10. Trim the metal on both sides at 100 mm/s to a depth of
200 μm.
11. Trim through the frozen dextran on each side in stages, each
approximately one-fourth of the width of the diamond blade,
until the final width of the block-face is reached.
The section collection technique is precisely described in reference [35] and thus will not be repeated here. The nominal cryosection thickness should be around 100 nm for tomography.
2.5 Other Samples
These include any purified organelle, compartment [37], in vitro
assembly [38], or purified protein complex. The sample should be
purified and concentrated as gently as possible to avoid damage
[28, 39]. As many of these samples are thin, they can be suitable for
high-resolution tomography. These samples are compatible with
the different material sections discussed above. However, for highresolution samples, we typically use Protochip C-flat 2/2 holey
carbon grids with 300 mesh, and mix the sample with 10 nm
gold fiducials.
3 Data Collection
A detailed protocol for data collection and processing is available in
reference [40].
3.1 Data Acquisition
Using SerialEM:
Conventional
Tilt-Series
After loading the samples to the microscope (in our lab a Titan
Krios) and assuming that the microscope is already aligned, data
collection can be started. Currently, we use SerialEM [41] to collect
our data. We refer the reader to the tutorial of SerialEM for detailed
discussion of what the software can do. In general, we start by
making an atlas of the grid at low magnification (82Â). This will
give a general idea of how the grid looks and the ice gradient.
Subsequently, a montage of some good squares (at magnification
2245Â) is collected providing a better view of the cells on these
squares. Subsequently, points are added on cells in the holes
(or other interesting cells in the montage). Then, the stage is
moved to each of these points and a preview image (with low
dose) is taken for that cell. The position of the cell is optimized,
for instance, centered or offset if one wants to collect data from a
pole of the cell). After that, a view image (at medium magnification,
2245Â) is taken and is saved and used as the anchor state. View
images are used to help tracking and aligning the target during data
collection. One can further proceed with the process for all the
required targets by pressing the “Anchor map” button. At the end,
click the “TS” (tilt-series) option for the targets where data will be
collected and tilt-series parameters are filled in as required (e.g.,
starting and ending angles, increment degrees, defocus, how to
change exposure time with tilting angle). Once that is done, click
98
Mohammed Kaplan et al.
200 μm.
11. Trim through the frozen dextran on each side in stages, each
approximately one-fourth of the width of the diamond blade,
until the final width of the block-face is reached.
The section collection technique is precisely described in reference [35] and thus will not be repeated here. The nominal cryosection thickness should be around 100 nm for tomography.
2.5 Other Samples
These include any purified organelle, compartment [37], in vitro
assembly [38], or purified protein complex. The sample should be
purified and concentrated as gently as possible to avoid damage
[28, 39]. As many of these samples are thin, they can be suitable for
high-resolution tomography. These samples are compatible with
the different material sections discussed above. However, for highresolution samples, we typically use Protochip C-flat 2/2 holey
carbon grids with 300 mesh, and mix the sample with 10 nm
gold fiducials.
3 Data Collection
A detailed protocol for data collection and processing is available in
reference [40].
3.1 Data Acquisition
Using SerialEM:
Conventional
Tilt-Series
After loading the samples to the microscope (in our lab a Titan
Krios) and assuming that the microscope is already aligned, data
collection can be started. Currently, we use SerialEM [41] to collect
our data. We refer the reader to the tutorial of SerialEM for detailed
discussion of what the software can do. In general, we start by
making an atlas of the grid at low magnification (82Â). This will
give a general idea of how the grid looks and the ice gradient.
Subsequently, a montage of some good squares (at magnification
2245Â) is collected providing a better view of the cells on these
squares. Subsequently, points are added on cells in the holes
(or other interesting cells in the montage). Then, the stage is
moved to each of these points and a preview image (with low
dose) is taken for that cell. The position of the cell is optimized,
for instance, centered or offset if one wants to collect data from a
pole of the cell). After that, a view image (at medium magnification,
2245Â) is taken and is saved and used as the anchor state. View
images are used to help tracking and aligning the target during data
collection. One can further proceed with the process for all the
required targets by pressing the “Anchor map” button. At the end,
click the “TS” (tilt-series) option for the targets where data will be
collected and tilt-series parameters are filled in as required (e.g.,
starting and ending angles, increment degrees, defocus, how to
change exposure time with tilting angle). Once that is done, click
98
Mohammed Kaplan et al.