Inhibition of mouse hepatocyte gap junctional intercellular communication by phenobarbital correlates with strain-specific hepatocarcinogenesis

The inhibition of gap junctional intercellular communication (GJIC) is a common effect of nongenotoxic carcinogens and might be a biomarker for these agents. To further test this relationship, we hypothesized that phenobarbital would inhibit mouse hepatocyte GJIC and this would correlate with strain-specific hepatocarcinogenicity. Phenobarbital is a strong nongenotoxic hepatocarcinogen in B6C3F1 mice, but not in C57BL/6 mice. Hepatocytes were isolated from males of both strains, placed in coculture with rat liver epithelial cells, and treated with phenobarbital for up to 14 days. Male mice were also administered PB by single intraperitoneal injection (0.1 mg/kg), then sacrificed 24 h later, or given phenobarbital in the drinking water (500 ppm) for 14 days before sacrifice. GJIC was assayed in cocultures by fluorescent dye microinjection and in isolated liver tissue by fluorescent dye "cut-loading." Phenobarbital decreased GJIC only in cultured B6C3F1 hepatocytes; this was dose-responsive and temporary, because hepatocyte GJIC returned to control levels within 24 h of phenobarbital exposure. Administration of phenobarbital to mice for 14 days also decreased hepatocyte dye coupling in B6C3F1 liver, but this effect was not seen in C57BL/6 mice or observed after a single administration of the drug. Phenobarbital did not alter connexin32 and connexin26 expression, but increased hepatic Cyp2b1 expression and the liver weight:body weight ratio in both strains. In summary, phenobarbital inhibited mouse hepatocyte GJIC in vivo and in vitro and in correlation with strain-specific hepatocarcinogenicity. These data support the hypothesis that decreased GJIC is a biomarker for nongenotoxic carcinogens and involved in their carcinogenic mechanism.

Positive regulation of connexin32 transcription by hepatocyte nuclear factor-1alpha

Connexin32 (Cx32) encodes the predominant gap junction protein expressed by hepatocytes. We investigated the transcriptional control of Cx32 in expressing and nonexpressing rat liver cell lines and hypothesized that a putative hepatocyte nuclear factor-1 (HNF-1) binding site (centered at mp -187) in the liver-active, P1 promoter is essential for transcription of Cx32. HNF-1alpha was expressed by Cx32-expressing rat liver cell lines and bound the promoter at the -187 site, but was not expressed by non-Cx32-expressing hepatic lines. Stable transfection of non-Cx32-expressing WB-F344 rat liver epithelial cells with HNF-1alpha stimulated a transfected Cx32 promoter element (mp -244 to -33), binding of HNF-1alpha to the -187 site, and expression of endogenous Cx32. Site-directed mutagenesis of this HNF-1 binding site abolished HNF-1alpha binding and proximal promoter activity. Hepatic Cx32 expression was also significantly decreased in HNF-1alpha(-/-) mice. These data indicate that HNF-1alpha is a positive regulator of Cx32 expression in hepatic cells.

Neoplastic reversal of human ovarian carcinoma cells transfected with connexin43

Gap junctional intercellular communication and expression of gap junction proteins (connexins) are decreased frequently in neoplastic cells including human ovarian carcinoma cells. In order to test the hypothesis that these changes contribute to the neoplastic phenotype of ovarian carcinoma cells, we transfected human ovarian carcinoma SKOV-3 cells with connexin43. Stable, connexin43-expressing transfectants were characterized for cell proliferation in vitro in normal, low-serum, and serum-free culture medium, for tumorigenicity in nude mice, and for sensitivity to adriamycin in vitro. Transfected clones expressed higher levels of connexin43 and gap junctional intercellular communication, reduced proliferation and greater dependence upon serum for growth in vitro, decreased tumor formation, increased sensitivity to adriamycin, and reduced expression of p-glycoprotein. These data suggest that gap junctional intercellular communication and/or connexin43 expression suppresses the neoplastic phenotype of ovarian carcinoma cells and their downregulation is involved in neoplastic transformation of ovarian epithelial cells. The increased sensitivity to adriamycin and elevated expression of p-glycoprotein by the transfected cells also suggest that gap junctional intercellular communication and connexin43 expression are involved in drug sensitivity and might be manipulated to enhance the clinical response.

Liver cell-specific transcriptional regulation of connexin32

Gap junctional intercellular communication facilitates liver homeostasis and growth control in the liver. The major gap junction protein expressed by hepatocytes is connexin32 (Cx32) and non-parenchymal hepatic cells do not express this gene. We investigated the regulation of Cx32 transcription by trans-activating factors in liver cells. Transient transfection assays using deletions of the rat Cx32 promoter (nt -753 to -33) linked to the luciferase gene were performed in MH1C1 rat hepatoma cells that express endogenous Cx32 compared with WB-F344 rat liver epithelial cells that do not. The basal promoter element was located within nt -134 to -33 and was 1.4-fold more active in MH1C1 cells than WB-F344 cells whereas the entire promoter fragment (nt -754 to -33) was four-fold more active in MH1C1 cells. Specific nuclear protein-DNA complexes that bound to Sp1 consensus sites within the basal promoter were formed using nuclear extracts from both types of cells. Additional promoter sequences increased promoter activity more strongly in MH1C1 cells than WB-F344 cells and this was correlated with the binding of hepatocyte nuclear factor-1 (HNF-1) to two HNF-1 consensus sites centered at -187 and -736. Expression of HNF-1 and binding to these elements was only observed with MH1C1 cells. Other specific protein-DNA complexes were formed, however, that included YY-1- and NF-1-containing complexes, but these were not related to promoter activity. Dexamethasone increased Cx32 promoter activity and expression in MH1C1 cells, but had little effect in WB-F344 cells and did not alter protein-DNA complex formation. These data suggest that Sp1 is responsible for Cx32 promoter basal activity, that HNF-1 determines the cell-specific expression of Cx32, and that dexamethasone increases Cx32 expression through other mechanisms.

Gap junctional intercellular communication and connexin43 expression in human ovarian surface epithelial cells and ovarian carcinomas in vivo and in vitro

Gap junctional intercellular communication (GJIC) and the expression of gap junction proteins (connexins) are frequently decreased in neoplastic cells and have been increased by cAMP and retinoids. GJIC and connexin expression were investigated in early passage normal human ovarian surface epithelial (HOSE) cells, human ovarian adenocarcinoma cell lines (CaOV-3, NIH:OVCAR-3, SK-OV-3 and SW626) and surgical specimens of human serous cystadenocarcinomas. We hypothesized that GJIC and connexin expression would be decreased in neoplastic cells and would be increased by cAMP and retinoic acid. Cultured HOSE cells exhibited extensive fluorescent dye-coupling and connexin43 (Cx43) expression; other connexins were not detected. The ovarian adenocarcinoma cell lines had little dye-coupling or connexin expression. Deletions and rearrangements of the Cx43 gene were not detected by Southern blotting in the carcinoma lines. N6, 2'-O-dibutyryladenosine 3',5'-cyclic monophosphate and all-trans-retinoic acid inhibited cell proliferation, but did not enhance GJIC or Cx43 expression. Surface epithelial cells of benign ovaries expressed Cx43, but this expression was barely detectable in ovarian serous cystadenocarcinomas. Thus, normal HOSE cells had extensive GJIC and Cx43 expression whereas ovarian carcinoma cells had less and cAMP and retinoic acid did not change these, although both agents inhibited cell growth.

Frequent reduction of gap junctional intercellular communication and connexin43 expression in human and mouse lung carcinoma cells

The reduced gap junctional intercellular communication (GJIC) and gap junction protein (connexin) expression that have been noted in many neoplastic cell types may contribute to the neoplastic phenotype. We assessed GJIC (by fluorescent dye micro-injection) and connexin expression (by Northern blotting, Western blotting and immunohistochemistry) in five mouse and 17 human lung carcinoma cell lines; both measures were lower in neoplastic cells compared to non-transformed lung epithelial cells. Other connexins were not detected in these cells. Co-culture experiments indicated that carcinoma cell lines able to transfer dye among themselves (homologous GJIC) had little capacity for dye-coupling with non-transformed cells (heterologous GJIC). Southern blot analyses indicated that reductions in GJIC and connexin43 expression were not due to deletions or rearrangements of this gene, but were more likely accounted for by transcriptional down-regulation and/or post-transcriptional factors. No correlations between GJIC and known oncogene and tumor suppressor gene alterations in the human lung carcinoma cells were apparent, suggesting that other mechanisms down-regulate GJIC in these cells. Since the neoplastic cell lines exhibited low GJIC (either homologous or heterologous), this characteristic may be involved in expression of the neoplastic phenotype.

Altered gap junctional intercellular communication in neoplastic rat esophageal epithelial cells

Gap junctional intercellular communication (GJIC) is reduced in many neoplastic cells, but few data exist for esophageal neoplasms. GJIC was examined by fluorescent dye microinjection in two nontumorigenic and two highly tumorigenic rat esophageal epithelial cell lines. All lines expressed high levels of dye coupling in homologous cell culture. In cocultures of nontumorigenic and tumorigenic cells, however, only one of six cell combinations displayed significant heterologous GJIC. Northern, Western, and immunohistochemical analyses indicated that all four cell lines expressed comparable levels of connexin43 (Cx43), but not connexin32 or connexin26, and formed Cx43-containing gap junction plaques at cell-cell interfaces. Immunostaining of rat esophageal frozen sections demonstrated that esophageal epithelial cells expressed Cx43 in vivo. In normal epithelium, the highest expression was seen in the basal cells and little suprabasal staining was evident. In preneoplastic and neoplastic lesions of the esophageal epithelium which were induced by treating rats with N-nitrosomethylbenzylamine, Cx43 staining of the basal layer was also seen but appeared to be more diffuse compared to normal epithelium. In addition, suprabasal Cx43 staining was apparent in dysplastic and papillomatous lesions. These results indicate that Cx43 is expressed in normal and neoplastic rat esophageal cells and that the cells exhibit extensive homologous GJIC, but little heterologous GJIC. This lack of heterologous GJIC may be due to differences in cell adhesion proteins or other factors.

Down-regulation by butylated hydroxytoluene of the number and function of gap junctions in epithelial cell lines derived from mouse lung and rat liver

The mouse pneumotoxicant and lung and liver tumor promoter butylated hydroxytoluene (BHT) was examined for its effects on gap junctional intercellular communication (GJIC) in mouse lung epithelial (C10) and rat liver epithelial (WB-F344) cell lines. GJIC, as measured by fluorescent dye microinjection, was inhibited in both types of cells by BHT in dose- and time-dependent fashions. Inhibition was detected in WB-F344 cells at BHT concentrations > or = 62.5 microM and in C10 cells at concentrations > or = 150 microM after 4 h treatment. Inhibition occurred within 15-30 min and was reversed by removing BHT from the culture medium. The highly toxic BHT metabolite 6-t-butyl-2-(hydroxy-t-butyl)-4-methylphenol (BHTOH) and the non-toxic BHT metabolite, 2,6-di-t-butyl-4-hydroxymethylphenol (BHTBzOH) were also tested. In both cell lines BHTOH was a more potent inhibitor of GJIC than BHT, whereas BHTBzOH was ineffective. The mechanisms of inhibition of GJIC by BHT were also examined. The initial rapid inhibition detected within 15-30 min may have been due to gap junction channel closure or blockage, since no changes in gap junction number, connexin (Cx) 43 levels or Cx43 phosphorylation were observed. By 2-4 h, however, gap junctions were internalized into the cytoplasm, the number of immunodetectable plasma membrane gap junctions was reduced and phosphorylated Cx43-P2 was decreased. Treatment of the cells for 24 h with 12-O-tetradecanoylphorbol-13-acetate (TPA) prevented inhibition of GJIC by TPA, but not by BHT. Western blot analyses of TPA-treated WB-F344 or C10 cells revealed the presence of a hyperphosphorylated form of Cx43 (Cx43-P3) and no reduction in Cx43-P2, in contrast to BHT-treated cells. These data suggest that BHT and TPA inhibit lung and liver epithelial cell GJIC through distinct mechanisms.

Oxytocin receptors: triggers for parturition and lactation?

Specific binding of tritiated oxytocin to uterine receptors of pregnant rats increases dramatically at term and is maximal during labor. In mammary glands the increase in binding is gradual, reaching a maximum during the lactation period. Concomitant changes in the sensitivity of the uterus and mammary gland to oxytocin indicate that the receptor concentration is of functional significance. Oxytocin receptors, therefore, may regulate the response of the target organs to circulating oxytocin and thereby control the onset of labor and lactation. Ovarian steroids participate in the regulation of oxytocin receptors in a manner as yet unclarified.