Immunohistochemical staining for desmogleins 1 and 2 in keratinocytic neoplasms with squamous phenotype: actinic keratosis, keratoacanthoma and squamous cell carcinoma of the skin.

Al Krunic,Re Clark,S Madani,Dr Garrod,Md Buchanan

doi:10.1038/bjc.1998.213

Abstract

Desmosomes are intercellular junctions that have been shown to be down-regulated in certain types of carcinoma and that may play a role in suppression of invasion and metastasis. This paper describes an immunohistochemical study of three types of epidermal neoplasms with monoclonal antibody to desmoglein in order to determine how desmosomal staining correlates with the clinical, biological and histopathological features of these neoplasms. Actinic keratosis (AK) is the most common keratinocytic premalignant neoplasm that was reported to have a 10-20% rate of malignant transformation into squamous cell carcinoma (SCC). Keratoacanthoma (KA) is a benign neoplasm that involutes spontaneously after a few months of rapid growth. SCC is a malignant tumour capable of metastasis. Electron microscope studies of KA and SCC showed significantly reduced staining for desmosomes in SCC but not in KA. We have examined staining for desmoglein using the monoclonal antibody 33-3D, a mouse IgM monoclonal antibody, that recognizes the cytoplasmic domains of desmoglein (Dsg)1 and Dsg2 on frozen sections. Immunohistochemical staining of normal skin with this antibody revealed strong pericellular localization of the antigen, outlining the cell membranes of the keratinocytes. A series of 30 AKs, 12 KAs and 24 SCCs was stained immunohistochemically with 33-3D monoclonal antibody. All examined KAs showed extensive pericellular staining for Dsg. By contrast, juxtanuclear staining for Dsg was noted in 12 SCCs, and completely negative staining in seven SCCs. The five remaining SCCs showed focal pericellular staining for the Dsg marker. The most common finding in AK was focal pericellular staining for Dsg, with complete absence of staining in dysplastic areas (25 cases). In five cases negative pericellular staining in dysplastic areas was associated with juxtanuclear accumulation of the Dsg marker. A strong negative correlation between Dsg staining and degree of dysplasia was obtained. The Dsg pattern in KA is similar to normal epidermis and shows a clear difference between KA and SCC. AK has a limited loss of Dsg expression in a SCC-like pattern that is congruent with its premalignant nature. As the stain works on frozen tissue, it may be helpful for rapid differentiation in selected cases in cutaneous oncology and Mohs micrographic surgery. This antibody may also have great potential for the detection of the effects of chemopreventive agents in skin cancer.

Highlights

Desmosomes are intercellular junctions that have been shown to be down-regulated in certain types of carcinoma
We have examined the staining for Dsg using the monoclonal antibody 33-3D, a mouse IgM monoclonal antibody, that recognizes the cytoplasmic domains of human Desmoglein 1 (Dsgl) and Dsg2 and can be used on frozen sections (Vilela et al, 1995; DR Garrod, unpublished observations)
The staining pattern was entirely similar to the Dsg distribution detected in the normal epidermis and to that previously found in KAs with monoclonal antibody 32-2B

Summary

Methods

Patients and specimensBiopsy samples of 12 KAs, 24 SCCs and 30 AKs were obtained from predominantly sun-exposed sites in 36 patients from the Dermatologic Surgery Unit, Duke University Medical Center, Durham, NC, USA. Strict clinical and histological criteria were used to differentiate between KAs, SCCs and AKs (Fisher et al, 1972; Janecka et al, 1978; Schwartz, 1994; Schwartz, 1996). Immunohistochemical staining procedure (Vilela et al, 1987; Burge and Garrod, 1991). The sections (5-6 gm) were mounted on Superfrost plus slides (Fisher Scientific, Pittsburgh, PA, USA), fixed in acetone for S min and air dried. The sections were incubated for 30 min with the 33-3D antidesmoglein monoclonal antibody (Vilela et al, 1995; Krunic et al, 1996; Schafer et al, 1996), rinsed in phosphate-buffered saline (PBS) and incubated for 15 min with the secondary biotinylated anti-mouse IgM. The sections were rinsed with PBS and incwtbated for 5 min with avidin-biotin-peroxidase conjugate, and with diaminobenzidine as a chromogen. The slides were washed in tap water and counterstained with haematoxylin

Results

Discussion

Conclusion