Background

Lymphatic filariasis (LF), a neglected tropical disease (NTD), is as a result of Wuchereria bancrofti, Brugia malayi or Brugia timori, transmitted through the bites of mosquitoes (from the genera Culex, Anopheles, Mansonia and Aedes) [1]. Since the inception of the Global Programme to Eliminate Lymphatic Filariasis (GPELF), efforts have been made at eliminating LF through mass drug administration (MDA) and morbidity management and disability prevention [MMDP] [2]. This has seen the likes of Palau meeting the criteria for eliminating LF as a public health concern [3]. Although this is welcoming, it falls short of the schedule of 2020 being the year to eradicate LF or be in the post MDA stage [4]. MMDP continues to remain one of the measures of managing LF patients including the complications that come with acute dermatolymphangioadenitis (ADLA). In the progression of LF, ADLA results in fever, chills, malaise, diffuse inflammation, swelling of the limbs, lymphangitis, adenitis and eventually, skin peeling [5]. This exposes LF patients to secondary bacterial infections which worsen their prognosis through increased morbidity and mortality [6,7,8,9,10,11].

Secondary infections occur during or after an infection of a primary pathogen [6]. In recent years, pathogenic bacteria have become problematic due to increasing antimicrobial resistance (AMR) [12]. Bacterial infections associated with other diseases are four times more likely to result in the death of a patient compared with bacterial infections alone [13]. Escherichia coli (E. coli), Staphylococcus aureus (S. aureus) and Haemophilus influenza remain some of the commonest bacteria isolates implicated in secondary infections [14]. While these have been mainly due to nosocomial infections, the resurgence of the One Health approach is giving evidence of community infections as well [15, 16]. Some S. aureus found in hospital and community environments are known to be penicillinase producers [17]. E. coli is located in different niches aside its intestinal habitat [18]. Pseudomonas aeruginosa (P. aeruginosa) is second to E. coli in terms of being the most significant MDR Gram-negative pathogen [19]. Virulence in S. aureus, E. coli and P. aeruginosa has thus been potentiated by mobile genetic elements and the innate resistances they possess [20]. Co-infections with these agents could result in abscess formation and necrotizing infections in LF patients [21,22,23].

Bacterial resistance may occur due to transient changes in bacterial permeability to antibiotics, drug indifference, biofilm formation or presence of persistent cells [24]. Acquired resistance has received attention when it comes to resistance in S. aureus, E. coil and P. aeruginosa bacteria strains due to their virulence [25,26,27,28,29]. Among these are mecA and ESBL genes. The mecA gene codes for resistance in S. aureus by producing penicillin-binding proteins (PBPs) that have low affinity for beta-lactam antibiotics. The mecA gene evolution is associated with anthropogenic activities that resulted in a core gene in Staphylococcus sciuri becoming a resistance determinant for broad spectrum beta-lactams in S. aureus [30]. Extended-spectrum beta-lactamases (ESBLs) are a group of enzymes known for their ability to hydrolyse ESBL antibiotics and are inhibited by beta-lactamase inhibitors [31]. ESBLs are known to have evolved from point mutations in the beta-lactamase (bla) encoding genes [32]. Among the many families of ESBLs are blaTEM, blaSHV and blaCTX-M genes with the latter being the predominant ESBL currently [32,33,34].

Aside the virulent nature of these organisms and genes, selective pressure due to antimicrobial usage has evolved a cocktail of resistance mechanisms that evade current antibiotics including penicillin, cephalosporins and carbapenem [35, 36]. Here, we applied a cross-sectional study and systematically collected bacteria isolates from eight endemic communities in the Ahanta West district, Ghana. MDR S. aureus, E. coli and P. aeruginosa were isolated from LF patients presenting with wounds. We assessed participants antibiotic usage, screened isolates resistance to commonly used antibiotics and presence of mecA, blaSHV, blaTEM and blaCTX-M genes. We found evidence of self-medication as a possible determinant of the antimicrobial resistance seen in the bacteria isolates. Our analysis informs the usage of antibiotics and MDR patterns among LF patients.

Results

Demography and wound characteristics of patients

Sixty (60) individuals were recruited for the study with the female sex having the highest representation (n = 42, 70%). Other demographic data on participants including age, employment status and wound characteristics (cause of wound, dimension, age (month) of wound existence and signs/symptoms of infections) were also captured (Table 1). Majority (n = 53, 70.67%) of the wounds participants presented were caused by the underlying condition of LF while fifteen (n = 15, 20%) were due to trauma suffered by the patients.

Table 1 Sociodemography of participants
Full size table

Antibiotics usage in participants

To gain insight on commonly used antibiotic in the study communities, participants were asked to respond to whether or not they use antibiotics and to name the antibiotic (if they had used one in the last three months) (Table 2). The antibiotics commonly used were amoxicillin, chloramphenicol, ampicillin, tetracycline, penicillin, flucloxacillin and clindamycin. Majority of the participants taking antibiotics (n = 36, 72%) were on beta-lactam antibiotics with the rest (n = 14, 28%) on chloramphenicol, tetracycline and lincosamide.

Table 2 Antibiotic (class) used by participants who knew the antibiotic they use
Full size table

Resistance pattern of all isolates

A total of seventy-five (75) wound swabs were collected during the study. Participants with multiple wounds were sampled with different swabs to prevent cross contamination. From the swabs, a total of thirty-nine (39) isolates were obtained; comprising E. coli (n = 5), P. aeruginosa (n = 8) and S. aureus (n = 26) (Table 3). All E. coli strains were resistant to sulfamethoxazole-trimethoprim, ampicillin, tetracycline and ampicillin-sulbactam. Sensitivity tests for P. aeruginosa were performed but not reported for sulfamethoxazole-trimethoprim, ampicillin, tetracycline, ampicillin-sulbactam, chloramphenicol, ceftriaxone and cefuroxime as there are no disk and minimum inhibitory concentration (MIC) breakpoints in European Committee on Antimicrobial Susceptibility Testing (EUCAST) [61] and Clinical & Laboratory Standards Institute (CLSI) protocols (Table 4). However, it is recommended they are reported as resistant when these antibiotics are tested although a small proportion of P. aeruginosa isolates are known to be sensitive to them. The S. aureus isolates were mostly resistant to penicillin (n = 23, 88.46%) and least resistant to erythromycin (n = 2, 7.69%). Eleven (11) isolates were MDR (resistant to three or more classes of antibiotics) (Table 4).

Table 3 Resistance pattern of S. aureus, E. coli and P. aeruginosa in LF patients
Full size table
Table 4 Distribution of MDR S. aureus, E. coli and P. aeruginosa in LF patients
Full size table

Distribution of MDR MRSA, E. coli and P. aeruginosa in LF patients

The eleven (11) MDR isolates comprised four (n = 4, 36.36%) MRSA, two (n = 2, 18.18%) E. coli and five (n = 5, 45.45%) P. aeruginosa (Table 4). Of the twenty-six (26) S. aureus isolates, four (n = 4, 15.38%) were resistant to cefoxitin and one antibiotic in 2 or more classes (Tables 3 and 4). Respectively, the 11 multi-drug resistant isolates were made up of 15.38, 40.00 and 62.50% of MRSA, E. coli, and P. aeruginosa (Table 4).

Sensitivity profile of MDR isolates in LF

Sensitivity profile of MDR MRSA

The MRSA isolates were all sensitive to sulfamethoxazole-trimethoprim, clindamycin, gentamicin and ciprofloxacin. All the MRSA isolates were resistant to cefoxitin, tetracycline and penicillin (Fig. 1). The American Type Culture Collection [ATCC] 25923 S. aureus strain was used as control.

Fig. 1
figure 1

Sensitivity profile of MRSA against different antibiotics

Key: SXT Sulfamethoxazole-trimethoprim, FOX Cefoxitin, C Chloramphenicol, CIP Ciprofloxacin, CC Clindamycin, E Erythromycin, CN Gentamicin, TET Tetracycline, P Penicillin. The breaks (white line) in the graph indicate resistant isolates (where the bars are below the break) or sensitive isolates (bars extend beyond the break) to the respective antibiotics listed

Full size image

Sensitivity profile of E. coli

MDR E. coli isolates were resistant to all antibiotics, including the third generation cephalosporins (cefuro-xime, ceftazidime and ceftriaxone), except meropenem. One isolate was also susceptible to gentamicin (Fig. 2). The ATCC 25922 strain was used as control.

Fig. 2
figure 2

Sensitivity profile of E. coli against different antibiotics

Key: SXT Sulfamethoxazole-trimethoprim, TET Tetracycline, AM Ampicillin, SAM Ampicillin-sulbactam, C Chloramphenicol, CRO Ceftriaxone, CXM Cefuroxime, CAZ Ceftazidime, CIP Ciprofloxacin, GM Gentamicin, MEM Meropenem. The breaks (white line) in the graph indicate resistant isolates (where the bars are below the break) or sensitive isolates (bars extend beyond the break) to the respective antibiotics listed

Full size image

Sensitivity profile of P. aeruginosa

All isolates were resistant to tetracycline, ampicillin, cefuroxime, ampicillin-sulbactam, sulfamethoxazole-trimethoprim and chloramphenicol because they did not show any in-vitro inhibition for P. aeruginosa isolates (zones of inhibition was ≤ 6 mm) (Fig. 3). EUCAST and CLSI have no guidelines for either zone of inhibition or MIC testing as P. aeruginosa is resistant to them (tetracycline, ampicillin, cefuroxime, ampicillin-sulbactam, sulfamethoxazole-trimethoprim, ceftriaxone and chloramphenicol). Both CLSI and EUCAST recommend that these antibiotics can be reported as resistant even without antibiotic testing. However for ceftriaxone, two P. aeruginosa isolates that were resistant to the other six antibiotics (tetracycline, ampicillin, cefuroxime, ampicillin-sulbactam, sulfamethoxazole-trimethoprim and chloramphenicol) exhibited very high zones of inhibition similar to the ATCC strains (Fig. 3). For interpretation of gentamicin, the CLSI breakpoint was used as EUCAST has no breakpoints (Fig. 3).

Fig. 3
figure 3

Sensitivity profile of P. aeruginosa against different antibiotics

Key: SXT Sulfamethoxazole-trimethoprim, AM Ampicillin, SAM Ampicillin-sulbactam, C Chloramphenicol, CRO Ceftriaxone, CXM Cefuroxime, CAZ Ceftazidime, CIP Ciprofloxacin, CN Gentamicin, MEM Meropenem, TET Tetracycline, AN Amikacin, AZT Aztreonam. The breaks (white line) in the graph indicate resistant isolates (where the bars are below the break) or sensitive isolates (bars extend beyond the break) to the respective antibiotics listed

Full size image

Minimum inhibitory concentration (MIC) and minimum bactericidal concentration (MBC) assays

MIC assay was performed for ciprofloxacin against the isolates. MRSA isolates had MIC and MBC < 0.48 µg/mL. E. coli isolates had MIC and MBC ≤ 31.25 µg/mL whereas P. aeruginosa isolates was ≤ 1.95 µg/mL. E. coli isolates exhibited the highest MIC and MBC at 31.25 µg/mL for ciprofloxacin which was one-fold higher than the E. coli ATCC strain used (15.63 µg/mL) (Table 5). The ATCC strain for S. aureus had similar MIC and MBC as that of the MRSA isolates. M064 isolate of P. aeruginosa was two fold higher than it’s ATCC strain (Table 5).

Table 5 Minimum inhibitory concentration for MDR isolates and ATCC strains
Full size table

Genotype distribution

The mecA and ESBL (blaCTX-M, blaSHV and blaTEM) genes were amplified in the multi-drug resistant isolates using PCR (Table 6, Fig. 4A, B). The bla-CTX-M ESBL gene was found in one of the MDR E. coli isolates (Fig. 4A). However, no ESBL genes were detected in the MDR P. aeruginosa isolates. All MDR MRSA isolates were positive for the mecA gene (Fig. 4B).

Table 6 Forward and reverse primers used
Full size table
Fig. 4
figure 4

Presence of resistance genes in MDR

Full size image

Discussion

Recent estimate puts the number of people infected with LF at 58 million, with majority of that population in developing countries. Long term exposure to LF results in damage of the lymphatics leading to lymphoedema and elephantiasis [37]. These result in mobility challenges, worsened financial status due to inability to engage in economic activities and put LF patients at risk of developing secondary infections due to ADLA [38]. In this study it is evident that majority of the participants reported with varied degrees of swelling and wounds that affect their ability to engage in economic activities. This also puts them at risk of bacterial infections (Table 1).

Previously, we have shown a positive correlation of the disease burden and pain or discomfort [39]. In managing the pain and skin peelings (as measures for morbidity management), LF patients resort to the use of antibiotics for self-medication as a preferred alternative to travelling long distance and the cost in accessing primary healthcare [5, 37]. Long exposure to antibiotics is known to be associated with resistance and this is supported by the fact that countries with higher antibiotic usage show higher resistance rates [40]. Antibiotic misuse due to self-medication has also resulted in the emergence of multi-drug resistance strains (including MRSA, cephalosporin-resistant E. coli and carbapenem resistant P. aeruginosa) with greater risk of infection and reduced therapeutic options [41, 42].

Bacterial infection in LF could worsen disease prognosis; hence knowledge on self medication and sensitivity patterns of co-infecting bacteria is essential in better management strategies (Tables  3, 4 and 5; Figs. 1, 2 and 3). Results indicated that participants have been self-medicating for a long time (evident in how easily they could assess and recall the antibiotics they used). Among the seven reported frequently used antibiotics by participants, the beta-lactam groups of antibiotics (amoxicillin, ampicillin, penicillin, flucloxacillin) represented 72% (n=36) followed by chloramphenicol (n=11, 22%), tetracycline (n=2, 4%) and clindamycin (n=1, 2%) (Table 2). This trend is similar to reports of amoxicillin, flucloxacillin, tetracycline and ampicillin being the commonest pharmacy antibiotic medicines used in Ghana [43]. In a similar report, amoxicillin is one of the most prescribed antibiotics in the United States of America [44].

Morbidity and mortality become more pronounced when polymicrobial bacterial infection are found in immunosuppressed individuals as in LF [6, 11]. Majority of participants presented with wounds that have been on them for ≥ 7 months with varied signs/symptoms of infections. Kathamuthu et al. [11] reported higher burdens of Mycobacterium tuberculosis in filarial infections as a result of immunomodulation (through upregulation of pathogenic chemokines and cytokines and downregulating key protective chemokines and cytokines of patients). Against the backdrop of the difficulty in curbing antimicrobial resistance, it was important to assess the phenotypic and genotypic characteristics of S. aureus, E. coli and P. aeruginosa isolates in LF patients. Among the Gram-negatives isolates, E. coli isolates were more resistant to the number of antibiotics tested compared to P. aeruginosa isolates. The results also showed that Gram-negative isolates from LF are beginning to exhibit higher MIC values when compared to typed strains that are not exposed to selective antibiotic pressure.

Persistent consumption and misuse of antibiotics enhance the acquisition and dissemination of drug resistant organisms (and genes) through environmental bacteria [40, 45,46,47]. This seems to be the phenomenon observed in this study; where the consumption of antibiotics (especially the beta-lactam antibiotcs) could have accounted for the development of MDR strains. In addition, easy access to antibiotics and lack of qualified personnel to dispense and counsel on appropriate antibiotic use in remote regions (study site) could account for the results obtained [43, 48].

S. aureus isolates exhibited the highest resistance to penicillin (88.46%) with erythromycin (7.69%) recording the lowest (Table 3). Similar reports have been observed in other studies because beta-lactam antibiotics continue to remain the mainstay of antibiotics treatment, hence the emergence of mecA and blaZ carrying MRSA strains [49,50,51,52,53,54]. Wolters et al. reported all isolates from chronic wounds with methicillin resistance to be positive for the mecA gene and resistant to other commonly used antibiotics (penicillin, tetracyline and trimethoprim/sulfamethoxazole). Aggarwal et al. [53] and Schaumburg et al. [54] have reported on the multi-drug resistant nature and high resistance of S. aureus for penicillin (97%), tetracycline (74%) and tremethiprim/sulfamethoxazole (77%).

Our findings show that all the MRSA (resistant to two or more additional antibiotics) were positive for the mecA gene as have been reported [51, 55]. MRSA percentage (15.38%) in this study is in line with the 15.15% reported in surgical wounds [56]. An earlier report by Wireko et al. [50] hints at a possible horizontal gene transfer for mecA in lymphatic filariasis MRSA isolates, owning to the presence of mecA gene is Staphylococcus epidermidis, Staphylococcus haemolyticus and Staphylococcus hominis isolates [50]. However, transposons have also been shown to be responsible for the translocation of resistance genes in the staphylococcal cassette chromosome, making MRSA resistant to other antibiotics [57].

All the E. coli isolates were resistant to tetracycline, ampicillin, sulbactam-ampicillin and sulfamethoxazole-trimethoprim and this could result in therapeutic failures (or higher concentrations will be needed) in morbidity management of LF patients (Table 3). In addition, more than half of the isolates were resistant to ciprofloxacin, ceftriaxone, cefuroxime and ceftazidime (Table 3). One MDR E. coli isolate was resistant to all the eleven [11] antibiotics tested with the exception of meropenem and paints a worrying sight of a possibility of extended drug-resistant strain (Fig. 2). Sensitivity tests for P. aeruginosa isolates were done but not reported for tetracycline, sulfamethoxazole-trimethoprim, chloramphenicol, ampicillin, ampicillin-sulbactam,ceftriaxone and cefuroxime as EUCAST and CLSI have no breakpoints for either the disk or MIC tests (Table 4). It was realized that the zones of inhibition for these antibiotics on MDR P. aeruginosa were ≤ 6 mm except ceftriaxone that had some isolates having zones > 15 mm (Fig. 3).

P. aeruginosa has been shown to acquire resistance under therapy and this has selected for multi-drug resistant strains [58,59,60]. The results agree with the selection of multi-drug resistant strains for P. aeruginosa as the isolates were found to be resistant to the antibiotics tested. The result (as shown with ceftriaxone) also corroborate reports by EUCAST and CLSI that some percentage of wild strains of P. aeruginosa are susceptible either due to a mutation or low levels of resistance [61, 62]. This further supports the antibiotic usage of the participants (none reported using any cephalosporin antibiotic) as selective pressure for cephalosporin resistance is low in this study.

P. aeruginosa has an innate resistance to the beta-lactam groups as such antipseudomonal penicillins in combination with beta-lactamase inhibitors is recommended [63]. In this study, it was evident that P. aeruginosa was not sensitive to ampicillin and ampicillin-sulbactam, which are part of the beta-lactam group. The observed innate resistance of P. aeruginosa for ampicillin, ampicillin-sulbactam and sulfamethoxazole-trimethoprim in this study has also been reported [64, 65]. Meropenem resistance occurred in 37.50% (n = 3) of P. aeruginosa isolates (40% in MDR isolates), which is similar to that reported by Bediako-Bowan and colleagues (Table 3; Fig. 3) [56].

Current evidence has shown that MDR bacteria have become prevalent, even in community-acquired infections [66, 67]. A total of 11 MDR were identified, representing 28.2% of bacteria isolated from the patients (Tables 4, 7, 8 and 9). 40.00% of isolates of E. coli were MDR and this is in line with the 33.33% that was reported by Nigussie et al. [65]. While Nigussie and colleagues reported 80.0% MDR P. aeruginosa isolates and 10.0% MRSA isolates, this study reports 62.5 and 15.38% as the proportions of MDR P. aeruginosa and MRSA isolates, respectively (Table 4) [65]. While findings from other diseases and niches have shaped the discourse of the genes driving these community infections, few have come from studies related to lymphatic filariasis [68].

Table 7 Phenotypic characterization of MDR methicillin resistant S. aureus
Full size table
Table 8 Phenotypic characterization of MDR E. coli and P. aeruginosa
Full size table
Table 9 Zone of inhibition for MDR isolates against antibiotics
Full size table

Our findings showed the presence of blaCTX-M ESBL gene in one of the MDR E. coli isolates, while none of the tested ESBL genes were found in the P. aeruginosa isolates. The presence of blaCTX-M in E. coli may explain the resistance (antibiotic resistance profile) that was evident in the MDR isolates as this gene is involved in co-resistance with fluoroquinolones, extended activity against cefotaxime compared to ceftazidime and being predominant in community infections (Fig. 4B) [69, 70]. Furthermore, physiological activities such as low expression of porins (OmpC and OmpF in E. coli) that are transporters of antibiotics (quinolones, tetracycline, β-lactams or chloramphenicol) may also result in resistance due to low accumulation of antibiotics in bacteria [24]. Most studies including this, have focused on inheritable traits as the cause of antibiotic resistance. However, the absence of the ESBL genes in P. aeruginosa isolates points to the absence of genetic change from these group of genes and the resistance seen here is possibly due to bacterial permeability or other genetic factors [24].

The fact that the ciprofloxacin MIC of all the MDR isolates tested in this study falls within the sensitive range for each bacterium is welcoming considering the report by Newman and colleagues that puts the MIC for MDR S. aureus at > 0.256 mg/mL for ciprofloxacin. The ones obtained here had MIC < 0.00048 mg/mL (Table 5) [71]. All the MDR isolates in this study fell within the range given by EUCAST and Clinical & Laboratory Standard Institute (CLSI) as breakpoints [P. aeruginosa (0.001—0.5 mg/mL), S. aureus (0.001 – 1 mg/mL) and E. coli (0.25 – 0.5 mg/mL)] [61]. Comparing the ATCC strains with isolates from this study, we see a higher MIC value for some Gram-negative isolates. MDR isolates being sensitive to ciprofloxacin gives credence that these isolates have not been extensively subjected to selection pressure from some antibiotics tested in this study, as was indicated by the response from the antibiotic usage questionnaire. In essence, most inhabitants resort less to ciprofloxacin in treating their wounds and during filarial attacks.

This study has highlighted the phenotypic and genotypic (mecA, blaSHV, blaTEM and blaCTX-M) characteristics of MDR methicillin-resistant S. aureus, cephalosporin resistant E. coli and P. aeruginosa prevalence (Tables 1, 2, 3, 4, 5, 6, 7, 8 and 9; Figs. 1, 2, 3 and 4A, B). This is worrying, and it is expected that the findings of this study direct clinicians in the management of LF patients with secondary bacterial infections.

Conclusion

This study reports the presence of MDR MRSA, cephalosporin resistant E. coli and carbapenem resistant P. aeruginosa in LF patients in the Ahanta West district of Ghana and the need for clinicians to be peculiar in how they administer antibiotics to these patients. In addition, this study has highlighted the antibiotic usage of participants and how it drives both phenotypic and genotypic modifications in the MDR isolates. These findings, especially the genetic mechanisms identified for the MDR resistance require further exploration such as sequencing to understand better the diversity occurring in these isolates and their impact on therapeutic failures in LF patients.

Methods

Study design and participants

Study participants were recruited from 8 LF endemic communities (Busia, Butre, Achowa, Princess Town, Akatakyi, Ampatano, Dixcove and Asemkow) in the Ahanta West Municipal. Ahanta West Municipal remains one of the hotspots in Ghana with a microfilariae prevalence of 2.2% and has been part of clinical trials to test doxycycline [72, 73]. This study followed all protocols regarding human participants and ethical clearance was given by the Committee of Human Research and Publications and Ethics, School of Medicinal Science, KNUST (CHRPE/AP/649/19). Participants who consented to the study included individuals with different stages of LF, who had lived in the endemic community for ten (10) years or more and were between the ages of eighteen (18) and sixty (60) years with wounds. These potential participants were enlisted by health extension workers prior to the study. Participants were given informed consent, which they consented to either by signing or thumb printing and countersigned by an independent witness.

Sociodemography and wound characteristics

Sociodemographic data on age, occupation, marital status and sex were taken. Leg staging of participants were done as has been reported by [50]. Information of cause of wound and age of wound in months were done using a simple structured questionnaire. The size of wound, appearance of wound and signs of infection were reported by measuring and observing the wounds during the sample taking process.

Data on antibiotic usage

To determine the possible impact of antibiotic usage on antimicrobial resistance in LF patients, a simple structured questionnaire was used to get their response as to whether they had used antibiotics within the last three months for treatment and, if any, to name the antibiotic.

Collection of samples

Swabs were taken from LF wounds patients using Becton Dickson and Company (BBL) culture swab (plus amies gel medium without charcoal). Swabs were stored between 2—8 °C on the field, transferred to a liquid nitrogen tank (liquid nitrogen submersion) at the end of the day’s collection and transported to Kumasi Centre for Collaborative Research (KCCR) for culture and isolation of bacteria. The culture and isolation have been described previously [50]. S. aureus (ATCC 25923), E. coli (ATCC 25922) and P. aeruginosa (ATCC 27853) strains were used as controls and performed to specifications as is reported for sensitivity testing [74,75,76].

Tests for identification of isolated organisms

Biochemical identification of MRSA

To identify S. aureus, coagulase test and catalase test were done on the Gram-positive isolates while observing for beta-hemolysis on blood agar (BA). Organisms were cultured on BA or Mueller–Hinton agar (MHA) and incubated at 35 ± 2 °C for 18 ± 2 h. In determining catalase production in the isolates, colonies of bacteria from a fresh culture were picked from Mueller–Hinton agar and placed in droplets of catalase reagent (3% H2O2) and observed for bubble formation [77]. Bubble formation was indicative of a catalase positive isolate while no bubble formation indicated a negative test. The coagulase test was used to differentiate between S. aureus and coagulase-negative staphylococci. The coagulase test was performed by using fresh colonies from a culture. This was placed in a rabbit serum in a tube and incubated for 4 h, where the presence of clumping indicated a positive test, while no clumping indicated a negative test [78]. Beta-hemolysis was recorded as positive if there were formation of clear zones around the S. aureus isolates growth on BA that had been incubated at 35 ± 2 °C for 18 ± 2 h [78].

Biochemical identification of P. aeruginosa and E. coli

Gram-negative organisms were subjected to oxidase, indole, citrate and triple sugar iron (TSI) tests to aid in the identification and characterization of the phenotypes of P. aeruginosa and E. coli isolates. Bacteria isolates were cultured on MacConkey agar (MAC) or MHA and incubated at 35 ± 2 °C for 18 ± 2 h before all the tests were done. The oxidase test was performed to identify isolates that catalyze oxidase-reduction via the cytochrome C oxidase enzyme. To perform the oxidase test BBL Dryslide and Remel oxidase were used. Isolates grown on MHA were streaked on BBL Dryslide. For a positive test a bluish colouration was seen after 20 s, while no colour change indicated negative test [79]. This was confirmed by using the Remel oxidase reagent [80]. Indole production was used to differentiate Enterobacteria that degrade tryptophanase. Indole production test was done using 4–5 colonies of isolate grown on non-selective media (MHA) placed on a sterile swab stick [81]. The indole Remel reagent was then poured on the swab and within a period of 120 s, bluish colouration of the colonies indicated indole production (positive test), while pink colouration indicated no indole production (negative test). In determining bacteria isolates that possess citrate permease, the citrate utilization test was performed. Citrate utilization test employed the use of 5–6 colonies streaked on the slant of citrate agar. This was later incubated (35 ± 2 °C for 18 ± 2 h) and a bluish colouration was indicative of a positive test for citrate utilization [82]. TSI was done to differentiate Gram-negative enteric bacilli as previously described by [83].

MALDI-TOF for confirmation of bacteria

Isolates of bacteria were grown on BA and later prepared for inoculation on the MALDI Biotyper [Bruker Daltonic GmbH] [50]. To confirm the isolates using ribosomal and housekeeping proteins, the MALDI-TOF was used. Generated spectra were compared with current libraries (MALDI Flex Control software system Server Version: 4.1.31 and 60 Hz Nitrogen Laser (337 nm wavelength) and log scores generated (Scores of the log were classified as: > 2.0 = High confidence identification, 1.7–2.0 = Low confidence and < 1.7.0 = Unreliable).

Sensitivity testing and identification of multi-drug resistant MRSA, E. coli and P. aeruginosa

Having confirmed the isolates, it was important to assess the antimicrobial resistance of S. aureus, P. aeruginosa and E. coli in LF patients with wounds. Colonies of bacteria were picked from MAC/BA and inoculated in 1 mL of normal saline to give 0.5 McFarland standard. Cotton swab sticks (sterile) were dipped in the inoculum and spread on MHA plates that had passed quality control [61]. The plates were left to dry for 5 min and incubated at 35 ± 2 °C for 18 ± 2 h after appropriate antibiotic disks had been placed on the plates. The following antibiotics were used for S. aureus: cefoxitin (30 µg), erythromycin (15 µg), chloramphenicol (30 µg), tetracycline (30 µg), ciprofloxacin (5 µg), sulfamethoxazole-trimethoprim (23.75 µg /1.25 µg), clindamycin (2 µg), gentamicin (10 µg) and penicillin (10 UI). The following antibiotics were used for P. aeruginosa: ciprofloxacin (5 µg), ampicillin (10 µg), aztreonam (30 µg), sulfamethoxazole-trimethoprim (23.75 µg /1.250 µg), amikacin (30 µg), ampicillin-sulbactam (10 µg /10 µg), ceftriaxone (30 µg), gentamicin (10 µg), cefuroxime (30 µg), meropenem (10 µg), ceftazidime (30 µg) and tetracycline (30 µg). For E. coli, all the antibiotics used for P. aeruginosa were used with the exception of aztreonam (30 µg) and amikacin (30 µg). Zones of inhibition were read after incubation and breakpoints were recorded [61]. MDR S. aureus, P. aeruginosa and E. coli were identified as resistant to at least one antibiotic in three or more classes. The following antibiotic classes were used in this study: aminoglycosides, carbapenems, cephalosporins, fluoroquinolones, macrolides, monobactams, penicillins, sulfonamides, lincosamides, phenicols and tetracyclines.

Minimum inhibitory concentration (MIC) assay

Preparation of inoculum

To gain insight into how the isolates from LF patients differed from typed strains (ATTC) in terms of inhibitory concentrations of antibiotics, MIC was performed using a broad-spectrum antibiotic (ciprofloxacin). Distinct colonies (5—6) grown on MHA were picked from an 18 ± 2 h culture and transferred to normal saline to give approximately 1.5 × 108 cfu/mL (0.5 McFarland standard). A volume of 0.1 mL of 1.5 × 108 cfu/mL inoculum was transferred to 9.99 mL of broth to give 1 × 106 cfu/mL [84]. This was homogenized on a roller. An equal volume (100 µL) of an antimicrobial agent was added to the same volume of broth containing 1 × 106 cfu/mL to give a final concentration of 5 × 105 cfu/mL [85]. Tubes were inoculated within 30 min of standardizing the inoculum to maintain viable cell density.

Viable cell count and purity check

Viable cell counts were performed to ensure the final concentration of inoculum was 5 × 105 cfu/mL. This was performed by taking 3 µL of the stock of inoculum (5 × 105 cfu/mL) and diluting it in 3 mL of saline. A volume of 100 µL was taken and then spread on MHA plates, where an average of 50 colonies were counted from the final concentration. Purity plating was done by taking an inoculum and plating on an unselective medium (MHA) to check if growth is uniform throughout the plate.

Determination of MIC of ciprofloxacin using microdilution assay

Labelled wells of microtitre plates (1—11) were filled with 100 µL of broth (1 × 106 cfu/mL), while well 12 was filled with broth without any bacteria inoculum. Test solutions of 100 µL were added to each well (1–8) in the microtitre plates to give concentrations in the range (128 mg/mL to 1 mg/mL). Wells 9 and 10 were filled with 100 µL of positive control (ciprofloxacin), while wells 11 and 12 were filled with normal saline [85]. This was then incubated at 35 ± 2 °C for 18 ± 2 h. To determine MIC, 20 µL of 3-(4,5-dimethylthiazol-2-yl)-2–5-diphenyltetrazolium bromide [MTT] (1.25 × 103 mg/mL) was dispensed in each well and the sample re-incubated for 30 min [86]. MIC for ciprofloxacin against test organisms were indicated by the presence of a violet colouration (due to the production of insoluble formazan from MTT by viable cells). MIC was taken as the immediate well that did not show a violet colouration after 30 min of adding MTT.

Minimum bactericidal concentration

To determine the concentration at which ciprofloxacin was bactericidal, an inoculum was taken from the MIC wells prior to the dispensing of MTT. The inoculum was plated on MHA plates and incubated at 35 ± 2 °C for 18 ± 2 h. The value of MBC was recorded after incubation.

Phenotypic identification of MRSA and extended-spectrum beta-lactamases (ESBL)

MRSA was identified as S. aureus with a zone of inhibition < 22 mm for cefoxitin [30 µg] [61, 87]. Screening of suspected extended-spectrum beta-lactamases for E. coli and P. aeruginosa were done by plating the confirmed isolates on MacConkey plates with a concentration of 1 µg/mL of ceftazidime or cefotaxime. Isolates that grew on the ceftazidime or cefotaxime MAC agar underwent the double-disk synergy test [88].

Extraction of DNA

ESBL-positive and MRSA isolates were examined for the presence of blaTEM, blaSHV, blaCTX-M and mecA using conventional PCR. DNA extractions were performed from overnight bacterial cultures using the boiling method [50, 89]. A loop full of bacteria was emulsified in 1 mL of phosphate-buffered saline and vortexed vigorously. The isolates were then centrifuged at 8000 rpm for 3 min and the supernatant discarded. A volume of 100 µL of nuclease-free water was added to the bacterial cells, vortexed vigorously and then incubated at 95 °C for 5 min. After incubation, centrifugation was done at 14,000 rpm for 5 min. The supernatant was pipetted into a well labelled 1.5 mL Eppendorf tube as DNA. The extracted DNA was stored at -20 °C until use. The extracted DNA were used as templates for the detection of the genes of interest.

Amplification of gene

To determine possible genes potentiating resistance in the MDR isolates, PCR amplification was performed. Detection of blaTEM, blaSHV, blaCTX-M and mecA were performed using primers in Table 1. Amplification using the Veriti® thermal cycler was conducted using the following PCR conditions; an initial denaturation at 95.0 °C for 15 min; 35 cycles at 94 °C for 30 s, 60 °C for 1 min and 72.0 °C for 1 min. With a final extension of 72.0 °C for 7 min, the reaction was put on hold at 4 °C until ready to be viewed for bands. Each reaction volume was 25 µL and was done with modifications from Lin and colleagues protocol [90]. The following PCR conditions were used for mecA gene; an initial denaturation at 94.0 °C for 30 s; 30 cycles of 94 °C for 15 s, 45 °C for 1 min and 68.0 °C for 1 min. With a final extension of 68.0 °C for 5 min, the reaction was put on hold at 4 °C. Each reaction volume was 25 µL [91]. PCR products were visualized by agarose gel electrophoresis using UV-transilluminator (Vilber Lourmat, Collegien, France) after staining with ethidium bromide. The stained gel was captured onto a desktop computer using the infinity® software.

Data analysis

Data were analyzed using Python (v 3.8.10), Matplotlib (v 3.3.3), Numpy (v 1.19.4) and Scipy (v 1.5.4).