ELECTRON DIFFRACTION OF MITOCHONDRIAL MEMBRANE PROTEINS

点击翻译按钮获取中文摘要

(Support NSF MCB 9506113 to C.A. Mannella) VDAC is the predominant channel in the outer membrane of mitochondria, formed by a single copy of a 30-KD protein. There is considerable evidence the channel is a beta-barrel-like bacterial porin but, unlike the porins, this channel gates (open and closes) with several stimuli, including low-amplitude transmembrane potential and pH below 5. The mechanism of gating is unknown and of considerable interest. We have been able to grow cylindrical 2D crystals of this protein by treating outer membranes of N. crassa mitochondria with phospholipase A2. Using low-dose electron microscopy and image processing, we have generated a low-resolution (1.7nm) 3D structure of the channel in aurothioglucose. We want to extend this structure to higher resolution by using electron diffraction on frozen-hydrated specimens to collect Fourier amplitude information, which would be phased by low-resolution images and phase extension procedures being worked out in collaboration with Doug Dorset (Hauptmann-Woodward Foundation, Buffalo, NY). The cylindrical crystals of VDAC are small, and reflections from the two membrane layers often overlap, which limits the usefulness of these crystals for electron diffraction (ED) experiments. However, we have found ways to open these vesicles into flat single-layer sheets that approach 1 micrometer in linear dimension. We will determine the quality of the ED data from these sheets in frozen-hydrated state and assess whether they are suitable for a serious attempt at 3D data collection. The quality of ED data collected at 100, 400 and 1000 KV will be compared to determine whether direct-phasing approaches are feasible, i.e., whether expected improvements in Ewald sphere flattening and reduced dynamical scattering make a significant difference when attempting to apply direct phasing approaches to protein structure determination. Mitochondrial outer membranes are isolated, and VDAC crystals are prepared by removal of some of the membrane lipid using phospholipase A (PLA, from bee ven om). The membrane crystals are changed from their vesicular form into flat sheets by lowering the pH to 3.0 using either phosphate, citrate and glycine buffers. Fractions with high enzyme levels form crystals first, and crystals are harvested from fractions with lower enzyme levels on subsequent days. The crystals must be negatively stained or plunge-frozen on EM grids soon after they form, otherwise the residual enzyme will destroy them. Tests of the buffers, and of enzyme inhibitors to stabilize the crystals in the sheet form, are underway. Crystals in both vesicular and sheet forms are examined first with negative stain on a CTEM and the HVEM, then using cryo-EM on the CTEM. When results look promising, comparisons will be made using cryo-EM on the HVEM and IVEM. The diffraction and imaging conditions for the crystals were worked out for both the CTEM and the HVEM, and diffraction spots were observed on negatively-stained crystals. The first cryo-EM work is now being done.