Susan Banks‐Schlegel

Different stratified squamous epithelia, whether they bear a stratum corneum or not, are shown by immunofluorescence to possess the precursor protein of the cross-linked envelope that is characteristic of epidermal s. corneum. This protein, involucrin, is not present in the deepest epithelial cells but appears in the course of their outward migration. The boundary at which involucrin first appears can sometimes by correlated with a visible boundary between zones of large and small cells. Cultured keratinocytes, derived from all stratified squamous epithelia (epidermal, corneal, conjuctival, esophageal, lingual, and vaginal), form colonies that grow together to form a stratified epithelium. The cells of the basal layer are nearly always free of detectable involucrin, but, in contrast to the natural epithelium, this protein usually makes its appearance in the cells immediately above the basal layer. When a cultured epithelium derived from epidermal keratinocytes is detached and applied as a graft to animals, the cells flatten and the distinctness of the basal layer is at first reduced; but with time the organization of the epithelium becomes more characteristic of epidermis. Cell size and shape become more orderly along the cell migration pathway, and involucrin first appears at some distance from the basal layer, instead of in immediately suprabasal cells, as in the cultured epithelium. The progeny of dissociated and cultured keratinocytes are therefore able, when grafted, to reassemble an epidermis in which the timing of specific gene expression is restored to that of the original tissue.

Immunohistochemical identification of intracellular keratin was achieved using an indirect antibody technique on paraffin-embedded human tissue. A study of numerous tissues confirms that keratins are abundant in all layers of squamous epithelia, in the ducts of epithelial-derived glands, and in the epithelia of the respiratory and urinary tracts. Using an immunoperoxidase technique which offers increased histologic resolution, we have shown that the basal or reserve cells of the tracheal, bronchial, prostatic, and cervical gland epithelia are the predominant keratin-containing cells in these tissues. The normal differentiation of basal cells into nondividing, superficial columnar cells is accompanied by the loss of cytoplasmic keratin proteins. Foci of epithelial squamous metaplasia stain intensely with antikeratin antibodies and presumably represent an exaggerated proliferation of the keratin-containing basal cells. Alveolar respiratory epithelium, acinar cells of various glands, and many mesodermal tissues (muscle, hematopoietic, and lymphoid tissue, nerve, and connective tissue) were devoid of keratin proteins. The ability to identify keratin proteins within fixed, embedded tissue (including those known to lack tonofilament bundles) may prove useful in the study of tissue histogenesis and carcinogenesis, and in the pathologic assessment of poorly differentiated malignant neoplasms and tumors of controversial cellular origin.

The distribution of intracellular keratin was studied in a variety of human tumors using a previously described immunoperoxidase technique employing antikeratin antibodies. Squamous cell carcinomas, transitional cell tumors, and mesotheliomas exhibited strong reactivity with antikeratin antibodies. Mammary adenocarcinomas were either negative or weakly positive. In the lung, an organ which can give rise to several morphologically distinct forms of carcinoma, only the squamous cell type stained strongly for keratin; undifferentiated lung carcinomas were negative, and adenocarcinomas were either negative or weakly positive. Colonic, renal, and prostatic adenocarcinomas were negative. Sarcomas, lymphomas, and neural tumors were uniformly negative. The analysis of intracellular keratin by the immunoperoxidase technique appears helpful in establishing the epithelial nature of primary or metastatic poorly differentiated neoplasms.