× Medicines Discovery Catapult

This collaborative project between the Medicines Discovery Catapult and the Lewandowski Group at the University of Warwick has resulted in high quality data obtained using the new ultra-fast magic angle spinning solid state NMR probe.

The results will be published in a peer reviewed journal in combination with biochemical data obtained by the Lewandowski Group.


Many structurally complex antibiotics are synthesised inside bacterial cells by large modular machineries such as Non-Ribosomal Peptide Synthetases (NRPSs). NRPSs generally have a multi-modular architecture where each module is responsible for the incorporation of an amino acid in a growing peptide chain.

A module consists of several domains that exhibit the functions necessary for recruiting the next building block and appending it to the polypeptide by formation of a peptide bond. However, the modules performing these steps in such assembly lines are often not encoded in one polypeptide and the quality of the product relies, therefore, on the flawless interplay between different modules (ref 1, ref 2).

In NRPS, this highly specific interaction between modules is mediated by communication domains (COM domains) at the termini of the respective polypeptide chains, a feature that has been exploited to design new compounds with the aim to fight antibiotic resistance (ref 3). An atomic resolution view of the interactions between modules would provide valuable information for engineering biosynthetic pathways to develop effective novel antibiotics using a synthetic biology approach.


The antibiotic peptide tyrocidin is synthesized by a NRPS that consists of three polypeptides TycA, TycB and TycC that encode ten modules together. Recognition between the TycA and TycB is mediated by a short intrinsically disordered C-terminal COM domain (TycACOM) which is predicted to form an α-helical structure when bound to its cognate communication domain located on the first condensation domain of the next module (ref 4).

The high flexibility of the TycACOM and dynamic nature of the interaction impeded the collection of X-ray crystallisation data of sufficient quality to obtain atomic resolution structure of the complex. And although it was possible to study TycACOM alone and bound to DPC micelles by solution NMR spectroscopy, the complex between TycACOM and the downstream condensation domain (C domain) is too large to use a similar approach.

A model of the complex was produced based on X-Ray crystallography, solution NMR and Mass Spectrometry data, however solid state NMR was the only method that could give atomic resolution information on TycACOM bound to the C-domain of TycB. A sample containing 13C and 15N labelled TycACOM (~3 kDa) and natural abundance TycB C-domain (~60 kDa) was sedimented by ultracentrifugation and packed into a 0.7 mm solid state NMR rotor.

The small volume of the rotor allows for a total sample amount of < 0.5 mg and since most of the space in the rotor is filled up with the unlabeled TycB C domain, only ~10-20 µg of labelled protein is used, making this approach attractive even for proteins that are very difficult to produce.

1H-13C spectrum of isotope labeled TycA bound to unlabeled TycB C-domain obtained at 110 kHz magic angle spinning and a 1H Larmor frequency of 700 MHz
1H-13C spectrum of isotope labeled TycA bound to unlabeled TycB C-domain obtained at 110 kHz magic angle spinning and a 1H Larmor frequency of 700 MHz

High quality spectra could be obtained and the chemical shifts of the assigned residues (marked in orange on the docking model) are consistent with α-helical structure and fits very well with the docking model.

TycA COM TycB C-Domain
Docking model of the complex formed between TycACOM and the C-domain of TycB
University of Warwick logo