MICROSCOPY RESEARCH AND TECHNIQUE 39:71–84 (1997) Monolayer and Three-Dimensional Cell Culture and Living Tissue Culture of Gallbladder Epithelium YASUNI NAKANUMA,* KAZUYOSHI KATAYANAGI, YASUHITO KAWAMURA, AND KAZUYOSHI YOSHIDA Second Department of Pathology, Kanazawa University School of Medicine, Kanazawa 920, Japan KEY WORDS gallbladder; monolayer cell culture; three-dimensional cell culture; explant; tissue culture; collagen gel ABSTRACT Several models for preparing and isolating human and animal gallbladder epithelial cells, including low-grade gallbladder carcinoma cells, as well as proposed systems for culturing these isolated epithelial cells are reviewed here. Several reports concerning tissue culture of the gallbladder are also reviewed. The cell culture systems are divided into monolayer cell culture on collagen-coated or uncoated culture dishes or other culture substrate and three-dimensional cell culture in collagen gel. To prepare and isolate gallbladder epithelial cells, digestion of the gallbladder mucosa, abrasion of the mucosal epithelial cells, and excision of epithelial outgrowth of mucosal explants are applied. In monolayer cell culture, most of the specific biological features of isolated and cultured cells characteristic to the gallbladder are gradually lost after several passages, though quantitative and objective analyses of the pathophysiology of cultured cells and their secretory substances can be performed. Tissue culture using explants of the gallbladder has mainly been used for physiological studies of the gallbladder, such as investigating the transport of water and electrolytes. In this tissue culture system, quantitative assessment is difficult, though the original and specific biological and histological characteristics of the gallbladder are retained. Three-dimensional collagen gel culture could be an ideal model combining monolayer cell culture and tissue culture systems, and create controllable conditions or environments when several biologically active substances, such as growth factors, proinflammatory cytokines and adhesion molecules, are added to the culture medium. Advantages and shortcomings of individual cultivation models are discussed, and selecting the culture model most appropriate to the purpose of the study will facilitate investigations of the biology and pathogenetic mechanisms of gallbladder diseases such as cholelithiasis. Microsc. Res. Tech. 39:71–84, 1997. r 1997 Wiley-Liss, Inc. INTRODUCTION The gallbladder performs many physiologic functions, namely absorption and secretion of water and electrolytes, control of the content and concentration of bile (Carey and Cahalane, 1988; Hoerl et al., 1992; Sherlock and Dooley, 1992; Tilvis et al., 1982). The characteristic morphologic and phenotypic properties and developmental processes of the gallbladder are also known (MacSween and Scothorne, 1994; Rhodin, 1974). The mucus coat on the luminal surface is thought to be a barrier protecting the mucosa; but it is also providing an environment for microcaliculi formation in cholelithiasis (Carey and Cahalane, 1988). Mucosubstances are secreted from the surface epithelial cells and to a lesser degree the cervical glands. This organ also has a role related to local mucosal immunity mediated by secretion of immunoglobulins, particularly Ig A and other nonspecific bactericidal substances like lactoferrin and lysozyme (Duitton et al., 1985; Nagura et al., 1987). In addition, the gallbladder is affected by a number of diseases. In particular, gallstone disease remains a leading cause of human morbidity and mortality in developed countries including Japan. Gallstone pathogenesis is supposed to involve alterations in the gallbladder wall (Carey and Cahalane, 1988; Conter et al., r 1997 WILEY-LISS, INC. 1992; Moore, 1990), affecting local circulation and mucus secretion, motor functions, and bile composition (Purdum et al., 1993). Yet, despite extensive research into the physiology and pathophysiology of the gallbladder, little is known about mucosal modifications that precede and contribute to gallstone disease. In vitro techniques are required for further investigations on the pathologic and physiologic properties of the gallbladder wall and of its epithelium. Epithelial cell cultures represent one of the adequate in vitro approaches. Several methods of monolayer and threedimensional cell culture after preparation and isolation of biliary epithelial cells from normal or pathologic biliary tract are available (Gall and Bathal, 1987; Hoerl et al., 1992; Joplin et al., 1990; Kawamura et al., 1989). Tissue culture using explants of the gallbladder mucosa is an interesting alternative to cell culture (Elhamady et al., 1983). In this review, several representative models for the preparation and isolation of gallbladder epithelial cells and for their use in monolayer and three-dimensional cell culture are examined. Finally, tissue culture of the *Correspondence to: Yasuni Nakanuma, MD, Second Department of Pathology, Kanazawa University School of Medicine, Kanazawa 920, Japan. Accepted in revised form 5 June 1995. 72 Y. NAKANUMA ET AL. biliary tract, particularly the gallbladder, will be described. The respective advantages, shortcomings and applications of these cultivation methods will also be considered. IDENTIFICATION OF GALLBLADDER EPITHELIAL CELLS FOR IN VITRO CULTURE Several approaches are available for identifying cells isolated from the gallbladder which are proliferating and growing in the culture system. The morphologic features or phenotypic markers described in this chapter are most frequently used for recognizing a biliary epithelial lineage. It is, however, noteworthy that the expression and intracellular distribution of the phenotypic characters described here are dependent upon the degree of in vivo differentiation and the anatomical location or level of the biliary epithelial cells that have been isolated and cultured (Berdichevsky et al., 1994; Nakanuma et al., 1997). The methods of cell isolation and of culture used, specifically the use of supporting or of coated matrices, whether growth factors and serum are added to the medium or not, and the duration and number of passages of cultures, are also important factors. Therefore, the morphologic and phenotypic characteristics described in this paragraph are not constantly expressed in the cultured biliary epithelial cells, while they are in vivo. However, when positive, the cells are most probably originating from biliary epithelium. Morphologic Characteristics Several morphologic characteristics, particularly ultrastructural ones, can be used to distinguish biliary epithelial cells from other cell lineages (Fig.1). In addition to the usual cytologic appearance of epithelial cells themselves, the presence of dense microvilli and cilia on the luminal surface, of complicated intercellular digitation or infoldings, of basally or centrally located nuclei and mucus droplets are in favor of a biliary epithelial lineage. Basement membrane is specific for biliary epithelial cells though its presence depends on the culture model. Mucin Histochemistry Cytoplasmic mucin droplets and a layer of mucinous substance on the luminal surface including a glycocarix can be demonstrated by several histochemical stainings. For example, alcian blue (AB) at pH 1.0 and 2.5 (Fig. 2A), mucicarmine stain, colloidal iron, high iron diamine (HID) and diastase-digested periodic acid Schiff (d-PAS) stainings (Sano, 1976) are routinely used for this purpose. Ruthenium red staining is another method to identify this luminal acid glycoprotein, which represents a surface marker (Nagase, 1988) of biliary epithelial cells. Other Phenotypic Markers of Gallbladder Epithelial Cells Expression of the phenotypic markers described here is highly suggestive of a biliary epithelial cell lineage. However, their expression largely depends on the animal species, the ontogeny and the degree of maturation of the biliary epithelial cells, as well as on the material and methods used for cultures. These makers are Fig. 1. Ultrastructure of the gallbladder epithelium. There are dense microvilli on the luminal surface and basally and centrally located nuclei. Note developed intercellular infoldings (asterisks). L, lumen of the gallbladder; N, nucleus. Normal rabbit gallbladder epithelium, 32,000. detectable by using immunoperoxidase, immunofluorescence, and lectin histochemistry with avidin-biotinperoxidase complex, or by using enzyme histochemistry (Hsu et al.,1981; Nagase, 1988). Cytokeratin (CK). Biliary type CK such as CK 7, 9 and 19 is known to be expressed on human biliary epithelial cells. Particularly, CK 19 is useful in the identification of biliary epithelial cells. These CKs are detected by using commercially available monoclonal antibodies such as AE1, AE3, KL1, Ks19.1 and CAM5.2 (Auth et al., 1993; Van Eyken and Desmet, 1993). CK detected by polyclonal anti-CK antibodies may also be used as a marker of biliary epithelial cells (Nakanuma and Ohta, 1986). Lectin Binding Properties. Concanavalin A (Con A), Pisum sativum agglutinin (PSA), wheat germ agglutinin (WGA), Dolichos biflorus agglutinin (DBA), and soy bean agglutinin (SBA) bind to the apical surface and/or cytoplasm of gallbladder epithelial cells in vivo (Fig. 2B). These lectins are known to bind to the specific carbohydrate residues of glycoproteins (Terada and Nakanuma, 1994; Yoshida et al., 1996). Therefore, it seems reasonable to propose that such carbohydrate residues which are detectable by lectin histochemistry CULTURE OF GALLBLADDER EPITHELIAL CELLS 73 Fig. 2. A: Normal gallbladder. There is a layer of mucinous substance (arrows) on the luminal surface. L, gallbladder lumen. Normal rabbit gallbladder. Alcian blue staining at pH 2.5. B: Normal gallbladder. Dolichos biflorus agglutinin is bound to the luminal surface of the normal rabbit gallbladder (arrows). L, lumen of the gallbladder. Lectin immunohistochemistry using avidin-biotin-peroxidase complex and Dolichos biflorus agglutinin and hematoxylin. be used as a phenotypic marker of the glycoprotein of the gallbladder epithelium in cultures. Mucin Core Protein. Recently, mucus core proteins (MUC 1-7) have been described as the backbone of glycoprotein, and antibodies to these core proteins are becoming available. Among these MUC proteins, MUC 1 is now available commercially (Sasaki and Nakanuma, 1994), and the expression of this protein can be used as a marker of the glycoprotein in biliary epithelial cells. Enzyme Histochemistry. Gamma-glutamyltranspeptidase is a well known as a marker of biliary epithelial cells in man and in experimental animals (Hoerl et al., 1992; Yoshida et al., 1995). Other Phenotypic Markers. Epithelial membrane antigen and a specific glycoprotein which are both constantly expressed in the biliary epithelial cells in vivo including the gallbladder (Momburg et al., 1987) may be used as markers of cultured biliary epithelial cells. Hepatocytes. Expression of albumin or glucose-6phosphatase and the presence of abundant glycogen and canalicular structures (Rhodin, 1974) are in favor of hepatocyte lineages. Cellular Markers to Detect Contaminated Mesenchymal Cells and Hepatocytes Mesenchymal Cells. Vimentin, an intermediate filament found in mesenchymal cells, and factor VIII– related antigens are constantly negative in biliary epithelial cells but positive in mesenchymal cells and in endothelial cells of blood vessels, respectively. So, they are available for identification of these contaminated cells, though vimentin is also expressed on secondary damaged bile ducts (Nakanuma and Kono, 1992) and possibly cultured biliary epithelial cells. Evaluation of Dynamic Changes of Cultured Epithelial Cells Autoradiographic analysis of [3H] thymidine incorporation into DNA (Hoerl et al., 1992) and immunostaining of proliferating cell nuclear antigen and BrdU may be used for investigating the proliferative activities. Flow cytometric analysis is also an interesting method for this purpose (Oda et al., 1991). Uptake of [3H] uridine is used for demonstrating RNA synthesis. Glycoprotein synthesis and secretion can be observed by following the incorporation of [14C] glycosamine into tissue glycoprotein and its subsequent secretion into the culture medium (Oda et al., 1991). MONOLAYER AND THREE-DIMENSIONAL CELL CULTURE Monolayer cell culture is the most usual method for culturing epithelial cells from the biliary tract, namely the gallbladder of humans and experimental animals (Joplin et al., 1990; Parola et al., 1988). Threedimensional cell culture is also challenging in the analysis of biological and pathological characteristics of the gallbladder, particularly with respect to stereotypical analysis (Kawamura et al., 1989). As for the human gallbladder and biliary tract, autopsy materials are not suitable for cultivation of epithelial cells because of prominent and rapid autoly- 74 Y. NAKANUMA ET AL. sis postmortem. Gallbladders without or with only minimal necrotic or inflammatory changes resected by the surgeon for gallstones or for other reasons like prophylactic resection or liver transplantation are recommended for cultivation (Katayanagi et al., submitted; Yoshitomi et al., 1987). For optimal results it is important that the surgically resected gallbladder specimens should be cultured within the first 2 hours (Auth et al., 1993). In general, both monolayer and three-dimensional cell culture systems involve two successive stages, namely the preparation and isolation of the biliary epithelial cells and the cultivation process(es). These stages will be successively described. Viability of isolated epithelial cells is usually tested by trypan blue staining. Preparation and Isolation of Normal and Malignant Gallbladder Epithelial Cells In addition to adequate culture conditions, cell purity is mandatory for a reliable in vitro culture study. There have been numerous reports on obtaining viable and purified gallbladder epithelial cells. Techniques can be subdivided into the following three categories. Preparation of Desquamated Epithelial Cells from the Mucosa Using a Whole Gallbladder or Parts of a Gallbladder. There are several approaches to this procedure (Auth et al., 1993; Conter et al., 1992; Hoerl et al., 1992; Kawamura et al., 1989; Oda et al., 1991; Yoshitomi et al., 1987). Our method using rabbit gallbladder (Kawamura et al., 1989; Yoshida et al., 1996) will be presented first. After the gallbadder was resected in aseptical conditions, the bile was gently aspirated from the lumen with a syringe. T/E mixed medium containing 0.02% ethylenediaminetetraacetic acid (EDTA) and 0.25% trypsin was injected into the empty gallbladder. The organ was then immersed in Hank’s culture medium in a moist chamber at 37oC for 50–70 minutes. Thereafter, T/E medium was recovered and centrifuged. Approximately 80% of all cells in smears from the pellets were isolated epithelial cells but there were also several epithelial cell clumps (12% of clumps consisted of 6 to 10 cells and 4% of clumps consisted of more than 11 cells, respectively). Usually, almost all isolated and clumped cells were viable as shown by trypan blue staining, while contamination by mesenchymal cells was minimal. After the pellets were dispersed and resuspended in culture medium with 15% fetal calf serum, penicillin G, streptomycin, and epidermal growth factor, the cells could be used in either monolayer or three-dimensional cell culture. This procedure appeared to be a simple and reproducible way to prepare singly isolated mucosal cells. However, separation of single cells from small clumps of epithelial cells or from some contaminating mesenchymal cells is not possible with this method. Recently, Oda et al. (1991) have reproduced this procedure using canine gallbladders. Yoshitomi et al. (1987) and Auth et al. (1993) developed other but similar methods for isolation of gallbladder epithelial cells. The former used functional gallbladders of patients undergoing cholecystectomy for gallstones and incubation of the specimens with protease of dispase, while the latter used gallbladders taken at the time of a prophylactic cholecystectomies and have proposed a 0.125% collagenase IV incubation. The mucosa was rinsed with culture medium through a small incision in the gallbladder wall and wiped several times with gauze in order to remove adherent bile. After prolonged incubation with protease or collagenase, careful mechanical abrasion of the gallbladder mucosa was carried out. After the abrasion the same process was repeated several times, the culture fluid containing epithelial cell suspensions being centrifuged twice at 85g for 5 minutes each time. This appeared to be the most efficient way to remove bile sludge and peripheral blood cells while preparing adequate cell pellets. According to Auth et al. (1993), freshly isolated gallbladder epithelial cells form characteristic clusters of 10 to 150 cells (organoid bodies). Hoerl et al. (1992) also developed a method using fragments or explants (approximately 1 cm2) of human surgical specimens with minimal pathological alterations. They removed as much as possible muscle and perimuscular connective tissue from the gallbladder fragments in the culture medium. The roughly denuded mucosal fragments of the gallbladder were floated in 0.17% trypsin in phosphate buffer saline at 4oC overnight and the epithelial cell layer was then separated from the mucosa using curved forceps. The isolated epithelial cells were then processed for cultivation. Several modifications of this method have been proposed. According to our most recent method, samples from which the major part of the muscle layer and connective tissue is removed are immersed in a solution of T/E mixed medium (see above). This fluid containing the fragments of gallbladder mucosa is then gently stirred for 24 hours at 37oC and the supernatant is centrifuged. The obtained cell pellets contain many isolated and clumped gallbladder epithelial cells as well as a considerable number of mesenchymal cells. Kaminsky et al. (1985) described another method using a 25 gauge needle to inject the submucosa of the gallbladder with iced 0.25% NaCl and 2.5% EDTA pH 7.5. at 4oC. The gallbladder was then shaken in the same ice-cold solution for 5 minutes. While shaking, the mucosal cells became separated from the muscle layer. The solutions were pooled and mucosal cells collected by centrifugation at 2,500g for 5 minutes. The pellets were washed twice with iced Krebs-Ringer phosphate solution pH 7.5, and recentrifuged to remove blood cells. This separation technique yielded approximately 500 mg of mucosal cells per gallbladder. Preparation and Isolation of Outgrowing Epithelial Cells from Explant of the Gallbladder on Collagen Gel. This procedure is applicable to human and animal gallbladders (Katayanagi et al., submitted). After removing the gallbladder aseptically, the organ was subdivided into several small pieces in culture fluid. These small explants were incubated in a culture fluid containing 0.05% collagenase and dispase at 37oC for 30 minutes and then thoroughly rinsed in culture fluid. The explants were then placed on collagen gel and cultured in an incubator for 1 or 2 weeks. Using repeated observations under phase contrast microscopy, it was observed that epithelial cells of the small explants were proliferating and were spreading like a sheet on the collagen gel surface (Fig.3A). Although mesenchymal cells were actively growing and spreading simultaneously in the collagen gel or under the CULTURE OF GALLBLADDER EPITHELIAL CELLS 75 Fig. 3. A: Explant of the mouse gallbladder. Phase contrast microscopy of explant (E) of the mouse gallbladder cultured on collagen gel for 1 week and epithelial sheet (asterisks) growing out from the explant. The edge of spreading epithelial sheet is shown by arrowheads. B: Explant of the mouse gallbladder cultured on collagen gel for 5 days. The mucosal epithelium of the explant (E) is viable. There is outgrowing epithelial sheet from the explant. Epithelial cells of the sheet is columnar (arrow) near the explant, while they become flat (arrowhead) at the front of this sheet. Mesenchymal cells are also growing into the collagen gel (asterisks). H&E staining. superficially spreading epithelia cells (Fig.3B), some sites of the culture, particularly at the periphery, were predominantly or purely composed of epithelial cells. Such areas of collagen gel on which a pure epithelial cell sheet was being observed were cut into fragments which were then plated onto other collagen gel surfaces. The latter secondary explants were cultured in a similar way for up to 1 to 2 weeks. From these secondary microexplants, actively proliferating gallbladder epithelial cells without mesenchymal contamination grew and spread on the collagen gel. Up to seven passages of cultured epithelial cells using subsequent explants which form such epithelial cell sheets are possible. Preparation of Carcinoma-Derived Gallbladder Epithelial Cells. Several cell lines have been derived from human gallbladder carcinomas (Johzaki et al., 1989; Koyama et al., 1980; Morgan et al, 1981; Purdum et al., 1993). Because of their adequate of differentiation and because they share many properties with normal gallbladder epithelial cells they have been proposed as a substitute for non-malignant cells in several studies. Confluent Monolayer Cell Culture and Supporting Systems There have been several reports describing methods of monolayer cell culture. This procedure offers a reliable and convenient source of experimentation and has been widely used in research. Hoerl et al. (1992) and Oda et al. (1991) developed and partially characterized a confluent monolayer of cultured canine gallbladder epithelial cells. Kawamura et al. (1989) reported monolayer cell culture of rabbit gallbladder in tissue culture flasks (Fig.4). Monolayer cell cultures can be passaged 76 Y. NAKANUMA ET AL. Fig. 4. A: Monolayer culture of rabbit gallbladder epithelial cells on collagen coated plastic dish for 1 week. There is a monolayer of epithelial cells with occasional vacuoles, suggesting mucus droplets. Phase contrast microscopy. B: Ultrastructure of monolayer culture of mouse gallbladder epithelial cells on collagen gel for 5 days. The epithelial cells are of low columnar epithelial cells resembling the in vivo gallbladder epithelial cells. There are dense microvilli on the luminal surface (arrowheads) and well-developed intercellular lateral infoldings. N, nucleus. L, lumen facing the culture medium, 35,000. Fig. 5. Three-dimensional culture of rabbit gallbladder epithelial cells in collagen gel under phase contrast microscopy. A: One day after being cultured, a small epithelial clump seems to contain small cyst (arrow). B: Three days after being cultured, epithelial cell clump increases in size and seems to be a large cyst (arrow). A and B, 3100. CULTURE OF GALLBLADDER EPITHELIAL CELLS several times (primary, secondary, tertiary cell culture, and so on). These cultured cells after several passages under all conditions eventually enlarge, show vacuolization, and demonstrate irreversible growth arrest. Methods of Monolayer Cell Culture. Monolayer cultivation of isolated gallbladder epithelial cells consists of the following several components and steps. Culture supporting systems including culture vessels and substrate. Several supporting systems are now being applied to monolayer cell culture. It is well known that extracellular matrix component(s) are necessary for the culture of gallbladder epithelial cells, particularly long-term ones requiring successive replications of epithelial cells. For example, collagen, a major constituent of the extracellular matrix in vivo, enhances the growth as well as differentiation of many epithelial cells (Kawamura et al., 1989; Yang et al., 1979, 1980), and has been used for decades as a culture substrate (Elhamady et al., 1983; Kumar and Jordan, 1986). These extracellular matrix components are also important as an attachment factor in cultured epithelial cells. Therefore, it is recommended to coat culture vessels like plastic dishes, flasks or plates with the following factors or substances: collagen type I, fibronectin, Matrigel (containing laminin, type IV collagen, and other basement membrane components), poly-L-lysine, vitrogen, or other components of basement membrane, polycarbon supporting matrix, or Kaiser’s glycerol gelatin. Culture vessels with a feeder layer of irradiated contact-inhibited Swiss albino 3T3 mouse fibroblasts are also preferable for long-term culture up to 6 weeks and provide optimal culture rates, differentiation and proliferation. Collagen gel matrix (see below) itself is another example of culture substrate. Uncoated plastic dishes in serum-free cell culture may also be used for short-term culture of gallbladder epithelial cells for special experimental study such as immunological study, though proliferation and differentiation of cultured cells is limited (Auth et al., 1993). Culture medium. A variety of culture media are now being used for the culture of gallbladder epithelial cells according to the aim of the study. Dulbecco’s modified Eagle’s medium with Ham’s F12 in a 1:1 mixture, William’s medium E, Eagle’s minimum essential medium, MCDB 153 medium, and serum-free keratinocyte growth medium are examples (Auth et al., 1993; Hoerl et al., 1992; Kawamura et al., 1989; Oda et al., 1991; Yoshitomi et al, 1987). The first four are usually supplemented with 10% fetal calf serum with or without growth factors. Several antibiotics such as penicillin G and streptomycin are added in the culture media. These cultures are usually incubated in a 5% CO2 atmosphere at 37oC. Culture media are usually changed every 2 to 3 days, though the frequency depends on the aim of the study. Procedures. Singly isolated or clumped gallbladder epithelial cells are suspended or dispersed in the culture medium and plated on coated or uncoated culture vessels supplemented with an appropriate culture medium. They are cultured for up to 2 weeks (primary culture). Subsequent monolayer cell culture of the primary culture (secondary culture) and that of secondary culture (tertiary culture) and so on are also frequently done. Observations. Attachment, growth and spread of cultured epithelial cells on coated or uncoated culture 77 vessels are usually surveyed and confirmed periodically and routinely by phase contrast microscopy. Morphologic observations by light microscopy, histochemistry, immunohistochemistry, and electron microscopy are also available, when necessary, and proliferative and other dynamic activities are also evaluable by several methods (see above). Growth and Morphologic and Phenotypic Differentiation of Monolayer Cultured Cells. Epithelial cells which adhere to the surface of the culture dish, culture flask or collagen gel within 24 hours or between 24 and 72 hours after being plated immediately start to proliferate and spread. Attached cells have oval nuclei with one or more nucleoli and polygon-shaped cytoplasm with some granules around the nucleus. After 2 to 5 days, they show a cobble stone morphology and gradually form a confluent monolayer after approximately 1 week. Under phase-contrast microscopy, the proliferating cells appear flattened and show a sheetlike growth pattern (Fig.4A). There are occasional vacuoles (cytoplasmic mucous vacuoles) among the cell sheets. These growth patterns and the occurrence of vacuoles become clearer in the secondary culture. Histologically, the outgrowing epithelial cells in culture vessels are largely spindle-shaped or elliptical, and their borders indistinct. The cytoplasm is slightly eosinophilic, and the nuclei are ovoid or elliptical. Occasional multinucleation and mitoses suggest active proliferation. Variably sized vacuoles in the cultured epithelial cytoplasm are positive for PAS, AB at pH 1.0 and 2.5, colloidal iron, HID, and mucicarmine stainings. Such cultured epithelial cells exhibit biliary-type CK, namely CK 7 and CK 19, consistent with their gallbladder epithelial origin; they secrete mucosubstances into the culture medium. Carbohydrate residues detectable by immunohistochemistry and lectin histochemistry are also detected in the cytoplasm of cultured biliary epithelial cells. Transmission electron microscopy shows that the growth and spread of epithelia on the supporting matrices predominantly form a single layer of cuboidal or low columnar cell epithelial cells (Fig.4B). This layer gradually acquires cell polarity towards the culture medium and towards the culture dish, respectively. These patterns are similar to those of the in vivo gallbladder epithelial layer. By transmission and scanning electron microscopy, cultured epithelial cells show variable amounts of surface microvilli, and a few organelles such as mitochondria and Golgi apparatus are seen in cytoplasm. There are variably developed intercellular adhesive devices and intercellular digitations between neighboring cells. These findings suggest the development of cell polarity. Occasionally it has been noted that monolayer cell cultures of normal human gallbladder do not reach confluency (Kumar and Jordan 1986; Purdum et al., 1993). Interestingly, Purdum et al. (1993) observed that well-differentiated carcinoma-derived gallbladder epithelial cells easily formed a confluent monolayer and acquired cellular characteristics and functions similar to those of the normal cells, namely apical orientation and predominantly monolayer growth with cellular interdigitations. However, they noted that timing was critical for obtaining preparations composed mainly of monolayers suitable for variable experimentation. If 78 Y. NAKANUMA ET AL. overgrowth is allowed to develop, the cells tend to head-up or pile up, presumably reflecting their carcinomatous origin. Advantages, and Shortcomings and Applications. Monolayer cell culture is an easy cultivation method for examining the characteristics of biliary epithelium. Quantitative assessment of the characteristics of cultured cells and their alterations with the in vivo pathology is possible. In addition, measuring the membrane permeability to water and electrolytes, secretory activities and the amount of substances secreted in the culture media is also possible (Lu et al, 1988; Okumura et al., 1988; Purdum et al., 1993; Tilvis et al., 1982). Furthermore, in vitro studies of immunological and biological characteristics, such as expression of adhesion molecules and their ligands or receptors, can be performed by adding chemical mediators and proinflammatory cytokines, like interferon or interleukin-2 to the medium (Auth et al., 1933; Joplin et al., 1990). The direct or indirect effects of carcinogenic substances on the biliary epithelium can also be evaluated using this system (Lu et al., 1989). It is also possible to multiply the number of pure gallbladder epithelial cells for biochemical and immunological studies of adequate samples. For example, several passages of explant culture on collagen gel forming pure epithelial sheets treated with collagenase can provide a great number of isolated cells (McAteer et al., 1988). Other approaches with monolayer cell culture yielding pure epithelial cells are also possible. Finally, gallbladder mucosal epithelial cells offer a unique model for the study of epithelial cells in general (Oda et al., 1991). As to the shortcomings, cultured biliary epithelial cells tend to become detached after long-term culture, particularly after more than seven passages. However, this problem depends on the conditions of the cell culture, particularly on the substances used to coat the culture vessels and on the culture medium. In addition, the biological and morphologic features characteristic of the gallbladder epithelium are gradually vanishing with the increased number of passages. Therefore there are only few applications of monolayer cell cultures for the study of the biological characteristics of gallbladder epithelium. However, addition of biologically active substances to the culture medium and the coating of culture vessels with several extracellular substances can improve the culture conditions so that the cells at least partly regain their original characteristics. Three-Dimensional Cell Culture in Collagen Gel Matrix This culture system was first established by Yang et al. (1979, 1980) using mammary carcinoma tissue. This method is needed for analyzing three-dimensional growth of cultured cells and stereotypical changes, namely of cell polarity, under several culture conditions (Berdichevsky et al., 1994). It has been applied to several other organs such as pulmonary alveolar epithelium, thyroid follicular epithelium and adipose tissue (Sugihara et al., 1993; Toda et al., 1990). Our group also applied this method to the cultivation of isolated rabbit gallbladder epithelial cells for up to 6 weeks (Kawamura et al., 1989; Yoshida et al., 1995). In this system, individually isolated epithelial cells or cell clumps start to proliferate while elongating and branching eventually forming multicellular cysts, ramified tubes or other organoid structures covered by one or more cuboidal or low columnar epithelial cells layers. Secretion of several substances into the cystic or tubular lumens by the epithelial cells is observed. This methodology may provide important information on physiological and pathological conditions in the gallbladder. Methods. The collagen gel matrix is usually supplemented with a modified Eagle’s minimum medium, fetal calf serum, hormones, growth factors, and antibiotics, in order to promote sustained growth of normal or cancerous gallbladder epithelial cells in three-dimensional culture. In our studies, the isolated epithelial cells or epithelial cell clumps were dispersed and immersed in a fluid collagen gel matrix (0.3% acid soluble collagen gel (type I collagen:fetal calf serum:concentrated William’s medium (7.5:1.5:1) containing EGF). This collagen fluid in which cells were embedded was gently planted on plastic culture bottles which had been previously coated with solidified collagen gel. This fluid soon became gelatinous. Thereafter the cultures were covered with William’s medium E supplemented with fetal calf serum, penicillin G, streptomycin, and EGF, which was changed every 2 or 3 days according to the experimental design. Culture was maintained at 37oC in humid air containing CO2 for 6 weeks or more. The growth and spread or cyst formation of the cultured epithelial cells in collagen gel was repeatedly observed by phasecontrast microscopy (Fig. 5). Their morphologic changes can be followed by repeated examination as in the monolayer cell culture method. Growth and Phenotypic and Morphologic Differentiation. Cultured cells in collagen gel show increasing changes between the initial stage of culture and the formation of multicellular cysts lined by biliary epithelial cells with gradual acquisition of many in vivo characteristics. Time sequential morphological changes of isolated and cultured epithelial cells in collagen gel can be comprehensively categorized into three phases: loss of cell polarity (up to 12 hours), early reestablishment of cell polarity with proliferation and/or conglomeration of epithelial cells (12–24 hours), and reestablishment of cell polarity with formation of multicellular cysts (after 24 hours)(Kawamura et al., 1989; Yoshida et al., 1995). The loss of cell polarity was characterized by the loss of distribution of organelles and by the transient appearance of autophagic vacuoles. While many round epithelial cells were present singly in collagen gel, there were also considerable numbers of clumps of two to several epithelial cells. These cells contained cytoplasmic vacuoles of various sizes, which were negative for PAS, mucicarmine, HID, and AB (pH 2.5). They were autophagic vacuoles (Gautam et al., 1987) and they failed to bind to the five lectins noted above and disappear by 24 hours. Ultrastructurally, the cell volume of the cultured cells and the numbers of cellular organelles were considerably reduced. The cellular adhesion apparatus was not clearly detected in the singly isolated cells; however, clumps composed of several cells retained this apparatus. A few mucin droplets in the cytoplasm, the CULTURE OF GALLBLADDER EPITHELIAL CELLS foci of microvilli, and several infoldings on the cell surfaces were occasionally identifiable. In most cells, the remaining cell organelles were poorly polarized, and the lateral infoldings were also considerably reduced. Instead, autophagic vacuoles of various sizes which were positive for acid phosphatase (autophagic vacuoles) were frequently found. These vacuoles were positive for acid phosphatase. The early reestablishment of cell polarity with an increased number of clumps of more than four cells was characterized by the appearance of mucin dots (Fig.6A). One or more small dots positive for PAS, AB (pH 2.5), mucicarmine and HID were present in these cell clumps. Some of these dots showed variable binding to Con A, WGA, and DBA after 24 hours, and binding to PSA and SBA was shown after 24 hours. Ultrastructurally, the mucin dots contained small lumens sealed by a few epithelial cells or by cytoplasmic inclusions, both being covered with microvilli and containing an amorphous substance (Fig.7A). Sometimes, a few mucin droplets and a centriole were seen in the vicinity of the small lumens or cytoplasmic inclusions. Some cytoplasmic inclusions were found to be continuous with the small lumens, appearing as diverticula. These inclusions or lumens increased in size, and small lumens consisting of several cells became enlarged, forming multicellular cysts (see below). The microvilli of these inclusions and microcysts were ultrastructurally positive for DBA, WGA, SBA, and ruthenium red. In the late reestablishment of cell polarity, the multicellular clumps began to organize into distinct multicellular cysts (Figs.5A and 6B,C). Their luminal borders and contents were positive for PAS, AB (pH 2.5), HID, and mucicarmine. The lining epithelia of these cysts were composed of one or two layers of cuboidal epithelial cells, while there were secondary mucin dots within the cytoplasm of the epithelial cell layers. The luminal borders of almost all the multicellular cysts showed strong binding to Con A, PSA, WGA, DBA, and SBA. Mucin ‘‘vacuoles’’ in the epithelial layer of the cysts and in the epithelial cell clumps showed similar staining and lectin binding patterns. Ultrastructurally, epithelial cells of the multicellular cysts began to show many features of epithelial cell polarity: dense luminal microvilli, mucin droplets in the subluminal cytoplasm, Golgi apparatus, and abundant mitochondria and rough endoplasmic reticulum in the supranuclear cytoplasm (Fig.7B). The formation of junctional complexes between neighboring cells in these cysts was clear. The microvilli of these epithelial cells were positive for DBA, WGA, SBA, and ruthenium red. After 72 hours, the multicellular cysts were still enlarging with mucinous substances filling the lumen. The structure of the cells in these cysts became more and more similar to those of in vivo gallbladder epithelial cells. The phenotypic and structural features of the cells in these multicellular cysts were active production and secretion of mucinous substances via the apical surfaces into their cystic lumens. No basal lamina was found around the basal side of the multicellular cysts of pure gallbladder epithelial cells. Advantages, Shortcomings and Applications. As to the advantages, three-dimensional outgrowth and morphologic changes of isolated cells and of their polarity (loss and restoration) could be observed and 79 followed sequentially. Moreover, this method is reproducible and is relatively easy to perform. Addition of biological active substances such as growth factors or proinflammatory cytokines produces morphological and phenotypical changes in these cells. Co-culture with other mesenchymal cells or immunocompetent cells from patients or addition of proinflammatory cytokines such as interleukin-2 or interferon-g to the culture media may be applied to investigate the pathogenesis of immune-mediated biliary tract diseases (Auth et al., 1993). Such cultured gallbladder mucosal epithelial cells exhibit parameters similar to those of native gallbladder epithelium and may offer a convenient new research tool for the study of the pathophysiology of gallstone formation and of other gallbladder disease. Loss and restoration of epithelial cell polarity are evaluable using this model (Yoshida et al., 1996). As to the shortcomings, quantitative assessment of cultured cells and biochemical analysis of secreted substances from cultured cells are rather difficult or impossible in this system. Modifications of Monolayer and Three-Dimensional Cell Culture of Gallbladder Epithelial Cells by Addition of Biologically Active Substances Recent studies indicate that many different agents are involved in the stimulation or inhibition of in vitro growth and morphogenesis of normal epithelial cells. Evidence is accumulating that agents such as fetal calf serum, insulin, cortisol, prostaglandin E1 and growth factors also act by regulating the increased rate of initiation of DNA synthesis (Berdichevsky et al., 1994). It is possible to clarify which substance(s) or factor(s) influence the outgrowth and morphogenesis of cultured gallbladder epithelial cells. Receptor(s) on the cultured cells for these agents and morphogens could determine their effects. Effects of Growth Factors as a Morphogen. Yet there are a number of growth factors, like transforming growth factor b1 (TGF-b1) and hepatocyte growth factor (HGF), and of adhesion molecules like a2b1, a3b1, and a6b4 integrins, that act as morphogens (Berdichevsky et al., 1994; Yoshida et al., 1996). For example, the addition of TGF-b1, which is able to influence the proliferation and maturation of epithelial cells, caused the epithelial cells in the multicellular cysts cultured in the collagen gel to become taller and more multilayered (Fig.8A). Their nuclei became more or less hyperchromatic and their cytoplasm was more acidophilic, whereas the mucin dots or vacuoles in the epithelial layer became almost unrecognizable. Ultrastructurally, these cells were acquiring characteristics suggestive of greater maturation than the controls without TGF-b1, and their polarization closely resembled that of the normal cells in vivo; the tight intercellular infoldings and cellular attachment devices, such as tight junctions, intermediate junctions, and desmosomes, were well developed. Mitochondria and rough endoplasmic reticulum were abundant. Microvilli were dense and pinocytic vesicles or invaginations were increased beneath the microvilli. The luminal surfaces of the multicellular cysts after the addition 80 Y. NAKANUMA ET AL. Fig. 6. Three-dimensional culture of rabbit gallbladder epithelial cells in collagen gel. A: One day after being cultured, epithelial cell clump shows PAS-negative vacuoles (small arrows) and positive dot (large arrow). B: Two days after being cultured, PAS-positive vacuoles (arrows) increase in size and become multicellular cysts. C: Three days after being cultured, a multicellular cyst is formed (arrow). A, B, PAS staining after diastase digestion and hematoxylin; C, H&E staining. Fig. 7. Ultrastructures of the rabbit gallbladder epithelial cells cultured in collagen gel. A: One day after being cultured, there is a small intracytoplasmic microcyst covered by microvilli (C), 34,000. B: Four days after being cultured, the wall of the multicellular cyst is composed of cuboidal or low columnar epithelial cells showing dense microvilli on the luminal surface (L) and basally situated nuclei (N), 33,000. of TGF-b1 showed the same mucin-histochemical reactions and lectin-binding patterns as the nontreated multicellular cysts. HGF is also known to cause branching and tree-like growing patterns in such cultured biliary epithelial cells (Nakamura, 1994), though it is unclear whether this property is applicable in vivo in any species. Effects of Cytochalasin B (CB) on the Multicellular Cysts. After the addition of CB which disrupts intracellular microfilaments (Phillips et al., 1983) in the three-dimensional culture of gallbladder epithelial cells (Yoshida et al., 1995), the multicellular cysts disintegrated and within 48 hours some epithelial cells became more or less separated (Fig.8B). These epithe- CULTURE OF GALLBLADDER EPITHELIAL CELLS 81 Fig. 8. Multicellular cyst of rabbit gallbladder epithelial cells cultured in collagen gel for three days. A: After addition of transforming growth factor B1, the epithelial cells of the cyst wall become taller and resemble more the in vivo gallbladder epithelium. H&E staining. B: After addition of cytochalasin B, the epithelial cells consisting of the cyst become disintegrated and PAS-positive and -negative vacuoles appear in their cytoplasm. C: Ultrastructure of multicellular cyst after addition of cytochalasin B show many small intracytoplasmic cysts covered by microvilli (A) and also those covered by long villi resembling intercellular lateral infoldings (B). L, cystic lumen. 34,500. D: Ultrastructure of multicellular cyst after addition of cytochalasin B, showing of intracytoplasmic covered by microvilli (A). There is a thick zone of amorphous substances around these intracytoplasmic cysts. Arrows denote intercellular space covered by infoldings, 38,000 from Yoshida, K., Katayanagi, K., Kawamura, Y., Saito, K., and Nakanuma, Y. (1996) Re-establishment of rabbit gallbladder epithelial cells in collagen gel cultures and their alterations by cytochalasin B and transforming growth factor beta-1: A morphologic study. Pathol. Res. Pract. 192:634–645. lial cells also showed a vacuolated cytoplasm after 48 hours. About half of these vacuoles were mucinpositive, while the remainder were not. Ultrastructurally, the intercellular lateral spaces were wide, and only those cellular adhesion structures facing the lumen were visible. Some of the vacuoles were covered by dense microvilli, resembling those of the luminal surface, and were surrounded by a thick amorphous zone containing glycocalyceal substances, while other vacuoles were covered by a few longer and more slender infoldings resembling the lateral cell surface as if they were empty (Fig.8C,D). The former correspond to mucin- 82 Y. NAKANUMA ET AL. positive vacuoles and the latter to mucin- negative vacuoles. Interestingly, the mucin-positive vacuoles resembled the microvillous inclusions of enterocytes in microvillous inclusion disease (MID)(Bell et al., 1991; Carruthers et al., 1986; Rhoads et al., 1991), in which poorly developed surface brush border microvilli and the formation of intracytoplasmic inclusions lined by microvilli are found. MID is also known to involve the gallbladder (Rhoads et al., 1991). In a preliminary experiment, similar cytoplasmic inclusions were reproduced in cultured human fetal intestinal epithelium, using CB to disrupt the microfilaments. Carruthers et al. (1986) speculated that MID may be attributable to a defect in the cytoskeletal elements of the brush border. It seems likely that microfilaments are essential for the maintenance of the microvilli of the brush border of cultured biliary epithelial cells, and the impairment of these microfilaments may lead to the formation of microvillous or related cytoplasmic inclusions. Addition of Cytotoxic Agents. In a preliminary experimental study we showed that several cytotoxic agents may cause cytopathic damage in gallbladder epithelial cell cultured in collagen gel (Kawamura et al., 1989). This culture procedure might be valuable as an experimental model for clinical conditions with important epithelial damage. For example, gallbladder epithelial cells cultured in the collagen gel for 1 week could be exposed to mitomycin C for 48 hours, then and the epithelial cells could be histologically examined. Mitomycin C is an alkylating agent (Ito, 1979). At concentrations of 100 and 1,000 µg/ml mitomycin C, almost all epithelial cells lining the cysts became necrotic and dropped into the cystic cavities. Almost all cysts cultured in the collagen gel showed similar changes. LIVING TISSUE (ORGAN) CULTURE OF GALLBLADDER Tissue cultures (also called organ cultures) offer a convenient and reliable source of living tissue for adequate experimentation. The use of culture system is widely accepted and has been applied to the biological, physiological, developmental and pathological analysis of the gastrointestinal and urogenital tracts (Menard et al., 1993; Okumura et al., 1988). Similar systems have been applied to human and animal gallbladders. They allow longer time periods during which steady-state conditions can be maintained (Gall and Bathal, 1987; Grant and Billings, 1977; Johzaki et al., 1989; Kumar and Jordan, 1986; Okumura et al., 1988). In studies of the gallbladder, tissue culture is neither popular nor satisfactory when compared to cell culture. One of the reasons is that degenerative changes are likely to occur in tissue-cultured explants of the gallbladder. However, there have been several interesting reports on tissue culture of the biliary tract, including the gallbladder (Gall and Bathal, 1987; Grant and Billing, 1977; Kumar and Jordan, 1986; Okumura et al., 1988). In this review, the most representative experimental models will be examined. For the preparation of explants or fragments of the gallbladder for tissue culture, the gallbladders are aseptically opened and washed gently several times with iced 0.15 M NaCl solution or phosphate buffered saline. Then, adequate samples of the gallbladder showing no or minimal necroinflammatory changes are selected and placed in iced culture medium (see below) and then sectioned into small pieces (explants). Okamura’s Model For the culture media, L-15 and/or Connaught Medical Research Laboratories 1066 with added fetal calf serum, growth factors and antibiotics are used (Okumura et al., 1988). The explants are gently plated on Gelform gelatin sponges which are previously placed on the culture dishes with the mucosal surface upwards. The culture dishes are kept in a shaking incubator in a 5% CO2 atmosphere and the culture media are changed twice a week. In these conditions the gallbladder epithelial cells grew and spread in all directions from the explants; the epithelium was generally flat 7 days after being cultured, then became cuboidal or columnar. Growth also extended into the gelatin sponge. However, regressive changes such as cytoplasmic vacuoles were also detected 14 days after the beginning of culture. Okamura et al. (1988) subsequently transplanted these cultured tissues into the subcutaneous tissue of nude mice. The viability and proliferative activities of these transplanted tissues could be maintained for 29 weeks. Okamura et al. concluded that their method may be applicable for studies on experimental carcinogenesis. Elhamady’s Model Elhamady et al. (1983) reported a tissue culture model of guinea pig gallbladder mucosa that could be maintained up to 2 weeks. The gallbladder of other experimental animals may also be used. The gallbladder was placed in sterile medium, was opened and vigorously rinsed in a series of culture dishes containing one or other of the culture media described below. After the last rinse, the gallbladder was minced to 1 mm2 pieces with a scalpel. The pieces were placed in tissue culture dishes covered by 2 ml complete culture medium and with a coverslip. The dishes were placed in plastic containers, gassed with 5% CO2, sealed with insulating tape and placed in an incubator at 37oC. The following five different culture media were tried: Dulbecco’s modified Eagle’s medium, Eagle’s minimum essential medium with Hank’s salts, NCTC 135 medium, Medium 199, and Ham’s F12. In each instance, penicillin, streptomycin and glutamine were added to the medium. There was morphological evidence of satisfactory survival of the gallbladder epithelium for up to 14 days of culture in all of the media used without marked morphological difference between the media. Only the surface epithelium survived as tall columnar cells and some thinner epithelium appeared to be migrating and growing over the surface of the underlying connective tissue after 2 days. The areas of outgrowth were mainly composed of epithelial-like cells with only a small percentage of fibroblasts. Elhamady et al. (1983) observed outgrowth between days 1 and 2. At the end of week 2, however, the cells underwent dedifferentiation. The overall surface structure became generally simplified: the microvilli became sparse with a few very long ones left between bald areas. This aspect was clearly put in evidence by CULTURE OF GALLBLADDER EPITHELIAL CELLS scanning electron microscopic (SEM). The glycocalyx was observed until day 14. The SEM confirmed that new cells colonized the free surface of the explant. Transmission electron microscopy showed good preservation of the original tall epithelial cells for up to 2 weeks. The new migrating cells were flatter than before but they kept the morphologic features of the columnar cells. Secretory granules were absent after day 1, but increased amounts of glycogen and lipids began to appear later. According to our preliminary study of tissue culture using human or mouse explants with a slight modification of Elhamady’s model, the epithelial cells grew and spread on collagen gel (Fig.3A). These spreading cells near the explant were taller and resembled the epithelial cells on the microexplants, while the epithelial cells at the front of spreading epithelial cell sheet were flat. On the contrary, the stromal cells began to grow mainly into the collagen gel (Fig.3B). Other Models To date, short-term incubation of tissue fragments from the biliary tract including the gallbladder has been used in the experimental study of physiological aspects of the gallbladder. Such studies require demonstration of physiologic functions similar to those found in the native organ, but the proliferation and redifferentiation of gallbladder epithelial cells are not mandatory. For example, Hosono (1950) cultured fragments of guinea pig gallbladder mucosa in blood plasma or Ringer’s solution. Blom and Helander (1977) used pieces of rabbit gallbladder which were incubated in vitro for 1 hour in Ringer’s solution at 37oC and studied correlations of the ultrastructures and water or electrolytes transport and demonstrated the pericellular channels in living rabbit gallbladder epithelium. Wright and Dimond (1968) and Geralden and Rose (1974) showed that 95% of the potential difference in the intact organ is lost upon removal of the mucosa. In vitro preparations of the gallbladder have significantly advanced our knowledge of mucosal function, particularly with respect to both monovalent ion and divalent ion transport. In terms of secretory functions, it is well-recognized that the gallbladder epithelium secretes mucin and other glycoproteins, and in the past decade it has been recognized that the epithelium also secretes H ions. The latter contributes significantly to acidification of bile during the concentration process. The tissue culture method has certain inherent shortcomings. Such studies necessarily involve anesthesia, surgery, manipulation, and care of animals at a relatively high cost. In addition, these experiments require further manipulation and diligent attention to maintain mucosal cell viability. Advantages, Shortcomings and Applications of Tissue Culture Short-term quantitative observation of biliary epithelial cells is possible using this tissue culture model. In particular, a more physiological approach is possible with this system than with the monolayer or threedimensional cell culture. However, the long-term maintenance of this tissue culture of the gallbladder is generally uneasy and limited to 1 week. 83 Although Elhamady et al. (1983) failed to detect significant differences in epithelial growth by morphological observation after changing the culture medium or adding several hormones like insulin and hydrocortisone, further studies with growth factors could be interesting. Adhesion molecules and proinflammatory cytokines which have been recently discovered could also be used in this model to determine whether they significantly affect the morphology of the gallbladder mucosa. OVERVIEW Both in vivo and in vitro experiments and studies have not only advantages but also shortcomings. The use of monlayer and three-dimensional cell culture as well as tissue culture systems has become an accepted method in hepatobiliary research namely for investigating the gallbladder. For cultivation, several adequate models of preparing and isolating epithelial cells from human and animal gallbladders are available and are widely used. Monolayer and three-dimensional cell culture have their own advantages and shortcomings. Considering the various models of isolating gallbladder epithelial cells and the variety of culture methods, the aim and design of the experiment should guide the selection of the most appropriate isolation and cultivation technique. Appropriate selection will facilitate analysis of the physiology of the gallbladder as well as the pathophysiology and pathogenesis of gallbladder disease including gallstone disease. REFERENCES Auth, M.K.H., Keitzer, R.A., Scholz, M., Blaheta, R.A., Hottenrott, E.C., Herrmann, G., Encke, A., and Markus, B.H. (1993) Establishment and immunological characterization of cultured gallbladder epithelial cell. Hepatology, 18:546–555. Bell S.E., Kerner, J.A., and Sibley, R.K. (1991) Microvillous inclusion disease. The importance of electron microscopy for diagnosis. Am. J. Surg Pathol., 15:1157–1164. Berdichevsky, F., Alford, D., D’Souza, B., and Taylor-Papadimitriou, J. (1994) Branching morphogenesis of human mammary epithelial cells in collagen gel. J. Cell. Sci., 107:3557–3568. Blom, H., and Helander, H.F. (1977) Quantitative electron microscopical studies on in vitro incubated rabbit gallbladder epithelium. J. Membr. Biol., 37:45–61. Carey, M.C., and Cahalane, M.J. (1988) Whither biliary sludge? Gastroenterology, 95:508–523. Carruthers, L., Dourmashkin, R., and Phillips, A. (1986) Disorders of the enterocytes. Clin. Gastroenterol., 15:105–129. Conter R.L., Washington, J.L., Liao, C.C., and Kauffmann, G.L. (1992) Gallbladder mucosal blood flow increases during early cholesterology, gallstone formation. Gastroenterology, 102:1764–1770. Duitton, D.A., Seilles, E., and Cause, P. (1985) Gall bladder: The predominant source of bile IgA in man? Clin. Exp Immunol., 62:185–192. Elhamady, M.S., Hopwood, D., Milne, G., Ross, P., and Bouchier, I.A.D. (1983) Tissue culture of guinea pig gall-bladder epithelium. J. Pathol., 140:221–235. Gall, J.A.M., and Bathal, P.S. (1987) Isolation and culture of intrahepatic bile ducts and its application in assessing putative inducers of biliary epithelial hyperplasia. Br. J. Exp. Pathol., 68:501–510. Gautam, P., Ng, O.C., and Boyer, J.L. (1987) Isolated rat hepatocytes couplets in short-term culture: Structural characteristics. Hepatology, 7:216–223. Geralden, R.T., and Rose, R.C. (1974) Electrical properties and diffusion potentials in the gallbladder of man, monkey, dog, goose and rabbit. J. Membr. Biol., 19:37–54. Grant, A.G., and Billings, B.H. (1977) The isolation and characterization of a bile ductule cell population from normal and bile-duct ligated rat livers. Br. J. Exp. Med., 53:301–310. Hoerl, B.J., Vroman, B.T., Kasperbauer, J.L., LaRusso, N.F., and Scott, R.E. (1992) Biological characteristics of primary cultures of human gallbladder epithelial cells. Lab. Invest., 66:243–250. 84 Y. NAKANUMA ET AL. Hosono, S. (1950) Cultivation in vitro of the epithelium of the gall bladder from guinea pig. III. Report. Acta Soc. Pathol. Jpn., 15:430– 431. Hsu, S.M., Raine, L., and Fanger, H. (1981) Use of avidin-biotinperoxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabelled (PAP) procedures. J. Histochem. Cytochem., 29:557–580. Ito, H. (1979) Phamacology. 6th ed. Tokyo, Eikodo (in Japanese). Joplin, R., Strain, A.J., and Neuberger, J.M. (1990) Biliary epithelial cells from the liver of patients with primary biliary cirrhosis: Isolation, characterization, and short-term culture. J. Pathol., 162: 255–260. Johzaki, H., Iwasaki, H., Nishida, T., Isayama, T., and Kikuchi, M. (1989) A human gallbladder adenocarcinoma cell line. Cancer, 64:2262–2268. Kaminsky, D.L., Deshpande, Y., Thomas, L., Qauly, J., and Blank, W. (1985) Effect of oral ibuprofen on formation of prostaglandins E and F by human gallbladder muscle and mucosa. Dig. Dis. Sci., 30:933– 940. Kawamura, Y., Yoshida, K., and Nakanuma, Y. (1989) Primary culture of rabbit gallbladder epithelial cells in collagen gel matrix. Lab. Invest., 61:350–356. Koyama, S., Yoshioka, A., Kawakita, I., Yamagata, S., Fukutomi, H., Sakita, T., Kondo, I., and Kikuchi, M. (1980) Establishment of a cell line (G-415) from a human gallbladder carcinoma. Jap. J. Cancer Res. 71:574–575. Kumar, U., and Jordan, T.W. (1986) Isolation and culture of biliary epithelial cells from the biliary tract of normal rats. Liver, 6:369– 378. Lu, M.D., Miyazaki, K., Yoshitomi, S., and Nakanuma, F. (1988) DNA repair synthesis in primary culture of bovine bile duct epithelial cells induced by chemical agents in relation to bile duct cancer. Mutat. Res., 194:73–79. MacSween, R.N.M., and Scothorne, R.J. (1994) Developmental anatomy and normal structure. In: Pathology of the Liver. R.N.M. MacSween, P.P. Anthony, P.J. Scheuer, A.D. Burt, and B.C. Portmann, eds. 3rd ed. Churchill Livingstone, Edinburgh, pp.1–49. McAteer, J.A., Carone, F.A., Grantham, J.J., Kempson, S.A., Gardner, K.D., and Evan, A.P. (1988) Explant culture of human polycystic kidney. Lab. Invest., 59:126–136. Menard, D., Arsenault, P., and Monfils, S. (1993) Maturation of human fetal stomach in organ culture. Gastroenterology, 104:492–501. Momburg, F., Moldenhauer, G., and Hammerling, G.H.(1987) Immunohistochemical study of the surface expression of a M4 34000 human epithelium-specific glycoprotein in normal and malignant tissues. Cancer Res., 47:2883–2891. Moore, E.W. (1990) Biliary calcium and gallstone formation. Hepatology, 12:206s–218s. Morgan, R.T., Woods I.K., Moore, G.E., McGavron, G.E., Guinn, L.A., and Semple, T.V. (1981) A human gall bladder adenocarcinoma cell line. In Vitro, 17:503–510. Nagase, M. (1988) Tracer and electron charge. In: Experimental Histochemistry. H. Hirano and M. Yokoyama, eds. Maruzen, Tokyo, pp.608–645 (in Japanese). Nagura, H., Smith, P.D., and Nakane P. (1987) IgA in human bile and liver. J. Immunol., 126:587–595. Nakanmura, T. (1994) Hepatocyte growth factor and liver cell regeneration. Acta Hepatol. Jpn. (suppl.) 35:33 (in Japanese). Nakanuma, Y., and Ohta, G. (1986) Immunohistochemical study on bile ductular proliferation in various hepatobiliary diseases. Liver, 6:205–211. Nakanuma, Y., and Kono, N. (1992) Expression of vimentin in proliferating and damaged bile ductules and interlobular bile ducts in nonneoplastic hepatobiliary diseases. Mod. Pathol., 5:550–554. Nakanuma, Y., Hoso, M., Sanzen, T., and Sasaki, M. (1997) Microstructure and development of the normal and pathologic biliary tract in humans, including blood supply. Microsc. Res. Tech. (in press). Oda, D., Lee, S., and Hayashi, A. (1991) Long-term culture and partial characterization of dog gallbladder epithelial cells. Lab. Invest., 64:682–692. Okumura, H., Sakamoto, K., Nakano, G., Ikeda, H., Nagashima, K., Nagamachi, Y., and Yuasa, Y. (1988) Long-term maintenance of human gall-bladder epithelia as xenografts following explant organ culture. Jpn. J. Surg., 89:388–393 (in Japanese). Parola, M., Cheeseman, K.H., Biocca, M.E., Dianzani, M.U., and Slater, T.F. (1988) Isolation and characterization of biliary epithelial cells from normal rat liver liver. J. Hepatol., 6:175–186. Phillips, M.J., Oshio, C., Miyairi, M., and Smith, C.R. (1983) Intrahepatic cholestasis as a canalicular motility disorder. Evidence using cytochalasin. Lab. Invest., 48:205–211. Purdum, P.P., Ulissi, A., Hylemon, P.B., Shiffman, M.L., and Moore, E.W. (1993) Cultured gallbladder epithelia. Methods and partial characterization of a carcinoma-derived model. Lab. Invest., 68:345– 353. Rhoads, J.M., Vogler, R.C., Lacy, S.R., Reddick, R.L., Keku, E.O., Azizkhan, R.G., and Berschneider, H.M. (1991) Microvillous inclusion disease. In vitro jejunal electrolyte transport. Gastroenterology, 100:811–817. Rhodin, J.A.G. (1974) Histology. A Test and Atlas. New York and London, Oxford University Press, 1974. Sano, Y. (1976) Histological Techniques; Theoretical and Applied. 5th ed. Tokyo, Nanzando (in Japanese). Sasaki, M., and Nakanuma, Y. (1994) Expression of mucin core protein of mammary type in primary liver cancer. Hepatology, 20:1192– 1197. Sherlock, S., and Dooley, J. (1993) Diseases of the Liver and Biliary System. 9th ed. Blackwell, London. Sugihara, H., Toda, S., Miyabara, S., Fujiyama, C., and Yonemitsu, N. (1993) Reconstruction of alveolus-like structure from alveolar type II epithelial cells in three-dimentional collagen gel matrix culture. Am. J. Pathol., 142:783–792. Terada, T., and Nakanuma, Y. (1994) Profiles of expression of carbohydrate chain structures during human intrahepatic bile duct development and maturation: A lectin-histochemical and immunohistochemical study. Hepatology, 20:388–397. Tilvis, R.S., Aro, J., Strandberg, T.E., Lempinen, M., and Miettinen, T.A. (1982) In vitro synthesis of triglycerides and cholesterology, in human gallbladder mucosa. Scand. J. Gastroenterol., 17:335– 340. Toda, S., and Sugihara, H. (1990) Reconstruction of thyroid follicles from isolated porcine follicle cells in three-dimensional collagen gel culture. Endocrinology, 126:2027–2034. Van Eyken, P., and Desmet, V.J. (1993) Cytokeratin and the liver. Liver, 13:113–122. Wright, E.M., and Diamond, J.M. (1968) Effects of pH and polyvalent cations on the selective permeability of gall-bladder epithelium to monovalent ions. Biochim. Biophys. Acta, 163:57–74. Yang, J., Richards, J., Bowmann, P., Guzman, R., Enami, J., McCormick, H., Hamamoto, S., Pitelka, D., and Nandi, S. (1979) Sustained growth and three-dimensional organization of primary mammary epithelial cells embedded in collagen gel. Proc. Natl. Acad. Sci. U.S.A., 76:3401–3405. Yang, J., Guzman, R., Richards, J., Jentoft, V., Devault, M.R., Wellings, S.R., and Nandi, S. (1980) Primary culture of human mammary epithelial cells embedded in collagen gels. J. Natl. Cancer Inst., 65:337–343. Yoshida, K., Katayanagi, K., Kawamura, Y., Saito, K., and Nakanuma, Y. (1996) Re-establishment of rabbit gallbladder epithelial cells in collagen gel culture and their alterations by cytochalasin B and transforming growth factor beta-1: A morphologic study. Pathol. Res. Pract. 192:634–645. Yoshitomi, S., Miyazaki, K., and Nakayama, F. (1987) Demonstration and maintenance of mucus secretion in cultured human gallbladder epihtelial cells. In Vitro Cell Dev. Biol., 23:559–566.