- Email: [email protected]
Development of gap junctional channels and intercellular communication in rat liver during ontogenesis
Journal of Hepatology2000; 32: 11-18 Printed in Denmark . AN rights reserved Mtmksgaard Copenhagen
Copyright 0 European Association for the Study of the Liver 2000 Journal of Hepatology ISSN 0168-8278
Development of gap junctional channels and intercellular communication in rat liver during ontogenesis Masaki Iwai’, Yoshinori
Harada’, Fumitaka
Akira Muramatsu’, Katoh2, Toshifumi
Saiyu Tanaka’, Takahiro Mori’, Takeshi Okanoue’, 0hkusa3 and Kei Kashima’
‘Third Department of Internal Medicine, Kyoto Prefectural University of Medicine, Kamigyo-ku and 2Research Laboratory, Nippon Shinyaku Co., Minami-ku, Kyoto and 3First Department of Internal Medicine, Tokyo Medical and Dental University, School of Medicine, Bunkyo-ku, Tokyo, Japan
Backgrounac/Aims: We investigated the expression of connexin (Cx) 32 and 26 subunit proteins of the gap junction (GJ) in the rat liver during ontogenesis to clarify their roles in control of growth and differentiation, and observed their channels in association with development of gap junctional ihtercellular communication (GJIC). Methods: The expression of Cx32 and 26 in prenatal and postnatal livers was examined by Western blot and hnmunofluorescence. GJ channels were investigated not only by double immunofluorescence study but also by immunogold electron microscopy. The spread of lucifer yellow 5 min after its microinjection was examined in the cultured liver tissues. Results: 1) Western blot shawed the expression of both Cx from the late stage of gestation and their peak a week after birth. 2) Cx32- or 26-positive plaques were scattered on hepatocytes of the fetal liver and some of them were colocalized, both were in-
G
which constitute adherent regions of plasma membranes, have direct intercellular communication (1,2) and in addition, ions, second messengers and small molecules move through junctional channels (2,3). It has been suggested that cellto-cell communication through junctional channels may play a crucial role in control of growth, development and differentiation in multicellular organisms (l&6). However, it has not yet been clarified at what AP JUNCTIONS,
Received
19 March; revised 19 July; accepted 27 July 1999
Correspondence:
Masaki Iwai, Third Department of Internal Medicine, Kyoto Prefectural University of Medicine, Kamigyo-Ku, Kyoto, 602-0841, Japan. Tel: 81 75 251 5519. Fax: 81 75 2510710. e-mail: [email protected]
creased just after birth. On day 7 after birth, Cx32positive plaques were present on all hepatocytes within a lobule, and Cx26-positive plaques were distributed in the periportal area. 3) Double-immunogold electron microscopy just after birth showed that most GJ channels were homotypic type of Cx32 or 26, and that few were heterotypic. On day 7 after birth, most channels had the homotypic type of type of Cx32 in the middle and pericentral areas, and there was a heterotypic type of Cx32 and 26 in the periportal area. 4) The dye transfer of lucifer yellow showed a wider spread in the liver tissues on day 7 after birth than on day 1. Conclusion: Increased GJ formation and compatibility or incompatibility of GJ channels are closely associated with development of GJIC, and GJIC may develop at cytodifferentiation during ontogenesis. Key words: Connexin 26; Connexin 32; Cytodifferentiation; Gap junction; Liver; Ontogenesis.
time point gap junctions are formed in the development of liver during ontogenesis or when gap junctional intercellular communication (GJIC) is developed. Gap junctions are made up of 6 units of connexin (Cx) with a hydrophobic channel (7,8). In the liver, at least two homologous Cx molecules, Cx32 and 26, are present. Cx26 is distributed mainly in the periportal area of rat and human adult liver lobules (9,10), whereas Cx32 is evenly spread throughout the lobules (9,ll). In the present study, to examine the formation of gap junctional channels in the development of the rat liver, we investigated the expressions of Cx32 and 26 in the liver during ontogenesis, by Western blot and immunofluorescence analyses, using antibodies to Cx32 (12) and 26 (8,13). To determine whether gap junctional channels are homotypic or heterotypic in
11
M. Zwai et al.
the liver during ontogenesis and whether connexons are homomeric or heteromeric (14) and whether GJIC has developed, we investigated Cx32 and 26 molecules at the plaque level by double immunofluorescence and double-immunogold electron microscopic studies, and we observed the dye transfer of lucifer yellow.
Materials and Methods Five prenatal rats of the Wistar strain were sacrificed on days 15, 16, 18, 20 and 21 of gestation, and 5 postnatal rats were sacrificed on days 1, 2, 3, 5, 6, 7, 10, 14 and 21 after birth; adult rats were used as controls.
i) Expressions of Cx32 and Cx26 1) Western blot analysis of Cx32 and 26. Total protein was extracted from the organic phase of homogenized liver tissue by Trizol (GibcoBRL) according to the method of Wu (15,16). Protein precipitant was re-suspended in sample buffer (150 mM NaCl, 2 mM EDTA, 1% Triton-X 100, 0.25% SDS, 1% sodium deoxycholate and 100 mM DTT in 10 mM phosphate buffer pH 7.2). Ten micrograms of protein was fractionated with SDS-PAGE, and transferred onto PVDF membrane (Millipore) using the semi-dry method. The membrane was im-
a
615
G18
Pl
P2
P3
PS
P7
Pi0
A
pi,
b
Gl5
(318
Pl
P2
P3
P5
W
PlO
A
Fig. 1. Western blot of Cx32 and 26 and their quantitative analysis in the rat liver during ontogenesis. a. Monoclonal antibody for Cx32 was specifically reacted on 32 kD protein in the liver during ontogenesis, and antibody for Cx26 was reacted on 26 kD protein. Both of them were gradually increased after birth. b. Cx32 (open asterisk) and 26 (solid asterisk) were slightly expressedfrom the late stage of gestation and they were gradually increased and reached a peak on day 7 or 5 after birth. n=3.
I2
mersed in blocking buffer (5% non-fat dry milk and 100 mM NaCl in 10 mM Tris pH 7.2) for 2 h, and reacted with mouse anti-Cx32 monoclonal antibody (12) (dilution x300) or rabbit anti-Cx26 antibody (Zymed, CA, USA) (dilution X500) for 4 h. Reacted antibodies were detected with alkaline phosphatase-conjugated anti-mouse IgG antibody (Leinco Technologies Inc., USA) (dilution x30) or anti-rabbit IgG antibody (Chemicon International, CA, USA) (dilution x50) using 5-bromo-4-chloro-3-indol phosphate and nitro blue tetrazolium as chromogenic substrates. Immunoreactive bands of Cx32 and 26 were quantitatively analyzed by Macintosh II (Apple Computer, Cupertino, CA, USA) using the software program, NIH Image 1.4 Densitometric Analysis; each Cx value evaluated by the Densitometry was compared to the highest value. 2) Immunofluorescence and double-immunofluorescence microscopy. Liver was frozen in OCT compounds with liquid nitrogen, and frozen liver sections were cut into 6 pm slices by cryostat. Sliced liver tissues on glass were fixed with acetone for 2 h at 4°C and washed with PBS (0.1 M, pH 7.4, phosphate buffered saline). They were incubated in mouse monoclonal antibody to rat connexin 32” solution (dilution Xl00 in PBS) for 48 h at 4°C then with fluorescence isothionate sheep-conjugated anti-mouse IgG (Boehringer Mannheim, Germany) (dilution X30 in PBS) for 8 h at 4°C and the same or serial sections were incubated in rabbit antibody to rat Cx26 (IgG fraction) (13) (provided by Prof. H. Ohta, Kyorin University School of Medicine and Dr. T. Ohkusa, Tokyo Medical and Dental University School of Medicine) solution (dilution X500 in PBS) for 48 h at 4°C then with rhodamine isothionate sheep conjugated anti-rabbit IgG (Boehringer Mannheim, Germany) (dilution X50 in PBS) for 8 h at 4°C. The sections were examined by fluorescent microscopy using filters (UMNIBA, U-MWIG and U-DM-FIfTR, Olympus, Tokyo, Japan). For controls, each primary antibody was replaced by the corresponding fraction of the preimmune sera, and the findings in the double immunofluorescence study were compared with those in the single immunofluorescence study. 3) The double-immunogold electron microscopy. Four rats were sacrificed on day 18 of gestation and day 1 or 7 after birth. Unfixed sections treated with the same immunoreactivity protocol of primary antibody for Cx32 and 26 as that used for immunofluorescence study were incubated with gold (10 nm)-labeled anti mouse IgG (Funakoshi Co., Tokyo, Japan) (dilution x30 in PBS) and gold (5 nm)-labeled anti rabbit IgG (Funakoshi Co.) (dilution X50 in PBS) as the second antibody. They were fixed with 2% 0~0, in PB solution for 45 min at 4°C dehydrated by graded concentrations of ethanol and embedded in Epon 812, and ultrathin sections were examined by transmission electron microscope JEM 200 CX (Japan Electron Company, Tokyo, Japan). Twenty hepatocytes in ultrathin sections of gestational day 18 and day 1 after birth were randomly selected and observed to have plaques containing Cx32 or 26-positive gold particles, and 20 hepatocytes in the central and periportal area on day 7 after birth were observed to contain two different types of gold particles.
ii) Microinjectionldye transfer assay of gap junctional intercellular communication (GJIC) in cultured liver slices All slices (0.4 mm thickness) from the liver of 3 rats each on gestational day 18 and day 1 or 7 after birth were prepared by means of a tissue chopper (The Mickle Laboratory Engineering Co., Surrey, UK) with a parallel blade. They were incubated in a culture medium of Dulbecco’s modified Eagle’s medium with F-12 (No. 11331430, Gibco BRK, Rockville, USA,) supplemented with insulin, dexamethasone, epidermal growth factor, proline and nicotinamide at 37°C for 1 h. Several injections (5-7 per slice) of Lucifer Yellow (LY) (Sigma Chemicals, St. Louis, MO, USA) solution (7% in 0.33 M lithium chloride) were performed by the slow insertion of a loaded microneedle with a constant flow rate into the liver slice at 37°C. Five minutes after the injection, excess dye that did not enter the cells was removed by washing the samples in several changes of PBS for 15 min. Preliminary studies revealed that the dye spread reached a plateau level after 5 min of injection. The injected liver slices were fixed with 4% paraformaldehyde (PH 7.4, diluted with distilled water and NaOH) in PBS, dehydrated by graded concentrations of ethanol,
Gap junctional channels in ontogenetic liver cleared with xylene and embedded in paraffin. Embedded tissues were cut into 5 jnn serial sections and they were observed using an Olympus (Tokyo, Japan) microscope BX50 with epifluorescence equipment. For the estimation of GJIC permeability in the liver, areas of 5 injected spots per slice were selected and micrographed; the pictures were analyzed for measurement of dye-transfer surface area with a Macscope scanner (Mitani Co., Chiba, Japan).
Results i) Expression
of Cx32 and 26
1) Western blot analysis of Cx32 and 26.
Antibody for Cx32 was specifically reacted on 32 kD protein, and the immunoblot was already seen from the late stage of gestation and had increased after birth. Antibody for Cx26 was reacted on 26 kD protein, and the immunoblot was faintly seen from the late stage of gestation, and had also increased just after birth (Fig. la). Densitometry analysis of both Cx32- and 26-positive bands showed a peak on day 7 and 5 after birth and a plateau level (Fig. lb). 2) Immunofluorescence study of Cx32 and Cx26. On
day 18 of gestation, small basophilic hepatocytes were intermingled with hematopoietic cells. The immunofluorescent study showed Cx3Zpositive plaques in some hepatocytes; some of them were aggregated and the large type of Cx32-positive plaque was present (Fig. 2a). Cx26-positive plaques were scattered on some hepatocytes (Fig. 2d). Just after birth, there were few hematopoietic cells; hepatocytes had loose contact with each other, but no lobular formation was seen. Cx3Zpositive plaques were distributed on the cytoplasmic membranes of all hepatocytes; there were few aggregated plaques and the number of large-type Cx32positive plaques was decreased (Fig. 2b). The number of Cx26-positive plaques was also increased (Fig. 2e). On day 7 after birth, the hepatocytes formed a lobule with trabecular structures. The number of Cx32-positive plaques was increased on all hepatocytes, and the numbers of these plaques were almost equal on each hepatocyte. No aggregated plaques were present and
Fig. 2. ImmunoJuorescent study of Cx32 and Cx26 in the rat liver during ontogenesis. a. Cx32-positive plaques were already present on some hepatocytes of the liver on day 18 of gestation, and there were two types, small and large plaques. b. The number of Cx32-positive plaques was increased on all hepatocytes on day I after birth, and the large type was decreased in number. c. Cx32-positive plaques were seen on the plasma membranes of all hepatocytes on day 7 after birth, and they were almost equal in number on each hepatocyte. The large type of plaque was hardly detected. P; portal tract. d. Cx26-positive plaques (small arrow) were present in a scattered pattern on day 18 of gestation. e. Cx26-positive plaques were increased in number on day I after birth. f Cx26-positive plaques were distributed around the portal tract (P) on day 7 after birth. Cental area (arrow head). Bar=50 pm.
13
M. Zwai et al,
I t I I I
there were no large-type plaques (Fig. 2~). Cx26-positive plaques were distributed in the periportal area (Fig. 20.
Fig. 3. Double immunofluorescent study for Cx32 and 26. a. Cx32 (green)- and Cx26 (red)-positive plaques were present at random on each hepatocyte on day 1 after birth; their colocalization (yellow) could be seen in large plaques. b. Cx32-positive plaques (green) on day 7 after birth were distributed in the pericentral (left or right lower corner) and middle area of a lobule; they were colocalized with Cx26-positive plaques (yellow) in the pertportal area. P: portal tract. Bar=50 pm.
Fig. 4. Double immunogold electron microscopy of Cx32 and 26 on day 18 of gestation. There were aggregates of Cx32-positive gold particles (10 nm) in a large plaques. Bar=O.l pm.
Fig. 8. Spreading of Lucifer Yellow dye in the liver 5 min after its microinjection. a. The dye was spread in the liver tissue on day 1 after birth. b. The dye was spread more widely on day 7 after birth compared to day 1. Bar=100 pm
Gap junctional channels in ontogenetic liver
-b Fig. 5. Double immunogold electron microscopy on day I after birth. a. Cx32-positive gold particles (10 nm) were symmetrically seen in submembranous area, and Cx26-positive gold particles (5 nm) (small arrow head) were symmetrically present in other submembranous area of the same hepatocyte. b. In colocalized plaques, a few Cx26-positive gold particles (small arrow head) were mixed at random with many Cx32-positive gold particles. Bar=O.l ,um.
Fig. 6. Double immunogold electron microscopy on day 7 after birth. a. Cx32-positive gold particles (10 nm) were symmetrically seen in submembranous area of opposite hepatocytes in the pericentral or middle area. b. Cx32- and Cx26-positive gold particles (5 nm) were detected symmetrically on opposite hepatocytes of the periportal area. A few Cx32-positive gold particles were mixed with many Cx26positive gold particles and vice versa. Bar=O.I ,um.
”
3) Cx26.
Double
immunofiuorescence
study
of Cx32
4
and
Cx32-
and Cx26-immunoreactivities just after birth were colocalized in some large plaques, and Cx32-and Cx26-positive plaques were independently present (Fig. 3a), as they were on the day of gestation.
=I $3 -ii p 1
Fig. 7. Double immunogold electron microscopic analysis of Cx-positive plaques. Average number of plaque containing Cx32- or Cx26-positive gold particles was increased in a hepatocyte during ontogenesis. The colocalizedplaques containing symmetry of Cx32- and 26-positive gold particle or their mixture were also increased. Numberlcell: the number of Cx-positive plaque per hepatocyte. The number was in MkSE. n=20.
0 G18
D7CZl)
D7(Z3)
microscopic analysis of Cx-positive plaques
hmunogo~~'e~ectron
G18;gestational day Dl;day 1 after birth D7;day 7 after birth Zl;zone 1 Z3;zone 3 15
M. Zwai et al.
On day 7 after birth, Cx3Zpositive plaques were independently distributed in the central area; many Cx32positive plaques in the periportal area were colocalized with Cx26-positive plaques, and colocalized plaques and Cx3Zpositive plaques were observed on hepatocytes in the middle zone (Fig. 3b). Cx26-positive plaques were hardly seen independently. 4) The double-immunogold electron microscopic study of Cx32 and 26. Immunogold electron microscopy on day 18 of gestation revealed aggregates of Cx3Zpositive gold particles in large plaques (Fig. 4). A symmetrical array of Cx32-positive gold particles was observed in plaques on day 1 after birth and the symmetry of small Cx26-positive gold particles was seen in other plaques of the same hepatocyte (Fig. 5a). In colocalized plaques Cx32-positive gold particles were mixed at random with a few Cx26-positive gold particles (Fig. 5b). Seven days after birth, there was symmetrical localization of Cx3Zpositive gold particles in most plaques of the middle or pericentral area (Fig. 6a); there were symmetrical pairs of Cx32- and Cx26positive gold particles in colocalized plaques of the periportal area, and a few Cx32-positive gold particles were mixed at random with many Cx26-positive gold particles and vice versa (Fig. 6b). Immunogold electron microscopic analysis revealed an increase in the number of Cx-positive plaques per hepatocyte from gestational day 18 to day 7 after birth. The number of plaques containing both Cx32- and 26-positive gold particles was increased from gestational day 18 to day 7 after birth, but they could be hardly seen in the pericentral and middle area of lobules on day 7 after birth (Fig. 7). ii) Dye transfer assay in cultured liver tissues
Five minutes after the microinjection of LY into fetal livers, a few hepatocytes were labeled with fluorescence. The injected LY was spread in the liver tissue on day 1 after birth (Fig. 8a). The area of spread of LY was wider on day 7 after birth (Fig. 8b) than on day 1 after birth. LY-positive areas were measured with an image analyzer. The average LY-positive area in the liver was 0.1320.06 and 0.35?0.08* (meanrtSD, n=5) mm2 on day 1 and 7 after birth and there was a significant difference between them (The values were analyzed with Student’s t-test. *p
Discussion In the fetal rat livers small hepatocytes were observed to be intermingled with hematopoietic cells and there were few contacts between hepatocytes. Western blot analysis revealed the expression of Cx32 and Cx26 in the fetal liver from the late stage of gestation, but less 16
frequently than after birth. Cx32- and Cx26-positive plaques were seen on some hepatocytes before birth, and there were Cx3Zpositive plaques of different sizes. Immunogold electron microscopy revealed aggregates of Cx32-positive gold particles in large plaques, but the dye transfer of LY was limited in the fetal liver tissue. Gap junctional channels were aggregated to maintain intercellular communication (GJIC) on limited areas of cell-to-cell contact in the fetal liver. The large type of Cx3Zpositive plaque was reported to be an aggregate of the precursor of the gap junction (17,18), and to be seen at the time of cytodifferentiation (19); the large type detected in the fetal liver may be premature for GJIC, and was not detected in the established lobules. It was observed in the present study that there were few hematopoietic cells just after birth, the hepatocytes were in contact with each other, and expressions of Cx32 and 26 were increased. These results may have been obtained in part because plasma glucagon and glucocorticoid, in establishing portal circulation, may influence the expression of Cx32 and Cx26 mRNA (20). The numbers of Cx32- and Cx26-positive plaques were increased, while the number of aggregated or large-type Cx3Zpositive plaques was decreased. On day 7 after birth, Cx32-positive plaques were diffusely distributed on the plasma membranes of all hepatocytes in a lobule, and no large type was present, as in adult rats. Cx26-positive plaques were seen in the periportal area because glucagon in established portal vein has a greater influence on the expression of Cx26 mRNA than on that of Cx32 mRNA (21), and glucagon induces more Cx26 mRNA in the periportal area than in the pericentral area (22,23). The double immunofluorescence study showed Cx32- and Cx26-positive plaques independently localized or colocalized on the plasma membrane of each hepatocyte on day 18 of gestation or just after birth, and neither plaque was regularly distributed in the liver tissues. Gap junctions are made up of 6 units of connexin, a connexon, with a hydrophobic channel (8), and their channels are homotypic or heterotypic (14). To examine the homotypic or heterotypic type of gap junctional channel and the homomeric or heteromeric connexon at the ultrastructural level, we used double immunogold electron microscopy Cx-positive gold particles were symmetrically detected in plaques on opposite hepatocytes, but there seemed to be a relatively low ratio of antibody binding to Cx. This may be due to the large size of the antibody of IgG molecules compared with the dense packing of Cx proteins (23). In addition, Cx26-positive gold particles could not be detected by double immunogold electron microscopy as expected after the immunoreactive procedure with
Gap junctional channels in ontogenetic liver
Cx32 molecules. On gestational day 18 and just after birth, there were symmetrical pairs of Cx32- or Cx26positive gold particles in plaques, and some plaques contained many Cx32-positive gold particles mixed at random with a few Cx26-positive gold particles. Therefore many gap junctional channels within a plaque are the homotypic type of Cx32 or Cx26 (23) and some of them may be heterotypic or some connexons may be heteromeric (24). On day 7 after birth, Cx3Zpositive plaques were distributed in the pericentral or middle area, and Cx32positive plaques in the periportal area were colocalized with Cx26-positive plaques. Symmetrical pairs of Cx3Zpositive gold particles were observed in plaques of the central and middle areas. Symmetrical pairs of Cx32- and Cx26-positive gold particles were seen in plaques of the periportal area, and a few Cx32-positive gold particles were mixed with many Cx26-positive gold particles in one side of the submembranous area and vice versa. Thus, gap junctional channels in the central or middle area are homotypic and connexons are homomeric, and many gap junctional channels in the periportal area are heterotypic, and connexons are homomeric and a few of them may be heteromeric. GJIC can be estimated by dye transfer assay. LY for the assay may have spread into the extracellular space after its microinjection and its transfer assay may have some validity (25,26). We washed injected liver slices with PBS to eliminate non-specific spread. The dye transfer of LY was limited in the fetal liver because the areas of cell-to-cell contact were small, there were few gap junctional structures, and premature gap junctions were present; however, Cx32 and 26 were expressed. LY is reported to pass through the heterotypic gap junction of Cx32 and Cx26 (27) as well as the homotypic gap junction. The LY injected spread gradually in the liver tissues on day 1 after birth, and it spread more widely on day 7 after birth compared to day 1. Thus, GJIC developed in the liver during ontogenesis because the gap junctions increased in number, or open connexons might have increased. There were differences in the gap junctional structures of hepatocytes in fetal and neonatal livers, and their structures were heterogeneous within established lobules. Cx32 has phosphorylation sites. No reports have shown the presence of different bands in Western blot. Cx26 is devoid of phosphorylated sites (28,29). The closure mechanism in gap junctional channels composed of Cx32 may, therefore, differ from that in gap junctional channels composed of Cx26 or a mixture of Cx32 and Cx26. Thus there may be heterogeneity in the GJIC in each hepatocyte of a fetus or neonate and within a lobule. Our findings suggest that
homotypic or heterotypic incompatibilities of the gap junctions and homomeric or heteromeric interactions may contribute to the development of GJIC in the liver during ontogenesis.
Acknowledgements We thank Dr. Y. Ibata (Department of Anatomy, Kyoto Prefectural University of Medicine), Dr. M. Oyamada (Department of Pathology, Kyoto Prefectural University of Medicine) and Dr. T. Ueda (Research Laboratory, Nippon Shinyaku, Co.) for critical and helpful advice with our research work.
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. Spray DC. Physiological and pharmacological regulation of gap junction channels. In: Citi S, editor. Molecular Mechanisms of Epithelial Cell Junctions: From Development to Disease. Austin, Texas: R.G. Landes Company; 1994. p. 195-215. 3. Dermietzel RD, Hwang TK, Spray DS. The gap junction family: structure, function and chemistry. Anat Embryo1 1990; 182: 5178. 4. Gimlich RL, Kumar NM, Gilula NB. Sequence and development expression of mRNA coding for a gap junction protein in xenopus. J Cell Biol 1988; 107: 1065-73. 5. Guthrie SC, Gilula NB. Gap junctional communication and development. Trends Neurosci 1989; 12: 12-6. 6. Yamasaki H, Naus CC. Role of connexin genes in growth control. Carcinogenesis 1996; 17: 1199-213. 7. Nicholson BJ, Dermietzel R, Teplow D, Traub 0, Willecke K, Revel J-P Two homologous protein components of hepatic gap junctions. Nature 1987; 329: 732-3. 8. Revel JP Nicholson BJ, Yancy SB. Chemistry of gap junctions. Ann Rev Physiol 1985; 47: 263-79. 9. Traub 0, Look J, Dermietzel R, Brummer F, Hulser D, Willecke K. Comparative characterization of the 21-kD and 26-kD gap junction proteins in murine liver and cultured hepatocytes. J Cell Biol 1989; 108: 1039-51. 10. Neveu MJ, Hully JR, Babcock KL, Vaughan J, Hertzberg EL, Nicholson BJ, et al. Proliferation-associated differences in the spatial and temporal expression of gap junction genes in rat liver. Hepatology 1995; 22: 202-12. 11 Dermietzel R, Leibstein A, Frixen U, Janssen-Timmen U, Traub 0, Willecke K. Gap junctions in several tissues share antigenic determinants with liver gap junctions. EMBO J 1984; 3: 226170. 12. 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. 13. Paul DL. Monoclonal cloning of cDNA for rat liver gap junction. J Cell Biol 1986; 103: 123-34. 14. Kumar NM, Gilula NB. The gap junction communication channel. Cell 1996; 84: 381-8. 15. Maniatis T, Fritsch EF, Sambrook J. Molecular Cloning: A Laboratory Manual, 2nd ed. Cold Spring Harbor, NY Cold Spring Harbor Laboratory; 1999: 60(18)-75(18). 16. Wu LC. Isolation and long-term storage of proteins from tissues and cells using Trizol reagent. Focus 1995; 17: 98-100. 17. Decker RS. Hormonal regulation of gap junction. J Cell Biol 1976; 69: 669-85. 18. Johnson R, Hammer M, Sheridan J, Revel J-P Gap junction formation between reaggregated Novikoff hepatoma cells. Proc Nat1 Acad Sci USA 1974; 71: 453640. 19. Yamamoto M, Kataoka K. Large particles associated with gap
17
M. Iwai et al.
20.
21.
22.
23.
24.
25.
18
junctions of pancreatic exocrine cells during embryonic and neonatal development. Anat Embryo1 1985; 171: 305-10. Kojima T, Mitaka T, Shibata Y, Mizoguchi Y. Induction and regulation of connexin by glucagon in primary cultures of adult rat hepatocytes. J Cell Sci 1995; 108: 2771-80. Saez JC, Berthoud VM, Iwanji V Regulation of the abundance of connexin 32 and glucagon receptors in primary cultures of adult rat hepatocytes [abstract]. J Cell Biol 1991; 115: 189A. Berthoud VM, Iwanji V, Garcia AM, Saez JC. Connexins and glucagon receptors during development of rat hepatic acinus. Am J Physiol 1992; 26: 650-8. Kuraoka A, Iida H, Hatae T, Shibata Y, Itoh M, Kurita T. Localization of gap junction proteins, connexin 32 and 26, in rat and guinea pig liver as revealed by quick-freeze, deep-etch immunoelectron microscopy. J Histochem Cytochem 1993; 41: 971-80. Stauffer KA. The gap junction proteins /II-connexin (connexin32) and &connexin (connexin-26) can form heteromeric hemichannels. J Biol Chem 1995; 24: 6768-72. Krutovskih VA, Oyamada M, Yamasaki H. Sequential change of
26.
27.
28.
29.
gap-junctional intercellular communications during multistage rat liver carcinogenesis: direct measurement of communication in viva Carcinogenesis 1991; 12: 17014. Tateno C, Itoh S, Tanaka M, Oyamada M, Yoshitake A. Effect of DDT on hepatic gap junctional intercellular communication in rats. Carcinogenesis 1994; 15: 517-21. Elfang C, Eckert R, Lichtenberg-Frate H, Butterweck A, Traub 0, Klein RA, et al. Specific permeability and selective formation of gap junction channels in connexin-transfected HeLa cells. J Cell Biol 1995; 129: 805-17. Saez JC, Nairn AC, Czernik AJ, Spray DC, Hertzberg EL, Greengard P et al. Phosphorylation of connexin 32, a hepatocyte gap junction protein, by CAMP-dependent protein kinase, protein kinase C and Ca 2+/calmodulin-dependent protein kinase II. Eur J Biochem 1990; 192: 263-73. Zhang J-T, Nicholson BJ. Sequence and tissue distribution of a second protein of hepatic gap junctions, Cx26, deduced from its cDNA. J Cell Biol 1989; 109: 3391-6.
Copyright 0 European Association for the Study of the Liver 2000 Journal of Hepatology ISSN 0168-8278
Development of gap junctional channels and intercellular communication in rat liver during ontogenesis Masaki Iwai’, Yoshinori
Harada’, Fumitaka
Akira Muramatsu’, Katoh2, Toshifumi
Saiyu Tanaka’, Takahiro Mori’, Takeshi Okanoue’, 0hkusa3 and Kei Kashima’
‘Third Department of Internal Medicine, Kyoto Prefectural University of Medicine, Kamigyo-ku and 2Research Laboratory, Nippon Shinyaku Co., Minami-ku, Kyoto and 3First Department of Internal Medicine, Tokyo Medical and Dental University, School of Medicine, Bunkyo-ku, Tokyo, Japan
Backgrounac/Aims: We investigated the expression of connexin (Cx) 32 and 26 subunit proteins of the gap junction (GJ) in the rat liver during ontogenesis to clarify their roles in control of growth and differentiation, and observed their channels in association with development of gap junctional ihtercellular communication (GJIC). Methods: The expression of Cx32 and 26 in prenatal and postnatal livers was examined by Western blot and hnmunofluorescence. GJ channels were investigated not only by double immunofluorescence study but also by immunogold electron microscopy. The spread of lucifer yellow 5 min after its microinjection was examined in the cultured liver tissues. Results: 1) Western blot shawed the expression of both Cx from the late stage of gestation and their peak a week after birth. 2) Cx32- or 26-positive plaques were scattered on hepatocytes of the fetal liver and some of them were colocalized, both were in-
G
which constitute adherent regions of plasma membranes, have direct intercellular communication (1,2) and in addition, ions, second messengers and small molecules move through junctional channels (2,3). It has been suggested that cellto-cell communication through junctional channels may play a crucial role in control of growth, development and differentiation in multicellular organisms (l&6). However, it has not yet been clarified at what AP JUNCTIONS,
Received
19 March; revised 19 July; accepted 27 July 1999
Correspondence:
Masaki Iwai, Third Department of Internal Medicine, Kyoto Prefectural University of Medicine, Kamigyo-Ku, Kyoto, 602-0841, Japan. Tel: 81 75 251 5519. Fax: 81 75 2510710. e-mail: [email protected]
creased just after birth. On day 7 after birth, Cx32positive plaques were present on all hepatocytes within a lobule, and Cx26-positive plaques were distributed in the periportal area. 3) Double-immunogold electron microscopy just after birth showed that most GJ channels were homotypic type of Cx32 or 26, and that few were heterotypic. On day 7 after birth, most channels had the homotypic type of type of Cx32 in the middle and pericentral areas, and there was a heterotypic type of Cx32 and 26 in the periportal area. 4) The dye transfer of lucifer yellow showed a wider spread in the liver tissues on day 7 after birth than on day 1. Conclusion: Increased GJ formation and compatibility or incompatibility of GJ channels are closely associated with development of GJIC, and GJIC may develop at cytodifferentiation during ontogenesis. Key words: Connexin 26; Connexin 32; Cytodifferentiation; Gap junction; Liver; Ontogenesis.
time point gap junctions are formed in the development of liver during ontogenesis or when gap junctional intercellular communication (GJIC) is developed. Gap junctions are made up of 6 units of connexin (Cx) with a hydrophobic channel (7,8). In the liver, at least two homologous Cx molecules, Cx32 and 26, are present. Cx26 is distributed mainly in the periportal area of rat and human adult liver lobules (9,10), whereas Cx32 is evenly spread throughout the lobules (9,ll). In the present study, to examine the formation of gap junctional channels in the development of the rat liver, we investigated the expressions of Cx32 and 26 in the liver during ontogenesis, by Western blot and immunofluorescence analyses, using antibodies to Cx32 (12) and 26 (8,13). To determine whether gap junctional channels are homotypic or heterotypic in
11
M. Zwai et al.
the liver during ontogenesis and whether connexons are homomeric or heteromeric (14) and whether GJIC has developed, we investigated Cx32 and 26 molecules at the plaque level by double immunofluorescence and double-immunogold electron microscopic studies, and we observed the dye transfer of lucifer yellow.
Materials and Methods Five prenatal rats of the Wistar strain were sacrificed on days 15, 16, 18, 20 and 21 of gestation, and 5 postnatal rats were sacrificed on days 1, 2, 3, 5, 6, 7, 10, 14 and 21 after birth; adult rats were used as controls.
i) Expressions of Cx32 and Cx26 1) Western blot analysis of Cx32 and 26. Total protein was extracted from the organic phase of homogenized liver tissue by Trizol (GibcoBRL) according to the method of Wu (15,16). Protein precipitant was re-suspended in sample buffer (150 mM NaCl, 2 mM EDTA, 1% Triton-X 100, 0.25% SDS, 1% sodium deoxycholate and 100 mM DTT in 10 mM phosphate buffer pH 7.2). Ten micrograms of protein was fractionated with SDS-PAGE, and transferred onto PVDF membrane (Millipore) using the semi-dry method. The membrane was im-
a
615
G18
Pl
P2
P3
PS
P7
Pi0
A
pi,
b
Gl5
(318
Pl
P2
P3
P5
W
PlO
A
Fig. 1. Western blot of Cx32 and 26 and their quantitative analysis in the rat liver during ontogenesis. a. Monoclonal antibody for Cx32 was specifically reacted on 32 kD protein in the liver during ontogenesis, and antibody for Cx26 was reacted on 26 kD protein. Both of them were gradually increased after birth. b. Cx32 (open asterisk) and 26 (solid asterisk) were slightly expressedfrom the late stage of gestation and they were gradually increased and reached a peak on day 7 or 5 after birth. n=3.
I2
mersed in blocking buffer (5% non-fat dry milk and 100 mM NaCl in 10 mM Tris pH 7.2) for 2 h, and reacted with mouse anti-Cx32 monoclonal antibody (12) (dilution x300) or rabbit anti-Cx26 antibody (Zymed, CA, USA) (dilution X500) for 4 h. Reacted antibodies were detected with alkaline phosphatase-conjugated anti-mouse IgG antibody (Leinco Technologies Inc., USA) (dilution x30) or anti-rabbit IgG antibody (Chemicon International, CA, USA) (dilution x50) using 5-bromo-4-chloro-3-indol phosphate and nitro blue tetrazolium as chromogenic substrates. Immunoreactive bands of Cx32 and 26 were quantitatively analyzed by Macintosh II (Apple Computer, Cupertino, CA, USA) using the software program, NIH Image 1.4 Densitometric Analysis; each Cx value evaluated by the Densitometry was compared to the highest value. 2) Immunofluorescence and double-immunofluorescence microscopy. Liver was frozen in OCT compounds with liquid nitrogen, and frozen liver sections were cut into 6 pm slices by cryostat. Sliced liver tissues on glass were fixed with acetone for 2 h at 4°C and washed with PBS (0.1 M, pH 7.4, phosphate buffered saline). They were incubated in mouse monoclonal antibody to rat connexin 32” solution (dilution Xl00 in PBS) for 48 h at 4°C then with fluorescence isothionate sheep-conjugated anti-mouse IgG (Boehringer Mannheim, Germany) (dilution X30 in PBS) for 8 h at 4°C and the same or serial sections were incubated in rabbit antibody to rat Cx26 (IgG fraction) (13) (provided by Prof. H. Ohta, Kyorin University School of Medicine and Dr. T. Ohkusa, Tokyo Medical and Dental University School of Medicine) solution (dilution X500 in PBS) for 48 h at 4°C then with rhodamine isothionate sheep conjugated anti-rabbit IgG (Boehringer Mannheim, Germany) (dilution X50 in PBS) for 8 h at 4°C. The sections were examined by fluorescent microscopy using filters (UMNIBA, U-MWIG and U-DM-FIfTR, Olympus, Tokyo, Japan). For controls, each primary antibody was replaced by the corresponding fraction of the preimmune sera, and the findings in the double immunofluorescence study were compared with those in the single immunofluorescence study. 3) The double-immunogold electron microscopy. Four rats were sacrificed on day 18 of gestation and day 1 or 7 after birth. Unfixed sections treated with the same immunoreactivity protocol of primary antibody for Cx32 and 26 as that used for immunofluorescence study were incubated with gold (10 nm)-labeled anti mouse IgG (Funakoshi Co., Tokyo, Japan) (dilution x30 in PBS) and gold (5 nm)-labeled anti rabbit IgG (Funakoshi Co.) (dilution X50 in PBS) as the second antibody. They were fixed with 2% 0~0, in PB solution for 45 min at 4°C dehydrated by graded concentrations of ethanol and embedded in Epon 812, and ultrathin sections were examined by transmission electron microscope JEM 200 CX (Japan Electron Company, Tokyo, Japan). Twenty hepatocytes in ultrathin sections of gestational day 18 and day 1 after birth were randomly selected and observed to have plaques containing Cx32 or 26-positive gold particles, and 20 hepatocytes in the central and periportal area on day 7 after birth were observed to contain two different types of gold particles.
ii) Microinjectionldye transfer assay of gap junctional intercellular communication (GJIC) in cultured liver slices All slices (0.4 mm thickness) from the liver of 3 rats each on gestational day 18 and day 1 or 7 after birth were prepared by means of a tissue chopper (The Mickle Laboratory Engineering Co., Surrey, UK) with a parallel blade. They were incubated in a culture medium of Dulbecco’s modified Eagle’s medium with F-12 (No. 11331430, Gibco BRK, Rockville, USA,) supplemented with insulin, dexamethasone, epidermal growth factor, proline and nicotinamide at 37°C for 1 h. Several injections (5-7 per slice) of Lucifer Yellow (LY) (Sigma Chemicals, St. Louis, MO, USA) solution (7% in 0.33 M lithium chloride) were performed by the slow insertion of a loaded microneedle with a constant flow rate into the liver slice at 37°C. Five minutes after the injection, excess dye that did not enter the cells was removed by washing the samples in several changes of PBS for 15 min. Preliminary studies revealed that the dye spread reached a plateau level after 5 min of injection. The injected liver slices were fixed with 4% paraformaldehyde (PH 7.4, diluted with distilled water and NaOH) in PBS, dehydrated by graded concentrations of ethanol,
Gap junctional channels in ontogenetic liver cleared with xylene and embedded in paraffin. Embedded tissues were cut into 5 jnn serial sections and they were observed using an Olympus (Tokyo, Japan) microscope BX50 with epifluorescence equipment. For the estimation of GJIC permeability in the liver, areas of 5 injected spots per slice were selected and micrographed; the pictures were analyzed for measurement of dye-transfer surface area with a Macscope scanner (Mitani Co., Chiba, Japan).
Results i) Expression
of Cx32 and 26
1) Western blot analysis of Cx32 and 26.
Antibody for Cx32 was specifically reacted on 32 kD protein, and the immunoblot was already seen from the late stage of gestation and had increased after birth. Antibody for Cx26 was reacted on 26 kD protein, and the immunoblot was faintly seen from the late stage of gestation, and had also increased just after birth (Fig. la). Densitometry analysis of both Cx32- and 26-positive bands showed a peak on day 7 and 5 after birth and a plateau level (Fig. lb). 2) Immunofluorescence study of Cx32 and Cx26. On
day 18 of gestation, small basophilic hepatocytes were intermingled with hematopoietic cells. The immunofluorescent study showed Cx3Zpositive plaques in some hepatocytes; some of them were aggregated and the large type of Cx32-positive plaque was present (Fig. 2a). Cx26-positive plaques were scattered on some hepatocytes (Fig. 2d). Just after birth, there were few hematopoietic cells; hepatocytes had loose contact with each other, but no lobular formation was seen. Cx3Zpositive plaques were distributed on the cytoplasmic membranes of all hepatocytes; there were few aggregated plaques and the number of large-type Cx32positive plaques was decreased (Fig. 2b). The number of Cx26-positive plaques was also increased (Fig. 2e). On day 7 after birth, the hepatocytes formed a lobule with trabecular structures. The number of Cx32-positive plaques was increased on all hepatocytes, and the numbers of these plaques were almost equal on each hepatocyte. No aggregated plaques were present and
Fig. 2. ImmunoJuorescent study of Cx32 and Cx26 in the rat liver during ontogenesis. a. Cx32-positive plaques were already present on some hepatocytes of the liver on day 18 of gestation, and there were two types, small and large plaques. b. The number of Cx32-positive plaques was increased on all hepatocytes on day I after birth, and the large type was decreased in number. c. Cx32-positive plaques were seen on the plasma membranes of all hepatocytes on day 7 after birth, and they were almost equal in number on each hepatocyte. The large type of plaque was hardly detected. P; portal tract. d. Cx26-positive plaques (small arrow) were present in a scattered pattern on day 18 of gestation. e. Cx26-positive plaques were increased in number on day I after birth. f Cx26-positive plaques were distributed around the portal tract (P) on day 7 after birth. Cental area (arrow head). Bar=50 pm.
13
M. Zwai et al,
I t I I I
there were no large-type plaques (Fig. 2~). Cx26-positive plaques were distributed in the periportal area (Fig. 20.
Fig. 3. Double immunofluorescent study for Cx32 and 26. a. Cx32 (green)- and Cx26 (red)-positive plaques were present at random on each hepatocyte on day 1 after birth; their colocalization (yellow) could be seen in large plaques. b. Cx32-positive plaques (green) on day 7 after birth were distributed in the pericentral (left or right lower corner) and middle area of a lobule; they were colocalized with Cx26-positive plaques (yellow) in the pertportal area. P: portal tract. Bar=50 pm.
Fig. 4. Double immunogold electron microscopy of Cx32 and 26 on day 18 of gestation. There were aggregates of Cx32-positive gold particles (10 nm) in a large plaques. Bar=O.l pm.
Fig. 8. Spreading of Lucifer Yellow dye in the liver 5 min after its microinjection. a. The dye was spread in the liver tissue on day 1 after birth. b. The dye was spread more widely on day 7 after birth compared to day 1. Bar=100 pm
Gap junctional channels in ontogenetic liver
-b Fig. 5. Double immunogold electron microscopy on day I after birth. a. Cx32-positive gold particles (10 nm) were symmetrically seen in submembranous area, and Cx26-positive gold particles (5 nm) (small arrow head) were symmetrically present in other submembranous area of the same hepatocyte. b. In colocalized plaques, a few Cx26-positive gold particles (small arrow head) were mixed at random with many Cx32-positive gold particles. Bar=O.l ,um.
Fig. 6. Double immunogold electron microscopy on day 7 after birth. a. Cx32-positive gold particles (10 nm) were symmetrically seen in submembranous area of opposite hepatocytes in the pericentral or middle area. b. Cx32- and Cx26-positive gold particles (5 nm) were detected symmetrically on opposite hepatocytes of the periportal area. A few Cx32-positive gold particles were mixed with many Cx26positive gold particles and vice versa. Bar=O.I ,um.
”
3) Cx26.
Double
immunofiuorescence
study
of Cx32
4
and
Cx32-
and Cx26-immunoreactivities just after birth were colocalized in some large plaques, and Cx32-and Cx26-positive plaques were independently present (Fig. 3a), as they were on the day of gestation.
=I $3 -ii p 1
Fig. 7. Double immunogold electron microscopic analysis of Cx-positive plaques. Average number of plaque containing Cx32- or Cx26-positive gold particles was increased in a hepatocyte during ontogenesis. The colocalizedplaques containing symmetry of Cx32- and 26-positive gold particle or their mixture were also increased. Numberlcell: the number of Cx-positive plaque per hepatocyte. The number was in MkSE. n=20.
0 G18
D7CZl)
D7(Z3)
microscopic analysis of Cx-positive plaques
hmunogo~~'e~ectron
G18;gestational day Dl;day 1 after birth D7;day 7 after birth Zl;zone 1 Z3;zone 3 15
M. Zwai et al.
On day 7 after birth, Cx3Zpositive plaques were independently distributed in the central area; many Cx32positive plaques in the periportal area were colocalized with Cx26-positive plaques, and colocalized plaques and Cx3Zpositive plaques were observed on hepatocytes in the middle zone (Fig. 3b). Cx26-positive plaques were hardly seen independently. 4) The double-immunogold electron microscopic study of Cx32 and 26. Immunogold electron microscopy on day 18 of gestation revealed aggregates of Cx3Zpositive gold particles in large plaques (Fig. 4). A symmetrical array of Cx32-positive gold particles was observed in plaques on day 1 after birth and the symmetry of small Cx26-positive gold particles was seen in other plaques of the same hepatocyte (Fig. 5a). In colocalized plaques Cx32-positive gold particles were mixed at random with a few Cx26-positive gold particles (Fig. 5b). Seven days after birth, there was symmetrical localization of Cx3Zpositive gold particles in most plaques of the middle or pericentral area (Fig. 6a); there were symmetrical pairs of Cx32- and Cx26positive gold particles in colocalized plaques of the periportal area, and a few Cx32-positive gold particles were mixed at random with many Cx26-positive gold particles and vice versa (Fig. 6b). Immunogold electron microscopic analysis revealed an increase in the number of Cx-positive plaques per hepatocyte from gestational day 18 to day 7 after birth. The number of plaques containing both Cx32- and 26-positive gold particles was increased from gestational day 18 to day 7 after birth, but they could be hardly seen in the pericentral and middle area of lobules on day 7 after birth (Fig. 7). ii) Dye transfer assay in cultured liver tissues
Five minutes after the microinjection of LY into fetal livers, a few hepatocytes were labeled with fluorescence. The injected LY was spread in the liver tissue on day 1 after birth (Fig. 8a). The area of spread of LY was wider on day 7 after birth (Fig. 8b) than on day 1 after birth. LY-positive areas were measured with an image analyzer. The average LY-positive area in the liver was 0.1320.06 and 0.35?0.08* (meanrtSD, n=5) mm2 on day 1 and 7 after birth and there was a significant difference between them (The values were analyzed with Student’s t-test. *p
Discussion In the fetal rat livers small hepatocytes were observed to be intermingled with hematopoietic cells and there were few contacts between hepatocytes. Western blot analysis revealed the expression of Cx32 and Cx26 in the fetal liver from the late stage of gestation, but less 16
frequently than after birth. Cx32- and Cx26-positive plaques were seen on some hepatocytes before birth, and there were Cx3Zpositive plaques of different sizes. Immunogold electron microscopy revealed aggregates of Cx32-positive gold particles in large plaques, but the dye transfer of LY was limited in the fetal liver tissue. Gap junctional channels were aggregated to maintain intercellular communication (GJIC) on limited areas of cell-to-cell contact in the fetal liver. The large type of Cx3Zpositive plaque was reported to be an aggregate of the precursor of the gap junction (17,18), and to be seen at the time of cytodifferentiation (19); the large type detected in the fetal liver may be premature for GJIC, and was not detected in the established lobules. It was observed in the present study that there were few hematopoietic cells just after birth, the hepatocytes were in contact with each other, and expressions of Cx32 and 26 were increased. These results may have been obtained in part because plasma glucagon and glucocorticoid, in establishing portal circulation, may influence the expression of Cx32 and Cx26 mRNA (20). The numbers of Cx32- and Cx26-positive plaques were increased, while the number of aggregated or large-type Cx3Zpositive plaques was decreased. On day 7 after birth, Cx32-positive plaques were diffusely distributed on the plasma membranes of all hepatocytes in a lobule, and no large type was present, as in adult rats. Cx26-positive plaques were seen in the periportal area because glucagon in established portal vein has a greater influence on the expression of Cx26 mRNA than on that of Cx32 mRNA (21), and glucagon induces more Cx26 mRNA in the periportal area than in the pericentral area (22,23). The double immunofluorescence study showed Cx32- and Cx26-positive plaques independently localized or colocalized on the plasma membrane of each hepatocyte on day 18 of gestation or just after birth, and neither plaque was regularly distributed in the liver tissues. Gap junctions are made up of 6 units of connexin, a connexon, with a hydrophobic channel (8), and their channels are homotypic or heterotypic (14). To examine the homotypic or heterotypic type of gap junctional channel and the homomeric or heteromeric connexon at the ultrastructural level, we used double immunogold electron microscopy Cx-positive gold particles were symmetrically detected in plaques on opposite hepatocytes, but there seemed to be a relatively low ratio of antibody binding to Cx. This may be due to the large size of the antibody of IgG molecules compared with the dense packing of Cx proteins (23). In addition, Cx26-positive gold particles could not be detected by double immunogold electron microscopy as expected after the immunoreactive procedure with
Gap junctional channels in ontogenetic liver
Cx32 molecules. On gestational day 18 and just after birth, there were symmetrical pairs of Cx32- or Cx26positive gold particles in plaques, and some plaques contained many Cx32-positive gold particles mixed at random with a few Cx26-positive gold particles. Therefore many gap junctional channels within a plaque are the homotypic type of Cx32 or Cx26 (23) and some of them may be heterotypic or some connexons may be heteromeric (24). On day 7 after birth, Cx3Zpositive plaques were distributed in the pericentral or middle area, and Cx32positive plaques in the periportal area were colocalized with Cx26-positive plaques. Symmetrical pairs of Cx3Zpositive gold particles were observed in plaques of the central and middle areas. Symmetrical pairs of Cx32- and Cx26-positive gold particles were seen in plaques of the periportal area, and a few Cx32-positive gold particles were mixed with many Cx26-positive gold particles in one side of the submembranous area and vice versa. Thus, gap junctional channels in the central or middle area are homotypic and connexons are homomeric, and many gap junctional channels in the periportal area are heterotypic, and connexons are homomeric and a few of them may be heteromeric. GJIC can be estimated by dye transfer assay. LY for the assay may have spread into the extracellular space after its microinjection and its transfer assay may have some validity (25,26). We washed injected liver slices with PBS to eliminate non-specific spread. The dye transfer of LY was limited in the fetal liver because the areas of cell-to-cell contact were small, there were few gap junctional structures, and premature gap junctions were present; however, Cx32 and 26 were expressed. LY is reported to pass through the heterotypic gap junction of Cx32 and Cx26 (27) as well as the homotypic gap junction. The LY injected spread gradually in the liver tissues on day 1 after birth, and it spread more widely on day 7 after birth compared to day 1. Thus, GJIC developed in the liver during ontogenesis because the gap junctions increased in number, or open connexons might have increased. There were differences in the gap junctional structures of hepatocytes in fetal and neonatal livers, and their structures were heterogeneous within established lobules. Cx32 has phosphorylation sites. No reports have shown the presence of different bands in Western blot. Cx26 is devoid of phosphorylated sites (28,29). The closure mechanism in gap junctional channels composed of Cx32 may, therefore, differ from that in gap junctional channels composed of Cx26 or a mixture of Cx32 and Cx26. Thus there may be heterogeneity in the GJIC in each hepatocyte of a fetus or neonate and within a lobule. Our findings suggest that
homotypic or heterotypic incompatibilities of the gap junctions and homomeric or heteromeric interactions may contribute to the development of GJIC in the liver during ontogenesis.
Acknowledgements We thank Dr. Y. Ibata (Department of Anatomy, Kyoto Prefectural University of Medicine), Dr. M. Oyamada (Department of Pathology, Kyoto Prefectural University of Medicine) and Dr. T. Ueda (Research Laboratory, Nippon Shinyaku, Co.) for critical and helpful advice with our research work.
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. Spray DC. Physiological and pharmacological regulation of gap junction channels. In: Citi S, editor. Molecular Mechanisms of Epithelial Cell Junctions: From Development to Disease. Austin, Texas: R.G. Landes Company; 1994. p. 195-215. 3. Dermietzel RD, Hwang TK, Spray DS. The gap junction family: structure, function and chemistry. Anat Embryo1 1990; 182: 5178. 4. Gimlich RL, Kumar NM, Gilula NB. Sequence and development expression of mRNA coding for a gap junction protein in xenopus. J Cell Biol 1988; 107: 1065-73. 5. Guthrie SC, Gilula NB. Gap junctional communication and development. Trends Neurosci 1989; 12: 12-6. 6. Yamasaki H, Naus CC. Role of connexin genes in growth control. Carcinogenesis 1996; 17: 1199-213. 7. Nicholson BJ, Dermietzel R, Teplow D, Traub 0, Willecke K, Revel J-P Two homologous protein components of hepatic gap junctions. Nature 1987; 329: 732-3. 8. Revel JP Nicholson BJ, Yancy SB. Chemistry of gap junctions. Ann Rev Physiol 1985; 47: 263-79. 9. Traub 0, Look J, Dermietzel R, Brummer F, Hulser D, Willecke K. Comparative characterization of the 21-kD and 26-kD gap junction proteins in murine liver and cultured hepatocytes. J Cell Biol 1989; 108: 1039-51. 10. Neveu MJ, Hully JR, Babcock KL, Vaughan J, Hertzberg EL, Nicholson BJ, et al. Proliferation-associated differences in the spatial and temporal expression of gap junction genes in rat liver. Hepatology 1995; 22: 202-12. 11 Dermietzel R, Leibstein A, Frixen U, Janssen-Timmen U, Traub 0, Willecke K. Gap junctions in several tissues share antigenic determinants with liver gap junctions. EMBO J 1984; 3: 226170. 12. 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. 13. Paul DL. Monoclonal cloning of cDNA for rat liver gap junction. J Cell Biol 1986; 103: 123-34. 14. Kumar NM, Gilula NB. The gap junction communication channel. Cell 1996; 84: 381-8. 15. Maniatis T, Fritsch EF, Sambrook J. Molecular Cloning: A Laboratory Manual, 2nd ed. Cold Spring Harbor, NY Cold Spring Harbor Laboratory; 1999: 60(18)-75(18). 16. Wu LC. Isolation and long-term storage of proteins from tissues and cells using Trizol reagent. Focus 1995; 17: 98-100. 17. Decker RS. Hormonal regulation of gap junction. J Cell Biol 1976; 69: 669-85. 18. Johnson R, Hammer M, Sheridan J, Revel J-P Gap junction formation between reaggregated Novikoff hepatoma cells. Proc Nat1 Acad Sci USA 1974; 71: 453640. 19. Yamamoto M, Kataoka K. Large particles associated with gap
17
M. Iwai et al.
20.
21.
22.
23.
24.
25.
18
junctions of pancreatic exocrine cells during embryonic and neonatal development. Anat Embryo1 1985; 171: 305-10. Kojima T, Mitaka T, Shibata Y, Mizoguchi Y. Induction and regulation of connexin by glucagon in primary cultures of adult rat hepatocytes. J Cell Sci 1995; 108: 2771-80. Saez JC, Berthoud VM, Iwanji V Regulation of the abundance of connexin 32 and glucagon receptors in primary cultures of adult rat hepatocytes [abstract]. J Cell Biol 1991; 115: 189A. Berthoud VM, Iwanji V, Garcia AM, Saez JC. Connexins and glucagon receptors during development of rat hepatic acinus. Am J Physiol 1992; 26: 650-8. Kuraoka A, Iida H, Hatae T, Shibata Y, Itoh M, Kurita T. Localization of gap junction proteins, connexin 32 and 26, in rat and guinea pig liver as revealed by quick-freeze, deep-etch immunoelectron microscopy. J Histochem Cytochem 1993; 41: 971-80. Stauffer KA. The gap junction proteins /II-connexin (connexin32) and &connexin (connexin-26) can form heteromeric hemichannels. J Biol Chem 1995; 24: 6768-72. Krutovskih VA, Oyamada M, Yamasaki H. Sequential change of
26.
27.
28.
29.
gap-junctional intercellular communications during multistage rat liver carcinogenesis: direct measurement of communication in viva Carcinogenesis 1991; 12: 17014. Tateno C, Itoh S, Tanaka M, Oyamada M, Yoshitake A. Effect of DDT on hepatic gap junctional intercellular communication in rats. Carcinogenesis 1994; 15: 517-21. Elfang C, Eckert R, Lichtenberg-Frate H, Butterweck A, Traub 0, Klein RA, et al. Specific permeability and selective formation of gap junction channels in connexin-transfected HeLa cells. J Cell Biol 1995; 129: 805-17. Saez JC, Nairn AC, Czernik AJ, Spray DC, Hertzberg EL, Greengard P et al. Phosphorylation of connexin 32, a hepatocyte gap junction protein, by CAMP-dependent protein kinase, protein kinase C and Ca 2+/calmodulin-dependent protein kinase II. Eur J Biochem 1990; 192: 263-73. Zhang J-T, Nicholson BJ. Sequence and tissue distribution of a second protein of hepatic gap junctions, Cx26, deduced from its cDNA. J Cell Biol 1989; 109: 3391-6.