(Support the NIH/NCRR P41 RR 01219 grant and NSF DMB 8916315 to C.A. Mannella) Optimal fixation of both isolated and in-situ mitochondria is being investigated by comparing conventional EM preparation techniques to cryo-based techniques. Various protocols of high-pressure freezing followed by freeze-substitution are being evaluated. In the case of isolated mitochondria, rapid plunge-freezing followed by cryo-EM is also being used. In all cases, final evaluation of results is made after electron tomography. An extensive series of experiments using high-pressure freezing and freeze substitution was done using fresh rat liver tissue. Freeze-substitution protocols from Martin M?ller's lab (ETH, Z?rich), and the Boulder HVEM lab (Mary Morphew) were compared. Hexadecene, fish gelatin, and ficol were all tried as fillers for the freezing step. Modifications of the freeze-substitution protocols involving different concentrations of osmium and uranyl acetate were tried. The quality of fixation was evaluated by imaging thin sections using a CTEM, and tomographic reconstructions were made of the best examples of each experiment, using 0.25(m sections on the IVEM or 0.5(m sections on the HVEM. Because of the increased density of specimens prepared with minimum extraction, the HVEM was beneficial even at 0.5(m thickness. We found that using the M?ller protocol, the intracristal space was very narrow, and the matrix was heavily stained, making the cristae membranes difficult to see. However, post-staining with lead alone sometimes made the membranes dense and quite visible. With the Morphew protocol, the matrix was also very dense, and the cristae membranes not visible, but the intracristal space was clear and a bit more open than in the case of the M?ller protocol. The variations in osmium concentration and en-bloc staining had little effect. Hexadecene and ficol seemed to be better than fish gelatin as a cryo-filler. The conclusions from the high-pressure freezing work are that mitochondrial architecture is essentially the same as that seen with conventional fixation techniques on the same tissue. However, good high-pressure freezing and freeze substitution results impose stringent demands on speed of tissue collection and the use of optimal media for the tissue during mincing and transfer to the high-pressure freezer. In addition, even in the best-frozen blocks, there is considerable variability in appearance of the mitochondria with respect to the openness of the intracristal space, a feature that seems to be more sensitive to local freezing conditions than, for example, the general appearance of the cytoplasm or nuclear membranes. Examples of the various experiments were shown at the Microscopy and Microanalysis '98 meeting in Atlanta in August. To avoid the problematic high-pressure freezing technique, but still study mitochondria in a more native state than conventional EM procedures permit, we turned to plunge-frozen whole, isolated mitochondria. In this way, we are able to make tomographic reconstructions of mitochondria frozen in vitreous ice, without any fixation or satin. Both Neurospora crassa and rat liver mitochondria are being studied. We did considerable experimentation with specimen preparation, in particular pre-plunge blotting methods and methods of application of colloidal gold markers for tomography alignment. We also did much experimentation with low-dose imaging techniques for tomography and recognition of the ideal ice thickness. The results we are now obtaining are excellent, much better than we imagined possible. We have made several reconstructions of both Neurospora crassa and rat liver mitochondria. All the mitochondrial membranes are very clearly visible, and the details of their organization a re clear. We find both """"""""orthodox"""""""" and """"""""condensed"""""""" mitochondria, as we did in our tomographic studies of conventionally-prepared plastic-embedded isolated mitochondria. The feature of """"""""tubular connecting regions"""""""", or restricted access between the inter-membrane and intracristal spaces, common to all mitochondria which have been studied by electron tomography is still observed. This feature was first described in our early reports (Mannella, et al. (1994) J. Microsc. Res. Tech. 27,278-283; Marko, et al. (1992) Proc. 50th Ann. Meet. Elec. Microsc. Soc. Am., 932-933). We also observe cross-bridges between the uniformly-parallel inner and outer membranes, which presumably become the puckered membrane connections we observe in conventionally-fixed material. Cross-bridges between the outer membrane and attached fragments of endoplasmic reticulum are also seen, again as observed in conventionally-fixed material by ourselves and others. The main problem still to be worked is avoiding the flattening (30-50%) observed in the reconstructions. We expect to increase the 3-D resolution (now about 7nm) and reduce the reconstruction artifacts by applying our double-tilt tomography technique, which will require a tilt-rotation cryo stage for the IVEM. Mannella, C.A., Buttle, K., Tessitore, K., Rath, B.K., Hsieh, C., D'Archangelis, D., Marko, M. (1998) Electron microscopic tomography of cellular organelles: Chemical fixation vs. cryo-substitution of rat-liver mitochondria Microscopy and Microanalysis vol 4, suppl. 2 (Proc. Microscopy and Microanalysis '98) Ed. G. Bailey, et al., Springer , New York, pp. 430-31.
Showing the most recent 10 out of 252 publications
