http://bloodjournal.hematologylibrary.org/content/101/3/907.full.html Updated information and services can be found at: (1930 articles) Signal Transduction • (2497 articles) Hemostasis, Thrombosis, and Vascular Biology • Articles on similar topics can be found in the following Blood collections http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#repub_requests Information about reproducing this article in parts or in its entirety may be found online at: http://bloodjournal.hematologylibrary.org/site/misc/rights.xhtml#reprints Information about ordering reprints may be found online at: http://bloodjournal.hematologylibrary.org/site/subscriptions/index.xhtml Information about subscriptions and ASH membership may be found online at:

It was the aim of this study to elucidate whether intraportally transplanted pancreatic islets were reinnervated after transplantation and whether the secretion of insulin from pancreatic islets might be modulated by the vegetative innervation of recipient livers. Two weeks after intraportal transplantation of 2000 neonatal pancreatic islets recipient rats completely recovered from streptozotocin-induced diabetes. Predominantly catecholaminergic, but also cholinergic nerve fibers were detected in islet cell complexes between beta-cells. Corresponding electron micrographs showed beta-cells in close contact with axons of nonmyelinated nerve fibers. Perfusion studies with livers of recipient rats revealed that the inhibition by hepatic sympathetic nerves of insulin secretion was mediated via alpha(2)-receptors as in normal pancreatic islets.

In contrast to conventionally used immunoassays, thin layer chromatography (TLC)—by prelabeling of cells with radioactive arachidonic acid (AA)—allows to differentiate between cellularly built and added prostanoids and thus to investigate feedback effects of prostanoids on their own release. PGD2, TXB2 and PGE2 released from zymosan-stimulated Kupffer cells were separated with distinct RF-values, corresponding to those of the pure substances. Quantification of PGD2 and PGE2 gave comparable results with TLC and immunoassays, but measurement in the presence of added prostanoids was only possible with TLC. Moreover TLC was superior to immunoassays in having a longer linear range while being comparably sensitive. Cellularly built TXB2 in its radioactively labeled form was not detectable by TLC. Inhibition of TXB2 release by externally added AA or technical artifacts were excluded, suggesting that the cellular AA-pools used for prostaglandin and thromboxane synthesis differ in their accessibility for added AA. Thus, TLC is a simple, sensitive and precise method for the quantification of cellularly built prostaglandins but not of thromboxane even in the presence of added prostanoids.

It has been proposed that hydroxyl radicals (·OH) generated in a perinuclear iron-dependent Fenton reaction are involved in O2-dependent gene expression. Thus, it was the aim of this study to localize the cellular compartment in which the Fenton reaction takes place and to determine whether scavenging of ·OH can modulate hypoxia-inducible factor 1 (HIF-1)-dependent gene expression. The Fenton reaction was localized by using the nonfluorescent dihydrorhodamine (DHR) 123 that is irreversibly oxidized to fluorescent rhodamine 123 while scavenging ·OH together with gene constructs allowing fluorescent labeling of mitochondria, endoplasmic reticulum (ER), Golgi apparatus, peroxisomes, or lysosomes. A 3D two-photon confocal laser scanning microscopy showed ·OH generation in distinct hot spots of perinuclear ER pockets. This ER-based Fenton reaction was strictly pO2-dependent. Further colocalization experiments showed that the O2-sensitive transcription factor HIF-1α was present at the ER under normoxia, whereas HIF-1α was present only in the nucleus under hypoxia. Inhibition of the Fenton reaction by the ·OH scavenger DHR attenuated HIF-prolyl hydroxylase activity and interaction with von Hippel–Lindau protein, leading to enhanced HIF-1α levels, HIF-1α transactivation, and activated expression of the HIF-1 target genes plasminogen activator inhibitor 1 and heme oxygenase 1. Further, ·OH scavenging appeared to enhance redox factor 1 (Ref-1) binding and, thus, recruitment of p300 to the transactivation domain C because mutation of the Ref-1 binding site cysteine 800 abolished DHR-induced transactivation. Thus, the localized Fenton reaction appears to impact the expression of hypoxia-regulated genes by means of HIF-1α stabilization and coactivator recruitment.

Glucokinase (GK) is the key enzyme of glucose utilization in liver and is localized in the less aerobic perivenous area. Until now, the O-2-responsive elements in the liverspecific GK promoter are unknown, and therefore the aim of this study was to identify the O-2-responsive element in this promoter.We found that the GK promoter sequence -87/-80 matched the binding site for hypoxia inducible factor 1 (HIF-1) and upstream stimulatory factor (USF). In primary rat hepatocytes we could show that venous pO(2) enhanced HIF-1alpha and USF-2a levels, both of which activated GK expression. Furthermore, transfection experiments revealed that the GK sequence -87/-80 mediated the HIF-1alpha or USF-2-dependent activation of the GK promoter. The binding of HIF-1 and USF to the GKHRE was corroborated by electrophoretic mobility shift assay (EMSA). However, the maximal response to HIF-1alpha or USF was only achieved when constructs with the -87/ -80 sequence in context with a 39-36 bp native GK promoter sequence containing a hepatocyte nuclear factor 4 (HNF-4) binding site were used. HIF-1alpha and HNF-4 additively activated the GK promoter, while USF-2 and HNF-4 together did not show this additive activation. Thus, HIF-1 and USF may play differential roles in the modulation of GK expression in response to O-2.

Einleitung. Lipopolysaccharide (LPS) gehören zu den wichtigsten Entzündungsauslösern im Säugerorganismus. Eigene Vorarbeiten haben gezeigt, dass Hepatozyten der Maus den LPS-Rezeptor-Komplex aus membranständigem CD14 (mCD14) und Toll-like-Rezeptor 4 (TLR4) nicht exprimieren und dementsprechend auch nicht direkt durch LPS stimulierbar sind. Durch IL-6 oder IL-1β sind in Ratten-Hepatozyten Rezeptoren für Anaphylatoxin C5a als Endprodukt der durch LPS ausgelösten Komplementaktivierung de novo induzierbar. Es sollte daher untersucht werden, ob auch mCD14/TLR4 in Hepatozyten induzierbar ist und ob dieser Rezeptor-Komplex funktionell aktiv ist. Methodik. Hepatozyten wurden aus Mäuselebern isoliert, mit IL-6, IL-1β oder TNFα stimuliert und die Expression von CD14-und TLR4-mRNA durch quantitative Real-Time-PCR, von CD14- und TLR4-Protein durch Western-Blots von Triton X114-Lysaten analysiert. Zur Überprüfung der Funktionalität der de novo exprimierten Rezeptoren wurden mit IL-1β vorstimulierte Hepatozyten mit LPS inkubiert und die α2-Macroglobulin-Synthese mittels Northern- und Western-Blot analysiert. Ergebnisse. Hepatozyten exprimierten konstitutiv geringe Mengen an CD14-mRNA aber kein mCD14-Protein und keine TLR4-mRNA. Die Stimulation mit proinflammatorischen Cyto? kinen verstärkte die Expression von CD14-mRNA und induzierte die Expression von TLR-4-mRNA und von mCD14-Protein in der Membran. Alle 3 Parameter wurden durch IL-1β am stärksten, durch IL-6 etwas weniger und durch TNFα am wenigsten stark beeinflusst. In mit IL-1β vorstimulierten Hepatozyten induzierte LPS – anders als in unbehandelten Hepatozyten – die mRNA- und Protein-Expression des Typ-2-Akut-Phase-Proteins α2-Macroglobulin. Schlussfolgerung: Durch Stimulation mit proinflammatorischen Cytokinen wird die Expression des LPS-Rezeptor-Komplexes aus mCD14 und TLR4 in Hepatozyten der Maus induziert. Durch diese Behandlung erlangen die Zellen eine direkte Reaktivität für LPS.

Various inflammatory stimuli such as anaphylatoxin C5a, zymosan, and lipopolysaccharides (LPSs) have been reported both to enhance glucose output in the perfused rat liver and to induce prostanoid (ie, prostaglandin and thromboxane) release from Kupffer cells, the resident liver macrophages. Because prostanoids can enhance glucose output from hepatocytes, it was the aim of this study to compare the possible roles of prostanoids released after C5a, zymosan, and LPS in the mediation of hepatic glucose output. In perfused livers both C5a and zymosan immediately enhanced glucose output, reduced flow, and induced prostanoid overflow into the hepatic vein, but with different quantities and kinetics. Only the C5a-induced but not the zymosan-induced effects were abrogated by inhibitors of prostanoid signaling as the prostanoid synthesis inhibitor indomethacin and the thromboxane receptor antagonist daltroban. In contrast to C5a and zymosan, LPS had no effect on glucose output, flow rate, or prostanoid overflow. In isolated Kupffer cells, C5a and zymosan induced maximal release of prostaglandins D 2 and E 2 and of thromboxane A 2 within a period of 0 to 2 minutes and 5 to 15 minutes, respectively. In pulse-chase experiments, maximal prostanoid release was already observed after 2 minutes of continuous stimulation with C5a, but only after 10 to 15 minutes of continuous stimulation with zymosan. LPS-dependent prostanoid release was not seen before 1 hour. Thus, even though C5a, zymosan, and LPS induced prostanoid release from Kupffer cells, only C5a quickly regulated hepatic glucose metabolism in a prostanoid-dependent manner (due to the kinetics and quantities of prostanoids released).

Mitogen-activated protein kinases (MAPKs) and protein kinase B (PKB) mediate growth and stress signals and have been implicated in the hypoxic response. Under hypoxic conditions, the expression of plasminogen activator inhibitor-1 (PAI-1) is mainly controlled by the hypoxia-inducible factor HIF-1. However, the role of MAPKs and PKB in HIF-1-mediated PAI-1 regulation is not clear.Treatment with the p38 inhibitor SB203580 and the PI3K inhibitor LY294002, but not with the MEK1 inhibitor PD98059, abrogated hypoxia-dependent PAI-1 induction in HepG2 cells. Consistently, overexpression of PKB or of the p38 upstream kinases MKK6 and MKK3 and of JNK, but not of ERK, enhanced PAI-1 mRNA levels. In MKK3-, MKK6- and PKB-expressing cells luciferase (Luc) activities from a hypoxia-inducible PAI-1-Luc construct or from a HIF-dependent Luc construct and, concomitantly, HIF-1alpha protein levels were enhanced. These findings indicate that p38- and PKB-dependent signalling pathways contribute to enhanced PAI-1 levels in the hypoxic response.

Prostanoids, i.e. prostaglandins and thromboxane, regulate liver-specific functions both in homeostasis and during defense reactions. For example, prostanoids are released from Kupffer cells, the resident liver macrophages, in response to the inflammatory mediator anaphylatoxin C5a, and mediate an enhanced glucose output from hepatocytes as energy supply. In perfused rat livers, the thromboxane receptor antagonist daltroban enhanced C5a-induced prostanoid overflow and reduced glucose output. It was the aim of this study to elucidate whether daltroban interfered with prostanoid release from Kupffer cells or prostanoid clearance by hepatocytes, and/or whether it directly influenced prostanoid-dependent glucose metabolism in these cells. In perfused rat livers, daltroban enhanced prostaglandin (PG)D(2) overflow not only after infusion of C5a (15-fold), but also after PGD(2) (10-fold). Neither daltroban nor another receptor antagonist, ifetroban, or the thromboxane synthase inhibitor furegrelate enhanced prostanoid release from Kupffer cells. In contrast, all inhibitors reduced clearance, i.e. uptake and degradation, of PGD(2) by hepatocytes: within 5 min uptake of 1 nmol/L PGD(2) was reduced from 43+/-5 fmol (controls) to 22+/-6 fmol (daltroban), 24+/-6 fmol (ifetroban) and 21+/-6 fmol (furegrelate). PGD(2) in the medium was reduced to 39+/-7% in the controls, but remained at 93+/-9%, 93+/-11% and 60+/-3% in the presence of the inhibitors. PGD(2)-dependent glucose output in the perfused liver or activation of glycogen phosphorylase in isolated hepatocytes remained unaffected by daltroban. These data clearly demonstrate that the thromboxane-inhibitors reduced PGD(2) clearance by hepatocytes, presumably by inhibition of prostanoid transport into the cells. In contrast, they did not interfere with PGD(2)-dependent glucose metabolism, suggesting an independent mechanism for the inhibition of glucose output from the liver.

The C5a-anaphylatoxin which is generated by limited proteolysis upon activation of the fifth component of complement may be induced by the classical, the alternative or the lectin pathway. C5a has been shown, under normal conditions, to induce the release of prostanoids from Kupffer cells (KC) and hepatic stellate cells (HSC) and thereby indirectly to increase glucose output from hepatocytes (HC). A direct action of C5a on HC would require the expression of the specific C5a receptor (C5aR). In studies using quantitative RT-PCR it was shown that non-stimulated HC lack C5aR, in contrast to KC, HSC and sinusoidal endothelial cells (SEC) all of which contained mRNA for the C5aR in decreasing amounts. FACS analyses, immunohisto- and immunocytochemistry as well as functional analyses confirmed the results of the RT-PCR assays. Under inflammatory situations the C5aR was found to be upregulated in various organs and tissues which included the liver. Interleukin-6 (IL-6) as a main inflammatory mediator in the liver induced a de novo expression of functional C5aR in HC in-vitro and in-vivo. In contrast, LPS failed to induce C5aR directly in cultured HC in-vitro but induced C5aR in HC in vivo and in co-cultures of HC and KC which release IL-6 upon stimulation with LPS. So far, the only known effector function of C5a on HSC was the induction of prostanoid release. In an approach to reveal new functions of C5aR in HSC, the cells responsible for liver fibrosis, it could be shown that C5a upregulated fibronectin-specific mRNA five-fold whereas entactin, collagen IV and the structure protein smooth muscle actin were not affected. In addition, C5a did not upregulate specific mRNA for the profibrotic cytokine TGF-beta1 in either isolated KC or HSC. Thus, C5a alone appears to have only a limited role in the induction of liver fibrosis.

SummaryThe hypoxia-inducible transcription factor HIF-1 mediates upregulation of plasminogen activator inhibitor-1 (PAI-1) expression under hypoxia. Reactive oxygen species (ROS) have also been implicated in PAI-1 gene expression. However, the role of ROS in HIF-1-mediated regulation of PAI-1 is not clear. We therefore investigated the role of the GTPase Rac1 which modulates ROS production in the pathway leading to HIF-1 and PAI-1 induction.Overexpression of constitutively activated (RacG12V) or dominant-negative (RacT17N) Rac1 increased or decreased, respectively, ROS production. In RacG12V-expressing cells, PAI-1 mRNA levels as well as HIF-1α nuclear presence were reduced under normoxia and hypoxia whereas expression of RacT17N resulted in opposite effects. Treatment with the antioxidant pyrrolidinedithiocarbamate or coexpression of the redox factor-1 restored HIF-1 and PAI-1 promoter activity in RacG12V-cells. In contrast, NFκB activation was enhanced in RacG12V-cells, but abolished by RacT17N. Thus, these findings suggest a mechanism explaining modified fibrinolysis and tissue remodeling in an oxidized environment.

The expression of the plasminogen activator inhibitor-1 (PAI-1) gene is enhanced by insulin both in vivo and in various cell types. Because insulin exerts a number of its biologic activities via the phosphatidylinositol 3-kinase and protein kinase B (PI3K/PKB) signaling pathway, it was the aim of the present study to investigate the role of the PI3K/PKB pathway in the expression of the PAI-1 gene and to identify the insulin responsive promoter sequences. It was shown that the induction of PAI-1 mRNA and protein expression by insulin and mild hypoxia could be repressed by the PI3K inhibitor wortmannin. Overexpression of a constitutively active PKB led to induction of PAI-1 mRNA expression and of luciferase (Luc) activity from a gene construct containing 766 bp of the rat PAI-1 promoter. Mutation of the hypoxia response elements (HRE-1 and HRE-2) in rat PAI-1 promoter, which could bind hypoxia inducible factor-1 (HIF-1), abolished the induction of PAI-1 by insulin and PKB. Insulin and the constitutive active PKB also induced Luc expression in cells transfected with the pGl3EPO-HRE Luc construct, containing 3 copies of the HRE from the erythropoietin gene in front of the SV40 promoter. Furthermore, insulin and the active PKB enhanced all 3 HIF alpha-subunit protein levels and HIF-1 DNA-binding activity, as shown by electrophoretic mobility shift assays (EMSAs). Thus, the insulin-dependent activation of the PAI-1 gene expression can be mediated via the PI3K/PKB pathway and the transcription factor HIF-1 binding to the HREs in the PAI-1 gene promoter.

Glucokinase (GK) is a key enzyme for glucose utilization in liver and shows a higher expression in the perivenous zone. In primary rat hepatocytes, the GK gene expression was activated by HNF (hepatic nuclear factor)-4alpha via the sequence -52/-39 of the GK promoter. Venous pO2 enhanced HNF-4 levels and HNF-4 binding to the GK-HNF-4 element. Thus, HNF-4alpha could play the role of a regulator for zonated GK expression.

BACKGROUND & AIMS:In normal rat liver, anaphylatoxin C5a induces glucose output from hepatocytes indirectly via prostanoids released from Kupffer cells. Correspondingly, it was found that hepatocytes, in contrast to Kupffer cells, did not express C5a receptors. Lipopolysaccharide (LPS) has been reported to enhance C5a receptor expression in murine livers. This might be the result of de novo expression in hepatocytes.METHODS:C5a receptor expression was investigated in hepatocytes after in vivo treatment of rats with LPS and in vitro stimulation of isolated cells with LPS and proinflammatory cytokines on messenger RNA (mRNA) and protein level, and functionally in isolated hepatocytes and perfused liver.RESULTS:In vivo treatment of rats with LPS induced C5a receptor mRNA and protein in hepatocytes with a maximum after 8-10 hours. At this time-point, C5a directly activated glycogen phosphorylase in isolated hepatocytes and enhanced glucose output in perfused livers without the involvement of prostanoids. LPS failed to induce C5a receptors in cultured hepatocytes in vitro, whereas interleukin (IL) 6 and IL-1beta, which are known to be released from Kupffer cells on stimulation with LPS, did so. In cocultures of hepatocytes with Kupffer cells, LPS induced C5a receptors in hepatocytes in an IL-6-dependent manner.CONCLUSIONS:Thus, IL-6 from Kupffer cells appears to be the main mediator of LPS-induced de novo expression of C5a receptors in hepatocytes.

The proinflammatory anaphylatoxin C5a induces the release of prostanoids, ie, prostaglandins (PG) and thromboxane (TX), from the resident liver macrophages (Kupffer cells [KC]). Because KC themselves express prostanoid receptors, prostanoids--besides having paracrine functions--might regulate their own release in an autocrine loop. So far, such a possible feedback regulation has not been investigated systematically, probably because of methodological difficulties to measure newly synthesized prostanoids in the presence of added prostanoids. Here, after prelabeling of phospholipids with [(14)C]arachidonate, cellularly formed [(14)C]prostanoids were determined in the presence of added unlabelled prostanoids by thin layer chromatography. In cultured KC, recombinant rat C5a (rrC5a) rapidly increased PGD(2), PGE(2), and TXA(2) release, which was strongly reduced by PGE(2), but neither by PGD(2) nor by the TXA(2) analog U46619. The inhibitory effect of PGE(2) was mimicked by cAMP, indicating that the G(s)-coupled PGE(2) receptors type 2 or 4 were involved. Zymosan also enhanced prostanoid release from KC, but with slightly slower kinetics; this action was neither inhibited by PGE(2) nor by cAMP. Also in perfused rat livers, rrC5a enhanced prostanoid release from KC as shown by prostanoid overflow and thereby indirectly increased glucose output from hepatocytes. Again, PGE(2), but not PGD(2), inhibited rrC5a-elicited prostanoid overflow. This resulted in a complete inhibition of rrC5a-induced, prostanoid-mediated glucose output. Thus, PGE(2) can inhibit specifically the C5a-induced prostanoid release from KC via a feedback mechanism and thereby limit prostanoid-mediated hepatocellular defense reactions, eg, glucose release.

The substrates oxygen and glucose are important for the appropriate regulation of metabolism, angiogenesis, tumorigenesis and embryonic development. The knowledge about an interaction between these two signals is limited. We demonstrated that the regulation of glucagon receptor, insulin receptor and L-type pyruvate kinase (L-PK) gene expression in liver is dependent upon a cross-talk between oxygen and glucose. The periportal to perivenous drop in O2 tension was proposed to be an endocrine key regulator for the zonated gene expression in liver. In primary rat hepatocyte cultures, the expression of the glucagon receptor and the L-PK mRNA was maximally induced by glucose under arterial pO2 whereas the insulin receptor was maximally induced under perivenous pO2. It was demonstrated for the L-PK gene that the modulation by O2 of the glucose-dependent induction occured at the glucose-responsive element (Glc(PK)RE) in the L-PK gene promoter. The reduction of the glucose-dependent induction of the L-PK gene expression under venous pO2 appeared to be mediated via an interference between hypoxia-inducible factor 1 (HIF-1) and the glucose-responsive transcription factors at the Glc(PK)RE. The glucose response element (GlcRE) also functioned as a hypoxia response element and, vice versa, a hypoxia-responsive element was functioning as a GlcRE. Thus, our findings implicate that the cross-talk between oxygen and glucose might have a fundamental role in the regulation of several physiological and pathophysiological processes.

Complement factor I (FI) is a regulatory serine protease of the complement system which cleaves three peptide bonds in the alpha-chain of C3b and two bonds in the alpha-chain of C4b and thus prevents the assembly of the C3 and C5 convertases. We have investigated the proinflammatory cytokines IL-6, IL-1beta, TNF-alpha and IFN-gamma for their potential role in the regulation of FI expression. Of the investigated cytokines, only IL-6 increased the FI-specific RT-PCR signal in isolated hepatocytes, in the two rat hepatoma-derived cell lines FAO and H4IIE or in HUVECs. Quantitative competitive RT-PCR showed an IL-6 induced upregulation of FI-specific mRNA by about ten-fold. These data are in accord with Northern blot analyses in which the FI-mRNA was upregulated by IL-6 between five- and seven-fold. IL-6, but not IL-1beta, TNF-alpha or IFN-gamma also increased FI-protein levels in cell culture supernatants by about five-fold as determined by a semiquantitative immunoblot using a novel monoclonal antibody specific for rat FI.

The cDNAs of three hypoxia-inducible factor (HIF) α-subunits were cloned from RNA of primary rat hepatocytes by reverse transcriptase PCR. All three cDNAs encoded functionally active proteins, of 825, 874 and 662 amino acids. After transfection they were able to activate luciferase activity of a luciferase gene construct containing three HIF-responsive elements. The mRNAs of the rat HIF α-subunits were expressed predominantly in the perivenous zone of rat liver tissue; the nuclear HIFα proteins, however, did not appear to be zonated. The sequence data reported here have been deposited in the EMBL Nucleotide Sequence Database under the following accession numbers: rHIF1α, Y09507; rHIF2α, AJ277828; rHIF3α, AJ277827.

The signals oxygen and glucose play an important role in metabolism, angiogenesis, tumorigenesis, and embryonic development. Little is known about an interaction of these two signals. We demonstrate here the cross-talk between oxygen and glucose in the regulation of L-type pyruvate kinase (L-PK) gene expression in the liver. In the liver the periportal to perivenous drop in O(2) tension was proposed to be an endocrine key regulator for the zonated gene expression. In primary rat hepatocyte cultures the expression of the L-PK gene on mRNA and on protein level was induced by venous pO(2), whereas its glucose-dependent induction occurred predominantly under arterial pO(2). It was shown by transient transfection of L-PK promoter luciferase and glucose response element (Glc(PK)RE) SV40 promoter luciferase gene constructs that the modulation by O(2) of the glucose-dependent induction occurred at the Glc(PK)RE in the L-PK gene promoter. The reduction of the glucose-dependent induction of the L-PK gene expression under venous pO(2) appeared to be mediated via an interference between hypoxia inducible factor-1 (HIF-1) and upstream stimulating factor at the Glc(PK)RE. The glucose response element also functioned as an hypoxia response element which was confirmed in cotransfection assays with Glc(PK)RE luciferase gene constructs and HIF-1alpha expression vectors. Furthermore, it was found by gel shift and supershift assay that HIF-1alpha and USF-1 or USF-2 could bind to the Glc(PK)RE. Our findings implicate that the cross-talk between oxygen and glucose might have a fundamental role in the regulation of several physiological and pathophysiological processes.

Gap junction channels in the rodent liver are composed of connexin26 (Cx26) and connexin32 (Cx32) proteins. Gap junctional intercellular communication in the mouse liver enhances the effects of hormonal or sympathetic stimulation of glucose release from glycogen stores. To determine whether contraction of bile canaliculi and bile secretion are dependent on the function of gap junction channels, we compared wild-type and connexin32-deficient mice. Confocal laser scanning microscopy of the wild-type mouse liver confirmed the close association of connexin26 and -32 proteins with the zona occludens-1 protein and actin filaments of the bile canaliculi. The decrease of bile flow after electrical stimulation of sympathetic nerves in the perfused liver was attenuated in the Cx32-deficient liver compared with wild-type controls. The amount of secreted bile, however, was similar in wild-type and Cx32-deficient livers. Furthermore, Cx32-deficient mice exhibited dilated bile canaliculi, suggesting that the contraction of bile canaliculi could be impaired in these animals.