AntiMicrobial Agents from LIchen Associated bacteria – AMALIA

AMALIA was organized into 6 different workpackages (WP) and tasks with their coordination and connections presented in the attached figure. Workpackage (WP)0 represented project management. In WP1 "Rapid screening", we screened the collection in batches of 50 strains using a single medium, growth condition and two extraction solvents (originally, we proposed a single solvent). Extracts were tested against multi-resistant strains at 100 µg/mL in liquid culture assays (modified from a more stringent original concentration) to quickly identify the "low hanging fruit" in the collection that were further developed in WP3-6. WP2 represented "Methods development". We used S. cyanofuscatus BBCC1488 and other Streptomyces strains to develop and optimize a method for fast screening of strains by Mass Spectrometry Imaging (MSI) on colonies interacting other organisms and the detection of metabolites and small molecules. We also developed a method to block the biosynthesis of major the bioactive metabolites N-methyl-dactinomycin for MOLA1488 and other Streptomyces strain (in place of the originally proposed Micromonospora BBCC1924) using CRISPR-BEST (in place of CRISPR-CAS9) gene editing. In WP3 "Secondary screening", we performed an further in-depth analysis of strains to exclude strains and extracts capable of producing known antibiotics by 1) analyzing the extracts by untargeted metabolomics 2) sequencing their genomes and by 3) screening for anti-persistence activity of extracts (the originally proposed activity “1) analyzing their interactome in confrontation assays inhibition halo with the multi-resistant strains by the technique developed in WP2” was abandoned due to difficulties in the development of the MSI on colonies). WP4 represented "Metabolite production and identification". Here we use different techniques to improve the production of molecules using pathway modification by ribosomal engineering and gene editing, and chemostat cultivation. This optimization was followed by large-scale molecule isolation and structural determination by NMR spectroscopy. WP5 originally planned as "Cell assays" to determine minimal inhibitory concentrations (MIC) of fractions and purified molecules against several multi-resistant and persistent strains, as well as toxicity against seven human cell lines to evaluate the suitability of the molecules for pharmacological development was not realized since we did not find a novel molecule suitable for development. Finally, we realized activities of knowledge and scientific dissemination in WP6.

We generated raw chemical extracts for 417 bacterial strains using two solvents for a total of 1007 extracts. These extracts were tested as whole (or as fractions) against 7 pathogenic bacteria (ESKAPE plus Salmonella enterica) for a total of 8697 single tests and 3483 triplicated tests. The results of these tests allowed us to identify 45 strains of interest for further development. In three cases multiple (2, 3 and 8) strains belonging to single species and presented different activity, confirming that dereplication would miss putative bioactive molecules.

In parallel we obtained genome sequences for 31 strains. We finished untargeted metabolomic analysis (LC-MS2 annotation and molecular network for 168 fractions/7 strains) as well as MSI imaging for 11 strains. Results indicated that most active fractions contained previously described classes of molecules of molecules such as surugamides lichenisin and collismicin among others, in accordance with BGCs in genomes. In some case however active fractions contained previously unidentified molecules and thus we selected 4 such strains for the remaining of the pipeline.

These strains identified as Micrococcus alloeverae (1917), Micromonsopora chalcea 1924 and S. cyanofuscatus (1493 and 1615) were grown in bioreactors as chemostats. As we lost bioactivity for 1493, we opted to use semi continuous culture that can be an alternative to accelerate biomass production, albeit not for all strains. Extracts for two of the strains were subject to compound isolation and NMR structural determination. The structure of 14 compounds was determined, all already described in the literature, while a putatively novel peptide currently is being isolated. The structure of some of these compounds was determined as soon as mo. 28 indicating that the main goal of streamlining the pipeline was achieved.

Methods development yielded interesting results. We applied a CRISPR-BEST genome editing of strain 1488 followed by growth in different media (OSMAC) and LC-MS2 and we were able to modify the metabolome of the strain suppressing the production of several actinomycin analogs. Several “novel” demethylated derivatives of actinomycin D and X2 were produced in ISP2 medium, but they were subject of recent patent applications. We were also able deploy the screening for anti-persistent E. coli for 6 bioactive extracts but we were unable to observe an effect. We were also able to deploy a method to profile colonies using MSI and a in house database and we were able to detect several compounds predicted by the genomic sequence and opening the possibility to the mapping on compounds in solid culture.

One of the most important roadblocks encountered was the fact that despite the novelty of the microorganism source (marine lichens) and the sheer number of strain and extracts tested, the “low hanging fruit” approach of bioactive screening and chemical characterization still yielded in many cases molecules in previously known classes, and it seems clear that focusing on a more limited number of high potential strains and attempting to active silent BGCs might be a better alternative path. The development of sequencing techniques that allow the sequencing of bacterial genomes, might provide a faster alternative to the discovery of rarer and novel molecules by selecting such high potential strains.

A second challenge is related to the screening of anti-persistence agents. It appears that presence of antibiotic molecules are again poor predictors of anti-persistence, and thus a more “random” screening with a large panel of novel and patentable molecules is the path to be taken. However as of now the throughput out anti-persistence platform is limited and so future developments increasing this throughput should be attempted.

We intend to keep our collaboration and follow up with future project submissions. The raw extracts from AMALIA were deposited in the Chimiothèque Nationale, and they will be available for screening for other purposes and open perspectives to collaborations (academic and industrial with other partners). We have dozens of genomes to be explored and with a large drop in sequencing costs and in house capabilities for genome assembly and annotation, we intend to continue to sequence strains from the AMALIA collection that open interesting perspectives both in terms of genome mining as well as the development of CRISPR-BEST tools to create novel molecules by genetic engineering approaches such as those from strain 1488 above. We also have several active fractions with putatively novel compounds, and we intend to continue purification and structural determination.

Antibiotic resistance (AR) and persistence of pathogenic microorganisms are emerging threats to human health, as the number of antibiotics effective against resistant bacteria is increasingly scarce. Microbe-host symbiotic interactions often depend on bioactive molecules with an antibiotic potential. Through the ANR project MALICA, we acquired a collection of >500 bacterial strains from marine lichens with yet unexploited potential for novel antibiotics. In AMALIA, the entire collection will be screened by the development of a novel streamlined approach by applying new methodologies using mass spectrometry imaging (MSI) and molecular networking to prioritize strains and extracts with novel molecules. As genomes of microorganisms code for a larger diversity of silent biosynthetic pathways we will develop ribosomal engineering and genome editing to change strain metabolomes and increase the diversity of produced antibiotics. To produce enough mass of compounds for structural determination and bioactivity testing we will combine the metabolome modifications with continuous cultures improving yields of molecules of interest. AMALIA is a 48-month project organized in 6 Workpackages. WP0 is project management, where we will guarantee that deliverables are reached and assure open access of data and resources. In WP1 we will realize a preliminary screening of all strains. In WP2 we will develop MSI screening and gene editing techniques. In WP3 we will realize in-depth screening and characterization of promising strains and extracts, including their action against persistent strains. WP4 represents large scale biomass production and structural determination of metabolites that will be tested in WP5. WP6 includes dissemination activities, including to the public at large. AMALIA will be led by Pr. Suzuki who has ample experience in project coordination. The consortium emerges from a long existing collaboration between LBBM and ISCR which resulted in 6 publications, and includes NUMECAN who has collaborated with ISCR to find new antibacterial agents resulting in 3 publications. We added Dr. N. Desbenoit from CBMN, a young CNRS researcher specialized in MSI who brings an innovative approach. Olgram is a private SME that has developed new strategies for the screening of persistent microorganisms and is interested in developing new antipersistant drugs. LBBM’s expertise is in environmental microbiology and genomics and recently we developed expertise in microbial genetics (N. West). LBBM is also specialized in analytical and natural product chemistry, metabolomic profiling and antibacterial research (D. Stien, A. Rodrigues). ISCR (Prof. S. Tomasi, responsible) expertise is in the chemistry of natural products (isolation, characterization, profiling, biological evaluation). NUMECAN (Dr. L. Bousarghin responsible) brings expertise on hypermutable and antibiotic resistant bacteria. We will involve permanent staff, including researchers and technical personnel. A major impact of the project will be the future development of effective antibiotics. Approaches to accelerate the discovery pipeline will contribute to other biodiscovery activities (e.g. antiviral, cancer therapies, etc.). Other impacts will be through publication and dissemination of research breakthroughs participation in international meetings. We will disseminate the project and its outcomes to a broader audience through the Biodiversarium of Banyuls that hosts a public starting from pre-school ages. AMALIA will contribute to skill building though the training of young scientists via internships or the training of early-career engineers hired by the project. The involvement of many permanent engineers and technicians will benefit their career development. Many senior participants are faculty members and will all be able to use the results and data of the project to transmit recent scientific information to university students from undergraduate to PhD levels.