Reproducibility challenges in the search for antibacterial compounds from nature

Reproducibility of reported antibacterial activities of plant extracts has long remained questionable. Although plant-related factors should be well considered in serious pharmacognostic research, they are often not addressed in many research papers. Here we highlight the challenges in reproducing antibacterial activities of plant extracts.

Plants with reported antibacterial activities of interest were obtained from a literature review. Antibacterial activities against Escherichia coli and Klebsiella pneumoniae were tested using extracts' solutions in 10% DMSO and acetone. Compositions of working solutions from both solvents were established using LC-MS analysis. Moreover, the availability of details likely to affect reproducibility was evaluated in articles which reported antibacterial activities of studied plants.

Inhibition of bacterial growth at MIC of 256-1024 μg/mL was observed in only 15.4% of identical plant species. These values were 4-16-fold higher than those reported earlier. Further, 18.2% of related plant species had MICs of 128-256 μg/mL. Besides, 29.2% and 95.8% of the extracts were soluble to sparingly soluble in 10% DMSO and acetone, respectively. Extracts' solutions in both solvents showed similar qualitative compositions, with differing quantities of corresponding phytochemicals. Details regarding seasons and growth state at collection were missing in 65% and 95% of evaluated articles, respectively. Likewise, solvents used to dissolve the extracts were lacking in 30% of the articles, whereas 40% of them used unidentified bacterial isolates.

Reproducibility of previously reported activities from plants' extracts is a multi-factorial aspect. Thus, collective approaches are necessary in addressing the highlighted challenges.