Cell Biology and Toxicology, Vol. 7, No. 4, 1991
315
GLUTAMATE UPTAKE IN PRIMARY CULTURES OF BILIARY EPITHELIAL CELLS FROM NORMAL RAT LIVER IRIS EISENMANN-TAPPE, SUSANNE WIZIGMANN, AND ROLF GEBHARDT Physiologisch-chemisches Institut Universitiit Tiibingen, Germany
Biliary epithelial cells (BEC) were isolated from normal rat liver with ,high purity (> 95%) as revealed by morphological criteria as well as staining for gamma-glutamyl transferase and cytokeratin 19. During cultivation fi2r 96 hr flattening of the cells and a loss of microvilli was apparent, while the cytokeratin 19-positive phenotype was maintained. The BEC contained a sodium-dependent as well as a sodium-independent uptake system for glutamate with high capacity. Both activities increased transiently during cultivation peaking after 72 and 48 hr, respectively. After 72 hr, apparent kinetic constants could be calculated for the sodium dependent (Krn = 13.6 mM; Vmax = 388 nmoles/min/mg protein) and for the sodium-independent system. (Km = 10.8 mM; Vma,x= 132 nmoles/min/mg protein). The transient increase of both transport systems was suppressed by dexamethasone. The sodium-dependence showed a threshold concentration of about 35 mM sodium. Inhibition by kainate was much less potent for BEC than for hepatocytes. These data indicate that BEC contain transport systems for glutamate different from those in hepatocytes and which may be involved in the intrahepatic reabsorbtion of glutamate from bile. INTRODUCTION Biliary epithelial cells (BEC) are small, cuboidal cells which line the intrahepatic bile ducts. It is known from morphological studies that BEC exhibit typical features of secretory as well as absorbing epithelia, and they are therefore thought to play a major role in modifying the primary bile secreted by hepatocytes. During the last few years appropriate methods to obtain
1. Direct all correspondence to: Dr. R. Gebhardt, Physiologisch-chemisches Institut, HoppeSeyler-Strasse 4, D-7400 Ttibingen 1, Germany. Tel: 49/7071/293357. Fax: 49/7071/293361. 2. Key words: cytokeratin 19, gamma-glutamyl transferase, dexamethasone, cholehepatic shunt. 3. Abbreviations: BEC, biliary epithelial cells; DMEM, Dulbecco's Modified Eagle's Medium; GGT, gamma-glutamyl transferase; Dex, dexamethasone; Glu, glutamate; N-Me-AIB, N-methylaminoisobutyrate; Hep, hepatocytes FBS, Fetal bovine serum. Cell Biology and Toxicology, Vol. 7, No. 4, pp. 315-325 Copyright © 1991 Princeton Scientific Publishing Co., Inc. ISSN: 0742-2091
316 Eisenmann-Tappe et al.
reasonably pure preparations of these cells have been established (Parola et al., 1988; Ishii et al., 1989; Alpini et al., 1989). Thus, isolated and cultivated BEC are now available as an in vitro model for studies of their biochemical properties as well as those of secretion and absorption. Besides cholesterol and bile salts, bile contains an abundance of organic solutes, including sugars, proteins, and free amino acids. Whereas the occurrence of fluid-phase endocytosis in BEC has recently been reported (Ishii et al., 1990), very little information exists to date about specific transport systems in BEC that could be involved in ductular absorption of solutes from bile. Such a process has been shown for glucose (Guzelian and Boyer, 1974; Olson and Fujimoto, 1980), and a "cholehepatic shunt pathway" has been postulated for several unconjugated mono- and dihydroxy bile acids (Yoon et al., 1986). Among the free amino acids in bile, L-glutamate is the most abundant (Ballatori et al., 1986). It is liberated from glutathione within the biliary tree by the action of gamma-glutamyl transferase, an enzyme located in the luminal membranes of BEC. In the present study we have characterized the uptake of L-glutamate in primary cultures of BEC derived from normal rats, and its behavior under different culture conditions. MATERIALS AND METHODS Animals. Male Sprague-Dawley rats (240-300g) were used as cell donors. They were kept in a controlled 12 hr-light and -dark cycle on a standardized diet of Alma H 1003 (Botzenhardt, Kempten, Germany) and tap water ad libitum. Isolation and cultivation of biliary epithelial cells. Biliary epithelial cells were isolated by combining the collagenase perfusion technique of Seglen (1976) and a modification of the procedures described by Parola et al. (1988) and Gall and Bhathal (1985).
After liver perfusion with 0.2% collagenase (No. 17446, Serva, Heidelberg) and 0.1% hyaluronidase (Sigma, Mtinchen) for 10 min, the liver capsule was opened and the great majority of hepatocytes was removed by gently shaking the liver in 100 ml of "isolation medium" (Hank's solution containing 2% glucose, (Seglen, 1976)). The remaining tissue containing the bile ducts was minced using scalpels and transferred to a small stoppered flask containing 25 ml of 0.3% trypsin (Serva, Heidelberg) in isolationmedium to allow dispersion of bile ducts. The flask was agitated continuously at room temperature for 40 min. During this period, the duct material was dispersed several times using a pasteur pipette. 25 ml of 0.3% trypsin-inhibitor in isolation medium was then added and the suspension was filtered through a 100 ~ mesh nylone gauze to remove clumped material. Following centrifugation (300g, 3 min) the cells were resuspended in 2 ml of isolation medium and loaded onto a discontinuous Percoll gradient (Pharmacia, Freiburg), prepared exactly as described by Gall and
Cell Biology and Toxicology, Vol. 7, No. 4, 1991
317
Bhathal (1985). The sample was centrifuged at 1,800 g for 20 min in a swing-out rotor (Labofuge III, Heraeus-Christ, Osterode). Cells at the 1.06-1.08 density interface were routinely collected and washed once in 50 ml of Hank's solution in order to remove the percoll. The cells were then transferred to a 90 mm culture plate precoated with a thin film of collagen as described by Gebhardt and Jung (1982) and incubated for 60 min in Dulbecco' s Modification of Eagle's medium (DMEM, Flow Laboratories) containing 10% FBS. This procedure allows the sequestration of remaining hepatocytes, Kupffer cells, and other contaminating cells since they adhere much more quickly to the substratum than do BEC. After the incubation period the medium was collected by gently shaking the culture plate and regained BEC were counted using the trypan blue exclusion test. The cells were resuspended in DMEM containing 10% FBS, 2 mM L-glutamine, penicillin (50 U/ml), and streptomycin (50 gg/ml) at the density of approx. 0.2 x 10 6 cells/ml and inoculated into costar 24-well cluster trays (0.2 ml/well) for transport measurements. For all other experiments cells were cultured at the same cell density. All supporting material was precoated with collagen (see above). Cultures were incubated at 37 °, 90% humidity, and 5% CO2 in air for the periods described for the different experiments. The medium was renewed daily.
Histochemical staining. Histochemimcal staining for gamma-glutamyl transferase was performed on methanol-fixed cell smears of freshly isolated cells using the method of Rutenburg et al. (1969). Cells were washed with saline prior to staining and incubated with the reaction mixture for 30 min to develop the dye. Immunocytochemistry. For identification of the cells we used a specific antiserum to cytokeratin 19 (Arnersham, Braunschweig) which has been shown to bind specifically to BEC in liver sections (Moll et al., 1982). Immnnohistochemical staining was performed on cultured cells fixed with methanol at -20*C for 6 rain using a modification (Gebhardt et al., 1.986) of the unlabeled antibody peroxidase-anti-peroxidase method by Sternberger et al. (1976). Glutamate transport. Glutamate uptake was measured using the cluster-tray method of Gazzola et al. (1981) in a modification described by Gebhardt and Kleeman (1987). Uptake rates were determined at an incubation interval of 1 rain in a modified Krebs-Ringer bicarbonate buffer (pH 7.4) containing sodium or choline, respectively. The Na+-dependent transport of 0.1 mM L-(U-a4C)-glutamate (Amersham Buchler) was determined as the cysteate-inhibitable part of glu uptake after substraction of the Na+-independent transport values. Na+-independent uptake was determined in the absence of specific inhibitors. Uptake rates are given in nmoles/min/mg protein. Statistical evaluation of the data was performed by Student's t-test. RESULTS
AND DISCUSSION
Cell isolation and cell culture. Cell yield usually ranged between 3 and 5 x 106 cells, and the cell viability estimated by the trypan blue exclusion test was always > 95%. Application of
318
Eisenmann-Tappe et al.
hyaluronidase during the isolation procedure proved to be necessary since satisfactory cells isolates were not obtained in its absence. Within the first 4 hr after isolation attachment of the cells to the collagen coated plates could not be observed. Even after 24 hr of cultivation the cells were only loosely adhering to the subslratum (Figure 1). When the cells had flattened completely during day 2 of cultivation, they proliferated rapidly usually forming a confluent monolayer after 96 hr of cultivation. The cells could be cultured at good viability for up to 120 hr.
FIGURE 1. Immunocytochemical detection of cytokeratin 19 by means of the unlabeled peroxidase-antiperoxidase method in BEC cultured for 24 (A) and 72 hr (B). Contaminating hepatocytes (arrow) do not stain. Magnification x 230.
Cell identification. Gamma-glutamyl transferase has been shown to be localized on the canalicular membranes of BEC. Methanol-fixed cells smears of the freshly isolated cells were stained for GGT using the histochemical procedure of Rutenburg et al. Usually more than 95% of the cells were positive for GGT at the end of the isolation procedure. As an even more specific marker for BEC we tested the presence of cytokeratin 19. Unlike GGT, which is also present in some hepatocytes around the portal tract, cytokeratin 19 has been shown to be localized exclusively in BEC in normal rat liver (Moll et al., 1982). Immunocytochemical detection of CK-19 was performed in the freshly isolated state, and after 24, 72, and 120 hr of cultivation. The cells were stained homogeneously and usually contained less than 5% of contaminating cells, which were mainly hepatocytes and unidentified fibroblast-like cells (Figure 1). Cytokeratin 19 was expressed constantly up to 120 hr.
Cell Biology and Toxicology, Vol. 7, No. 4, 1991
319
The electon micrographs of freshly isolated BEC show cells still partially attached to each other containing a highly convoluted nucleus, few mitochondria and a great number of vacuoles beneath the plasma membrane (Figure 2 A). Microvilli are numerous and very long. After 24 hr (Figure 2 B) the cells were attached to the substratum, but had not yet flattened. The number of microvilli seemed slightly diminished. Figure 2 C shows cells from a confluent monolayer after 72 hr of cultivation. Along with flattening of the cells the nuclei had assumed a more regular shape. Tight junctions were formed. Most remarkably° only very few and short microvilli could be seen at the cell surface. The loss of microvilli may be due to adaptation to culture conditions. Its influence on membrane functions, in particular substrate transport, is not known.
FIGURE 2. Electron micrographs showing fleshly isolated and cultured BEC. (A) Freshly isolated BEC. The cells in the lower left are still attached to each other by fight junctions (arrow). (B) BEC cultured for 24 hr. The smooth surface on the lower left represents the contact side with the substratum. (C). BEC cultured for 72 hr. Cells have flattened and tight junctions are reformed (arrow). Bars : 2 !am.
320
Eisenmann-Tappe et al. 5
i, ®
U--] - d e x I~R + d e x
A
3
0
i 1,
0 Z4
48
"/2
cultivation period
96
(h)
5
4 O2
I--] - d e x I~q + d e x
]3
3
0
0 24
45
72
96
c u l t i v a t i o n p e r i o d (h) FIGURE 3. Influence of cultivation period and dexamethasone supply on the sodium dependent (A) and the sodium independent (B) uptake of glutamate in cultivated BEC. Values are means + SD from 3-5 different experiments. Glutamate transport. The uptake rates for L-glutamate were measured after 24, 48, 72, and 96 hr of cultivation. The results are given in Figures 3A and 3B. As can be seen, a sodium-dependent as well as a sodium-independent transport system for glutamate existed in BEC, and transport values for both systems depended on the cultivation period. Na÷-dependent specific uptake rates are given in Figure 3A (open bars). Values increased simultaneously with the proliferation of the cells up to 72 hr and declined thereafter when the cultures reached confluency.
Cell Biology and Toxicology, Vol. 7, No. 4, 1991
321
Comparable data were obtained for the sodium-independent transport of glutamate (Figure 3 B, open bars), with a somewhat different induction pattern. Maximum transport rates were measured after 48 hr and values did not change significantly thereafter. The sodium-dependent transport system generally amounted to about 40% of the total glutamate uptake of the cells, except for the extreme 72 hr-induction maximum, when its part increased even to 66%. This behavior is in marked contrast to glutamate transport in hepatocytes. Generally, uptake rates are very much lower in cultured hepatocytes (Na+-dependent: 0.3+0.1 nmoles/min/mg; Na+-independent: 0.1+0.03 nmoles/min/mg; N=4) than in cultured BEC (Na+-dependent: 3.0+0.03 nmoles/min/mg; Na+-independent: 1.3+0.7 nmoles/min/mg; N=5) when measured under the same culture conditions. It should be noted, however, that hepatocytic glutamate uptake is highly zonated (Burger et al., 1989; Stoll et al., 1991), particularly the sodium-dependent glu uptake (Gebhardt, 1989). Furthermore, uptake via the sodium-dependent transport system in isolated hepatocytes amounts only to about 30% of total glu uptake (Gebhardt and Mecke, 1983; Burger et al., 1989), but can be significantly induced by dexamethasone (dex) during cultivation, while the sodium-independent transport of glu in hepatocytes is not affected by dex. We therefore tested the effect of dex on the glu transport systems in BEC. The cultivation medium was supplemented by 10-7 mM dex from the beginning of cultivation and transport rates were measured under the same conditions. Surprisingly, dex did not have an inductory effect on glu transport in BEC. Instead both transport systems were simultaneously reduced by dex (Figures 3 A and 3 B, hatched bars). Sodium-dependent transport rates decreased to 45-70% of control values without dex, sodium-independent transport to 50-65% of control values (Figures 3 A and 3 B, open bars) except for the data measured after 24 hr of cultivation, when the dex incubation period was obviously too short to be effective. Furthermore, dex did not change the induction pattern of glu uptake during cultivation. Interestingly, the adult rat liver epithelial cell lines (ARL) (Williams and Gunn, 1974) also display a high glutamate transport rate (Gebhardt and Williams, 1986) which suggests a relationship of the clonigenic precursors of these lines (Furukawa et al., 1986) to BEC. Glutamate transport in ARL lines also was not increased by dex (Gebhardt and WiUiams, 1986). To exclude an influence of the glutamate supply on transport rates, we raised the glu concentration in the cultivation medium to 2 mM, which should be about twice the concentration in bile (Ballatori et al., 1986) and incubated the cells up to 96 hr. This did not have any effect on the induction phenomenon, suggesting that in normal BEC, just as in hepatocytes, an adaptive control of glutamate transport does not exist. In order to characterize further the transport systems involved in glutamate uptake in BEC, we tested the inhibitory effect of several L-amino acids on glu transport. L-cysteate, L-aspartate and the cyclic glutamate analogon kainate are known to be potent inhibitors of the sodium dependent transport system G- in hepatocytes. On the other hand, N-methyl-tx-aminoisobutyrate
322
Eisenmann-Tappe et al.
(N-Me-AIB), which is transported via system A, a transport system for short neutral amino acids in hepatocytes, should not have any effect on glu uptake. Glu concentration in these experiments was 0.1 mM, all inhibitors were added at a concentration of 10 mM. Table 1 gives the inhibition rates measured in BEC and hepatocytes after 72 hr of cultivation. Cysteate and aspartate were inhibitors of sodium-dependent and sodium-independent glu transport in BEC, and data are very similar to those obtained in hepatocytes. As expected, N-Me-AIB had only a very weak inhibitory effect. Kainate, however, which is a potent inhibitor of sodium-dependent glu transport in hepatocytes, had only a weak effect on sodium-dependent glu transport and practically did not inhibit the sodium-independent transport in BEC. Obviously, this cyclic analogon is less accepted by the transport systems in BEC which may indicate some differences in the molecular organization of the transport proteins in BEC and hepatocytes.
TABLE 1.
Inhibitor
Inhibition of the uptake of 0.1 mM glutamate by several glutamate analogues given in a concentration of 10 mM each
Inhibition of sodium-dependent uptake (%) BEC
cysteate aspartate kainate N-Me-AIB
93.3 + 93.4 + 33.4 + 9.4 +
15.1 (5) 12.0 (5) 30.3*(5) 11.0 (3)
Inhibition of sodium-independent uptake (%)
Hep
BEC
95.8 + 7.2 (4) 98.5 + 3.7 (4) 98.6 + 2.0 (4) 7.0 (2)
19.5 18.7 5.0 9.4
Hep + 6.5 (5) + 9.6 (5) + 7.1"(5) + 12.3 (3)
19.4 + 9.7 (4) 14.7 + 15.0 (4) 26.3 + 6.8 (4) 0 (2)
BEC = biliary epithelial cells; Hep = hepatocytes; N-Me-AIB = N-methyl-aminoisobutyrate Data are mean values + S.D. The number of experiments is given in parenthesis * statistical evaluation: significantly different from hepatocytes" P
315
GLUTAMATE UPTAKE IN PRIMARY CULTURES OF BILIARY EPITHELIAL CELLS FROM NORMAL RAT LIVER IRIS EISENMANN-TAPPE, SUSANNE WIZIGMANN, AND ROLF GEBHARDT Physiologisch-chemisches Institut Universitiit Tiibingen, Germany
Biliary epithelial cells (BEC) were isolated from normal rat liver with ,high purity (> 95%) as revealed by morphological criteria as well as staining for gamma-glutamyl transferase and cytokeratin 19. During cultivation fi2r 96 hr flattening of the cells and a loss of microvilli was apparent, while the cytokeratin 19-positive phenotype was maintained. The BEC contained a sodium-dependent as well as a sodium-independent uptake system for glutamate with high capacity. Both activities increased transiently during cultivation peaking after 72 and 48 hr, respectively. After 72 hr, apparent kinetic constants could be calculated for the sodium dependent (Krn = 13.6 mM; Vmax = 388 nmoles/min/mg protein) and for the sodium-independent system. (Km = 10.8 mM; Vma,x= 132 nmoles/min/mg protein). The transient increase of both transport systems was suppressed by dexamethasone. The sodium-dependence showed a threshold concentration of about 35 mM sodium. Inhibition by kainate was much less potent for BEC than for hepatocytes. These data indicate that BEC contain transport systems for glutamate different from those in hepatocytes and which may be involved in the intrahepatic reabsorbtion of glutamate from bile. INTRODUCTION Biliary epithelial cells (BEC) are small, cuboidal cells which line the intrahepatic bile ducts. It is known from morphological studies that BEC exhibit typical features of secretory as well as absorbing epithelia, and they are therefore thought to play a major role in modifying the primary bile secreted by hepatocytes. During the last few years appropriate methods to obtain
1. Direct all correspondence to: Dr. R. Gebhardt, Physiologisch-chemisches Institut, HoppeSeyler-Strasse 4, D-7400 Ttibingen 1, Germany. Tel: 49/7071/293357. Fax: 49/7071/293361. 2. Key words: cytokeratin 19, gamma-glutamyl transferase, dexamethasone, cholehepatic shunt. 3. Abbreviations: BEC, biliary epithelial cells; DMEM, Dulbecco's Modified Eagle's Medium; GGT, gamma-glutamyl transferase; Dex, dexamethasone; Glu, glutamate; N-Me-AIB, N-methylaminoisobutyrate; Hep, hepatocytes FBS, Fetal bovine serum. Cell Biology and Toxicology, Vol. 7, No. 4, pp. 315-325 Copyright © 1991 Princeton Scientific Publishing Co., Inc. ISSN: 0742-2091
316 Eisenmann-Tappe et al.
reasonably pure preparations of these cells have been established (Parola et al., 1988; Ishii et al., 1989; Alpini et al., 1989). Thus, isolated and cultivated BEC are now available as an in vitro model for studies of their biochemical properties as well as those of secretion and absorption. Besides cholesterol and bile salts, bile contains an abundance of organic solutes, including sugars, proteins, and free amino acids. Whereas the occurrence of fluid-phase endocytosis in BEC has recently been reported (Ishii et al., 1990), very little information exists to date about specific transport systems in BEC that could be involved in ductular absorption of solutes from bile. Such a process has been shown for glucose (Guzelian and Boyer, 1974; Olson and Fujimoto, 1980), and a "cholehepatic shunt pathway" has been postulated for several unconjugated mono- and dihydroxy bile acids (Yoon et al., 1986). Among the free amino acids in bile, L-glutamate is the most abundant (Ballatori et al., 1986). It is liberated from glutathione within the biliary tree by the action of gamma-glutamyl transferase, an enzyme located in the luminal membranes of BEC. In the present study we have characterized the uptake of L-glutamate in primary cultures of BEC derived from normal rats, and its behavior under different culture conditions. MATERIALS AND METHODS Animals. Male Sprague-Dawley rats (240-300g) were used as cell donors. They were kept in a controlled 12 hr-light and -dark cycle on a standardized diet of Alma H 1003 (Botzenhardt, Kempten, Germany) and tap water ad libitum. Isolation and cultivation of biliary epithelial cells. Biliary epithelial cells were isolated by combining the collagenase perfusion technique of Seglen (1976) and a modification of the procedures described by Parola et al. (1988) and Gall and Bhathal (1985).
After liver perfusion with 0.2% collagenase (No. 17446, Serva, Heidelberg) and 0.1% hyaluronidase (Sigma, Mtinchen) for 10 min, the liver capsule was opened and the great majority of hepatocytes was removed by gently shaking the liver in 100 ml of "isolation medium" (Hank's solution containing 2% glucose, (Seglen, 1976)). The remaining tissue containing the bile ducts was minced using scalpels and transferred to a small stoppered flask containing 25 ml of 0.3% trypsin (Serva, Heidelberg) in isolationmedium to allow dispersion of bile ducts. The flask was agitated continuously at room temperature for 40 min. During this period, the duct material was dispersed several times using a pasteur pipette. 25 ml of 0.3% trypsin-inhibitor in isolation medium was then added and the suspension was filtered through a 100 ~ mesh nylone gauze to remove clumped material. Following centrifugation (300g, 3 min) the cells were resuspended in 2 ml of isolation medium and loaded onto a discontinuous Percoll gradient (Pharmacia, Freiburg), prepared exactly as described by Gall and
Cell Biology and Toxicology, Vol. 7, No. 4, 1991
317
Bhathal (1985). The sample was centrifuged at 1,800 g for 20 min in a swing-out rotor (Labofuge III, Heraeus-Christ, Osterode). Cells at the 1.06-1.08 density interface were routinely collected and washed once in 50 ml of Hank's solution in order to remove the percoll. The cells were then transferred to a 90 mm culture plate precoated with a thin film of collagen as described by Gebhardt and Jung (1982) and incubated for 60 min in Dulbecco' s Modification of Eagle's medium (DMEM, Flow Laboratories) containing 10% FBS. This procedure allows the sequestration of remaining hepatocytes, Kupffer cells, and other contaminating cells since they adhere much more quickly to the substratum than do BEC. After the incubation period the medium was collected by gently shaking the culture plate and regained BEC were counted using the trypan blue exclusion test. The cells were resuspended in DMEM containing 10% FBS, 2 mM L-glutamine, penicillin (50 U/ml), and streptomycin (50 gg/ml) at the density of approx. 0.2 x 10 6 cells/ml and inoculated into costar 24-well cluster trays (0.2 ml/well) for transport measurements. For all other experiments cells were cultured at the same cell density. All supporting material was precoated with collagen (see above). Cultures were incubated at 37 °, 90% humidity, and 5% CO2 in air for the periods described for the different experiments. The medium was renewed daily.
Histochemical staining. Histochemimcal staining for gamma-glutamyl transferase was performed on methanol-fixed cell smears of freshly isolated cells using the method of Rutenburg et al. (1969). Cells were washed with saline prior to staining and incubated with the reaction mixture for 30 min to develop the dye. Immunocytochemistry. For identification of the cells we used a specific antiserum to cytokeratin 19 (Arnersham, Braunschweig) which has been shown to bind specifically to BEC in liver sections (Moll et al., 1982). Immnnohistochemical staining was performed on cultured cells fixed with methanol at -20*C for 6 rain using a modification (Gebhardt et al., 1.986) of the unlabeled antibody peroxidase-anti-peroxidase method by Sternberger et al. (1976). Glutamate transport. Glutamate uptake was measured using the cluster-tray method of Gazzola et al. (1981) in a modification described by Gebhardt and Kleeman (1987). Uptake rates were determined at an incubation interval of 1 rain in a modified Krebs-Ringer bicarbonate buffer (pH 7.4) containing sodium or choline, respectively. The Na+-dependent transport of 0.1 mM L-(U-a4C)-glutamate (Amersham Buchler) was determined as the cysteate-inhibitable part of glu uptake after substraction of the Na+-independent transport values. Na+-independent uptake was determined in the absence of specific inhibitors. Uptake rates are given in nmoles/min/mg protein. Statistical evaluation of the data was performed by Student's t-test. RESULTS
AND DISCUSSION
Cell isolation and cell culture. Cell yield usually ranged between 3 and 5 x 106 cells, and the cell viability estimated by the trypan blue exclusion test was always > 95%. Application of
318
Eisenmann-Tappe et al.
hyaluronidase during the isolation procedure proved to be necessary since satisfactory cells isolates were not obtained in its absence. Within the first 4 hr after isolation attachment of the cells to the collagen coated plates could not be observed. Even after 24 hr of cultivation the cells were only loosely adhering to the subslratum (Figure 1). When the cells had flattened completely during day 2 of cultivation, they proliferated rapidly usually forming a confluent monolayer after 96 hr of cultivation. The cells could be cultured at good viability for up to 120 hr.
FIGURE 1. Immunocytochemical detection of cytokeratin 19 by means of the unlabeled peroxidase-antiperoxidase method in BEC cultured for 24 (A) and 72 hr (B). Contaminating hepatocytes (arrow) do not stain. Magnification x 230.
Cell identification. Gamma-glutamyl transferase has been shown to be localized on the canalicular membranes of BEC. Methanol-fixed cells smears of the freshly isolated cells were stained for GGT using the histochemical procedure of Rutenburg et al. Usually more than 95% of the cells were positive for GGT at the end of the isolation procedure. As an even more specific marker for BEC we tested the presence of cytokeratin 19. Unlike GGT, which is also present in some hepatocytes around the portal tract, cytokeratin 19 has been shown to be localized exclusively in BEC in normal rat liver (Moll et al., 1982). Immunocytochemical detection of CK-19 was performed in the freshly isolated state, and after 24, 72, and 120 hr of cultivation. The cells were stained homogeneously and usually contained less than 5% of contaminating cells, which were mainly hepatocytes and unidentified fibroblast-like cells (Figure 1). Cytokeratin 19 was expressed constantly up to 120 hr.
Cell Biology and Toxicology, Vol. 7, No. 4, 1991
319
The electon micrographs of freshly isolated BEC show cells still partially attached to each other containing a highly convoluted nucleus, few mitochondria and a great number of vacuoles beneath the plasma membrane (Figure 2 A). Microvilli are numerous and very long. After 24 hr (Figure 2 B) the cells were attached to the substratum, but had not yet flattened. The number of microvilli seemed slightly diminished. Figure 2 C shows cells from a confluent monolayer after 72 hr of cultivation. Along with flattening of the cells the nuclei had assumed a more regular shape. Tight junctions were formed. Most remarkably° only very few and short microvilli could be seen at the cell surface. The loss of microvilli may be due to adaptation to culture conditions. Its influence on membrane functions, in particular substrate transport, is not known.
FIGURE 2. Electron micrographs showing fleshly isolated and cultured BEC. (A) Freshly isolated BEC. The cells in the lower left are still attached to each other by fight junctions (arrow). (B) BEC cultured for 24 hr. The smooth surface on the lower left represents the contact side with the substratum. (C). BEC cultured for 72 hr. Cells have flattened and tight junctions are reformed (arrow). Bars : 2 !am.
320
Eisenmann-Tappe et al. 5
i, ®
U--] - d e x I~R + d e x
A
3
0
i 1,
0 Z4
48
"/2
cultivation period
96
(h)
5
4 O2
I--] - d e x I~q + d e x
]3
3
0
0 24
45
72
96
c u l t i v a t i o n p e r i o d (h) FIGURE 3. Influence of cultivation period and dexamethasone supply on the sodium dependent (A) and the sodium independent (B) uptake of glutamate in cultivated BEC. Values are means + SD from 3-5 different experiments. Glutamate transport. The uptake rates for L-glutamate were measured after 24, 48, 72, and 96 hr of cultivation. The results are given in Figures 3A and 3B. As can be seen, a sodium-dependent as well as a sodium-independent transport system for glutamate existed in BEC, and transport values for both systems depended on the cultivation period. Na÷-dependent specific uptake rates are given in Figure 3A (open bars). Values increased simultaneously with the proliferation of the cells up to 72 hr and declined thereafter when the cultures reached confluency.
Cell Biology and Toxicology, Vol. 7, No. 4, 1991
321
Comparable data were obtained for the sodium-independent transport of glutamate (Figure 3 B, open bars), with a somewhat different induction pattern. Maximum transport rates were measured after 48 hr and values did not change significantly thereafter. The sodium-dependent transport system generally amounted to about 40% of the total glutamate uptake of the cells, except for the extreme 72 hr-induction maximum, when its part increased even to 66%. This behavior is in marked contrast to glutamate transport in hepatocytes. Generally, uptake rates are very much lower in cultured hepatocytes (Na+-dependent: 0.3+0.1 nmoles/min/mg; Na+-independent: 0.1+0.03 nmoles/min/mg; N=4) than in cultured BEC (Na+-dependent: 3.0+0.03 nmoles/min/mg; Na+-independent: 1.3+0.7 nmoles/min/mg; N=5) when measured under the same culture conditions. It should be noted, however, that hepatocytic glutamate uptake is highly zonated (Burger et al., 1989; Stoll et al., 1991), particularly the sodium-dependent glu uptake (Gebhardt, 1989). Furthermore, uptake via the sodium-dependent transport system in isolated hepatocytes amounts only to about 30% of total glu uptake (Gebhardt and Mecke, 1983; Burger et al., 1989), but can be significantly induced by dexamethasone (dex) during cultivation, while the sodium-independent transport of glu in hepatocytes is not affected by dex. We therefore tested the effect of dex on the glu transport systems in BEC. The cultivation medium was supplemented by 10-7 mM dex from the beginning of cultivation and transport rates were measured under the same conditions. Surprisingly, dex did not have an inductory effect on glu transport in BEC. Instead both transport systems were simultaneously reduced by dex (Figures 3 A and 3 B, hatched bars). Sodium-dependent transport rates decreased to 45-70% of control values without dex, sodium-independent transport to 50-65% of control values (Figures 3 A and 3 B, open bars) except for the data measured after 24 hr of cultivation, when the dex incubation period was obviously too short to be effective. Furthermore, dex did not change the induction pattern of glu uptake during cultivation. Interestingly, the adult rat liver epithelial cell lines (ARL) (Williams and Gunn, 1974) also display a high glutamate transport rate (Gebhardt and Williams, 1986) which suggests a relationship of the clonigenic precursors of these lines (Furukawa et al., 1986) to BEC. Glutamate transport in ARL lines also was not increased by dex (Gebhardt and WiUiams, 1986). To exclude an influence of the glutamate supply on transport rates, we raised the glu concentration in the cultivation medium to 2 mM, which should be about twice the concentration in bile (Ballatori et al., 1986) and incubated the cells up to 96 hr. This did not have any effect on the induction phenomenon, suggesting that in normal BEC, just as in hepatocytes, an adaptive control of glutamate transport does not exist. In order to characterize further the transport systems involved in glutamate uptake in BEC, we tested the inhibitory effect of several L-amino acids on glu transport. L-cysteate, L-aspartate and the cyclic glutamate analogon kainate are known to be potent inhibitors of the sodium dependent transport system G- in hepatocytes. On the other hand, N-methyl-tx-aminoisobutyrate
322
Eisenmann-Tappe et al.
(N-Me-AIB), which is transported via system A, a transport system for short neutral amino acids in hepatocytes, should not have any effect on glu uptake. Glu concentration in these experiments was 0.1 mM, all inhibitors were added at a concentration of 10 mM. Table 1 gives the inhibition rates measured in BEC and hepatocytes after 72 hr of cultivation. Cysteate and aspartate were inhibitors of sodium-dependent and sodium-independent glu transport in BEC, and data are very similar to those obtained in hepatocytes. As expected, N-Me-AIB had only a very weak inhibitory effect. Kainate, however, which is a potent inhibitor of sodium-dependent glu transport in hepatocytes, had only a weak effect on sodium-dependent glu transport and practically did not inhibit the sodium-independent transport in BEC. Obviously, this cyclic analogon is less accepted by the transport systems in BEC which may indicate some differences in the molecular organization of the transport proteins in BEC and hepatocytes.
TABLE 1.
Inhibitor
Inhibition of the uptake of 0.1 mM glutamate by several glutamate analogues given in a concentration of 10 mM each
Inhibition of sodium-dependent uptake (%) BEC
cysteate aspartate kainate N-Me-AIB
93.3 + 93.4 + 33.4 + 9.4 +
15.1 (5) 12.0 (5) 30.3*(5) 11.0 (3)
Inhibition of sodium-independent uptake (%)
Hep
BEC
95.8 + 7.2 (4) 98.5 + 3.7 (4) 98.6 + 2.0 (4) 7.0 (2)
19.5 18.7 5.0 9.4
Hep + 6.5 (5) + 9.6 (5) + 7.1"(5) + 12.3 (3)
19.4 + 9.7 (4) 14.7 + 15.0 (4) 26.3 + 6.8 (4) 0 (2)
BEC = biliary epithelial cells; Hep = hepatocytes; N-Me-AIB = N-methyl-aminoisobutyrate Data are mean values + S.D. The number of experiments is given in parenthesis * statistical evaluation: significantly different from hepatocytes" P
