CORDIS - EU research results

Cocultivation of microorganisms as a strategy for discovery of new natural products with antimicrobial properties

Final Report Summary - COCULMICRO (Cocultivation of microorganisms as a strategy for discovery of new natural products with antimicrobial properties)

New antibiotics are needed to combat persistent and emerging infectious diseases. Microbial natural products have been a prolific source of anti-bacterials, and the screening of microbial extracts has yielded many classes of antibiotics. This approach was largely abandoned by pharmaceutical companies because of the general belief that there were very few new anti-bacterial compounds to be discovered from this source.

The sequencing of bacterial genomes has changed that view completely. In many bacteria, especially actinomycetes, there are far more gene clusters directing the biosynthesis of antibiotic-like metabolites than natural products known to be produced by the microorganism. These so-called “cryptic” gene clusters are often silent (not expressed) under standard laboratory conditions. Thus, new strategies need to be implemented in order to “awake” these gene clusters to further exploit microbial metabolomes.

Under standard laboratory conditions bacteria are grown in axenic conditions (i.e. one bacterium at a time). In this project we aimed to exploit bacterial interactions to identify and isolate new microbial natural products. Our hypothesis is that the co-culture of microorganisms will create (due to the presence of signaling molecules or physical interactions) the conditions required to “awake” some “cryptic” antibiotic-like gene clusters so new microbial natural products can be identified.

To test our hypothesis, a collection of actinomycetes (over 300 isolates) was generated from sampling at different spots in southern Spain. We then carried out a study of pairwise combinations employing micro-fermentations in 96 deep-well plates, as well as classical agar-based assays to look for pairs of microorganisms sharing some kind of cross-stimulation phenomena. Pairwise fermentations of the isolates with microorganisms highly likely to induce the production of secondary metabolites (e.g. mycolic-acid containing bacteria) were also carried out. High-throughput antibiosis assays against a panel of human pathogenic bacteria (methicillin-resistant Staphylococcus aureus, Acinetobacter baumanii and Candida albicans) were then used to identify pairs of microorganisms displaying an antibiosis activity not detected in the corresponding axenic controls.

An important drawback of employing high-throughput antibiosis assays to identify interacting pairs of microorganisms is that non-active induced secondary metabolites will not be detected. Thus we also carry out extensive studies based on UHPLC-comparative metabolite profiling in order to identify pairs of microorganisms producing metabolites in the co-culture and absent in the corresponding axenic controls.
The identified pairs of microorganisms with significant differences in the antibiosis profile of the co-culture relative to the axenic controls, as well as those with significant changes observed by HPLC-comparative metabolite profiling were double checked in new small-scale fermentations and dereplicated by LC-MS using an internal microbial natural product database.

Fourteen co-cultures were shortlisted and are currently being scale-up in order to isolate the compound(s) whose biosynthesis is induced during the co-culture. The typical approach that we follow involves the scale-up of the co-culture and axenic controls (up to 2 L fermentation broth), followed by SPE (solid phase extraction) and low pressure chromatography. The obtained fractions are assayed for activity and/or subjected to UHPLC-comparative metabolite profiling to identify the desired fractions. Active/selected fractions undergone successive rounds of reverse-phase chromatography following a classic bioactivity-guided fractionation approach, or are analysed by HPLC-comparative metabolite profiling if the co-culture was selected based in differences in the HPLC-metabolite profile. So far we have purified and isolated three compounds from two different co-cultures which are not produced in axenic fermentations, and we are working on their structural elucidation.

Our results strongly suggest that the co-culture of actinomycetes, the identification of new signaling molecules and the inherent mechanisms responsible of these interactions open new avenues for the exploitation of their biosynthetic potential. However, the strategy is not without important drawbacks. From our experience the most important drawback is the little control we can have on the interactions between the two microorganisms in the co-culture, which can have a profound impact on the development of this complex-cycle microorganisms and in their metabolome. As a consequence reproducibility in the production of secondary metabolites is hard to achieve at the levels required for isolation and identification of new molecules.