Samples for electron microscopy do not usually undergo routine formalin fixation and paraffin embedding, but formalin-fixed paraffin-embedded (FFPE) samples can be processed for ultrastructural examination and immuno-electron microscopy (Table 22.1), allowing the use of stored material for the characterization of amyloid proteins. Comparison of the paraffin block with a Congo red-stained slide can help to select the most relevant areas for ultrastructural study. Small portions of tissue are then extracted from the paraffin block, deparaffinized in xylene, rehydrated in a graded series of ethyl alcohols, washed in cacodylate buffer, post-fixed in osmium tetroxide, dehydrated, and embedded in epoxy resin as usual (see below).

The processing of fresh biopsy samples for electron microscopy is summarized in Table 22.2. A small sample size is important in order to obtain good fixation of the tissue. Thus, specimens greater than 2–3 mm³ show poor morphology due to poor penetration of the fixing solution. The specimens are fixed by immersion in a modified Karnovsky solution [14] (see below), then washed in cacodylate buffer, dehydrated through a graded series of ethyl alcohols, and embedded in epoxy resin (Table 22.2). After polymerization at 60 ° C, ultrathin sections 600–800 Å (Angstrom) thick are cut with an ultramicrotome, stained with uranyl acetate and lead citrate (Reynolds’ solution [15]), and observed with an electron microscope.

Immuno-electron microscopy can be applied to virtually every tissue, either specifically fixed for ultrastructural examination or previously formalin fixed and paraffin embedded. The only limitations include good tissue fixation and the availability of antibodies to the amyloid protein under test. Many protocols have been developed since the introduction of colloidal gold to ultrastructural immunohistochemistry. Immuno-electron microscopy, using antibody probes conjugated with gold particles, permits high-resolution detection and localization of antigens either within the cells, on their surface, or in the interstitium. The two techniques most widely used in transmission electron microscopy consist of either immunolabeling before the specimens are embedded in resin (preembedding immunogold labeling) or immunolabeling after embedding in resin (postembedding immunogold labeling). Nonetheless, immunogold histochemistry has some general limitations, such as antigen preservation and antibody specificity. Successful detection and localization depends on the antigen-recognition specificity of the primary antibodies for the investigated antigen, on the preservation of the antigenicity of the target proteins, and on the ability of the antibodies to bind to the antigens. Antigenicity is frequently lost during the dehydration and embedding procedures. Suppression of immunoreaction in resin-embedded tissues may be due to intra- and inter-cross-links within aldehyde-treated proteins and the destruction of secondary and tertiary protein structures during dehydration [16]. Specimens embedded in acrylic resins without osmium tetroxide fixation show poor preservation of membranes and low contrast of tissue components, thus making it difficult to correlate the immunostains with tissue details.

For all the above reasons, fixation is one of the most important aspects of sample preparation [17] for immuno-electron microscopy because it affects the strength of the immunoreaction and the preservation of fine cellular structures. In general, fixatives that provide good morphology cross-link macromolecules rapidly and tightly, forming a gel-like structure in the tissues, and, thus, directly modify epitopes and severely inhibit immunoreactions. Due to these undesirable effects, many kinds of fixatives, fixation conditions, and procedures have been employed to label each antigen. Although glutaraldehyde is an excellent fixative for the preservation of morphology, it severely inactivates the immunoreactivity of many antigens and, hence, a better fixative for immuno-electron microscopy is a mixture of paraformaldehyde and a low concentration of glutaraldehyde (modified Karnovsky’s solution [7]).

In our Pathology Unit of IRCCS Policlinico San Matteo/University of Pavia, electron microscopy is routinely used to confirm the diagnosis of amyloidosis. When ultrastructural examination demonstrates amyloid fibrils, immuno-electron microscopy is subsequently performed to identify the amyloidogenic protein. We use a postembedding method, i.e., immunostaining is performed on ultrathin sections on nickel grids, as summarized in Table 22.3.

Usually, for diagnostic purposes, a panel of four primary antibodies is sufficient. This panel covers the commonest forms of amyloidosis: AL (kappa and lambda), AA, and ATTR; in selected cases we also perform immunogold labeling with antibodies directed against apolipoprotein AI and beta-2-microglobulin. Fibrinogen and lysozyme have also been investigated in rare instances with negative results. We have never so far encountered rarer forms of kidney amyloidosis such as ALect2. Enzymatic predigestion with trypsin can be used in some cases to unmask antigenic epitopes modified by fixation-induced protein cross-links, as is commonly performed in light immunohistochemistry. The optimal conditions (antibody source and dilutions, type and duration of pretreatment) must be determined in each laboratory. Table 22.4 gives an example of the conditions used in our laboratory.