This study was approved by the joint Catholic University of Health and Allied Sciences and Bugando Medical Centre Research Ethics and Review Committee with research clearance certificate number: CREC/298/2023. The current study used bacteria species previously isolated from clinical samples at Bugando Medical Centre and stored at –80 °C in the Microbiology Laboratory at Catholic University of Health and Allied Sciences in Mwanza, Tanzania. Participants’ informed consent forms were not applicable because archived bacteria were used. Moreover, patient-related information such as socio-demographic data were not used in the current study, to ensure data confidentiality.

This was a laboratory-based study conducted in May 2020 in Mwanza, Tanzania. Surface soil samples were collected in sterile Falcon™ 50 mL conical tubes (Corning, Glendale, Arizona, United States) from 20 different locations of the same site at Bugando Hill in Mwanza, Tanzania. Soil samples were sent to the Microbiology Laboratory at the Catholic University of Health and Allied Sciences and processed within 1 h of collection.

Soil samples were serially diluted in 0.85% sterile saline as described previously¹⁷ before being inoculated onto 5% sheep blood agar (BA; HiMedia, Mumbai, India) plates. Briefly, 1 g of soil sample was suspended in 5 mL of sterile saline, which was then vortexed for 15 s and allowed to settle for 5 min. One millilitre of the resulting supernatant was mixed with 4 mL of sterile saline, vortexed for 15 s, and then allowed to settle for 5 min. This step was repeated three times to obtain five-fold serial dilutions. From the fifth dilution, 1 mL of supernatant was inoculated on a BA plate using the pour plate technique. Plates were incubated aerobically at 37 °C for 20 h. Presumptive Bacillus species colonies (large-sized, dry, flat, greyish to whitish, and beta-haemolytic colonies) were selected and sub-cultured on fresh BA plates to obtain pure colonies. These plates were incubated aerobically at 37 °C for 20 h. Gram staining was used for the preliminary identification of Bacillus species before further analysis.

A loopful (10 µL) of each presumptive Bacillus isolate was suspended in 1 mL of sterile nuclease-free water in an Eppendorf tube (1.5 mL safe lock microcentrifuge tube; Sigma-Aldrich Chemie GmbH, Taufkirchen, Germany). The tubes were incubated at 50 °C in a heating block for 30 min with intermittent vortex-mixing for 15 s every 10 min. We selected 50 °C for the extraction of bacterial metabolites because this temperature is insufficient for the destruction of metabolites. After incubation, the tubes were centrifuged at 12 000 revolutions per minute for 10 min, after which 0.8 mL of the supernatant from each tube was transferred into sterile Eppendorf tubes. The supernatants containing metabolites from soil-isolated presumptive Bacillus species were used immediately for antibacterial activity determination.

The well diffusion method, as described by Yilmaz and colleagues,¹⁸ was used to test the antibacterial activity of the extracted metabolites against one isolate each of extended-spectrum beta-lactamase-producing and non-producing E. coli and K. pneumoniae, MRSA and methicillin-susceptible S. aureus, inducible clindamycin-resistant S. aureus, coagulase-negative staphylococci, and Streptococcus pyogenes. These bacterial pathogens were isolated from clinical samples of previous studies conducted in the same setting^19,20 and archived at –80 °C in 20% glycerol stocks. The recovery of these bacterial strains was performed by subculture on BA plates which were incubated at 37 °C for 24 h. Bacterial suspensions of these test strains were prepared using sterile saline and adjusted to 0.5 McFarland standard solution (Remel, Lenexa, Kansas, United States). The bacterial suspensions were then inoculated onto the surface of Mueller Hinton agar (MHA; HiMedia, Mumbai, India) plates. Within 15 min of inoculation, wells of 6 mm in diameter were created in the inoculated plates using a cork borer (Sigma Aldrich, Darmstadt, Germany), and 100 µL of the suspension containing the extracted metabolites was then pipetted into the bored wells. The inoculated MHA plates were incubated in an upright position in ambient air at 37 °C for 20 h. The diameters of zones of inhibitions were measured in millimetres from the edge of the inhibition zone to the margin of Bacillus species metabolites. The complete absence of inhibition was referred to as ‘negative antibacterial activity’ and the presence of a clear zone of inhibition was referred to as ‘positive antibacterial activity’. This experiment was performed in duplicate to ensure the validity of our results. In cases of positive antibacterial activity, an average zone of inhibition was calculated and recorded as the final result.

Only one soil-isolated presumptive Bacillus species with a clear zone of inhibition (‘positive antibacterial activity’) was further identified at the species level. We first conducted biochemical identification tests, including catalase production, coagulase production, oxidase production, urease production, DNase production, indole production, citrate utilisation, lactose fermentation, hydrolysis of bile aesculin, and motility tests. We also tested the capacity of the isolates to grow on BA supplemented with 7% NaCl, as well as at higher incubation temperatures of 50 °C and 70 °C. The biochemical identification tests were not conclusive. For further species identification, we also used the VITEK MS (BioMérieux, Baden, Germany) system, which is an automated microbial identification system that uses the Matrix Assisted Laser Desorption Ionization Time-of-Flight technology. The latter was performed at the Microbiology Laboratory, National Public Health Laboratory in Dar es Salaam, Tanzania.

Laboratory data were documented and analysed using Microsoft Excel (Microsoft Corporation, Redmond, Washington, United States). Categorical data were expressed as percentages.

Due to limited resources and funds, we were not able to identify the presumptive Bacillus species with negative antibacterial activity beyond the genus level. We also could not isolate, purify, or characterise the active metabolites from the Bacillus altitudinis/pumilus complex isolate with positive antibacterial activity.

Our findings suggest that metabolites from soil-isolated Bacillus altitudinis/pumilus complex can be a potential source of antimicrobial agents with activity against Gram-positive bacteria, including MRSA and inducible clindamycin-resistant strains.