DTU Chemistry - PhD 2016

Structures of Natural Anti-cancer Agents

The project investigates how nuclear magnetic resonance (NMR) can be an efficient tool for solving 3D structures of novel compounds from fungal and synthetic sources.
The effect of pharmaceuticals is not only connected to the utilized chemical elements, nor their numbers or sequence, but also to the physical shape of the molecule. Substances which appear to be chemically nearly identical may be very different if the molecules inhabit different 3D structural spaces. The best example is enantiomers; two mirror-image forms of an otherwise identical structure, where only one may have the desired effect. The need for this structural information is increasing with e.g. the demand for enantiomeric information regarding drug candidates of, or inspired by, naturally occurring substances.

The project investigates how nuclear magnetic resonance (NMR) can be an efficient tool for solving 3D structures of novel compounds from fungal and synthetic sources, which in case of the latter have been suggested as potential anti-cancer agents. NMR has been used in analytical chemistry for more than half a century, with constant developments always increasing the applicability of the technique.

The stereochemistry and 3D structural space of several compounds from fungal sources was determined, using NOEs, yielding distance information between nuclei, coupled to another method, 3JHH-coupling constants, yielding dihedral information. 3D structural information of e.g. a bicyclic non-ribosomal peptide (with a novel structural motif), a steroid and several polyketides was thus gained. Furthermore, structural insights were gained for potential anti-cancer agents, the azumamides, including synthetic analogues. Differences in the conformational space of solution state compounds were identified experimentally between structural analogues, and compared to the in vitro potency of the compounds. The structures of two peptides that exhibited a high degree of molecular recognition were also investigated, resulting in the elucidation of a possible mode of interaction.

In parallel, new NMR experiments were developed based on spin-state selective (S3) methods. Thus, the S3 HMBC (Heteronuclear Multiple-Bond Correlation) experiment resulted in spectra with nJCH correlated cross-peaks, from which n+1JHH-coupling constants were sign-selectively determined with high accuracy. Very small coupling constants, including previously unreported coupling constants from strychnine, were extracted, with all experimental values correlating very well to theoretical coupling constants. The technique was extended to yield nJCH-coupling constants accurately by changing the polarization in methine pairs. The experiments have great potential for extensive usage in for example carbohydrate chemistry where methines are abundant.

In a final project, residual dipolar coupling (RDCs), a relatively new technique in small molecular NMR is investigated. Homonuclear RDCs were extracted from the homonuclear S3 HMBC that correlated well to alignment tensors from 1DCH-coupling constants, thus increasing the number of inter-nuclear vectors. Also, novel alignment media, chiral polymers, were synthesized and their ability to differentiate enantiomers investigated. Finally, a new method of back-calculation of RDCs from 3D structures was developed. The approach coped better with multiple conformers than commonly used methodologies, and resulted in good conformer populations for several small molecules, including multiple cinchona alkaloids.

DTU Chemistry - PhD 2016

Charlotte H. Gotfredsen

Funded by:
The project is funded by DTU Chemistry.