Recently a new classification of primary cutaneous B-cell lymphomas (PCBCLs) has been proposed by the European Organization for Research and Treatment of Cancer (EORTC)--Cutaneous Lymphoma Project Group. The marginal zone B-cell lymphomas (MZLs) were not included as a distinct entity because of insufficient experience and controversial opinions. We have studied 32 patients (M:F ratio 1.5:1; age range 25-93 years; mean age 49.6 years; median age 50 years) to determine the diagnostic criteria of primary cutaneous MZL and the relationship with other low-grade malignant PCBCLs. For comparison, three patients with immunocytoma were included in the study. Clinically, patients presented with solitary or clustered reddish or red-brown papules, nodules, and plaques, sometimes surrounded by an erythematous halo. Histopathologic sections showed nodular or diffuse infiltrates involving the dermis and subcutaneous fat. Cytomorphologically small to medium-sized cells with indented nuclei and abundant pale cytoplasm (marginal zone cells, centrocyte-like cells) predominated. In addition, scattered blasts, lymphoplasmacytoid cells, and plasma cells were observed below the epidermis and at the periphery of the infiltrates. Reactive germinal centers were present in 75% of the cases. The three cases of immunocytoma showed a more monomorphous pattern with predominance of lymphoplasmacytoid cells. The marginal zone cells showed a CD20+, CD79a+, CD5- and Bcl-2+ immunophenotype. They expressed immunoglobulin G in the majority of the cases. Staining with the monocytoid B cell-related antibody KiM1p gave positive results in all specimens with a typical intracytoplasmic granular pattern. A monoclonal distribution of immunoglobulin light chains was observed in marginal zone cells in 75% of the cases. Germinal centers, when present, were either polyclonal or negative for both kappa and lambda light chains. Monoclonal rearrangement of the JH gene was detected via polymerase chain reaction (PCR) in 18 of 26 investigated specimens. Analysis in 12 patients of the bcl-2/immunoglobulin heavy chain gene rearrangement using PCR yielded negative results. Lesions were treated by surgical excision followed in some patients by local radiotherapy. Systemic antibiotic therapy was administered to three patients, with good response in two. The prognosis is excellent. After a mean follow-up of 47.9 months (range 6-252; median 24) all patients are alive without signs of systemic lymphoma. Primary cutaneous MZL represents a distinct clinicopathologic subtype of low-grade malignant PCBCL.

Mallory bodies are characteristic morphological features of alcoholic hepatitis in man and can be produced in the mouse by chronic griseofulvin intoxication. The appearance of Mallory bodies in hepatocytes is associated with derangement of the cytokeratin intermediate filament cytoskeleton, at least as revealed by immunofluorescence and suggested by immunoelectron microscopy. Immunohistochemical studies were performed to answer the question whether Mallory body formation and cytoskeleton alterations finally lead to cell death or are reversible phenomena. Chronically griseofulvin-intoxicated mice killed at different stages of recovery on a normal diet served as experimental animals. It could be shown that (a) Mallory bodies are very durable structures and are found for up to 6 mo after griseofulvin withdrawal as a result of persistence and neoformation; (b) new Mallory bodies can appear even several months after cessation of griseofulvin feeding; (c) Mallory body formation and cytoskeletal changes by themselves do not lead to irreversible cell damage; (d) the cytoskeletal changes are reversible within 7 mo after griseofulvin withdrawal; (e) a dissociation between disappearance of Mallory bodies and restoration of a regularly immunostained cytoplasmic cytokeratin meshwork is observed.

To identify Mallory body (MB) constituents, monoclonal antibodies to murine MBs induced by long-term griseofulvin (GF) feeding were produced. One of these, antibody MM 120-1, specifically reacted in immunofluorescence microscopy with MBs in all developmental stages but not with other cell structures of human and mouse liver and other organs. The MM 120-1 antigen was present in murine MBs induced by griseofulvin or 3,5-diethoxycarbonyl-1,4-dihydrocollidine feeding and also in human MBs in livers of patients with alcoholic hepatitis. In immunoblots, the MM 120-1 antigen was detectable in the high molecular weight fraction of MB preparations, most of which remained in the well and at the interphase between stacking and resolving gel. No reactivity with cytokeratin polypeptides of different conformational states (i.e., isolated cytoker atin components A and D, heterotetramers A2D2, reconstituted intermediate filaments) was found. It is concluded that the antibody MM 120-1 is a highly specific and sensitive marker for murine and human MBs recognizing a high molecular weight nonkeratin component. This component could play a central role in the pathogenesis of MBs.

42 renal cell carcinomas and 1 oncocytoma were investigated by means of immunofluorescence (including double immunofluorescence) using a panel of mono- and polyclonal antibodies to vimentin and cytokeratins. In all tumors except chromophobe cell renal carcinoma (CCRC) and oncocytoma generally a coexpression of vimentin and cytokeratins could be demonstrated; however, the intermediate filament expression was often very heterogeneous with regard to the distribution of vimentin and cytokeratins in general, depending on the mono- and polyclonality of the antibodies and on the areas of a tumor investigated. In CCRC and oncocytoma all tumor cells contained cytokeratin filaments. In addition, as revealed by double immunofluorescence, in only occasional tumor cells we could demonstrate vimentin.

To examine whether hepatocellular cytokeratins can serve as substrates of transglutaminases (TG) TG-catalyzed incorporation of [3H]putrescine into, as well as cross-linking of, cytokeratins was studied. Purified guinea pig liver TG and mouse liver TG present in 105,000 x g supernatants were used as enzymes. Isolated mouse liver cytokeratin filaments, heterotypic tetramers (A2D2), as well as cytokeratin filaments reconstituted from isolated and column-purified liver cytokeratin polypeptides A and D served as substrates. Moreover, to more closely mimic the situation within the cell, mouse liver homogenate containing cytokeratins was also used. Cross-linked proteins were identified as cytokeratins by immunoblotting after sodium dodecyl sulfate-polyacrylamide gel electrophoresis using cytokeratin antibodies. The results indicated that mouse liver cytokeratins can serve as substrates of homologous and heterologous TG. However, liver cytokeratin components A and D differed in this respect. Depending on the experimental conditions either components D or components A were better substrates of TG-mediated cross-linking as revealed by increased high molecular weight aggregates, which failed to enter the gels, concomitant with a decrease of the monomer band suggesting a more intimate relationship between homologous cytokeratin polypeptides within the filament. The results presented provide the basis for studies of TG-induced cross-linking of cytoskeletal components in hepatocytes that may occur during liver cell injury associated with increased intracellular Ca2+ concentrations.

Freshly isolated mouse hepatocytes were tested with respect to the induction of heat shock (stress) proteins by elevated temperature, sodium arsenite and ethanol treatment. With heat, arsenite and ethanol treatments, the synthesis of a protein with a molecular weight of 68 kD (heat shock protein 68) was predominantly elevated; arsenite and ethanol exerted their effects on heat shock protein synthesis in a dose-dependent manner. Hepatocytes derived from livers of chronically griseofulvin-pretreated mice differed in their response from normal hepatocytes in that ethanol was ineffective in these cells. These results indicate that different modes and pathways of the stress response exist, depending on the nature of the inducing agent but also on pretreatment conditions. In vivo, pathologic alterations of cells and organs (e.g., in the course of chronic diseases) can, therefore, be expected to modulate the stress response.

The turnover of cytokeratin polypeptides A (equivalent to No. 8 of the human cytokeratin catalog) and D (equivalent to human cytokeratin No. 18) of mouse hepatocytes was studied by pulse-labeling of mouse liver proteins after intraperitoneal injection of l-[guanido-14C]arginine and [14C]sodium bicarbonate. At various times after injection cytoskeletal proteins were prepared and separated by SDS-polyacrylamide gel electrophoresis, and the specific radioactivities of polypeptides recovered from excised gel slices were determined. With l-[guanido-14C]arginine a rapid increase in the specific radioactivity of both cytokeratins was observed which reached a plateau between 12 and 24 h. With [14C]sodium bicarbonate maximal specific radioactivity was obtained at 6 h followed by a rapid decrease to half maximum values within the subsequent 6 h and then a slower decrease. Half-lives were determined from the decrease of specific radioactivities after pulse-labeling by least-squares plots and found to be 84 h (for cytokeratin component A) and 104 h (component D) for arginine labeling. Values obtained after bicarbonate labeling were similar (95 h for A and 98 h for D). These results show that liver cytokeratins are relatively stable proteins and suggest that components A and D are synthesized and degraded at similar rates, probably in a coordinate way.

Seminomas and non-seminomatous testicular germ cell tumours were studied for the presence of cytokeratin and vimentin filaments and desmosomes using immunohistochemical methods. In the majority of the classical seminomas and in seminomatous areas of mixed tumours most tumour cells appeared to lack cytokeratin filaments. Some seminomas contained a focally variable proportion of cells exhibiting cytokeratin-positive structures while other cases contained only few seminoma cells with a well developed fibrillar cytokeratin network. Gel electrophoresis of cytoskeletal proteins from microdissected regions revealed cytokeratin polypeptides nos. 8 and 18 typical of simple epithelia. In one seminoma, however, all, or almost all, tumour cells contained cytokeratin filaments. This finding is in line with the assumption of transitional forms between seminoma and embryonal carcinoma. Despite the lack - or variable expression - of cytokeratin filaments most seminoma cells contained desmosomes, although often few in number and irregularly distributed at the circumference of the cells. Loosely arranged and often very sparse vimentin fibrils were found in many, but not all seminoma cells. Double label immunofluorescence microscopy suggested that the majority of desmosomes was associated with intermediate filaments of the vimentin type. In contrast, in carcinoma cells of malignant teratomas, in well differentiated epithelial cells of intermediate-type malignant teratomas and in trophoblastic cells present in trophoblastic-type malignant teratomas cytokeratin filament bundles as well as desmosomes were decorated. The arrangement and density of the cytokeratin filament skeleton and of desmosomes varied with degree of maturation of the tissue. The most regular distribution and intensive staining of cytokeratin filaments and desmoplakin was found in “mature” tissues. Vimentin was demonstrated in mesenchymal areas and stroma cells. The results show that seminomas are distinguished from most other germ cell and non-germ cell tumours by the presence of true desmosomes together with scanty vimentin filaments in most tumour cells. In addition, they indicate that seminoma cells can be heterogenous in their cytoskeletal complement and may include cells with cytokeratin expression, indicative of a multi-potential character of the initially transformed cell(s).

The presence and distribution of intermediate filament proteins, such as cytokeratins, vimentin, neurofilament proteins and glial fibrillary acidic protein were assessed immunohistochemically in pituitary adenomas, medullary thyroid carcinomas, endocrine pancreatic tumours, gastric, intestinal and bronchial carcinoids, parathyroid adenomas, pheochromocytomas, paragangliomas and related non-neoplastic tissues. In some cases, immunohistochemical results were correlated with cytoskeletal proteins as analysed by SDS-polyacrylamide gel electrophoresis. Cytokeratin antibodies with broad range of immunoreactivity (i.e. to murine liver cytokeratin component D) reacted with epithelial cells in all non-neoplastic endocrine tissues and related neuroendocrine tumours studied, except for adrenal medulla, pheochromocytoma and paraganglioma, independently of hormone production and biological behaviour. In contrast, antibodies to epidermis-derived cytokeratins failed to stain endocrine tissues and tumours. Paranuclear cytokeratin accumulations were seen in bronchial, gastric, and intestinal carcinoids and seem to be a common feature of neuroendocrine tumours. One- and two-dimensional SDS-polyacrylamide gel electrophoresis of non-neoplastic endocrine tissues and related tumours revealed two major keratin polypeptides corresponding to cytokeratins No. 8 and 18 of the cytokeratin catalog of human cells (Moll et al. 1982). According to this cytokeratin polypeptide composition, endocrine tissues and related tumours conform to the "simple type" of epithelia. Vimentin-related immunoreactivity was restricted to stromal cells and to folliculo-stellate cells in normal pituitary gland, Schwann cells in carcinoids and satellite cells in normal adrenal medulla and in pheochromocytomas. Neurofilament protein- (70 kD)-antibodies only stained nerve fibers in normal tissues and at the periphery of carcinoid tumour cell complexes, and to a variable degree, cells in nontumorous adrenal medulla, pheochromocytomas and paragangliomas. Furthermore, neurofilament reactivity was observed along with cytokeratin expression in two bronchial carcinoids.

A monoclonal murine antibody (KM 54-5) was produced against Mallory body (MB) material isolated from liver tissue of griseofulvin treated mice. The antigen was identified by positive immunofluroescence microscopy of MBs and by the immunoblotting technique on polypeptides separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis. In immunoblotting experiments, antibody KM 54-5 reacted with cytokeratins A (human no. 8) and D (human no. 18) of murine, bovine, and human hepatocytes as well as with cytokeratin A (no. 8) and its degradation products present in isolated murine MB. In immunofluorescence microscopy the antibody did not react with cytokeratin filaments of normal liver but showed a positive reaction with MBs after a certain stage in MB development had been reached. In a dot blot assay, using individual cytokeratin polypeptides isolated from murine liver and purified by ion exchange chromatography in pH 8 buffer containing 8 M urea, the antibody reacted with the individual polypeptides A (no. 8) and D (no. 18) but not with the heterotypic tetramer (A2D2) reconstituted from these polypeptides in 4 M urea. These findings confirm the cytokeratin nature of MB filaments. In addition, they show that the pathologic process of MB formation involves changes in cytokeratin organization and conformation, resulting in the accessibility of a specific antigenic determinant which is inaccessible ("masked") in the heterotypic tetramer subunit and in the cytokeratin filaments of normal cells. Hence this study presents an example of a pathological change of cytokeratin filaments and illustrates the value of monoclonal antibodies in the detection of such changes.

The presence and distribution of tissue polypeptide antigen (TPA) were assessed in gastrointestinal carcinomas of different origin, morphology and degree of differentiation. Immunocytochemistry was employed, using the PAP technique on formalin-fixed, paraffin-embedded material and compared with the results obtained with antibodies to cytokeratins. Like cytokeratins, TPA was a reliable marker of epithelial differentiation and showed tissue distribution patterns similar to cytokeratins, as revealed by antibodies with broad-range cytokeratin immunoreactivity. In most carcinomas, TPA-specific immunostaining was less intense than in non-neoplastic tissue. No direct relationship between intensity of TPA staining and morphological degree of differentiation and proliferation was found. TPA staining was most pronounced at the periphery of the cells. In stratified epithelium, i.e. oesophageal mucosa, basally located cells exceeded superficial cells in TPA immunoreactivity in contrast to the cytokeratin antibodies which decorated the more superficially placed cell layers. TPA and cytokeratin staining patterns were similar in neoplastic and non-neoplastic gastric, intestinal mucosa, as well as in biliary tract epithelium. Antral and cardial mucoid glands of the stomach as well as gastric carcinomas of the pylorocardial type remained unstained with both types of antibodies. Similar staining with TPA and cytokeratin antibodies was also observed in pancreatic and liver tissue. In this study, hepatocytes were, although weakly, stained by TPA antibodies and an identical staining was found with benign and malignant hepatocellular neoplasms. Ductal and ductular TPA-staining was most conspicuous and so was the immunoreactivity of cholangiocellular carcinomas. A comparison between TPA and cytokeratins was also made by immunoblotting which revealed immunoreactivity of antibodies to TPA with cytokeratin polypeptides of different species (man, mouse) and organs (epidermis, liver), particularly with the cytokeratin component 8 of human liver and the related component A of mouse liver. The significance of this finding is uncertain until the pertinent epitopes have been revealed by monoclonal mapping of the components which exhibit similar molecular weights by SDS polyacrylamide gel electrophoresis.

The presence and distribution of cytokeratins, neurofilament proteins, vimentin and glial fibrillary acidic protein were studied in 10 cutaneous neuroendocrine carcinomas (CNEC) by immunohistochemical techniques, using specific antibodies. In all cases tumour cells were specifically stained with antibodies to mouse liver cytokeratin component D in paraffin-embedded formalin-fixed and frozen sections. Moreover, one- and two-dimensional SDS-polyacrylamide gel electrophoresis of high salt/detergent resistant cytoskeletal residues from tumour material, isolated by microdissection from frozen sections, revealed the presence of cytokeratin components 8 and 18 which are characteristic constitutents of cytokeratin filaments of simple epithalia. Neurofilament proteins were detected by immunocytochemistry in tumour cells from 2 patients, from whom frozen material was available, and their presence was also positively identified in cytoskeletal residues by immunoblotting using specific antibodies. Glial fibrillary acidic protein and vimentin could not be demonstrated in tumour cells. Our studies did not confirm the suggested origin of CNEC from epidermal Merkel cells.

Mallory bodies are a morphological key feature of severe alcoholic liver cell injury (alcoholic hepatitis) and the morphological expression of dysregulation and derangement of the intermediate filament cytoskeleton of the hepatocyte. Their pathogenesis is still unclear. Studies on Mallory body formation may not only help to elucidate the mechanisms of liver cell injury associated with alcoholic hepatitis, but may also contribute to our understanding of the regulation and function of the intermediate filament cytoskeleton.

ABSTRACT— Mallory bodies induced by long‐term griseofulvin feeding in mouse liver were isolated and analyzed by one‐ and two‐dimensional gel electrophoresis and reaction of the separated polypeptides with cytokeratin antibodies using the blotting technique. Comparison with normal intermediate filament cytoskeletons from mouse hepatocytes revealed that Mallory bodies contain two polypeptides: Component II (Mr: 55,000; apparent isoelectric point values: 6.45, 6.1, 5.9) and component III (Mr: 48,000; apparent isoelectric point values: 5.7, 5.5, 5.43, 5.38, 5.2) which appear to be similar, if not identical, to liver cytokeratins A and D, respectively. By contrast, component I of Mallory bodies (Mr: 65,000; apparent isoelectric point values: 5.4, 5.38, 5.2) was not found in appreciable amounts in normal hepatocytes. Component II was positive in immunoreaction with antibodies to murine hepatocyte keratins A and D as well as epidermal prekeratin. Component III showed reaction with the antibodies to murine hepatocyte keratins A and D but not with those raised against epidermal prekeratins. By contrast, the unusual component I reacted with antibodies to murine hepatocyte keratin D and to epidermal prekeratins. The results prove that cytokeratin polypeptides are major constituents of Mallory bodies and suggest that the pattern of liver cytokeratin polypeptides is altered during the toxic treatment and/or Mallory body formation.