As the promise of biomolecular engineering grows, so too does the need for methods and tools to evaluate the efforts. In particular, robust methods to identify and quantitate metabolites in unfractionated cultures and extracts are needed. Recently, we described the development of an atom based algorithm, Atomic Sort, to guide the discovery of unusual chemical structures within a microbial extract or fractionated extract. Modification of this algorithm to identify known compounds has allowed us to evaluate and direct several biomolecular engineering efforts. Notably, we were able to identify and quantify the desired products, as well as their precursors, fragments, and shunt products. Two case studies, explored for methods development and application, will be described.
Pladienolide B is a potent spliceosome inhibitor originally isolated from an engineered strain of Streptomyces platensis, Mer-11107. Since this strain was no longer available alternatives were sought. Three strains of S. platensis with different phenotypes were obtained and cultured as described previously. Ethyl acetate extracts of whole cultures were used to record 1H-13C ASAP-HSQC spectra on a Bruker Avance III 600 MHz spectrometer fitted with a 1.7 mm triple resonance cryoprobe. Atomic Sort was able to identify Pladienolide B, and its olefinic precursor, and determine which strain was the most productive. Engineering efforts were then able to proceed using the strain producing the highest titer of Pladienolide B.
Didemnin B is a cyclic didepsipeptide well known for its anti-cancer properties. Biosynthetic engineering is likely to afford easier access to it, and its analogues, than synthesis. To identify the optimal biosynthetic gene cluster, our team conducted a detailed analysis of the production of didemnin B within all known strains of the microbial producer, Tistrella sp. using the Atomic Sort approach. Didemnin B and only didemnin B (no congeners obtained) was identified in several strains along with peptide fragments of the target, and shunt products such as maculosin. Quantification directed further engineering efforts to maximise production.
Atomic Sort was able to not only identify entire molecules, but also, highlight areas of difference in closely related molecules. This is extremely useful when attempting to construct a series of analogues. We anticipate Atomic Sort will be a useful tool to evaluate biomolecular engineering efforts, and could also be applied to metabolomics studies and quality control of biologics.
Session #3: Metabolomics/Mixture analysis: to isolate or not to isolate … lets go with not!