Transmission Electron Microscopy


Electron Microscopy (EM) is the method used to obtain high resolution images of plant and animal tissue. There are two basic types of electron microscopy: scanning electron microscopy, or SEM, and transmission electron microscopy, or TEM. SEM yields high resolution surface morphology from a secondary electron beam, which ‘bounces off’ the sample, is collected in a scintillator, and displayed onto a monitor as a relief image showing the surface features of the sample. TEM yields high resolution ultrastructural detail from a transmitted electron beam through the tissue, allowing morphometric analyses of the internal detail of sectioned cells or direct preps of macromolecules, localizing where exogenous substances reside within a cell, or immunologically marking specific antigens in a cell. In the above montage on the left, the image in the lower left-hand corner is an SEM image, showing the surface morphology of the sample. The rest of the montage contains TEM images, showing the internal structural detail of the cells. Although there are many more aspects of electron microscopy than presented here, this tutorial will focus on basic TEM and immunolocalization applications. 

Different steps in the preparation of samples for electron microscopy usually include tissue fixation, embedding, sectioning, and staining, procedures that take several days to complete. Fixation, a process that basically kills the cells and “locks” in place the cellular components, may take place before or after the tissue is excised from the subject (plant or animal), and most often utilizes an aldehyde, preferably glutaraldehyde, as the fixation agent. Embedding the tissue involves the removal of moisture and the subsequent infiltration with resins, usually epoxy or an acrylic.  After the sample is embedded, ultrathin sections are cut and placed on small (3mm) grids. The staining may be with electron-dense stains, such as uranyl acetate, or may include immunological procedures using antibodies and, usually, gold particles.  Once the tissue preparation has been completed, the imaging takes place in a transmission electron microscope. The UF CoM EM Core uses a Hitachi H7600 TEM (Hitachi High Technologies America, Inc., Schaumburg, IL), shown above on the right. 

An alternative preparation, often used for bacteria, viral particles, and other macromolecules are the whole mount negative staining procedure. In this method, a suspension of the sample is placed on a formvar-coated grid, stained with an electron-dense stain and allowed to dry before viewing in the TEM, a process taking only minutes before completion.  

Sample Preparation

Tissue collection and Fixation:  For most tissues, in vivo perfusion fixation of tissue produces superior tissue morphology compared to immersion fixation. Depending on the tissue being collected and the experimental animal used, specific perfusion protocols are available. In vivo perfusion-fixation removes blood from the tissues during a pre-fixation rinse and allows rapid fixation of cells and retention of tissue architecture by delivering the fixative directly while maintaining open vessels, open lumens, and interstitial spaces. Many times it is not convenient to perfuse the animal, so collections of the tissue may be made by biopsy or excision. It is vitally important to work quickly and place the excised tissue immediately into the fixative, where it should be dissected into small (~1mmpieces) so it can more effectively absorb the fixative.   

Many fixation solutions and techniques have evolved over the years, and there are numerous articles and books dedicated to fixation. It is vitally important to select the best fixative for the tissue you are preparing for electron microscopy. For immunolocalization studies by light microscopy or electron microscopy, we typically recommend 2 to 4% paraformaldehyde-lysine-periodate (PLP) as the primary fixative. Both the fixative and the carrier agent (buffer) have profound effects on the ultimate results you will see at magnifications of 1,000 to 200,000 times. Paraformaldehyde, a fixative regularly used in light microscopy, infiltrates tissue rapidly, but breaks down quickly into formic acid and does not maintain fixation qualities for a long period of time, thus allowing supposedly-fixed sensitive cellular components (for example, mitochondria) to deteriorate quickly, sacrificing morphology. It is recommended that paraformaldehyde be made fresh within hours of use. Glutaraldehyde, however, penetrates the tissue more slowly, but crosslinks more securely with the cell structures, and so becomes a more satisfactory fixative for structures to be observed at high magnifications. It is important to use an EM-grade glutaraldehyde, which can be purchased from vendors of EM products.  Many EM fixative solutions, therefore, contain a mixture of both aldehydes, thereby utilizing the ‘good’ features and reducing the ‘bad’ effects of both. Two published and commonly used fixatives for EM are Karnovsky’s fixative1 and Trump’s, also known as McDowell-Trump’s fixative2

Buffers are a necessary component of every fixative, in order to keep the tissue at a pH and an osmolarity physiologically beneficial to the tissue being studied. Most plant and animal tissues are more ‘comfortable’ at a pH of 7.3 to 7.4. Using a fixative with an osmolarity that is too high or too low can have an adverse effect on the morphology of the tissue. Thus the osmolarity of the buffer used for rinsing and as a base for the fixative should be near the ambient osmolarity of the tissue being collected. Aldehydes (paraformaldehyde, glutaraldehyde) add to the measured osmolarity of the fixative solution, but because they cross cell membranes freely, they do not add to the effective osmolarity in the tissue. To avoid using fixative that was mixed incorrectly or with a bad reagent, we always check and record the final pH and osmolarity of the finished fixative to make sure that they are within the desired range before continuing the experiment. We use a μOsmette osmometer (Precision Systems, Inc., Natick, MA) to measure our solution osmolarity by freezing point depression; we have found that vapor pressure osmometers do not read aldehyde solutions correctly. 


  1. Karnovsky, MJ. A formaldehyde-glutaraldehyde fixative of high osmolality for use in electron microscopy.  J. Cell Biol. 27:137-138A. 1965. 
  2. EM McDowell, BF Trump. Histologic fixatives suitable for diagnostic light and electron microscopy.  Arch Pathol Lab Med. Aug;100(8):405-14. 1976. 

Embedding:  Both plant and animal tissues exist in an aqueous environment in nature.  In general, embedding media are not compatible with water, therefore all traces of water must be removed from the tissue before it will absorb the media. This is done through a process known as dehydration, where the tissue is immersed in gradients of a transition fluid that is compatible with the media, usually ethanol or acetone. There is often a step, before or during the dehydration, when it is desirable to introduce electron-dense contrast enhancers (en-block staining). The most commonly used enhancers are osmium tetroxide, uranyl acetate, and ruthenium tetroxide.  When the dehydration steps are finished, the tissue samples are placed in the media for several hours to become completely infiltrated.  Finally, the tissue sample and media are placed in a mold and polymerized, usually by heat or by UV light.  (LINKS to embedding tutorials coming soon) 


 Conventional epoxy blocks – Black tips are tissue, ~1mm3 

Sectioning:   To view the embedded tissue on the TEM, ultra-thin slices must be placed on 3mm grids that will fit into the microscope’s sample holder. This involves using an ultramicrotome, an instrument designed specifically for cutting slices from 40nm to 2 μm thick. The microtomes in the EM Core are from Leica Microsystems, Inc. (Buffalo Grove, IL). 


In order to get the somewhat flawless images that we need, a diamond knife is used (see next page, left image) to cut the sections, although glass knives may also be used. The diamond, usually 2.5 to 3mm wide, is attached to a ‘boat’, which holds water on which the sections will float. We usually cut 0.5 to 1.0 μm thick sections to preview the sample, so we can verify that the sections will contain the subject we wish to image on the TEM. These semi-thin sections are placed on a microscope slide, stained with 1% Toluidine Blue-O, and viewed on a light microscope.   

Rat kidney, 1% glutaraldehyde, osmium

Rat kidney tol blue

Epoxy embedding, Tol. Blue stain

Mouse kidney, no osmium

mouse kidney tol blue

Acrylic embedding, Tol. Blue stain 

After sample verification, ultrathin sections ranging from 50 to 70nm thick are cut, carefully removed from the water and placed on the 3mm TEM grids (see right image below). Ultrathin epoxy sections are placed on copper grids for ultrastructural transmission electron microscopy, while ultrathin sections of acrylics on grids are used for immunogold labeling. Since the immunological procedure involves many steps which can ultimately weaken and tear the sectioned tissue, and some of the reagents used in the procedure may react with the grid, it is advisable to use nickel grids that have been coated with a formvar polymer that has been strengthened with a thin layer of carbon. These grids can be coated by the investigator in his or her lab, but it requires access to a carbon evaporator and it is a tedious process with a high failure rate; therefore, we purchase our formvar-coated TEM grids ready-made from Ladd Research Industries, Inc. (Williston, VT).

Diamond knife
copper grid zoomed in

Staining:  When staining grids for imaging on the TEM, it is important to work in a clean area, and to use recently-made stains in order to prevent artifacts attributed to stain precipitates. As an initial stain, we routinely place the grid – tissue side down – for 4 minutes on a drop of 8% uranyl acetate (UAc) (Ladd Research Industries), in 50% ethanol, filtered immediately before use. UAc is sensitive to light, so this step should be carried out in the dark (cover with foil). To avoid precipitate on the sample, we do not use a stain solution older than 30 days. (Caution: Uranyl acetate, made from ‘depleted’ uranium, is still radioactive. Handle with care under a hood, wear gloves, and dispose of all contaminated waste in a properly labeled radioactive waste container.) We usually follow the UAc stain with one minute on a drop of 0.25% aqueous lead citrate (Electron Microscopy Sciences, Hatfield, PA). Again, we make the lead stain fresh monthly, or whenever the solution becomes cloudy. The grids should be allowed to air dry thoroughly before inserting them into the TEM. 

EM kidney

Example of Transmission Electron Microscopy – tissue processed for optimal ultrastructure 

Immunolocalization:   As a general rule, samples that are intended for immunolocalization should initially be fixed in a very low concentration of glutaraldehyde, preferably 0.25 to 0.5%, to maintain antigenicity.  Mixtures of glutaraldehyde with paraformaldehyde are commonly used since paraformaldehyde does not seem to interfere with antigenicity.  Another obstacle to the immune response is the embedding media; acrylic resins usually work well, while epoxy resins are resistant and seldom show good results.  An alternative would be to avoid the use of resin and aldehydes altogether, using unfixed frozen samples that have been immersed in a cryoprotectant. 

In the UF CoM EM Core, tissues destined for immunolocalization are usually preserved and processed with a combination of 2 to 4% paraformaldehyde and 0.25 to 1% glutaraldehyde to balance immunoreactivity with ultrastructural preservation. Methods are customized for each project. Options include pre- and post-embedding methods, using either acrylic-embedded samples for on-grid immunogold labeling, free-floating vibratome sections of preserved tissues for pre-embedding immunoperoxidase or immunogold labeling, and on-grid labeling of ultrathin cryosections.

Example of Transmission Electron Microscopy with on-grid immunogold labeling, high magnification of the apical region of  Type A intercalated cells in isolated, perfused mouse cortical collecting duct labeled for H+ ATPase. 





ANGII causes a dramatic difference in morphology of type A ICs and apical plasma membrane H+-ATPase.