3.2 ECT: Overview and Key Concepts
In essence, ECT involves calculating the 3-D structure of a flash-frozen specimen to
nanometre-resolution (for an overview workflow, see Fig. 3.1). A well-chosen
specimen is first flash-frozen, vitrifying it to preclude the formation of ice crystals.
The specimen is then inserted into an appropriately configured microscope, and a
‘tilt series’ of 2-D projection images of the specimen are acquired over a range of
angles. Although in 2-D projection images the 3-D information is collapsed along
the Z-axis, the full tilt series can be used to computationally extrapolate this 3-D
information. Identical subvolumes within the tomogram(s) can additionally be
subjected to subtomogram averaging. To design imaging sessions and interpret
images it is critical to understand a number of key concepts outlined below.
It is important to understand aspects of electron scattering relevant to ECT.
ECT’s ability to image biological material without stains or fixatives is facilitated
by the strong scattering of electrons by matter, enabling discernible imaging contrast from individual molecules [7]. The flipside of this strong interaction is that
many electrons deposit energy into the biological material. When an electron is
scattered by the specimen it may be scattered elastically or inelastically [7];
inelastically scattered electrons break bonds and lose energy, as a result contributing incoherently to the image as noise. Additionally, minimizing the damage
caused by inelastically scattered electrons necessitates deliberately keeping electron
exposure of the specimen to a minimum, resulting in very low signal-to-noise
ratios, and making interpretation of cryo-tomograms challenging. Furthermore the
strong interaction with matter means that samples thicker than a few hundreds of
nanometres become intractable to imaging due to the uninterpretable inelastic (or
multiple) electron scattering.
Another aspect of electron scattering important to interpretation of electron
microscopy images is the effect of the considerable spherical aberration in conventional electron lenses. This aberration alters the phase of scattered electrons; the phase
alteration is a function of the angle of scatter—i.e., the resolution of the image
component. This unintuitively leads to oscillation of the contrast of the image as a
function of resolution, a function referred to as the Contrast Transfer Function, or
CTF. As a result, features of certain resolutions have zero contrast, while others have
negative contrast—obviously a hindrance to interpretation of the image. A positive
aspect of the CTF is that it is a function of the focus of the lens (together with the
electron acceleration voltage and the spherical aberration coefficient of the objective
lens), conveniently meaning that appropriate, deliberate defocusing of the lens can be
used to produce phase contrast in specimens, and until recently this was the only
method to achieve phase contrast because phase plates (as used in conventional visible
light microscopy) had not been developed sufficiently (and still remain rare). The
resolution at which the first contrast inversion occurs in the oscillating CTF is colloquially referred to as the “first zero”, a particularly important concept in ECT. It is
possible to mathematically correct for the CTF contrast inversions in the image past
this “first zero”, although some data is lost, and some attenuated, as a result.
64
J. L. Ferreira et al.
In essence, ECT involves calculating the 3-D structure of a flash-frozen specimen to
nanometre-resolution (for an overview workflow, see Fig. 3.1). A well-chosen
specimen is first flash-frozen, vitrifying it to preclude the formation of ice crystals.
The specimen is then inserted into an appropriately configured microscope, and a
‘tilt series’ of 2-D projection images of the specimen are acquired over a range of
angles. Although in 2-D projection images the 3-D information is collapsed along
the Z-axis, the full tilt series can be used to computationally extrapolate this 3-D
information. Identical subvolumes within the tomogram(s) can additionally be
subjected to subtomogram averaging. To design imaging sessions and interpret
images it is critical to understand a number of key concepts outlined below.
It is important to understand aspects of electron scattering relevant to ECT.
ECT’s ability to image biological material without stains or fixatives is facilitated
by the strong scattering of electrons by matter, enabling discernible imaging contrast from individual molecules [7]. The flipside of this strong interaction is that
many electrons deposit energy into the biological material. When an electron is
scattered by the specimen it may be scattered elastically or inelastically [7];
inelastically scattered electrons break bonds and lose energy, as a result contributing incoherently to the image as noise. Additionally, minimizing the damage
caused by inelastically scattered electrons necessitates deliberately keeping electron
exposure of the specimen to a minimum, resulting in very low signal-to-noise
ratios, and making interpretation of cryo-tomograms challenging. Furthermore the
strong interaction with matter means that samples thicker than a few hundreds of
nanometres become intractable to imaging due to the uninterpretable inelastic (or
multiple) electron scattering.
Another aspect of electron scattering important to interpretation of electron
microscopy images is the effect of the considerable spherical aberration in conventional electron lenses. This aberration alters the phase of scattered electrons; the phase
alteration is a function of the angle of scatter—i.e., the resolution of the image
component. This unintuitively leads to oscillation of the contrast of the image as a
function of resolution, a function referred to as the Contrast Transfer Function, or
CTF. As a result, features of certain resolutions have zero contrast, while others have
negative contrast—obviously a hindrance to interpretation of the image. A positive
aspect of the CTF is that it is a function of the focus of the lens (together with the
electron acceleration voltage and the spherical aberration coefficient of the objective
lens), conveniently meaning that appropriate, deliberate defocusing of the lens can be
used to produce phase contrast in specimens, and until recently this was the only
method to achieve phase contrast because phase plates (as used in conventional visible
light microscopy) had not been developed sufficiently (and still remain rare). The
resolution at which the first contrast inversion occurs in the oscillating CTF is colloquially referred to as the “first zero”, a particularly important concept in ECT. It is
possible to mathematically correct for the CTF contrast inversions in the image past
this “first zero”, although some data is lost, and some attenuated, as a result.
64
J. L. Ferreira et al.