Transmission electron microscopy (TEM) has long been the best available method for the diagnosis of epidermolysis bullosa. Today, TEM is largely superseded by immunofluorescence microscopy mapping, which is generally more available. This article discusses its continuing role in confirming or refining results obtained by other methods, or in establishing the diagnosis where other techniques have been unsuitable or have failed. It covers key steps for optimizing tissue preparation, features of analysis, recently classified epidermolysis bullosa disorders, and strengths and weaknesses of TEM.
Since the early application of transmission electron microscopy (TEM) to the study of epidermolysis bullosa (EB), TEM has for many years retained the premier position among diagnostic techniques for establishing or confirming the diagnosis of the major types of EB. Even today, TEM remains the most powerful method for determining the levels of skin separation which underlie the current classification of EB and certain related disorders. Other laboratory techniques, including immunofluorescence (IF) microscopy and DNA analysis, have become increasingly important in EB diagnosis. However, no single method has yet fully superseded TEM, although it is now used mainly in a reference laboratory and is less applicable as a primary diagnostic procedure.
This article aims to provide a summary of the main TEM findings in the major types of EB and to describe briefly the ultrastructural features of some of the disorders that have been recently added to the EB classification. Finally, it includes a short discussion on the current advantages and disadvantages of using TEM for EB diagnosis.
Key steps for optimizing tissue preparation for TEM
Compared with other laboratory applications such as routine diagnostic histopathology, TEM is very costly and time-consuming. The specialized equipment is expensive to purchase and maintain, and the laboratory staff is required to spend several hours of highly skilled technical work in the preparation of each specimen for analysis. Adherence to a few key guidelines helps avoid wasting time, effort, and expense in obtaining and processing skin biopsies for TEM diagnosis. All skin samples are precious and often it is possible to have only one attempt at procuring a suitable diagnostic biopsy from an EB patient, especially in the case of a neonate or young infant. It is usually preferable not to contemplate taking the biopsy unless all the appropriate steps are known to be available and achievable beforehand.
As soon as the biopsy is planned, it is essential to ensure that the laboratory is notified and the necessary reagents, mainly the TEM fixative, are available. An effective primary fixative for TEM contains 2.5% glutaraldehyde and 2% formaldehyde in 0.1M phosphate buffer at pH 7.4. This fixative can be stored conveniently in a −20°C freezer as premeasured aliquots of double-strength buffer and double-strength aldehyde solutions, ready for thawing and mixing before use. The reactivity of the aldehydes is preserved better if kept separate from the buffer. Upon notification of an imminent biopsy, the laboratory can supply the clinician with this prepared fixative, with a disposable plastic minipipette for mixing the solutions. The fixative is adequately preserved during overnight transport at ambient temperature.
If a sample for IF antigen mapping is to be taken at the same time, the biopsy can usually be subdivided with minimal trauma by placing it, epidermis down, on the inverted lid of a 5cm plastic Petri dish and cutting it using a large (no. 22) scalpel blade with a rocking motion, thus minimizing shearing forces. For preservation of antigenicity, it is essential to avoid contamination of the IF sample with TEM fixative. Alternatively, a separate sliver of skin can be taken from the patient for IF studies. Once the TEM sample has been immersed in fixative, it can be dispatched to the laboratory at ambient temperature, suitably packaged to comply with local regulations for the transport of biologic material. If not dispatched on the day of the biopsy, the sample should be refrigerated (2–8°C).
Biopsies from established blisters or lesions more than an hour old will usually produce conflicting results arising from tissue necrosis and repair. Therefore, the authors recommend that the clinician induce skin separation (a clinically obvious blister is not necessarily best) by gently rubbing the skin in the proposed biopsy site. Ideally, the biopsy should contain the edge of a new lesion so that split and unseparated skin can be examined in the microscope. In extremely fragile skin, as encountered, for example, in Herlitz junctional EB, the very act of cutting the skin during the biopsy procedure is often sufficient to cause shearing of the epidermis from the dermis. In contrast, in certain subtypes of EB, notably localized EB simplex (EBS) (formerly called EBS-Weber-Cockayne), it can be very difficult to induce a blister, even after subjecting the skin to localized warming and extensive rubbing, so diagnostically useful biopsies are usually impractical and often unobtainable. Fortunately, in most cases of mild localized EBS a strong family and personal medical history, and relevant clinical features, are sufficient to support the diagnosis.
TEM fixatives, containing the powerful cross-linking agent glutaraldehyde, penetrate dense tissue such as skin relatively poorly. The addition of formaldehyde, a smaller molecule that diffuses into the tissue more rapidly, improves the penetration rate, but the biopsy sample still has to be subdivided to allow for adequate fixation. Shave biopsy samples are usually preferred since only shallow skin samples are required and the shearing motion that often accompanies the punch biopsy technique may result in total separation of the dermis from the epidermis if the skin is very fragile.
The sample should be immersed immediately in the TEM fixative since any delay will cause it to dry out and result in irreversible cellular and tissue damage, rendering it partly or totally useless for subsequent TEM analysis. The divided sample or subsamples must be small enough for adequate fixation and should also be large enough to allow for the provision of informative semithin (∼0.5–1.0 μm thick) sections for light microscopy (see below), an essential step before trimming the resin blocks further for ultramicrotomy and TEM examination. Semithin sections will enable the level of blistering or tissue splitting to be ascertained ( Fig. 1 ), thus complementing IF-antigen mapping, which will often be undertaken concurrently.