Role of connexins in microvascular dysfunction during inflammation

In arterioles, a locally initiated diameter change can propagate rapidly along the vessel length (arteriolar conducted response), thus contributing to arteriolar hemodynamic resistance. The response is underpinned by electrical coupling along the arteriolar endothelial layer. Connexins (Cx; constituents of gap junctions) are required for this coupling. This review addresses the effect of acute systemic inflammation (sepsis) on arteriolar conduction and interendothelial electrical coupling. Lipopolysaccharide (LPS; an initiating factor in sepsis) and polymicrobial sepsis (24 h model) attenuate conducted vasoconstriction in mice. In cultured microvascular endothelial cells harvested from rat and mouse skeletal muscle, LPS reduces both conducted hyperpolarization-depolarization along capillary-like structures and electrical coupling along confluent cell monolayers. LPS also tyrosine-phosphorylates Cx43 and serine-dephosphorylates Cx40. Since LPS-reduced coupling is Cx40- but not Cx43-dependent, only Cx40 dephosphorylation may be consequential. Nitric oxide (NO) overproduction is critical in advanced sepsis, since the removal of this overproduction prevents the attenuated conduction. Consistently, (i) exogenous NO in cultured cells reduces coupling in a Cx37-dependent manner, and (ii) the septic microvasculature in vivo shows no Cx40 phenotype. A complex role emerges for endothelial connexins in sepsis. Initially, LPS may reduce interendothelial coupling and arteriolar conduction by targeting Cx40, whereas NO overproduction in advanced sepsis reduces coupling and conduction by targeting Cx37 instead.     

Flow regulates intercellular communication in HAEC by assembling functional Cx40 and Cx37 gap junctional channels

Direct cell-to-cell transfer of ions and small signaling molecules via gap junctions plays a key role in vessel wall homeostasis. Vascular endothelial gap junctional channels are formed by the connexin (Cx) proteins Cx37, Cx40, and Cx43. The mechanisms regulating connexin expression and assembly into functional channels have not been fully identified. We investigated the dynamic regulation of endothelial gap junctional intercellular communication (GJIC) by fluid flow and the participation of each vascular connexin in functional human endothelial gap junctions in vitro. Human aortic endothelial cells (HAEC) were exposed for 5, 16, and 24 h to physiological flows in a parallel-plate flow chamber. Connexin protein expression and localization were evaluated by immunocytochemistry, and functional GJIC was evaluated by dye injection. Connexin-mimetic peptide inhibitors were used to assess the specific connexin composition of functional channels. HAEC monolayers in culture exhibited baseline functional communication at a striking low level despite abundant expression of Cx43 and Cx40 localized at cell-to-cell appositions. Upon exposure to flow, GJIC by dye spread demonstrated a significant time-dependent increase from baseline levels, reaching 7.5-fold in 24 h. Inhibition studies revealed that this response was mediated primarily by Cx40, with lesser contributions of the other two vascular connexins assembled into functional homotypic and/or heterotypic channels. This is the first study to demonstrate that flow simultaneously and differentially regulates expression of the Cx37, Cx40, and Cx43 proteins and their involvement in the augmentation of intercellular communication by dye transfer in human endothelial cells in vitro.     

A Neuroprotective Role for Gap Junctions

Glial-neuronal interactions have been implicated in both normal information processing and neuroprotection. One pathway of cellular interactions involves gap junctional intercellular communication (GJIC). In astrocytes, gap junctions are composed primarily of the channel protein, connexin43 (Cx43), and provide a substrate for formation of a functional syncytium implicated in the process of spatial buffering in the CNS. Thus gap junctional communication may be neuroprotective following a CNS insult that entails glutamate cytotoxicity (i.e. ischemia). We have shown that blocking gap junctions during a glutamate insult to co-cultures of astrocytes and neurons results in increased neuronal injury. To assess the effect of reduced Cx43 and GJIC on neuroprotection, we examined brain infarct volume in wild type and Cx43 heterozygote null mice following focal ischemia. Cx43 heterozygous null mice exhibited a significantly larger infarct volume compared to wild type. At the cellular level, a significant increase in TUNEL positive cells was observed in the penumbral region of the Cx43 heterozygote mice. These results suggest that augmentation of GJIC in astrocytes may contribute to neuroprotection following ischemic injury. These findings support the hypothesis that gap junctions play a neuroprotective role against glutamate cytotoxicity.     

A neuroprotective role for gap junctions

Glial-neuronal interactions have been implicated in both normal information processing and neuroprotection. One pathway of cellular interactions involves gap junctional intercellular communication (GJIC). In astrocytes, gap junctions are composed primarily of the channel protein, connexin43 (Cx43), and provide a substrate for formation of a functional syncytium implicated in the process of spatial buffering in the CNS. Thus gap junctional communication may be neuroprotective following a CNS insult that entails glutamate cytotoxicity (i.e. ischemia). We have shown that blocking gap junctions during a glutamate insult to co-cultures of astrocytes and neurons results in increased neuronal injury. To assess the effect of reduced Cx43 and GJIC on neuroprotection, we examined brain infarct volume in wild type and Cx43 heterozygote null mice following focal ischemia. Cx43 heterozygous null mice exhibited a significantly larger infarct volume compared to wild type. At the cellular level, a significant increase in TUNEL positive cells was observed in the penumbral region of the Cx43 heterozygote mice. These results suggest that augmentation of GJIC in astrocytes may contribute to neuroprotection following ischemic injury. These findings support the hypothesis that gap junctions play a neuroprotective role against glutamate cytotoxicity.     

Deletion of a single allele of Cx43 is associated with a reduction in the gap junctional intercellular communication and increased cell proliferation of mouse lung pneumocytes type II

OBJECTIVES: Connexins (Cx) are proteins that form the gap junctional channels at neighbouring plasma membranes between adjacent cells. Cxs are involved in cell communication, which is reportedly correlated with cell proliferation and differentiation. Alterations in connexin expression and/or gap junctional intercellular communication (GJIC) capacity have long been postulated to be important in a number of pathological conditions including cancer. This study was performed to determine the consequences of the deletion of a single allele of Gja1 (Cx43 gene) in Alveolar Type II cells (APTIIs), and its impact on GJIC and cell proliferation. MATERIAL AND METHODS: In order to do so, APTIIs from wild type (Cx43(+/+)) and heterozygous (Cx43(+/-)) mice were harvested and cultured for 4 days. The GJIC capacity was evaluated by scrape-loading method, with the transfer of lucifer yellow dye. The expression of Cx43 was evaluated by immunofluorescence method and Western blotting. Cell proliferation was evaluated by 3-(4,5-dimethylthazol-2-yl)-2,5-diphenyltetrazolium bromide assay. Results: It was observed that GJIC capacity was significantly reduced and cell proliferation index was significantly higher in Cx43(+/-) cells compared to Cx43(+/+) cells. CONCLUSIONS: These results show that knocking out one allele of Cx43 leads to a lower cell to cell communication capacity, and consequently induces a higher cell proliferation. Because chemically induced lung adenomas in mice are known to originate from APTIIs, these alterations may play a critical role in their susceptibility to lung carcinogenesis.     

ATP-sensitive potassium channels may participate in the coupling of neuronal activity and cerebrovascular tone

K(+) dilate and constrict cerebral vessels in a dose-dependent fashion. Modest elevations of abluminal K(+) cause vasodilatation, whereas larger extracellular K(+) concentration ([K(+)](out)) changes decrease cerebral blood flow. These dilations are believed to be mediated by opening of inward-rectifier potassium channels sensitive to Ba(2+). Because BaCl(2) also blocks ATP-sensitive K(+) channels (K(ATP)), we challenged K(+) dilations in penetrating, resistance-size (<60 mmu) rat neocortical vessels with the K(ATP) channel blocker glibenclamide (1 microM). Glibenclamide reduced K(+) responses from 138 +/- 8 to 110 +/- 0.8%. K(+) constrictions were not affected by glibenclamide. The Na(+)-K(+)-pump inhibitor ouabain (200 microM) did not significantly change resting vessel diameter but decreased K(+) dilations (from 153 +/- 9 to 99 +/- 2%). BaCl(2) blocked K(+) dilations with a half-maximal dissociation constant of 2.9 microM and reduced dilations to the specific K(ATP) agonist pinacidil with equal potency. We conclude that, in resistance vessels, K(+) dilations are mediated by K(ATP); we hypothesize that [K(+)](out) causes activation of Na(+)-K(+) pumps, depletion of intracellular ATP concentration, and subsequent opening of K(ATP). This latter hypothesis is supported by the blocking effect of ouabain.     

Role of connexin (gap junction) genes in cell growth control and carcinogenesis

Gap junctional intercellular communication (GJIC) is considered to play a key role in the maintenance of tissue independence and homeostasis in multicellular organisms by controlling the growth of GJIC-connected cells. Gap junction channels are composed of connexin molecules and, so far, more than a dozen different connexin genes have been shown to be expressed in mammals. Reflecting the importance of GJIC in various physiological functions, deletion of different connexin genes from mice results in various disorders, including cancers, heart malformation or conduction abnormality, cataract, etc. The possible involvement of aberrant GJIC in abnormal cell growth and carcinogenesis has long been postulated and recent studies in our own and other laboratories have confirmed that expression and function of connexin genes play an important role in cell growth control. Thus, almost all malignant cells show altered homologous and/or heterologous GJIC and are often associated with aberrant expression or localization of connexins. Aberrant localization of connexins in some tumour cells is associated with lack of function of cell adhesion molecules, suggesting the importance of cell-cell recognition for GJIC. Transfection of connexin genes into tumorigenic cells restores normal cell growth, supporting the idea that connexins form a family of tumour-suppressor genes. Some studies also show that specific connexins may be necessary to control growth of specific cell types. We have produced various dominant-negative mutants of Cx26, Cx32 and Cx43 and showed that some of them prevent the growth control exerted by the corresponding wild-type genes. However, we have found that connexins 32, 37 and 43 genes are rarely mutated in tumours. In some of these studies, we noted that connexin expression per se, rather than GJIC level, is more closely related to growth control, suggesting that connexins may have a GJIC-independent function. We have recently created a transgenic mouse strain in which a mutant Cx32 is specifically overexpressed in the liver. Studies with such mice indicate that Cx32 plays a key role in liver regeneration after partial hepatectomy. A decade ago, we proposed a method to enhance killing of cancer cells by diffusion of therapeutic agents through GJIC. Recently, we and others have shown that GJIC is responsible for the bystander effect seen in HSV-tk/ganciclovir gene therapy. Thus, connexin genes can exert dual effects in tumour control: tumour suppression and a bystander effect for cancer therapy.     

In eNOS knockout mice skeletal muscle arteriolar dilation to acetylcholine is mediated by EDHF

The mechanisms that account for acetylcholine (ACh)-induced responses of skeletal muscle arterioles of mice lacking endothelial nitric oxide (NO) synthase (eNOS-KO) were investigated. Isolated, cannulated, and pressurized arterioles of gracilis muscle from male eNOS-KO (74.1 +/- 2.3 microm) and wild-type (WT, 87.2 +/- 2.1 microm) mice developed spontaneous tone accounting for 63 and 61% of their passive diameter (116.8 +/- 3.4 vs. 143.2 +/- 2.8 microm, respectively) and dilated dose-dependently to ACh (10(-9)-10(-7) M). These dilations were significantly smaller in vessels of eNOS-KO compared with WT mice (29.2 +/- 2.0 microm vs. 46.3 +/- 2.1 microm, at maximum concentration) but responses to the NO donor, sodium nitrite (NaNO(2), 10(-6)-3 x 10(-5) M), were comparable in the vessels of the two strains. N(G)-nitro-L-arginine (L-NNA, 10(-4) M), an inhibitor of eNOS, inhibited ACh-induced dilations by 60-90% in arterioles of WT mice but did not affect responses in those of eNOS-KO mice. In arterioles of eNOS-KO mice, dilations to ACh were not affected by indomethacin but were essentially abolished by inhibitors of cytochrome P-450, clotrimazole (CTZ, 2 x 10(-6) M) or miconazole (MCZ, 2 x 10(-6) M), as well as by either high K(+) (40 mM) or iberiotoxin [10(-7) M, a blocker of Ca(2+)-dependent K(+) channels (K(Ca) channels)]. On the other hand, in WT arterioles CTZ or MCZ inhibited ACh-induced dilations only by approximately 10% and only in the presence of L-NNA. These results indicate that in arterioles of eNOS-KO mice, endothelium-derived hyperpolarizing factor (EDHF), synthesized via cytochrome P-450, accounts entirely for the mediation of ACh-induced dilation via an increase in K(Ca)-channel activity. In contrast, in arterioles of WT mice, endothelium-derived NO predominantly mediates ACh-induced dilation in which participation of EDHF becomes apparent only after inhibition of NO synthesis.     

Involvement of gap junctions in placental functions and development

Connexin (Cx) expression and gap junctional intercellular communication (GJIC) are involved in development and differentiation processes. Mediating exchanges between mother and fetus, the placenta is formed when fetal membranes are apposed or even fusing or destroying the uterine mucosa. Therefore, an extraordinary variability of placental structures is observed throughout the mammalian species. This variability affect mainly, the maternofetal blood flow interrelationships, the kind and number of tissue layers separating maternal and fetal bloods, the trophoblast invasiveness and the formation of a syncytium (syncytiotrophoblast). Here, the expression, the localisation and the possible role of Cx and GJIC in placental functions and development are discussed. In rodents, gene knock out in mice have vastly improved our understanding of the role of Cx genes in mouse placental development: Cx26 in transplacental uptake of glucose, Cx31 in the proliferative process of trophoblastic cells and Cx45 in placental vascularisation. In human, it appears that Cx43 allows a GJIC required for the fusion process of cytotrophoblastic cells leading to the formation of the syncytiotrophoblast, the site of the numerous placental functions. On other hands, Cx40 plays a critical role in the switch from a proliferative to an invasive phenotype of the trophoblastic cells invading the endometrium. Owing to the striking diversity of Cx expression in placental structures, we must be careful when extrapolating findings from one species to another.     

Nitric oxide specifically reduces the permeability of Cx37-containing gap junctions to small molecules

Gap junction intercellular communication (GJIC) plays a significant role in the vascular system. Regulation of GJIC is a dynamic process, with alterations in connexin (Cx) protein expression and post-translational modification as contributing mechanisms. We hypothesized that the endothelial autacoid nitric oxide (NO) would reduce dye coupling in human umbilical vein endothelial cells (HUVECs). In our subsequent experiments, we sought to isolate the specific Cx isoform(s) targeted by NO or NO-activated signaling pathways. Since HUVEC cells variably express three Cx (Cx37, Cx40, and Cx43), this latter aim required the use of transfected HeLa cells (HeLaCx37, HeLaCx43), which do not express Cx proteins in their wild type form. Dye coupling was measured by injecting fluorescent dye (e.g., Alexa Fluor 488) into a single cell and determining the number of stained adjacent cells. Application of the NO donor SNAP (2 microM, 20 min) reduced dye coupling in HUVEC by 30%. In HeLa cells, SNAP did not reduce dye transfer of cells expressing Cx43, but decreased the dye transfer from Cx37-expressing cells to Cx43-expressing cells by 76%. The effect of SNAP on dye coupling was not mediated via cGMP. In contrast to its effect on dye coupling, SNAP had no effect on electrical coupling, measured by a double patch clamp in whole cell mode. Our results demonstrate that NO inhibits the intercellular transfer of small molecules by a specific influence on Cx37, suggesting a potential role of NO in controlling certain aspects of vascular GJIC.     

Pathological Significance of Intracytoplasmic Connexin Proteins: Implication in Tumor Progression

A considerable amount of evidence has established that gap junctional intercellular communication (GJIC) suppresses tumor development by halting the stage of tumor promotion. Consistently, GJIC is downregulated in tumors. The downregulation of GJIC is caused by not only the reduced expression level of connexin proteins but also their aberrant cytoplasmic localization. Although it has long been thought that cytoplasmic localization of connexin proteins is merely one of the mechanisms of the downregulation of GJIC, careful studies with human tumor samples have indicated that the expression level of intracytoplasmic connexin proteins correlates well with the grade of malignancy and the progression stage of tumors. Hypothesizing that intracytoplasmic connexin proteins should have their proper functions and that their increase should facilitate tumor progression such as cell migration, invasion and metastasis, we examined the effects of overexpressed connexin32 (Cx32) protein on the phenotype of human HuH7 hepatoma cells, which express a basal level of endogenous Cx32 only in cytoplasm. The cells were retrovirally transduced with the Tet-off Cx32 construct so that withdrawal of doxycycline from the culture medium could induce overexpression of Cx32 protein in cytoplasm. Even when overexpressed, Cx32 protein was retained in cytoplasm, i.e., Golgi apparatuses, and did not induce GJIC. However, overexpression of Cx32 protein in cytoplasm enhanced both the motility and the invasiveness of HuH7 cells and induced metastasis when the cells were xenografted into SCID mice. Taken together, cytoplasmic accumulation of connexin proteins may exert effects favorable for tumor progression.     

Action potential modulation of connexin40 gap junctional conductance

Connexin40 (Cx40) is abundantly expressed in the atrial myocardium, ventricular conduction system, and vascular endothelial and smooth muscle cells of the mammalian cardiovascular system. Rapid conduction through cardiac tissues depends on electrotonic transfer of the action potential between neighboring cells. To determine whether transjunctional voltages (Vj) elicited by an action potential can modulate conductance of Cx40 gap junctions, simulated myocardial action potentials were applied as voltage-clamp waveforms to Cx40 gap junctions expressed in mouse neuro2A (N2A) cells. Junctional currents resembled the action potential morphology but declined by >50% from peak to near-constant plateau values. Kinetics of Cx40 voltage gating were examined at peak voltages > or =100 mV, and decay time constants changed e-fold per 17.6 mV for Vj > +/-40 mV. Junctional conductance recovered during phase 3 repolarization and early diastole to initial values. These phasic changes in junctional conductance were due to rapid decay kinetics, increasing to tens of milliseconds at peak Vj of 130 mV, and the increase in the steady-state conductance curve as Vj returned toward 0 mV. Time-dependent conductance curves for Cx40 were modeled with one inactivation and two recovery Vj-dependent components. There was a temporal correlation between development of conduction delay or block and the inactivation phase of junctional conductance. Likewise, recovery of junctional conductance was coincident with recovery from refractoriness, suggesting that gap junctions may play a role in the genesis and propagation of cardiac arrhythmias.     

Endothelin-1 decreases gap junctional intercellular communication by inducing phosphorylation of connexin 43 in human ovarian carcinoma cells

Endothelin-1 (ET-1) is overexpressed in ovarian carcinoma and acts as an autocrine factor selectively through the ETA receptor (ETAR) to promote tumor cell proliferation, survival, neovascularization, and invasiveness. Loss of gap junctional intercellular communication (GJIC) is critical for tumor progression by allowing the cells to escape growth control. Exposure of HEY and OVCA 433 ovarian carcinoma cell lines to ET-1 led to a 50-75% inhibition in intercellular communication and to a decrease in the connexin 43 (Cx43)-based gap junction plaques. To investigate the phosphorylation state of Cx43, ovarian carcinoma cell lysates were immunoprecipitated and transient tyrosine phosphorylation of Cx43 was detected in ET-1-treated cells. BQ 123, a selective ETAR antagonist, blocked the ET-1-induced Cx43 phosphorylation and cellular uncoupling. Gap junction closure was prevented by tyrphostin 25 and by the selective c-Src inhibitor, PP2. Furthermore, the increased Cx43 tyrosine phosphorylation was correlated with ET-1-induced increase of c-Src activity, and PP2 suppressed the ET-1-induced Cx43 tyrosine phosphorylation, indicating that inhibition of Cx43-based GJIC is mainly mediated by the Src tyrosine kinase pathway. In vivo, the inhibition of human ovarian tumor growth in nude mice induced by the potent ETAR antagonist, ABT-627, was associated with a reduction of Cx43 phosphorylation. These findings indicate that the signaling mechanisms involved in GJIC disruption on ovarian carcinoma cells depend on ETAR activation, which leads to the Cx43 tyrosine phosphorylation mediated by c-Src, suggesting that ETAR blockade may contribute to the control of ovarian carcinoma growth and progression also by preventing the loss of GJIC.     

EDHF,but not NO or prostaglandins,is critical to evoke a conducted dilation upon ACh in hamster arterioles

Vasomotor reactions upon focal stimulation of arterioles have been shown to be conducted along the vascular wall. Such a conduction, which is assumed to reflect the spread of electrical signals, may contribute to coordination of responses within a vascular segment. We aimed to identify which endothelial autacoid(s) act as mediators of the local and conducted dilator responses, respectively. To this end, arterioles in the hamster cremaster microcirculation were locally stimulated with endothelium-dependent [acetylcholine (ACh)] or endothelium-independent dilators [sodium nitroprusside (SNP)], and the resulting changes in diameter were measured using a videomicroscopy technique at the site of application and up to 1.4 mm upstream at distant sites. Experiments were also performed after blockade of nitric oxide (NO) synthase, cyclooxygenase, P-450 monooxygenase, or K(+) channels. Dilations upon ACh (71 +/- 3%) were conducted rapidly (<1 s) to upstream sites (at 1.4 mm: 37 +/- 5%). Although the NO donor SNP induced a similar local dilation (71 +/- 7%), this response was not conducted. Maximal amplitudes of ACh-induced dilations were not attenuated after inhibition of NO synthase and cyclooxygenase at the local and remote sites. However, additional treatment with a P-450 monooxygenase blocker (sulfaphenazole) strongly attenuated the local response (from 62 +/- 9 to 17 +/- 5%) and abrogated dilations at distant sites (at 0.67 mm: from 23 +/- 4% to 4 +/- 3%). Likewise, 17-octadecynoic acid strongly attenuated local and remote responses. Blockers of Ca(2+)-dependent K(+) channels (charybdotoxin or iberiotoxin) attenuated dilations at the local and remote sites after focal application at the ACh stimulation site. In marked contrast, treatment of the upstream site with these blockers was without any effect. We conclude that upon local stimulation with ACh, a cytochrome P-450 monooxygenase product is generated that induces local dilation via the activation of Ca(2+)-dependent K(+) channels and initiates conduction of the dilation. In contrast to the local site, neither activation of these K(+) channels nor the synthesis of NO or prostaglandins is necessary to dilate the arterioles at remote, distant sites. This suggests that endothelium-derived hyperpolarizing factor serves as an important mediator to initiate conducted dilations and, by doing so, may act as a key player in the coordination of arteriolar behavior in the microcirculatory network.     

Inhibition of gap junctional intercellular communication in heptachlor- and heptachlor epoxide-treated normal human breast epithelial cells

Based on the concern of organochlorides in the environment and in human tissue, this study was designed to determine whether various noncytotoxic levels of heptachlor and heptachlor epoxide could inhibit, reversibly, gap junctional intercellular communication in human breast epithelial cells (HBEC). Cytotoxicity and gap junctional intercellular communication (GJIC) were evaluated by lactate dehydrogenase assay and fluorescence redistribution after photobleaching analysis, respectively. Both heptachlor and heptachlor epoxide were noncytotoxic up to 10 μg/ml. At this concentration, heptachlor and heptachlor epoxide inhibited GJIC of normal human breast epithelial cells after 1 h treatment. Within a 24 h treatment with heptachlor and heptachlor epoxide at 10 μg/ml, recovery of GJIC had not returned. GJIC completely recovered after a 12 h treatment of 1 μg/ml heptachlor epoxide, but it did not recover after a 24 h treatment of 1 μg/ml heptachlor. RT-PCR and Western blots were analyzed to determine whether the heptachlor or heptachlor epoxide might have altered the steady-state levels of gap junction mRNA and/or connexin protein levels or phosphorylation state. No significant difference in the level of connexin 43 (Cx43) message between control and heptachlor-treated cells was observed. Western blot analyses showed hypophosphorylation patterns in cells treated with 10 μg/ml heptachlor and heptachlor epoxide for 1 h with no recovery within 24 h. Immunostaining of Cx43 protein in normal HBEC indicated that heptachlor and heptachlor epoxide caused a loss of Cx43 from the cell membranes at noncytotoxic dose levels. Taken together, these results suggest that heptachlor and heptachlor epoxide can alter GJIC at the post-translational level, and that, under the conditions of exceeding a threshold concentration in the breast tissue containing ‘initiated’ cells for a long time and not being counteracted by anti-tumor-promoting chemicals, they could act as breast tumor promoters.     

A novel function of connexin 32: marked enhancement of liver function in a hepatoma cell line

Connexin 32 (Cx32) is the main gap junction protein in hepatocytes and plays an important role in the regulation of signal transfer and growth control in the liver by constructing gap junction channels and gap junctional intercellular communication (GJIC). In this study, the human Cx32 gene was transfected into a hepatoma cell line (HepG2) that showed aberrant expression of Cx32 and was deficient in GJIC. Cx32-transfected HepG2 not only expressed a higher level of Cx32 mRNA, but also showed increased GJIC compared with HepG2 and vector-transfected HepG2. Furthermore, the liver functions of ammonia removal and albumin secretion of HepG2 were markedly enhanced with Cx32 gene transfection. It may be expected to improve the cellular functions of the hepatoma cell line by Cx32 gene transfection and serve to develop an efficacious bioartificial liver.