Profiling of meropenem-resistant bacteria in a river receiving wastewater effluent from a pharmaceutical industrial unit

The aim of the present study was to understand the seasonal occurrence and diversity of species of meropenem-resistant bacteria in the Gumuncheon river receiving effluents from a pharmaceutical industry in Seoul, Korea.Water samples were collected from the Gumuncheon river in Kyoung-gi province during winter (January), spring (April), summer (August), and fall (November) of 2018. Water samples were plated in triplicate on tryptic soy agar plates containing 16 mg/L meropenem. Meropenem-resistant bacteria were isolated and genetically identified using 16S rRNA analysis. The predominant bacterial genera identified were Elizabethkingia, Pseudomonas, Chryseobacterium and Stenotrophomonas. Among these; Pseudomonas species Pseudomonas chengduensis and Pseudomonas taiwanesis showed resistance against 15 antibiotics. To prevent the occurrence and spread of meropenem-resistant bacteria in rivers, it is necessary to implement methods that can simultaneously kill multi-drug resistant bacteria and remove antibioticsfrom pharmaceutical industry effluent discharge. Further, to stop the spread of meropenem-resistant bacteria in environment, effluent discharge water should be stringently assessed for their risk of being an environmental hazard.

Carbapenem antibiotics enter GNB through outer membrane proteins (OMPs) known as porins (Martinez-Marinez, 2008). Carbapenem antibiotics pass through the periplasmic space and subsequently inhibit peptide cross-linking by permanent acylation of penicillin-binding proteins (PBPs). PBPs facilitate the synthesis of peptidoglycans present in the bacterial cell wall (Hashizume et al., 1984). Cell wall synthesis is a dynamic three-dimensional process with synthesis and autolysis occurring simultaneously. As a result, inhibition of PBPs causes weakening of peptidoglycans, thereby resulting in cell rupture due to osmotic pressure (Van Dam et al., 2009).
The pharmaceutical industry produces effluent discharge which carries various types of antibiotics that are not degraded during the wastewater treatment process. Antibiotics that are discharged without being completely degraded can cause disturbances in aquatic ecosystems and will be hazardous for human life (Kim and Kim, 2016). Pharmaceutical industry wastewater treatment plants may also discharge wastewater directly into rivers if they satisfy the effluent standards for general metrics such as biochemical oxygen demand (BOD), chemical oxygen demand (COD), suspended solids (SS), total nitrogen (T-N), and total phosphorus (T-P), similar to non-pharmaceutical wastewater treatment plants (Hwang and Kim, 2018).
The purpose of this study was to identify the seasonal frequency of occurrence, species variety, and antibiotic resistance spectrum of meropenem-resistant bacteria in rivers receiving wastewater discharged from a pharmaceutical industrial park.

Collection area and methods
Gumuncheon river; from where the study samples were collected, is a tributary of the Balancheon river. The Hyangnam Pharmaceutical Industrial Park, located in Hwaseong-si (Gyeonggi Province, South Korea), houses 10 pharmaceutical companies that discharge wastewater into the Gumuncheon river. We sampled 1L of river water at 37.097598N and 126.902025 E in January, April, August, and November of 2018. Samples were collected in sterile collection bottles and refrigerated during transportation to the laboratory (Hwang et al., 2018).
Water temperature, dissolved oxygen (DO), and the pH of the river water were measured using a DO/pH meter (DH-32P, Toa, Japan). Prior to analyzing T-N, T-P, and total organic carbon (TOC), a 0.45 μm pore filter (Advnatec, Japan) was used to remove any suspended solids and temperature, DO, and pH were measured at the sampling point. T-N and T-P were measured using Integral Futura continuous flow analyzer (Alliance, USA), while TOC was measured using TOC-L (Shimadzu, Japan). BOD and COD were analyzed in accordance with the water pollution standard method (NIER, 2018).

Isolation of heterotrophic bacteria
To measure the heterotrophic bacterial count within each sample, the samples were diluted 100-fold using sterilized saline solution and 1.0 ml from each dilution each was spread 3 times on 3 agar plates (pour plate agar, Difco, USA) and incubated at 35°C for 48 h.

Identification of meropenem-resistant bacteria and antibiotic resistance test
To identify the meropenem-resistant bacteria, samples were diluted by 10-fold using sterilized saline solution and 0.1 ml each was spread on three tryptic soy (TS) agar plates (Difco, USA) containing 16 mg/L of meropenem (Daewoong Pharmaceutical, Korea) and cultured at 35°C for 48 h. Meropenem-resistant bacteria were subjected to 16S rRNA analysis for species identification. Colony polymerase chain reaction (PCR) was used to amplify the 16S rRNA gene of resistant bacteria from pure colonies (Kim and Kim, 2016). To determine the antibiotic sensitivities of the identified bacteria, disk diffusion assays were performed with ampicillin, ceftizoxime, vancomycin, imipenem, clindamycin, gentamicin, erythromycin, ciprofloxacin, nitrofurans, rifampin, ampicillin/ sulbactam, aztreonam, spectinomycin, trimethoprim, tetracycline, and chloramphenicol. The diameter of the inhibition zone was measured in mm and interpreted according to the Clinical Laboratory Standards Institute (CLSI) guidelines (CLSI, 2013;Hwang et al., 2018).
The numbers of heterotrophic and meropenemresistant bacteria identified are shown in Table 2. The counts of heterotrophic bacteria in the winter, spring, summer, and fall seasons were 3.4, 4.0, 5.6 and 5.1 (×10 4 CFU/ml respectively, while the numbers of meropenem-resistant bacteria were 14.0, 8.3, 1.2 and 9.0 (×10 2 CFU/ml) respectively.
The prevalence of each strain by season is shown in   Table 3. The Elizabethkingia genus was the dominant genus in the winter season, and was also partially identified in the spring and fall when the water temperature was low; however, it was not identified in summer. The Pseudomonas genus was not identified in the winter and summer but was identified during the spring and the fall. The Chryseobacterium genus was identified as the dominant genus in the spring, summer, and winter; the Stenotrophomonas genus was also identified in the same periods. The Cupriavidus, Acinetobacter, and Pandoraea genera were identified in the summer season only. The results of the antibiotic resistance tests of the isolated meropenem-resistant bacteria are shown in Tables 4, 5, and 6. The Elizabethkingia genus commonly showed resistance to eight antibiotics (ampicillin, ceftizoxime, vancomycin, imipenem, meropenem, nitrofuratoin, aztreonam, and trimethoprim), while the Pseudomonas genus showed resistance to thirteen antibiotics (ampicillin, ceftizoxime, vancomycin, imipenem, meropenem, clindamycin, erythromycin, nitrofurantoin, rifampin, ampicillin/sulbactam, trimethoprim, tetracycline, and chloramphenicol). The Chryseobacterium genus, which was the dominant genus in the spring, summer, and fall, showed resistance to five antibiotics (colistin, ampicillin, ceftizoxime, ampicillin/sulbactam, and aztreonam). The S. maltophilia, which appeared in the spring, summer, and fall, showed resistance to ten antibiotics (ampicillin, ceftizoxime, vancomycin, imipenem, meropenem, clindamycin, ampicillin/sulbactam, spectinomycin, aztreonam, and trimethoprim). Among bacteria that appeared only in the summer season, C. plantarum showed resistance to seven antibiotics (ampicillin, vancomycin, imipenem, meropenem, clindamycin, gentamicin, and aztreonam), A. junii showed resistance to nine antibiotics (ampicillin, ceftizoxime, vancomycin, imipenem, meropenem, clindamycin, nitrofurantoin, rifampin, and trimethoprim), and P. pnomenusa showed resistance to eleven antibiotics (colistin, ampicillin, vancomycin, meropenem, clindamycin, gentamicin, erythromycin, nitrofurantoin, rifampin, spectinomycin, and aztreonam).

DISCUSSION
We evaluated the occurrence and season-wise prevalence of meropenem-resistant bacteria in Gumuncheon river receiving wastewater effluents from a pharmaceutical industrial unit. The average water temperature during the winter was 7.5°C, which was lower than the temperature of other sessions for the river.   100  100  100  100  Ceftizoxime  100  100  100  100  Vancomycin  21  100  100  100  Imipenem  100  100  100  100  Meropenem  100  100  100  100  Clindamycin  57  0  0  100  Gentamicin  100  0  100  33  Erythromycin  79  0  100  100  Ciprofloxacin  57  0  100  100  Nitrofurantoin  100  100  100  100  Rifampin  0  0  0  100  Ampicillin/Sulbactam  100  0  100  100  Aztreonam  100  100  100  100  Spectinomycin  43  0  100  100  Trimethoprim  57  100  100  100  Tetracycline  0  0  0  100  Chloramphenicol  0  0  0  100 The increase in water temperature was assumed to be the result of an inflow of effluents from the sewage treatment plant. The concentration of T-P, a limiting nutrient for microbial growth on water surface, was found to be 0.103-0.182 mg/L, which was exceedingly higher than the concentration needed for microbial growth (100 μg T-P/L). It showed that the sampled river water satisfied the conditions for growth of antibiotic resistant bacteria (Correll, 1999). Our results clearly show that the present allowed limit of T-P concentration (4 mg/L) for pharmaceutical industrial wastewater treatment plant effluents should be lowered, considering the additional  amount of T-P that could be introduced into the river from nearby farmland or domestic sewage (Hwang and Kim, 2018). The percentage of meropenem-resistant bacteria among the heterotrophic bacteria identified in the river samples was 1.18-2.14%, which was higher than the percentage measured in a similar river which did not receive the wastewater effluents (Kim and Kim, 2015). It is known that as the number of non-pathogenic bacteria increases, the number of antibiotic-resistant bacteria could also increase through gene transfer from pathogenic to non-pathogenic bacteria (Levy and Marshall, 2004).
The Elizabethkingia genus contains Gram-negative, obligate aerobic bacillus species, and is an emerging healthcare threat as it has been reported to be associated with various life-threatening infections including sepsis, neonatal meningitis, and nosocomial pneumonia. Moreover, improperly processed animalderived food and companion animals are known reservoirs for this antibiotic resistant bacterial pathogen. Elizabethkingia anophelis isolated in clinical practice (Figueroa Castro et al., 2017;Lee et al., 2021) and E. anophelis isolated from horses (Johnson et al., 2018) were reportedly susceptible to ciprofloxacin; however, the E. anophelis isolated in the present study was ciprofloxacin-resistant. Further studies are required to determine whether this difference was due to mutation of one or more genes involved in bacterial DNA separation (Drlica and Zhao, 1997). Similarly, E. meningoseptica isolated from clinical samples has shown resistance to various β-lactams and colistin, but is reported to be susceptible to vancomycin (Ratnamani and Rao, 2013). However, the E. meningoseptica isolated in the present study was resistant to not only ampicillin, imipenem, meropenem, and the β-lactam/β-lactamase inhibitor ampicillin/Sulbactam, but also to colistin and vancomycin. On the other hand, the E. meningoseptica isolated in the present study showed greater susceptibility to tetracycline, clindamycin, erythromycin, and gentamicin than E. meningoseptica from hospital effluents described previously. Thus, more in-depth studies are requiredto determine whether this difference could be attributed to the difference in their source (NIER, 2013;Gullberg et al., 2011). E. miricola, which is an opportunistic oral pathogen, has exhibited resistance to many antibiotics, including imipenem, meropenem, carbapenem, colistin, and gentamicin (Zdziarski et al., 2017;Howard et al., 2020); the E. miricola isolated in the present study also showed resistance to those antibiotics.
Pseudomonas pseudoalcaligenes, another member of the Pseudomonas genus, showed resistance to six antibiotics (ampicillin, ceftizoxime, vancomycin, clindamycin, erythromycin, and chloramphenicol), similar to P. pseudoalcaligenes strains isolated from hospital effluents.
Stenotrophomonas maltophilia, isolated in the spring, summer, and fall, is strongly associated with human respiratory infection. It is a multi-drug resistant bacterium commonly found in environment. Due to its low permeability, it is known to be resistant to cephem antibiotics, such as cefepime and ceftazidime, as well as to β-lactams. Moreover, presence of genes encoding βlactamases, multi-drug resistant efflux pumps, and antibiotic-modifying enzymes confers it with resistance to various antibiotics (Brook, 2012;Adegoka et al., 2017). Similarly, the S. maltophilia isolated in the present study showed resistance to not only β-lactam and cephem antibiotics, such as ampicillin and ceftizoxime, but also to vancomycin, imipenem, meropenem, clindamycin, ampicillin/sulbactam, spectinomycin, aztreonam, and trimethoprim.
Cupriavidus plantarum has been isolated from the plant rhizosphere (Estrada-de Los Santos et al., 2014), but there are no known reports of this bacteria in clinical samples. The C. plantarum identified in the present study showed resistance toampicillin, vancomycin, imipenem, meropenem, clindamycin, gentamicin, and aztreonam.
Biofilms, sewage effluent sediments, wastewater treatment plant effluents, sewage sludge, pharmaceutical manufacturing plants, liquid manure tanks, and manurefertilized soil are known to act as hot spots for the occurrence and spread of antibiotic resistance (Berkner et al., 2014). In South Korea, the industrial effluents are generally checked for total coliforms and ecotoxicity according to the classification of effluent discharge zones. Our results show that there is a need to perform toxicity tests and environmental hazard assessments on effluents containing antibiotics or antimicrobial substances, such as pharmaceutical industry discharge water, to prevent the occurrence and spread of antibiotic resistance (Hernando et al., 2006;Escher et al., 2011). In Korea, the ecotoxicity testing is performed by using water fleas (Hwang and Kim, 2018). It is believed that the toxicity testing of the effluents that include trace amounts of antimicrobial substances, too should be performed using microorganisms. The pharmaceutical industrial park that discharges wastewater into the river sampled in the present study has a wastewater treatment facility; however, there is a high probability of occurrence of antibiotic resistance due to an inflow of processed water containing various clinical disinfectants and discharge from antibiotic-manufacturing pharmaceutical factories. In general, the concentrations of antibiotics present in pharmaceutical effluents are significantly lower than the ones used in clinical practice (NIER, 2013). These concentrations, though not high enough to kill bacteria, are sufficient to exert selective pressure on bacteria to develop antimicrobial resistance (Gullberg et al., 2011). Therefore, it is necessary to implement proper treatment methods, including membrane filtration, ozonation, and UV disinfection, to completely remove the antibiotics and antibiotic-resistant bacteria present in pharmaceutical effluents to reduce their flow into the environment (Pruden et al., 2013).

Conclusion
The present study identified the species, contamination level, seasonal distribution, and antibiotic-resistance spectrum of meropenem-resistant bacteria in a river receiving pharmaceutical industry discharge. Our study showed that the presence of multi-drug resistant bacteria in the river water poses a threat to human health due to wider reach and use of river water. The outcomes of our study highlight the need to implement methods that can simultaneously disinfect multi-drug resistant bacteria and remove antibiotics from effluent containing discharge water from pharmaceutical and industrial units.