Immunohistochemical analysis of rat liver using a monoclonal antibody (HAM8) against gap junction.код для вставкиСкачать
THE ANATOMICAL RECORD 235335-341 (1993) lmmunohistochemicalAnalysis of Rat Liver Using a Monoclonal Antibody (HAM8) Against Gap Junction YOSHIHISA FUJIKURA, HIDEHIKO OHTA, TOSHIO HIRAI, AND TETSUO FUKUMOTO Department of Anatomy, Yamaguchi University School of Medicine, Ube, Yamaguchi 755, Japan (Y.F., T.H., T.F.);Department of Biochemistry, Kyorin University, School of Health Sciences, Hachioji, Tokyo 192, Japan (H.O.) ABSTRACT Four monoclonal antibodies were raised against crude gap junction fractions of rat liver to clarify the distribution of gap junctions during animal development and to analyze gap junction expression in vivo and the polarity of hepatocytes in vitro. Among the monoclonal antibodies obtained, HAM8 antibody recognized the 27-kDa rat liver gap junction protein connexin 32. This antibody recognized gap junctions at the contiguous faces of hepatocytes, and the antigen was also observed in exocrine pancreas and salivary gland but not in kidney, heart, esophagus, or thymus. HAM8 did not react with amphibian or fish liver, heart, esophagus, stomach, or intestine as assessed via the immunofluorescence method on frozen sections. A few hepatocytes and many hemopoietic cells were seen in rat fetal liver at 15 days of gestation. HAM8 antigen was expressed on some hepatocytes but not on any hemopoietic cells. As the fetus grew, the number of hepatocytes in the liver increased gradually, together with the amount of HAM8 antigen. The distribution of HAM8 antigen at 25 days after birth was similar to that in adult liver. When the expression of HAM8 antigen was examined in primary cultured hepatocytes using the immunofluorescence method, the antigen was observed clearly between the hepatocytes. However, most of the HAMS antigen on the free surface of hepatocytes disappeared within 4 hr. HAM8 antigen was not expressed on AH7974 rat hepatoma cells when they formed small islets in the rat peritoneal cavity or within the liver. When HAM8 IgG antibody was injected intravenously, the HAM8 signal was expressed in the liver. o 1993 Wiley-Liss, Inc. Key words: Connexin 32, Hepatocyte, Immunohistochemistry Gap junctions play a n important role in forming connections between adjacent cells and permitting the transfer of ions and other molecules from one cell to another. Each gap junction consists of connexons, which are constructed from six connexins (Cx),with a central pore. Four kinds of Cx have been reported in rats: Cx-26 (Nicholson et al., 19871, Cx-32 (Hertzberg and Gilula, 19791, Cx-43 (Manjunath et al., 19841, and Cx-46 (Kistler et al., 1986). These have organ specificity, and only Cx-26 and Cx-32 are found in the liver. There are many reports of polyclonal antibodies (Abs) raised against crude or synthesized gap junction protein (Aoumari et al., 1990; Beyer et al., 1989; Sugiyama and Ohta, 19901, but few monoclonal antibodies (mAbs) have been used to analyze the morphology and functions of gap junctions (Dermietzel et al., 1987; Takeda et al., 1988). In the present study, we isolated rat liver crude gap junction protein and raised a n mAb, HAM8, against it. mAb HAM8 recognized only Cx-32 in r a t liver. Many researchers have used polyclonal Abs against gap junction protein to analyze transformed hepatocytes and have reported the presence of little or no gap junction 0 1993 WILEY-LISS. INC. signal in the focus (Miyashita et al., 1991; Trosko et al., 1990). We have reported previously other mAbs (HAM1-5) against rat hepatocyte membrane antigens (Fujikura et al., 1985; Fukumoto e t al., 1984; Tamakoshi et al., 19851, their characterization (Fukumoto et al., 1986; Yamaguchi e t al., 1991), and also their application for analysis of hepatocyte physiology and development (Matsumoto et al., 1988; Sawada et al., 1992; Yamaguchi et al., 1991). We have also described mAbs (HAM6 and 7) showing specificity for the regions of chemically induced hepatocarcinoma and their usefulness for pathological analysis (Okita et al., 1988).In the present report, we describe the usefulness of HAM8 mAb for analyzing the distribution and development of gap junctions, especially the gap junction protein Cx- Received April 3, 1992; accepted June 23, 1992. Address reprint requests to Dr. Yoshihisa Fujikura, Department of Anatomy, Yamaguchi University School of Medicine, 1144, Kogushi, Ube, Yamaguchi 755, Japan. Y. FUJIKURA ET AL. 336 32, and also for determining the polarity of hepatocytes and analysis of transformed hepatocytes. MATERIALS AND METHODS Animals Sprague-Dawley female rats were purchased from Shizuoka Laboratory Animal Center (Hamamatsu, Japan). DA rats (newborns and young adults), LodM rats, and BALB/c mice (2-6 months old), were from stocks maintained at the Institute of Laboratory Animals, Yamaguchi University School of Medicine. These animals were inbred and have been maintained conventionally. To examine the expression of antigen in fetal liver, fetuses of DA rats were used. To determine the gestational age, the date when sperm were recognized in the vagina of DA female rats was used as day 0 of gestation, Other animals used included a n adult Rana catesbeiana from Seiwa Experimental Animal Center (Fukuoka, Japan) and specimens of Halichoeres poeilopterus and Lagocephalus lunaris spadiceus, 10 cm long, caught in the Seto Inland Sea. on a feeder layer of mouse thymocytes. After cell growth, the culture supernatants from each single colony well were tested, and clones in positive wells were considered to be mAb-producing hybridomas. Isolated Rat Hepatocyfes and a Rat Hepatoma Cell Line Isolated hepatocytes were prepared using collagenase following the method described previously (Berry and Friend, 1969; Seglen, 1976) with minor modification. Briefly, a DA rat was anesthetized and perfused with prewarmed Krebs-Henseleit bicarbonate buffer, pH 7.4, at 37°C through the portal vein for 10 min, followed by buffer containing 0.05%collagenase (Wako Chemical Co. Ltd., Osaka, Japan), pH 7.5, saturated with 95% 0, and 5%CO, a t a flow rate of 25 ml/min for 30-45 min. Recovered hepatocytes were then washed twice with Hank’s balanced salt solution (viability >95% by trypan blue exclusion) and cultured in Williams E medium containing 10% FCS, M insulin, and M dexamethasone on round coverslips for 1, 2, 3, or 4 h r in a GO, incubator at 37°C. Preparation of Antigen AH-7974 rat hepatoma cells (Fukumoto et al., 1984) Gap junction membrane was purified by the method maintained in our laboratory were cultured in 10% of Hertzberg (1984). As minor modifications, RPR-12, 1640 medium. These cells were washed three times in -16, and -20 rotors (Hitachi, Tokyo) and SW-27, 40Ti, PBS, then injected directly into the liver of Lou/M rats and type 50 rotors (Beckman California) were used for using a 27-gauge needle. The liver implants were then centrifugation alternatively, Livers from five to ten examined 3 days later. Ascitic-type AH-7974 cells, which formed small islands in the peritoneal cavity of Sprague-Dawley female rats were used per run. Lou/M rats, were also examined. - Immunization Thirty micrograms of antigen in 100 11.1of phosphatebuffered saline (10 mM Na,HP04, 150 mM NaCl, pH 7.4; PBS) were mixed with an equal volume of complete Freund’s adjuvant and injected into a female BALB/c mouse subcutaneously. Three weeks later, the mouse was given a booster injection in a similar manner. The third immunization was given intraperitoneally 3 weeks after the second booster. Final immunization was done intravenously without adjuvant. Cell Fusion and Hybridoma Isolation Antibodies Anti-Cx-32 polyclonal Ab was raised in rabbit using a synthesized dodecapeptide (P229-239; Ser-Arg-LysGly-Ser-Gly-Phe-Gly-His-Arg-Leu) with three separate immunizations (Sugiyama and Ohta, 1990). Fluorescein isothiocyanate (F1TC)-conjugated affinitypurified goat F(ab’), fragment antimouse IgG (0.7 mg/ ml) (Caltag Lab., San Francisco, CA) and horseradish peroxidase (HRPI-conjugated rabbit F(ab’), antimouse IgG (heavy- and light-chain specific) (1mg/ml) (Cappel Laboratories Inc., Cochranville, PA) were used as second-layer antibodies for immunohistochemical staining. NS-1 mouse myeloma cells were grown in RPMI1640 (Nissui Pharmaceutical Co., Ltd., Tokyo, Japan) supplemented with 10% heat-inactivated fetal calf serum (FCS) (Whittaker Bioproducts Inc., Walkersville, MD; lot No. 9M0604) (10% 1640) at 37°C in humidified Gel Electrophoresis and lmmunoblofting 5% co2/95% air. The myeloma cells (2 x lo7) were fused with spleen cells (2 x 10’) from an immunized Samples were boiled in 100 11.1sodium dodecyl sulfate mouse 3 days after the last antigen boost, using poly- (SDSI-sample buffer (2% SDS, 10% glycerol, 100 mM ethylene glycol 4000 (Merck, Darmstadt, Federal Re- Tris HC1, pH 6.8, and 0.02% bromophenol blue) for 5 public of Germany), as described elsewhere (Fujikura min. The eluted samples were electrophoresed on an et al., 1985; Fukumoto et al., 1984; Galfre et al., 1979) 8-14% gradient polyacrylamide slab gel according to in the above-described medium. One milliliter of 10% Laemmli (1970). Following the electrophoresis, the 1640 containing lop4 M hypoxanthinei4 x M proteins were transferred from the gel onto a nitrocelaminopterin/l.6 x M thymidine medium was lulose membrane. After treatment of the membrane added to each well after the fused cells had been sus- with Tris HC1 buffer (10 mM Tris HC1, pH 7.5,150 mM pended at 2 x lo6 cells/ml/well in 24-well tissue cul- NaC1, 0.05% Tween 20) containing 5% skim milk, the ture plates and confirmed to be growing well. Ten to blot was incubated with HAM8 mAb (10 pg/ml) for 1hr fourteen days after fusion, when hybrid clones had ap- a t room temperature. After washing with PBS, the peared, the culture media were screened for antibodies membrane was reacted with HRP-conjugated antiby a n immunofluorescence method. Cultured cells that mouse IgG for 1h r at 4°C. Then, it was developed with secreted antibodies were diluted with 10% 1640 me- 3,3’-diaminobenzidine and 0.02% hydrogen peroxide dium containing h poxanthine (1 x lop4M) and thy- solution following the method of Graham and Karmidine (1.6 x 10- B M) in 96-well tissue culture plates novsky (19661, with minor modification. MONOCLONAL ANTIBODY AGAINST RAT GAP JUNCTION 337 Indirect lmmunohistochemical Staining of Tissues Rats were anesthetised with diethyl ether and sacrificed by exsanguination. Frogs were killed by destruction of the spinal cord and the fish by exsanguination. Tissues (liver, kidney, pancreas, salivary glands, heart, spleen, thymus, esophagus, and lens of eyeball) and AH-7974 rat hepatoma cells implanted into liver were removed, embedded in OCT compound, frozen, and sectioned at a thickness of 6 p m on a Bright OT/FAS cryostat (U.K.) a t -25°C. The 6-pm sections of tissues, freshly isolated hepatocytes, or AH-7974 cells on microscope slides were air dried for 30 min at room temperature and immersed in ice-cold acetonelmethanol (1:l)or methanol only for 10 min for fixation. Cultured cells on round coverslips were fixed by direct immersion. After washing three times in ice-cold PBS, these specimens were incubated with 100 pl of first-layer antibodies or control at room temperature for 1hr in a humid chamber and at 4°C for 10 min. The majority of these tissue sections were then washed again in a similar manner and incubated with 30-40 pl of 1:20-diluted FITC-conjugated goat F(ab’), antimouse IgG a s a second-layer antibody for 1-2 h r a t 4°C in a humid chamber. After washing three times in ice-cold PBS, the sections were mounted in glycerol and observed using a fluorescence microscope (Nikon, XFEF). HAM8 Antigen Expression In Vivo Ascitic-type HAM8 mAb was purified by the precipitation method using saturated sodium sulfate (Good et al., 1980). The Ab was dissolved in PBS, and the final concentration was adjusted to 20 mg/ml. A rat was given a n injection of the HAM8 Ab (100 mg/kg body weight) intravenously and was then killed by exsanguination 10 min later. The liver was removed immediately, and cut into frozen thin sections. These sections were stained with FITC-conjugated antimouse IgG and observed using a fluorescence microscope. RESULTS Production and Characterization of Hybridomas Four hybridomas were established and were designated HAM8, JC4-264, JC4-781, and JC4-902. Their immunoglobulin subclasses were IgG1, IgM, IgG2b, and IgM, respectively as examined by the double diffusion precipitation technique using antimouse IgG1, IgGaa, IgG2b, IgG3, and IgM antibodies (Cappel Laboratories Inc.). lmmunoblot Analysis (KDa) 126.96.36.19931 21.514.4- Fig. 1. Specificity of antibodies was determined by Western blot analysis of rat liver cell membrane extract. The proteins were electrophoretically separated by SDS-polyacrylamide gels and transferred to nitrocellulose paper, and then transfers of individual lanes were reacted with different antibodies as indicated. Lane 1: AntiCx-32 polyclonal Ab. Lane 2: HAM& Lane 3 JC4-264. Lane 4 Control. The following standard proteins were used as mobility markers: phosphorylase b (92.5 kDa), bovine serum albumin (66.2 kDa), ovalbumin (45 kDa), carbonic anhydrase (31 kDa), trypsin inhibitor (21.5 kDa), and a-lactalbumin (14.4 kDa). lane 4). When the crude proteins from mouse liver cell membrane were subjected to electrophoresis, HAM8 recognized 27- and 45-kDa proteins. This pattern was similar to that of anti-Cx-32 Ab. No similar band was stained by the JC4-264, -781, and -902 mAbs (data not shown). Expression of HAM8 Antigen in Rat Liver The staining pattern of HAM8 mAb on rat liver frozen sections was specific; it recognized only the intercellular borders of hepatocytes, producing short linear and dotted patterns (Fig. 2a). The number of HAM8 antigen spots between hepatocytes on 6-p.m-thick cryosections was usually two to four (Fig. 2a). No specific staining was observed on control sections (Fig. 2b). HAM8 IgG Ab was injected into rats intravenously, and then cryosections from the rat liver were stained with second Ab. HAM8 signals were clearly observed at the intercellular border of hepatocytes, giving the same pattern as that by indirect immunof luorescence (data not shown). After SDS-PAGE and immunoblotting of rat liver membrane extract, the membrane was stained immunohistochemically using affinity purified anti-Cx-32 polyclonal Ab (Fig. 1, lane 1)and HAM8 and JC4-264 mAb (Fig. 1,lanes 2 and 3). HAM8 recognized a 27-kDa protein and its dimer, showing a staining pattern identical to that obtained with anti-Cx-32 polyclonal Ab (Fig. 1, lane 1).The staining patterns of JC4-264 and -902 were similar to each other, and these Abs recogHAM8 Antigen in Various Organs of Raf and nized a 43-kDa protein and its polymers (Fig. 1,lane 3). Other Animals The positions of their bands on the membrane were In the cryosections from DA rats, HAM8 antigen was different from these of anti-Cx-32 Ab. Two faint bands were seen in all lanes, including the control (Fig. 1, recognized in salivary glands and in the exocrine pan- 338 Y. FUJIKURA ET AL. Fig. 2. Rat liver cryosections 6 pm thick were fixed with ice-cold methanol and stained via the indirect immunofluorescence method. a: HAM8 was used as the first-layer Ab. HAM8 antigen was observed a t the intercellular borders of hepatocytes as short lines and dots. x 280. b: No staining was seen on the control section. x 280. Fig. 3. Salivary gland and pancreas from DA rat were cryosectioned, and the 6-pm-thick specimens were stained by the indirect immunofluorescence method using HAM8 Ab. a: In salivary gland, HAM8 antigen was observed between acinar cells and adjacent cells but not on the duct portion (D). x310. b In pancreas, HAM8 antigen was recognized in exocrine portions but not in the islet of Langerhans (L). x 310. creas, but not in heart, thymus, spleen, esophagus, lens, or kidney. In salivary glands, HAM8 antigen was usually recognized as small spots between acinar cells and adjacent cells. The number of spots a t the intercellular border was generally one or two and rarely zero or three on 6-pm-thick sections. There was no HAM8 antigen in various portions of ducts, and none or little between neighboring acini (Fig. 3a). In the pancreas, HAM8 antigen was observed only in exocrine portions and not in endocrine portions (islets of Langerhans). The number of HAM8-positive spots between the acinar cells was usually one, sometimes two, and rarely zero or three in thin sections (Fig. 3b). HAM8 antigen was expressed in liver sections from BALB/c mice but not in those from Rana catesbeiana, Halichoeres poeilopterus, OT Lagocephalus lunaris spadiceus. The staining pattern of HAM8 antigen in mouse liver was the same as that in rat liver (data not shown). HAM8 Antigen Expression in Perinatal Rat Liver The liver a t day 15 of gestation consists mainly of hemopoietic cells (90-95%) and of a few hepatocytes. HAM8 antigen was observed only on some hepatocytes but not on hemopoietic cells (data not shown). The number of HAM8 antigen spots between hepatocytes on 6-pm-thick sections appeared to be similar to that in adult liver. However, the ratio of short linear and dotshaped HAM8 antigen deposits was slightly higher than that in adult liver. The percentage of hepatocytes in the liver at day 17 of gestation was increased to 30%, and the frequency of HAM8 antigen observed in sections was also higher (data not shown). After birth, the majority of liver cells were hepatocytes and Kupffer cells, and hemopoietic cells had mostly disappeared. Many HAM8 signals were observed on cryosections of the liver at 2, 5, and 8 days after birth. HAM8 antigen was more plentiful a t the centers of lobules (around the central veins) than at their periphery (data not shown). The hepatic cords were not well developed a t this stage. Hepatic structure and HAM8 antigen expression in the liver of 12and 25-day-old rats were similar to those in adult liver. HAM8 Antigen on Rat Primary Cultured Hepatocytes A considerable amount of HAM8 antigen was present on the surface of cultured hepatocytes as dots or short linear profiles just after perfusion with collagenase (Fig. 4b, arrows). On the free surface of hepatocytes, the antigens were arranged as parallel dots and/or broken lines (Fig. 4b, large arrows). When the hepatocytes were not completely separated by the colIagenase treatment, the above-mentioned parallel MONOCLONAL ANTIBODY AGAINST RAT GAP JUNCTION 339 Fig. 4. DA rat liver was perfused with 0.05% collagenase solution through the portal vein; then, free hepatocytes were smeared or cultured on round coverslips in a 24-well culture plate. The specimens were washed with PBS and then fixed with methanol after culture for 0 (a, b), 2 (c), and 4 (d) hr. After washing, the cells were reacted with HAM8 antibody (b-d) or control (a),followed by FITC-conjugated antimouse IgG antibody. a: No staining was seen in the control. X 450. b: Many HAM8 antigen spots were seen on the surface of hepatocytes (large arrows) and between hepatocytes as dots or short linear profiles (small arrow). x 450. c: HAM8 antigen was recognized around cells, especially between hepatocytes, as dots or short lines on the cell surface connected to a neighboring hepatocyte. x 375. d HAM8 antigen was usually recognized between hepatocytes (small arrows), but it had disappeared almost completely from the free surface of hepatocytes at 4 hr of culture. x 375. lines on the hepatocyte surface were connected to the surface of a neighboring hepatocyte (Fig. 4b, large arrow a t bottom). No antigen signal was Seen on control cells (Fig. 4a). HAM8 antigen on the free surface Of hepatocfles gradually became reduced and disappeared by 2 h r of culture (Fig. 4 ~ 1whereas , the antigen Present between hepatocytes was retained with hr after seeding (Fig. 4d, arrows). HAM8 Antigen Expression In Vivo HAM8 Ab IgG was injected into a rat intravenously, and then cryosections of the rat liver were stained with second Ab. HAM8 signals were clearly observed a t the intercellular border of hepatocytes, with a pattern corresponding to that produced by the indirect immunofluorescence method on normal rat liver (data not shown). HAM8 Antigen Expression on Rat Hepatoma Cells DISCUSSION AH-7974 hepatoma cells were transplanted into the peritoneal cavity of a LodM rat. One week after transfer, the cells formed many small islets in the peritoneal cavity. HAM8 antigen was not observed on frozen sections of the small masses upon examination via the immunofluorescence method (data not shown). On the other hand, free hepatoma cells that were injected into LouIM rat liver formed many miliary-sized hepatoma foci 3 days later. HAM8 antigen was observed on normal liver tissue, as on normal DA r a t liver sections, but not in the area of the transplanted hepatoma foci or on the border between normal tissue and hepatoma tissue (data not shown). Gap junctions, showing various forms, are recognized in many organs and play important roles, including maintenance of cell-to-cell electrical continuity and intercellular adhesion. To analyze the functions and structures of gap junctions, some polyclonal (Aoumari et al., 1990; Beyer et al., 1989; Sugiyama and Ohta, 1990) and monoclonal Abs (Dermietzel et al., 1987; Takeda et al., 1988) have been raised against rat cardiac or brain gap junction protein (Cx-43) (Aoumari et al., 1990; Beyer et al., 1989) and rat liver gap junction protein (Cx-32) (Beyer et al., 1989; Dermietzel et al., 1987; Sugiyama and Ohta, 1990; Takeda et al., 1988) using synthesized polypeptides (Aoumari et al., 1990; 340 Y. FUJIKURA ET AL. Beyer et al., 1989; Sygiyama and Ohta, 1990) or crude the level of mRNA specific for murine Cx-43 in mouse protein (Dermietzel et al., 1987; Takeda et al., 1988) as heart at different stages of development. The level of mRNA increased from day 11 of gestation to 1 week antigens. We prepared four mAbs against crude gap junction after birth and then gradually decreased until the protein. One of them, HAM8 was used to clarify the adult stage. In the present study, we examined HAM8 distribution of Cx-32. mAbs (6-3Gll and 7-3H6) re- antigen expression in perinatal rat liver and found that ported previously by Takeda et al. (1988) reacted with the number of antigen-positive cells gradually ina 27-kDa gap junction protein in rat liver and exocrine creased with age. This result was similar to that for acinar cells but not in brain, heart, lung, kidney, ad- HAM4 antigen, which was first expressed on fetal herenal gland, or uterus. HAM8 may therefore recognize patocytes a t 18 days of gestation and increased until 4 an epitope similar to that reactive with 6-3Gll or 7- weeks after birth (Matsumoto et al., 1988).HAM5 mAb 3H6. However, the antigenic epitopes defined with was raised against rat fetal liver cells a t day 15 of their antibodies were not yet reported. In our prelimi- gestation (Fujikura et al., 1985). This Ab reacted with nary study, using dot-blot analysis, HAM8 Ab reacted both fetal and adult hepatocytes, but not with hemopoietic cells in fetal liver. The changes in level of mRNA with an 11-peptide (Ser-Arg-Lys-Gly-Ser-Gly-Phe-GlyHis-Arg-Leu)sequence of Cx-32, P229-239 (Ohta et al., might be different from those in antigen expression. in preparation). This sequence is located in the cyto- HAMS, HAM4, and HAM5 mAbs might be useful for plasmic region of Cx-32 near the C terminal. However, characterizing the differentiation of hepatocytes durthe results obtained after injection of HAM8 Ab in vivo ing the perinatal period. The half-life of hepatocyte gap junction protein is suggested that the Ab might bind to the epitope facing the outer surface of the cell. Thus there is a discrepancy reported to be 5 hr in situ (Fallon and Goodenough, between the epitope localization (in the cytoplasm) and 1981). Many researchers have examined the period of the result of the HAM8 antigen-antibody reaction in expression of gap junction protein using cultured hevivo. This discrepancy may now be considered in terms patocytes. Traub et al. (1989) measured the volume of of gap junction synthesis and hepatocyte endocytosis 27-kDa gap junction protein using immunoblot analydescribed by Evans and Graham (1989). sis, autoradiography, and intercellular dye transfer HAM8 antigen was not observed in cryosections of methods, and Spray et al. (1987) used electrical couvarious tissues from lower vertebrates such as amphib- pling. However, very few investigations have employed ians and fish. A few previous studies have obtained immunohistochemistry. One such study, by Saez et al. similar results, but there are no reports indicating the (19891, found that the gap junction signal in cultured presence of Cx-32 in lower animals (Peracchia, 1991; hepatocytes was decreased 5-8 hr after plating, apRyerse, 1989). The expression of Cx-32 antigen is re- pearing as a diffuse granular deposit. We noticed the ported to be decreased in regenerated liver after partial disappearance of gap junctions in cultured hepatocytes hepatectomy (Sugiyama and Ohta, 1990) and in chem- at an early stage, 0-4 hr, and observed two kinds of ically induced neoplastic nodules or hepatocellular car- signal at the cell surface. When hepatocytes were culcinoma (Beer et al., 1988). Furthermore, Fitzgerald et tured in vitro, Cx-32, which had lost its binding partal. (1989) have reported that the level of gap junction ner on the surface of hepatocytes, had almost disapmRNA was markedly reduced during rat liver carcino- peared within 2 hr, whereas HAM8 antigen was genesis. Therefore, we examined the expression of Cx- relatively well conserved at the intercellular border 32 antigen in a rat hepatoma cell line, AH-7974. until 4 hr of culture. In our previous study, the periphHAM8 antigen was not seen in or on AH-7974 cells. ery of primary-cultured hepatocytes was stained with These results are similar to the results of Beer et al. HAM4 Ab uniformly at the beginning. Within 1-3 hr (1988), although Oyamada et al. (1990) reported that of seeding, HAM4 staining became localized in the arthe number of gap junction spots in human hepatocel- eas of attachment to neighboring cells, but diminished lular carcinomas stained with anti-Cx-32 Ab was not on other surfaces. When inhibitors of the cytoskeleton, less than that in the surrounding noncancerous tissue. such as colchicine, colcemid, and nocodazole, were Thus there might be a difference in the mechanism of added t o the hepatocyte culture medium, the localizagap junction disappearance in carcinogenetic liver be- tion of HAM4 antigen molecules at the bile-canalicular tween human and rat. The level of mRNA for the c-raf surface was disrupted (Sawada et al., 1992). Saez et al. protooncogene in chemically induced hepatocarcinoge- (19891, in their experiment, used nocodazole, a micronetic lesions was higher than that in nontumor sites tubule disruptor, and found that hepatocyte coupling in (Beer et al., 1988). Previously, we reported several vitro was prolonged. Although these inhibitors were other mAbs (HAM1-7) against rat hepatocyte surface not used in the present study, it seems from these reantigens or transformed liver cells. HAM4 mAb recog- sults that HAM8 antigen moved into the cytoplasm nized the bile-canalicular face of the cells and bound to after losing its partner, then was digested by an enAH-44 rat hepatoma cells very weakly but not to zyme or moved to sites of adhesion through the cytoAH66F cells (Tamakoshi et al., 1985). The chemically plasm or cell surface from the sinusoidal to the lateral induced hepatocarcinogenic lesions had HAM6 anti- face with the aid of cytoskeletal components. To congen, whereas HAM7 reacted with the hepatocytes sur- firm this, we are currently analyzing changes in the rounding hyperplastic nodules (Okita et al., 1988). As expression of HAM8 antigen on cultured hepatocytes was mentioned previously, these HAM-series mAbs using cytoskelton inhibitors. might be very useful for analyzing rat hepatoma or The present study has thus demonstrated the utility hepatocarcinogenesis. of HAM8 mAb for localization of Cx-32. This Ab might There have been very few developmental studies of be useful for analyzing the polarity of rat hepatocytes, murine gap junctions. Formaget et al. (1990)measured along with HAM4 and -5 mAbs, for clarifying the pro- MONOCLONAL ANTIBODY AGAINST RAT GAP JUNCTION cess of formation and transfer of Cx-32 in or on the hepatocytes and for demonstrating the ontogenic development of Cx-32 in liver during the perinatal period. LITERATURE CITED Aoumari, A.E., C. Fromaget, E. Dupont, H. Reggio, P. Durbec, J.P. Briand, K. Boller, B. Kreitman, and D. Gros 1990 Conservation of a cytoplasmic carboxy-terminal domain of connexin 43, a gap junction protein, in mammal heart and brain. J . Membrane Biol., 115:229-240. Beer, D.G., M.J. Neveu, D.L. Paul, U.R. Rapp, and H.C. Pitot 1988 Expression of the c-ruf protooncogene, y-glutamyltranspeptidase and gap junction protein in rat liver neoplasms. Cancer Res., 48t1610-1617. Berry, M.N., and D.S. Friend 1969 High-yield preparation of isolated rat liver parenchymal cells. A biochemical and fine structural study. J . Cell Biol., 43506-520. Beyer, E.C., J . Kistler, D.L. Paul, and D.A. Goodenough 1989 Antisera directed against Connexin 43 peptides react with a 43-kD protein localized to gap junctions in myocardium and other tissues. J . Cell Biol., 108t595-605. Dermietzel, R., B. Yancey, U.J. Timmen, 0. Traub, K. Willecke, and J.P. Revel 1987 Simultaneous light and electron microscopic observation of immunolabeled liver 27 kD gap junction protein on ultra-thin cryosections. J . Histochem. Cytochem., 35:387-392. Evans, W.H., and J.M. Graham 1989 Membrane biogenesis and trafficking. In: Membrane Structure and Function. D. Rickwood ed. IRL Press, Oxford, pp. 52-69. Fallon, R., and D.A. Goodenough 1981 Five-hour half-life of mouse liver gap junction protein. J . Cell Biol., 90521-526. Fitzgerald, D.J., M. Mesnil, M. Oyamada, H. Tsuda, N. Ito, and H. Yamasaki 1989 Changes in gap-junction protein (connexin 32) gene expression during rat liver carcinogenesis. J . Cell Biochem., 41:97-102. Formaget, C., A.E. Aoumari, E. Dupont, J.P. Briand, and D. Gros 1990 Changes in the expression of connexin 43, a cardiac gap junctional protein, during mouse heart development. J . Mol. Cell Cardiol., 22r1245-1258. Fujikura, Y., H. Kuniki, and T. Fukumoto 1985 Monoclonal antibodies against fetal rat liver cells. Bull. Yamaguchi Med. School, 32:21-26. Fukumoto, T., H. Kimura, M. Naito, M. Miyamoto, A. Yamashita, and H. Sugiyama 1984 Monoclonal antibodies to rat liver cell membrane glycoproteins. Mol. Immunol., 21:285-291. Fukumoto, T., K. Tamakoshi, H. Ohta, A. Yamashita, and K. Maeda 1986 Purification and characterization of MHC class I antigen from rat liver with monoclonal antibody. Int. J. Biochem., 18: 971-977. Galfre, G., C. Milstein, and B. Wright 1979 Rat x rat hybrid myelomas and a monoclonal a n t i d d protein of mouse IgG. Nature, 277:131-133. Good, A.H., L. Wofsy, J . Kimura, and C. Henry 1980 Purification of immunoglobulins and their fragments. In: Selected Methods in Cellular Immunology. B.B. Mishell and S.M. Shiigi, eds. W.H. Freeman and Company, San Francisco, pp. 278-280. Graham, R.C., and M.J. Karnovsky 1966 The early stages of absorption of injected horseradish peroxidase in the proximal tubules of absorption of mouse kidney: Ultrastructural cytochemistry by a new technique. J . Histochem., 14r291-302. Hertzberg, E.L. 1984 A detergent-independent procedure for the isolation of gap junctions from rat liver. J . Biol. Chem. 259:99369943. Hertzberg, E.L., and N.B. Gilula 1979 Isolation and characterization of gap junctions from rat liver. J. Biol. Chem. 254r2138-2147. Kistler, J., B. Kirkland, and S. Bullivant 1986 Identification of a 341 70,000-D protein in lens membrane junctional domains. J. Cell Biol.., 707t2L.15. Laemmli, U.K. 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227t680-685. Manjunath, C.K., G.E. Goings, and E. Page 1984 Cytoplasmic surface and intramembrane components of rat heart gap junctional proteins. Am. J . Physiol., 246:H865-H875. Matsumoto, M., K. Tamakoshi, K. Kanai, M. Kako, and T. Fukumoto 1988 Expression of a hepatocyte membrane antigen during hepatocarcinogenesis and in the developing - - liver of the rat. Int. J. Cancer, 41:583-588. Miyashita, T., A. Takeda, M. Iwai. and T. Shimazu 1991 Sinale ad- ministration of hepatotoxic chemicals transiently decreases the gap junction protein levels of connexin 32 in rat liver. Eur. J. Biochem., 196:37-42. Nicholson, B.J., R. Dermietzel, D. Teplow, 0. Traub, K. Willecke, and J.P. Ravel 1987 Two homologous protein components of hepatic gap junctions. Nature, 329:732-734. Okita, K., T. Esaki, F. Kurokawa, T. Takemoto, Y. Fujikura, and T. Fukumoto 1988 An antigen specific to hyperplastic liver nodules defined with monoclonal antibody: A new marker for preneoplastic cells in rat chemical hepatocarcinogenesis. Tumor Biol., 9: 170-177. Oyamada, M., V.A. Krutovskikh, M. Mesnil, C. Partensky, F. Berger, and H. Yamasaki. 1990 Aberrant expression of gap junction gene in primary human hepatocellular carcinomas: increased expression of cardiac-type gap junction gene connexin 43. Mol. Carcinogen., 3:273-278. Peracchia, C. 1991 Effects of the anesthetics heptanol, halothane and isoflurane on gap junction conductance in crayfish septate axons: A calcium- and hydrogen-independent phenomenon potentiated by caffeine and theophylline and inhibited by 4-aminopyridine. J. Membrane Biol., 121:67-78. Ryerse, J.S. 1989 Electron microscope immunolocation of gap junctions in Drosophilu. Tissue Cell, 21:835-839. SBez, J.C., W.A. Gregory, T. Watanabe, R. Dermietzel, E.L. Hertzberg, L. Reid, M.V.L. Bennett, and D.C. Spray 1989 CAMP delays disappearance of gap junctions between pairs of rat hepatocytes in primary culture. Am. J. Physiol., 257:Cl-Cll. Sawada, T., H. Itai, Y. Fujikura, H. Kuniki, M. Tamechika, and T. Fukumoto 1992 Distribution of the surface antigen HAM-4 and cytoskeleton during reformation of bile-canalicular structures in rat primary cultured hepatocytes. Exp. Cell Res., 199t50-55. Seglen, P.O. 1976 Preparation of isolated rat liver cells. Methods. Cell Biol., 13r29-83. Spray, D.C., M. Fujita, J.C. Saez, H. Choi, T. Watanabe, E. Hertzberg, L.C. Rosenberg, and L.M. Reid 1987 Proteoglycans and glycosaminoglycans induce gap junction synthesis and function in primary liver cultures. J . Cell Biol., 105.541-551. Sugiyama, Y., and H. Ohta 1990 Changes in density and distribution of gap junction after partial hepatectomy: Immunohistochemical and morphometric studies. Arch. Histol. Cytol., 53:71-80. Takeda, A,, M. Kanoh, T. Shimazu, and N. Takeuchi 1988 Monoclonal antibodies recognizing different epitopes of 27-kDa gap junction protein from rat liver. J . Biochem., 104t901-907. Tamakoshi, K., T. Fukumoto, K. Kanai, and A. Yamashita 1985 A monoclonal antibody to a rat hepato-renal membrane antigen. Clin. Exp. Immunol., 60r373-380. Traub, O., J . Look, R. Dermietzel, F. Briimmer, D. Hiilser, and K Willecke 1989 ComDarative characterization of the 21-kD and 26-kD gap junction 'proteins in murine liver and cultured hepatocytes. J. Cell Biol., 108:1039-1051. Trosko, J.E., C. Chang, B.V. Madhukar, and J.E. Klaunig 1990 Chemical, oncogene a<d growth factor inhibition of gap iunctional intercellular communication: An integrative hypothesis of carcinogenesis. Pathobiology, 58:265-278. Yamaguchi, K., Y. Fujikura, H. Kuniki, K. Itoh, K. Tamakoshi, and T. Fukumoto 1991 Immunoelectron microscopic localization of cell surface antigens on rat hepatocytes detected with monoclonal antibodies (HAM-2 and HAM-4). Cell Struct. Funct., 16t303-313.