3.1.2 Observation
and Trouble Shooting
Ideally the sample should be evenly distributed with little interactions between individual particles (compare Figs. 1f and 3a). The
macromolecules should be included in a layer of stain of the same
thickness as the particle to sustain optimally the sample, to reveal all
parts and to obtain the best contrast. It is often necessary to screen
several places on the grids to find optimal staining. The following
indications may help to improve staining.
1. Poor staining. In a standard bright field transmission electron
microscope, your protein sample should appear as a lighter area
embedded in a darker background. If your sample is poorly
stained, you will only see fingerprints of your protein (Fig. 3b)
or worse, and your sample may be positively stained and appear
as dark spots on a lighter background. You may adjust the glow
discharge conditions to render the support more hydrophilic
and make sure that the staining solution is not removed too fast
by wetting the filter paper beforehand. Adding 0.5% glycerol
may help to reduce the drying speed and improve staining.
2. Excess staining. Your sample may be well embedded, but a
thick layer of stain may be superimposed to the particles, thus
reducing the contrast and obscuring fine details of your sample
(Fig. 3c). Sometimes wiping away the stain faster can help.
Check that your sample buffer does not contain large excess
of sucrose of glycerol which increases its viscosity. In this case,
an additional washing step can be added.
3. Presence of stain crystals. Uranyl salts precipitate at neutral pH
or in contact with phosphate buffer and form small crystals
visible in EM. Make sure that these reagents are absent or
sufficiently dilute in the last staining step and that the pH is
suitable.
4. Damaged/destroyed proteins. Some fragile protein complexes
may be altered during grid preparation and even fall apart
during the dilution, adsorption, or staining steps. Chemical
cross-linkers such as glutaraldehyde or BS3 may stabilize the
sample. As a standard procedure, glutaraldehyde is added at a
final concentration of 0.1–0.5% for 2 min, before adsorption on
the EM grid. Cross-linkers may lead to protein aggregation,
and therefore, cross-linking time and concentration need to be
adjusted. Adding 1% glycerol in the staining solution may help
sustaining fragile samples and prevent their collapse during the
drying step.
5. Small number of particles. When a smaller number of particles
than expected is observed, the sample may form aggregates.
Aggregates are not always detected because of their small interaction interface with the carbon support which can lead their
dissociation during the washing or staining steps. Alternatively,
Specimen Optimization for Single Particle cryo-EM
251
and Trouble Shooting
Ideally the sample should be evenly distributed with little interactions between individual particles (compare Figs. 1f and 3a). The
macromolecules should be included in a layer of stain of the same
thickness as the particle to sustain optimally the sample, to reveal all
parts and to obtain the best contrast. It is often necessary to screen
several places on the grids to find optimal staining. The following
indications may help to improve staining.
1. Poor staining. In a standard bright field transmission electron
microscope, your protein sample should appear as a lighter area
embedded in a darker background. If your sample is poorly
stained, you will only see fingerprints of your protein (Fig. 3b)
or worse, and your sample may be positively stained and appear
as dark spots on a lighter background. You may adjust the glow
discharge conditions to render the support more hydrophilic
and make sure that the staining solution is not removed too fast
by wetting the filter paper beforehand. Adding 0.5% glycerol
may help to reduce the drying speed and improve staining.
2. Excess staining. Your sample may be well embedded, but a
thick layer of stain may be superimposed to the particles, thus
reducing the contrast and obscuring fine details of your sample
(Fig. 3c). Sometimes wiping away the stain faster can help.
Check that your sample buffer does not contain large excess
of sucrose of glycerol which increases its viscosity. In this case,
an additional washing step can be added.
3. Presence of stain crystals. Uranyl salts precipitate at neutral pH
or in contact with phosphate buffer and form small crystals
visible in EM. Make sure that these reagents are absent or
sufficiently dilute in the last staining step and that the pH is
suitable.
4. Damaged/destroyed proteins. Some fragile protein complexes
may be altered during grid preparation and even fall apart
during the dilution, adsorption, or staining steps. Chemical
cross-linkers such as glutaraldehyde or BS3 may stabilize the
sample. As a standard procedure, glutaraldehyde is added at a
final concentration of 0.1–0.5% for 2 min, before adsorption on
the EM grid. Cross-linkers may lead to protein aggregation,
and therefore, cross-linking time and concentration need to be
adjusted. Adding 1% glycerol in the staining solution may help
sustaining fragile samples and prevent their collapse during the
drying step.
5. Small number of particles. When a smaller number of particles
than expected is observed, the sample may form aggregates.
Aggregates are not always detected because of their small interaction interface with the carbon support which can lead their
dissociation during the washing or staining steps. Alternatively,
Specimen Optimization for Single Particle cryo-EM
251