inadvertently eliminate the feature of interest due to an inability to determine the
position of the feature relative to the z axis.
1.13 Conclusion
The functions of viruses and cells are linked inextricably with structure. Electron
tomography is therefore pivotal to an enhanced understanding of biology. This
chapter hopefully makes clear that the practise of electron tomography has matured
to the point of being useful to specialist biologists who have minimal training in
physics and computer programming. A capable microscopist, however, will be able
to troubleshoot effectively and adjust suboptimal beam conditions that would
otherwise lead to the collection of inferior data. They will also be more wary of the
numerous traps of image analysis [126].
Visualisation using electron tomography must always be considered in the light
of structural preservation. Tomograms of plastic-embedded, stained samples provide
information about the architecture and arrangements of organelles: ultrastructure.
These amplitude contrast -dominated (stained) specimens are easier to segment, and
all other factors considered, they supplant the need for stereological techniques and
their assumptions. Importantly, quantitation of e.g. the volume fraction of mitochondria must take into account anisotropic (and frequently disregarded) shrinkage
and mass loss that occurs during pre-irradiation and tilt-series acquisition. These
considerations do not attempt to take into account the shrinkage, swelling, extraction, or redistribution of components that precedes image acquisition.
Tomograms of frozen-hydrated cells are effectively low-resolution snapshots of a
cell’s proteome. They are mandatory for studies of structural biology. Studies of
function depend on the maintenance of spatial relationships between macromolecular
assemblies. This is afforded by vitrification, which is routine for viruses and small
cells [127, 128], and has become feasible for mammalian cells thanks to cryo- electron
tomography of vitreous sections furnished by focussed ion beam milling [67, 129].
The majority of organelles are easily recognisable, and we have now entered an era of
being able to make this claim for many large molecular assemblies in situ, where
unambiguous structural signatures allow tomograms to be searched via sophisticated
3D cross-correlation functions. In the absence of improved resolution, smaller and/or
less distinctive structures require correlative techniques for meaningful localisation.
Thankfully, fluorescent and/or electron-dense markers for this purpose have been
identified that are compatible with vitrification and even freeze-substitution strategies
[17]. Also, photobleaching is less problematic at vitreous temperatures [130, 131].
The importance of recent developments in phase plate technology and nanofabrication cannot be understated. The next major technical advances in electron
tomography are likely to emerge as a result of improvements to sample fabrication that
will in turn facilitate faster and more objective 3D segmentation. Applications will
benefit most from cryo-correlative microscopy, particularly sub-diffraction-limit
1 Electron Tomography: A Primer
25
position of the feature relative to the z axis.
1.13 Conclusion
The functions of viruses and cells are linked inextricably with structure. Electron
tomography is therefore pivotal to an enhanced understanding of biology. This
chapter hopefully makes clear that the practise of electron tomography has matured
to the point of being useful to specialist biologists who have minimal training in
physics and computer programming. A capable microscopist, however, will be able
to troubleshoot effectively and adjust suboptimal beam conditions that would
otherwise lead to the collection of inferior data. They will also be more wary of the
numerous traps of image analysis [126].
Visualisation using electron tomography must always be considered in the light
of structural preservation. Tomograms of plastic-embedded, stained samples provide
information about the architecture and arrangements of organelles: ultrastructure.
These amplitude contrast -dominated (stained) specimens are easier to segment, and
all other factors considered, they supplant the need for stereological techniques and
their assumptions. Importantly, quantitation of e.g. the volume fraction of mitochondria must take into account anisotropic (and frequently disregarded) shrinkage
and mass loss that occurs during pre-irradiation and tilt-series acquisition. These
considerations do not attempt to take into account the shrinkage, swelling, extraction, or redistribution of components that precedes image acquisition.
Tomograms of frozen-hydrated cells are effectively low-resolution snapshots of a
cell’s proteome. They are mandatory for studies of structural biology. Studies of
function depend on the maintenance of spatial relationships between macromolecular
assemblies. This is afforded by vitrification, which is routine for viruses and small
cells [127, 128], and has become feasible for mammalian cells thanks to cryo- electron
tomography of vitreous sections furnished by focussed ion beam milling [67, 129].
The majority of organelles are easily recognisable, and we have now entered an era of
being able to make this claim for many large molecular assemblies in situ, where
unambiguous structural signatures allow tomograms to be searched via sophisticated
3D cross-correlation functions. In the absence of improved resolution, smaller and/or
less distinctive structures require correlative techniques for meaningful localisation.
Thankfully, fluorescent and/or electron-dense markers for this purpose have been
identified that are compatible with vitrification and even freeze-substitution strategies
[17]. Also, photobleaching is less problematic at vitreous temperatures [130, 131].
The importance of recent developments in phase plate technology and nanofabrication cannot be understated. The next major technical advances in electron
tomography are likely to emerge as a result of improvements to sample fabrication that
will in turn facilitate faster and more objective 3D segmentation. Applications will
benefit most from cryo-correlative microscopy, particularly sub-diffraction-limit
1 Electron Tomography: A Primer
25