A comparative study of the desquamation of urothelial cells during gestation and in adults mice following moderate stress or endotoxin treatment

The present paper deals with the comparative EM study of the detachment of mouse urinary bladder epithelial cells under various physiological conditions, namely, during gestation and in adult mice following induction by endotoxin or by exposure to moderate stress. It has been shown that desquamation during gestation involves two distinct modes, shedding of single cells and formation of apoptotic bodies. Moderate stress in adult female mice induced by constant illumination for 72 or 96 hours, results in desquamation of superficial and intermediate cells. Application of LPS is followed by desquamation of single cells and whole sheets of cells. Cell detachment involves interruption of tight junctions between neighbouring cells and formation of numerous cup shaped vesicles, multivesicular bodies and vacuoles at the base of desquamating cells. LPS induces desquamation entails delivery of lysosomal enzymes extracellulary into the intercellular space. In the areas of sloughing the bladder epithelium lack permeability barrier. Desquamated cells in the bladder lumen are mostly alive. These results clearly demonstrate the existence of specific adhesion mechanisms of detachment and desquamation of uroepithelial cells.     

The response of junctional complexes to induced desquamation in mouse bladder urothelium

During desquamation, the cells of mouse urinary bladder epithelium undergo detachment. In this process we examined the disconnection of cell adhesion molecules. Two proteins of cell junctions were studied: ZO1 of tight junctions and desmoplakin of desmosomes. Desquamation was induced by intravesical injection of LPS, constant illumination of mouse for 96 h, application of a combination of stress hormones hydrocortisone and norepinephrine or by removal of calcium with EGTA. All the inducers caused penetration of lanthanum tracer through the tight junctions, indicating paracellular permeability. Dilatation of extracellular spaces between neighboring cells was seen whenever desquamation was induced in bladders containing urine. Desquamation of single cells as well as groups of cells was observed. Contrary to obvious disconnection of cell junctions, as a precondition for desquamation, the distribution of junctional proteins did not change either in urothelial tissue or in desquamated cells. This study demonstrates that all the inducers of desquamation cause first an extensive dysfunction of a blood urine barrier and after that an occasional mechanical disconnection of adhesive junctions which consequently leads to desquamation.     

Rapid differentiation of superficial urothelial cells after chitosan-induced desquamation

Superficial cell desquamation followed by differentiation of newly exposed superficial cells induces regeneration of the urinary bladder epithelium, urothelium. In the present work, chitosan was evaluated as a new inducer of urothelial cell desquamation, in order to study the regeneration of mouse urothelial cells in vivo. Intravesical application of chitosan dispersion caused complete removal of only the superficial layer of cells within 20 min of treatment. Differentiation of the new superficial layer was followed by the appearance and distribution of three urothelial differentiation markers, tight junction protein ZO1, cytokeratin 20 and the maturation of the apical plasma membrane. The arrangement of ZO1 into continuous lines in individual cells of the intermediate layer was already found after 10 min of chitosan application, when desquamation had just started. The appearance of the apical membrane changed from microvillar to typically scalloped within 20 min of regeneration, while complete arrangement of the cytokeratin 20 network took 60 min. These findings provide a new perspective on the rate of the differentiation process in the urothelium and make chitosan a new and a very controllable tool for studies on urothelial regeneration.     

Succession of events in desquamation of superficial urothelial cells as a response to stress induced by prolonged constant illumination

The effect of moderate stress induced by prolonged illumination was analysed on urothelial cells of female mouse urinary bladders at ultrastructural and cytochemical levels. This study demonstrates that the urothelium responds to moderate stress with desquamation which involves two subsequent steps. The first step includes a local detachment of tight junctions and consequently the loss of the permeability barrier leading to expanded intercellular spaces among urothelial cells. During the second step, the disjunction of desmosomes accompanied by exocytosis of lysosomal enzymes (NADPase) in the intercellular space results in exfoliation of superficial cells. It is evident that moderate stress elicits an enhanced desquamation of only superficial cells by a subsequent dysfunction of first tight junctions and after that adherens-type junctions. A rapid restoration of the new tight junctions prevents a long-term malfunction of the blood-urine barrier.     

Autophagic activity in the mouse urinary bladder urothelium as a response to starvation

The urinary bladder urothelium is subjected to mechanical forces during cycles of distension and contraction, and its superficial cells are constantly flushed by toxic urine. Yet, the urothelium shows a very slow turnover of cells and superficial cells are extremely long lived. Autophagy has a well-known role in tissue homeostasis and serves as a protective mechanism against cellular stress. Therefore, the presence of autophagy as one of possible processes of survival in an unpleasant environment and during long lifetime of superficial cells was examined in mouse urothelium. We detected and evaluated autophagic activity of superficial urothelial cells under normal and stress conditions, caused by short-term starvation of newborn and 24-h-starved adult mice. Immunolabeling and Western blotting of essential effectors of autophagy, LC3 and Beclin 1, showed a weak signal in superficial urothelial cells. On the other hand, ultrastructural analysis, which proved to be the most reliable method in our study, revealed the presence of autophagic vacuoles, some of them containing specific urothelial structures, fusiform vesicles. Quantitative analysis showed increased autophagy in newborn and starved mice in comparison to a low basic level of autophagy in the urothelium of normal mice. Interestingly, some superficial cells of adults and neonates exhibit intense immunoreactions against LC3 and Beclin 1 and the typical ultrastructural characteristics of autophagy-dependent cell death. We conclude that autophagy, despite low basic activity under physiological conditions, plays an important role in urothelial homeostasis and stability under stress.     

Postnatal restoration of the mouse urinary bladder urothelium

Mouse urothelium is disrupted just before birth, followed by a postnatal restoration process which includes cell proliferation, death and differentiation. We assessed urothelial proliferation by the expression of proliferating cell nuclear antigen (PCNA), desquamation by electron microscopy, and apoptosis by TUNEL staining and urothelial differentiation by the expression of uroplakins and cytokeratin 20 (CK20) as well as the apical plasma membrane maturation. Our results indicated that urothelial proliferation was high from birth until about the 14th postnatal day. A majority of basal cells and even occasional superficial cells were PCNA positive during the first 5 postnatal days. Cell death occurred during the first 9 postnatal days. Between birth and day 5, single cells underwent apoptosis, whereas between days 6 and 9 cells mainly desquamated. CK20 and uroplakins were expressed in all superficial cells in postnatal urothelium. Their subcellular distribution characteristically changed in accordance with the progressive differentiation of superficial cells. During the urothelial postnatal development, proliferation activity slowly decreases to the proliferatively quiescent urothelium of the adult animal. Apoptosis is present in the first 9 postnatal days and within a few days of this period it appears simultaneously with desquamation. Superficial urothelial cells gradually differentiate, which is reflected in the changeable morphology of the apical plasma membrane.     

Electron tomography of fusiform vesicles and their organization in urothelial cells

The formation of fusiform vesicles (FVs) is one of the most distinctive features in the urothelium of the urinary bladder. FVs represent compartments for intracellular transport of urothelial plaques, which modulate the surface area of the superficial urothelial (umbrella) cells during the distension-contraction cycle. We have analysed the three-dimensional (3D) structure of FVs and their organization in umbrella cells of mouse urinary bladders. Compared to chemical fixation, high pressure freezing gave a new insight into the ultrastructure of urothelial cells. Electron tomography on serial sections revealed that mature FVs had a shape of flattened discs, with a diameter of up to 1.2 μm. The lumen between the two opposing asymmetrically thickened membranes was very narrow, ranging from 5 nm to 10 nm. Freeze-fracturing and immunolabelling confirmed that FVs contain two opposing urothelial plaques connected by a hinge region that made an omega shaped curvature. In the central cytoplasm, 4-15 FVs were often organized into stacks. In the subapical cytoplasm, FVs were mainly organized as individual vesicles. Distension-contraction cycles did not affect the shape of mature FVs; however, their orientation changed from parallel in distended to perpendicular in contracted bladder with respect to the apical plasma membrane. In the intermediate cells, shorter and more dilated immature FVs were present. The salient outcome from this research is the first comprehensive, high resolution 3D view of the ultrastructure of FVs and how they are organized differently depending on their location in the cytoplasm of umbrella cells. The shape of mature FVs and their organization into tightly packed stacks makes them a perfect storage compartment, which transports large amounts of urothelial plaques while occupying a small volume of umbrella cell cytoplasm.     

Superficial cell differentiation during embryonic and postnatal development of mouse urothelium

After drastic urothelial destruction around birth and around postnatal day 6, mouse urothelial renewal starts each time de novo. The differentiation of superficial cells during urothelial restoration was followed for the first time from embryonic day 15 to postnatal day 6 by the detection of differentiation markers: cytokeratins, uroplakins and apical membrane specialization. The differentiation markers of short-lived superficial cells were studied before and after urothelial destruction. Three distinctive types of superficial cells, typical for certain developmental period, were characterised: cells at low differentiation stage with microvilli and cilia, expressing CK7 and CK18, detected on embryonic day 15; cells at advanced differentiation stage with star-like arrangement of prominent membrane ridges, expressing CK7 and CK20, present between the two urothelial destruction events; highly differentiated cells with typically jagged apical surface, expressing CK7 and CK20, found twice during development. This cell type appears for the first time on embryonic day 18 as the terminal stage of embryonic differentiation. It was found again on postnatal day 6 as an initial stage of differentiation, leading toward terminally differentiated cells of the adult urothelium. Our work proves that apical membrane specialization is the most valuable differentiation marker of superficial cells.     

小鼠膀胱上皮的中间层细胞中存在Uroplakins

本文运用超薄切片、冰冻蚀刻及免疫胶体金标记等多种电镜技术并结合免疫组化、免疫荧光染色技术,直观地显示出小鼠膀胱上皮的中间层细胞存在Uroplakins,并在梭形泡膜上形成了与表层细胞类似的AUM结构,而且梭形泡的AUM结构也结合在中间纤维上。蛋白质免疫印迹反应进一步证实中间层细胞含有与表层细胞相同的Uroplakin Ⅰ和UroplakinⅢ等AUM蛋白的主要成份,从而为AUM的发生及其与细胞分化关系的研究提供了重要的实验证据。     

12-O-tetradecanoylphorbol-13-acetate induces selective desquamation of superficial cells in rat urinary bladder epithelium

Summary 12-O-tetradecanoylphorbol-13-acetate (TPA) is known to affect the proliferation and/or differentiation of several types of cells. We injected TPA directly into the lumen of rat bladder to determine, using scanning and transmission electron microscopy, its effects on the bladder epithelium in vivo. At 1 h after TPA injection (1g/ml), the superficial cells of the epithelium had changed their morphology, and large spherical vacuoles occupied their cytoplasm. In some areas, the underlying intermediate cells were exposed by the desquamation of the superficial cells. During the next few hours, TPA was excreted from the bladder lumen by voluntary micturition, but the desquamation of the superficial cells proceeded further. All the superficial cells were lost from the luminal surface by 24 h after TPA injection. The changes noted were specific for the superficial cells and were not observed in the intermediate or basal cells. After 24h, part of the epithelium had a three-layer structure, indicating that regeneration was taking place. These results demonstrate that TPA selectively affects and desquamates superficial cells in a short period of time. This experimental system may be useful for studying in vivo cell proliferation and/or differentiation.     

Cell cycle of normal bladder urothelium in developing and adult mice

The present research has employed a novel, nonradioactive technique to quantitatively study normal urothelial proliferation in foetal, neonatal, juvenile and adult mouse bladder. Using whole mount histological preparations, the total number of urothelial nuclei per mouse bladder, and per given urothelial cell layer, have been assessed to provide data of the (unstimulated) kinetic behaviour of basal urothelial cells (the proliferative population), to analyse characteristics of the normal urothelial cell cycle. The urothelial cell cycle time increases from 30.6 h (foetal) to 40 weeks (adult), the duration of mitosis from 0.23 h (foetal) to 2.71 h (adult) and the duration of DNA synthesis from 2.52 h (neonatal) to 10.83 h (adult). These are average values for the urothelial cell cycle, which do not preclude the possible existence of proliferative units. The ratio of superficial nuclei to basal and intermediate nuclei, possibly indicative of a urothelial proliferative unit, declines to reach a plateau (1:40) in adult mice. These findings indicate that rapid urothelial proliferation during early murine development was likely to be a) biologically useful, since intrauterine foetal metabolic activity may require a functional bladder urothelium at an early stage, b) kinetically similar to acutely regenerating adult urothelial cells after cytotoxic insult. During murine life, the range of durations of mitosis and DNA synthesis is much less than the range of cell cycle times. Normal unstimulated urothelium of adult mice was confirmed to proliferate slowly.     

Endocytotic activity of bladder superficial urothelial cells is inversely related to their differentiation stage

The composition of the apical plasma membrane of bladder superficial urothelial cells is dramatically modified during cell differentiation, which is accompanied by the change in the dynamics of endocytosis. We studied the expression of urothelial differentiation-related proteins uroplakins and consequently the apical plasma membrane molecular composition in relation to the membrane-bound and fluid-phase endocytosis in bladder superficial urothelial cells. By using primary urothelial cultures in the environment without mechanical stimuli, we studied the constitutive endocytosis. Four new findings emerge from our study. First, in highly differentiated superficial urothelial cells with strong uroplakin expression, the endocytosis of fluid-phase endocytotic markers was 43% lower and the endocytosis of membrane-bound markers was 86% lower compared to partially differentiated cells with weak uroplakin expression. Second, superficial urothelial cells have 5–15-times lower endocytotic activity than MDCK cells. Third, in superficial urothelial cells the membrane-bound markers are delivered to lysosomes, while fluid-phase markers are seen only in early endocytotic compartments, suggesting their kiss-and-run recycling. Finally, we provide the first evidence that in highly differentiated cells the uroplakin-positive membrane regions are excluded from internalization, suggesting that uroplakins hinder endocytosis from the apical plasma membrane in superficial urothelial cells and thus maintain optimal permeability barrier function.     

Urothelial injuries and the early wound healing response: tight junctions and urothelial cytodifferentiation

Using primary explant cultures of mouse bladder, the early response of the urothelium after superficial and full-thickness injuries was investigated. In such an in vitro wound healing model, explant surfaces with a mostly desquamated urothelial superficial layer represented superficial wounds, and the exposed lamina propria at the cut edges of the explants represented full-thickness wounds. The urothelial cell ultrastructure, the expression and subcellular distribution of the tight junctional protein occludin, and differentiation-related proteins CK 20, uroplakins, and actin were followed. Since singular terminally differentiated superficial cells remained on the urothelium after superficial injury (i.e., original superficial cells), we sought to determine their role during the urothelial wound-healing process. Ultrastructural and immunocytochemical studies have revealed that restored tight junctions are the earliest cellular event during the urothelial superficial and full-thickness wound-healing process. Occludin-containing tight junctions are developed before the new superficial cells are terminally differentiated. New insights into the urothelium wound-healing process were provided by demonstrating that the original superficial cells contribute to the urothelium wound healing by developing tight junctions with de novo differentiated superficial cells and by stretching, thus providing a large urothelial surface with asymmetric unit membrane plaques.     

Cyclic AMP-regulated exocytosis of Escherichia coli from infected bladder epithelial cells

The superficial bladder epithelium is a powerful barrier to urine and also serves as a regulator of bladder volume, which is achieved by apical exocytosis of specialized fusiform vesicles during distension of the bladder. We report that type 1 fimbriated uropathogenic Escherichia coli (UPEC) circumvents the bladder barrier by harboring in these Rab27b/CD63-positive and cAMP-regulatable fusiform vesicles within bladder epithelial cells (BECs). Incorporation of UPEC into BEC fusiform compartments enabled bacteria to escape elimination during voiding and to re-emerge in the urine as the bladder distended. Notably, treatment of UPEC-infected mice with a drug that increases intracellular cAMP and induces exocytosis of fusiform vesicles reduced the number of intracellular E. coli.     

The effect of thioglycolate on intermediate filaments and membrane translocation in rat urothelium during the expansion-contraction cycle

Summary The functional role of cytokeratin intermediate filaments in the translocation of asymmetric membrane plaques between cytoplasm and surface of apical urothelial cells was investigated during contraction and expansion of rat urinary bladders. A stereological investigation of electron micrographs provided estimations of surface area, volume, and number of discoidal vesicles and infoldings per unit volume of urothelial apical cell cytoplasm. Contracted and distended bladders incubated in 0.01 M sodium bicarbonate were compared to identical preparations experimentally incubated in 5 mM thioglycolic acid. The latter reagent disrupts the intermediate filament network by reducing sulfhydryl bridges. Densities of discoidal vesicles in cells contracted after incubation in thioglycolate were similar to density estimations in cells expanded under control conditions. Similarly, densities of vesicles in cells expanded after exposure to thioglycolate were comparable in number to those in normally contracted cells. Thus, membrane translocation to and from the luminal surface was blocked by thioglycolate treatment. The lack of normal membrane transfer at the luminal surface induces apical cells exposed to experimental conditions to undergo extraordinary adjustments in response to external pressures of bladder contraction and distension. During contraction, the apical-intermediate cell interface unfolded while the luminal surface ballooned out into the lumen. In distended bladders, large intercellular spaces formed between apical cells along their lateral margins. The results support a model published earlier implicating the filament network as a critical mediator of membrane translocation.     

Urothelium-specific antibody and lectin surface mapping of bladder urothelium

Summary Coupled ligand-colloidal gold complexes were found to provide a convenient approach for the localization by scanning electron microscopy of cell surface membrane antigens and lectin-binding sites on bladder urothelium and for the immunocytochemical identification of urothelial cell populations at different stages of differentiation. The ligands used to probe the membrane were a urothelium-specific rabbit antibody raised to a urothelial membrane-associated antigen (UMA), and two lectins: Concanavalin A (Con A) and peanut agglutinin (PNA). A complex luminal surface distribution pattern was demonstrated by the UMA antigen related to the stage of urothelial cell maturation and differentiation. UMA could be detected on the surface of immature and early differentiating intermediate cells, but was absent from the late differentiation stage, becoming re-expressed as the cells matured and was found in greatest abundance on the terminally differentiated superficial cells. It was absent on cells in benign hyperplasia of the urothelium. Cellular and regional differences in lectin binding to the urothelial cell surface was suggested with Con A receptors localized uniformly over the superficial cells, and PNA receptors confined to linear arrays or occasional clusters over the apical surface but evenly dispersed over the lateral surface of these cells.     

Uroplakins and cytokeratins in the regenerating rat urothelium after sodium saccharin treatment

A sodium saccharin (NaSac) diet was used to induce cell damage and regeneration in the urothelium of the male rat urinary bladder. Foci of terminally differentiated superficial cell exfoliation were detected after 5 weeks and their number increased after 10 and 15 weeks of the diet. At the sites of superficial cell loss, regenerative simple hyperplasia developed. Within 5 weeks of NaSac removal, regeneration re-established normal differentiated urothelium. In order to follow urothelial differentiation during regeneration we studied the expression of uroplakins and cytokeratins by means of immunocytochemistry and immunohistochemistry, respectively. Normal urothelium was characterised by terminally differentiated superficial cells which expressed uroplakins in their luminal plasma membrane and cytokeratin 20 (CK20) in the cytoplasm. Basal and intermediate cells were CK20 negative and cytokeratin 17 (CK17) positive. In hyperplastic urothelium all cells synthesised CK17, but not CK20. Differentiation of the superficial layer was reflected in three successive cell types: cells with microvilli, cells with rounded microridges and those with a rigid-looking plasma membrane on the luminal surface. The cells with microvilli did not stain with anti-uroplakin antibody. When the synthesis of uroplakins was detected rounded microridges were formed. With the elevated expression of uroplakins the luminal plasma membrane becomes rigid-looking which is characteristic of asymmetric unit membrane of terminally differentiated cells. During differentiation, syn-thesis of CK17 ceased in superficial cells while the synthesis of CK20 started. These results indicate that during urothelial regeneration after NaSac treatment, specific superficial cell types develop in which the switch to uroplakin synthesis and transition from CK17 to CK20 synthesis are crucial events for terminal differentiation. Accepted: 19 August 1997     

Detection of urinary bladder cancer with flow cytometry and hexaminolevulinate in urine samples

OBJECTIVES: Urinary bladder urothelial carcinoma is diagnosed by a combination of cystoscopy and biopsy, with cytology as a valuable additional technique. The accuracy of cytological diagnosis depends on the experience of the cytologist and can inevitably vary from one cytologist to another. There is a need for an easy, reliable and objective diagnostic method. In the present study a new method was designed for the detection of bladder cancer cells in urine. METHODS: Flow cytometry was utilized to detect protoporphyrin IX in an artificial model consisting of normal urinary bladder transitional epithelial cells (NBECs) from healthy volunteers' urine and an established human urinary bladder carcinoma cell line, TCCSUP, after incubation with hexaminolevulinate (HAL). In addition, urine samples from 19 patients with histopathologically confirmed superficial bladder cancer were examined. RESULTS: Incubation of NBECs or TCCSUP cells with HAL for 1 hour resulted in production of protoporphyrin IX only in the TCCSUP cells. Incubation of a mixture of NBECs and TCCSUP cells with HAL gave rise to a separated subpopulation of cells with protoporphyrin IX fluorescence. After cell sorting by flow cytometry the protoporphyrin IX-containing subpopulation of cells was confirmed as TCCSUP cells on cytological examination. It was possible to detect 5% TCCSUP cells in the mixture of NBECs/TCCSUP cells. To test the feasibility of the method in clinica diagnosis, urine samples from patients with bladder cancer were also measured with comparable, although preliminary and limited, results to those of cytological examination. CONCLUSIONS: The preliminary results show that the technique may be feasible for the detection of bladder cancer cells in urine with possible advantages of simplicity, reliability and objectivity.     

MAL facilitates the incorporation of exocytic uroplakin-delivering vesicles into the apical membrane of urothelial umbrella cells

The apical surface of mammalian bladder urothelium is covered by large (500-1000 nm) two-dimensional (2D) crystals of hexagonally packed 16-nm uroplakin particles (urothelial plaques), which play a role in permeability barrier function and uropathogenic bacterial binding. How the uroplakin proteins are delivered to the luminal surface is unknown. We show here that myelin-and-lymphocyte protein (MAL), a 17-kDa tetraspan protein suggested to be important for the apical sorting of membrane proteins, is coexpressed with uroplakins in differentiated urothelial cell layers. MAL depletion in Madin-Darby canine kidney cells did not affect, however, the apical sorting of uroplakins, but it decreased the rate by which uroplakins were inserted into the apical surface. Moreover, MAL knockout in vivo led to the accumulation of fusiform vesicles in mouse urothelial superficial umbrella cells, whereas MAL transgenic overexpression in vivo led to enhanced exocytosis and compensatory endocytosis, resulting in the accumulation of the uroplakin-degrading multivesicular bodies. Finally, although MAL and uroplakins cofloat in detergent-resistant raft fractions, they are associated with distinct plaque and hinge membrane subdomains, respectively. These data suggest a model in which 1) MAL does not play a role in the apical sorting of uroplakins; 2) the propensity of uroplakins to polymerize forming 16-nm particles and later large 2D crystals that behave as detergent-resistant (giant) rafts may drive their apical targeting; 3) the exclusion of MAL from the expanding 2D crystals of uroplakins explains the selective association of MAL with the hinge areas in the uroplakin-delivering fusiform vesicles, as well as at the apical surface; and 4) the hinge-associated MAL may play a role in facilitating the incorporation of the exocytic uroplakin vesicles into the corresponding hinge areas of the urothelial apical surface.     

What determines differentiation of urothelial umbrella cells?

Cytokeratins, uroplakins and the asymmetric unit membrane are biochemical and morphological markers of urothelial differentiation. The aim of our study was to follow the synthesis, subcellular distribution and supramolecular organization of differentiation markers, cytokeratins and uroplakins, during differentiation of umbrella cells of mouse bladder urothelium. Regenerating urothelium after destruction with cyclophosphamide was used to simulate de-novo differentiation of cells, which was followed from day 1 to day 14 after cyclophosphamide injection. Cytokeratin 7 and uroplakins co-localized in the subapical cytoplasm of superficial cells from the early stage of differentiation on. At early stages of superficial cell differentiation cytokeratin 7 was filamentary organized, and rare uroplakins were found on the membranes of relatively small cytoplasmic vesicles, which were grouped in clusters under the apical membrane. Later, cytokeratin 7 gradually reorganized into a continuous trajectorial network, and uroplakins became organized into plaques of asymmetric unit membrane, which formed fusiform vesicles. After insertion of fusiform vesicles into the apical plasma membrane, the surface acquired microridged appearance of umbrella cells. Cytokeratin 20 appeared as the last differentiation marker of umbrella cells. Cytokeratin 20 was incorporated into the pre-existing trajectorial cytokeratin network. These results indicate that differentiation of urothelial cells starts with the synthesis of differentiation-related proteins i.e., cytokeratins and uroplakins, and later with their specific organization. We consider that the umbrella cell has reached its final stage of differentiation when uroplakins form plaques of asymmetric unit membrane that are inserted into the apical plasma membrane and when cytokeratin 20 becomes included in a trajectorial cytokeratin network in the subapical area of cytoplasm.