Growth of normal human hepatocytes in primary culture: effect of hormones and growth factors on DNA synthesis

Although the ability of hormones and growth factors to stimulate DNA synthesis in rat hepatocytes has been investigated extensively, no such study of human hepatocytes has been reported. Here we describe a series of experiments to identify those factors that regulate human hepatocyte DNA synthesis in vitro and which therefore may play a role in the control of human liver regeneration. Human hepatocytes were isolated from normal liver tissue obtained after graft reduction for transplantation into pediatric recipients. Cells were maintained in culture for up to 5 days, and DNA synthesis was determined. Hydroxyurea reduced [3H]thymidine incorporation into DNA by 95%, indicating replicative DNA synthesis. As previously found with rat hepatocytes, epidermal growth factor and transforming growth factor-alpha stimulated DNA synthesis at low nanomolar concentrations; transforming growth factor-alpha was slightly more potent. Although the overall rate of thymidine incorporation was lower than that for rodent cells, human hepatocytes were sensitive to lower concentrations of these growth factors, and the degree of stimulation was similar. Conversely, transforming growth factor-beta inhibited DNA synthesis at low picomolar levels. By contrast (unlike rat hepatocytes), arginine-vasopressin failed to initiate or potentiate DNA synthesis in human cells. These data indicate that normal human hepatocytes can respond to low concentrations of growth promoters or inhibitors, previously shown to have activity on rat hepatocytes. These factors may then play a role in control of human liver growth. However, important species differences are apparent, highlighting the limitations of extrapolating from animal studies to humans.     

Cell cycle progression proteins (cyclins), oncogene expression,and signal transduction during the proliferative response of human hepatocytes to hepatocyte growth factor

Human hepatocytes stimulated with human recombinant hepatocyte growth factor (h-rHGF) (10 ng/mL) displayed a characteristic lag period before entering into the S phase. The duration of this delay was dependent on the timing of h-rHGF addition to cultures. The highest peak of DNA synthesis was observed at 120 hours of culture when hepatocytes were stimulated with h-rHGF at 72 hours of culture. This was accompanied by an early peak of c-jun and c-fos synthesis (3 hours after addition of h- rHGF) followed by c-myc (6 hours) and increased expression of cyclins A, B, D, and E (12 hours after h-rHGF). A significant dose-dependent increase in inositol 1,4,5-P₃ was observed within 45 seconds after stimulation with the factor. This was followed by an immediate increase in the cytosolic-free calcium. Cyclic adenosine monophosphate (cAMP) levels did not change after stimulation with the factor. Tyrosine phosphorylation seems to be an early event in the course of the stimulatory effect of h-rHGF on DNA synthesis of hepatocytes; genistein, a tyrosine kinase inhibitor, impaired the stimulatory effect of h-rHGF on DNA synthesis dose dependently. On the other hand, the action of the factor was negatively regulated by protein kinase C activation, as shown by the increased stimulatory effect of h-rHGF on DNA synthesis upon inhibition of protein kinase C by H7. (Hepatology 1996 May;23(5):1012-9)     

Intracellular calcium as a second messenger for human hepatocyte growth factor in hepatocytes.

Human hepatocyte growth factor is a newly discovered substance that stimulates DNA synthesis in vitro. In this study, we examined intracellular Ca2+ movement as one of the second messengers for human hepatocyte growth factor in primary-cultured hepatocytes. The addition of hHGF induced Ca2+ oscillation, but the frequency of oscillations varied from cell to cell. We also saw marked intercellular heterogeneity in the initial latent period for the Ca2+ response; the mean latent period was rather longer than those seen with phenylephrine and vasopressin. This difference in the initial latent period may be due to the difference in the pathways of Ca2+ elevation. Duration of culture determined the number of human hepatocyte growth factor-responsive cells; their number peaked at 2 to 5 hours of confluent culture, whereas the peak was earlier in a low-density culture. These changes in responsiveness during culture can be explained by the cell cycle-dependent sensitivity to human hepatocyte growth factor of hepatocytes. The Ca2+ response to human hepatocyte growth factor was dose dependent; 10(-10) mol/L hHGF gave the highest Ca2+ response, similar to the dose-response curve of DNA synthesis. We even observed the Ca2+ response in the Ca(2+)-free buffer, so the increase in Ca2+ was considered due to release from intracellular Ca2+ stores. These results suggest that human hepatocyte growth factor causes the intracellular Ca2+ elevation in the early stage of the cell cycle and that it plays important roles in the signal transduction systems for human hepatocyte growth factor and the proliferation of hepatocytes.     

Effect of multiplication-stimulating activity (MSA) on the cyclic AMP level and proteoglycan synthesis in cultured chondrocytes

Multiplication-stimulating activity (MSA), a somatomedin purified from conditioned medium of Buffalo rat liver cells, had little effect on the intracellular level of cyclic AMP when it markedly enhanced the synthesis of sulphated glycosaminoglycans in rabbit chondrocytes in culture. In addition, MSA did not inhibit prostaglandin E1- or parathyroid hormone-induced accumulation of cyclic AMP in the chondrocytes. On the contrary, MSA slightly decreased stimulation of cyclic AMP accumulation by prostaglandin in human fibroblasts. Dibutyryl cyclic AMP and MSA increased the incorporation of [35S]sulphate and [3H]serine into proteoglycans synthesized by rabbit chondrocytes, and their effects were additive. These findings suggest that somatomedin and dibutyryl cyclic AMP enhance sulphate proteoglycan synthesis through different mechanisms. The lack of inhibitory effect of MSA on cyclic AMP accumulation may be favourable for producing additive effects with cyclic AMP on proteoglycan synthesis and DNA synthesis in chondrocytes.     

Studies on the Mechanism of Action of Nerve Growth Factor

In this study, explanted chick-embryo sensory ganglia were treated with nerve growth factor, and cellular concentrations of neurotubule protein were measured chemically by a colchicine-binding assay. Even after brief time periods, ganglia treated with growth factor were enriched in neurotubule protein, as compared with untreated (control) ganglia. Furthermore, studies with ganglia treated with both vincristine and growth factor demonstrated that neurotubule protein synthesis can occur even though neurite outgrowth is abolished. Several lines of evidence indicate that the growth factor stimulates de novo synthesis of neurotubule subunit protein, and that this effect precedes neurite extension. Like nerve growth factor, dibutyryl cyclic AMP also stimulates neurite outgrowth from embryonic sensory ganglia, yet it does not increase neurotubule protein levels. Available information suggests that the ability of growth factor to elicit rapid neurite outgrowth is closely related to its ability to increase cellular neurotubule levels. Cyclic AMP appears to stimulate neurite outgrowth by a different mechanism.     

The biosynthesis of thrombomodulin and its enhancement by dibutyryl cAMP in a human megakaryoblastic cell line, UT-7.

Thrombomodulin (TM) is an endothelial cell membrane glycoprotein which modulates coagulation via the formation of thrombin-TM complexes. We investigated the human megakaryoblastic cell line (UT-7) for the presence of functional TM on the cell surface and in cell lysates using a specific enzyme-linked immunosorbent assay, a functional assay, and analysis by fluorescent activated cell sorter. We also examined the effect of cyclic nucleotides on TM in UT-7 cells. Quiescent UT-7 cells contained TM protein in cell lysates, but no TM antigen was observed on the cell surface. Dibutyryl cyclic AMP up-regulated TM: UT-7 cells transiently expressed functional TM antigen on the cell surface via de novo synthesis of TM protein resulting from increased TM mRNA levels. In contrast, dibutyryl cyclic GMP did not significantly affect TM antigen levels. The results suggest that megakaryocytes produce TM antigen and protein kinase A are involved in cellular mechanisms of TM expression.     

Alteration of Nucleoside Transport of Chinese Hamster Cells by Dibutyryl Adenosine 3′:5′-Cyclic Monophosphate

Cultured Chinese hamster ovary cells showed no significant change in generation time or fraction in the S-phase in the presence of 1 mM N(6),O(2')-dibutyryl adenosine 3':5'-cyclic monophosphate. Growth continued for at least two generations after expression of the morphological transformation induced by this cyclic AMP analog. Despite identical growth rates, apparent rates of DNA and RNA synthesis (incorporation of [(3)H]-thymidine or [(3)H]uridine) were reduced up to 15-fold in log phase by 1 mM cyclic nucleotide. [(3)H]Deoxycytidine incorporation was much less sensitive to dibutyryl cyclic AMP. Uptake studies with [(3)H]thymidine demonstrated an inhibition of transport rate dependent on the concentration of dibutyryl cyclic AMP in the growth medium. The rate of thymidine uptake at 1 degrees was decreased 21-fold by 1 mM cyclic nucleotide; half-maximal inhibition occurred at 6 muM. At 37 degrees , the pool size of acid-soluble thymidylate was strongly reduced by 1 mM cyclic nucleotide, and synergistic reduction of the pool size was found with 0.5 mM aminophylline. Phosphorylation of the acid-soluble intracellular label was unaffected by dibutyryl cyclic AMP. Inhibition of thymidine uptake is attributed to an observed decrease in thymidine kinase activity caused by growth in 1 mM dibutyryl cyclic AMP, and possibly to a simultaneous alteration in membrane permeability. Kinase-facilitated uptake of other metabolites may be regulated in a similar fashion by cyclic AMP.     

Oxygen dependency of epidermal growth factor receptor binding and DNA synthesis of rat hepatocytes

Background/Aims: Changes in oxygen availability modulate replicative responses in several cell types, but the effects on hepatocyte replication remain unclear. We have studied the effects of transient nonlethal hypoxia on epidermal growth factor receptor binding and epidermal growth factor-induced DNA synthesis of rat hepatocytes.Methods: Lactase dehydrogenase activity in culture supernatant, intracellular adenosine triphosphate content, ²⁵I-epidermal growth factor specific binding, epidermal growth factor receptor protein expression, and ³H-thymidine incorporation were compared between hepatocytes cultured in hypoxia and normoxia.Results: Hypoxia up to 3 caused no significant increase in lactate dehydrogenase activity in the culture supernatant, while intracellular adenosine triphosphate content decreased time-dependently and was restored to normoxic levels by reoxygenation (nonlethal hypoxia). Concomitantly, ¹²⁵I-epidermal growth factor specific binding to hepatocytes decreased time-dependent (to 54.1% of normoxia) and was restored to control levels by reoxygenation, although ¹²⁵I-insulin specific binding was not affected. The decrease in ¹²⁵I-epidermal growth factor specific binding was explained by the decrease in the number of available epidermal growth factor receptors (21.37±3.08 to 12.16±1.42 fmol/10⁵ cells), while the dissociation constant of the receptor was not affected. The change in the number of available receptors was not considered to be due to receptor degradation-resynthesis, since immunodetection of the epidermal growth factor receptor revealed that the receptor protein expression did not change during hypoxia and reoxygenation, and since neither actinomycin D nor cycloheximide affected the recovery of ¹²⁵I-epidermal growth factor binding by reoxygenation. Inhibition of epidermal growth factor-induced DNA synthesis after hypoxia (to 75.4% of normoxia by 3 h hypoxia) paralleled the decrease in ¹²⁵I-epidermal growth factor binding.Conclusions: Transient hypoxia, which caused no increase in lactase dehydrogenase leakage but affected intracellular adenosine triphosphate levels, did, however, modulate the number of available epidermal growth factor receptors without affecting the receptor protein expression, and inhibit the epidermal growth factor-induced DNA synthesis of hepatocytes. This suggests that even transient nonlethal hypoxia affects the epidermal growth factor-induced DNA synthesis of rat hepatocytes through reversible changes in the epidermal growth factor receptor molecule, which depends on oxygen availability.     

Transport of Methotrexate into Normal Haemopoietic Cells and into Leukaemic Cells and its Effects on DNA Synthesis

Summary . Methotrexate uptake by, and action on normal and leukaemic haemopoietic cells has been studied. Peak uptake occurred just prior to peak DNA synthesis, at 48 hr for normal, and at 6–7 days for chronic lymphocytic leukaemic lymphocytes stimulated by phytohaemagglutinin (PHA). Uptake of methotrexate into these cells and into normal bone marrow cells was consistently inhibited by dibutyryl cyclic 3′,5’adenosine monophosphate (dibutyryl cyclic AMP) at concentrations of 10^?3 to 10^?5m and by inhibitors of phosphodiesterase (theophylline, puromycin aminonucleoside and puromycin). Lower external concentrations of dibutyryl cyclic AMP (10^?5 to 10^?6m ), however, slightly increased the uptake of methotrexate. Dibutyryl cyclic AMP and the phosphodiesterase inhibitors at high concentrations also inhibited cellular uptake of 5-methyltetrahydrofolate and of thymidine into DNA. Conversely, none of these compounds inhibited uptake of pteroylglutamic acid (folic acid). These results suggest that the rate of entry of methotrexate and reduced folates (but not of folic acid) into haemopoietic cells may be governed by the intracellular concentration of cyclic AMP and that uptake of methotrexate by cells is related to uptake of natural folates but not to that of folic acid. Methotrexate uptake was also rapidly inhibited by actinomycin D suggesting that it is closely coupled to DNA synthesis. Rate of entry of methotrexate into acute leukaemic blast cells was roughly proportional to their rate of proliferation, measured by ³H-thymidine (³H-TdR) incorporated into DNA. Resistant cells were not found to show a defect in methotrexate transport in relation to their degree of proliferation. Methotrexate was found to inhibit mitosis, but not transformation of PHA-stimulated lymphocytes and this was presumably due to blocking of de novo thymidylate synthesis from deoxyuridylate, as shown by inhibition of incorporation of tritiated deoxyuridine (³H-dU) into DNA. Studies on human leukaemic blast cells showed differences between sensitive and resistant cases in the effect of methotrexate in vitro on ³H-dU and ³H-TdR incorporation into DNA and on dU blocking of ³H-TdR incorporation into DNA but none of these tests gave reliable prediction of resistance or sensitivity to methotrexate.     

Angiogenic growth factors inhibit chondrocyte ageing in osteoarthritis: potential involvement of catabolic stress‐induced overexpression of caveolin‐1 in cellular ageing

Objective: Recently, attention has been attracted by the finding that overexpression of caveolin‐1 induces cellular senescence in age‐related diseases. We aimed to ascertain whether angiogenic growth factors (AGFs) can inhibit interleukin (IL)‐1β‐induced senescence in human chondrocytes by downregulation of caveolin‐1. Methods: We investigated the intracellular signalling pathways involved in chondrocyte ageing. Human chondrocytes were isolated from the articular cartilage of patients undergoing arthroplastic knee surgery in osteoarthritis (OA). Chondrocytes were stimulated with or without IL‐1β (10 ng/mL) in the presence or absence of vascular endothelial growth factor, basic fibroblast growth factor or hepatocyte growth factor (20 ng/mL). After 72‐h incubation, we observed the expression of caveolin‐1 in human chondrocytes by immunohistochemistry, and analysed the protein levels of caveolin‐1 by Western blot. We examined the time‐course of phosphorylation patterns of mammalian mitogen‐activated protein kinase (MAPK) and phosphatidylinositol 3‐kinase (PI3‐K) by Western blot, and used several specific protein kinase inhibitors to evaluate the involvement of the intracellular signalling pathways. Also, chondrocyte replicative lifespan was analyzed in the presence or absence of AGFs. Results: Treatment with AGFs inhibited IL‐1β‐induced overexpression of caveolin‐1 in human OA chondrocytes. Treatment with AGFs all down‐regulated protein levels of IL‐1β‐accelarated expression of caveolin‐1 in chondrocytes. IL‐1β significantly decreased the cellular replicative lifespan in chondrocytes. Treatment with AGFs prevented the IL‐1β‐induced shortening of chondrocyte replicative lifespan. The specific inhibitors for MAPK/extracellular signal‐regulated kinase and PI3‐K cancelled the AGF‐induced downregulation of overexpression of caveolin‐1. Conclusion: Our results suggest that AGFs downregulated IL‐1β‐induced chondrocyte ageing and overexpression of caveolin‐1 in human chondrocytes, which is mediated by kinase cascades involving the p42/44 MAP kinase and PI3‐K/Akt signalling pathways.     

Effects of hepatocyte growth factor on the growth and metabolism of human hepatocytes in primary culture

The effect of recombinant human hepatocyte growth factor (h-rHGF), a potent mitogen for hepatocytes, was investigated in primary cultures of human hepatocytes. Here, we describe a series of experiments to investigate the kinetics of its mitogenic action, as well as its metabolic effects on cultured human hepatocytes. The hrHGF is a potent signal for initiating DNA synthesis in human hepatocytes, with maximal stimulatory effects at 10 ng/mL (0.1 pmol/L). The kinetics of DNA synthesis showed a lag of about 48 to 72 hours, followed by a maximum at 96 hours. At least 48 hours of continuous exposure to h-rHGF are required to initiate DNA synthesis in quiescent human hepatocytes. Cell cycle analysis by flow cytometry showed that most of quiescent 2c cells have left GO/G1 and entered the cell cycle (S and G2/M phases) by 96 hours of continuous exposure to h-rHGF. When compared with other growth factors, h-rHGF was a much more potent mitogen. The effects of 10 ng/mL (0.1 pmol/L) h-rHGF on DNA synthesis were only achieved by 1.5 pmol/L epidermal growth factor (EGF), 0.1 μmol/L insulin, or 1 μmol/L glucagon. It is noteworthy that the effect of h-rHGF was potentiated by glucagon but not by insulin or EGF. The stimulatory effect of HGF on DNA synthesis was gradually inhibited by h-rHGF transforming growth factor beta (TGF-β) in the range 1 to 10 ng/ml. The HGF also influenced the expression of other hepatic genes. This mitogenic factor stimulated the synthesis of the negative acute-phase plasma protein, albumin, and inhibited the synthesis of the positive one, α₁-antichymotrypsin, whereas it moderately influenced the synthesis of fibrinogen. Although glucagon per se had no effects on cell growth, the synergism between glucagon and h-rHGF could be of relevance facilitating the mobilization of glycogen and the stimulation of human hepatocyte growth.     

Phosphoinositide 3-kinase, but not mitogen-activated protein kinase, pathway is involved in hepatocyte growth factor-mediated protection against bile acid-induced apoptosis in cultured rat hepatocytes

We have previously shown that cAMP protects against hydrophobic bile acid-induced apoptosis in cultured rat hepatocytes through pathways dependent on activation of phosphoinositide 3-kinase and inhibition of mitogen activated protein kinase. Hepatocyte growth factor protects epithelial cells against apoptosis and activates both of these kinases in hepatocytes. We studied the effect of hepatocyte growth factor on glycochenodeoxycholate-induced apoptosis to determine whether hepatocyte growth factor protects hepatocytes against bile acid-induced apoptosis and whether the protective effect is mediated via phosphoinositide 3-kinase and/or mitogen-activated protein kinase pathways. Two-hour exposure of cultured rat hepatocytes to glycochenodeoxycholate resulted in apoptosis in 12.5 +/- 0.49% of the cells. Pretreatment with hepatocyte growth factor (50 ng/mL) decreased apoptosis by 50% to 70%. Hepatocyte growth factor cytoprotection was prevented by pretreatment with the phosphoinositide 3-kinase inhibitors, wortmannin (50 nmol/L) or Ly 294002 (40 micromol/L). Hepatocyte growth factor activated phosphoinositide 3-kinase dependent protein kinase B and mitogen-activated protein kinase. Pretreatment of hepatocytes with a mitogen-activated protein kinase inhibitor, U0126 (40 micromol/L) or an inhibitor of pp70(s6) kinase, rapamycin (100 nmol/L), had no effect on the growth factor's anti-apopotic effect. Treatment with hepatocyte growth factor resulted in mitogen-activated protein kinase-dependent phosphorylation of BAD on serine(112). In summary, hepatocyte growth factor protection against bile acid-induced apoptosis occurs via a phosphoinositide 3-kinase pathway and is not dependent on the mitogen-activated protein kinase pathway, phosphorylation of BAD on serine(112), or activation of p70(S6) kinase.     

A novel mechanism for mitogenic signaling via pro-transforming growth factor alpha within hepatocyte nuclei

Transforming growth factor (TGF) alpha, an important mediator of growth stimulation, is known to act via epidermal growth factor receptor (EGF-R) binding in the cell membrane. Here we show by immunohistology, 2-dimensional immunoblotting, and mass spectrometry of nuclear fractions that the pro-protein of wild-type TGF-alpha occurs in hepatocyte nuclei of human, rat, and mouse liver. Several findings show a close association between nuclear pro-TGF-alpha and DNA synthesis. (1) The number of pro-TGF-alpha+ nuclei was low in resting liver and increased dramatically after partial hepatectomy and after application of hepatotoxic chemicals or the primary mitogen cyproterone acetate (CPA); in any case, S phase occurred almost exclusively in pro-TGF-alpha+ nuclei. The same was found in human cirrhotic liver. (2) In primary culture, 7% of hepatocytes synthesized pro-TGF-alpha, which then translocated to the nucleus; 70% of these nuclei subsequently entered DNA replication, whereas only 2% of pro-TGF-alpha- hepatocytes were in S phase. (3) The frequency of hepatocytes coexpressing pro-TGF-alpha and DNA synthesis was increased by the hepatomitogens CPA or prostaglandin E(2) and was decreased by the growth inhibitor TGF-beta1. (4) Treatment with mature TGF-alpha increased DNA synthesis exclusively in pro-TGF-alpha- hepatocytes, which was abrogated by the EGF-R tyrosine kinase inhibitor tyrphostin A25. In conclusion, TGF-alpha gene products may exert mitogenic effects in hepatocytes via 2 different signaling mechanisms: (1) the "classic" pathway of mature TGF-alpha via EGF-R in the membrane and (2) a novel pathway involving the presence of pro-TGF-alpha in the nucleus.     

Oval cells compensate for damage and replicative senescence of mature hepatocytes in mice with fatty liver disease

Hepatic steatosis may have a generally benign prognosis, either because most hepatocytes are not significantly injured or mechanisms to replace damaged hepatocytes are induced. To determine the relative importance of these mechanisms, we compared hepatocyte damage and replication in ethanol-fed and ob/ob mice with very indolent fatty liver disease to that of healthy control mice and PARP-1(-/-) mice with targeted disruption of the DNA repair enzyme, poly(ADP-ribose) polymerase. Compared to the healthy controls, both groups with fatty livers had significantly higher serum alanine aminotransferase values, hepatic mitochondrial H(2)O(2) production, and hepatocyte oxidative DNA damage. A significantly smaller proportion of the hepatocytes from fatty livers entered S phase when cultured with mitogens. Moreover, this replicative senescence was not reversed by treating cultured hepatocytes with agents (i.e., betaine or leptin) that improve liver disease in intact ethanol-fed or leptin-deficient mice. Hepatocytes from PARP1(-/-) mice also had more DNA damage and reduced DNA synthesis in response to mitogens. However, neither mice with fatty livers nor PARP-1-deficient mice had atrophic livers. All of the mice with senescent mature hepatocytes exhibited hepatic accumulation of liver progenitor (oval) cells and oval cell numbers increased with the demand for hepatocyte replacement. Therefore, although hepatic oxidant production and damage are generally increased in fatty livers, expansion of hepatic progenitor cell populations helps to compensate for the increased turnover of damaged mature hepatocytes. In conclusion, these results demonstrate that induction of mechanisms to replace damaged hepatocytes is important for limiting the progression of fatty liver disease.     

Induction of Cytolysis of Cultured Lymphoma Cells by Adenosine 3′:5′-Cyclic Monophosphate and the Isolation of Resistant Variants

Cultured mouse lymphosarcoma cells are killed on exposure to 0.1 mM N(6),O(2')-dibutyryl-adenosine 3':5'-cyclic monophosphate. A population of cells resistant to the killing effect of dibutyryl cyclic AMP at concentrations as high as 1 mM was selected. The growth characteristics of the resistant cells were similar to those of the sensitive parental line. However, the resistant cells contain less cytoplasmic cyclic AMP-binding proteins and decreased cyclic AMP-stimulated protein kinase activity. It is proposed that transition from sensitivity to resistance to dibutyryl cyclic AMP in lymphoma cells is connected with a modification of the cyclic AMP-binding protein, which appears to be the regulatory subunit of the cyclic AMP-activated protein kinase.     

Studies of the Influence of Cyclic Nucleotides on in Vitro Haemoglobin Synthesis

The influence of various cyclic nucleotides on in vitro haemoglobin synthesis has been examined in suspension cultures of mammalian marrow cells. Over a wide range of concentrations, dibutyryl cyclic AMP (db-cAMP) was either ineffective or inhibited haemoglobin synthesis by marrow cells from rat, mouse and guinea-pig. However, 10(-3) M db-cAMP consistently stimulated haemoglobin synthesis in cultures of human, sheep, rabbit and canine cells, with the latter being most responsive. This effect, which approached in magnitude that of erythropoietin (ESF) itself, was specific for cAMP and its mono- and dibutyryl derivatives and was not inhibited by anti-ESF. Adenosine, AMP, ADP, ATP, cGMP, db-cGMP, cCMP, cIMP and sodium butyrate were either inactive or inhibitory at similar concentrations. Enhancement of haemoglobin synthesis was also observed with the phosphodiesterase inhibitor, RO-20-1724. The susceptibility to ionizing radiation of the response to ESF and db-cAMP was marked, indicating that the increased haemoglobin synthesis in this system was proliferation dependent, although the response to db-cAMP was less radiosensitive. Studies with tritiated thymidine showed that about 50% of the cells which were responding to either db-cAMP or ESF were actively engaged in DNA synthesis. However, the physical characteristics of db-cAMP-and ESF-responsive cells were dissimilar as analysed by their velocity sedimentation properties. These studies demonstrate that cAMP has a major stimulatory effect on haemoglobin synthesis with cells from selected mammalian species with activity approaching that of ESF, but the target cells most responsive to these agents appear different. The results suggest that cyclic nucleotide-related mechanisms may modulate in vitro erythropoiesis.     

Glucagon induces the plasma membrane insertion of functional aquaporin-8 water channels in isolated rat hepatocytes

Although glucagon is known to stimulate the cyclic adenosine monophosphate (cAMP)-mediated hepatocyte bile secretion, the precise mechanisms accounting for this choleretic effect are unknown. We recently reported that hepatocytes express the water channel aquaporin-8 (AQP8), which is located primarily in intracellular vesicles, and its relocalization to plasma membranes can be induced with dibutyryl cAMP. In this study, we tested the hypothesis that glucagon induces the trafficking of AQP8 to the hepatocyte plasma membrane and thus increases membrane water permeability. Immunoblotting analysis in subcellular fractions from isolated rat hepatocytes indicated that glucagon caused a significant, dose-dependent increase in the amount of AQP8 in plasma membranes (e.g., 102% with 1 micromol/L glucagon) and a simultaneous decrease in intracellular membranes (e.g., 38% with 1 micromol/L glucagon). Confocal immunofluorescence microscopy in cultured hepatocytes confirmed the glucagon-induced redistribution of AQP8 from intracellular vesicles to plasma membrane. Polarized hepatocyte couplets showed that this redistribution was specifically to the canalicular domain. Glucagon also significantly increased hepatocyte membrane water permeability by about 70%, which was inhibited by the water channel blocker dimethyl sulfoxide (DMSO). The inhibitors of protein kinase A, H-89, and PKI, as well as the microtubule blocker colchicine, prevented the glucagon effect on both AQP8 redistribution to hepatocyte surface and cell membrane water permeability. In conclusion, our data suggest that glucagon induces the protein kinase A and microtubule-dependent translocation of AQP8 water channels to the hepatocyte canalicular plasma membrane, which in turn leads to an increase in membrane water permeability. These findings provide evidence supporting the molecular mechanisms of glucagon-induced hepatocyte bile secretion.     

NO sensitizes rat hepatocytes to proliferation by modifying S-adenosylmethionine levels

BACKGROUND & AIMS: Liver regeneration is a fundamental response of this organ to injury. Hepatocyte proliferation is triggered by growth factors, such as hepatocyte growth factor. However, hepatocytes need to be primed to react to mitogenic signals. It is known that nitrous oxide (NO), generated after partial hepatectomy, plays an important role in hepatocyte growth. Nevertheless, the molecular mechanisms behind this priming event are not completely known. S-adenosylmethionine (AdoMet) synthesis by methionine adenosyltransferase is the first step in methionine metabolism, and NO regulates hepatocyte S-adenosylmethionine levels through specific inhibition of this enzyme. We have studied the modulation of hepatocyte growth factor-induced proliferation by NO through the regulation of S-adenosylmethionine levels. METHODS: Studies were conducted in cultured rat hepatocytes isolated by collagenase perfusion, which triggers NO synthesis. RESULTS: The mitogenic response to hepatocyte growth factor was blunted when inducible NO synthase was inhibited; this process was overcome by the addition of an NO donor. This effect was dependent on methionine concentration in culture medium and intracellular S-adenosylmethionine levels. Accordingly, we found that S-adenosylmethionine inhibits hepatocyte growth factor-induced cyclin D1 and D2 expression, activator protein 1 induction, and hepatocyte proliferation. CONCLUSIONS: Together our findings indicate that NO may switch hepatocytes into a hepatocyte growth factor-responsive state through the down-regulation of S-adenosylmethionine levels.