RESEARCH AR T I C L E Recovery of clinically relevant multidrug-resistant Klebsiella pneumoniae lineages from wastewater in Kumasi Metropolis, Ghana Amen Ekhosuehi1 | Odion O. Ikhimiukor2,3 | Helen Michelle Korkor Essandoh1,4 | Nana Yaw Asiedu5 | Isoken Tito Aighewi1 | Gabriel Temitope Sunmonu2 | Erkison Ewomazino Odih2 | Anderson O. Oaikhena2 | Dorothy Cyril-Okoh2 | Clara Yeboah6 | Iruka N. Okeke2 1Regional Water and Environmental Sanitation Centre, Kumasi, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana 2Department of Pharmaceutical Microbiology, Faculty of Pharmacy, University of Ibadan, Ibadan, Nigeria 3Department of Biological Sciences, University at Albany, State University of New York, Albany, New York, USA 4Department of Civil Engineering, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana 5Department of Chemical Engineering, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana 6Department of Parasitology, Noguchi Memorial Institute for Medical Research, University of Ghana, Accra, Ghana Correspondence Amen Ekhosuehi, Regional Water and Sanitation Centre Kumasi, Deaprtment of Civil Engineering, Kwame Nkrumah University of Science and Technology, Ghana. Email: amen.ekhosuehi@uniben.edu Funding information Bill and Melinda Gates Foundation, Grant/Award Number: INV-036234; World Bank; SEQAFRICA; Department of Health and Social Care’s Fleming Fund Abstract Antimicrobial resistance (AMR) is under-monitored in Africa, with few reports characterizing resistant bacteria from the environment. This study examined physicochemical parameters, chemical contaminants and antibiotic-resistant bacteria in waste stabilization pond effluents, hospital wastewater and domestic wastewater from four sewerage sites in Kumasi. The bacteria isolates were sequenced. Three sites exceeded national guidelines for total suspended solids, biochemical oxygen demand, chemi- cal oxygen demand and electrical conductivity. Although sulfamethoxazole levels were low, the antibiotic was detected at all sites. Multi-drug-resistant Klebsiella pneumoniae and Pseudomonas aeruginosa were isolated with multi-locus sequence typing identifying K. pneumoniae strains as ST18 and ST147, and P. aeruginosa as ST235, all of clinical relevance. A comparison of ST147 genomes with isolates from human infections in Africa showed remarkable similarity and shared AMR profiles. Thirteen of the twenty-one plasmids from ST147 harbored at least one AMR gene, including blaCTX- M-15 linked to copper-resistance genes. Our study demonstrated high bac- terial counts and organic matter in the analysed wastewater. The recovery of clinically significant isolates with multiple antibiotic and heavy metal resis- tance genes from the wastewater samples raises public health concerns. INTRODUCTION Addressing antimicrobial resistance (AMR) requires a comprehensive understanding of its scope, particularlyAmen Ekhosuehi and Odion O. Ikhimiukor contributed equally. Received: 6 March 2024 Accepted: 10 September 2024 DOI: 10.1111/1758-2229.70018 ENVIRONMENTAL MICROBIOLOGY REPORTS This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2024 The Author(s). Environmental Microbiology Reports published by John Wiley & Sons Ltd. Environmental Microbiology Reports. 2024;16:e70018. wileyonlinelibrary.com/journal/emi4 1 of 16 https://doi.org/10.1111/1758-2229.70018 https://orcid.org/0000-0003-2033-8810 https://orcid.org/0000-0002-3738-4584 https://orcid.org/0000-0002-1694-7587 mailto:amen.ekhosuehi@uniben.edu http://creativecommons.org/licenses/by/4.0/ http://wileyonlinelibrary.com/journal/emi4 https://doi.org/10.1111/1758-2229.70018 http://crossmark.crossref.org/dialog/?doi=10.1111%2F1758-2229.70018&domain=pdf&date_stamp=2024-11-08 in low-income settings, where the burden is severe (Ikhimiukor et al., 2022; Okeke et al., 2024). The World Health Organization’s (WHO) Global Action Plan has prompted many countries to enhance AMR surveillance and return data to the WHO Global Antimicrobial Resis- tance and Use Surveillance System (GLASS). In many African settings, where resources are limited, surveil- lance primarily targets human clinical isolates, focusing on WHO priority pathogens, including the ESKAPE-E organisms: Enterococcus feacium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter bauman- nii, Pseudomonas aeruginosa, Enterobacter spp. and Escherichia coli. These pathogens are categorized by the urgency of new antibiotics development into critical, high and medium groups (Tacconelli et al., 2018). The surveillance often leverages on diagnostic microbiology at sentinel sites to identify these and other pathogens from sick patients. To effectively combat the spread of priority patho- gens, surveillance needs to extend beyond human clini- cal settings to include a holistic examination of antibiotic resistance genes and resistant isolates within a One Health context. Despite recognizing this need, AMR monitoring in low and middle-income countries (LMICs) remains insufficient, typically skewed towards human and, to a lesser extent, animal health sectors (Ikhimiukor & Okeke, 2023; Munk et al., 2022; Okeke et al., 2024). In Ghana, for example, AMR monitoring is predominantly focused on human and animal samples, with limited research on environmental isolates (Osei Sekyere & Reta, 2020; Yevutsey et al., 2017). Further- more, few studies or surveillance programs employ whole genome sequencing (WGS), which can provide detailed insights into the genetic basis of AMR and pathogen evolution (Vegyari et al., 2020). Therefore, characterizing the presence and distribution of AMR genes in environmental settings is increasingly crucial for effective AMR management. Kumasi, Ghana’s second largest city, grapples with significant challenges due to poorly maintained and dilapidated sanitation systems (UNICEF, 2016). Like many cities in LMICs, Kumasi’s wastewater treatment infrastructure is inadequate, resulting in the discharge of untreated and poorly treated effluents into the envi- ronment. Wastewater represents a critical environment for the evolution of AMR, as it harbours a complex mix of pathogens, commensals, organic matter, heavy metals, nutrients, chemicals and antibiotic residues. While empirical evidence directly linking antibiotic resi- dues pollution to proliferation of AMR genes in waste- water is scarce and challenging to obtain, the accumulation of these residues, coupled with favour- able physicochemical conditions such as high nutrient levels and abundant microorganisms, likely create selective pressure that fosters the survival and prolifer- ation of antibiotic resistant bacteria (Martínez, 2008; Osi et al., 2019). Additionally, the presence of heavy metals and disinfectants can also exert selective pres- sure on wastewater flora, complicating the assessment of antibiotic residues’ specific impact on resistance genes development (Stanton et al., 2022; Tello et al., 2012). Some studies have demonstrated correla- tions between concentrations of antibiotic residues and the prevalence of antibiotic resistance genes. For instance, Kristiansson et al. (2011) discovered high levels of antibiotics, as well as elevated levels of resis- tomes and mobilomes, in river sediments receiving effluents from pharmaceutical industries. Similarly, Tello et al. (2012) used models to predict that concen- trations of ciprofloxacin, erythromycin and tetracycline in river sediments and swine faeces could inhibit wild- type bacterial populations by 60%–92%, thus favouring resistant bacteria that can colonize or infect humans. These data strongly suggest that wastewater can pro- mote the selection of antibiotic resistant bacteria and genes, highlighting the need to track wastewater and other selection hotspots to support policy-making (Booth et al., 2020; Fouz et al., 2020). Priority pathogens that are opportunistic pathogens, such as Klebsiella, are well known sources and sinks of mobile resistance genes that can be transferred to other organisms (Wyres & Holt, 2018). This study assessed sulfamethoxazole (SMX) residues and inves- tigated the presence of resistant bacteria and identified clinically relevant K. pneumoniae and P. aeruginosa species in raw wastewater and waste stabilization pond effluents in Kumasi, Ghana. Additionally, we measured selected physicochemical parameters of the raw waste- water and waste stabilization pond effluents. EXPERIMENTAL PROCEDURES Study site description In this study, treated wastewater samples were col- lected from Asafo (6.67336N, �1.61305E) and Chira- patre (6.654105N, �1.57824E) waste stabilization ponds (WSPs). Additionally, untreated wastewater samples were collected from the sewers of Komfo- Anokye Teaching Hospital (KATH) (6.689105N, �1.62716E) and Kwame Nkrumah University of Sci- ence and Technology’s sewage treatment plant (KSTP) (6.66833N, �1.5756E) before their discharge into a wetland (Figure 1). The selection of these sites was based on the presence of sewers infrastructure and accessibility. Asafo WSP (AWSP) is situated in a densely popu- lated, low-income area in Kumasi. Originally designed in 1994 to serve 320 households with an estimated population of around 20,000, the AWSP initially com- prised two anaerobic ponds, one facultative pond and two maturation ponds. In nearly 30 years of operation, the facility has experienced some minor breakdowns. 2 of 16 EKHOSUEHI ET AL.ENVIRONMENTAL MICROBIOLOGY REPORTS 17582229, 2024, 6, D ow nloaded from https://envirom icro-journals.onlinelibrary.w iley.com /doi/10.1111/1758-2229.70018 by U niversity of G hana - A ccra, W iley O nline L ibrary on [04/03/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense Recent upgrades, including the addition of a second anaerobic pond to accommodate new connections from Kumasi Polytechnic hostels, reflect its ongoing adapta- tion (Salifu, 2013). AWSP discharges its treated efflu- ent into the Subin river which flows through the city centre and joins the Oda river where it supports irriga- tion for local farming communities (Azanu et al., 2018). The Chirapatre WSP (CWSP) receives both sewage and household contaminants. Initially designed for 300 households with an estimated population of about 1800, CWSP has been expanded due to community growth (Darko et al., 2016; Tenkorang et al., 2012). It currently consists of an anaerobic pond, two facultative ponds and two maturation ponds. The latter of which are used for aquaculture. Effluents from CWSP flows into a nearby stream (Table 1). KATH is the only teaching hospital in Kumasi with a capacity of 1200 beds (Azanu et al., 2018). It channels its sewage through a network of sewers that ultimately discharges into a natural wetland (see Table 1). Mean- while, KSTP receives mainly sewage from eight halls of residence and annexes on the campus, serving over 7000 students (Awuah, 2014). Notably, during the sam- pling period, the trickling filter was undergoing mainte- nance, necessitating the rerouting of wastewater to a natural wetland. F I GURE 1 Study area (A) Ghana, (B) Ashanti region and (C) Kumasi metropolitan district. TAB LE 1 Site characteristics. Sites Wastewater contributors Sampled effluent Discharge point AWSP 20,000 households Treated River CWSP 300 households Treated Stream KATH KATH sewage Raw Natural wetland KSTP Student halls Raw Natural wetland Abbreviations: AWSP, Asafo waste stabilization pond; CWSP, Chirapatre waste stabilization pond; KATH, sewage from Komfo Anokye Teaching Hospital Sewers; KSTP, sewage from Kwame Nkrumah University of Science and Technology’s Sewage treatment plant. RESISTANT KLEBSIELLA IN KUMASI WASTEWATER 3 of 16ENVIRONMENTAL MICROBIOLOGY REPORTS 17582229, 2024, 6, D ow nloaded from https://envirom icro-journals.onlinelibrary.w iley.com /doi/10.1111/1758-2229.70018 by U niversity of G hana - A ccra, W iley O nline L ibrary on [04/03/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense Sampling and sample collection Flow-proportional composite samples were manually collected for chemical and physicochemical analysis. To achieve this, three 6-hourly aliquots were taken, each calculated based on flowrate at the time of sam- pling. These aliquots were then combined to create a 1 L sample for the analysis. The generated samples were kept on ice and transported to the laboratory within 12 h for processing. For microbiological analysis, 500 mL grab samples were collected in sterile autoclavable bottles. Duplicate samples were collected monthly from each sampling sites, resulting in a total of 120 samples collected between July and December 2019. Chemical and physicochemical analysis SMX residues were analysed using solid phase extrac- tion (SPE) followed by High Performance Liquid Chro- matography coupled with a diode array detector (HPLC-DAD) (Cecil Adept Binary Pump HPLC with WaveQuest DAD Detector). The analytical investigation adhered to the procedures outlined by Zhou and Jiang (2014) with some modifications. Wastewater samples (1 L) were first filtered using 240 mm Whatmann filter paper before undergoing cleanup via SPE. The hydrophilic–lipophilic balance (HLB) sorbents (Oasis HLB, 6 cc/500 mg, Waters) were preconditioned with 10 mL methanol and equilibrated with 15 mL distilled water. The prepared wastewater samples were then loaded onto the HLB sorbent fitted on a manifold con- nected to a vacuum pump. Elution of the sorbents was carried out using 10 mL methanol collected into 15 mL centrifuge tubes and dried under a nitrogen stream. The dried eluents were reconstituted with 2 mL metha- nol and vortexed at 3500 rpm for 10 min before analy- sis by HPLC-DAD. The mobile phase constituted 70% of acetonitrile (A) and 30% of 0.1% formic acid in distilled water (B). Column temperature was maintained at 35�C with a pump flow rate of 1 mL/min and detection was per- formed at a wavelength of 280 nm. Fifty microliters of the samples were injected into a reversed phase sta- tionary phase column (Phenomenex, Synergi MAX-RP 150 � 4.60, 4 μm). Additionally, samples were also analysed for pH and electrical conductivity (EC) on-site using a multi-parameter (Pc Testr 35, Eutech Instru- ment). Total suspended solids (TSS) were determined according to American Public Health Association’s standards methods for examination of Wastewater (APHA, AWWA, WEF, 2017). Chemical oxygen demand (COD), total phosphorus (TP) and total nitro- gen (TN) were determined using a spectrophotometer (HACH, DR 3900). Biochemical oxygen demand (BOD) was assessed using the manometric method (Velp scientifica BOD Evo Sensor and Lovibond BOD system). Chemical method validation Chemical method validation was conducted in accor- dance with standard guidelines (Booth & Simon, 2016; International Council for Harmonisation, 2022). A 5-point calibration curve for SMX (Sigma Aldrich) was constructed with concentrations ranging from 0.44 mg/L to 7.04 mg/L, yielding a regression coefficient of 0.99%. The limit of quantification (LOQ) and limit of detection (LOD) were calculated using a signal- to-noise ratio of 10 and 3 respectively. The calculated LOD and LOQ values for SMX were 0.02 mg/L and 0.1 mg/L respectively. Absolute recovery studies were also conducted using 1 L wastewater and 1 L distilled water in duplicates at the lowest and highest concentra- tions within the method calibration ranges. The abso- lute recovery was calculated as the ratio of the peak areas from wastewater to distilled water samples, were 68.75% and 60.97% respectively. Isolation and characterization of bacteria isolates Total heterotrophic counts (THC) were determined according to American Public Health Association’s standards methods for examination of Wastewater (APHA, AWWA, WEF, 2017). Samples were mixed to ensure uniform microbial dispersion, and serial dilutions ranging from 10�2 to 10�6 were prepared in sterile dis- tilled water due to the high microbial load in sewage- impacted wastewater. The diluted samples were then spread onto Plate Count Agar (PCA, Oxoid) in triplicate and incubated at 37�C for 24–48 h. Plates containing 25–300 colonies were used to compute the microbial counts. Single and separate colonies from each plate were subsequently sub-cultured onto PCA and later into nutrient broth for cryopreservation. Isolates were subjected to Gram staining and a selected subset were further identified using the Gram-negative (GN) test kit (Ref: 21341) on the VITEK 2 system (ver- sion 2.0, Marcy-l’Etoile, France, Biomérieux). Isolates that had low-confidence identities or were unidentified on VITEK2, as well as Gram positive strains, under- went 16S rRNA gene sequencing. DNA extraction was performed using MP Biomedicals FastDNA™ spin kit for soil following the manufacturer’s protocol. Polymer- ase chain reaction (PCR) amplification of the 16S rRNA gene was conducted using 16S_1492r (5-0GGTTACC TTGTTAGACTT-30) and 16S_27f (50GAGAGTTTGAT CCTGGCTCAG-30) primers designed by Turner et al. (1999) and Lane (1991) respectively. The PCR amplicons were visualized on a 1% agarose gel 4 of 16 EKHOSUEHI ET AL.ENVIRONMENTAL MICROBIOLOGY REPORTS 17582229, 2024, 6, D ow nloaded from https://envirom icro-journals.onlinelibrary.w iley.com /doi/10.1111/1758-2229.70018 by U niversity of G hana - A ccra, W iley O nline L ibrary on [04/03/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense (CSL-AG500, Cleaver Scientific Ltd.) stained with EZ- vision® Bluelight DNA Dye and then Sanger-sequenced using the Applied Biosystems ABI 3500XL Genetic Analyser. The sequencing outputs were converted from ABI chromatogram to FASTA format using the DNA Baser Assembler (https://www.dnabaser.com/ download/download.html) and the sequences were assembled using the BioEdit Sequence Alignment Edi- tor (https://bioedit.software.informer.com/7.2/). Bacteria identification was conducted by comparing the sequences to the 16S rRNA/ITS database nucleotide database of the National Center for Biotechnology Information (NCBI) using the BLASTN tool (https:// blast.ncbi.nlm.nih.gov/Blast.cgi). Isolates from genera commonly associated with human infections were sub- jected to antimicrobial susceptibility testing and WGS. Antimicrobial susceptibility testing Antimicrobial susceptibility testing was performed on eight clinically relevant isolates, comprising K. pneumoniae (n = 7) and P. aeruginosa (n = 1), using Biomérieux VITEK AST N280 cards (Ref: 413432) following the man- ufacturer’s instructions. The VITEK AST N280 cards include a panel of 85 antimicrobials, such as imipenem- relebactam, meropenem-vaborbactam and an extended- spectrum beta-lactamase (ESBL) test. As the aim was to determine clinical significance of resistances, minimum inhibitory concentrations obtained for each antibiotic were interpreted using the breakpoints established by the Clini- cal and Laboratory Standards Institute (CLSI, 2021). Whole genome sequencing DNA from K. pneumoniae (n = 7) and P. aeruginosa (n = 1) identified as clinically significant priority patho- gens via VITEK2, was extracted using the Wizard DNA extraction kit (Promega; Wisconsin, USA) following the manufacturer’s instructions. Quantification of the extracted DNA was performed on a Qubit fluorome- ter (Invitrogen; California, USA) utilizing the dsDNA broad range assay. Double-stranded DNA libraries were prepared using NEBNext Ultra II FS DNA library prep kit for Illumina incorporating 96 unique indexes (New England Biolabs, Massachusetts, USA; Cat. No: E6609L). Subsequent quantification of the DNA librar- ies employed the dsDNA High Sensitivity assay on a Qubit fluorometer, and average fragment length of the DNA libraries were determined using 2100 Bioanalyzer (Agilent). The libraries were sequenced using 150 bp paired- end chemistry on an Illumina MiSeq (Illumina, Califor- nia, USA). Sequence reads were assembled using the GHRU assembly pipeline (https://gitlab.com/cgps/ghru/ pipelines/dsl2/pipelines/assembly) implemented in a Nextflow workflow. Briefly, the reads were trimmed using trimmomatric v0.39 (Bolger et al., 2014), and Cutadapt v3.2 (https://github.com/marcelm/cutadapt) was used to remove adapter sequences from the reads. Trimmed reads were de novo assembled with SPAdes v3.15.3 (Bankevich et al., 2012) and the qual- ity of the assembled genomes was evaluated using QUAST v5.0.2 (Gurevich et al., 2013), ConFindr v0.7.2 (Low et al., 2019), qualifyr v1.4.4 (https://gitlab.com/ cgps/qualifyr). Genomes with contamination levels <5%, number of contigs <300 and N50 > 25,000 were subjected to downstream analyses. Strain identities were confirmed using Bactinspector v0.1.3 (https:// gitlab.com/antunderwood/bactinspector). Klebsiella typing and determination of genetic determinants for AMR and virulence We used Kleborate v2.2.0 (Lam et al., 2021) to confirm Klebsiella species identities and determine multi-locus sequence types. Determination of Klebsiella K- and O-locus types were done using Kaptive v2.0.0 (Lam et al., 2021). Genotypic mechanisms of AMR were determined by screening the genome assemblies for the presence of AMR genes using AMRFinderPlus v.3.10.23 (Feldgarden et al., 2021) and its accompany- ing NCBI AMR database. The presence of genetic determinants for virulence in the genomes was detected by following GHRU protocols implemented on a nextflow workflow (https://gitlab.com/cgps/ghru/ pipelines). Plasmid replicons and plasmid sequence reconstruction Plasmid replicons in the genomes were detected using PlasmidFinder implemented in a Nextflow workflow (https://gitlab.com/cgps/ghru/pipelines). To investigate the potential presence of AMR genes on plasmids, we used the mob-recon tool available in the MOB-suites software to reconstruct plasmid sequences from the draft genome assemblies (Robertson & Nash, 2018). The plasmid contigs generated were used as input to determine presence of AMR genes and heavy metal resistance genes using AMRFinderPlus v3.10.23 (Feldgarden et al., 2021). Phylogenetic analysis of genomes To elucidate the evolutionary relationships between our strains and others from same clonal lineage, we downloaded all publicly available K. pneumoniae ST147 genomes (n = 29) from Pathogenwatch RESISTANT KLEBSIELLA IN KUMASI WASTEWATER 5 of 16ENVIRONMENTAL MICROBIOLOGY REPORTS 17582229, 2024, 6, D ow nloaded from https://envirom icro-journals.onlinelibrary.w iley.com /doi/10.1111/1758-2229.70018 by U niversity of G hana - A ccra, W iley O nline L ibrary on [04/03/2025]. See the T erm s and C onditions (https://onlinelibrary.w iley.com /term s-and-conditions) on W iley O nline L ibrary for rules of use; O A articles are governed by the applicable C reative C om m ons L icense https://www.dnabaser.com/download/download.html https://www.dnabaser.com/download/download.html https://bioedit.software.informer.com/7.2/ https://blast.ncbi.nlm.nih.gov/Blast.cgi https://blast.ncbi.nlm.nih.gov/Blast.cgi https://gitlab.com/cgps/ghru/pipelines/dsl2/pipelines/assembly https://gitlab.com/cgps/ghru/pipelines/dsl2/pipelines/assembly https://github.com/marcelm/cutadapt https://gitlab.com/cgps/qualifyr https://gitlab.com/cgps/qualifyr https://gitlab.com/antunderwood/bactinspector https://gitlab.com/antunderwood/bactinspector https://gitlab.com/cgps/ghru/pipelines https://gitlab.com/cgps/ghru/pipelines https://gitlab.com/cgps/ghru/pipelines (https://pathogen.watch/), including isolates from human (n = 28) and environmental (n = 1) sources across Africa (Algeria = 2, Egypt = 5, Kenya = 4, Nigeria = 13, South Africa = 1 and Zambia = 4) (Argim�on et al., 2021). For K. pneumoniae ST18, we included all available genomes from Pathogenwatch in our phylogeny analysis. The genomes were first annotated using Prokka v1.14.6 (Seemann, 2014) and then subjected to pangenomes analysis using Panaroo v1.2.7 (Tonkin-Hill et al., 2020). Multiple sequence alignment was performed with MAFTT v7.4.472 (Katoh & Standley, 2013) to generate a core genome alignment. Single nucleotide polymorphisms (SNPs) were extracted from this alignment using snp- sites v2.5.1 (Page et al., 2016) and subsequently used in RAXML v8.2.12 (Stamatakis, 2014) to construct a phylo- genetic tree employing GTR nucleotide substitution model and GAMMA distribution of heterogeneity. Pairwise SNP distances were calculated using snp-dists v0.8.2 (https:// github.com/tseemann/snp-dists), and the phylogeny was visualized on iTOL (Letunic & Bork, 2019). Statistical analysis All data were entered in an Excel Spreadsheet and analysed using RStudio 2022.02.1. To compare the number of genetic determinants such as plasmid repli- cons, AMR and virulence genes identified in the K. pneumoniae ST147 strains from our study with those detected in human associated K. pneumoniae ST147 strains across Africa, we applied the Wilcoxon Rank Sum test. Differences in selected physicochemical parameters across various locations were evaluated using One Way Analysis of Variance (ANOVA). Where applicable, post-hoc tests were conducted, with signifi- cance determined at p < 0.05 using SPSS software (version 25). RESULTS Wastewater pollution levels in sampling sites The mean physicochemical parameters for all sampled sites are presented in Table S1. AWSP, KSTP and KATH demonstrated comparable physicochemical pro- files, with no significant differences in pH, BOD/COD ratio, BOD, COD, EC, TN and TSS (p ≤ 0.05). The parameter ranges across the sites were as follows: pH, 7.77 ± 0.11–8.30 ± 0.24; BOD/COD ratio, 0.51 ± 0.12– 0.60 ± 0.03; BOD, 96.17 ± 18.08–506 ± 39.99; COD, 162.41 ± 31.71–883.17 ± 71.28; EC, 1437.5 μS/cm ± 196.62–2074.83 μS/cm ± 241.53; TN, 73.93 mg/L ± 26.14–139.10 mg/L ± 31.06 and TSS, 70.17 mg/L ± 20.62–278 mg/L ± 44.25. When compared to the Ghana Environmental Protection Agency’s (GEPA) limits, BOD, COD, EC, TN and TSS concentrations at AWSP, KSTP and KATH exceeded the respective thresholds of 50 mg/L, 50 mg/L, 1500 μS/cm, 50 mg/L and 50 mg/L. Notably, KSTP recorded the highest TP content of 8.05 mg/L ± 1.21 while TP levels at other sites were relatively the same. In contrast, CWSP dis- played significantly lower values for most examined parameters, including BOD and COD. SMX was consistently detected across all sites throughout the 6 months sampling period, albeit most concentrations were