Protein leakage into the extracellular environment is a key challenge for in-cell NMR studies. In the present work, the intracellular localization of 
the observed resonances was verified directly by observation of the restricted translational diffusion behavior characteristic of intracellular species. One-dimensional heteronuclear diffusion measurements were recorded before and after all 3D NMR measurements, using a 300 ms diffusion delay as previously described. Using this non-invasive method, when the fraction of extracellular aSyn exceeded 5%, data acquisition was halted and a fresh sample was prepared. NMR analysis of an expression time course showed no discernable lag phase, indicating that the species being observed constitutes the major state of aSyn within the cell, in agreement with previous spectroscopic and biochemical analyses. To determine the HN, N, CO, Ca and Cb backbone chemical shifts of intracellular aSyn a series of triple-resonance BESTHNCO and BEST-HNCOCACB experiments were recorded. Each 3D spectrum was acquired in just 1�C2 hours, which is an important factor as samples were typically found to contain significant levels of extracellular species after just a few hours. Spectra were repeatedly acquired from a total of four samples, and were then summed to produce a final spectrum. Analysis of the chemical shift of the single histidine resonance showed that within 30 min of sample preparation, the intracellular pH was 6.260.1, indicating acidification of the cytoplasm consistent with that observed previously under nutrient-depleted conditions. Therefore, for comparison purposes HNCO and HNCOCACB spectra were also acquired for a sample of isolated aSyn in bulk solution at the same pH. Because of the magnetically inhomogeneous environment characteristic of the dense cell samples studied here, having cell volume fractions of ca. 30%, aSyn resonances exhibit a strong inhomogeneous line AbMole Isoforskolin broadening giving rise to diagonal lineshapes. This effect can result in severe resonance overlap even within 3D spectra. As this broadening arises from variations in the magnetic field strength within the sample, its effect is constant on a ppm scale irrespective of the type of nucleus. This is therefore a particular problem for the determination of 1H chemical shifts in disordered states, due to the small chemical shift dispersion of ca. 1 ppm. The observed spectrum can be represented in the frequency domain as the ��true’ spectrum convolved with a threedimensional point spread function that reflects the distribution of magnetic field strengths found across the sample, and which is therefore the same for all residues. We note that such line broadening cannot be eliminated using non-uniform sampling methods, although for folded proteins where resonance overlap is a less significant problem NUS sampling schemes have been demonstrated to be very effective for the rapid acquisition of in-cell NMR spectra.