Benchmarking ideal sample thickness in cryo-EM using MicroED

Abstract

AbstractThe relationship between sample thickness and quality of data obtained by microcrystal electron diffraction (MicroED) is investigated. Several EM grids containing proteinase K microcrystals of similar sizes from the same crystallization batch were prepared. Each grid was transferred into a focused ion-beam scanning electron microscope (FIB/SEM) where the crystals were then systematically thinned into lamellae between 95 nm and 1650 nm thick. MicroED data were collected at either 120, 200, or 300 kV accelerating voltages. Lamellae thicknesses were converted to multiples of the calculated inelastic mean free path (MFP) of electrons at each accelerating voltage to allow the results to be compared on a common scale. The quality of the data and subsequently determined structures were assessed using standard crystallographic measures. Structures were reliably determined from crystalline lamellae only up to twice the inelastic mean free path. Lower resolution diffraction was observed at three times the mean free path for all three accelerating voltages but the quality was insufficient to yield structures. No diffraction data were observed from lamellae thicker than four times the calculated inelastic mean free path. The quality of the determined structures and crystallographic statistics were similar for all lamellae up to 2x the inelastic mean free path in thickness, but quickly deteriorated at greater thicknesses. This study provides a benchmark with respect to the ideal limit for biological specimen thickness with implications for all cryo-EM methods.SignificanceA systematic investigation of the effects of thickness on electron scattering from protein crystals was previously not feasible, because there was no accurate method to control sample thickness. Here, the recently developed methods for preparing protein crystals into lamellae of precise thickness by ion-beam milling are used to investigate the effects of increasing sample thickness on MicroED data quality. These experiments were conducted using the three most common accelerating voltages in cryo-EM. Data across these accelerating voltages and thicknesses were compared on a common scale using their calculated inelastic mean free path lengths. It is found that structures may accurately be determined from crystals up to twice the inelastic mean free path length in thickness, regardless of the acceleration voltage.