Dynamic Nuclear Polarization (DNP) NMR at cryogenic temperatures provides one to two orders of magnitude in signal enhancement by transferring magnetization from excited free electrons to nuclear spins, often 1H. The radical source is a critical component for obtaining high enhancement factor and, in particular, its distribution through the sample has been shown to be important. This can become problematic for complex samples such as whole cells, which possess multiple environments (e.g. membranous vs. aqueous). Since we are interested in mapping antimicrobial peptide (AMP)-AMP self-assembly in live bacteria, a spin-labelled peptide has been designed to bring the radical source near the location of the label-free AMPs. A significant difference in signal enhancement was observed when comparing the addition of a mono-radical (TOAC) alone or the TOAC-AMP to model membranes, confirming that co-localization provides greater signal enhancement.
This approach was used to investigate how the pore-forming AMP maculatin 1.1 (Mac1) interacts with E. coli. Bacteria were prepared in a cryo-protectant solution containing the spin-labelled peptide (TOAC-Mac1) and less than a mg of 13C and 15N labelled Mac1 peptides. Signal enhancement for 31P, 13C and 15N of ca. 15-22 were obtained in-cell. In an attempt to determine how Mac1 peptides self-assemble to form a pore in bacteria, {15N}13C REDOR measurements were performed to measure the distance between the 13C=O and 15NH or 15N imidazole ring. A signal dephasing of ca. 13% was obtained at 12 ms which was primarily due to the NH atoms which showed a similar dephasing of 11% using a frequency-selective REDOR selection whereas none was observed for the protonated imidazole nitrogen atoms. This method of combining DNP and REDOR techniques allows identifying the structure of AMPs in whole cells.