | 4 | 2 | 4 | 1 |
| Proteus mirabilis (1/0) | 1 | – | 1 | – | 1 | – |
| Pseudomonas aeruginosa (1/0) | 1 | – | 1 | – | 1 | – |
| Pseudomonas spp. (1/0) | 1 | – | – | – | 1 | – |
| Serratia spp. (1/0) | 1 | – | – | – | 1 | – |
| Unidentified (4/1) | 3 | – | 2 | 1 | 4 | 1 |
| Total (17/14) | 16 | 7 | 13 | 14 | 17 | 9 |
| Species (number of strains H1/H2) | Resistance genes (number of strains) | |||||
|---|---|---|---|---|---|---|
| tetA | tetD | tetM | ||||
| H1 | H2 | H1 | H2 | H1 | H2 | |
| Acinetobacter spp. (1/0) | 1 | – | 1 | – | 1 | – |
| Aeromonas spp. (1/2) | 1 | 1 | 1 | 2 | 1 | 2 |
| Enterobacter spp. (0/3) | – | 1 | – | 3 | – | 1 |
| Escherichia coli (3/5) | 3 | 3 | 3 | 5 | 3 | 3 |
| Klebsiella oxytoca (0/1) | – | – | – | 1 | – | 1 |
| Klebsiella pneumoniae (4/2) | 4 | 2 | 4 | 2 | 4 | 1 |
| Proteus mirabilis (1/0) | 1 | – | 1 | – | 1 | – |
| Pseudomonas aeruginosa (1/0) | 1 | – | 1 | – | 1 | – |
| Pseudomonas spp. (1/0) | 1 | – | – | – | 1 | – |
| Serratia spp. (1/0) | 1 | – | – | – | 1 | – |
| Unidentified (4/1) | 3 | – | 2 | 1 | 4 | 1 |
| Total (17/14) | 16 | 7 | 13 | 14 | 17 | 9 |
‘–’ absence.
We observed a significant difference between H1 and H2 regarding tetA (p = 0.0109) and tetM (p = 0.0118).
A total of 39 strains were isolated from the selective medium with amikacin and 29 (74.4%) of them had at least one resistance gene – strB (n = 18) and strA (n = 11) were prevalent. We also found aacA (n = 10), aadA4 (n = 7), and aph (n = 5), but did not find aph2, aphA, aacC1, aac(6)Im, and aacA1 (Table 5).
Presence of aminoglycoside resistance genes in 39 strains isolated from the selective medium with amikacin
| Species (number of strains H1/H2) | Resistance genesa (number of strains) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| aacA | aadA4 | aph | strA | strBb | ||||||
| H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | |
| Acinetobacter baumannii (2/0) | 1 | – | 2 | – | 1 | – | 1 | – | 2 | – |
| Aeromonas hydrophila (0/1) | – | – | – | – | – | – | – | – | – | – |
| Aeromonas spp. (1/3) | 1 | 1 | – | – | – | – | – | 1 | 1 | 1 |
| Burkholderia cepacia (0/1) | – | – | – | – | – | – | – | – | – | – |
| Citrobacter spp. (1/0) | 1 | – | – | – | 1 | – | 1 | – | 1 | – |
| Enterobacter spp. (0/6) | – | 2 | – | 2 | – | – | – | 2 | – | – |
| Enterococcus faecalis (2/1) | 1 | – | – | 1 | 1 | – | 1 | – | 2 | – |
| Enterococcus faecium (1/0) | – | – | – | – | – | – | – | – | 1 | – |
| Proteus mirabilis (2/0) | – | – | – | – | – | – | – | – | 1 | – |
| Pseudomonas aeruginosa (3/2) | – | 1 | – | – | – | – | 1 | – | 3 | – |
| Pseudomonas alcaligenes (2/0) | – | – | 1 | – | – | – | 1 | – | 1 | – |
| Pseudomonas spp. (2/0) | 1 | – | 1 | – | 1 | – | 1 | – | 2 | – |
| Unidentified (5/4) | – | 1 | – | – | 1 | – | 2 | – | 2 | 1 |
| Total (21/18) | 5 | 5 | 4 | 3 | 5 | 0 | 8 | 3 | 16 | 2 |
| Species (number of strains H1/H2) | Resistance genesa (number of strains) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| aacA | aadA4 | aph | strA | strBb | ||||||
| H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | H1 | H2 | |
| Acinetobacter baumannii (2/0) | 1 | – | 2 | – | 1 | – | 1 | – | 2 | – |
| Aeromonas hydrophila (0/1) | – | – | – | – | – | – | – | – | – | – |
| Aeromonas spp. (1/3) | 1 | 1 | – | – | – | – | – | 1 | 1 | 1 |
| Burkholderia cepacia (0/1) | – | – | – | – | – | – | – | – | – | – |
| Citrobacter spp. (1/0) | 1 | – | – | – | 1 | – | 1 | – | 1 | – |
| Enterobacter spp. (0/6) | – | 2 | – | 2 | – | – | – | 2 | – | – |
| Enterococcus faecalis (2/1) | 1 | – | – | 1 | 1 | – | 1 | – | 2 | – |
| Enterococcus faecium (1/0) | – | – | – | – | – | – | – | – | 1 | – |
| Proteus mirabilis (2/0) | – | – | – | – | – | – | – | – | 1 | – |
| Pseudomonas aeruginosa (3/2) | – | 1 | – | – | – | – | 1 | – | 3 | – |
| Pseudomonas alcaligenes (2/0) | – | – | 1 | – | – | – | 1 | – | 1 | – |
| Pseudomonas spp. (2/0) | 1 | – | 1 | – | 1 | – | 1 | – | 2 | – |
| Unidentified (5/4) | – | 1 | – | – | 1 | – | 2 | – | 2 | 1 |
| Total (21/18) | 5 | 5 | 4 | 3 | 5 | 0 | 8 | 3 | 16 | 2 |
a‘–’ absence. We did not find aacA1, aac(6’)Im, aacC1, aphA, and aph2.
b: statistical difference in gene detection between H1 and H2.
We observed that strB was more significantly detected in H1 (p < 0.0001). Other genes presented no significant difference.
A total of 33 strains were isolated from the medium with levofloxacin and we did not find qnr, qnrA3, and qnrB1 in any strain. Considering n = H1/H2, the following species were isolated: Acinetobacter baumannii (n = 1/0), Alcaligenes faecalis (n = 0/2), Citrobacter freundii (n = 0/5), Corynebacterium spp. (n = 1/0), Enterobacter spp. (n = 0/1), Enterobacter cloacae (n = 0/1), Enterococcus faecalis (n = 1/0), Escherichia coli (n = 2/2), Klebsiella oxytoca (n = 1/0), Klebsiella pneumoniae (n = 2/3), Klebsiella spp. (n = 1/0), Morganella morganii (n = 1/0), Proteus mirabilis (n = 2/0), Pseudomonas aeruginosa (n = 2/1), Pseudomonas spp. (n = 1/0), and unidentified (n = 3/0).
Moreover, 36 strains were isolated from the medium with colistin: Aeromonas hydrophila (n = 1/0), Aeromonas spp. (n = 0/1), Chromobacterium violaceum (n = 0/1), Enterobacter spp. (n = 0/7), Escherichia coli (n = 1/0), Klebsiella pneumoniae (n = 1/1), Kluyvera spp. (n = 0/1), Ochrobactrum spp. (n = 0/1), Proteus mirabilis (n = 11/0), Providencia spp. (n = 0/1), and unidentified (n = 4/5). We did not find mcr1, mcr2, mcr3, mcr4, and mcr5 in any strain.
Five gram-positive strains were isolated from mannitol salt agar supplemented with oxacillin: Staphylococcus haemolyticus (n = 2/0), Enterococcus faecalis (n = 2/0), and Enterococcus faecium (n = 0/1). The two S. haemolyticus strains had mecA gene.
We found no Enterococcus in the selective bile-esculin agar with vancomycin. Thus, we did not study vanA and vanB genes.
DISCUSSION
The flow of effluents and solid waste with chemicals, pharmacological pollutants, and persistent organic compounds has been changing aquatic ecosystems. In this study, we assessed a wide variety of species with different ARGs. Enterobacterales were prevalent (>65%) in a selective medium supplemented with antibiotics. Other studies found higher rates for this group in hospital wastewater (Picão et al. 2013; Haller et al. 2018). In our study, Enterobacter spp., K. pneumoniae, P. mirabilis, E. coli, and P. aeruginosa were the most isolated species. This group of bacteria is very relevant due to their association with serious infections and antibiotic resistance in health care (Bush & Bradford 2020). We observed significant differences between some prevalent species isolated in the two hospitals. We found most Pseudomonas species and P. mirabilis only in H1. On the other hand, we found most Enterobacter spp. in H2. The reasons for this difference are unknown; however, it may reflect the environmental characteristics and circulating microbiota in each hospital. E. coli and Klebsiella species presented no significant difference between hospitals.
β-lactams are among the most frequently prescribed antibiotics and are important in the treatment of numerous types of infections; however, the dissemination of β-lactamase-encoding genes, mainly the TEM, SHV, and CTX-M extended-spectrum β-lactamases and carbapenemases, such as KPC and NDM, are serious concerns for public health (Lee et al. 2016; Wang et al. 2018; CDC 2019). In Brazil, Picão et al. (2013) found KPC-2-producing Aeromonas spp. and Enterobacterales, including Kluyvera spp., in hospital effluents and different WWTP sites. The authors showed that ARB are continually discharged in urban rivers, even after a secondary sewage treatment. Similarly, in our study, blaKPC was predominant in several Enterobacterales, including Kluyvera spp. and Aeromonas spp. A high concentration of blaKPC in multispecies was found in the hospital wastewater and forepart stages of the WWTP in the United States (Loudermilk et al. 2022).
The Burkholderia cepacia complex is a group of gram-negative and glucose nonfermentative bacteria found in the environment that was isolated from opportunistic infection in immunocompromised patients (Uehlinger et al. 2009). Furlan et al. (2018) were the first to assess blaKPC in the B. cepacia complex isolated from Brazilian soil. The authors also found blaOXA-1-like in these strains. We found blaKPC and blaOXA-1 like in only one B. cepacia strain. To our knowledge, this is the first study showing B. cepacia from hospital raw sewage systems with these genes isolated. These results are worrying and show the possibility of transfer to other glucose-nonfermentative bacteria, such as Pseudomonas spp. and Acinetobacter spp.
Conte et al. (2017) described ESBL and quinolone-resistant Enterobacterales from a hospital effluent, a sanitary effluent, an inflow sewage, an aeration tank, and an outflow sewage within a WWTP. Results show the high presence of ESBLs, mainly blaCTX-M. Moreover, they found quinolone resistance genes at all sites, except in the inflow sewage and aeration tank. In our study, the rates of blaCTX-M were low and blaTEM was the most isolated gene. We did not find any quinolone resistance gene (qnr, qnrA3, or qnrB1). Haller et al. (2018) evaluated effluents from two large hospitals in Singapore and found several β-lactamase-encoding genes: blaSHV (41.1%), blaNDM1 (35.6%), blaCTX-M (35.6%), and blaKPC (28.8%) were prevalent. In our study, we found no strains with blaNDM1. A pan-European survey analyzing ARGs in treated wastewater and in the receiving water bodies showed that the absolute abundance (ARGs/100 mL) of the analyzed genes could be ranked according to the following order: intI1 > sul1 > tetM > blaOXA-58 > blaTEM > blaOXA-48 > blaCTX-M-32 > mcr-1 > blaCTX-M-15 > blaKPC (Cacace et al. 2019). We did not find blaOXA-58, blaOXA-48, and mcr-1 in our study, however, we found tetM in most strains.
We described ARGs in various species that were predominantly isolated from aquatic environments and soil or have been involved in opportunistic human infections. Alcaligenes faecalis is a sporadic and opportunistic cause of infection in immunocompromised patients and extensively drug-resistant A. faecalis infections emerged (Huang 2020). Ngbede et al. (2020) found mobile colistin resistance genes (mcr-1.1) in A. faecalis of animal origin in Nigeria and Al Laham et al. (2017) found strains with different blaVIM in Gaza, Palestine. In our study, A. faecalis strains had only blaTEM. Ochrobactrum spp., another traditionally environmental pathogen, can also cause human infection (Dharne et al. 2008). In Brazil, Furlan & Stehling (2017) showed a high level of resistance to β-lactam antibiotics and found different β-lactamase-encoding genes (blaCTX-M-gp1, blaSHV, blaOXA-1-like, and blaKPC) in Ochrobactrum spp. and Achromobacter spp. In our study, the Ochrobactrum spp. strain, when isolated from a selective medium with β-lactam antibiotics, had blaTEM, which confirms the emergence of resistance genes in these species.
Tetracyclines are used in agriculture, fish farming, and in the treatment of infectious diseases in humans and animals. Tetracycline resistance genes were found in different bacterial species isolated from aquatic environments in different areas of the world (Szczepanowski et al. 2009; Xu et al. 2015; Tuo et al. 2018). In our study, we found tetA, tetD, and tetM in most strains. Tuo et al. (2018) showed the prevalence of tetM and tetA in strains isolated from the Funan River, China. Szczepanowski et al. (2009) also found these genes in WWTP in Germany. Xu et al. (2015) evaluated the distribution of antibiotics and ARGs in WWTP in Beijing, China, and of the nine tetracycline genes studied, three were related to the efflux pump (tetA, tetB, and tetE) and three to ribosomal protection (tetM, tetZ, and tetW). Their rates were high, which is in accordance with our findings for tetA, tetD, and tetM.
Various authors demonstrated high rates of macrolide genes in WWTP and hospital effluents (Szczepanowski et al. 2009; Rodriguez-Mozaz et al. 2015; Milaković et al. 2020). We also found three macrolide resistance genes (mphA, ermB, and mel) among the six studied. In Germany, Szczepanowski et al. (2009) found 15 macrolide resistance genes out of 20 — mphA, mphB, mph, ermA, ermB, mefA, and mel were the most prevalent genes in activated sludge from WWTP and mphA, mphB, mph, ermB, and mel in its final effluents. In Spain, the presence of ermB in hospital effluents and WWTP effluents and influents was associated with the presence of antibiotics, such as clarithromycin and azithromycin (Rodriguez-Mozaz et al. 2015). Similarly to our study, two different Chinese studies found high ermB concentrations in hospital wastewater from three tertiary public hospitals (Wang et al. 2018) and various sites throughout the Funan River (Tuo et al. 2018). Macrolides are used to treat gram-positive human infections and their activity against Enterobacterales is poor (Gomes et al. 2017). However, the activity of azithromycin against most common diarrhoeagenic pathogens, such as E. coli, Shigella spp., and Salmonella spp., was excellent (Lübbert 2016; Gomes et al. 2017). Similarly to our study, Gomes et al. (2019) showed that mph(A) was the most frequent macrolide resistance gene in E. coli isolated from stool samples of children under five years of age in periurban areas in Lima, Peru. Moreover, in this study, erm(A), erm(B), and mef(A) were less frequent. In our study, we did not find ermA and mefA, but found erm(B) in more than 50% of strains (16/28).
Regarding aminoglycoside resistance genes, Tuo et al. (2018) showed that aac(6′)-Ib-cr and aph(3′)-IIIa were prevalent among 24 mcr-positive strains. Aminoglycosides were the most abundant genes in all ARG groups (24.57%) from an urban general hospital sewage in Shantou, China – aph(6)-Id, aph(3′′)-Ib, and aac(6′)-Ib7 were the prevalent genes (Cai et al. 2021). In our study, we found only five out of the 10 genes analyzed and strA, strb, and aacA were the most isolated. These genes circulate in aquatic and animal strains. Lu et al. (2022) showed that, among the streptomycin-resistant E. coli isolated from chickens, 99.76% had the strA aminoglycoside resistance gene and 63.27% had strB.Graves et al. (2011) compared the presence of ARGs in E. coli isolated from swine manure, lagoon effluent, and soil collected from lagoon waste and showed that the most frequent aminoglycoside resistance genes were ARGs – aadA, strA, and strB were prevalent.
Methicillin resistance in staphylococci occurs by the acquisition of staphylococcal cassette chromosome mec (SCCmec), which has mecA gene and its homologues, such as mecC (Lakhundi & Zhang 2018). Several studies have already found mecA or mecC in hospital effluents and aquatic environments. Basode et al. (2018) found mecA in S. aureus in three out of four samples collected from a municipal sewage, including those of an animal slaughterhouse and a fish market in Saudi Arabia. Silva et al. (2022) found mecA in 45 methicillin-resistant staphylococci (28 S. aureus and 17 coagulase-negative staphylococci from the Hospital Center Trás-os-Montes e Alto Douro, Portugal). Silva et al. (2021) found mecC in methicillin-resistant Staphylococcus aureus (MRSA) from surface water in Portugal. MRSA is a serious global threat and is related to hospital and community infections (Lakhundi & Zhang 2018). In our study, we did not study mecC and only two strains had mecA (both were Staphylococcus haemolyticus). Another studied already showed that this species can serve as a reservoir of resistance genes for other staphylococcci species (Rossi et al. 2016). The presence of this gene in the environment may favor its dissemination to other strains of the genus, especially S. aureus, considering that the amount of community-associated methicillin-resistant strains is increasing (Fritz et al. 2020).
Although some Brazilian laws and regulations about waste and wastewater management in hospital and health care units exist, no detailed study links the discharge of hospital raw sewage with the main urban sewage collection line. This could represent an open risk for the environmental dissemination of antibiotic resistance among the urban population and highlights the need for specific legislation about proper hospital sewage treatment and waste disposal.
CONCLUSION
This study showed that a diversity of traditionally environmental bacterial species and Enterobacterales has relevant ARGs. Enterobacter spp., Proteus mirabilis, Klebsiella pneumoniae, and Escherichia coli had various ARGs, including β-lactams (blaTEM and blaKPC), macrolides (mphA, mel, and ermB), tetracycline (tetA, tetD, and tetM), and aminoglycosides (strB and strA). Studies on the presence of antibiotic-resistant bacteria in hospital raw sewage systems are essential to assess and establish a specific legislation for their treatment.
ACKNOWLEDGEMENTS
We thank Julianne Soares Jardim Lacerda Batista for her contributions to the statistical analysis.
FUNDING
This study was funded by the Coordination of Superior Level Staff Improvement (CAPES) (Finance Code 001).
AUTHORS’ CONTRIBUTIONS
M.P.B.B.: draft of the article, data collection and interpretation, and methodology; F.S.C.: data interpretation and methodology; S.T.A.C. and R.P.S.: supervision, data analysis and interpretation, project administration, revision and editing of the article. All authors contributed to the study conception and design.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
The authors declare there is no conflict.
%20cropped.png?versionId=5945)