Sputter shadowing improved by using a tungsten target

This work builds upon a previous paper (W. Colquhoun, 1984, J. Ultrastruct. Res. 87, 97) in which a sputter shadowing device was briefly described. The device allowed TEM specimens to be shadowed in a conventional sputter coater. Images obtained by sputter shadowing with a standard Au/Pd target were of good quality but were slightly inferior to the best that could be obtained by e--beam evaporation of tungsten. Here we show that construction and use of a tungsten target greatly improves the quality of the sputter shadowed deposit. Images of DNA and ribosomal subunits contrasted by sputter shadowing with tungsten are shown. The DNA images indicate that sputter shadowing with tungsten is a gentle contrasting technique. The sputter shadowed images of the 30 S ribosomal subunits show the major features of the particle revealed by evaporation shadowing using the most sophisticated of methods in that technology. Advantages of sputter shadowing are discussed and a rationale for the improved grain obtained by sputtering tungsten is suggested.     

THE SIZE OF THE CELLULOSE MICROFIBRIL

Recently the lateral width of the cellulose microfibril has been estimated as 30 A rather than about 150 to 200 A, by extrapolation of data from model shadowing experiments. The difference was attributed to a layer of metal deposited during shadowing. However, direct photographs of the same microfibrils parallel and perpendicular to the direction of shadowing, of unshadowed portions of microfibrils compared with shadowed portions of the same microfibrils, of silver-stained unshadowed microfibrils, and of unshadowed, unstained segments of microfibrils give no evidence of a layer of metal of this thickness in material shadowed under normal conditions. Furthermore, the evidence for microfibril strands of about 35 A in width from negative-staining experiments is subject to a bias from the form of the filaments and from variable positive adsorption of phosphotungstic acid by cellulose. Consequently, the conclusion that the true lateral width of native cellulose microfibrils is about one-fifth of the presently accepted value is not yet justified by unequivocal direct experimental evidence.     

An initial test of a method for the estimation of real mean particle size from shadowed samples

During shadowing, a "cap" of metal develops on small particles. This cap increases apparent particle with (measured normal to the shadowing direction) by an extent which cannot be predetermined. The extent of this increase in particle size (here defined as the "cap," X) is estimated in the present method by using opposite (180 degrees sample rotation) bidirectional shadowing. It is argued that the bidirectional cap is the sum of the two unidirectional caps, and therefore that X = 2A - (B + C), where A is the mean bidirectionally shadowed particle size, and B and C are the two mean unidirectionally shadowed particle sizes. As a validation of the method, the mean diameter of air-dried ferritin was estimated and the results appear to confirm the hypothesis (mean diameter by present method, 10.7 +/- 0.2 nm; mean diameter by previous methods, 10.89 nm).     

Hochauflösende Gefrierätzung

Zusammenfassung Die Anwendungsmöglichkeit der Gefrierätztechnik auf molekularbiologische und cytochemische Probleme ist vielfach durch das zu geringe Auflösungsvermögen der derzeit üblichen Simultanbeschattung mit Platin-Kohle begrenzt. Zwar lassen sich durch Beschatten mit höchstschmelzenden Metallen Abdrucke mit höherer Auflösung erzielen, jedoch scheinen bisher alle Versuche, diese Methode in der Gefrierätzung anzuwenden, fehlgeschlagen zu sein, da sie zu Präparatveränderungen führten. Außer hoher Auflösung müssen nämlich Beschattungsmethoden, um für die Gefrierätzung geeignet zu sein, auch folgende Bedingungen erfüllen: kurze Aufdampfzeit, geringe thermische Präparatbelastung und große chemische Stabilität des Abdruckes. Durch die Verwendung von Tantal-Wolfram als Aufdampfmaterial und die Konstruktion eines geeigneten Elektronenstrahl-Verdampfers, der eine Reduzierung der Wärmestrahlung und ein Ausschalten des Ionenbeschusses ermöglicht, ist es nunmehr gelungen, obige Bedingungen zu erfüllen und die Beschattung mit höchstschmelzenden Metallen auch in der Gefrierätzung erfolgreich einzusetzen. Die mit Tantal-Wolfram beschatteten Präparate sind entsprechenden Platin-Kohle-Abdrucken an Auflösung deutlich überlegen.Da die Wärmebelastung des Präparates in der Gefrierätztechnik von entscheidender Bedeutung ist, wurde sie für Ta/W und Pt/C-Beschattungen untersucht. Bei Ta/W-Be-dampfung liegt der für unsere Anordnung berechnete Wärmefluß bei 16 mW·cm^–2, der gemessene beträgt 23 mW·cm^–2. Bei Pt/C-Verdampfung ist die berechnete Wärmebelastung gleich hoch wie bei Ta/W. Der an einem üblichen Verdampfer gemessene Wärmefluß beträgt aber 41 mW·cm^–2. Rechnung und Experiment stimmen darin überein, daß die Versuchsbedingungen, wie Aufdampfrate, Verdampfergröße und -abstand auf die thermische Präparatbelastung einen stärkeren Einfluß haben, als die Art des aufgedampften Materials.

---

Improved resolution in freeze-etching

Summary Resolution in freeze-etching is primarily limited by the need for shadowing the replica. In addition to giving high resolution, any technique suitable for freeze-etching must allow short shadowing times, a low thermal load for the specimen, and must provide a final film capable of surviving the drastic chemical procedures used for cleaning the replica.Simultaneous evaporation of platinum and carbon is at present the standard shadowing method in freeze etching. Replicas of higher resolution can in principle be obtained by using very high melting metals such as tungsten or tantalum. Extensive specimen damage caused by long evaporation times and excessive thermal load have however prevented the successful application of such ultrashadowing methods to freeze etching.Since neither long shadowing times nor a high thermal load are properties intrinsic to ultrashadowing, a suitable electron beam evaporator for high melting metals was built and thus ultrashadowing successfully applied to freeze-etching. All parts of the gun are water cooled and can be outgased by electron bombardment. The source can thus be operated reproducibly at high rates and without affecting the vacuum. The actual source (3 mm Ø) is the only hot part of the gun not shielded from the specimen. The ions which are generated during evaporation, and can cause considerable specimen damage, are deflected from the specimen by an electric field. The evaporation is done in 7–10 seconds, the source-specimen distance is 200 mm. For the shadowing material we use a tantalum-tungsten alloy. The resultant films are stable in 70% sulfure acid.At lower magnification freeze-etched specimen which are shadowed with Ta/W look just like Pt/C replicas which indicates that no additional artification took place. High magnification micrographs of ultrashadowed objects show a resolution considerably higher than those of platinum-carbon replicas published in the literature.Since heat damage is a crucial problem in freeze-etching the thermal load for tantalum-tungsten-and platinum-carbon-shadowing was calculated. For both methods a theoretical value of approximately 16 mW·cm^–2 was obtained provided the above shadowing conditions are observed and the ions are deflected from the specimen. Furthermore, the calculations show that the load caused by thermal radiation can be reduced drastically, without increasing the shadowing time, if the gun is operated at a higher rate thus allowing the use of either a smaller source or a longer source-specimen distance.Measurements of the thermal load agreed basically with the calculations. The values measured using the tantalum-tungsten gun were about half the ones obtained during platinum-carbon shadowing with a commercial evaporator.

     

Measurement of Globular Protein Molecules by Electron Microscopy

A series of molecular species with approximately spherical shape and with molecular weights between 35,000 and 250,000 were shadowed with platinum while resting on a cleaved mica surface. They were backed, stripped from the surface, and examined by electron microscopy. Materials examined were: pepsin, liver alcohol dehydrogenase, yeast alcohol dehydrogenase, glutamic dehydrogenase, polyhedral virus protein (insect), fibrinogen substructure, alkaline phosphatase, and microsomal particles from Escherichia coli. Measurements were made of widths perpendicular to the shadowing direction and heights were deduced from shadow lengths. For those molecular species with well established molecular weights the average heights correlate very well with the diameter of the theoretical sphere but the average widths are too great by 50 to 80 A due to the lateral growth of the deposited metal. Although the distortion in shape of shadowed particles is relatively large, with standardized conditions for shadowing, it is possible to make allowance for the distortion and to obtain reasonably reliable estimates of the dimensions of spherical organic particles down to a molecular weight of about 35,000.     

Surface topography of microtubule walls decorated with monomeric and dimeric kinesin constructs

The surface topography of opened-up microtubule walls (sheets) decorated with monomeric and dimeric kinesin motor domains was investigated by freeze-drying and unidirectional metal shadowing. Electron microscopy of surface-shadowed specimens produces images with a high signal/noise ratio, which enable a direct observation of surface features below 2 nm detail. Here we investigate the inner and outer surface of microtubules and tubulin sheets with and without decoration by kinesin motor domains. Tubulin sheets are flattened walls of microtubules, keeping lateral protofilament contacts intact. Surface shadowing reveals the following features: (i) when the microtubule outside is exposed the surface relief is dominated by the bound motor domains. Monomeric motor constructs generate a strong 8 nm periodicity, corresponding to the binding of one motor domain per alpha-beta-tubulin heterodimer. This surface periodicity largely disappears when dimeric kinesin motor domains are used for decoration, even though it is still visible in negatively stained or frozen hydrated specimens. This could be explained by disorder in the binding of the second (loosely tethered) kinesin head, and/or disorder in the coiled-coil tail. (ii) Both surfaces of undecorated sheets or microtubules, as well as the inner surface of decorated sheets, reveal a strong 4 nm repeat (due to the periodicity of tubulin monomers) and a weak 8 nm repeat (due to slight differences between alpha- and beta-tubulin). The differences between alpha- and beta-tubulin on the inner surface are stronger than expected from cryo-electron microscopy of unstained microtubules, indicating the existence of tubulin subdomain-specific surface properties that reflect the surface corrugation and hence metal deposition during evaporation. The 16 nm periodicity visible in some negatively stained specimens (caused by the pairing of cooperatively bound kinesin dimers) is not detected by surface shadowing.     

The molecular structure of human erythrocyte spectrin. Biophysical and electron microscopic studies.

Molecules of human erythrocyte spectrin have been examined by electron microscopy after low-angle shadowing. Spectrin heterodimers and tetramers were first purified and characterized by polyacrylamide gel electrophoresis and analytical ultracentrifugation under conditions which minimize proteolysis and aggregation. The heterodimers and tetramere were separated for low-angle shadowing by gel filtration in ammonium acetate buffer at physiological ionic strength, in which they showed sedimentation coefficients of 8.9 S and 12.5 S, respectively, similar to those values reported for heterodimers and tetramers in non-volatile buffers. The ammonium acetate buffer promoted the dissociation of spectrin tetramers into heterodimers under conditions in which tetramers in NaCl or KCl buffers are stable. When visualized by low-angle unidirectional and rotary shadowing, spectrin heterodimers appeared as long flexible molecules with a mean shadowed length of 97 nm. Each heterodimer, composed of the two polypeptide chains, band 1 (240,000 M_r) and band 2 (220,000 M_r), often appeared as two separate strands which lay partially separated from one another or coiled round each other in a loose double helix. The association between these polypeptides appears to be weak, except at both ends of the molecule where there are sites of strong binding. Tetramers are formed by the end-to-end association of two spectrin heterodimer molecules without measurable overlap, and have a mean shadowed length of 194 nm. This association to form tetramers probably involves head-to-head binding of the heterodimers, since the higher oligomers to be expected from a head-to-tail binding mode are not observed. The molecular shape of spectrin is quite distinct from that of myosin, to which it has often been likened.     

AN ultrastructural study of cross-bridge arrangement in the frog thigh muscle thick filament.

We have developed thick filament isolation methods that preserve the relaxed cross-bridge order of frog thick filaments such that the filaments can be analyzed by the convergent techniques of electron microscopy, optical diffraction, and computer image analysis. Images of the filaments shadowed by using either unidirectional shadowing or rotary shadowing show a series of subunits arranged along a series of right-handed near-helical strands that occur every 43 nm axially along the filament arms. Optical filtrations of images of these shadowed filaments show 4-5 subunits per half-turn of the strands, consistent with a three-stranded arrangement of the cross-bridges, thus supporting our earlier results from negative staining and computer-image analysis. The optical diffraction patterns of the shadowed filaments show a departure from the pattern expected for helical symmetry consistent with the presence of cylindrical symmetry and a departure of the cross-bridges from helical symmetry. We also describe a modified negative staining procedure that gives improved delineation of the cross-bridge arrangement. From analysis of micrographs of these negatively stained filament tilted about their long axes, we have computed a preliminary three-dimensional reconstruction of the filament that clearly confirms the three-stranded arrangement of the myosin heads.     

Intrathalline and size-dependent patterns of activity in Lasallia pustulata and their possible consequences for competitive interactions

1. In dense populations of the saxicolous lichen Lasallia pustulata the margins of adjacent thalli overlap each other in intraspecific competition for space and light.
2. In situ non-destructive monitoring of hydration-dependent potential photosynthetic activity by modulated fluorescence systems in different parts of the thallus shows that the activity is structured by a centre-to-margin gradient, with the centre of the thallus remaining active for substantially longer periods than the margins when the thalli dry up after being activated by wetting. The pattern reflects the water status of different parts of the thallus; the margins which are thin and exposed dry up first.
3. The activity pattern within individual lichen thalli suggests that marginal overlapping between neighbours may have a less detrimental effect on the shadowed individuals than expected from a pure consideration of the amount of area shadowed. Because the centre of the lichen thallus is active for longer periods, shadowing of this region may possibly be more harmful per area unit than an overlap at the less active margins.
4. Larger thalli are active for substantially longer periods than small ones. Even the margins of larger thalli tend to be active for a longer period than the centre of small thalli.     

Heavy-meromyosin-decorated actin filaments: a simple method to preserve actin filaments for rotary shadowing

It has become accepted that deep-freeze-drying at or below -90 degrees C is necessary to preserve the structure of supramolecular assemblies such as actin filaments (AFs) for metal shadowing. This has kept the metal shadowing technique from widespread use in the study of proteins complexed with AFs because of the limited availability of the apparatus for deep-freeze-drying. I report here that adsorption to freshly cleaved mica, treatment with buffered uranyl acetate in glycerol solution, rinsing, and removal of liquid eliminate the need of freeze-drying to preserve the structure of AFs. This technique, in combination with metal shadowing, was applied to the study of AFs decorated with heavy meromyosin (HMM). It was observed that (1) when HMM molecules are associated with single AFs in the majority of cases only one head of each HMM molecule makes contact at the point furthest from the neck region; (2) binding of HMM causes bundling of AFs, probably by the two heads of each molecule binding different filaments; and (3) the binding of HMM to the bundled AFs appears to be more stable than that to a single AF. This method of specimen preparation requires no freeze-drying and is therefore easily applicable to other large protein complexes.     

Electron microscopy of subnanometer surface features on metal-decorated protein crystals

Crystals of heavy riboflavin synthase from Bacillus subtilis were freeze-etched and vacuum-coated at normal incidence with 0.1 to 0.4 nm of gold and silver, respectively. This decoration technique was applied to probe the protein surface for preferential nucleation sites. Image processing of the electron micrographs revealed two particular decoration sites for silver and a different one for gold. According to X-ray crystallography, the riboflavin synthase molecules are spherical and smooth except for a surface corrugation of less than 1 nm, which can not be depicted by heavy-metal shadowing. Thus the decoration sites represent sites of specific physical-chemical interactions between the condensing metal and the protein. The decoration pattern correctly reflects the icosahedral symmetry of the almost spherical protein molecules. Owing to the molecule's symmetry, the position of these topochemical sites with respect to the symmetry axes can be localized within 5A. The packing of the molecules in the crystal can be directly observed on shadowed replicas. Only decoration, however, makes it possible to observe the exact orientation of the molecules within the crystal planes and to derive the true lattice constant along the 6-fold screw axis. This proves decoration to be a technique suitable for studying crystal packing and the molecular symmetry of protein complexes at high resolution. The technique can be applied to crystals that are not large enough or insufficiently ordered for X-ray crystallography.     

Shadowing of elongated helical molecules (myosin, tropomyosin, collagen, and DNA) yields regular molecule-dependent heavy metal grain patterns

Myosin and other alpha-helical molecules (tropomyosin, collagen) can now directly be adsorbed on EM support films, washed, air-dried, or frozen and freeze-dried. Using this method, the molecules were rotary or unidirectionally shadowed with different heavy metals (Pt/C, Ta/W, Ag) or with C alone. After shadowing at low elevation angles with Ta/W or Ag, myosin, tropomyosin, collagen, and DNA showed strikingly regular patterns of either single or coalesced heavy metal grains (bands) along their entire lengths. Even after shadowing with C alone, repetitive, granular accumulations or bands of C were found along the molecules. The different heavy metals and C displayed distinctive banding patterns on the molecules examined, all of which are characterized by different surface charge periodicities and pitch values. The patterns were quantified on the basis of the distances between grains or bands. Two most frequently measured distances between bands were found after shadowing with heavy metals. After shadowing with Ag the prevalent distances between grains were about twice as large as those after Ta/W shadowing. By evaporating a thin layer of carbon on the molecules before shadowing with heavy metals or by evaporating C alone (with no heavy metal) at 6 degrees, one of these two most prevalent distances between bands was attenuated or disappeared. It was demonstrated that the remaining most frequently measured distances between grains seemed to be related to relief periodicities, to the pitch of the double-coiled (myosin, tropomyosin) and triple-coiled alpha-helices (collagen) and fractions thereof. The attenuated distances between grains agreed very well with distances of periodic surface charges on the molecules examined. The investigation of the grain or band patterns showed that their characteristics appearance was molecule-dependent and caused both by periodic chemical (repeats of positive and negative surface charges) and periodic structural features (coiling of the helical strands). The examination confirmed the existence of periodic positive and negative surface charges along the myosin rod and suggested a value of about 17.0 nm for the hitherto undetermined pitch of the double-coiled myosin rod.     

Rotary and unidirectional metal shadowing of VAT: localization of the substrate-binding domain

AAA-ATPases have important roles in manifold cellular processes. VAT (valosine-containing protein-like ATPase of Thermoplasma acidophilum), a hexameric archaeal member of this family, has the tripartite domain structure N-D1-D2 that is characteristic of many members of this family. N, the N-terminal domain of 20.5 kDa, has been implicated in substrate binding. We have applied rotary and unidirectional shadowing to VAT and an N-terminally deleted mutant, VAT(Delta N), in order to map the location of this domain. For the analysis of data derived from unidirectionally shadowed samples we used a new approach combining eigenvector analysis with surface relief reconstruction. Averages of rotary shadowed particles as well as relief reconstructions map the N-terminal domains to the periphery of the hexameric complex and reveal their bipartite structure. Thus, this method appears to be well suited to study the conformational changes that occur during the functional cycle of the protein.     

Spotiton: a prototype for an integrated inkjet dispense and vitrification system for cryo-TEM

Over the last three decades, Cryo-TEM has developed into a powerful technique for high-resolution imaging of biological macromolecules in their native vitrified state. However, the method for vitrifying specimens onto EM grids is essentially unchanged - application of ～3 μL sample to a grid, followed by blotting and rapid plunge freezing into liquid ethane. Several trials are often required to obtain suitable thin (few hundred nanometers or less) vitrified layers amenable for cryo-TEM imaging, which results in waste of precious sample and resources. While commercially available instruments provide some level of automation to control the vitrification process in an effort to increase quality and reproducibility, obtaining satisfactory vitrified specimens remains a bottleneck in the Cryo-TEM pipeline. We describe here a completely novel method for EM specimen preparation based on small volume (picoliter to nanoliter) dispensing using inkjet technology. A first prototype system (Spotiton v0.5) demonstrates feasibility of this new approach for specimen vitrification. A piezo-electric inkjet dispenser is integrated with optical real-time cameras (100 Hz frame rate) to analyze picoliter to nanoliter droplet profiles in-flight and spreading dynamics on the grid, and thus provides a method to optimize timing of the process. Using TEM imaging and biochemical assays we demonstrate that the piezo-electric inkjet mechanism does not disrupt the structural or functional integrity of macromolecules. These preliminary studies provide insight into the factors and components that will need further development to enable a robust and repeatable technique for specimen vitrification using this novel approach.     

The Leeuwenhoek lecture 2006. Microscopy goes cold: frozen viruses reveal their structural secrets

The electron microscope provides a powerful tool for investigating the structure of biological complexes such as viruses. A modern instrument is fully capable of atomic resolution on suitable non-biological specimens, but biological materials are difficult to preserve, owing to their fragility, and to image, owing to their radiation, sensitivity. The act of imaging the specimen severely damages it. Originally, samples were prepared by staining with a heavy metal salt, which provides a stable specimen but limits the amount of details that can be retrieved. Now particulate specimens, such as viruses, are prepared by rapid freezing of unstained material and observed in a frozen state with low doses of electrons. The resulting images require extensive computer processing to extract fully detailed three-dimensional information about the specimen. The whole process is referred to as single-particle electron cryomicroscopy. Using this approach, the structure of the human hepatitis B virus core was solved at the level of the protein fold. By comparing maps of RNA- and DNA-containing cores, it was possible to propose a model for the maturation and control of the envelopment of the virus during assembly. These examples show that cryomicroscopy offers great potential for understanding the structure and function of complex biological assemblies.     

A simple method for freeze-drying of macromolecules and macromolecular complexes

We present a simple approach for effective freeze-drying and rotary shadowing of large molecules, molecular assemblies, and cell organelles. Simply, a suspension of specimen is adsorped to a glass coverslip, stabilized, and rinsed with 30% methanol. A second coverslip is "sandwiched" on top, and excess methanol is withdrawn from the edges then frozen by plunging into liquid nitrogen and split. Following either rotary or unidirectional shadowing and replication, the coverslip is dissolved in hydrofluoric acid. In addition to avoiding the problems encountered with air-drying specimens for rotary shadowing, the technique also reproducibly provides the thin layer of solution necessary for proper freeze-drying, regardless of how hydrophobic the sample is. The "glass sandwich" technique allows modification of the glass substrate (making it hydrophobic with carbon or hydrophilic by soaking it in alcian blue) which clearly alters the shape of macromolecular assemblies such as myosin filaments and decorated thin filaments.     

Rotary-Shadowed Freeze-Fracture Replicas: A Simple Method for the Analysis of Fracture Faces

In rotary-shadowed freeze-fracture replicas, intramembrane particles on the periphery of a membrane fracture face are not uniformly shadowed from all sides. Those eccentrically positioned intramembrane particles with a net centripetally directed shadowing are on a convex fracture face. In contrast, those eccentrically positioned intramembrane particles with a net centrifugally directed shadowing are on a concave fracture face. Centrally positioned intramembrane particles on convex faces are uniformly shadowed from all sides; however, central depressions of concave faces are often unshadowed.     

Rotary-shadowed freeze-fracture replicas: a simple method for the analysis of fracture faces

In rotary-shadowed freeze-fracture replicas, intramembrane particles on the periphery of a membrane fracture face are not uniformly shadowed from all sides. Those eccentrically positioned intramembrane particles with a net centripetally directed shadowing are on a convex fracture face. In contrast, those eccentrically positioned intramembrane particles with a net centrifugally directed shadowing are on a concave fracture face. Centrally positioned intramembrane particles on convex faces are uniformly shadowed from all sides; however, central depressions of concave faces are often unshadowed.     

Electron microscopy of cholesterol

Negative staining and platinum-carbon shadowing have been used to prepare electron microscope specimens from aqueous colloidal suspensions of cholesterol microscrystals and from crystalline suspensions in methanol and ethanol. Microcrystals prepared by injection of alcoholic solutions of cholesterol into water exhibit angular conformations of varying regularity which contain a number of parallel cholesterol bilayers. The electron optical images of the cholesterol microcrystals, oriented horizontally and ‘on-edge’, obtained by both negative staining and metal shadowing, are in good agreement. Metal shadowing does, however, reveal greater detail within microcrystal clusters than does negative staining, as well as of the bilayer steps at microcrystal edges. The needle-like crystals (from methanol) and plate-like crystals (from ethanol) present considerable difficulties for the negative staining technique, because of their thickness and the consequent depth of the surrounding negative stain. Small crystals are, nevertheless, shown to possess multiple cholesterol bilayers. Platinum-carbon shadowing of cholesterol crystals taken directly from methanol and ethanol provides more satisfactory images than negative staining. The large depth of focus of the transmission electron microscope enables the stacked cholesterol bilayers to be clearly defined at the edges of crystals. The results obtained are discussed in relation to the physicochemical and biological properties of cholesterol, which underlie the fundamental difficulty encountered when fixing and staining cholesterol for thin sectioning, and also the role of cholesterol insolubility in the formation of gallstones.     

Fundamental Technical Elements of Freeze-fracture/Freeze-etch in Biological Electron Microscopy

Freeze-fracture/freeze-etch describes a process whereby specimens, typically biological or nanomaterial in nature, are frozen, fractured, and replicated to generate a carbon/platinum “cast” intended for examination by transmission electron microscopy. Specimens are subjected to ultrarapid freezing rates, often in the presence of cryoprotective agents to limit ice crystal formation, with subsequent fracturing of the specimen at liquid nitrogen cooled temperatures under high vacuum. The resultant fractured surface is replicated and stabilized by evaporation of carbon and platinum from an angle that confers surface three-dimensional detail to the cast. This technique has proved particularly enlightening for the investigation of cell membranes and their specializations and has contributed considerably to the understanding of cellular form to related cell function. In this report, we survey the instrument requirements and technical protocol for performing freeze-fracture, the associated nomenclature and characteristics of fracture planes, variations on the conventional procedure, and criteria for interpretation of freeze-fracture images. This technique has been widely used for ultrastructural investigation in many areas of cell biology and holds promise as an emerging imaging technique for molecular, nanotechnology, and materials science studies.