Plasmid-mediated quinolone resistance determinants in clinical bacterial pathogens isolated from the Western Region of Ghana: a cross-sectional study

Introduction quinolones are critically important antibiotics that are reserved for treating very severe infections caused by multidrug-resistant bacterial pathogens. However, their indiscriminate uses have resulted in an increased number of resistant strains in many parts of the world including Ghana. We determined the quinolone resistance profile of Gram-negative bacterial pathogens and characterized the underlying molecular determinants of resistance. Methods Gram-negative pathogens obtained from clinical specimens at three hospital laboratories were tested for resistance to quinolones and other commonly used antibiotics. ESBL production among the Enterobacterial isolates was confirmed using the combined disc diffusion method. We then used PCR to determine seven types of plasmid-mediated quinolone resistance genes present in the isolates resistant to nalidixic acid and ciprofloxacin. Results in this study, 29.5% of the isolates were resistant to ciprofloxacin, with the highest of 50% among E. coli resistance to the other quinolones was levofloxacin (24.4%), norfloxacin (24.9%), and nalidixic acid (38.9%). Significant proportions of the quinolone-resistant isolates were ESBL producers (P-values < 0.001). The aac(6´)-Ib-cr, qnrS, oqxA, and qepA genes were present in 43 (89.6%), 27 (56.3%), 23 (47.9%), and one (2.1%) of the isolates, respectively. None of the isolates tested positive to qnrA, qnrB, and oqxB genes. The presence of the aac(6´)-Ib-cr gene positively correlated with resistance to ceftriaxone, cefotaxime, and gentamicin (P-values < 0.05). Conclusion high proportions of Gram-negative bacterial isolates were resistant to quinolones and most of these isolates possessed multiple PMQR genes. There is a need to implement measures to limit the spread of these organisms.


Introduction
Antibiotics are essential drugs for controlling and preventing morbidity and mortality from infection caused by bacteria. However, rising resistance levels in nearly all bacteria types that cause illnesses have substantially reduced their effectiveness and threatened a return to the pre-antibiotic era [1]. This has led to increased length of hospital stays, cost of healthcare, morbidity and death [2]. Quinolones are an important group of these essential drugs and were introduced into clinical practices in the mid-1980s. They are well known for their ease of absorption following oral administration, with higher concentrations in the urinary tract. This often makes them the drug of choice for urinary tract infections. Newer generations such as ciprofloxacin are mostly preferred in treating infections such as enteric fevers to other groups which are required to be administered parenterally, including the cephalosporins and aminoglycosides [3]. Quinolones are also one of the major options for the treatment of infections caused by dangerous and multidrug-resistant bacteria [1]. In Ghana, quinolones were introduced for the treatment of bacterial infections due to increased resistance levels in bacterial pathogens to commonly used, less expensive, and readily available antibiotics such as the penicillins, sulfonamides and first and second-generation cephalosporins [1,4]. However, the recurrence of bacterial resistance to this family of antibiotics is limiting their use globally, whiles alternatives are limited. Urgent measures are therefore required to reduce the spread of resistant bacteria. This will help reduce the already existing pressure on health systems in resource-limited countries and promote the achievement of the Sustainable Development Goals (SDGs) [2].
Resistance to the quinolones is generally higher in developing countries. A systematic review in 2015 recorded levels as low as 2% in USA and Ireland, compared to 80% in Nigeria [5]. In Ghana, up to 50% resistance to ciprofloxacin was reported among pathogens obtained from wound and blood cultures [6,7]. In particular, there has been increasing reports of reduced susceptibility of Salmonella species to fluoroquinolones [4,8,9].
Resistance to quinolone antibiotics is mediated by chromosomal mutations and the acquisition of transferable plasmid genes [3,10,11]. Several forms of plasmid-mediated quinolone resistance (PMQR) genes have been described and most of these have been detected in clinical isolates [3]. These genes facilitate the selection and spread of quinolone-resistant bacterial strains rapidly and pose a greater threat to public health [10,12]. Studies in most countries have determined the prevalence and characteristics of the various resistant determinants in clinical isolates. In Ghana, there is a dearth of studies on quinolone resistance determinants in clinical isolates and the few studies are centered and localized to teaching hospital in the two big cities in the country; Accra and Kumasi. This study aimed to determine the molecular determinants that mediate quinolone resistance among Gramnegative bacterial pathogens in the Western Region of Ghana. Data obtained from the study are essential to inform policy on antibiotic stewardship in the region and to monitor the effectiveness of existing interventions.

Methods
Study design, sites, and populations: a prospective cross-sectional study was conducted in the Western Region of Ghana from October 2020 to February 2021. The region is situated in the Southwestern part of Ghana. It shares borders with the Central Region in the East, Western North Region in the North, La Cote d´Ivoire in the West, and the Gulf of Guinea (Atlantic Ocean) in the South [13]. The region is divided into fourteen administrative divisions, including one Metropolitan, seven Municipalities, and six Districts [14]. Clinical specimens of patients conducting microbiology tests at the Sekondi Public Health Laboratory (SPHL), situated in the Western Regional capital; Tarkwa Municipal Hospital (TMH), in a semi-urban city; and Prestea Government Hospital (PGH), in a rural town were obtained for the study. Gramnegative bacterial pathogens isolated from Mid-stream urine, blood and wound swab specimens were included in the study.
Bacterial isolation and identification: the specimen were cultured using conventional techniques [15]. We used cysteine-lactoseelectrolyte-deficient (CLED) agar for culturing the urine specimens. Blood and wound specimens were cultured on MacConkey agar and blood agar. The inoculated plates were incubated aerobically at 35-37°C and examined after 18-24 hours for bacterial growth. For the urine cultures, a count of 10 5 CFU/mL or more bacterial growth was considered significant for further investigations. The bacterial pathogens were identified using conventional methods including their Gram staining reactions, colonial characteristics, and biochemical properties.
In this study, geographical differences in quinolone resistance were observed. Although, isolates from Sekondi Public Health Laboratory and Tarkwa Municipal Hospital showed no significant differences in resistance to the tested quinolones, they had significantly higher levels of resistance to all the four quinolones than isolates from Prestea Government Hospital (P values < 0.05) (data not shown).

Resistance patterns of the isolates to other
antimicrobials: the isolates were tested against eleven other antimicrobials. The highest proportion of the isolates were resistant to trimethoprim sulfamethoxazole 143 (64.7%), followed by amoxicillin clavulanic acid 103 (46.6%), ceftriaxone 95 (42.5%), and cefotaxime 94 (42.5%), the least proportions of resistance were to amikacin 5 (2.0%) and meropenem 12 (4.7%) ( Table 2). The presence of PMQR genes was investigated to determine whether it affects susceptibility to antibiotics other than quinolones (Table 3). We observed that the presence of qnrS and oqxA genes did not correlate with the proportion of the isolates that were resistant to any of the antimicrobials tested (p-values > 0.05). However, the presence of aac(6´)-Ib-cr gene positively correlated to the proportions of the isolates that were resistant to cefotaxime (P-value=0.006), ceftriaxone (p-value=0.018), and gentamicin (p-value=0.020).

Discussion
Antimicrobial resistance has emerged as an important threat to human and animal health worldwide, calling for intensive research to elucidate the complexity of underlying causes and possible ways of reducing, stopping, and possibly reversing the rate of its development [18,19]. In this study, we studied the antibiotic resistance patterns of 254 clinical Gram-negative bacterial isolates collected from three hospitals in the Western Region of Ghana. The presence of plasmid-mediated quinolone resistance genes was determined in selected ciprofloxacin and nalidixic acid-resistant isolates. . These observed differences were probably due to differences in geographical distribution of resistant bacterial strains [34], study population, study design [35], existing drug prescription policies, AMR surveillance, and antimicrobial stewardship programs [26,36]. Infection with a quinolone-resistant and/or ESBL producing organism has significant public health implications. These infections have limited treatment options, making it difficult for clinicians to prescribe successful treatment. They also make empiric prescriptions perilous since they are often associated with poorer treatment outcomes, including increased treatment failure, mortality, and morbidity [26,37]. They also increase the pressure on health facilities, which are often inadequate in our settings, due to increased hospital stays and the need for additional procedures due to complications [38]. Invariably, these translate into a higher social and economic burden on individuals, families, and the community.
In the present study, four PMQR genes were detected in decreasing order: aac(6´)-Ib-cr43 (89.6%), qnrS27 (56.2%), oqxA23 (47.9%), and qepA one (2.1%) among ciprofloxacin and nalidixic acid resistant isolates. The predominance of the aac(6´)-Ib-cr gene is supported by the work of Attipoe et al. involving E. coli isolated from 18 testing laboratories across Ghana. In their work, all 29 isolates tested were positive for the aac(6´)-Ib-cr gene. They also found a larger proportion of E. coli isolates to possess qnrS26 (89.6%), oqxA19 (65.5%), qnrA16 (55.1%), and qnrB15 (55.1%) [39]. Contrarily, the present study did not detect qnrA, qnrB and oqxB in the isolates. This may be due to differences in the geographical distribution of PMQR genes. Other studies in Nigeria [40] and Iran [41] confirmed the predominance of the aac(6´)-Ib-cr gene. These reports indicate the widespread presence of this quinolone-resistance gene.
In this study, no correlation was observed between the presence of qnrS and oqxA genes and the level of resistance to the non-quinolone antibiotics tested. This implies that, the presence of these genes alone may not be enough to affect the efficacy of the non-quinolone antibiotics.
Isolates possessing aac(6´)-Ib-cr genes were found to have significant levels of resistance to cefotaxime (P-value = 0.006), ceftriaxone (P-value = 0.018), and gentamicin (P-value = 0.020) as compared to their aac(6´)-Ib-cr negative counterparts. No correlation was found with the other antibiotics. The enzyme aminoglycoside acetyltransferase, encoded by the aac(6´)-Ib-cr gene is originally known to confer resistance to the aminoglycosides (tobromycin, kanamycin, amikacin and gentamicin). Hence resistance to gentamycin in the presence of aac(6´)-Ib-cr gene is expected. The aac(6´)-Ib-cr gene is found in various integrons, especially on IncF11 plasmids that express CTX-M-15 [12]. CTX-M-15 is a widespread ESBL that confers resistance to third generation cephalosporin such as ceftriaxone and cefotaxime. Hence, isolates with aac(6´)-Ib-cr genes are likely to be resistance to third generation cephalosporin. These may explain the observed association between the presence of aac(6´)-Ib-cr gene and resistance to ceftriaxone and cefotaxime [12].
Limitations of this study: sequencing of PMQR genes identified in the study was not done to determine the allelic forms present in the study areas. Minimum inhibition concentration (MIC) determination for quinolone antibiotics was not done. MIC could have helped to determine if there is difference in levels of resistance to quinolone among the isolates positive and negative for PMQR genes. In addition, this is a laboratory based study so risk factors such as previous admission to hospital, length of stay at the hospital, presence of indwelling devices, and history of antibiotic use could not be assessed.

Conclusion
In this study, high proportion of quinoloneresistant Gram-negative bacterial pathogens possessed multiple horizontally transferable PMQR genes. These genes promote the selection of highly quinolone-resistant bacterial strains that when spread can pose serious threat to public health in the region and Ghana. It is therefore recommended to the regional health authorities and stakeholders to institute measures to ensure adequate antimicrobial stewardship and surveillance, adequate infection prevention and control, and good prescription behavior among health workers to prevent further spread of quinolone resistant bacterial pathogens.

What is known about this topic
 The role of plasmid-mediated quinolone resistance genes in the development of bacterial resistance to quinolone antibiotics;  The circulation of plasmid-mediated quinolone genes in the gram-negative bacterial population of Ghana;  The production of extended-spectrum beta-lactamase as predictor of quinolone antimicrobial resistance in gram-negative bacterial.  Table 1: the distribution of bacterial pathogens isolated from the Western Region of Ghana, stratified by gender, patient category, age, specimen type, and testing laboratory Table 2: antibiotic resistance pattern of Gramnegative pathogens isolated from patients at three laboratories in the Western Region of Ghana Table 3: plasmid-mediated quinolone-resistant genes as predictors of resistance to nonquinolone antibiotics Figure 1: agarose gel electrophoresis of PCR products; A) oqxA; B) aac(6´)-Ib-cr    77(46.5) βInclude 6 strains of Acinetobacter spp., 4 Aeromonas spp., 3 Providencia spp., 2 Morganella spp., 1 Serratia sp. and 1 Hafnia sp. * Acinetobacter spp. was tested with ciprofloxacin, levofloxacin, Amikacin, gentamicin, piperacillin/tazobactam, ceftazidime, and meropenem only. **Tested against only urine isolates, *** Enterobacteriaceae isolates (except the Serratia sp) were tested