Scanning Transmission Electron Microscopy Facility
STEM Specimen Preparation
Freeze-dried specimens for mass measurements in the STEM are prepared by the
wet film method. 2.3 mm titanium grids are coated with a thick holey carbon
film. The open areas are 5-10 µm in diameter which will support the thin
(2-3 nm thick) carbon film substrate. The thin carbon film is prepared by
ultra-high vacuum evaporation onto a freshly cleaved crystal of rock salt.
The thin carbon film is then floated on a dish of clean water.
Grids covered with holey film, assembled in rings and caps for handling in the
microscope, are placed face down on the floating thin carbon film. The grids
are picked up from above one at a time such that the thin carbon film retains
a droplet of water.
This water is exchanged by washing and wicking, either with water or with
injection buffer for the specimen. Two µl of tobacco mosaic virus (TMV)
solution, at a concentration of 100 µg/ml, are injected into the drop and
allowed to stand for 1 min. The TMV is both a qualitative and a quantitative
internal control for all specimens.
After four washings, 2 µl of the specimen (its concentration is determined
primarily by its size) is injected into the drop and allowed to stand for 1
min. Ten more washings are performed. Half of these can be with injection
buffer, but the final few must be with a volatile buffer such as 20 mM
ammonium acetate or water.
After the final wash, the grid is pinched between two pieces of filter paper
(leaving a retained layer of solution less than 1 µm thick) and immediately
plunged into liquid nitrogen slush. Six grids are transferred under liquid
nitrogen to an ion-pumped freeze dryer with an efficient cold trap, freeze
dried overnight by gradually warming to -80°C and transferred under vacuum
to the STEM.
Examples of good STEM specimens are shown in this
annotated gallery of:
1. Freeze dried earth worm hemoglobin (stained and unstained)
2. Freeze dried bacteriophage T4 (with TMV standard)
3. and GroEL chaperone complexes (stained).
Although the STEM is an extraodinary machine, there are limitations to
the kinds of specimens that it can image with good results.
Because it is designed for applications in 'molecular biology" the
molecular weight of the specimens should be in the range of 40 kilo daltons
to 800 million daltons. Specimen purity, concentration and buffering, may
also make a sample unsuitable for use in the STEM.
Here, the following limitations are discussed:
- Buffers and Salts
- Other Problems
The STEM is capable of 0.25 nm resolution, but freeze-dried
biological specimens usually have approximately 2 nm resolution. The
STEM can provide valuable information on individual particles in the
molecular weight range from 40 kD to several hundred MD. In general,
the larger the particle, the more accurate the mass measurement can
be. Success in looking at small proteins (less than 100 kD) will depend
on their conformation. If parts of them are very extended, they may
suffer more mass loss and the mass measurements will be low.
Proteins of molecular weight less than 40 kD are likely to be too small to
yield good mass measurements unless they are unusually dense and compact.
tRNA can be seen in the STEM, but it is very compact and nucleic acids are
more radiation resistant than proteins. Double-stranded DNA is also visible
in the STEM for the same reasons.
At the other extreme, while viruses
(and sub-viral particles of large viruses) such as herpes simplex have given
excellent results, the STEM is probably not the best choice for
looking at very large structures such as bacteria, cells, or organelles.
They are essentially opaque, i.e. saturated, in the STEM and so yield
little information. For filaments or two- dimensional arrays, the actual
size is not a problem. It is only the mass density that is of concern, so
they are fine if they are spread out enough and not too dense.
The physical purity of the specimens is very important for STEM
preparations. Essentially everything is visible in the STEM, and
impurities may well not wash off the grid. Most biologists have some
assay for the biological activity of their specimens. However, many
preparations contain biologically inert but physically visible material
that may make it difficult to obtain good STEM data. Particulates from
columns or gels are a problem. Additives for activity such as BSA,
PEG, or trypsin are a problem. Precipitations can concentrate the
specimen, but also enrich contaminants. If a multi- subunit enzyme has
90% activity, but the remaining 10% has fallen apart, that can be very
visible in the STEM. Frequently, a final purification over an appropriate
sizing column (such as an A5M) where the intact specimen comes off in
the void volume gives adequate purity. At times, the particles of interest
can be distinguished from contaminants. However, it is clearly
preferable to eliminate contaminants when possible.
The necessary concentration of a given specimen to give a good
distribution of particles on a grid is somewhat empirical. However, for
data taking it is very important. If the particles are too close together,
they cannot be measured individually, but if they are too sparse, it is
very difficult to obtain both particles and the control TMV in an image.
Also, searching for rare events in the STEM is slow. As a guide, large
particles such as viruses need to be applied to the grid at a relatively
high concentration such as 200-300 µg/ml. For smaller proteins, a
concentration of 50 µg/ml might be adequate. For any given specimen, it
is not known how well it will adhere to the carbon grid. For a new
specimen, a series of dilutions is usually made. A highly concentrated
specimen can always be diluted, but there is little that can be done with
a specimen that is too dilute.
The ideal specimen for the STEM is one that can be applied to the
grid in water or 20 mM ammonium acetate which volatilizes during
freeze-drying. However, many biological specimens have additional
requirements. Most salts and buffers can be tolerated in the injection
buffer for the specimen. The assumption is that if the specimen is
absorbed to the carbon grid in a few washes of its own buffer, it will be
able to tolerate some final washes with a volatile solution such as water
or ammonium acetate.
For example, low levels (1-2 mM) of Mg++,
Ca++, ATP (or non-hydrolyzable analogues) are often needed for
biological activity and usually will wash away. Low concentrations of
glycerol or sucrose also usually wash off, as will DTT or
mercaptoethanol. NaCl at levels as high as 100 mM usually washes off
with ammonium acetate, but KCl at more than a few mM occasionally
causes problems. Phosphate buffers should be avoided if possible since
they usually leave a background high in "spots", which is a real problem
for interpreting site-specific cluster labeling experiments. Tris buffers
often, but not always, have a bad background, whereas Hepes, Mops,
and Pipes are usually better. For a new specimen in an unusual buffer,
a grid of TMV by itself washed with the buffer is made as a control.
Some specimens, such as membrane proteins, have to be solubilized in
a detergent and these are always difficult to work with although some
successful results have been obtained.
Some specimens which have met all the above criteria simply do not
work out. Harm can come to the specimen during its attachment to the
carbon film (a part can attach firmly while the rest of the molecule flails
around), during the freeze-drying process, and during the data-taking
process (from radiation damage in the electron beam). An occasional
specimen will not attach to the carbon film. Double-stranded DNA, for
example, does not attach reliably even at high concentrations.
However, with polylysine pre-treatment of the carbon grid, DNA will bind
at concentrations of less than 1 µg/ml. Polylysine pre-treatment will also
help some other types of specimens. All proteins denature to some extent
at the air-water interface of the drop on the grid. Usually, washing and
wicking removes this denatured film. However, some proteins continue to
denature at the interface until there is nothing left on the grid.
Sometimes washing with low concentrations of organic solvents such as
ethanol or acetonitrile will help. Some specimens which fall apart on the
carbon film benefit from a brief fixation with glutaraldehyde just prior to
injection. If this does help, the increase in mass is less than 10%.
Occasionally specimens will be all right on some parts of the grid but not
in other areas. The STEM staff is continually working on specimen
preparation, especially with "difficult" specimens, to try to solve these
Last Modified: June 12, 2009
Please forward all questions about this site to: