Connexin43 Signaling Contributes to

Download PDF
5 downloads 0 Views 625KB Size Report
Comment
Oct 14, 2011 - Both channel types displayed high activity near the .... environmental conditions with free access to food and water. .... after photobleaching (FRAP) analysis, cultured hepatocytes were loaded with ... (LDH) leakage was measured according to the Bergmeyer's procedure ..... In Methods of Enzymatic.
TOXICOLOGICAL SCIENCES 125(1), 175–186 (2012) doi:10.1093/toxsci/kfr277 Advance Access publication October 14, 2011

Connexin43 Signaling Contributes to Spontaneous Apoptosis in Cultures of Primary Hepatocytes Mathieu Vinken,*,1 Elke Decrock,† Tamara Vanhaecke,* Luc Leybaert,† and Vera Rogiers* *Department of Toxicology-Center for Pharmaceutical Research, Faculty of Medicine and Pharmacy, Vrije Universiteit Brussel, B-1090 Brussels, Belgium; and †Department of Basic Medical Sciences-Physiology group, Faculty of Medicine and Health Sciences, Ghent University, B-9000 Ghent, Belgium 1

To whom correspondence should be addressed at Department of Toxicology, Faculty of Medicine and Pharmacy, Vrije Universiteit Brussel, Laarbeeklaan 103, B-1090 Brussels, Belgium. Fax: þ32-2-477-45-82. E-mail: [email protected]. Received June 24, 2011; accepted October 3, 2011

Primary hepatocyte cultures suffer from the progressive occurrence of dedifferentiation followed by spontaneous apoptosis. This is associated with modifications in the expression of connexins (Cxs), which are the building stones of hemichannels that in turn form gap junctions between neighboring cells. Specifically, a shift is observed from the adult hepatocellular Cx32 species toward the fetal Cx43 isoform. The current study was set up to investigate the role of Cx43 in spontaneous apoptosis taking place in primary hepatocyte cultures. For this purpose, freshly isolated adult rat hepatocytes were cultivated in conventional conditions for 4 days with daily monitoring of Cx expression, Cx localization, and gap junction channel and hemichannel functionality. Gap junction activity was low shortly after isolation, whereas the inverse was observed for hemichannel functionality. Both channel types displayed high activity near the end stages of the cultivation period. The Cx32-to-Cx43 switch became progressively manifested at the translational level. At the transcriptional level, a fivefold decrease in Cx32 messenger RNA abundance and a twofold increase in Cx43 expression were noticed within the first 24 h of cultivation. Throughout the cultivation period, Cx32 was mainly located at the plasma membrane surface, whereas Cx43 immunostaining was more diffuse. Application of three Cx43 inhibitors resulted in the downregulation of both hemichannel functionality and gap junction activity. This was paralleled by decreased expression and activity of caspase 3 as well as by reduced expression of Bid. Collectively, these data show that Cx43 signaling actively contributes to the occurrence of spontaneous apoptosis in cultures of primary hepatocytes. Key Words: apoptosis; primary hepatocyte; connexin43; hemichannel; gap junction.

Cultures of primary hepatocytes, typically isolated from a freshly removed liver by means of a two-step collagenase perfusion technique, represent versatile in vitro tools that are widely used in the field of hepatic physiology, pathology, and pharmacotoxicology. As such, they provide a good reflection of the hepatic in vivo situation, at least during short-term

regimes (Elaut et al., 2006; Papeleu et al., 2006b). Indeed, long-term cultivation of primary hepatocytes is largely impeded by the progressive loss of the hepatocyte-specific phenotype both at the morphological and at the functional level. This dedifferentiation process is initiated during the isolation procedure, whereby cellular contacts are abolished and considerable ischemia/reperfusion injury occurs. These deleterious events, in turn, trigger an inflammatory reaction and a proliferative response, both of which negatively affect liver-specific gene expression patterns and hence the overall differentiated hepatocellular status (Elaut et al., 2006; Paine and Andreakos, 2004; Papeleu et al., 2006b; Vinken et al., 2006b). When seeded in a conventional monolayer configuration on a plastic culture dish, the dedifferentiation process manifests gradually and ultimately burgeons into the generalized onset of cell death by apoptosis after 4 days of cultivation (Bailly-Maitre et al., 2002; Elaut et al., 2006; Papeleu et al., 2006b; Vinken et al., 2006b). A prominent feature of hepatocyte dedifferentiation relates to modifications that occur in the expression of connexin (Cx) proteins (Elaut et al., 2006; Vinken et al., 2006a,b). The latter are the molecular constituents of hemichannels, which provide a pathway for exchange of critical messengers and metabolites, such as adenosine triphosphate (ATP), between the intracellular compartment and the extracellular environment. Hemichannels from adjacent cells can further interact to form gap junctions, which also convey essential homeostasis regulators, albeit in an extracellular circuit (Decrock et al., 2009; Vinken et al., 2009, 2011). Cxs are produced in a cell type–specific way. Thus, in the adult liver, Cx32 is the major Cx expressed by hepatocytes, whereas nonparenchymal liver cells, including Kupffer cells, stellate cells, endothelial cells, and cells of Glisson’s capsule, preferentially harbor Cx43 (Vinken et al., 2006b, 2008). Furthermore, Cx expression patterns undergo substantial changes during development. Early hepatic progenitor cells switch from Cx43 to Cx32 expression during differentiation into adult liver parenchymal cells (Naves et al., 2001; Neveu et al., 1995; Paku

Ó The Author 2011. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For permissions, please email: [email protected]

176

VINKEN ET AL.

et al., 2004; Zhang and Thorgeirsson, 1994). Upon dedifferentiation, however, as seen in several pathological conditions in vivo (Krutovskikh et al., 1994; Oyamada et al., 1990) as well as in cultures of primary hepatocytes (Kojima et al., 1995; Stutenkemper et al., 1992; Vinken et al., 2006a; Willecke and Haubrich, 1996), a reverse process takes place, namely, a shift from Cx32 to Cx43 expression. The functional relevance of this phenomenon is not clear. Recently, it has been reported that Cx43 expression becomes induced in hepatocytes of adult rats that have been administered acetaminophen. Cx43 is hereby colocalized with caspase 3 in the liver parenchymal cells, suggesting a function for this Cx species in hepatocyte apoptosis (Naiki-Ito et al., 2010). The present study was set up to point out whether Cx43 fulfills a similar function in cultures of primary hepatocytes. In the first part, an in-depth molecular scenario is provided with respect to the modifications that occur in Cx expression at the transcriptional, translational, and activity level. In the second part, Cx43 signaling in the hepatic in vitro setting is characterized with a set of Cx43 inhibitors and their outcome on apoptotic cell death is tested.

plexes was carried out as previously described (Vandenbroucke et al., 2008). The siRNA transfection was started 20 h after cell seeding by replacing the initial culture medium by hepatocyte culture medium containing the PbAE2/ siRNA complexes (final siRNA concentration 100nM). After 5 h, the culture medium was removed and replaced by regular hepatocyte culture medium. The hepatocyte culture medium was renewed another two times (19 and 43 h posttransfection), and sampling was performed 72 h posttransfection. Experiments using both unlabeled and fluorescein isothiocyanate–labeled nontargeting siRNA (Dharmacon) were performed in parallel. Cells transfected with the latter were analyzed by light and fluorescence microscopy (Cetec, Belgium) at 3400 magnification. Stained cells were counted, and transfection efficiency was averaged 79 ± 7.0% (n ¼ 3), and no cytotoxicity was observed after the 5-h transfection period (Supplementary fig. S1).

Hepatocyte isolation and cultivation. Procedures for the housing of rats, and isolation and cultivation of hepatocytes were approved by the local ethical committee of the Vrije Universiteit Brussel (Belgium). Male outbred SpragueDawley rats (Charles River Laboratories, Belgium) were kept under controlled environmental conditions with free access to food and water. Hepatocytes were isolated by use of a two-step collagenase method, including purification by serial differential centrifugation, and cell viability was assessed by trypan blue exclusion (Papeleu et al., 2006a). Viable ( 85%) hepatocytes were plated at a density of 0.56 3 105 cells per square centimeter in William’s medium E (Invitrogen, Belgium) supplemented with 7 ng/ml glucagon, 292 mg/ml L-glutamine, antibiotics (7.33 International Units of sodium benzyl penicillin, 50 lg/ml kanamycin monosulfate, 10 lg/ml sodium ampicillin, and 50 lg/ml streptomycin sulfate), and 10% fetal bovine serum. After 4, 24, 48, and 72 h, the cell culture medium was removed and replaced by serum-free medium supplemented with 25 lg/ml hydrocortisone sodium hemisuccinate and 0.5 lg/ ml insulin (‘‘hepatocyte culture medium’’). Samples were taken 4 h after cell seeding (T4) and on day 1 (D1), day 2 (D2), day 3 (D3), and day 4 (D4) of the cultivation period.

Immunoblotting. For the preparation of total protein lysates, hepatocytes were harvested from culture plates by scraping and washed twice with cold PBS. Cells were homogenized in lysis buffer (50mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid [Hepes], 150mM NaCl, 1mM ethylenediamine tetra-acetic acid, 2.5mM ethyleneglycol tetra-acetic acid, 0.1% Tween 20, and 10% glycerol) supplemented with 0.1mM phenylmethylsulfonylfluoride, 5mM dithiothreitol, and protease inhibitor cocktail (Roche, Germany). Following sonication for 10 s, samples were left on ice for 1 h. Cell lysates were centrifugated at 13,791 3 g for 5 min, and protein concentrations were determined in the supernatants according to the Bradford’s procedure (Bradford, 1976) using a Bio-Rad protein assay kit (Bio-Rad, Germany), with bovine serum albumin as a standard. For separation of Triton X-100 soluble and Triton X-100 insoluble fractions, cultured hepatocytes were washed with cold PBS and were harvested in cold 1% Triton X-100 in PBS supplemented with 50mM sodium fluoride, 1mM sodium orthovanadate, 1% protease inhibitor cocktail (Sigma), 1% phosphatase inhibitor cocktails (Sigma), and ethylenediamine tetra-acetic acid–free protease inhibitor cocktail (Roche). Separation into Triton X-100 soluble and Triton X-100 insoluble fractions was performed by centrifugation at 16,060 3 g at 4°C for 10 min. Triton X-100 insoluble fractions were resuspended in Laemmli sample buffer (125mM Tris, 80mM sodium dodecyl sulfate, and 10% glycerol) and sonicated on ice for 30 s. Protein concentrations were determined according to the Bradford’s procedure (Bradford, 1976) using a Bio-Rad protein assay kit. Proteins were fractionated on sodium dodecyl sulfate polyacrylamide and blotted afterward onto nitrocellulose membranes (Amersham, U.K.). Blocking of the membranes was performed with 5% nonfatty milk in Tris-buffered saline solution (TBS; 20mM Tris, 135mM NaCl) containing 0.1% Tween 20. Membranes were incubated overnight at 4°C with a primary antibody directed against Cx32 (Sigma), Cx43 (Sigma), caspase 3 (Calbiochem, Belgium), or Bid (R&D Systems, U.K.), followed by incubation for 1 h at room temperature with appropriate horseradish peroxidase–conjugated secondary antibodies (Dako, Denmark). Excessive antibody was removed by washing the membranes in Tween-supplemented TBS. Detection of the proteins was carried out by means of an enhanced chemiluminescence Western blotting system (Amersham). For semiquantification of the results, blots were further incubated with a monoclonal anti-mouse b-actin antibody (Abcam, U.K.). Blots were scanned, and densitometric analyses were performed by using the Quantity One software (Bio-Rad). For Cx32 and Cx43 in the cultivation time–related experiments, signals at D1, D2, D3, and D4 were normalized to the corresponding b-actin signals and were expressed as percentage of the normalized Cx32 and Cx43 signals at T4. For caspase 3, Bid, and Cx43 in the inhibitor experiments, signals of the treated conditions were normalized to the corresponding b-actin signals and were expressed as percentage of the normalized caspase 3 and Bid signals of the control condition.

Gene silencing by small interfering RNA transfection. Cx43 expression was suppressed by small interfering RNA (siRNA) treatment with ONTARGET plus SMARTpool siRNA from Dharmacon (Belgium), which contains four different siRNA duplexes directed against rat gja1 (sense strand sequences: siRNA number 1: 5#-CAACAACCUGGCUGCGAAAUU-3#; siRNA number 2: 5#-UGAUUGAAAUGUCGAGUUAUU-3#; siRNA number 3: 5#-CGUGAAGGGAAGAAGCGAUUU-3#; and siRNA number 4: 5#UUACUGAGAUUCUGCGAUAUU-3#). Preparation of PbAE2/siRNA com-

Immunocytochemistry. Hepatocytes, grown on glass coverslips, were fixed for 10 min with cold ethanol (70%). Following rehydratation with PBS, cells were permeabilized with 0.1% Triton X-100 and blocked with 0.1% nonfatty milk, each for 30 min. Cells were then incubated with antibodies directed against Cx32 and Cx43 for 2 h, washed with PBS, and exposed to a DyLight-conjugated secondary antibody (Jackson) for 45 min. After extensive rinsing with PBS, samples were mounted with diamidinophenylindolecontaining Vectashield (Vector Laboratories). Samples were kept at 4°C prior

MATERIALS AND METHODS Chemicals and reagents. Acetyl-Asp-Glu-Val-Asp-7-amino-4-trifluoromethylcoumarin (Ac-DEVD-AFC) and carbenoxolone were purchased from Merck (Belgium) and Sigma (Belgium), respectively. The 43Gap27 peptide (SRPTEKTIFII) was synthesized by ThermoFisher Scientific (Germany). The 1,6-hexanediol diacrylate–based poly-beta-aminoester (PbAE2) was synthesized as described elsewhere (Vandenbroucke et al., 2008). All other chemicals were commercially available products of analytical grade and were supplied by Sigma, unless specified otherwise.

CONNEXIN43 AND HEPATOCELLULAR APOPTOSIS

177

FIG. 1. Gap junction activity and hemichannel activity in cultures of primary rat hepatocytes. Freshly isolated rat hepatocytes were cultivated in a monolayer configuration. Four hours after cell seeding (T4) and on D1, D2, D3, and D4 of the cultivation period, fluorescence recovery after photobleaching (FRAP) analysis (A) and extracellular divalent ion–triggered cellular ATP release (B) were measured as described in the ‘‘Materials and Methods’’ section, indicating gap junction activity and hemichannel activity, respectively. For the gap junction assay, at least six cells per culture dish were examined. Fluorescence in the bleached cells was expressed as the percent recovery relative to the prebleach level. Data were expressed as mean ± SD of three independent experiments. For the hemichannel assay, the release of ATP was measured under basal (baseline: white bars) and induced (divalent-free medium: black bars) conditions and was expressed as the percentage of ATP release triggered by the divalent-free medium at T4. Data were expressed as mean ± SD of three independent experiments. Results were evaluated by oneway ANOVA followed by post hoc Bonferroni tests. Asterisks indicate significant differences between the divalent-free medium condition and the corresponding baseline condition per time point (*p < 0.05, ***p < 0.001). Number signs indicate significant differences between the divalent-free medium conditions at the different time points and T4 (#p < 0.05, ##p < 0.01).

to analysis. Detection was performed by fluorescence microscopy (Leica DMR/XA, Belgium). For negative controls, an identical procedure was followed, but primary antibodies were omitted (data not shown). Quantitative real-time PCR. Cells were harvested from culture plates by scraping and washed twice with cold PBS. Total cellular RNA was extracted using the SV Total RNA isolation system (Promega Corp.) according to the manufacturer’s instructions. Samples were subsequently subjected to DNase treatment (Ambion). Reverse transcription of approximately 2 lg of total RNA was carried out using the iScript complementary DNA (cDNA) synthesis kit

(Bio-Rad). For quantitative real-time PCR (qRT-PCR) analysis, a reaction mix was prepared in RNase-free water (final volume 25 ll) containing 2 ll cDNA preparation, 23 TaqMan Universal PCR Master Mix (Applied Biosystems), and 203 TaqMan Gene Expression Assays. Gene expression assays for Cx32 (accession number NM_017251.2), Cx43 (accession number NM_012567.2), glyceraldehyde-3-phosphate dehydrogenase (GAPDH; accession number NM_017008.3), 18S ribosomal RNA (accession number X03205.1), and b-actin (accession number NM_031144.2) were from Applied Biosystems with assay IDs Rn01641031_s1, Rn01433957_m1, Rn99999916_s1, Hs99999901_s1, and Rn00667869_m1, respectively. A number of housekeeping genes in

178

VINKEN ET AL.

FIG. 2. Cx32 and Cx43 protein expression in cultures of primary rat hepatocytes. Freshly isolated rat hepatocytes were cultivated in a monolayer configuration. Samples were taken 4 h after cell seeding (T4) and on D1, D2, D3, and D4 of the cultivation period and subjected to immunoblot analysis as described in the ‘‘Materials and Methods’’ section using primary antibodies that specifically recognize Cx32 and Cx43 (A). In this analysis, rat liver homogenate (LH) was included in order to allow comparison with hepatic Cx protein steady-state levels in vivo. For semiquantification of the results, Cx32 and Cx43 signals at D1, D2, D3, and D4 were normalized to the corresponding b-actin signals and were expressed as percentage of the normalized Cx32 (B) and Cx43 signals (C) at T4 (indicated with dashed lines). Data were expressed as mean ± SD of three independent experiments. (P2-Cx43, phosphorylated Cx43 variant; NP-Cx43, nonphosphorylated Cx43 variant.)

cultures of primary rat hepatocytes was tested, including GAPDH, 18S rRNA, and b-actin (Vinken et al., 2010), and their average expression was used for normalization in the DDCt method. At all time points, Ct values of Cx32 and Cx43 were normalized to those of the averaged housekeeping genes. The resulting DCt values of the D1, D2, D3, and D4 samples were then normalized to those of the T4 samples, yielding DDCt. Relative alterations (percentage) in messenger RNA (mRNA) levels were calculated according to the formula 2(DDCt).

was added and luminescence was measured. ATP release was expressed as the percentage of ATP release triggered by divalent-free medium.

Fluorescence recovery after photobleaching. For fluorescence recovery after photobleaching (FRAP) analysis, cultured hepatocytes were loaded with 10lM calcein-acetoxymethyl ester in Hank’s balanced salt solution buffered with Hepes (HBSS-Hepes; 0.81mM MgSO4.7H2O, 0.95mM CaCl2.2H2O, 137mM NaCl, 0.18mM Na2HPO4.2H2O, 5.36mM KCl, 0.44mM KH2PO4, 5.55mM D-glucose, and 25mM Hepes) for 30 min at room temperature. After extensive rinsing, cultures were kept for 10 min in HBSS-Hepes. Fluorescence within a single cell was photobleached by 1 s spot exposure to 488 nm Argon laser light, and dye influx from neighboring cells was recorded over the next 5 min with a 403 water immersion objective (Nikon, Japan). Fluorescence in the bleached cell was expressed as the percent recovery relative to the prebleach level. At least six cells per culture dish were examined.

Measurement of caspase 3–like activity. Cultured hepatocytes were washed twice with cold PBS, resuspended in lysis buffer (10mM Hepes, 2mM ethylenediamine tetra-acetic acid, 1mM phenylmethylsulfonylfluoride, 14.6lM pepstatin A, 21.6lM aprotinin, 3.1lM leupeptin, 2mM 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate, and 5mM dithiothreitol) and subjected to three freeze-thawing cycles. Cell lysates were centrifuged at 1559 3 g for 30 min at 4°C, and protein concentrations were determined according to the Bradford’s procedure (Bradford, 1976) using a Bio-Rad protein assay kit, with bovine serum albumin as a standard. Subsequently, 30 lg protein was added to reaction buffer (10mM Pipes, 2mM ethylenediamine tetra-acetic acid, 5mM dithiothreitol, and 1.6mM 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate). Then, 1.375mM of Ac-DEVD-AFC substrate was added, and after 1 h incubation at 37°C, fluorescence was measured at an excitation wavelength of 400 nm and an emission wavelength of 505 nm in a Victor plate reader (PerkinElmer, Belgium). The background readings from the buffer were subtracted from the readings from the samples, and the enzyme activity was expressed as nmol AFC/min 3 lg protein.

Measurement of extracellular ATP. ATP was measured using a commercial luciferin/luciferase assay kit (Sigma). Briefly, cultured hepatocytes were washed with divalent-free buffer (137mM NaCl, 0.18mM Na2HPO4.2H2O, 5.36mM KCl, 0.44mM KH2PO4, 4mM ethylene glycol tetra-acetic acid, 5.55mM D-glucose, and 25mM Hepes) and incubated for 2.5 min with divalentfree buffer at room temperature. For baseline measurements, divalent-free buffer was replaced by HBSS-Hepes. After 2.5 min, ATP assay mix in HBSS-Hepes

Measurement of lactate dehydrogenase leakage. Lactate dehydrogenase (LDH) leakage was measured according to the Bergmeyer’s procedure (Bergmeyer, 1974) using a commercial kit (Merck, Germany). Percentage of LDH leakage was calculated by the following equation: (100 3 LDH activity in supernatant)/[LDH activity in (supernatant þ cells)].

Statistical analysis. Data were expressed as means ± SDs of at least three independent experiments (n  3). Results were evaluated by one-way ANOVA (repeated measures when appropriate), followed by post hoc Bonferroni tests.

CONNEXIN43 AND HEPATOCELLULAR APOPTOSIS

179

FIG. 3. Connexin32 and connexin43 protein localization in cultures of primary rat hepatocytes. Freshly isolated rat hepatocytes were cultivated in a monolayer configuration. Samples were taken 4 h after cell seeding (T4) and on day 1 (D1), day 2 (D2), day 3 (D3), and day 4 (D4) of the cultivation period. (A) Samples were subjected to immunocytochemistry analysis as described in the ‘‘Materials and Methods’’ section, using primary antibodies that specifically recognize connexin32 (Cx32) and connexin43 (Cx43), followed by exposure to a DyLight-conjugated secondary antibody. Nuclear counterstaining was performed with diamidinophenylindole. Results shown are representative for at least three independent experiments (630x magnification). (B) Triton X-100 protein extraction was performed on samples, and the Triton X-100 soluble (cytosolic) fraction and the Triton X-100 insoluble (membrane) fractions were subjected to immunoblot analysis using primary antibodies that specifically recognize connexin32 (Cx32), connexin43 (Cx43), and b-actin as described in the ‘‘Materials and Methods’’ section. (P2-Cx43, phosphorylated Cx43 variant; NP-Cx43, nonphosphorylated Cx43 variant.

RESULTS

Gap Junction and Hemichannel Activity in Cultures of Primary Hepatocytes Upon two-step collagenase isolation of hepatocytes from the rat liver, the vast array of intercellular contacts becomes abolished, resulting in the loss of normal cytoarchitecture (Elaut et al., 2006; Papeleu et al., 2006b; Vinken et al., 2006b). Consequently, gap junctions are also negatively affected during this process, which is likely to explain the absence of gap junctional intercellular communication 4 h after cell seeding (T4), as judged on the low FRAP recovery (Fig. 1A). At this time point, however, extracellular divalent ion–triggered cellular ATP release, indicative for hemichannel activity, is relatively high (Fig. 1B), suggesting that disruption of gap junctions directly yields functional hemichannels in the plasma membrane surface of formerly neighboring cells. After 12 h of cultivation, the hepatocytes start to regain their normal morphology, and after about 24 h, de novo intercellular contacts reappear (Elaut et al., 2006; Vinken et al., 2006b). This is associated with the formation of active gap junctions at the expense of their half channel counterparts, which is particularly seen on D2 (Figs. 1A and 1B). Thereafter, gap junctional intercellular communication

remains at a relatively stable plateau level, whereas hemichannel activity increases toward the final day (D4) of the cultivation period (Figs. 1A and 1B). Cx32 Expression and Localization in Cultures of Primary Hepatocytes Cx32 is the major gap junction building stone in the liver, comprising as much as 90% of the total hepatic Cx amount and being located between hepatocytes (Vinken et al., 2006b, 2008). Western blot analysis indeed reveals large quantities of Cx32 in freshly prepared liver homogenate, which is representative for the hepatic in vivo situation (Fig. 2A). Upon isolation and subsequent cultivation of hepatocytes, however, Cx32 protein expression progressively deteriorates, with only minimal Cx32 immunoreactivity manifested on the last day of cultivation (Figs. 2A and 2B). Additionally, the deleterious outcome of the isolation procedure on gap junctions is also clearly visualized by immunocytochemistry analysis, showing several unopposed Cx32-positive spots at the cell plasma membrane surface at T4 and D1. Although the typical punctuate Cx32 staining pattern at the borders between cells is restored near D2, there seems to be tendency for a more cytoplasmic localization on D3 and D4 (Fig. 3A). The close

180

VINKEN ET AL.

FIG. 4. Cx32 and Cx43 gene expression in cultures of primary rat hepatocytes. Freshly isolated rat hepatocytes were cultivated in a monolayer configuration. Samples were taken 4 h after cell seeding (T4) and on D1, D2, D3, and D4 of the cultivation period and subjected to qRT-PCR analysis as described in the ‘‘Materials and Methods’’ section. At all time points, Ct values of (A) Cx32 and (B) Cx43 were normalized to averaged GAPDH, 18S rRNA, and b-actin Ct values. The resulting DCt values of the D1, D2, D3, and D4 samples were then normalized to those of the T4 samples, yielding DDCt. Relative alterations (percentage) in mRNA levels were calculated according to the formula 2(DDCt). Data were expressed as mean ± SD of three independent experiments.

association of Cx32 with the plasma membrane, especially at early time points of the cultivation regime, is also observed upon immunoblot analysis of Triton X-100 soluble and Triton X-100 insoluble fractions (Fig. 3B). Notably, the modifications in Cx32 protein content are not reflected at the transcriptional level, where a dramatic drop in Cx32 mRNA is observed between 4 and 24 h postplating. This is followed by a progressive though only partial recovery of the initial Cx32 mRNA abundance at T4 toward the final cultivation day (Fig. 4A).

Cx43 Expression and Localization in Cultures of Primary Hepatocytes In contrast to Cx32, Cx43 is the predominant Cx species that is expressed by the nonparenchymal cell population in adult liver as well as by hepatocytes in fetal liver (Vinken et al., 2008). The former is evidenced by Cx43 immunoblot analysis of liver homogenate of adult rat liver, showing the preferential presence of the nonphosphorylated Cx43 variant (Fig. 2A). It has been reported on several occasions that Cx43, especially its nonphosphorylated isoform, becomes increasingly detectable

CONNEXIN43 AND HEPATOCELLULAR APOPTOSIS

181

FIG. 5. Effects of Cx43 inhibitors on gap junction activity and hemichannel activity in cultures of primary rat hepatocytes. Freshly isolated rat hepatocytes were cultivated in a monolayer configuration. On D1, a subset of cell cultures was transfected with 100nM siRNA directed against Cx43, nontargeting siRNA, or mock for 5 h as described in the ‘‘Materials and Methods’’ section. On D3, another set of cell cultures was exposed to either 50lM carbenoxolone or 191lM 43 Gap27 for 24 h. For all three inhibitor strategies, sampling was performed on D4. (A) Gap junction activity was tested by means of fluorescence recovery after photobleaching (FRAP) analysis, as described in the ‘‘Materials and Methods’’ section, whereby at least six cells per culture dish were examined. Fluorescence in the bleached cells was expressed as the percent recovery relative to the prebleach level. Data were expressed as mean ± SD of three independent experiments. Results were evaluated by one-way ANOVA followed by post hoc Bonferroni tests. Asterisks indicate significant differences compared with untreated cells (*p < 0.05, **p < 0.01, ***p < 0.001). (B) Hemichannel activity was probed through extracellular divalent ion–triggered cellular ATP release as described in the ‘‘Materials and Methods’’ section. The release of ATP was measured under basal (baseline: white bars) and induced (divalent-free medium: black bars) conditions and was expressed as the percentage of ATP release triggered by the divalent-free medium in untreated cells. Data were expressed as mean ± SD of three independent experiments. Results were evaluated by one-way ANOVA followed by post hoc Bonferroni tests. Asterisks indicate significant differences between the divalent-free medium condition and the corresponding baseline condition per treatment (*p < 0.05, **p < 0.01, ***p < 0.001). Number signs indicate significant differences between the divalent-free medium condition of a given treatment and that of untreated cells (#p < 0.05, ##p < 0.01).

upon establishing cultures of freshly isolated adult hepatocytes (Kojima et al., 1995; Stutenkemper et al., 1992; Vinken et al., 2006a; Willecke and Haubrich, 1996), a feature that is also reproduced in the experimental setting used in the present

study. Indeed, a progressive elevation in Cx43 protein amounts is observed, reaching 412.6 ± 25.9% of the T4 level at the final day of cultivation (Figs. 2A–C). Unlike Cx32, the modifications in the Cx43 protein pattern are paralleled by identical

182

VINKEN ET AL.

FIG. 6. Effects of Cx43 inhibitors on Cx43 protein expression in cultures of primary rat hepatocytes. Freshly isolated rat hepatocytes were cultivated in a monolayer configuration. On D1, a subset of cell cultures was transfected with 100nM siRNA directed against Cx43, nontargeting siRNA, or mock for 5 h as described in the ‘‘Materials and Methods’’ section. On D3, another set of cell cultures was exposed to either 50lM carbenoxolone or 191lM 43Gap27 for 24 h. For all three inhibitor strategies, samples were taken on D4 and subjected to immunoblot analysis as described in the ‘‘Materials and Methods’’ section using primary antibodies that specifically recognize Cx32 and Cx43 (A). For semiquantification of the results, Cx43 signals of the different conditions were normalized to the corresponding b-actin signals and were expressed as percentage of the normalized Cx43 signal in the untreated condition (indicated with a dashed line). Data were expressed as mean ± SD of three independent experiments. Results were evaluated by one-way ANOVA followed by post hoc Bonferroni tests. Asterisks indicate significant differences compared with the untreated condition (***p < 0.001). Number signs indicate significant differences compared with the nontargeting siRNA condition (###p < 0.001) (B). The following conditions have been tested: C, control; 1, mock transfection; 2, nontargeting siRNA; 3, Cx43 siRNA; 4, 43Gap27; 5, carbenoxolone. (P2-Cx43, phosphorylated Cx43 variant; NP-Cx43, nonphosphorylated Cx43 variant.)

changes at the transcriptional platform, with as much as a sevenfold increase at D4 (Fig. 4B). Throughout the entire cultivation period, Cx43 protein is diffusely distributed among plasma membrane and intracellular structures, particularly in the perinuclear area, as evidenced through combined immunocytochemistry and Western blot analysis of Triton X-100 soluble and Triton X-100 insoluble cellular portions (Figs. 3A–B). Characterization of Cx43 Signaling and Inhibitors in Cultures of Primary Hepatocytes In order to unravel the functional relevance of Cx43 appearance in primary hepatocyte cultures, a set of three different inhibitors was addressed, namely, (1) carbenoxolone (50lM, 24-h treatment), which acts as a general inhibitor of Cx-related signaling, (2) 43Gap27 mimetic peptide (191lM,

24-h treatment) that specifically suppresses the activity of Cx43-based gap junctions and hemichannels, and (3) Cx43 siRNA duplexes (100nM, 5-h treatment), which silence the upstream expression of Cx43 (Decrock et al., 2009). These approaches were applied in parallel, and their outcome was evaluated on D4 of the cultivation period. The three experimental strategies were void of cytotoxic potential, as concluded from the unaltered extracellular release of LDH (Supplementary fig. S2), a cytosolic enzyme that leaks from the cell upon plasma membrane damage (Bergmeyer, 1974). As expected, all inhibitors downregulated calcein FRAP recovery (Fig. 5A). Simultaneously, low extracellular divalent ion–triggered cellular ATP release was negatively affected (Fig. 5B). Reduction of Cx43 expression, however, was only observed in Cx43 siRNA-treated primary hepatocyte

183

CONNEXIN43 AND HEPATOCELLULAR APOPTOSIS

cultures. This effect was not due to the RNA interference approach as such because Cx43 protein content was not modified in hepatocytes exposed to nontargeting siRNA duplexes. Moreover, other relevant Cxs, in casu Cx32, remained unchanged under both siRNA conditions, an outcome equally seen for the other Cx43 inhibitor strategies (Figs. 6A and 6B). Effects of Cx43 Inhibitors on Spontaneous Apoptosis in Cultures of Primary Hepatocytes Induction of Cx43 expression in hepatocytes in vivo was recently found to be associated with the occurrence of apoptosis (Naiki-Ito et al., 2010). To verify whether this also holds true in vitro, cultured hepatocytes were exposed to the optimized Cx43 inhibitors in parallel, and the impact on caspase 3–like activity was investigated on D4 of the cultivation regime. For this purpose, Ac-DEVD-AFC a prototypical though not exclusive caspase 3 substrate (Bai et al., 2003) was used, which was previously reported to be a reliable and sensitive marker of apoptosis in cultures of primary hepatocytes (Go´mez-Lecho´n et al., 2002). The three experimental approaches commonly downregulated this parameter (Fig. 7). This finding was further substantiated by immunoblot analysis of caspase 3, showing lowered expression of its active 18 kDa fragment (Figs. 8A and 8B). Similarly, the proapoptotic Bid protein was negatively affected by the Cx43 inhibitors

FIG. 7. Effects of Cx43 inhibitors on caspase 3–like activity in cultures of primary rat hepatocytes. Freshly isolated rat hepatocytes were cultivated in a monolayer configuration. On D1, a subset of cell cultures was transfected with 100nM siRNA directed against Cx43, nontargeting siRNA, or mock for 5 h as described in the ‘‘Materials and Methods’’ section. On D3, another set of cell cultures was exposed to either 50lM carbenoxolone or 191lM 43Gap27 for 24 h. For all three inhibitor strategies, samples were taken on D4 and subjected to measurement of caspase 3–like activity as described in the ‘‘Materials and Methods’’ section. Results are expressed as nmol AFC/min 3 lg protein and represent mean ± SD of three independent experiments. Results were evaluated by one-way ANOVA followed by post hoc Bonferroni tests. Asterisks indicate significant differences compared with the untreated condition (*p < 0.05, **p < 0.01).

(Figs. 8A–C), thus again pointing to the reduced presence of apoptosis under these circumstances.

DISCUSSION

The functional deterioration that progressively occurs in cultures of primary hepatocytes is accompanied by drastic changes in Cx expression. Protein amounts of Cx32, which typifies the differentiated hepatocyte phenotype, hereby rapidly decline. As shown in the current study, this process is not reflected at the transcriptional level, suggesting the involvement of mechanisms that directly affect the Cx32 protein as such. This is in contrast to Cx43, for which a gradual increase in protein and mRNA abundance go hand in hand. It has been reported very recently that Cx43 expression in primary hepatocyte cultures relies on Wnt signaling and thus on the transcriptional T-cell factor/lymphoid enhancer factor–related machinery (Yamaji et al., 2011), a finding that could underlie the results of the present study. Interestingly, Cx43 is not only present at the plasma membrane surface but is also abundantly associated with cytoplasmic membrane structures. Such atypical localization has been observed in other cell types as well, including human (Krutovskikh et al., 1994; Oyamada et al., 1990), mouse (Tekpli et al., 2010), and rat (Ionta et al., 2009) hepatoma cells, whereby Cx43 resides in the Golgi apparatus (Tekpli et al., 2010) and mitochondria (Goubaeva et al., 2007; Ruiz-Meana et al., 2008). In fact, mitochondrial Cx43 has been found to play a role in the cellular response to ischemia/reperfusion injury and subsequent cell death (RuizMeana et al., 2008), a process that also occurs during two-step collagenase isolation of hepatocytes from the liver (Elaut et al., 2006). At the activity level, minimal gap junction functionality but high hemichannel activity was noted shortly after establishing primary hepatocyte cultures. This is likely to result from the disruption of Cx32-based gap junctions during the isolation procedure, which directly gives rise to functional hemichannels. On the other hand, the increase in activity of both channel types observed toward the end stages of the cultivation period is closely related to the emerging expression of Cx43. This was confirmed by experiments with different Cx43 inhibitors that target Cx43 expression, hemichannels, and gap junctions. In an attempt to elucidate the functional relevance of this induced Cx43-based signaling, it was hypothesized that Cx43 plays a role in the occurrence of spontaneous apoptosis, which accompanies the dedifferentiation process in primary hepatocyte cultures. In this respect, Naiki-Ito et al. (2010) recently showed that upon administration of acetaminophen to rats, de novo synthesized Cx43 is colocalized with caspase 3 in hepatocytes, suggesting a function for Cx43 in apoptosis. In the in vitro setting addressed in the current study, the Cx43 inhibitor approaches collectively triggered downregulation of caspase 3–like activity and expression as well as a concomitant decrease in the expression

184

VINKEN ET AL.

FIG. 8. Effects of Cx43 inhibitors on caspase 3 and Bid protein expression in cultures of primary rat hepatocytes. Freshly isolated rat hepatocytes were cultivated in a monolayer configuration. On D1, a subset of cell cultures was transfected with 100nM siRNA directed against Cx43, nontargeting siRNA, or mock for 5 h as described in the ‘‘Materials and Methods’’ section. On D3, another set of cell cultures was exposed to either 50lM carbenoxolone or 191lM 43Gap27 for 24 h. For all three inhibitor strategies, samples were taken on D4 and subjected to immunoblot analysis as described in the ‘‘Materials and Methods’’ section using primary antibodies that specifically recognize caspase 3 (active 18 kDa fragment) and Bid (A). For semiquantification of the results, caspase 3 signals (B) and Bid signals (C) of the different conditions were normalized to the corresponding b-actin signals and were expressed as percentage of the normalized caspase 3 and Bid signals of the untreated condition. Data were expressed as mean ± SD of three independent experiments. Results were evaluated by one-way ANOVA followed by post hoc Bonferroni tests. Asterisks indicate significant differences compared with the untreated condition (*p < 0.05, **p < 0.01). The following conditions have been tested: C, control; 1, mock transfection; 2, nontargeting siRNA; 3, Cx43 siRNA; 4, 43Gap27; 5, carbenoxolone.

of the proapoptotic Bid protein. This points to the active involvement of Cx43-dependent cellular communication through both hemichannels and gap junctions in the unprompted cell death phenomenon that takes place in primary hepatocyte cultures. In conclusion, the results of the present study show that in cultures of primary hepatocytes, a transition occurs from a Cx32-related signaling program to a Cx43-based communication modus, whereby the latter contributes to spontaneous apoptosis. In turn, this suggests that counteracting Cx43 expression and channel functionality might be a potential approach to increase the survival of cultured primary hepatocytes and thus to generate a liver-based in vitro tool that can be used for long-term purposes. In this regard, a number of anti-dedifferentiation strategies have been explored over the years, namely, the addition of differentiation-promoting soluble compounds to the hepatocyte culture medium (Kojima et al., 1995, 1996; Ren et al., 1994), the reestablishment of cell–cell contacts by cocultivating hepatocytes with another cell type (Guguen-Guillouzo et al., 1983; Mesnil et al., 1993), and the reintroduction of an extracellular matrix scaffold by cultivating hepatocytes on one layer or

between two layers of extracellular matrix proteins (Dunn et al., 1989; Fujita et al., 1987; Spray et al., 1987). All these methodologies also positively affect gap junctions though with rather limited success. A possible explanation for this shortcoming is that these classical anti-dedifferentiation strategies are basically focused toward reducing the consequences of the dedifferentiation process by mimicking the natural hepatocyte microenvironment in vitro. The outcome of the current study lays a foundation for a more site-directed approach to counteract dedifferentiation and concomitant spontaneous cell death in primary hepatocyte cultures by targeting its actual causes.

SUPPLEMENTARY DATA

Supplementary data are available online at http://toxsci. oxfordjournals.org/. FUNDING

Fund for Scientific Research Flanders (FWO-Vlaanderen); Research Council of the Vrije Universiteit Brussel (OZR-VUB);

CONNEXIN43 AND HEPATOCELLULAR APOPTOSIS

European Union (FP6 project carcinoGENOMICS and FP7/ Colipa projects DETECTIVE and HeMiBio).

ACKNOWLEDGMENTS

The authors wish to thank Mr Paul Claes and Miss Kirsten Leurs for their excellent technical assistance.

REFERENCES Bai, L., Wang, J., Yin, X. M., and Dong, Z. (2003). Analysis of apoptosis: Basic principles and procedures. In Essentials of Apoptosis: A Guide for Basic and Clinical Research (X. M. Yin and Z. Dong, Eds.), pp. 239–251. Humana Press Inc., Secaucus, NJ. Bailly-Maitre, B., de Sousa, G., Zucchini, N., Gugenheim, J., Boulukos, K. E., and Rahmani, R. (2002). Spontaneous apoptosis in primary cultures of human and rat hepatocytes: Molecular mechanisms and regulation by dexamethasone. Cell Death Differ. 9, 945–955. Bergmeyer, H. U. (1974). Lactate dehydrogenase. In Methods of Enzymatic Analysis (H. U. Bergmeyer, Ed.), Vol. 2, pp. 574–579. Academic Press, New York, NY. Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254.

185

rat hepatocytes: Use of serum-free L-15 medium supplemented with EGF and DMSO. J. Cell Sci. 108, 1347–1357. Krutovskikh, V., Mazzoleni, G., Mironov, N., Omori, Y., Aguelon, A. M., Mesnil, M., Berger, F., Partensky, C., and Yamasaki, H. (1994). Altered homologous and heterologous gap-junctional intercellular communication in primary human liver tumors associated with aberrant protein localization but not gene mutation of connexin 32. Int. J. Cancer 56, 87–94. Mesnil, M., Piccoli, C., and Yamasaki, H. (1993). An improved long-term culture of rat hepatocytes to detect liver tumour-promoting agents: Results with phenobarbital. Eur. J. Pharmacol. 248, 59–66. Naiki-Ito, A., Asamoto, M., Naiki, T., Ogawa, K., Takahashi, S., Sato, S., and Shirai, T. (2010). Gap junction dysfunction reduces acetaminophen hepatotoxicity with impact on apoptotic signaling and connexin 43 protein induction in rat. Toxicol. Pathol. 38, 280–286. Naves, M. M., Silveira, E. R., Dagli, M. L., and Moreno, F. S. (2001). Effects of beta-carotene and vitamin A on oval cell proliferation and connexin 43 expression during hepatic differentiation in the rat. J. Nutr. Biochem. 12, 685–692. Neveu, M. J., Hully, J. R., Babcock, K. L., Vaughan, J., Hertzberg, E. L., Nicholson, B. J., Paul, D. L., and Pitot, H. C. (1995). Proliferation-associated differences in the spatial and temporal expression of gap junction genes in rat liver. Hepatology 22, 202–212. Oyamada, M., Krutovskikh, V. A., Mesnil, M., Partensky, C., Berger, F., and Yamasaki, H. (1990). Aberrant expression of gap junction gene in primary human hepatocellular carcinomas: Increased expression of cardiac-type gap junction gene connexin 43. Mol. Carcinog. 3, 273–278.

Decrock, E., Vinken, M., De Vuyst, E., Krysko, D. V., D’Herde, K., Vanhaecke, T., Vandenabeele, P., Rogiers, V., and Leybaert, L. (2009). Connexin-related signaling in cell death: To live or let die? Cell Death Differ. 16, 524–536. Dunn, J. C., Yarmush, M. L., Koebe, H. G., and Tompkins, R. G. (1989). Hepatocyte function and extracellular matrix geometry: Long-term culture in a sandwich configuration. FASEB J. 3, 174–177.

Paine, A. J., and Andreakos, E. (2004). Activation of signalling pathways during hepatocyte isolation: Relevance to toxicology in vitro. Toxicol. In Vitro 18, 187–193.

Elaut, G., Henkens, T., Papeleu, P., Snykers, S., Vinken, M., Vanhaecke, T., and Rogiers, V. (2006). Molecular mechanisms underlying the dedifferentiation process of isolated hepatocytes and their cultures. Curr. Drug Metab. 7, 629–660.

Papeleu, P., Vanhaecke, T., Henkens, T., Elaut, G., Vinken, M., Snykers, S., and Rogiers, V. (2006a). Isolation of rat hepatocytes. Methods Mol. Biol. 320, 229–237.

Fujita, M., Spray, D. C., Choi, H., Saez, J. C., Watanabe, T., Rosenberg, L. C., Hertzberg, E. L., and Reid, L. M. (1987). Glycosaminoglycans and proteoglycans induce gap junction expression and restore transcription of tissue-specific mRNAs in primary liver cultures. Hepatology 7, 1S–9S. Go´mez-Lecho´n, M. J., O’Connor, E., Castell, J. V., and Jover, R. (2002). Sensitive markers used to identify compounds that trigger apoptosis in cultured hepatocytes. Toxicol. Sci. 65, 299–308. Goubaeva, F., Mikami, M., Giardina, S., Ding, B., Abe, J., and Yang, J. (2007). Cardiac mitochondrial connexin 43 regulates apoptosis. Biochem. Biophys. Res. Commun. 352, 97–103. Guguen-Guillouzo, C., Cle´ment, B., Baffet, G., Beaumont, C., MorelChany, E., Glaise, D., and Guillouzo, A. (1983). Maintenance and reversibility of active albumin secretion by adult rat hepatocytes co-cultured with another liver epithelial cell type. Exp. Cell Res. 143, 47–57. Ionta, M., Ferreira, R. A., Pfister, S. C., and Machado-Santelli, G. M. (2009). Exogenous Cx43 expression decrease cell proliferation rate in rat hepatocarcinoma cells independently of functional gap junction. Cancer Cell Int. 9, 22. Kojima, T., Mitaka, T., Mizuguchi, T., and Mochizuki, Y. (1996). Effects of oxygen radical scavengers on connexins 32 and 26 expression in primary cultures of adult rat hepatocytes. Carcinogenesis 17, 537–544. Kojima, T., Mitaka, T., Paul, D. L., Mori, M., and Mochizuki, Y. (1995). Reappearance and long-term maintenance of connexin32 in proliferated adult

Paku, S., Nagy, P., Kopper, L., and Thorgeirsson, S. S. (2004). 2-Acetylaminofluorene dose-dependent differentiation of rat oval cells into hepatocytes: Confocal and electron microscopic studies. Hepatology 39, 1353–1361.

Papeleu, P., Vanhaecke, T., and Rogiers, V. (2006b). Histone deacetylase inhibition: A differentiation therapy for cultured primary hepatocytes? Curr. Enzym. Inhib. 2, 91–104. Ren, P., de Feijter, A. W., Paul, D. L., and Ruch, R. J. (1994). Enhancement of liver cell gap junction protein expression by glucocorticoids. Carcinogenesis 15, 1807–1813. Ruiz-Meana, M., Rodrı´guez–Sinovas, A., Cabestrero, A., Boengler, K., Heusch, G., and Garcia-Dorado, D. (2008). Mitochondrial connexin43 as a new player in the pathophysiology of myocardial ischaemia-reperfusion injury. Cardiovasc. Res. 77, 325–333. Spray, D. C., Fujita, M., Saez, J. C., Choi, H., Watanabe, T., Hertzberg, E., Rosenberg, L. C., and Reid, L. M. (1987). Proteoglycans and glycosaminoglycans induce gap junction synthesis and function in primary liver cultures. J. Cell Biol. 105, 541–551. Stutenkemper, R., Geisse, S., Schwarz, H. J., Look, J., Traub, O., Nicholson, B. J., and Willecke, K. (1992). The hepatocyte-specific phenotype of murine liver cells correlates with high expression of connexin32 and connexin26 but very low expression of connexin43. Exp. Cell Res. 201, 43–54. Tekpli, X., Rivedal, E., Gorria, M., Landvik, N. E., Rissel, M., DimancheBoitrel, M. T., Baffet, G., Holme, J. A., and Lagadic-Gossmann, D. (2010). The B[a]P-increased intercellular communication via translocation of connexin-43 into gap junctions reduces apoptosis. Toxicol. Appl. Pharmacol. 242, 231–240.

186

VINKEN ET AL.

Vandenbroucke, R. E., De Geest, B. G., Bonne´, S., Vinken, M., Vanhaecke, T., Heimberg, H., Wagner, E., Rogiers, V., De Smedt, S. C., Demeester, J., et al. (2008). Prolonged gene silencing in hepatoma cells and primary hepatocytes after small interfering RNA delivery with biodegradable poly(beta-amino esters). J. Gene Med. 10, 783–794. Vinken, M., Decrock, E., De Vuyst, E., De Bock, M., Vandenbroucke, R. E., De Geest, B. G., Demeester, J., Sanders, N. N., Vanhaecke, T., Leybaert, L., et al. (2010). Connexin32 hemichannels contribute to the apoptotic-tonecrotic transition during Fas-mediated hepatocyte cell death. Cell. Mol. Life Sci. 67, 907–918. Vinken, M., Decrock, E., De Vuyst, E., Ponsaerts, R., D’hondt, C., Bultynck, G., Ceelen, L., Vanhaecke, T., Leybaert, L., and Rogiers, V. (2011). Connexins: Sensors and regulators of cell cycling. Biochim. Biophys. Acta 1815, 13–25. Vinken, M., Doktorova, T., Decrock, E., Leybaert, L., Vanhaecke, T., and Rogiers, V. (2009). Gap junctional intercellular communication as a target for liver toxicity and carcinogenicity. Crit. Rev. Biochem. Mol. Biol. 44, 201–222.

Vinken, M., Henkens, T., De Rop, E., Fraczek, J., Vanhaecke, T., and Rogiers, V. (2008). Biology and pathobiology of gap junctional channels in hepatocytes. Hepatology 47, 1077–1088. Vinken, M., Henkens, T., Vanhaecke, T., Papeleu, P., Geerts, A., Van Rossen, E., Chipman, J. K., Meda, P., and Rogiers, V. (2006a). Trichostatin a enhances gap junctional intercellular communication in primary cultures of adult rat hepatocytes. Toxicol. Sci. 91, 484–492. Vinken, M., Papeleu, P., Snykers, S., De Rop, E., Henkens, T., Chipman, J. K., Rogiers, V., and Vanhaecke, T. (2006b). Involvement of cell junctions in hepatocyte culture functionality. Crit. Rev. Toxicol. 36, 299–318. Willecke, K., and Haubrich, S. (1996). Connexin expression systems: To what extent do they reflect the situation in the animal? J. Bioenerg. Biomembr. 28, 319–326. Yamaji, S., Droggiti, A., Lu, S. C., Martinez-Chantar, M. L., Warner, A., and Varela-Rey, M. (2011). S-Adenosylmethionine regulates connexins subtypes expressed by hepatocytes. Eur. J. Cell Biol. 90, 312–322. Zhang, M., and Thorgeirsson, S. S. (1994). Modulation of connexins during differentiation of oval cells into hepatocytes. Exp. Cell Res. 213, 37–42.