Extended Data Fig. 3: NMR dynamics of Lipid II-bound and free plectasin, and investigation of the plectasin – pyrophosphate interaction. | Nature Microbiology

Extended Data Fig. 3: NMR dynamics of Lipid II-bound and free plectasin, and investigation of the plectasin – pyrophosphate interaction.

From: Host defence peptide plectasin targets bacterial cell wall precursor lipid II by a calcium-sensitive supramolecular mechanism

Extended Data Fig. 3

A) ssNMR 15N-R measurement of plectasin in complex with Lipid II in membranes, derived from 1H-detected data6. ssNMR spectra were recorded at 800 MHz (1H frequency), 60 kHz MAS, and 305 K sample temperature. A 15N spin lock amplitude of 18 kHz and with durations of 0, 5, 10, 20, 40, 80 and 120 ms were applied. The heights of amide resonances were fit using a single exponential decay. The error bars show the standard error of the fit. Source data are provided. B) Backbone amide dynamics determined with solution NMR on free plectasin. Measurements were done at 600 MHz (1H frequency) and 298 K. Experiment details are provided in the methods section. The increase in the ratio of 15N R2/R1 points to slow conformational exchange dynamics of the αβ-loop in the free state. This means that, in the absence of Lipid II, the αβ-loop is plastic, sampling different conformations. Source data are provided. C,D) 31P ssNMR experiments of the plectasin – lipid II complex. C) 31P CP spectra of Lipid II in DOPC membranes titrated with plectasin. Spectra were acquired at 500 MHz (1H-frequency), a sample temperature of 270 K and 10 kHz MAS for spectrum in the absence of plectasin and 13.5 kHz MAS for the other spectra. D) ssNMR 31P{13C}-REDOR experiments show that plectasin directly interacts with the pyrophosphate of Lipid II. REDOR experiments were conducted at 18.8 T (800 MHz 1H frequency) with 60 kHz MAS and a sample temperature of 290 K. 31P resonances of the lipid II pyrophosphate dephase upon reintroduction of 13C-31P dipolar couplings, while the bulk phosphate remains largely unaffected. 1D REDOR spectra were deconvoluted with Lorentzian and Gaussian functions. We fitted three peaks at –0.31 ppm (bulk phosphate), –5.76 ppm (named PPi-1), and –8.12 ppm (named PPi-2) with Lorentzian functions. An additional broad peak at –10.23 ppm, due to little 31P additional species not fully characterized, was fitted with a gaussian function to improve the quality of the PPi-2 integral. The signal integration was performed with trapezoidal method with the function numpy.trapz from numpy on the deconvoluted peaks. To fit the experimental spectra in the deconvolution procedure, the least-square method was used. Source data are provided. The deconvolution software for fitting and analyzing the data are written using Python 3.9.127, using the libraries numpy8 1.23.3. nmrglue9 0.9. matplotlib10 3.5.2. lmfit11 1.0.3. The package is available upon request to the authors. Dephasing curves are superimposed over simulated REDOR profiles corresponding to a 31P-13C distance of 8.0 Å (grey), 7.5 Å (green), 7.0 Å (orange), and 6.5 Å (blue), corresponding to 31P-13C dipolar couplings of –23.9 Hz, –29.0 Hz, –35.7 Hz and –44.6 Hz, respectively.

Source data

Back to article page