Resolution as a function of accelerating voltage in electron microscopy of semithick biological specimens

In the past, biological sections ranging in thickness from 0.10- to 0.50-micron have usually been examined with high-voltage (greater than 500 kV) electron microscopes (HVEM). Now investigators are increasingly using intermediate voltage (200-500 kV) electron microscopes (IVEM), which are more readily available and demand less maintenance. In a study of "typical" plastic-embedded, stained sections of mouse liver ranging from 0.10 to 1.0 micron thick, we determined the resolution obtainable at 100, 200, and 1000 kV. At all three accelerating voltages the resolution (2.7 nm) for 0.10-micron sections was limited only by the sections stain granularity. For 0.25-micron thickness the resolutions were 5.8, 3.1, and 3.1 nm at 100, 200, and 1000 kV, respectively. The maximum usable thickness at 200 kV with resolution sufficient to resolve membranes clearly was between 0.75 and 1.0 micron, depending on the magnification. Resolution at 100 kV was adequate for screening sections up to 1.0-micron thick for preparation defects prior to examination with an IVEM or HVEM.     

Use of colloidal gold and ruthenium red in stereo high-voltage electron microscopic study of Con A-binding sites in mouse macrophages

Concanavalin A (Con A)-binding sites were labeled with colloidal gold (CG), stained with ruthenium red, and observed under a high-voltage electron microscope. Mouse peritoneal macrophages were labeled by the indirect Con A/CG labeling method at 0 degree C. After washing, some of the cells were incubated in phosphate-buffered saline (PBS) at 37 degrees C. The specimens were then stained with ruthenium red, to enhance the contrast of the cell surface, and embedded in Epon. Sections (0.3 approximately 3 micron thick) were cut and examined by high-voltage electron microscopy at accelerating voltages of 200 approximately 1,000 kV. Staining with ruthenium red provided a strong contrast of the cell surface and the invaginating tubules beneath it against the cytoplasm; in thick sections, both of them were clearly seen by stereomicroscopy. CG particles which represented Con A-binding sites were also sufficiently electron dense to be recognized by high-voltage electron microscopy of thick sections. The two- and three-dimensional distribution of CG particles on the ruthenium-red-positive cell surface was clearly visualized. At 0 degree C, Con A-binding sites were randomly distributed on the cell surface. The redistribution and endocytosis of Con A-binding sites were seen at 37 degrees C. The three-dimensional organization of membrane invagination, which represented the process of endocytosis, was clearly seen by stereomicroscopy. The combination of CG labeling and ruthenium red staining is a useful method for high-voltage electron microscopic analysis of the two- and three-dimensional distribution of CG-labeled ligands on the cell surface in thick sections.     

A Simple Technique for Osmicating and Flat Embedding Large Tissue Sections for Light and Electron Microscopy

Many investigators now use thin hand-sliced, tissue chopper, or Vibratome sections of fresh tissue in various procedures. In our experience brain and nerve sections varying in thickness from less than 40 to more than 300 μm, with or without prior embedding in agar, have a tendency to roll up or curl during aldehyde fixation and buffer washes. Once osmicated, such curled sections cannot be flattened. When the entire cut face of such thin slices is to be studied, sufficiently flat embedding so that some regions are not completely sectioned before others are even sampled is critical. This report describes fixation and flat embedding procedures, developed for light and electron microscopic autoradiographic studies of plastic embedded brain slices about 200 μm thick (Schwartz 1981), which can be applied to any comparable thin slice of nervous tissue (or potentially of many other tissues) to achieve maximally flat tissue faces. Since osmicated tissue slices are usually too thick to be transilluminated for direct examination with the light microscope, the methods described simplify preparation of the semithin sections required for this purpose.     

Use of colloidal gold and ruthenium red in stereo high-voltage electron microscopic study of Con A-binding sites in mouse macrophages

Summary Concanavalin A (Con A)-binding sites were labeled with colloidal gold (CG), stained with ruthenium red, and observed under a high-voltage electron microscope. Mouse peritoneal macrophages were labeled by the indirect Con A/CG labeling method at 0° C. After washing, some of the cells were incubated in phosphate-buffered saline (PBS) at 37° C. The specimens were then stained with ruthenium red, to enhance the contrast of the cell surface, and embedded in Epon. Sections (0.33 m thick) were cut and examined by high-voltage electron microscopy at accelerating voltages of 2001,000 kV. Staining with ruthenium red provided a strong contrast of the cell surface and the invaginating tubules beneath it against the cytoplasm; in thick sections, both of them were clearly seen by stereomicroscopy. CG particles which represented Con A-binding sites were also sufficiently electron dense to be recognized by high-voltage electron microscopy of thick sections. The two- and three-dimensional distribution of CG particles on the ruthenium-red-positive cell surface was clearly visualized. At 0° C, Con A-binding sites were randomly distributed on the cell surface. The redistribution and endocytosis of Con A-binding sites were seen at 37° C. The three-dimensional organization of membrane invagination, which represented the process of endocytosis, was clearly seen by stercomicroscopy. The combination of CG labeling and ruthenium red staining is a useful method for high-voltage electron microscopic analysis of the two- and three-dimensional distribution of CG-labeled ligands on the cell surface in thick sections.     

Cytological Staining Procedures Applicable to Methacrylate-Embedded Tissues

Experiments indicate that osmic-fixed, plastic-embedded sections are suitable for examination in the light microscope. Nuclei, mitochondia, cellular membranes and cytoplasmic granules are readily demonstrable by phase microscopy. Connective tissue stains permit the identification of elastic and collagenous fibers. Glycogen and other carbohydrate-containing structures are demonstrable by the periodic acid-Schiff and the ammoniacal silver nitrate procedures. It is, therefore, possible to cross-check individual structures by comparing alternate thick and thin sections, examined in the light microscope and electron microscope respectively. Several other advantages pertain to plastic embedded tissues. The sections compare favorably in translucency and in their lack of distortion with material embedded in celloidin, yet the procedure is simpler and much more rapid. Sections of any desired thinness can be prepared, and alternate thick and thin sections are easily forthcoming. When examined in the phase-contrast microscope, mitochondrial preparations become routinely available without the uncertainties of most of the mitochondrial staining methods. It appears, therefore, that plastic embedding should find a useful place among the methods for light microscopy as well as in the armamentarium of the electron microscopist.     

The myosin filament: V. Intermediate voltage electron microscopy and optical diffraction studies of the substructure

Using a 200 kV electron microscope (JEM 200 A), thick (up to 0.4 μm) crosssections of the myosin filaments of vertebrate striated muscle were studied. It was found that: (a) with increasing section thickness the cross-sectional profiles of the shaft of the filament were increasingly more triangular and in sections 0.4 μm thick each apex of the triangle was clearly blunted. This unique cross-sectional profile is predicted by the model proposed by Pepe (1966,1967) in which 12 parallel structural units are packed to form a triangular profile with a structural unit missing at each apex of the triangle. (b) With increasing section thickness the substructure of the myosin filament was enhanced, with the best substructure visible in sections 0.2 μm to 0.3 μm thick. This strongly supports parallel alignment of structural units in the shaft of the filament as proposed by Pepe (1966,1967). (c) The substructure spacing, determined by optical diffraction from electron micrographs of cross-sections of individual myosin filaments or groups of filaments is about 4 nm. (d) The different optical diffraction patterns observed from individual myosin filaments can be explained if the projection of each structural unit in the plane of the section has an elongated profile. With a substructure spacing of 4 nm an elongated cross-sectional profile could be produced by having two myosin molecules per structural unit. Models drawn with two myosin molecules per structural unit in the model proposed by Pepe (1966,1967) gave optical diffraction patterns similar to those observed from individual filaments. (e) The different optical diffraction patterns observed from individual myosin filaments can be explained if the elongated profiles in each structural unit are similarly oriented but with the orientation changing along the length of the filament. The change in orientation per unit length of the filament must be small enough to maintain an elongated profile for the projection of the structural unit in the plane of the sections 0.3 μm thick. All of these observations and conclusions strongly support the model for the myosin filament proposed by Pepe (1966,1967).     

Ultrastructure of frog muscle fiber thick filaments at rest and during potassium contracture]

Isolated slow and intermediate frog muscle fibres were fixed in the rest state and under potassium contracture (50-100 mM KC1). The longitudinal and cross sections of two types of fibres were investigated. It was shown that at the rest the thick filaments of different fibres had similar length (1.6-1.65 mum), diameter (160-165 A) and the amount of subunits (12-13). Under potassium contracture the length of the thick filaments of both fibre types was shortened by 25-30% of the rest-length, the diameter of the slow fibres increased to 180-185 A, the diameter of the intermediate fibres to 200-220 A. The amount of subunits increased to 14-15 in slow fibres and to 17-18 in intermediate fibres. We believe that the ultrastructural changes observed in the thick filaments are a result of molecular transformation in these filaments, which seems to be important for maintaining the contracture.     

Invertebrate myosin filament: Parallel subfilament arrangement in the wall of solid filaments from the honeybee, Apis mellifica

Transverse serial sections (100-140 nm thick) of solid myosin filaments of the honeybee, Apis mellifica, were photographed in a JEM-200 electron microscope at 200 kV. The images were digitized and computer processed by rotational filtering. 87% of the myosin filaments showed 6-fold symmetry in their power spectra, confirming the results of earlier works (Beinbrech et al., 1988, 1991). To determine if the subfilaments were arranged parallel to the filament backbone, two methods were used. First, the three images of each myosin filament in the three serial sections were superimposed. 85% of the resulting images showed a strong peak for 6-fold symmetry and the averaged images showed 6 pairs of subfilaments, which gives evidence for parallel arrangement of the subfilaments relative to the filament axis. This result was confirmed by the second method in which a 3-dimensional reconstruction was made. An average image was made from the images of the same 17 myosin filaments from each of the three sections. The data for the 3-dimensional reconstruction were collected by tracing the outlines of the structures in the three successive sections. The resulting stereo image shows a parallel arrangement of the subfilaments.     

The structure of the scales of Latimeria chalumnae

Scales from four specimens of Latimeria chalumnae were examined in a dissecting microscope and then X-rayed. Some were demineralized and prepared for routine histology. Others were cleared in cedarwood oil. Ground sections of plastic embedded scales were micro-radiographed and electronmicrographs made of araldite embedded frozen scales. Corresponding 1 μm thick serial sections were examined in the light microscope.
The greater part of the scale is composed of layers of unmineralized isopedine surmounted by the exposed portion of the scale which is pigmented and ornamented by a series of denticles of tubular dentine tipped with enameloid. Between these two parts is a thin ridged bone-like layer. In the electron micrographs the isopedine was seen to consist of layers of densely packed collagen fibres; the orientation of which was uniform in each layer but varied markedly from layer to layer. Only few cells were found between the fibre bundles.
The X-rays revealed numerous concentric annulae and, lying approximately at right angles to these, a further series of ridges radiating from the centre of the scales.
It is suggested that the basal unmineralized isopedine and the ridged layer of bone-like tissue covering it represents a highly modified cosmoid scale on which the denticles and pigmented layer have become superimposed.     

Quantitative comparison of zero-loss and conventional electron diffraction from two-dimensional and thin three-dimensional protein crystals

The scattering cross-section of atoms in biological macromolecules for both elastically and inelastically scattered electrons is approximately 100,000 times larger than that for x-ray. Therefore, much smaller (<1 microm) and thinner (<0.01 microm) protein crystals than those used for x-ray crystallography can be used to analyze the molecular structures by electron crystallography. But, inelastic scattering is a serious problem. We examined electron diffraction data from thin three-dimensional (3-D) crystals (600-750 A thick) and two-dimensional (2-D) crystals (approximately 60 A thick), both at 93 K, with an energy filtering electron microscope operated at an accelerating voltage of 200 kV. Removal of inelastically scattered electrons significantly improved intensity data statistics and R(Friedel) factor in every resolution range up to 3-A resolution. The effect of energy filtering was more prominent for thicker crystals but was significant even for thin crystals. These filtered data sets showed better intensity statistics even in comparison with data sets collected at 4 K and an accelerating voltage of 300 kV without energy filtering. Thus, the energy filter will be an effective and important tool in the structure analysis of thin 3-D and 2-D crystals, particularly when data are collected at high tilt angle.     

Membrane turnover in crab photoreceptors studied by high-resolution scanning electron microscopy and by a new technique of thick-section transmission electron microscopy

Summary Crab photoreceptors were examined after treatment by the osmium-DMSO-osmium method for high-resolution scanning electron microscopy. This technique of specimen preparation was also adapted for transmission electron microscopy, enabling sections up to 1 urn thick to be viewed in a conventional microscope at 75 kV. With appropriate pretreatment, some cytoskeletal elements can be visualised by both techniques. The methods were then used to investigate some of the daily changes known to occur in photoreceptor cell structure. Striking differences were found in the structure of Golgi bodies present in retinula cells during the synthesis and breakdown phases of the daily cycle of photoreceptor membrane turnover. Cyclic changes were also noticed in the mitochondria of retinula cells, and additional evidence was found for a previously proposed model of rhabdomeral microvillus formation.     

Superficial Warming of Epoxy Blocks for Cutting of 25-150 μm Sections to be Resectioned in the 40-90 nm Range

An improved method is described in which tissue areas can be initially identified in thick sections by light microscopy and isolated for subsequent ultrathin sections and observation by electron microscopy. This is achieved by embedding in hard Epon which can be sectioned at 25-150 μm on a sliding microtome after softening the blockface by applying a 60-70 C tacking iron to its surface immediately before each section is taken. The thick sections are then mounted on plastic slides to enable light microscopic selection of areas to be observed by electron microscopy. The selected areas are remounted on faced Epon blanks and resectioned at less than 50 nm. This technique makes it possible to obtain thick sections while maintaining an Epon hard enough for good serial ultrathin sections.     

A simple method using alizarin red S for the detection of calcium in epoxy resin embedded tissue

This report presents a simple procedure for staining 1-2 microns epoxy plastic sections of cells and mineralizing matrix present in fetal bovine bone tissue cultures. A 0.3% aqueous toluidine blue O solution was used as a cellular stain and was followed with 2% alizarin red S for the detection of calcium at sites of mineralization. Effects of concentration and pH of alizarin red S on the penetration of epon embedded thick sections were investigated. Optimal staining was achieved with a 2% aqueous alizarin red S solution adjusted to a pH of 5.5-6.5. This staining procedure provides unusually clear contrast between mineral and bone cells in plastic sections for light microscopy.     

High-order structure of metaphase chromosomes: evidence for a multiple coiling model

When chromosome preparations made by the conventional air-drying method were processed with the OsO₄/TCH technique and examined by scanning electron microscopy (SEM), spiral structures in chromatids, which have been frequently observed to be present by light microscopy, were found to be composed of 30 nm fibres. In some portions these fibres appeared to be arranged in coils to form thicker fibres. When chromosome preparations were processed for SEM without air drying, chromosomes appeared to consist of fairly homogeneous thick fibrous structures measuring about 200 nm in diameter. In relatively condensed chromosomes, these 200 nm fibres appeared to be arranged perpendicular to the long axis of the chromatid. These findings suggest that chromatid spiral structures represent a regularly loosened state of the compactly spiralized 200 nm fibres which in turn consist of spiralized 30 nm fibres.     

Preservation of fine structure in vibratome-cut sections of the central nervous system stained for light microscopy

A technique without negative effects on tissue preservation that allows precise identification and subsequent removal of central nervous system nuclei for ultrastructural analysis is described. The procedure uses 200 microns thick Vibratome-cut sections of glutaraldehyde fixed brains. These sections are stained for 25 seconds with a methylene blue solution and stored for 4 hours in 0.2 M pH 7.4 phosphate buffer in 4% sucrose for optimal visualization at the light microscopic level. The stock solution of 1 g methylene blue and 1 g sodium borate in 100 ml of distilled water, is filtered through a Millipore filter and diluted 5:95 with distilled water immediately prior to use. Regions of specific interest are then processed for electron microscopy.     

Progress in applying the high-voltage electron microscope to biomedical research

This review attempts a physical definition of the technical problems and achievements in applying the high-voltage electron microscope (HVEM) to biological and medical research. It is hoped that the review will summarize for biologists, funding agencies, and institutions the achievements of the HVEM, its future prospects, and the main problem areas that still need to be explored. At present it is not known whether future HVEMs will favor the fixed beam or the scanning transmission electron microscopy (STEM) mode. The STEM mode offers reduced radiation damage as a result of more efficient electron detection and ease of manipulation of the collected signals by separating the elastic and inelastic signals. Energy filtration to remove the inelastic signal provides a means to enhance the contrast and improve the resolution for thick specimens. Several prototype STEM-mode HVEMs are now under development and it is expected that, in a few years, comparisons of fixed beam and STEM modes will be possible. The review discusses several HVEM instrument features that remain poorly developed. In the area of image recording a photographic emulsion has been designed to give optimized performance at an acceleration voltage of 1 MV. However, this remains unavailable commercially. Conversion of the HVEM electron image to a usable light image by phosphors etc., involves some difficulties, making it difficult to obtain good performance from TV systems. Since the HVEM is particularly useful for three-dimensional imaging, the further development of improved goniometers for stereo viewing and image reconstruction is important. The large volume available in the objective specimen volume and the increased penetration at high acceleration voltages make the HVEM particularly suitable for the application of environmental chambers in the microscopy and electron diffraction of thick wet specimens. An improved signal-to-noise ratio improves the prospects for elemental analysis at high acceleration voltages. When carefully carried out, improved resolution can be obtained in dark-field over that obtainable at 100 kV. Dark-field provides the easiest way to obtain high contrast on weakly stained or unstained objects. Its further improvement requires the use of specially thick and shaped beam stops and apertures that are not penetrated by the 1 MV beam. Recent HVEM studies of whole cells and microorganisms are reviewed. These studies already show that the former thin-section approach led to some incorrect ideas about the shape of some organelles and their three-dimensional relationships. This new information is proving important in helping to establish the function of fibrillar and membranous components of the cell. The most important limitation in examining thick sections is the large depth of field that causes excessive overlap of in-focus structures in stereo views of thick sections. In a few cases special specific heavy metal stains have been developed to overcome this problem, but an optical solution would be more generally applicable. Attempts are now being made to unscramble overlapped detail by applying the image reconstruction techniques of tomography and holography. It is concluded that even with existing techniques, the HVEM examination of thick sections provides a very useful improvement in sampling statistics and in three-dimensional imaging of cell structures over that obtainable by examining thin sections at a lower acceleration voltage (100 kV). Randomized author sequence.     

The invertebrate myosin filament: subfilament arrangement of the solid filaments of insect flight muscles.

Transverse sections (approximately 140 nm thick) of solid myosin filaments of the flight muscles of the fleshfly, Phormia terrae-novae, the honey bee, Apis mellifica, and the waterbug, Lethocerus uhleri, were photographed in a JEM model 200A electron microscope at 200 kV. The images were digitized and computer processed by rotational filtering. In each of these filaments it was found that the symmetry of the core and the wall was not the same. The power spectra of the images showed sixfold symmetry for the wall and threefold symmetry for the core of the filaments. The images of the filaments in each muscle were superimposed according to the sixfold center of the wall. These averaged images for all three muscles showed six pairs of subunits in the wall similar to those found in the wall of tubular filaments. From serial sections of the fleshfly filaments, we conclude that the subunits in the wall of the filaments represent subfilaments essentially parallel to the long axis of the filament. In each muscle there are additional subunits in the core, closely related to the subunits in the wall. Evaluation of serial sections through fleshfly filaments suggests that the relationship of the three subunits observed in the core to those in the wall varies along the length of the filaments. In waterbug filaments there are three dense and three less dense subunits for a total of six all closely related to the wall. Bee filaments have three subunits related to the wall and three subunits located eccentrically in the core of the filaments. The presence of core subunits can be related to the paramyosin content of the filaments.     

Measuring the resolution of uncompressed plastic sections cut using an oscillating knife ultramicrotome

Thin sections of biological tissue embedded in plastic and cut with an ultramicrotome do not generally display useful details smaller than approximately 50 A in the electron microscope. However, there is evidence that before sectioning the embedded tissue can be substantially better preserved, which suggests that cutting is when major damage and loss of resolution occurs. We show here a striking example of such damage in embedded insect flight muscle fibres. X-ray diffraction of the embedded muscle gave patterns extending to 13A, whereas sections cut from the same block showed only approximately 50 A resolution. A possible source of this damage is the substantial compression that was imposed on sections during cutting. An oscillating knife ultramicrotome eliminates the compression and it seemed possible that sections cut with such a knife would show substantially improved preservation. We used the oscillating knife to cut sections from the embedded muscle and from embedded catalase crystals. Preservation with and without oscillation was assessed in Fourier transforms of micrographs. Sections cut with the knife oscillating did not show improved preservation over those cut without. Thus compression during cutting does not appear to be the major source of damage in plastic sections, and leaves unexplained the 50 A versus 13A discrepancy between block and section preservation. The results nevertheless suggest that improvements in ultramicrotomy will be important for bringing thin-sectioning and tomography of plastic-embedded cells and tissues to the point where macromolecule shapes can be resolved.     

Comparison of red and white muscles by cytophotometry of their muscle fibre populations

Synopsis Samples from two red muscles (vastus intermedius and vastus medialis) and two white muscles (biceps femoris and gluteus medius) were taken from four pigs. Serial transverse sections were reacted for ATPase and NADH oxidative activity. Sections were mapped with a projection microscope so that the staining intensity of individual fibres for the two reactions could be measured with a simple microscope photometer. Transmission values at 600 nm were converted to units of 0–10 for the range from darkest to lightest staining fibres on each section to cancel variation in staining intensity between sections. The aim of the study was to use simple cytophotometry instead of subjective judgement in the categorization of different histochemical types of muscle fibres. Cytophotometry enabled clear resolution of the major fibre types (types I and II using the ATPase reaction), partial resolution of more variable characteristics (NADH oxidative activity in type I and II fibres) and no resolution of subtle subtypes (IA and IB with the NADH oxidative reaction). However, between the major fibre types, cytophotometry revealed variable numbers of fibres with transitional characteristics. There were more of these fibres in red muscles. With sections reacted for ATPase, transmission values for low magnification fields containing 100 to 200 fibres were correlated (r=–0.91) with the ratio of type I:II fibres.     

A CORRELATED HISTOCHEMICAL AND ELECTRON MICROSCOPIC STUDY OF THE INTRANUCLEAR CRYSTALLINE AGGREGATES OF ADENOVIRUS (RI-APC VIRUS) IN HELA CELLS

HeLa cells in tissue cultures infected with types 3, 4, or 7 of adenovirus (RI-APC virus) were studied in order to correlate certain histochemical and electron microscopic findings. Adjacent thin (ca. 0.05 µ) and thick (2–4 µ) sections of osmium-fixed, methacrylate-embedded cells were cut; by mapping the sections the same cells could be identified with both the electron and the light microscope. Intranuclear crystalline aggregates seen with the electron microscope to be composed of ordered arrays of viral particles were found by means of the Feulgen reaction to contain DNA. DNA is therefore assumed to be a constituent of the viral particle. The virus appeared to develop from an osmiophilic Feulgen-negative matrix. Displacement of nuclear chromatin occurred during this process. A Feulgen-azure staining method was found to permit clear distinction between viral and nuclear (host) DNA in thick sections.