Keratin-like proteins in normal and neoplastic cells of human and rat mammary gland as revealed by immunofluorescence microscopy

Intermediate-sized Filament Proteins (Keratin, Vimentin, Desmin) in Metaplastic Carcinomas, Carcinosarcomas and Stromal Sarcomas of the Breast

The distribution of intermediate-filament (IF) proteins of the keratin, vimentin and desmin type in breast stromal sarcomas, carcinosarcomas, metaplastic carcinomas and phyllodes tumors has been compared using the avidin-biotin complex immunoperoxidase technique. Keratin reactivity was found in carcinomatous and pseudosarcomatous areas of all metaplastic carcinomas, in the cuboidal epithelial cells of carcinosarcomas and in the epithelial component of phyllodes tumors. Vimentin and desmin were detected in the sarcomatous portion of carcinosarcoma, focally in the stromal component of phyllodes tumors and not always in the stromal sarcomas. These data confirm that combined analysis of IF expression is a reliable and convincing way to differentiate stromal sarcomas, metaplastic carcinomas and carcinosarcomas in breast pathology.

Lactation affects expression of intermediate filaments in human breast epithelium

The human breast contains two epithelial lineages, luminal epithelial and myoepithelial. Specific patterns of expression of intermediate filaments have previously been demonstrated in the resting breast. To determine how terminal differentiation and lactation influenced expression of intermediate filaments in breast epithelial cells, we used Western blot analysis to measure the levels of vimentin, alpha-smooth muscle actin, keratin 14, and keratin 18 in the resting and lactating breast. Confocal immunofluorescence was used to determine the subcellular site of localization of the intermediate filaments. Vimentin was localised to myoepithelial cells in both the resting and lactating gland. There was a four-fold increase in vimentin protein levels in lactating tissue relative to resting tissue, and this may be related to increased cellular activity of the myoepithelial cells which surround secretory alveoli. Alpha-smooth muscle actin and keratin 14 were detected in myoepithelial cells, and similar levels of expression were found in lactating and resting tissue. In the resting breast, keratin 18 and keratin 8 were detected in luminal epithelial cells in a filamentous form, whereas in lactating tissue it was present in a punctate form in luminal cells and also seen as granules in the lumen of alveoli. Our results indicate that intermediate filament expression patterns are altered in the lactating human breast, and this may reflect their role in the fully functional gland.

Immunohistochemistry with keratin, vimentin, desmin, and alpha-smooth muscle actin monoclonal antibodies in canine mammary gland: normal mammary tissue

Normal canine mammary gland tissue was studied immunohistochemically with monoclonal antibodies (MoAbs) directed against various human keratin types, vimentin, desmin, and alpha-smooth muscle actin. Both ductal and alveolar luminal cells were immunoreactive with MoAbs recognizing respectively human keratins no. 7, 8, 18 and 19. In addition, some ductal luminal cells were labelled with a keratin 4 and a keratin 10 MoAb. Basal/myoepithelial cells were immunoreactive only with MoAbs directed against keratin 14, keratins 14 and 17, and alpha-smooth muscle actin. The vimentin MoAb merely labelled solitary loose intraluminal cells representing macro-phages or sloughed epithelial cells. These findings correspond largely to observations made in human breast tissue.

Intra-abdominal desmoplastic small cell tumor in an adolescent suggesting a neurogenic origin

A case of a desmoplastic small cell tumor of the large omentum associated with gross ascites that occurred in a male adolescent is reported. Light microscopic studies revealed that the tumor cells were small and epithelioid in nature with eosinophilic hyaline material located in the perinuclear area. They were surrounded by rich desmoplastic and myxoidal stromal bands. Immunohistochemical staining revealed globoid perinuclear positivity for desmin. Vimentin, cytokeratin (AE3, CaM 5.2), epithelial membrane antigen, tissue polypeptide antigen, neuron-specific enolase, chromogranin A, endocrine granule constituent and synaptophysin were also positive in the cytoplasm. Electron microscopy revealed whorled intermediate filaments and some dense core granules in the cytoplasm. Bundles of microtubules in the cytoplasmic process and occasional cell junctions of zonulae adherentes in the tumor cells were also observed. DNA analysis of the tumor cells showed the three-fold amplification of the N-myc gene. Although desmoplastic small cell tumors showed a heterogeneous pattern with immunohistochemical studies, it is suggested that the tumor may originate from neurogenic cells.

An immunohistochemical study of the breast using antibodies to basal and luminal keratins, alpha-smooth muscle actin, vimentin, collagen IV and laminin. Part I: Normal breast and benign proliferative lesions

The distribution of simple epithelial (K8/18/19) and basal (myoepithelial) (K5/14) keratins, alpha-smooth-muscle actin, vimentin, collagen IV and laminin in normal mammary glands and in benign proliferative lesions was studied using monoclonal antibodies (mAbs). These antibodies (Abs) identified myoepithelial cells and luminal cells specifically. In lesions with adenosis and papillomas, the two-layered formation resembled that of normal glands with a purely myoepithelial-epithelial differentiation. In scleradenotic lesions, the main cell was of myoepithelial immunophenotype with intermixed trabecular-tubular proliferations of simple-type epithelium. The sclerosis seems to be the result of an irregular basal lamina synthesis by the myoepithelial cells. In contrast to these lesions, epitheliosis represents a purely intraluminal cell proliferation of clearly simple epithelial immunophenotype and of cells with a basal keratin phenotype, lacking myoepithelial differentiation antigen actin. The basal keratin type epithelium may represent post-stem or intermediate cells developing into luminal epithelium. Epitheliosis appears to be a purely epithelial hyperplasia with striking similarity to the regeneration of normal breast epithelium. The different proliferative patterns may give an explanation for differences in potential cancer risks of patients with these lesions.

Intermediate filaments in DMBA-induced mammary adenocarcinomas in rats

Experimental DMBA (7,12-dimethylbenzanthracene)-induced mammary tumours in rats were studied. Microscopic examination revealed a high rate of adenocarcinomas (95.3% in 148 mammary tumours studied). An immunohistochemical study of cytoskeleton intermediate filaments (I.F.) showed cytokeratins in neoplastic epithelial cells; myoepithelial cells expressed cytokeratins, vimentin and desmin. Finally we propose the immunocytochemical localization of desmin as a useful method to recognize myoepithelial proliferation areas in mammary tumours.

Keratins as markers that distinguish normal and tumor-derived mammary epithelial cells

Keratin 5 (K5) mRNA and protein are shown to be expressed in normal mammary epithelial cells in culture and are absent from tumor-derived cell lines. To extend these findings, the full complements of keratins in normal, immortalized, and tumor cells were compared. It is shown here that normal cells produce keratins K5, K6, K7, K14, and K17, whereas tumor cells produce mainly keratins K8, K18, and K19. In immortalized cells, which are preneoplastic or partially transformed, the levels of K5 mRNA and protein are lower than in normal cells, whereas the amount of K18 is increased. Thus, K5 is an important marker in the tumorigenic process, distinguishing normal from tumor cells, and decreased K5 expression correlates with tumorigenic progression.

Keratin staining of canine epithelial tissues by a polyclonal antiserum

The keratin distribution pattern in various canine epithelial tissues has been studied using a commercially available rabbit antiserum, raised against human skin keratins, in an immunoperoxidase staining method with the peroxidase-antiperoxidase complex (PAP). The staining results of two fixation methods were compared. Paraffin sections after fixation in Carnoy's solution showed optimal results, whereas paraffin sections after fixation in 10% buffered formalin resulted in a strongly reduced keratin staining reaction. Keratinizing and non-keratinizing stratified epithelial showed a strong staining reaction. Glandular epithelium and the non-stratified epithelia of internal organs e.g. liver, kidney and intestine did not react with the antiserum. However, glandular ductal epithelium, myoepithelial cells and basal cells in various epithelial tissues showed a positive staining reaction. The results indicate the presence of different keratin types in these canine epithelial tissues. The keratin distribution pattern is compared with the distribution pattern observed in tissues of other species.

Immunohistologic heterogeneity of malignant tumors

Intratumor heterogeneity exists in most, if not all, human solid tumors and is reflected by differences in morphology, antigenicity, chromosomal and biochemical markers, invasion, metastatic spread, as well as sensitivity to therapy. Heterogeneity may be the result of genetic and/or environmental mechanisms. To be assessed by immunohistochemistry, heterogeneity of tumor cells has to be expressed by antigenic changes, i.e. quantitative or qualitative differences in antigen expression, secretion or uptake. Consequently, presence and distribution of hormones, steroid hormone receptors (e.g., in mammary carcinomas), blood group antigens (e.g., in diverse carcinomas), and cytoskeletal components, particularly intermediate filament-associated proteins (e.g., in neoplasms of diverse origin), can indicate heterogeneity within a tumor.

Morphogenesis of mouse mammary epithelial cells growing within collagen gels: ultrastructural and immunocytochemical characterization

Mammary epithelial cells from adult virgin mice have been cultured within collagen gels in totally serum-free medium containing either epidermal growth factor or the mammogenic hormones, progesterone and prolactin, or prolactin alone. The cellular organization, differentiation and cell-type composition of the colonies from the three culture conditions were assessed by transmission electron microscopy and light-microscope immunocytochemistry. The epithelial cells form branching duct-like structures and, when exposed to mammogenic hormones, assume a secretory morphology (including casein micelles) similar to that seen in the early to mid-pregnant mouse.

Immunohistochemistry and ultrastructure of myoepithelium and modified myoepithelium of the ducts of human major salivary glands: histogenetic implications for salivary gland tumors

The organization of salivary gland ducts, especially the presence or absence of myoepithelial cells, is central to histogenetic approaches to the classification of salivary gland tumors. Striated and excretory ducts are reported to be devoid of myoepithelial cells but do contain basal cells. To investigate the nature of such basal cells, tissue sections of normal human salivary glands were examined by means of immunohistochemical, ultrastructural, and fluorescent microscopic techniques. With the use of a mouse monoclonal anticytokeratin antibody (3 12C8-1) that, in salivary glands, is specific for myoepithelial cells, these cells associated with acini and intercalated ducts were strongly stained, as were the basal cells of striated and excretory ducts in each case. Ultrastructurally, some basal cells of both striated and excretory ducts had narrow, elongated cellular processes or the main portion of the cell containing parallel arrays of microfilaments with linear densities and micropinocytotic vesicles, whereas in other basal cells tonofilament bundles predominated. A similar range of cytoplasmic features existed in myoepithelial cells associated with acinar and intercalated duct cells. In addition, some duct basal cells have a complement of actin filaments similar to classic myoepithelium of acini and intercalated ducts. Striated and excretory ducts of human salivary glands, therefore, contain fully differentiated and modified myoepithelial cells, both of which express a specific cytokeratin polypeptide that is absent from duct luminal and acinar cells. Differentiation patterns in the intralobular and interlobular ducts suggest that these regions of salivary gland parenchyma cannot be excluded as histogenetic sites for the induction of salivary gland tumors in which neoplastic myoepithelial cells have been shown to have a major role.

Structural distinctions among human breast epithelial cells revealed by the monclonal antikeratin antibodies AE1 and AE3

Two monoclonal antikeratin antibodies, AE1 and AE3, were used in indirect immunocytochemistry to examine keratin expression in normal, benign proliferative, and malignant human breast epithelium. Both antibodies reacted strongly with most luminal cells in ducts and acini of normal gland. While AE1 did not stain myoepithelium, AE3 recognized myoepithelial cells of ducts but not acini, implying a cytoskeletal difference between the myoepithelium of these two components. Moreover, the antibodies reacted differently with the myoepithelium of intracanalicular as compared with pericanalicular types of fibroadenomas. Tumour cells of infiltrating ductal carcinomas with a prominent intraductal component stained more homogeneously with AE1 and AE3 than those without intraductal growth. The results provide evidence for two phenotypes of myoepithelial cells and for the presence of cryptic keratin epitopes in human breast epithelial cells. The finding that neither AE1 nor AE3 is a universal detector of these cells has important clinical and experimental implications.

Keratin polypeptide distribution in benign and malignant breast tumors: subdivision of ductal carcinomas using monoclonal antibodies

Monoclonal antibodies which recognize one or only a few keratin polypeptides have been used to study the distribution of different keratins in benign and malignant breast lesions by immunocytochemical methods. Seven monoclonal antibodies which recognized either different keratin polypeptides by immunoblotting techniques, or identified different epithelial cell types in complex tissues were used. In two mastopathies and three fibroadenomas the antibody lu5 stained luminal cells as well as myoepithelial cells. In contrast the antibodies CK7, Troma 1, CK2 and KA4 labeled only luminal cells, whereas antibody CKB1 decorated only myoepithelial cells. All 15 ductal carcinomas showed a uniform staining of tumor cells with the antibodies Troma 1, CK2, KA4 and lu5. The antibody CK7 also stained all ductal carcinomas, but in two specimens the staining was heterogeneous. The antibody CKB1 decorated only the pre-existing myoepithelial cells in 11 of 12 ductal carcinomas but in the remaining specimen the tumor cells were also strongly positive. Tumor cells in lobular carcinomas were labeled by antibodies CK7, Troma 1, CK2, KA4, bu not by CKB1. The antibody CKS1 showed no staining of any of the benign and malignant breast lesions.

Immunohistochemical localization of keratin in human prostate

The localization of keratin was investigated in normal human prostate, benign prostatic hyperplasia (BPH), and prostatic adenocarcinoma using immunohistochemical technique on frozen tissue sections. The purpose of this study was to identify changes in the distribution patterns of keratin due to malignancy. In normal and benign tissue specimens, keratin was detected in the cytoplasm of basal cells and glandular epithelial cells. In the glandular epithelial cells keratin was found as a deposit of fine granules. In the basal cells, the positive staining for keratin had a uniform distribution in the scanty cytoplasm. In specimens of prostatic adenocarcinoma, the basal cells retained a strong positive reaction for keratin. The shapes and the distributions of basal cells were markedly different in malignant specimens. Basal cells formed a discontinuous layer and surrounded the population of neoplastic cells in tissue sections containing the cribriform patterns. The cells expressed characteristic protrusions into extracellular spaces between the cancer cells. Keratin-positive granules were demonstrated in the adenocarcinoma cells as well. These granules had slightly smaller sizes and were distributed randomly. The study demonstrates that the immunohistochemical localization of keratin provides a refinement for the characterization of cells in tissue sections of prostate.

Distinction of mesothelioma from adenocarcinoma. An immunohistochemical approach

The authors investigated the expression of keratin, carcinoembryonic antigen (CEA), and an epithelial marker derived from milk fat globule membranes in 12 mesotheliomas and 100 diverse adenocarcinomas with immunohistochemical methods. The authors employed a monoclonal antibody to keratin designated as AE1, as well as the following commercially available antisera: rabbit anti-whole human keratin, rabbit anti-CEA, and a monoclonal antibody to an epithelial factor designated as MFG-2. Expression of keratin was found in all the mesotheliomas and adenocarcinomas with antibody AE1 as well as with the rabbit antiserum; CEA was detectable in 65% of the adenocarcinomas but two mesotheliomas also reacted weakly. With antibody MFG-2, positive results were obtained in 85% of the adenocarcinomas and in none of the mesotheliomas. All of 64 (100%) breast-, lung- and ovary-derived adenocarcinomas immunostained positively with antibody MFG-2. This is of particular significance because pulmonary and ovarian adenocarcinoma frequently may be indistinguishable clinically and histologically from epithelial mesothelioma. The authors conclude that antikeratin antibodies are not useful in the distinction of adenocarcinoma from mesothelioma. Because of its greater sensitivity and specificity, MFG-2 is superior to CEA in this differential diagnosis.

Expression of keratins and other cytoskeletal proteins in mouse mammary epithelium during the normal developmental cycle and primary culture

Mammary epithelium is composed of ductal, alveolar, and myoepithelial cells, and undergoes dramatic responses in growth, differentiation, and function to hormonal stimuli during the four stages of the mammary developmental cycle represented in virgin, pregnant, lactating, and involuting animals. To determine if progression of the epithelium through the cycle is accompanied by changes in cytoskeletal composition, particularly the keratins, the polypeptides in cytoskeletal extracts from BALB/c mouse mammary tissues were analyzed by one- and two-dimensional gel electrophoresis combined with immunoblots using polyclonal and monoclonal antikeratin antibodies. The major polypeptides in cytoskeletal fractions enriched in intermediate filaments included seven acidic and three basic components ranging in molecular weight from 40,000 to 90,000. Two major polypeptides of Mr 50,000 and 40,000, along with two minor components of Mr 57,000 and 55,000 were identified as keratins. The polypeptide profiles of mammary glands from virgin, pregnant, lactating, and involuting mice were very similar, indicating a remarkable stability of cytoskeletal composition during hormonal shifts and periods of minimal or maximal cell growth and differentiated function. The data also suggest that ductal and alveolar cells express the same set of cytoskeletal polypeptides, including keratins. Mammary cells grown in primary culture exhibited a loss or reduction in most of the basic polypeptides, a large increase in an acidic Mr 55,000 keratin, and the appearance of a prominent acidic polypeptide of Mr 46,000. The latter results demonstrate that keratin expression in mouse mammary epithelial cells is subject to regulation by certain environmental factors.

Developmental regulation of cytokeratins in cells of the rat mammary gland studied with monoclonal antibodies

We have isolated two monoclonal antibodies to cytokeratins and determined their cell specificities. They display interesting localization within the rat mammary gland. One (1A10) shows specificity for myoepithelial cells; the other (24B42) is specific for lumenal cells at various stages of development. These two monoclonal antibodies and three others to cytokeratin previously isolated were used in conjunction with antibodies to myosin and collagen IV to confirm and extend our previous findings on epithelial cell types and development within the mammary gland.

Antibodies to intermediate filament proteins in the diagnosis and classification of human tumors

Immunohistochemistry of intermediate filaments (IF) is a new and important way to evaluate the epithelial, mesenchymal, muscular, glial, or neural differentiation in tumors. This is based on the stable cell-type-specific expression of IF proteins in normal and neoplastic tissues. Immunohistochemical studies with antibodies to intermediate filaments have also given new perspectives in the histogenesis and biologic nature of many tumors. This article reviews both the recent findings and the authors' experience in the use of intermediate filament antibodies in tumor diagnosis and classification.

The organization of intermediate filaments in normal human colonic epithelium and colonic carcinoma cells

A comparative ultrastructural study of the organization of intermediate filaments (tonofilaments) within normal colonic epithelium and colonic carcinoma cells was carried out. The pattern of development of the specialized intermediate filament (IF) network, occurring during the differentiation of normal colonic epithelial cells, was examined to allow a more complete comparison with tumour cells of different degrees of differentiation. Essentially the IF system becomes more organized as the cells differentiate, resulting in the formation of a three-dimensional network loosely surrounding the nucleus, extending through the sub-terminal web and basal cytoplasm, and inserting into lateral desmosomes. Another system of IFs also runs directly between lateral desmosomes, lying parallel to the plasma-membrane and desmosomal plaque, without inserting directly into the junction. Alterations in the organization of the IF system were seen in tumour cells with the type and extent being dependent upon their location and degree of differentiation. Cells within areas of the tumour where some degree of glandular structure was retained exhibited major alterations in their microfilament-containing structures. However, their IF system was essentially intact and lateral desmosomes were still present in similar numbers to those seen in normal, partially differentiated mid-crypt cells. Apolar tumour cells within areas of the tumour where a gross loss of cryptal architecture had occurred exhibited a striking disorganization of their IF system. No parallel bundles of interdesmosomal tonofilaments were present beneath the cell surface and the majority of the three-dimensional network appeared to have collapsed around the nucleus, although some extensions to randomly distributed surface desmosomes still occurred. It appears that this disorganization of the IF system is associated with a loss of cell polarity and may involve alterations in putative IF-associated proteins important in the interaction of IFs with surface desmosomes or other cytoskeletal elements. A study of this phenomenon may shed light on both the means of IF organization and assembly and their role in normal cells as well as the possible role of alterations in cytoskeletal elements in the expression or maintenance of the malignant phenotype.

Antibodies to intermediate filament proteins in the immunohistochemical identification of human tumours: an overview

Intermediate-sized filament proteins (IFP) are tissue specific in that antibodies to keratin, vimentin, desmin, glial fibrillary acidic protein (GFAP) and the neurofilament proteins can distinguish between cells of epithelial and mesenchymal origin as well as of myogenic and neural origin respectively. Malignant cells retain their tissue-specific IFP, which makes it possible to use these antibodies in tumour diagnosis. Carcinomas are exclusively detected by antibodies to keratin. Monoclonal antibodies to keratin have allowed the differentiation between subgroups of epithelial tumours until now between adenocarcinomas and squamous cell carcinomas. Lymphomas, melanomas and several soft tissue tumours are distinctly recognized by antibodies to vimentin. On the other hand, rhabdomyosarcomas and leiomyosarcomas are positive for desmin, while astrocytomas give a strong reaction with GFAP antibodies. Thus, antibodies to IFP are useful tools for differential diagnosis in surgical pathology.

An epithelial cell line with elongated myoid morphology derived from bovine mammary gland. Expression of cytokeratins and desmosomal plaque proteins in unusual arrays

Cells of a clonal line (BMGE + HM) selected from bovine mammary gland epithelial cell cultures are described which, after reaching confluence, do not assume typical epithelioid morphology, but form elongated cells with long slender processes extending over the surfaces of other cells. However, cells of this line which display non-epithelioid morphology and are exceptionally rich in actin microfilaments are identified as epithelial cells by their synthesis of cytokeratins and desmosomal plaque proteins, as demonstrated by immunofluorescence and immunoelectron microscopy and by gel electrophoresis of cytoskeletal proteins. The cells do not produce vimentin and desmin filaments. The specific cytokeratin polypeptides of these myoid cells are identical to those present in normal epithelioid BMGE + H cells but are arranged in unusual arrays of meshworks of finely dispersed, non-fasciated filaments and granular structures. Desmosomal plaque proteins, notably desmoplakins, are abundant, but the electron microscopic appearance of the desmosomes is abnormal in that most of them are associated with a second accessory plaque formed at a distance of 0.1-0.15 micron from the normal desmosomal plaque. Both cytokeratin filaments and desmosomal structures are found throughout the whole cytoplasm, including the extended cell processes. The existence of an epithelial cell line with such an unusual morphology demonstrates the importance of non-morphological criteria in identifying epithelium-derived cells. Our findings also indicate that dramatic differences of cell shape and organization of epithelial cells need not necessarily be associated with changes in the expression of specific cytoskeletal proteins. The possible origin of this cell line from myoepithelial cells is discussed.

Intermediate filaments in malignant melanomas. Identification and use as marker in surgical pathology

Intermediate-sized filaments have been studied in human malignant melanomas and in normal melanocytes by immunofluorescence microscopy with antibodies directed against keratin, vimentin, desmin, neurofilament protein, and glial filament protein. Both human melanotic and amelanotic tumor cells and tumor metastases as well as normal melanocytes in human skin and in the rat eye contain exclusively intermediate filaments of the vimentin type. No reaction was seen with antibodies to keratin, desmin, neurofilaments, or glial filaments. These latter four antisera, however, gave strong reactions in epidermis and other epithelial tissues, muscle, or neural tissues, respectively. The results favor a mesenchymal character of melanocytes, although a neuroectodermal origin in an early developmental stage is possible. The finding that melanomas contain exclusively vimentin intermediate filaments may prove useful in differential diagnosis of melanomas from other tumor types.

Neuroendocrine carcinoma of the skin (Merkel cell carcinoma): ultrastructural and immunohistochemical demonstration of neurofilaments

In this study we characterized a skin tumor that grew in the temporal region of a 69-year-old woman. On the basis of tumor morphology, a metastasis from a small cell carcinoma of the lung was initially suggested, but X-ray and bronchoscopic studies were negative. The tumor recurred twice within a year, yet no tumors were found elsewhere in the body. Ultrastructurally, cytoplasmic organelles compatible with neuroendocrine storage granules and perinuclear aggregates of intermediate-sized (8-10 nm) filaments were found in many tumor cells. Indirect immunofluorescence microscopy revealed neurofilament-type intermediate filaments in the tumor cells but no keratin- or vimentin-type filaments. Our results further demonstrate neural properties of this tumor type, which is generally considered to have its origin from Merkel cells, the cutaneous neuroendocrine cells.

Immunocytochemical identification of epithelium-derived human tumors with antibodies to desmosomal plaque proteins

Epithelial cells contain desmosomes, special intercellular junctions providing sites of membrane attachment for intermediate-sized filaments of the cytokeratin type (tonofilaments). Such sites of anchorage of tonofilaments appear as dense plaques on the cytoplasmic side of the desmosomal membrane. We have isolated desmosome-enriched fractions from bovine snout epidermis and tongue mucosa and have characterized the major protein associated with the desmosomal plaque. This protein occurs in equimolar amounts of two polypeptides of Mr 250,000 (desmoplakin I) and Mr 215,000 (desmoplakin II) which are chemically and immunologically related. Antibodies raised against desmoplakins allow the identification and localization of this protein in epithelial cells grown in tissues or in vitro and show crossreaction in species as diverse as man, mouse, and chicken. Using immunolocalization at the light and electron microscope levels, we show that these antibodies bind specifically to desmosomal plaques. Antibodies to desmoplakins have been used successfully for detection of desmosomal proteins in a broad variety of epithelium-derived human tumors, including primary carcinomas and their metastases, irrespective of the morphology of the specific tumor. Nonepithelial tumors examined have been negative. We propose to use antibodies to desmoplakins and to cytokeratins in pathological diagnosis as two independent markers for the positive immunocytochemical identification and classification of epithelium derived tumors.

Tissue type-specific expression of intermediate filament proteins in a cultured epithelial cell line from bovine mammary gland

Different clonal cell lines have been isolated from cultures of mammary gland epithelium of lactating cow's udder and have been grown in culture media containing high concentrations of hydrocortisone, insulin, and prolactin. These cell (BMGE+H), which grow in monolayers of typical epithelial appearance, are not tightly packed, but leave intercellular spaces spanned by desmosomal bridges. The cells contain extended arrays of cytokeratin fibrils, arranged in bundles attached to desmosomes. Gel electophoresis show that they synthesize cytokeratins similar, if not identical, to those found in bovine epidermis and udder, including two large (mol wt 58,500 and 59,000) and basic (pH range: 7-8) and two small (mol wt 45,500 and 50,000) and acidic (pH 5.32 and 5.36) components that also occur in phosphorylated forms. Two further cytokeratins of mol wts 44,000 (approximately pH 5.7) and 53,000 (pH 6.3) are detected as minor cytokeratins in some cell clones. BMGE+H cells do not produce vimentin filaments as determined by immunofluorescence microscopy and gel electrophoresis. By contrast, BMGE-H cells, which have emerged from the same original culture but have been grown without hormones added, are not only morphologically different, but also contain vimentin filaments and a different set of cytokeratins, the most striking difference being the absence of the two acidic cytokeratins of mol wt 50,000 and 45,500. Cells of the BMGE+H line are characterized by an unusual epithelial morphology and represent the first example of a nonmalignant permanent cell line in vitro that produces cytokeratin but not vimentin filaments. The results show that (a) tissue-specific patterns of intermediate filament expression can be maintained in permanent epithelial cell lines in culture, at least under certain growth conditions; (b) loss of expression of relatively large, basic cytokeratins is not an inevitable consequence of growth of epithelial cells in vitro. Our results further show that, during culturing, different cell clones with different cytoskeletal composition can emerge from the same cell population and suggest that the presence of certain hormones may have an influence on the expression of intermediate filament proteins.

Intermediate filaments and their interaction with membranes. The desmosome-cytokeratin filament complex and epithelial differentiation

Intermediate-sized filaments represent a class of morphologically similar but biochemically and immunologically distinguishable cytoplasmic protein polymer structures. Five major filament types have been identified (cytokeratin, vimentin, desmin, neurofilament protein, glia filament protein) and antibodies to these proteins have been used for distinguishing different cell types and tumors derived therefrom. Epithelial and carcinoma cells are characterized by the presence of cytokeratin filaments and desmosomal elements identified by antibodies to certain high molecular weight proteins of desmosomal plaques. However, the specific pattern of cytokeratin polypeptides is different in different epithelia. The potential value of cell type identification by immunological reactions with antibodies to cytoskeletal proteins in tumor diagnosis is discussed.

Complex cytokeratin polypeptide patterns observed in certain human carcinomas

Human epithelial cells contain, intermediate-sized filaments formed by polypeptides related to epidermal alpha-keratin ("cytokeratins") which are expressed in different combinations in different epithelia. Using cytoskeletal proteins from human biopsies and autopsies we have examined, by two-dimensional gel electrophoresis and immunoblotting experiments, the cytokeratin polypeptide patterns of diverse primary and metastatic carcinomas and have compared them with those of corresponding normal epithelial tissues and cultured cells. Five groups of carcinoma cytokeratin patterns can be discriminated. (1) Cytokeratins typical of simple epithelia (polypeptides Nos. 7, 8, 18, 19) are expressed, in various combinations, by many adenocarcinomas, for example those of gastrointestinal tract. (2) Cytokeratins typical of stratified epithelia (Nos. 1, 5, 6, 10, 11, 14-17) are found, in various combinations, in squamous cell carcinomas of skin and tongue. (3) Complex patterns showing polypeptides Nos. 7, 8, 18, 19, and one basic component (No. 5 or 6) are detected in certain carcinomas of the respiratory tract and the breast. (4) Complex patterns containing cytokeratins widespread in stratified epithelia (Nos. 4-6, 14-17) as well as components Nos. 8 and 19 occur in diverse squamous cell carcinomas derived from non-cornified stratified epithelia, with or without additional small amounts of cytokeratin No. 18. (5) Patterns of unusually high complexity can be found in some rare tumors as is shown for a cloacogenic carcinoma. No significant qualitative changes of expression of cytokeratins were found when primary tumors and metastases were compared. When compared with cytokeratin patterns of normal epithelia, carcinomas of the first type usually display a high degree of relatedness to the tissue of origin. Other carcinomas do not express some of the cytokeratins present in the tissue of their origin and, vice versa, certain components which are minor or apparently absent in normal tissue are major cytokeratins in the corresponding tumor. These differences may be explained by cell type selection during carcinogenesis, but changes of expression during tumor development cannot be categorically excluded. The possibility of cell type heterogeneity within a given tumor is also discussed. Similarly complex patterns of cytokeratin polypeptides have been noted in certain cultured human carcinoma cell lines (e.g., A-431, RPMI 2650, Detroit 562, A-549) and can also be observed in cell clones. The possible value of analyses of cytokeratin patterns, by gel electrophoresis or specific monoclonal antibodies, in distinguishing different carcinomas by non-morphologic criteria is discussed.

Antibodies to intermediate filament proteins as molecular markers in clinical tumor pathology. Differentiation of carcinomas by their reaction with different cytokeratin antibodies

Antibodies to human and bovine epidermal prekeratin and antibodies to mouse liver cytokeratin component D (Mr 49 000) have been applied in indirect immunofluorescence microscopy on sections of human tumors of mammary gland and liver. In non-neoplastic mammary gland all epithelial cells were stained with these antibodies. In pre-invasive and invasive ductal and lobular carcinomas a cell population was observed which was not significantly stained with antibodies to epidermal prekeratin but did strongly react with antibodies to liver cytokeratin D. In the liver, the antibodies to epidermal prekeratin as well as those directed against liver cytokeratin D strongly decorated bile duct epithelia. In contrast, significant staining of the hepatocytes was only achieved with antibodies to liver cytokeratin D. This different staining reaction was maintained in liver tumors of hepatocellular and cholangiocellular origin. Antibodies to vimentin stained mesenchymal cells and tumors of mesenchymal derivation but reacted not significantly with any of the epithelial and carcinoma cells examined. The difference is of practical importance for the discrimination between anaplastic carcinomas and sarcomas of unknown origin. Cytokeratin could also be detected by antibody staining using the peroxidase-antiperoxidase (PAP) technique in formaldehyde-fixed and paraffin-embedded material of skin, gastrointestinal, respiratory, urinary and genital tract as well as various glands, liver and kidney. Examples of positive reactions were shown in a squamous cell carcinoma, a basalioma and a pleomorphic adenoma of the parotis. It is concluded that the immunohistochemical analysis of intermediate filament proteins has diagnostic potential in clinical pathology and may help to elucidate histogenesis and differentiation of tumors and possibly also prognosis of tumor growth. It is further suggested to use antibodies recognizing different subsets of proteins of the cytokeratin family in order to distinguish between different types of carcinomas.

Desmoplakins of epithelial and myocardial desmosomes are immunologically and biochemically related

Guinea pig antibodies against desmoplakins from bovine muzzle epidermis showed specific reaction in several epithelial tissues with desmoplakin I (Mr 250,000) and desmoplakin II (Mr 215,000). By immunofluorescence microscopy, prominent punctate staining was observed in various lines of cultured epithelial cells, revealing desmosomal junctions at sites of established cell-to-cell contacts as well as hemidesmosomes and internalized desmosome-derived membrane domains. On frozen tissue sections punctate staining was observed along plasma membranes of epithelial cells, and electron microscopy using the immunoperoxidase technique revealed that the antibodies were specifically localized at the plaques associated with desmosomes and hemidesmosomes. Of a large number of non-epithelial cells examined positive staining was only observed on desmosome-like junctions of myocardial cells and Purkinje fiber cells. In both epithelial and myocardial tissues the antibodies showed a broad range of cross-reactivity between diverse vertebrate species such as man, cow, rodent, and chicken, indicating that desmoplakins contain determinants strongly conserved during evolution. When binding of these antibodies to cytoskeletal polypeptides separated by gel electrophoresis and blotted on nitrocellulose paper sheets was examined, specific reaction was noted with desmoplakin I and, to a variable degree, also desmoplakin II from various epithelial cells. Reaction was also observed with a myocardial polypeptide from bovine and human hearts which had a similar Mr value (250,000) and isoelectric pH range as desmoplakin I. We conclude that desmoplakins are the major proteins present in the desmosomal plaques of both epithelial and myocardial cells and that the desmoplakin polypeptides present in these two different cell types are very similar, if not identical.

Spatial distribution of proteins specific for desmosomes and adhaerens junctions in epithelial cells demonstrated by double immunofluorescence microscopy

The spatial relationships between the protein constituents to two junctional structures, adhaerens junctions and desmosomes, were determined by double immunofluorescence microscopy using marker proteins specific for these structures. Adhaerens junctions were visualized by immunofluorescent labeling for the membrane-associated protein vinculin and by their association with actin filaments. Desmosomal components were identified by labeling with antibodies to a group of minor desmosomal plaque proteins (DP1 antigens) and their association with filaments stained by cytokeratin antibodies. Double immunofluorescence microscopy of these components was performed in several tissues and cultured cells, including intact intestine, dissociated intestinal cells, and two morphologically different types of epithelial cells, cultured bovine kidney (MDBK), and mammary gland (BMGE) epithelial cells. This allowed the direct demonstration that each filament system is associated exclusively with its specific membrane-bound junctional protein. Vinculin and DP1-protein were found in distinct sites in the subapical intercellular junctional complex of intestinal epithelium and MDBK cells. Cell-substrate focal contacts contained vinculin and actin and showed no apparent relationships to the tonofilament system whereas intercellular contacts of BMGE cells were characterized by positive staining for DP1-protein and associated cytokeratin filaments. Immunolabeling of the cultured cells at different intervals after plating for the cytoskeletal elements and their membrane anchorage proteins was used to determine the temporal sequence of their organization. We propose that this approach may be used for the molecular definition and identification of cellular contacts and junctions as well as for studies of junction topology, dynamics and junction-cytoskeleton interactions, and junction biogenesis.

An unusual type of cytokeratin filament in cells of a human cloacogenic carcinoma derived from the anorectal transition zone

Epithelia-derived tumors (carcinomas) can be distinguished from mesenchymally derived tumors by the presence of intermediate-sized filaments of the cytokeratin type, which usually coincides with the absence of other types of intermediate-sized filaments such as vimentin filaments. In the course of diagnostic examinations of human tumors, using immunofluorescence microscopy, we have come across a case of an unusual carcinoma (Primary tumor and lymph node metastasis) positively stained not only with cytokeratin antibodies but also with immunoglobulins present in vimentin antisera. Therefore, this tumor, a cloacogenic carcinoma apparently derived from the rectal-anal transitional region, has been examined in greater detail using both immunofluorescence microscopy and immuno-electron microscopy as well as gel electrophoretic analysis of cytoskeletal polypeptides from total tumor tissue and from microdissected nodules enriched in carcinoma cells. The unusual reaction of the carcinoma cells with immunoglobulins present in seven different (rabbit or guinea pig) antisera raised against vimentin, has been found to be diminished after absorption on purified cytokeratin or total epidermal cytoskeletal material, but not after absorption on purified vimentin. Gel electrophoretic analysis of tumor cytoskeletons showed an unusual complex pattern of cytokeratin polypeptides containing relatively large (Mr 68,000 and Mr 58,000) neutral-to-slightly basic cytokeratins, as are typically found in epidermis and other stratified squamous epithelia, as well as several smaller acidic cytokeratins, including a Mr 40,000 polypeptide found in certain nonstratified epithelial such as colon and small intestine. Total tumor also showed the inclusion of some vimentin which, however, was significantly decreased in analysis of excised carcinoma nodules. Examining antibody binding to polypeptides separated by gel electrophoresis and blotted on nitrocellulose paper, we have found that antisera raised against vimentin contained not only vimentin antibodies but also immunoglobulins which specifically bound to the largest cytokeratin component. We conclude that the unusual reaction of immunoglobulins present in vimentin antisera with cytokeratin filament bundles does not represent specific binding to vimentin in these carcinoma cells, but is due to a component obviously widespread in vimentin antisera which binds specifically to a cytokeratin present in this type of tumor but not in most other carcinomas. It is proposed that use is made in diagnostic examinations of vimentin antisera or affinity-purified vimentin antibodies that have been pre-absorbed on cytokeratin protein, in order to eliminate such disturbing reactions.

Detection of a cytokeratin determinant common to diverse epithelial cells by a broadly cross-reacting monoclonal antibody

A monoclonal antibody derived from a mouse immunized with bovine epidermal prekeratin has been characterized by its binding to cytoskeletal polypeptides separated by one- or two-dimensional gel electrophoresis and by immunofluorescence microscopy. This antibody (KG 8.13) binds to a determinant present in a large number of human cytokeratin polypeptides, notably some polypeptides (Nos. 1, 5, 6, 7, and 8) of the 'basic cytokeratin subfamily' defined by peptide mapping, as well as a few acidic cytokeratins such as the epidermis-specific cytokeratins Nos. 10 and 11 and the more widespread cytokeratin No. 18. This antibody reacts specifically with a wide variety of epithelial tissues and cultured epithelial cells, in agreement with previous findings that at least one polypeptide of the basic cytokeratin subfamily is present in all normal and neoplastic epithelial cells so far examined. The antibody also reacts with corresponding cytokeratin polypeptides in a broad range of species including man, cow, chick, and amphibia but shows only limited reactivity with only a few rodent cytokeratins. The value of this broad-range monoclonal antibody, which apparently recognizes a stable cytokeratin determinant ubiquitous in human epithelia, for the immunohistochemical identification of epithelia and carcinomas is discussed.