Ethical clearance was not required for this study, since there were no human subjects or animals models used in this research; however, this study was registered with the Institutional Research and Ethics Committee (ISERC) of the Institute of Primate Research (IPR/IRC/2014). The bacterial isolates used in this study were anonymised clinical isolates obtained from the Kenya Medical Research Institute (KEMRI) Center for Microbiology (SERU#2767), in Nairobi, Kenya. For environmental waste water collection, an approval was issued by Nairobi Water and Sewerage Company (NAWASCO#NCWSC/TRG14/109).

Environmental waste water collection was done by the Institute of Primate Research Phage Biology Group from July to August 2019. Samples were collected from six informal settlements: Kibera, Dandora, Kariobangi, Huruma, Mathare and Korogocho, all in Nairobi County, Kenya. Three samples per settlement were collected making a total of 18 samples. Dark screw cap containers (to prevent direct light) were used to collect the samples which were transported to the Institute of Primate Research Phage laboratory using cooler boxes as the secondary containment followed by storage at 4 °C and processed within three weeks. The three samples per settlement were named using number designations in the following order: 1, 2 and 3.

An MDR isolate of E. cloacae isolated from environmental waste water was used for phage isolation. This bacterium was identified using culture media, microscopic and biochemical examination using Vitek II machine (BioMérieux, Marcy-L’Etoile, France) for identification and antimicrobial susceptibility of bacteria.¹⁶ A panel of anonymised clinical bacterial isolates (not evaluated for antimicrobial susceptility test) of E. cloacae (n = 30) were subjected to the phages to assess cross infectivity.

Phages were isolated through an enrichment method according to the methods described by Akhtar et al. and Clokie et al. with slight modifications.^17,18 Briefly, 30 mL of each environmental waste water was centrifuged at 10 000 × gravitation for 10 min (Centrific™, Centrifuge Fisher Scientific, Waltham, Massachusetts, United States). The supernatant was mixed with an equivalent volume of tryptose soy broth (TSB) (HiMedia, Mumbai, India), and inoculated with 1 mL of 18 h-old MDR E. cloacae culture. The mixture was incubated overnight at 37 °C in a shaker incubator at 120 rotations per minute (Lab-line^® Incubator-Shaker, Waltham, Massachusetts, United States). The cultures were then centrifuged at 10 000 × gravitation for 10 min, and the supernatant sterilised using a 0.22 µm syringe filter (Millipore, Merck, Darmstadt, Germany) and stored at 4 °C for use in the spot test.

Phages were screened through spot test procedures according to Clokie and Kropinski (2010) with slight modifications.¹⁸ Briefly, a lawn of 24 h-old MDR E. cloacae isolate (100 µL) was prepared in soft agar (0.7%) with TSB on a tryptose soy agar (TSA) plate (HiMedia, Mumbai, India). Ten-fold serial dilutions of the phage filtrate were prepared and 5 µL of each dilution was spotted on a well-labelled plate with dilutions ranging from 10⁻¹ to 10⁻⁸ followed by overnight incubation at 37 °C. Observation of plaques (clear-patched regions) on the bacterial lawn was recorded as positive for the phage. For plates with distinct plaques, the well-isolated plaques were harvested using a sterile Pasteur pipette and suspended in 200 µL sterile saline magnesium (SM) buffer (100 mM sodium chloride, 10 mM magnesium sulphate, 50 mM Tris-HCl, pH 7.5 and 0.01% weight by volume gelatin), vortexed and incubated at room temperature for 1 h before centrifuging at 4000 gravitation for 5 min to remove any remaining debris, labelled and stored at 4 °C.

For plates without well-isolated plaques, the double agar overlay method was employed with slight modifications.¹⁹ Briefly, 100 µL of the 10-fold serial dilution of the lysate (from the spot with the least number of plaques) was mixed with 100 µL of 18 h-old E. cloacae in 6 mL molten agar (0.7% agar with TSB) and dispensed on TSA plate (1.5% agar with TSB) medium and allowed to solidify before incubation at 37 °C for 18 h. Well-isolated plaques were identified and marked before harvesting each plaque using a sterile Pasteur pipette, vortexed and incubated at room temperature for 1 h before centrifuging at 4000 gravitation for 5 min to remove any remaining debris, labelled well and stored at 4 °C.

In plates with distinct plaques for the spot test and plaque assay, the plaques were counted respective to their dilution factors employing the following formula:

Phage titre (pfu/mL)=plaques per plate×volume plated (mL)×dilution factor.[Eqn 1]¹⁸

Phage adsorption rates on the host bacterium were recorded at the temperatures 4 °C, 10 °C, 20 °C, 37 °C and 50 °C. Then 100 µL of actively growing host strain cultures in TSB to an optical density of 600 nm (OD600) of 0.6 was used to make an overlay on TSA plates. The selected phages were then incubated at these different temperatures for 1 h and then placed at room temperature (20 °C – 27 °C) for 30 min prior to performing spot tests. The outcomes were given as a log₁₀ of phage titre.^20,21,22

The effect of pH 2, 5, 7, 9, 11 and 13 on phage titre and viability of phages was studied in TSA plates by the spot test method.¹⁸ The SM buffer was adjusted to the desired pH by the use of NaOH and HCl. The actual pH of the SM was determined with a pH meter (Hanna Instruments Inc. Woonsocket, Rhode Island, United States) and this was used as a control or standard. The adjusted SM buffers were used for serial dilutions of the phages under study. Dilutions of the phage stocks were done to get the working dilution factors. Roughly, 10⁶ pfu/mL of 20 µL phages (individual phages) was added to 180 µL of SM buffer, after an earlier adjustment of pH (2–11), in Eppendorf tubes, followed by 30 min of incubation at 37 °C. The remaining phages were determined by spotting the phages in different pH and the outcomes indicated as pfu/mL.²³

The effect of the divalent cations on bacterial lysis and phage adsorption was investigated by varying the concentration of CaCl₂: 0.000 M, 0.005 M, 0.010 M and 0.015 M in soft agar during preparation. Plaque assays were conducted in duplicate, followed by overnight incubation at 37 °C. Phage titre was determined at the different salt concentrations as log₁₀ and comparisons made with the increase in salt concentrations.²¹

Stability of the phages during storage was investigated using a previously described method with slight modifications.²⁴ Briefly, 3 mL of the selected phages with known phage titre were aliquoted into 15 mL centrifuge tubes and wrapped with aluminium foil to prevent direct light and kept at –20 °C, 4 °C, and 37 °C for two weeks. The spot test method was used to determine the effectiveness and the efficiency of the phages after storage.

The spot test method was used to determine the effect of ultraviolet light on phage irradiation with various modifications.²⁵ Briefly, 10 µL of each phage of known titre from each site was aliquoted into five sterile PCR tubes and labelled as 0, 5, 10, 15 and 20 min of ultraviolet light exposure. The phages for each time period were placed in a Biosafety Cabinet Level 2 (290 nm – 320 nm, BSC-2, Haier, Tokyo, Japan) and the ultraviolet light turned on for the required time; the phages were then removed simultaneously. Polymerase chain reaction (PCR) tubes have been found to permeate ultraviolet light rays and affect the integrity of DNA; hence, the tubes were capped during this study.

To ensure the specificity of bacteriophages, their effect on other bacterial genera and species was investigated. The anti-bactericidal efficacy of individual phages (n = 29) was evaluated through the spot test method against each individual E. cloacae isolate (n = 30) as described by Kutter (2009).²³ Staphylococcus aureus, a bacteria from another genera, was used as negative control. A volume of 5 µL of individual phage stock was spotted on a TSA plate with a lawn of 100 µL overnight cultured host bacteria in soft agar, which was examined for bacterial lysis after 18 h – 24 h. The spot tests were performed in duplicate. A clear zone was considered as a positive infection in the tests and negative with no cross-reactivity.²⁶ A phage was termed ‘potent’ upon lysing any bacterial strain in the host range panel. A phage with the widest spectrum of lysis activity on the tested bacterial strains was termed the ‘most potent’, while a phage with the lowest spectrum of activity on the tested bacterial strains was termed the ‘least potent’. A bacteria that was sensitive to a phage infection was termed ‘susceptible’.

All the experiments were performed in triplicate and the mean values obtained. Statistical entry was carried out and given treatment in Microsoft Excel 20 for Windows (Microsoft, Redmond, Washington, United States) and Epidemiological Information (EPI info7^TM, Centers for Disease Control and Prevention, Atlanta, Georgia, United States). One-way analysis of variance using Statistical Package for Social Sciences version 20 (SPSS Inc., Chicago, Illinois, United States) was used to determine significant differences at p < 0.05. The surviving phage population obtained in each study were converted to log₁₀ PFU/mL. Data presentation was performed using GraphPad Prism version 5.00 for Windows (GraphPad Software, San Diego, California, United States).

This isolate was resistant to five classes of antibiotics namely: beta-lactams, penicillins (ticarcillin or clavulanic acid, and piperacillin), cephalosporins (cefuroxime, cefuroxime axetil, ceftriaxone, and cefepime), monobactams (aztreonam), chloramphenicols (amphenicol), tetracycline (tetracycline, minocycline), quinolones (levofloxacin) and sulphonamides (trimethoprim). It was susceptible only to: meropenem (carbapenem) and tigelcycline (glycylcycline).

A total of 29 phage strains were isolated, with four phages from Dandora, five phages from Huruma, five phages from Kibera, five phages from Kariobangi, seven phages from Korogocho and three phages from Mathare. These phages were named using number-letter designations according to the sample source. For example, from Kibera 1 settlement: the phages were named Kibera 1a, 1b, and 1c. All the phages were subjected to host range studies. From the 29 phages isolated from the six sources, one phage from each source that had complete lytic properties and a lytic zone diameter of ≥ 3 mm was selected for study of physicochemical properties.

The isolated phages were stable from 4 °C to 50 °C (Figure 1). There was a slight decrease in phage titre in four out of six phages at 50 °C, and two out of six phages had a slight increase in phage titre at 50 °C. In addition, one out of five phages had a constant titre from 4 °C to 30 °C.

No phages had any lytic activity at pH 2. All the phages were stable from pH 5 to 11 pH (slightly acidic to strong base). No phages had activity at pH 13 (Figure 2).

The addition of calcium chloride (0.002 M – 0.015 M) salt increased the adsorption rate and phage titre of phages in three out of five of the phages and a decrease in two out of five phages (Figure 3).

All the phages were stable and efficacious at 4 °C and 37 °C. There was minimal or no activity at −20 °C following two weeks of storage (Figure 4).

Exposure to ultraviolet light (290 nm – 320 nm) resulted in decreased phage titre from the 5th min to the 15th min. At the 20th min, there was no activity at all, which indicated that all the phages had been sterilised (Figure 5).

All the isolated E. cloacae phages (n = 29) showed cross-reactivity against E. cloacae strains (n = 30) (Figure 6). In order of individual phage potency from the highest to the least, the reactivity levels were: 67% (1 phage), 63% (3 phages), 60% (4 phages), 57% (5 phages), 53% (12 phages), 50% (1 phage), 47% (1 phage) and 27% (2 phages). According to the most susceptible bacteria to phages from the highest to the least, the following data were obtained: 100% (8), 97% (1), 93% (6), 86% (1), 79% (1), 31% (1), 21% (1), 14% (1), 10% (2), 7% (2) and 0% (7) (Figure 7).

In host range determination, a panel of 30 anonymised clinical isolates of E. cloacae and one Staphylococcus aureus (control) were used for cross-reactivity studies. However, no antimicrobial susceptibility studies for these bacterial isolates were carried out.

This study reveals the existence of the most potent lytic phages that are effective on MDR E. cloacae isolates found in Kenya. The existence of diverse phage strains from the sampled areas provided an effective cocktail of phages that could be used as antimicrobial agents. Findings from this study demonstrate that physicochemical properties influence the efficacy of phages in their antimicrobial activities and are worth considering.