that correct for the non-uniform changes that occur over the area of a large specimen during collection of images for single-axis and dual-axis tilt series [1, 2, 23].
In addition, tools for aligning and stacking tomograms from serial sections have
made this procedure routine. Finally, we describe the programs for stitching
together laterally adjacent tomograms, creating a “supermontage” of tomograms
themselves based on montaged images.
4.2 Specimen Preparation for Large-Scale Tomography
Large scale ET uses samples that are resin-embedded, stained with heavy metals,
and cut into sections of appropriate thickness for the microscope’s operating
voltage (200–300 nm thick). The use of high-pressure freezing followed by freeze
substitution results in excellent preservation of cell ultrastructure and is therefore
the sample preparation method of choice for study of mammalian cell ultrastructure
by ET. Methods for preparing cell monolayers for serial tomography have been
developed [24, 25] and an overview of the process is illustrated in Fig. 4.2. Cells
are grown on sapphire discs that have been carbon-shadowed with a grid pattern,
high pressure frozen and freeze substituted in 1% osmium and 0.1% uranyl acetate
in acetone at low temperature and subsequently embedded in a thin wafer of epon
resin [25]. Samples can be imaged in the LM and cells of interest identified, excised
and remounted for microtomy (Fig. 4.2a; arrow, inset). Serial 300 nm sections of
the cell are collected onto formvar-coated slot grids and post stained with 2% uranyl
acetate followed by lead citrate (Fig. 4.2b). Colloidal gold particles (15 nm) are
then affixed to each side of the grid to serve as fiducial markers for tilt series
alignment.
4.3 Using SerialEM for Automated Single Frame
or Montaged Serial Tilt Series Acquisition
Efficient, automated data collection for ET has been an important goal for large area
and volume reconstruction. A number of programs are available for automated
image acquisition for ET (e.g., commercial programs from Gatan and FEI; academic software from UCSF tomography [26] and the TOM toolbox [27]), but our
freely available SerialEM program has emerged as the tool of choice for many
laboratories (http://bio3d.colorado.edu/SerialEM). It contains a number of unique
features, including a prediction algorithm that allows rapid image acquisition. As a
result, it can routinely acquire tilt series from montaged images [1]. In addition, the
Navigator module in SerialEM assists with identifying and moving to areas of
interest to allow tilt series to be taken from multiple areas automatically, thus
increasing throughput. Figure 4.3 shows an example of a Navigator window that
4 Large-Scale Electron Tomography of Cells …
99
Précédent

- 117/339

Suivant