FREEZE-FRACTURE STUDY OF THE OCCURRENCE OF PLASMA MEMBRANE DIFFERENTIATIONS IN HUMAN BASAL CELL CARCINOMA

Freeze-fracture electron microscope technique was used to study the plasma membrane ultrastructure of basal cell carcinoma (BCC) cells biopsied from two patients. Freeze-fracture replicas of the plasma membrane of BCC cells indicated the presence of gap junctions, tight junctions and desmosomes. Tight junctions were found on almost all fracture faces of BCC cell plasma membranes, while they were seldom found in the normal keratinocytes or basal cells. These tight junctions appeared as an incomplete network of anastomosing strands with many free-ending spurs. The frequency of gap junctions seemed to increase in number but decrease in size as compared with normal keratinocytes. Many of the gap junctions were associated with tight junctions. Desmosomes varied in size from 0.1 to 1 µm and were found rather frequently. The results of this study indicated that these junctional complexes were very well developed and that these well differentiated junctions may explain the BCC cell's disinclination to metastasize.     

Ultrastructural localization of gap junction protein connexin 43 in normal human skin, basal cell carcinoma, and squamous cell carcinoma

The expression and localization of connexin 43 (Cx43) were investigated in normal human epidermis, pilosebaceous apparatus, basal cell carcinoma, and squamous cell carcinoma by immunofluorescence as well as by immunoelectron microscopy. In the normal epidermis the immunofluorescence was weak in the basal layer, increased in spinous layer and negative in the horny layer. In the sebaceous gland, peripheral lobular cells showed weak cell membrane dotted pattern. Cell membrane and cytoplasmic fluorescence was strong in the central lobular cells. In the lower hair follicle, the cortex, inner and outer root sheath cells showed cell membrane fluorecence. As cortical cells underwent keratinization, they lost Cx43 epitopes. Basal cell carcinoma and squamous cell carcinoma were poorly stained, and eccrine and apocrine glands were unstained. In immunoelectron microscopy, close membrane appositions of typical gap junctions were often observed in the spinous layers of the epidermis and the immunolabeling for Cx43 was seen along the gap junction structures. Circular and long gap junctions often were found in follicular root sheaths and sebaceous glands. Gold particles labeling Cx43 in these gap junctions were found on the gap junctions or localized in the cytoplasm near the gap junction membranes. Basal cell carcinoma and squamous cell carcinoma had a small number of small gap junctions, and gold particles were not only localized to gap junctions but scattered in the cytoplasm. No gap junctions were labeled in eccrine and apocrine glands. These findings confirmed that 1) long, curved or circular membrane appositions found in hair follicle and sebaceous gland are true gap junctions, 2) immature cells such as epidermal basal cells, peripheral germinative cells of sebaceous gland and basal and squamous cell carcinoma cells do not have fully developed gap junctions, and 3) Cx43 or its precursors are present in the cytoplasm as well as on poorly developed gap junctions in these immature cells. Immunofluorescence findings generally corresponded to ultrastructural distribution and structural maturity of gap junctions.     

Freeze-fracture cytochemical study of membrane systems in human epidermis using filipin as a probe for cholesterol

Filipin (a polyene antibiotic) interacts specifically with cholesterol in membranes, producing characteristic 25 nm-diameter deformation (pitlike lesions) within the membrane plane detectable by freeze-fracture electron microscopy. Utilizing this probe, the distribution of cholesterol molecules in membranes and in lamellar structures between horny cells was investigated in human skin. The plasma membranes of basal, spinous, and granular cells reacted extensively with filipin except for desmosomal membrane portions. However, the plasma membranes of horny cells were rarely labeled with filipin, while lamellar structures between horny cells were well labeled. These observations indicate the distinct difference in susceptibility to filipin among the plasma membranes of viable cells and horny cells, and the lipid lamellar structures. Whenever horny cell plasma membranes were affected with filipin, they revealed a low deformability showing shallow pits or low protrusions. This low deformability may be due to greater membrane rigidity rather than a lower content of cholesterol, although the possibility of a low amount of cholesterol cannot be excluded. Lamellar bodies in granular cells were well labeled in the limiting membranes but poorly labeled in the internal lamellar structures. The regions of gap junctions were absolutely unlabeled. Filipin-cholesterol complexes were produced very close to the junctional strands but did not appear to disrupt the junctional structure of tight junctions. Nuclear membranes were affected only in the outer membrane with filipin. These results suggest that keratinocytes undergo a distinctive reduction in membrane deformability or in free-cholesterol content at the transition from living to dead cells, and display a heterogeneity in cholesterol distribution in human epidermal cell membranes.     

Immunomolecular mapping of adherens junction and desmosomal components in normal human epidermis

Adherens junctions (AJs) are cell-cell and cell-matrix junctions that are known to comprise the transmembrane and cytoplasmic components linked to the f-actin cytoskeleton. Although the presence of AJs han been confirmed in normal human epidermis, previous studies immunolocalizing AJ-related antigens have been controversial. The purpose of this study was to produce a more precise molecular mapping of AJs and their constituents in relation to desmosomes in normal human epidermal keratinocytes. Using an electron microscope (EM) method to optimally fix plasma membranes. AJ structures were typically seen as a narrowing of the intercellular space between two keratinocytes that was distinct from desmosomes and gap junctions. Such structures were consistently found more frequently in the upper epidermis than in the basal layer. Immunogold electron microscopy showed an absence of the AJ components (E-cadherin and beta-catenin) from desmosomal areas but they were present at interdesmosomal areas at sites of close membrane association. Conversely, the desmosomal components plakoglobin and plakophilin 1 were restricted only to the outer attachment plaque of the desmosome. These results further confirm that AJs have a distinct molecular composition and distribution from desmosomes and that they regularly occur between desmosomes along the keratinocyte plasma membrane to provide alternative cell-cell adhesion mechanisms.     

Lectin-binding pattern in extramammary Paget's disease by horseradish peroxidase (HRP)-labeling method--specific staining with Dolichos biflorus agglutinin (DBA)

Lectin-binding pattern in extramammary Paget's disease was studied using seven different lectins (Con A, WGA, RCA-I, PNA, SBA, DBA, and UEA-I) by means of the horseradish peroxidase (HRP)-labeling method. By light microscopy it was observed that Con A, WGA, RCA-I, and DBA stained almost all the extramammary Paget cells, while PNA, SBA, and UEA-I stained only some of them. Normal keratinocytes and tumor cells from other diseases such as mammary Paget's disease, malignant melanoma, squamous cell carcinoma, basal cell epithelioma, Bowen's disease, and seborrheic keratosis were positively stained with Con A, WGA, and RCA-I, but not with DBA except in some of the mammary Paget's cells. By electron microscopy it was observed that DBA stained the cell membrane and the Golgi apparatus of the extramammary Paget cells. The present results suggest that DBA is a specific lectin for glycoconjugates in extramammary Paget cells.     

Mammary and Extramammary Paget''s Disease

Two cases of mammary and one case of vulval Paget's disease are studied histologically and electronmicroscopically. There are no substantial morphological differences between those two types. The Paget cell has a characteristic fine structure: it lacks tonofibrils and shows many of the features of apocrine cells. Less highly differentiated keratinocytes lie adjacent to Paget cells and desmosomes exist between them. These observations suggest that Paget cells differentiate in situ from epidermal cells.     

Mammary and extramammary Paget's disease: expression of Ca 15-3, Ka-93, Ca 19-9 and CD44 in Paget cells and adjacent normal skin

The histogenesis of mammary Paget's disease (MPD) and extramammary Paget's disease (EMPD) cells remains controversial. The purpose of this study was to investigate MPD and EMPD immunohistochemically with antibodies to some tumour markers (Ca 15-3, KA-93 and Ca 19-9), and a cell surface receptor for hyaluronate (CD44), as these have been shown to be expressed in normal eccrine or apocrine glands and/or the epidermis, as well as some tumours. Surgically excised, formalin-fixed, paraffin-embedded tissues, or frozen tissues, from seven mammary, five vulvar, two scrolal and two axillary lesions were studied. Paget cells stained strongly with antibodies to Ca 15-3 and KA-93, but did not stain with those to Ca 19-9 and CD44. Staining with the antibody to Ca 15-3 was also observed in the ductal and secretory portions of the eccrine and apocrine glands, and in the sebaceous gland cells. Staining with the antibody to KA-93 was also seen in the apocrine secretory coils, lactiferous duct, epidermal dendritic cells, and cells in the dermal inflammatory infiltrate. Staining with the antibody to Ca 19-9 was observed only in the eccrine duct, and that to CD44 was seen in eccrine secretory cells and epidermal keratinocytes. These findings suggest that the origin of Paget cells may be the secretory cells of apocrine sweat glands (in EMPD) or the luminal lactiferous ducts (in MPD). We found that the antibodies to Ca 15-3 and CD44 were useful in differentiating Paget cells from surrounding keratinocytes, by showing positive and negative immunoreactivity, respectively.     

Immunobead and Density Gradient Purification of Paget Cells: In vitro Studies of Proliferation

Previous studies using primary monolayer cultures of epithelial cells from the involved epidermis of patients with mammary and extramammary Paget's disease investigated whether Paget cells proliferate as other malignant cells do. Although epithelial monolayers from the involved skin were maintained for approximately 45 days, no permanent cell lines were established. The proportion of carcinoembryonic antigen (CEA)-positive cells did not increase in the long-term cultures. Herein, we report studies of whether there is a real reduction of Paget cell numbers or if this is merely a decrease in the expression of CEA by the cells. Furthermore, we investigated whether Paget cells survive longer when cultured free from any potential inhibitory keratinocytes or other epidermal cells. Skin samples were obtained from one patient with mammary Paget's disease and three with extramammary Paget's disease; epidermal cells were cultured in vitro. An enrichment of Paget cells was carried out from the cultured epidermal cells by combining an anti-epithelial membrane antigen monoclonal antibody, binding to immunobeads, and density gradient centrifugation in Nycodenz. The separated cells were re-cultured in Keratinocyte-SFM serum-free media. The proportion of CEA-positive cells did not increase in the cultures, and the purified cells did not show any increase in survival times compared to the non-purified cultured cells. These results suggest that the decrease of CEA-positive cells noted during culture results from a decline in expression of CEA in the Paget cells. Paget cells in the involved epidermis do not proliferate significantly and thus differ from many other malignant cells.     

Immunologic characteristics of keratins in extramammary Paget's disease

Six cases of extramammary Paget's disease were immunohistochemically investigated with several antikeratin monoclonal antibodies. Paget cells and surrounding epidermal keratinocytes were equally stained with an antikeratin monoclonal antibody, HKN-4, which recognizes a broad spectrum of keratins. However, Paget cells were clearly distinguished from the surrounding epidermal keratinocytes by HKN-2, which does not react with keratins of secretory cells but does react with keratins of ductal and myoepithelial cells of sweat glands and with epidermis and hair tissue of the normal skin. The HKN-2 did not bind to Paget cells, but the surrounding keratinocytes were positive. CK7, LE41, RGE53, and LP2K, which recognize simple epithelium-type keratins 7 (molecular weight [MW], 54,000; type II), 8 (MW, 52,500; type II), 18 (MW, 45,000; type I), and 19 (MW, 40,000; type I), respectively, stained Paget cells but not the surrounding keratinocytes. Two cases of Merkel cell carcinoma, examined as controls, showed positivity to LE41 and RGE53 but not to CK7 and LP2K. Since in the normal skin the secretory cells of sweat glands showed the same keratin expression as that of Paget cells, Paget cells of extramammary Paget's disease may be derived from or differentiate to the secretory cells of sweat glands.     

Qualitative and quantitative studies of Paget cells in the horny layer

One case of extramammary Paget's disease was studied by ultrastructural and morphometrical methods to clarify whether or not Paget cells change qualitatively or quantitatively in passing through the epidermis. Paget cells in living layers were observed as large cells with abundant clear cytoplasm, large nuclei, and prominent nucleoli. The cytoplasm was enriched with well developed rough-surfaced endoplasmic reticulum, mitochondria, free ribosomes, and glycogen particles. In contrast, filamentous structures were not developed. Paget cells are connected to each other and adjacent keratinocytes by small desmosomes. In contrast, in the horny layer, Paget cells contained shrunken, flattened nuclei. Cell organelles had degenerated and large lipid droplets were present in the cytoplasm. As disruption of the desmosomes took place, intercellular spaces between Paget cells and keratinocytes widened. Based on the morphometrical analysis by a sonic digitizer computer system, there were statistically significant decreases in areas and diameters of Paget cells between the horny layer and the living cell layers, including the basal, spinous and granular layers. In conclusion, it is suggested that Paget cells change qualitatively and quantitatively after they move into the horny layer from the living layers.     

Lectin-binding pattern in extramammary Paget's disease by horseradish peroxidase (HRP)-labeling method —Specific staining with Dolichos biflorus agglutinin (DBA)

Summary Lectin-binding pattern in extramammary Paget's disease was studied using seven different lectins (Con A, WGA, RCA-I, PNA, SBA, DBA, and UEA-I) by means of the horseradish peroxidase (HRP)-labeling method. By light microscopy it was observed that Con A, WGA, RCA-I, and DBA stained almost all the extramammary Paget cells, while PNA, SBA, and UEA-I stained only some of them. Normal keratinocytes and tumor cells from other diseases such as mammary Paget's disease, malignant melanoma, squamous cell carcinoma, basal cell epithelioma, Bowen's disease, and seborrheic keratosis were positively stained with Con A, WGA, and RCA-I, but not with DBA except in some of the mammary Paget's cells. By electron microscopy it was observed that DBA stained the cell membrane and the Golgi apparatus of the extramammary Paget cells. The present results suggest that DBA is a specific lectin for glycoconjugates in extramammary Paget cells.Supported in part by grants-in-aid for scientific research from the Ministry of Education, Science and Culture of Japan, and Japan Private School Promotion Foundations     

The expression of the gap junctional protein Cx43 is restricted to proliferating and non differentiated normal and transformed keratinocytes

Abstract The passage of specific growth modulating signals through gap junctions may regulate the proliferation and differentiation of human keratinocytes. To investigate this, we correlated the proliferation of normal human keratinocytes and a transformed squamous cell carcinoma cell line, SCC4. with the expression of the gap junctional proteins Cx43. 31 and 31.1, known to be expressed by keratinocytes. Proliferation was con-lined to preconfiuent and confluent cultures of normal keratinocytes, falling to undelectable levels once poslconfluency was achieved. Cx43. at both the message and protein levels, paralleled these changes, being elevated predominantly in prcconiluent and confluent cultures, and downregulated in postconfluency. Similar results were found for Cx31 and 31.1 at the message level. In contrast, the proliferation of SCC4 cells cultured in media supplemented with 5.0° FCS was maintained at n substantial level from preconfluency through 2 weeks postconfluency. Cx43, 31. and 31.1 RNA and Cx43 protein expression mirrored the levels of proliferation within SCC4 cultures. Cx26 and 32 were not found in normal keratinocytes or SCC4 cells at any stage of differentiation. These data, illustrating a tight correlation between proliferation and Cx43, 31 and 31.1 expression, suggest that these connexins may represent proliferation-specific gap junctions within keratinocytes. and may therefore transmit signals that control keratinocyle division.     

The specialized junctions between Merkel cell and neurite: an electron microscopic study.

Longitudinal serial sections of one half of the entire sinus hair of a mouse were examined by the electron microscope. Three neurites entering the outer root sheath from the perifollicular blood sinus were encountered. These were separate nerve trunks from those connected with perifollicular tactile nerve endings and exclusively innervated intrafollicular Merkel cells. Two types of specialized junctions were observed at the contact regions between Merkel cell plasma membrane and neurite plasma membrane: (i) desmosome-like structures in which small clear vesicles and/or the large cored vesicles of neurite and thicker membrane (post-synaptic?) of apposed Merkel cell were found ant (ii) synapse-like structures in which Merkel cell granules were concentrated near the plasma membrane and the membrane of the apposed neurite was usually thicker (post-synaptic?). In some of the synapse-like junctions the limiting membrane of Merkel cell granules fused with the Merkel cell plasma membrane and its content seemed to be discharged into the intercellular space. This suggested actual exocytotic secretion of Merkel cell granules. Juxtaposition of 2 types of junctions, i.e. (i) and (ii) above, was also found. This suggested the possibility that the reciprocal synapse would be present between Merkel cells and afferent neurites.     

Extramammary Paget's disease with superimposed herpes simplex virus infection: immunohistochemical comparison with cases of the two respective diseases

We describe an extremely rare case of genital Paget's disease with superimposed herpes simplex virus (HSV) infection. We also describe immunohistochemical comparison of this lesion with 19 cases of genital Paget's disease and 12 cases of skin lesions caused by HSV or varicella-zoster virus. The Paget cells expressed simple epithelial keratins (CK7 and CK19) and carcinoembryonic antigen (CEA), but did not express stratified epithelial keratins (CK1, CK2e, CK10, CK5/8, CK14). Conversely, the virus-infected keratinocytes were positive for stratified epithelial keratins but negative for simple epithelial keratins and CEA. In the present case, simple epithelial keratins, stratified epithelial keratins, CEA and HSV were heterogeneously expressed in the ballooning and multinucleated giant cells. These results suggest that these cells were derived from keratinocytes and Paget cells and that the production of many multinucleated giant cells resulted from the virus-mediated cell fusion between Paget cells and neighbouring keratinocytes.     

Plasma membrane fluidity of keratinocytes of normal and psoriatic skin: a study using fluorescence anisotropy of trimethylammoniumdiphenylhexatriene (TMA-DPH)

The aim of this study was to investigate plasma membrane fluidity in human keratinocytes using fluorescence anisotropy of 1,6-diphenyl-1,3,5-hexatriene (DPH) and its cationic derivative 1-[4-(trimethylamino)-phenyl]-6-phenyl-1,3,5-hexatriene (TMA-DPH). Keratinocytes from normal or psoriatic skin were isolated using trypsin-EDTA or dispase. In keratinocytes isolated from normal skin, TMA-DPH anisotropy values were higher than those observed using DPH; the difference must be related to the different localization of the two probes. In fact, DPH in whole cells localizes in plasma as well as intracellular membranes, yielding an average value of fluidity, while the cationic derivative TMA-DPH resides in the plasma membrane of the whole cells for a sufficient time for anisotropy measurements. Moreover, it has to be considered that plasma membrane is more ordered than intracellular membranes. The kinetics of incorporation of TMA-DPH was similar in keratinocytes isolated using trypsin-EDTA and those isolated using; dispase, however, the fluorescence anisotropy values were lower in keratinocytes isolated with dispase (0.260 ± 0.01 vs 0.270 ± 0.01, p = 0.029). This difference is probably related to modifications of lipid-protein interactions after trypsin treatment. Since no damage to plasma membrane after incubation with dispase seems to have been reported, we decided to use this separation procedure to study plasma membrane fluidity in psoriasis, a human pathological condition characterized by excessive cell proliferation and incomplete differentiation. Lower anisotropy values (0.260 ± 0.01 vs 0.270 ± 0.01, p = 0.001), indicating an increase in fluidity, were observed in keratinocytes isolated from skin of psoriatic patients than in epidermal cells isolated from normal human skin. We suggest that the measurement of fluorescence anisotropy in living cells is a convenient and useful tool to study membrane fluidity in human keratinocytes isolated from normal and diseased skin. Its application represents a technical advance because plasma membrane fluidity can be measured using very limited amounts of tissue, as obtained from biopsies. Received: 23 March 1995     

Expression of wild-type, but not mutant, loricrin causes programmed cell death in HaCaT keratinocytes

The epidermal cornified cell envelope is a complex protein-lipid composite that replaces the plasma membrane of corneocytes and is crucial for epidermal barrier function. Loricrin is a major constituent of the epidermal cornified cell envelope, contributing approximately 70% by mass. In order to explore novel function of wild-type (WT) loricrin other than the major component of the epidermal cornified cell envelope, we transiently expressed construct encoding human WT and mutant loricrin (730insG) in HaCaT keratinocytes. HaCaT cells transfected with WT or mutant loricrin were at differentiation level. WT loricrin in the transfected cells was seen diffusely in the cytoplasm and nuclei. Positive transferase deoxytidyl uridine end labeling staining was observed in the nuclei of WT loricrin-transfected HaCaT keratinocytes. Data from the DNA fragmentation assay showed that only WT loricrin induced DNA ladders compared with that of mutant loricrin. WT loricrin-transfected HaCaT keratinocytes were susceptible to programmed cell death (PCD). Activation of caspase-14 was also seen. In contrast, PCD or activation of caspase-14 did not occur in mutant loricrin-transfected HaCaT cells. These results suggest that the expression of WT loricrin facilitates induction of PCD in HaCaT keratinocytes.     

Plasma membrane fluidity of keratinocytes of normal and psoriatic skin: A study using fluorescence anisotropy of trimethylammoniumdiphenylhexatriene (TMA-DPH)

The aim of this study was to investigate plasma membrane fluidity in human keratinocytes using fluorescence anisotropy of 1,6-diphenyl-1,3,5-hexatriene (DPH) and its cationic derivative 1-[4-(trimethylamino)-phenyl]-6-phenyl-1,3,5-hexatriene (TMA-DPH). Keratinocytes from normal or psoriatic skin were isolated using trypsin-EDTA or dispase. In keratinocytes isolated from normal skin, TMA-DPH anisotropy values were higher than those observed using DPH; the difference must be related to the different localization of the two probes. In fact, DPH in whole cells localizes in plasma as well as intracellular membranes, yielding an average value of fluidity, while the cationic derivative TMA-DPH resides in the plasma membrane of the whole cells for a sufficient time for anisotropy measurements. Moreover, it has to be considered that plasma membrane is more ordered than intracellular membranes. The kinetics of incorporation of TMA-DPH was similar in keratinocytes isolated using trypsin-EDTA and those isolated using; dispase, however, the fluorescence anisotropy values were lower in keratinocytes isolated with dispase (0.260±0.01 vs 0.270±0.01,p=0.029). This difference is probably related to modifications of lipid-protein interactions after trypsin treatment. Since no damage to plasma membrane after incubation with dispase seems to have been reported, we decided to use this separation procedure to study plasma membrane fluidity in psoriasis, a human pathological condition characterized by excessive cell proliferation and incomplete differentiation. Lower anisotropy values (0.260±0.01 vs 0.270±0.01,p=0.001), indicating an increase in fluidity, were observed in keratinocytes isolated from skin of psoriatic patients than in epidermal cells isolated from normal human skin. We suggest that the measurement of fluorescence anisotropy in living cells is a convenient and useful tool to study membrane fluidity in human keratinocytes isolated from normal and diseased skin. Its application represents a technical advance because plasma membrane fluidity can be measured using very limited amounts of tissue, as obtained from biopsies.     

A case of erythrokeratoderma variabilis: loosened gap junctions in the acanthotic epidermis

A 15-year-old Japanese female without contributory personal or family medical history had demonstrated irregular, keratotic plaques in the lower extremities since infancy that had been gradually enlarging. The keratotic plaques showed partial erythematous change, which altered shape over a relatively short period, leaving pigmentation. The biopsy specimen taken from the erythematous, keratotic plaque showed typical church-spire-like papillomatosis with acanthosis, and thickening of granular and horny layers. Gene analysis targeting connexin 30.3 and 31, based on the diagnosis of erythrokeratoderma variabilis, did not demonstrate any abnormality of these genes. However, ultrastructural observation disclosed an increased amount of gap junctions, some of which showed four layers on high-powered electron microscopy, suggesting loosened connection of the plasma membrane of the keratinocytes through the gap junctions. This loosened gap junction structure was also observed in a case of lamellar ichthyosis, examined as a reference. The disturbed cell-to-cell interaction through latent damage to the gap junctions may be related to the keratotic changes of the epidermis in these skin diseases.     

Plasma membrane ultrastructure of the human skin keratinocyte as observed by freeze-fracture electron microscopy.

The ultrastructure of the plasma membrane of human keratinocytes was investigated by freeze-fracture electron microscopy. Desmosomes were identified as an intramembranous particle aggregation on both of the fracture faces of the plasma membrane. The particle density of desmosomes ranged from 1,100 to 1,500/μ², while those of the P and E faces of the non-desmosomal membrane areas were 350 and 100/μ², respectively. The size of desmosomes varied from 0.3 to 0.7 μ in diameter. The frequency of the desmosomes was one to two per 2 μ². Gap junctions were a hexagonal array of the intramembranous particles in freeze-fracture replicas, and found much less frequently than desmosomes. On the basal cell membranes, round pits due to pinocytosis and particle aggregates of hemidesmosomes were observed.