- Email: [email protected]
Expression of gap junction protein connexin 32 and E-cadherin in human hepatocellular carcinoma
Journal of Hepatology 1995; 22:536-539 Printed in Denmark . All rights reserved
Copyright© Journalof Hepatology1995 Journal of Hepatology
ISSN 0168.8278
Expression of gap junction protein connexm 32 and E-cadherin in human hepatocellular carcinoma Kazuaki Yamaoka 1'3, Toshihiko Nouchi 2, Junichi Tazawa 3, Susumu Hiranuma 4, Fumiaki Marumo 5 and Chifumi Sato 5,6 1Division of Internal Medicine, Hokushin General Hospital, Nagano, 2Division of Gastroenterology, Showa General Hospital, Tokyo, 3Division of Internal Medicine and 4Division of Surgery, Tsuchiura Kyodo Hospital, IbarakL 5Second Department of lnternal Medicine and 6Division of Health Science, Faculty of Medicine, Tokyo Medical and Dental UniversiLv, Tokyo, Japan
The expression of connexin 32, a major gap junction protein, and E-cadherin, an intercellular adhesion molecule that is supposed to be involved in the regulation of gap junctional intercellular communications, was examined immunohistochemically in seven specimens of human hepatocellular carcinoma and surrounding non-carcinomatous tissues. We found that the number of connexin 32-positive spots per mm 2 was significantly less in hepatocellular carcinoma tissues than in the surrounding non-carcinomatous cirrhotic tissues (4360±3390/mm 2 vs 10 030-3690/mm2; p<0.01). The number in the latter
was also significantly less than that in normal controls (23 560_+4170/mm2). E-cadherin was expressed in all non-carcinomatous hepatocytes as well as carcinomatous cells, except for one case of Edmondson's grade III hepatocellular carcinoma. These results suggest an impairment of cell-to-cell communications in human hepatocellular carcinomas.
AP junctions are collections of transmembrane channels that communicate directly between contacting cells, and are considered to maintain tissue homeostasis and control cell growth and differentiation (1). Each gap junction unit is composed of a hexamer of structural protein subunits called connexins (2). The major gap junction protein in human livers is connexin 32 (Cx32) (3). Recently, it has been reported that both Cx32 mRNA and protein decrease significantly during chemical hepatocarcinogenesis in rat livers (4-6): Its expression in human livers is not well known. On the other hand, it has been suggested that E-cadherin, a calcium-dependent intercellular adhesion molecule, is involved in the regulation of gap junctional intercellular communications (7). To compare cell-to-cell communications in tumorous
and non-tumorous tissues, we performed immunohistochemical staining of Cx32 as well as E-cadherin in specimens of human hepatocellular carcinomas (HCC) and surrounding non-carcinomatous cirrhotic liver tissues.
Received 27 May 1994
Correspondence: Chifumi Sato, M.D., Second Department of Internal Medicine, Faculty of Medicine, Tokyo Medical and Dental University, 5-45, Yushima 1-chome, Bunkyoku, Tokyo 113, Japan. 536
Key words: Adhesion molecule; Cell-to-cell communication; Cirrhosis; lmmunohistochemistry. © Journal of Hepatology.
Materials and Methods Seven specimens of human HCC and the surrounding non-carcinomatous liver were obtained by surgical resection. Four samples of normal livers were obtained at surgery from patients with suspected gallbladder carcinoma and from a patient with insulinoma. These samples were mounted in OCT compound (Tissue Tek II, Lab Tek, Miles Lab., Naperville, Illinois, USA) and snap frozen in liquid nitrogen for immunohistochemistry. Histological findings in non-carcinomatous tissues showed postnecrotic cirrhosis in all cases. The cases of HCC included two cases of Edmondson's grade I, four cases of Edmondson's grade II, and one case of Edmondson's grade III. The immunohistochemical detection of Cx32 and E-
Connexin 32 and E-cadherin in human hepatocellular carcinoma
cadherin was performed using an avidin-biotin complex peroxidase technique as described previously (8), with minor modifications. In brief, the H N C buffer (10 mM HEPES (pH 7.4), 150 mM NaC1 and 2 mM CaC12) was used for E-cadherin staining instead of 0.1 M phosphate-buffered saline (PBS) (9). Snap-frozen samples were cut on a cryostat at 7-/zm thickness and then fixed in cold acetone for 10 min. A mouse monoclonal anti-Cx32 antibody (clone 6-3G11, supplied by Nippon Shinyaku Co., Ltd.), which reacts specifically with intact Cx32 (10), was diluted 1:1000 in PBS containing 5% goat serum. A mouse monoclonal anti-human E-cadherin antibody (clone HECD-1, Takara biochemicals Co.) (11) was diluted 1:800 in H N C buffer containing 5% goat serum. Sections were incubated with these primary antibodies for 1 h. In the second step, biotin-conjugated goat antimouse IgG (1:100, Sigma) was applied on the sections for 30 min. In the final step, an avidin-biotin complex solution (Vectastain, Burlingame, California) was applied for 30 min, and the color was developed in freshly prepared Tris buffer (pH 7.6) containing 0.003% H202, 0.02% 3,3'diaminobenzidine, and 0.04% nickel chloride for 2 min. Each incubation step was done at room temperature and followed by 15-min washes with PBS. The sections were then counterstained with methyl green. Negative controls, with the application of control mouse ascites in place of the primary antibodies, were
employed, which uniformly demonstrated no reactions. The number of Cx32-positive spots/mm 2 in HCC tissues and in surrounding non-carcinomatous cirrhotic tissues was counted at random in 15 photographic fields (magnificationX400). Values are expressed as means+_SD and statistical significances were assessed by Student's paired t-test.
Results Macular Cx32-positive spots were observed in cirrhotic tissues at intercellular borders of hepatocytes (Fig. 1A). Gap junction plaques in HCC tissues were markedly decreased compared with those in surrounding non-carcinomatous liver tissues (Fig. 1B). No apparent tendency was observed among the differentiation of HCC. The number of gap junction plaques in HCC tissues was significantly less than that in surrounding non-carcinomatous livers with cirrhosis (4360_ + 3390 vs 10 030_+3690 spots/mm2; n=7, p<0.01) (Fig. 2). In normal livers (Fig. 1C), the number of gap junction plaques was 23 560+_4170 spots/mm z (n=4, p<0.05 compared with livers with cirrhosis). All non-carcinomatous hepatocytes expressed Ecadherin at the cell-to-cell boundary (Fig. 3A). HCC cells also expressed E-cadherin, except for one case with Edmondson's grade III HCC (Fig. 3B).
Fig. 1. Immunoperoxidase staining of connexin 32 in the surrounding cirrhotic liver ( A ), Edmondson "sgrade III hepatocellular carcinoma (B), and normal liver (C). (A): Macutar connexin 32-positive spots are observed at the intercellular border of hepatocytes, but are absent on the sinusoidal surface (x400). (B): The number of gap junction plaques is markedly decreased (×400). ( C) : Many plaques are observed at the intercellular border of hepatocytes ( x 400). 537
K. Yamaoka et al.
Discussion
(xlO 3/ram z ) 14-
1:>.
10-
N e~ X
6.
C 0
~J
4
i
I~l~toe~lta~ Cirrhotic Tissues
Fig. 2. A quantitative analysis of connexin 32-positive spots in human hepatocellular carcinoma and in the surrounding cirrhotic liver. The number of gap junction plaques in hepatocellular carcinoma tissues was significantly lower than that in surrounding non-carcinomatous cirrhotic" livers (4360+_3390 vs 10 030+_3690 spots/ram2; p
In the present study, we have shown that the expression of Cx32 is decreased in human HCC tissues compared with surrounding non-carcinomatous cirrhotic tissues. Furthermore, the expression of Cx32 in cirrhosis appears to be decreased compared with that in normal livers. On the other hand, E-cadherin is expressed not only in non-carcinomatous cirrhotic tissues but also in HCC cells, except for poorly differentiated HCC. Since Lowenstein & Kanno used electrophysiologic techniques to demonstrate a lack of communications between rat liver cancer cells (12), interest has been focused on changes in cell-to-cell communications in carcinoma tissues. In rats, decreased levels of liver gap junction proteins have been observed after partial hepatectomy and in acute liver injury induced by hepatotoxic chemicals such as dimethylnitrosamine and CC14 (13), suggesting a close association of the decreased expression of gap junction proteins with liver regeneration. Cx32 mRNA and protein are also decreased significantly during chemical hepatocarcinogenesis in rats (4-6). Cx32 is the major gap junction component in human livers. However, few studies have been carried out in human HCC. Oyamada et al. (14) reported that Cx32 mRNA levels and Cx32-positive spots in human HCC did not differ from those in the surrounding livers. The present study has clearly demonstrated that the number of Cx32-positive spots in HCC tissues was significantly decreased compared with that in the surrounding cirrhotic tissues. No apparent tendency was observed among the grade of differentiation. We used a mouse monoclonal antibody (63G11) that has been shown to react specifically with
O
Fig. 3. Immunoperoxidase staining of E-cadherin in the surrounding noncancerous tissue ( A ) and in hepatocellular carcinoma, Edmondson grade III (B). (A): Hepatocytes express Ecadherin at the cell-to-cell boundary (x400). (B): Cancer cells do not express E-cadherin (×400). 538
Connex& 32 and E-cadher& & human hepatocellular carcinoma
intact CX32 (10), whereas O y a m a d a et al. used rabbit anti-J-peptide antisera against Cx32 (15). The different source and the specificity of the primary antibody against Cx32 may account for these differences. G a p junction channels permit direct cell-to-cell communications, through which cells exchange ions and small molecules; they play an important role in metabolic cooperation, regulations of cell growth and differentiation (1). The present findings support the suggestion that cell-to-cell communications are impaired in human HCC. It is further suggested that cell-to-cell communications are also impaired in cirrhosis. In the present study, the expression of Cx32 was investigated by detecting gated plaques at the cell-to-cell b o u n d a r y immunohistochemically. It may also be necessary, however, to study the a m o u n t of gap junction proteins. E-cadherin is a Ca2+-dependent intercellular adhesion molecule that is expressed in epithelial cells. Shimoyama & Hirohashi reported a loss of Ecadherin expression in Edmondson's grade IV H C C (9). Recently, in mouse epidermal cells, it has been shown that E-cadherin may be involved in the regulation of gap junction function, and that cells may need to recognize each other through adhesion molecules before they can form functional gap junctions (7). In the present study, E-cadherin was stained diffusely at the cell-to-cell boundary. Ecadherin was demonstrated on the surface of not only non-carcinomatous cells but also carcinomatous cells, except for one case o f Edmondson's grade III HCC. Since the decrease in the number o f gap junctional spots in H C C was observed even in the presence of E-cadherin at the cell-to-cell boundary, the regulatory role of E-cadherin may not be important in these cases. In conclusion, the expression of Cx32 is decreased in h u m a n H C C tissues, suggesting impaired direct cellto-cell communications between H C C cells. The regulatory role of E-cadherin in the expression of Cx32 needs to be further clarified.
Acknowledgements We thank Miss Mieko G o t o for her excellent technical assistance.
References 1. Bennet MVL, Barrio LC, Bargiello TA, Spray DC, Hertzberg E, Saez JC. Gap junctions: new tools, new answers, new questions. Neuron 1991; 6: 305-20. 2. Upwin P, Zampighi G. Structure of the junction between communicating cells. Nature 1980; 283: 545-9. 3. Nicholson B, Dermietzel R, Teplow D, Traub O, Willecke K, Revel J. Two homologous protein components of hepatic gap junctions. Nature 1987; 329: 732-4. 4. Fitzgerald D J, Mesnil M, Oyamada M, Tsuda H, Ito N, Yamasaki H. Changes in gap junction protein (connexin 32) gene expression during rat liver carcinogenesis. J Cell Biochem 1989; 41: 97-102. 5. Neveu M J, Hully JR, Paul DL, Pitot HC. Reversible alteration in the expression of the gap junctional protein connexin 32 during tumor promotion in rat liver and its role during cell proliferation. Cancer Commun 1990; 2: 21-31. 6. Krutovskikh VA, Oyamada M, Yamasaki H. Sequential changes of gap-junctional intercellular communications during multistage rat liver carcinogenesis: direct measurement of communication in vivo. Carcinogenesis 1991; 12: 1701-6. 7. Jongen WMF, Fitzgerald DJ, Asamoto M, Piccoli C, Slaga TJ, Gros D, Takeuchi M, Yamasaki H. Regulation of connexin 43-mediated gap junctional intracellular communication by Ca2+ in mouse epidermal cells is controlled by E-cadherin. J Cell Biol 1991; 114: 545-55. 8. Yamaoka K, Nouchi T, Marumo E Sato C. Alpha-smoothmuscle actin expression in normal and fibrotic human livers. Dig Dis Sci 1993; 38: 1473-9. 9. Shimoyama Y, Hirohashi S. Cadherin intercellular adhesion molecule in hepatocellular carcinomas: loss of E-cadherin expression in an undifferentiated carcinoma. Cancer Lett 1991; 57: 131-5. 10. Takeda A, Kanoh M, Shimazu T, Takeuchi N. Monoclonal antibodies recognizing different epitopes of the 27-kDa gap junction protein from rat liver. J Biochem 1988; 104: 901-7. 11. Shimoyama Y, Hirohashi S, Hirano S, Noguchi M, Shimosato Y, Takeuchi M, Abe O. Cadherin cell-adhesion molecules in human epithelial tissues and carcinomas. Cancer Res 1989; 49: 2128-33. 12. Loewenstein WR, Kanno Y. Intercellular communication and the control of tissue growth: lack of communication between cancer cells. Nature 1966; 209: 1248-9. 13. Miyashita T, Takeda A, Iwai M, Shimazu T. Single administration of hepatotoxic chemicals transiently decreases the gap-junctional-protein levels of connexin 32 in rat liver. Eur J Biochem 1991; 196: 37-42. 14. Oyamada M, Krutovskikh VA, Mesnil M, Partensky C, Franqoise B, Yamasaki H. Aberrant expression of gap junction gene in primary human hepatocellular carcinomas: increased expression of cardiac-type gap junction gene connexin 43. Mol Carcinog 1990; 3: 273-8. 15. Milks L, Kumar N, Houghton R, Upwin N, Gilula N. Topology of the 32-kd liver gap junction protein determined by site-directed antibody localizations. EMBO J 1988; 7: 296775.
539
Copyright© Journalof Hepatology1995 Journal of Hepatology
ISSN 0168.8278
Expression of gap junction protein connexm 32 and E-cadherin in human hepatocellular carcinoma Kazuaki Yamaoka 1'3, Toshihiko Nouchi 2, Junichi Tazawa 3, Susumu Hiranuma 4, Fumiaki Marumo 5 and Chifumi Sato 5,6 1Division of Internal Medicine, Hokushin General Hospital, Nagano, 2Division of Gastroenterology, Showa General Hospital, Tokyo, 3Division of Internal Medicine and 4Division of Surgery, Tsuchiura Kyodo Hospital, IbarakL 5Second Department of lnternal Medicine and 6Division of Health Science, Faculty of Medicine, Tokyo Medical and Dental UniversiLv, Tokyo, Japan
The expression of connexin 32, a major gap junction protein, and E-cadherin, an intercellular adhesion molecule that is supposed to be involved in the regulation of gap junctional intercellular communications, was examined immunohistochemically in seven specimens of human hepatocellular carcinoma and surrounding non-carcinomatous tissues. We found that the number of connexin 32-positive spots per mm 2 was significantly less in hepatocellular carcinoma tissues than in the surrounding non-carcinomatous cirrhotic tissues (4360±3390/mm 2 vs 10 030-3690/mm2; p<0.01). The number in the latter
was also significantly less than that in normal controls (23 560_+4170/mm2). E-cadherin was expressed in all non-carcinomatous hepatocytes as well as carcinomatous cells, except for one case of Edmondson's grade III hepatocellular carcinoma. These results suggest an impairment of cell-to-cell communications in human hepatocellular carcinomas.
AP junctions are collections of transmembrane channels that communicate directly between contacting cells, and are considered to maintain tissue homeostasis and control cell growth and differentiation (1). Each gap junction unit is composed of a hexamer of structural protein subunits called connexins (2). The major gap junction protein in human livers is connexin 32 (Cx32) (3). Recently, it has been reported that both Cx32 mRNA and protein decrease significantly during chemical hepatocarcinogenesis in rat livers (4-6): Its expression in human livers is not well known. On the other hand, it has been suggested that E-cadherin, a calcium-dependent intercellular adhesion molecule, is involved in the regulation of gap junctional intercellular communications (7). To compare cell-to-cell communications in tumorous
and non-tumorous tissues, we performed immunohistochemical staining of Cx32 as well as E-cadherin in specimens of human hepatocellular carcinomas (HCC) and surrounding non-carcinomatous cirrhotic liver tissues.
Received 27 May 1994
Correspondence: Chifumi Sato, M.D., Second Department of Internal Medicine, Faculty of Medicine, Tokyo Medical and Dental University, 5-45, Yushima 1-chome, Bunkyoku, Tokyo 113, Japan. 536
Key words: Adhesion molecule; Cell-to-cell communication; Cirrhosis; lmmunohistochemistry. © Journal of Hepatology.
Materials and Methods Seven specimens of human HCC and the surrounding non-carcinomatous liver were obtained by surgical resection. Four samples of normal livers were obtained at surgery from patients with suspected gallbladder carcinoma and from a patient with insulinoma. These samples were mounted in OCT compound (Tissue Tek II, Lab Tek, Miles Lab., Naperville, Illinois, USA) and snap frozen in liquid nitrogen for immunohistochemistry. Histological findings in non-carcinomatous tissues showed postnecrotic cirrhosis in all cases. The cases of HCC included two cases of Edmondson's grade I, four cases of Edmondson's grade II, and one case of Edmondson's grade III. The immunohistochemical detection of Cx32 and E-
Connexin 32 and E-cadherin in human hepatocellular carcinoma
cadherin was performed using an avidin-biotin complex peroxidase technique as described previously (8), with minor modifications. In brief, the H N C buffer (10 mM HEPES (pH 7.4), 150 mM NaC1 and 2 mM CaC12) was used for E-cadherin staining instead of 0.1 M phosphate-buffered saline (PBS) (9). Snap-frozen samples were cut on a cryostat at 7-/zm thickness and then fixed in cold acetone for 10 min. A mouse monoclonal anti-Cx32 antibody (clone 6-3G11, supplied by Nippon Shinyaku Co., Ltd.), which reacts specifically with intact Cx32 (10), was diluted 1:1000 in PBS containing 5% goat serum. A mouse monoclonal anti-human E-cadherin antibody (clone HECD-1, Takara biochemicals Co.) (11) was diluted 1:800 in H N C buffer containing 5% goat serum. Sections were incubated with these primary antibodies for 1 h. In the second step, biotin-conjugated goat antimouse IgG (1:100, Sigma) was applied on the sections for 30 min. In the final step, an avidin-biotin complex solution (Vectastain, Burlingame, California) was applied for 30 min, and the color was developed in freshly prepared Tris buffer (pH 7.6) containing 0.003% H202, 0.02% 3,3'diaminobenzidine, and 0.04% nickel chloride for 2 min. Each incubation step was done at room temperature and followed by 15-min washes with PBS. The sections were then counterstained with methyl green. Negative controls, with the application of control mouse ascites in place of the primary antibodies, were
employed, which uniformly demonstrated no reactions. The number of Cx32-positive spots/mm 2 in HCC tissues and in surrounding non-carcinomatous cirrhotic tissues was counted at random in 15 photographic fields (magnificationX400). Values are expressed as means+_SD and statistical significances were assessed by Student's paired t-test.
Results Macular Cx32-positive spots were observed in cirrhotic tissues at intercellular borders of hepatocytes (Fig. 1A). Gap junction plaques in HCC tissues were markedly decreased compared with those in surrounding non-carcinomatous liver tissues (Fig. 1B). No apparent tendency was observed among the differentiation of HCC. The number of gap junction plaques in HCC tissues was significantly less than that in surrounding non-carcinomatous livers with cirrhosis (4360_ + 3390 vs 10 030_+3690 spots/mm2; n=7, p<0.01) (Fig. 2). In normal livers (Fig. 1C), the number of gap junction plaques was 23 560+_4170 spots/mm z (n=4, p<0.05 compared with livers with cirrhosis). All non-carcinomatous hepatocytes expressed Ecadherin at the cell-to-cell boundary (Fig. 3A). HCC cells also expressed E-cadherin, except for one case with Edmondson's grade III HCC (Fig. 3B).
Fig. 1. Immunoperoxidase staining of connexin 32 in the surrounding cirrhotic liver ( A ), Edmondson "sgrade III hepatocellular carcinoma (B), and normal liver (C). (A): Macutar connexin 32-positive spots are observed at the intercellular border of hepatocytes, but are absent on the sinusoidal surface (x400). (B): The number of gap junction plaques is markedly decreased (×400). ( C) : Many plaques are observed at the intercellular border of hepatocytes ( x 400). 537
K. Yamaoka et al.
Discussion
(xlO 3/ram z ) 14-
1:>.
10-
N e~ X
6.
C 0
~J
4
i
I~l~toe~lta~ Cirrhotic Tissues
Fig. 2. A quantitative analysis of connexin 32-positive spots in human hepatocellular carcinoma and in the surrounding cirrhotic liver. The number of gap junction plaques in hepatocellular carcinoma tissues was significantly lower than that in surrounding non-carcinomatous cirrhotic" livers (4360+_3390 vs 10 030+_3690 spots/ram2; p
In the present study, we have shown that the expression of Cx32 is decreased in human HCC tissues compared with surrounding non-carcinomatous cirrhotic tissues. Furthermore, the expression of Cx32 in cirrhosis appears to be decreased compared with that in normal livers. On the other hand, E-cadherin is expressed not only in non-carcinomatous cirrhotic tissues but also in HCC cells, except for poorly differentiated HCC. Since Lowenstein & Kanno used electrophysiologic techniques to demonstrate a lack of communications between rat liver cancer cells (12), interest has been focused on changes in cell-to-cell communications in carcinoma tissues. In rats, decreased levels of liver gap junction proteins have been observed after partial hepatectomy and in acute liver injury induced by hepatotoxic chemicals such as dimethylnitrosamine and CC14 (13), suggesting a close association of the decreased expression of gap junction proteins with liver regeneration. Cx32 mRNA and protein are also decreased significantly during chemical hepatocarcinogenesis in rats (4-6). Cx32 is the major gap junction component in human livers. However, few studies have been carried out in human HCC. Oyamada et al. (14) reported that Cx32 mRNA levels and Cx32-positive spots in human HCC did not differ from those in the surrounding livers. The present study has clearly demonstrated that the number of Cx32-positive spots in HCC tissues was significantly decreased compared with that in the surrounding cirrhotic tissues. No apparent tendency was observed among the grade of differentiation. We used a mouse monoclonal antibody (63G11) that has been shown to react specifically with
O
Fig. 3. Immunoperoxidase staining of E-cadherin in the surrounding noncancerous tissue ( A ) and in hepatocellular carcinoma, Edmondson grade III (B). (A): Hepatocytes express Ecadherin at the cell-to-cell boundary (x400). (B): Cancer cells do not express E-cadherin (×400). 538
Connex& 32 and E-cadher& & human hepatocellular carcinoma
intact CX32 (10), whereas O y a m a d a et al. used rabbit anti-J-peptide antisera against Cx32 (15). The different source and the specificity of the primary antibody against Cx32 may account for these differences. G a p junction channels permit direct cell-to-cell communications, through which cells exchange ions and small molecules; they play an important role in metabolic cooperation, regulations of cell growth and differentiation (1). The present findings support the suggestion that cell-to-cell communications are impaired in human HCC. It is further suggested that cell-to-cell communications are also impaired in cirrhosis. In the present study, the expression of Cx32 was investigated by detecting gated plaques at the cell-to-cell b o u n d a r y immunohistochemically. It may also be necessary, however, to study the a m o u n t of gap junction proteins. E-cadherin is a Ca2+-dependent intercellular adhesion molecule that is expressed in epithelial cells. Shimoyama & Hirohashi reported a loss of Ecadherin expression in Edmondson's grade IV H C C (9). Recently, in mouse epidermal cells, it has been shown that E-cadherin may be involved in the regulation of gap junction function, and that cells may need to recognize each other through adhesion molecules before they can form functional gap junctions (7). In the present study, E-cadherin was stained diffusely at the cell-to-cell boundary. Ecadherin was demonstrated on the surface of not only non-carcinomatous cells but also carcinomatous cells, except for one case o f Edmondson's grade III HCC. Since the decrease in the number o f gap junctional spots in H C C was observed even in the presence of E-cadherin at the cell-to-cell boundary, the regulatory role of E-cadherin may not be important in these cases. In conclusion, the expression of Cx32 is decreased in h u m a n H C C tissues, suggesting impaired direct cellto-cell communications between H C C cells. The regulatory role of E-cadherin in the expression of Cx32 needs to be further clarified.
Acknowledgements We thank Miss Mieko G o t o for her excellent technical assistance.
References 1. Bennet MVL, Barrio LC, Bargiello TA, Spray DC, Hertzberg E, Saez JC. Gap junctions: new tools, new answers, new questions. Neuron 1991; 6: 305-20. 2. Upwin P, Zampighi G. Structure of the junction between communicating cells. Nature 1980; 283: 545-9. 3. Nicholson B, Dermietzel R, Teplow D, Traub O, Willecke K, Revel J. Two homologous protein components of hepatic gap junctions. Nature 1987; 329: 732-4. 4. Fitzgerald D J, Mesnil M, Oyamada M, Tsuda H, Ito N, Yamasaki H. Changes in gap junction protein (connexin 32) gene expression during rat liver carcinogenesis. J Cell Biochem 1989; 41: 97-102. 5. Neveu M J, Hully JR, Paul DL, Pitot HC. Reversible alteration in the expression of the gap junctional protein connexin 32 during tumor promotion in rat liver and its role during cell proliferation. Cancer Commun 1990; 2: 21-31. 6. Krutovskikh VA, Oyamada M, Yamasaki H. Sequential changes of gap-junctional intercellular communications during multistage rat liver carcinogenesis: direct measurement of communication in vivo. Carcinogenesis 1991; 12: 1701-6. 7. Jongen WMF, Fitzgerald DJ, Asamoto M, Piccoli C, Slaga TJ, Gros D, Takeuchi M, Yamasaki H. Regulation of connexin 43-mediated gap junctional intracellular communication by Ca2+ in mouse epidermal cells is controlled by E-cadherin. J Cell Biol 1991; 114: 545-55. 8. Yamaoka K, Nouchi T, Marumo E Sato C. Alpha-smoothmuscle actin expression in normal and fibrotic human livers. Dig Dis Sci 1993; 38: 1473-9. 9. Shimoyama Y, Hirohashi S. Cadherin intercellular adhesion molecule in hepatocellular carcinomas: loss of E-cadherin expression in an undifferentiated carcinoma. Cancer Lett 1991; 57: 131-5. 10. Takeda A, Kanoh M, Shimazu T, Takeuchi N. Monoclonal antibodies recognizing different epitopes of the 27-kDa gap junction protein from rat liver. J Biochem 1988; 104: 901-7. 11. Shimoyama Y, Hirohashi S, Hirano S, Noguchi M, Shimosato Y, Takeuchi M, Abe O. Cadherin cell-adhesion molecules in human epithelial tissues and carcinomas. Cancer Res 1989; 49: 2128-33. 12. Loewenstein WR, Kanno Y. Intercellular communication and the control of tissue growth: lack of communication between cancer cells. Nature 1966; 209: 1248-9. 13. Miyashita T, Takeda A, Iwai M, Shimazu T. Single administration of hepatotoxic chemicals transiently decreases the gap-junctional-protein levels of connexin 32 in rat liver. Eur J Biochem 1991; 196: 37-42. 14. Oyamada M, Krutovskikh VA, Mesnil M, Partensky C, Franqoise B, Yamasaki H. Aberrant expression of gap junction gene in primary human hepatocellular carcinomas: increased expression of cardiac-type gap junction gene connexin 43. Mol Carcinog 1990; 3: 273-8. 15. Milks L, Kumar N, Houghton R, Upwin N, Gilula N. Topology of the 32-kd liver gap junction protein determined by site-directed antibody localizations. EMBO J 1988; 7: 296775.
539