In-Cell NMR of Intact Mammalian Cells Preserved with the Cryoprotectants DMSO and Glycerol Have Similar DNP Performance

Abstract

NMR has the resolution and specificity to determine atomic-level protein structures of isotopically-labeled proteins in complex environments and, with the sensitivity gains conferred by dynamic nuclear polarization (DNP), NMR has the sensitivity to detect proteins at their endogenous concentrations. Prior work established that DNP MAS NMR is compatible with cellular viability. However, in that work, 15% glycerol, rather than the more commonly used 10% DMSO, was used as the cellular cryoprotectant. Moreover, incubation of cells cryoprotected 15% glycerol with the polarization agent, AMUPol, resulted in an inhomogeneous distribution of AMUPol through the cellular biomass, which resulted in a spatial bias of the NMR peak intensities. Because 10% DMSO is not only the most used cryoprotectant for mammalian cells, but also because DMSO is often used to improve delivery of molecules to cells, we sought to characterize the DNP performance of cells that were incubated with AMUPol and cryoprotected with 10% DMSO. We found that, like cells preserved with 15% glycerol, cells preserved with 10% DMSO retain high viability during DNP MAS NMR experiments if they are frozen at a controlled rate. However, DMSO did not improve the dispersion of AMUPol throughout the cellular biomass. Cells preserved with 15% glycerol and with 10% DMSO had similar DNP performance for both the maximal DNP enhancements as well as the inhomogeneous dispersion of AMUPol throughout the cellular biomass. Therefore, 10% DMSO and 15% glycerol are both appropriate cryoprotectant systems for DNP-assisted MAS NMR of intact viable mammalian cells.

Keywords: AMUPol; DMSO (dimethyl sulphoxide); Dynamic nuclear polarization (DNP); HEK293; cryopreservation; glycerol; in-cell NMR.

FIGURE 1

HEK293 cells that are cryopreserved with 10% DMSO are viable throughout the DNP NMR process. (A). Experimental scheme of the DNP NMR sample preparation procedure. Colored arrows indicate points at which sample viability was assessed. Viability was assessed for cells after trypsinization and washing (dark red), after suspension in AMUPol and cryoprotectants (orange), after being frozen at 1 °C per min (green), and after the entire DNP MAS NMR experiment (blue). (B) Percentage of cells with trypan impermeable membranes at each sample assessment point, colored as in A. Each point represents an independent sample. Black bars indicate average and standard deviation. Brackets indicate results of two-tailed homoscedastic student’s t-tests. (n.s. p > 0.05, ****p < 0.0001). (C) Growth kinetics as assessed by confluency, colored as in A. The averages and standard deviations of three independent experiments are indicated by circles and error bars, respectively. The best fit of sigmoid is indicated in solid lines and the 95% confidence interval by the shaded area.

FIGURE 2

The polarization agent, AMUPol, effectively polarizes all the components of HEK293 cells cryoprotected with 10% DMSO. (A) ¹³C cross-polarization spectra of cryopreserved HEK293 cells grown on isotopically enriched media with 10 mM AMUPol at 100 K taken at 600 MHz with 12 kHz magic angle spinning and a recycle delay of 10 s. Displayed spectra are taken with (black) and without (grey) microwave irradiation. The microwave off spectrum is plotted on the same scale as the microwave on spectrum (bottom) and with the intensity multiplied by 10 (middle). Colored arrowheads indicate peaks that are representative of proteins (green), nucleotides (blue) and lipids (pink). (B) DNP enhancement and (C) T _B,on values from saturation recovery experiments are dependent upon the AMUPol concentration. Fits of the T _B,on data to a mono-exponential equation (black line) for different biomass components for cells incubated with 10 mM AMUPol with 10% DMSO as a cryoprotectant. (D) The protein component had a T _B,on value of 4.6 s with a regression error (lower plot) of 3.0%. (E) The nucleotide component had a T _B,on value of 4.4 s with a regression error (lower plot) of 2.9%. (F) The lipid component had a T _B,on value of 5.2 s with a regression error (lower plot) of 2.4%.

FIGURE 3

(A) 2D homonuclear correlation spectra (DARR) of cells cryoprotected with DMSO. Selected ¹³C–¹³C correlations from carbons in the ribose (purple) and deoxyribose (blue) rings of RNA and DNA are annotated. (B) 2D heteronuclear correlation spectra (zTEDOR) of cells cryoprotected with DMSO. Selected ¹³C-¹⁵N correlations from the protein back bone and sidechains (green), from RNA (purple), from DNA (blue) and from lipid (pink) are annotated. The signal to noise ratios of selected peaks are reported in Supplementary Table S1.