architecture of intact cells in a near-native state. As a result, ECT has made significant and previously unobtainable biological insights into cell biology [1–4]; we
illustrate this with examples at the end of this chapter.
ECT is not restricted to imaging unique individual cells and other specimens,
however: if multiple cryo-tomograms of identical structures are collected, these
identical single particles can be computationally extracted, aligned, and averaged, an
approach known as subtomogram averaging, subvolume averaging, or single-particle
electron tomography. This approach ‘averages out’ noise and reinforces signal, greatly
improving the signal-to-noise ratio, with the potential for boosting resolution from
nanometres to Ångstroms. Because cryo-tomograms are in 3-D, particles can be
extracted even from cryo-tomograms of crowded environments such as the cytoplasm,
enabling in situ structure determination [5]. Advances in hardware and software have
increasingly seen subtomogram averaging applied to structural biology problems, and
a growing number of subtomogram averages have been determined to sub-nanometre
resolutions, enabling fitting of crystal structures into densities of macromolecular
complexes in situ. The boundaries continue to blur between ECT and more traditional
single-particle cryo-EM approaches to structure determination, particularly in light of
the considerable advances in contemporary cryo-EM [6].
This chapter focuses on the workflow of a biological ECT research project,
discussing each step consecutively with emphasis on choices faced by the
researcher, and the rationale behind the decision. We start with an overview of the
process and underlying concepts, followed by in-depth discussion of specimen
choice, preparation, and vitrification; background on microscope hardware setup;
software and data collection parameters; and tomogram reconstruction, processing,
and interpretation. We intend this chapter to provide an overview and introduction
for those training in ECT and many topics are covered in greater depth in subsequent chapters of this book.
c
Fig. 3.1 An overview of the electron tomography workflow. ECT can be used to image a wide
range of biological specimens from purified protein complexes to large eukaryotic cells and
tissues. The sample preparation and vitrification method is primarily dictated by the size of the
specimen being imaged. Small specimens (up to a few µm thick) can be plunged into a liquid
cryogen, causing rapid vitrification before the formation of crystalline ice can occur. Large
specimens (more than a few µm thick) must be high pressure frozen. As imaging of specimens
thicker than *500 nm is greatly hindered by excessive inelastic scattering, a method of thinning
must be chosen for thicker samples. FIB-milling can be used to mill away part of the specimen,
leaving a thin lamellae to be imaged, or cryo-sectioning can be used to produce thin slices of a high
pressure-frozen sample. Grid preparation remains a major bottleneck in ECT and the process of
sample preparation and vitrification may need to be repeated multiple times before data-collection
can begin. When ready, the grids are loaded into the microscope and data-collection parameters are
setup according to the type of specimen and biological question. Areas of interest are targeted and
tilt series’ are collected. Images are then processed and tomograms reconstructed. Finally, single
tomograms can be analysed (segmented tomogram from [131]) or multiple sub-volumes can be
picked and averaged, producing a higher resolution sub-tomogram average
62
J. L. Ferreira et al.
illustrate this with examples at the end of this chapter.
ECT is not restricted to imaging unique individual cells and other specimens,
however: if multiple cryo-tomograms of identical structures are collected, these
identical single particles can be computationally extracted, aligned, and averaged, an
approach known as subtomogram averaging, subvolume averaging, or single-particle
electron tomography. This approach ‘averages out’ noise and reinforces signal, greatly
improving the signal-to-noise ratio, with the potential for boosting resolution from
nanometres to Ångstroms. Because cryo-tomograms are in 3-D, particles can be
extracted even from cryo-tomograms of crowded environments such as the cytoplasm,
enabling in situ structure determination [5]. Advances in hardware and software have
increasingly seen subtomogram averaging applied to structural biology problems, and
a growing number of subtomogram averages have been determined to sub-nanometre
resolutions, enabling fitting of crystal structures into densities of macromolecular
complexes in situ. The boundaries continue to blur between ECT and more traditional
single-particle cryo-EM approaches to structure determination, particularly in light of
the considerable advances in contemporary cryo-EM [6].
This chapter focuses on the workflow of a biological ECT research project,
discussing each step consecutively with emphasis on choices faced by the
researcher, and the rationale behind the decision. We start with an overview of the
process and underlying concepts, followed by in-depth discussion of specimen
choice, preparation, and vitrification; background on microscope hardware setup;
software and data collection parameters; and tomogram reconstruction, processing,
and interpretation. We intend this chapter to provide an overview and introduction
for those training in ECT and many topics are covered in greater depth in subsequent chapters of this book.
c
Fig. 3.1 An overview of the electron tomography workflow. ECT can be used to image a wide
range of biological specimens from purified protein complexes to large eukaryotic cells and
tissues. The sample preparation and vitrification method is primarily dictated by the size of the
specimen being imaged. Small specimens (up to a few µm thick) can be plunged into a liquid
cryogen, causing rapid vitrification before the formation of crystalline ice can occur. Large
specimens (more than a few µm thick) must be high pressure frozen. As imaging of specimens
thicker than *500 nm is greatly hindered by excessive inelastic scattering, a method of thinning
must be chosen for thicker samples. FIB-milling can be used to mill away part of the specimen,
leaving a thin lamellae to be imaged, or cryo-sectioning can be used to produce thin slices of a high
pressure-frozen sample. Grid preparation remains a major bottleneck in ECT and the process of
sample preparation and vitrification may need to be repeated multiple times before data-collection
can begin. When ready, the grids are loaded into the microscope and data-collection parameters are
setup according to the type of specimen and biological question. Areas of interest are targeted and
tilt series’ are collected. Images are then processed and tomograms reconstructed. Finally, single
tomograms can be analysed (segmented tomogram from [131]) or multiple sub-volumes can be
picked and averaged, producing a higher resolution sub-tomogram average
62
J. L. Ferreira et al.