Hepatic metabolism and gene expression are among the factors controlled by the cellular hydration state, which changes within minutes in response to aniso-osmotic environments, cumulative substrate uptake, oxidative stress and under the influence of hormones such as insulin. The signalling events coupling cell-volume changes to altered cell function were studied in H4IIE rat hepatoma cells. Hypo-osmotic cell swelling resulted within 1 min in a tyrosine kinase-mediated activation of the extracellular signal-regulated protein kinases Erk-1 and Erk-2, which was independent of protein kinase C and cytosolic calcium. Activation of mitogen-activated protein kinases was followed by an increased phosphorylation of c-Jun, which may explain our recently reported finding of an about 5-fold increase in c-jun mRNA level in response to cell swelling. Pretreatment of cells with pertussis or cholera toxin abolished the swelling-induced activation of Erk-1 and Erk-2, suggesting the involvement of G-proteins. Thus, a signal-transduction pathway resembling growth factor signalling is activated already by osmotic water shifts across the plasma membrane, thereby providing a new perspective for adaption of cell function to alterations of the environment.

Freshly isolated rat hepatocytes were plated for 4-6 h and either loaded with (2',7)-bis(carboxyethyl)-5(6)-carboxyfluorescein (BCECF) or allowed to endocytose fluorescein isothiocyanate (FITC)-coupled dextran in order to study the effects of aniso-osmotic exposure and oxidative stress on cytosolic (pHcyt) and apparent vesicular pH (pHves) by single-cell fluorescence recordings. In the presence of normo-osmotic (305 mosmol/l) medium pHcyt was 7.23 +/- 0.03 (n = 108), whereas an apparent pH of 6.07 +/- 0.02 (n = 156) was found in the vesicular compartment accessible to endocytosed FITC-dextran. Substitution of 60 mM NaCl against 120 mM raffinose had no effect on pHcyt or apparent pHves, whereas addition of NH4Cl increased both pHcyt and apparent pHves. Hypo-osmotic cell swelling lowered pHcyt, whereas simultaneously apparent pHves increased. These effects were rapidly reversible upon re-institution of normo-osmotic media. Similarly, an increase of apparent pHves was observed when cell swelling was induced by Ba2+, glutamine or histidine. Conversely, hyperosmotic cell shrinkage due to addition of NaCl or raffinose led to a cytosolic alkalinization and a vesicular acidification. Both, H2O2 (0.2 mmol/l) and t-butyl-hydroperoxide (0.2 mmol/l) were without effect on pHcyt, but lowered apparent pHves by about 0.2 pH units. Ba2+ (1 mmol/l) diminished the acidifying effect of the hydroperoxides by about 50%. Pretreatment of the cells with colchicine, but not with lumicolchicine, largely abolished the effects of aniso-osmolarity and hydroperoxides on pHves. The data suggest that hepatocellular hydration affects the proton gradients built up across the membranes of endocytotic FITC-dextran-accessible compartments in a microtubule-dependent way. They further suggest that hydroperoxides induce vesicular acidification in a colchicine- and Ba(2+)-sensitive way. Because hydroperoxides induce Ba(2+)-sensitive cell shrinkage [Hallbrucker, Ritter, Lang, Gerok and Häussinger (1992) Eur. J. Biochem. 211, 449-458], the results are compatible with the view that hydroperoxide-induced cell shrinkage mediates vesicular acidification. It is concluded that modulation of vesicular pH by the hepatocellular hydration state may play a role in triggering some metabolic changes in response to cell swelling/shrinkage.

The effect of aniso-osmotic exposure on the level of inducible cyclooxygenase (Cox-2) and on prostanoid synthesis was studied in cultured rat liver macrophages (Kupffer cells). In lipopolysaccharide (LPS)- or phorbol 12-myristate 13-acetate-stimulated Kupffer cells, hyperosmotic (355 mosmol/l) exposure, due to addition of NaCl or impermeant sugars, markedly increased prostaglandin (PG) E2, D2 and thromboxane B2 synthesis in a time- and osmolarity-dependent manner. Increased prostanoid production was observed about 8 h after exposure to LPS in hyperosmotic medium compared to Kupffer cells treated with LPS under normotonic (305 mosmol/l) conditions. A similar stimulatory effect of hyperosmolarity on PGE2 production was also seen when arachidonate was added exogenously. Hyperosmotic stimulation of PGE2 production was accompanied by a strong induction of Cox-2 mRNA levels and an increase in immunoreactive Cox-2, whereas the levels of immunoreactive phospholipase A2 and cyclooxygenase-1 did not change significantly. Dexamethasone, indomethacin and the selective Cox-2 inhibitor, NS-398, abolished the hypertonicity-induced stimulation of PGE2 formation; dexamethasone also prevented the increase in Cox-2 mRNA and protein. The increase of immunoreactive Cox-2 lasted for about 24 h and was also blocked by actinomycin D or cycloheximide, but not by brefeldin A. Tunicamycin or treatment with endoglucosidase H reduced the molecular mass of hypertonicity-induced Cox-2 by 5 kDa. Tunicamycin treatment also suppressed the hypertonicity-induced stimulation of PGE2 production. The hyperosmolarity/LPS-induced stimulation of prostaglandin formation was partly sensitive to protein kinase C inhibition but was not accompanied by an increase in the cytosolic free Ca2+ concentration. The data suggest that osmolarity may be a critical factor in the regulation of Cox-2 expression and prostanoid production in activated rat liver macrophages.

Cell swelling is shown to induce an increase in acridine orange fluorescence intensity, an effect pointing to the alkalinization of acidic vesicles. Since autophagic hepatic proteolysis is accomplished by pH-sensitive proteinases within acidic lysosomes, this effect may contribute to the well-known inhibitory effect of cell swelling on proteolysis. In the present study, the role ofmicrotubules in volume-dependent alterations of pH in acidic vesicles of rat and human hepatocytes was studied. Colcemid and colchicine were used to depolymerize microtubules and vesicular pH was monitored using two dif- Amino acids elicit cell swelling by their concentrative uptake and cellular accumulation, while insulin exerts its effect by activation of ion uptake via Na+, K+, 2C-cotransport and Na+/H+ exchange (1, 2). The antiproteolytic action of insulin and glutamine is fully mimicked by the respective osmotic alterations of cell volume and is abolished if alterations of cell volume are reversed by appropriate alterations of extracellular osmolarity. Further, inhibition of Na+, K+, 2C-cotransport by furosemide or bumetanide not only blunts the swelling effect of insulin but also leads to a proportional inhibition of proteolysis. In the presence of both furosemide and amiloride, insulin-induced cell swelling and its antiproteolytic action are completely abolished. The mechanism linking cell volume alterations to proteolysis remained elusive, however, until recently when it was observed that cell swelling leads to an increase in acridine orange fluorescence intensity. This effect indicated an alkalinization of acidic intracellular compartments (3). Since proteolysis resides largely within acidic lysosomes and is accomplished by pH-sensitive lysosomal proteinases (4), the alkalinization of acidic intracellular compartments could indeed couple cell swelling to the inhibition of proteolysis. The mechanism by which cell volume changes are communicated to the acidic cellular compartments remained, however, unknown. In view of the described interaction between microtubules and intracellular vesicles (5, 6), the present study has been performed to examine a possible involvement of the microtubule network. To this end, the effect of cell swelling on fluorescein isothiocyanate (FITC)-dextran or acridine orange fluorescence has been studied both in intact cells and in cells treated with either colchicine or colcemid, drugs known to depolymerize microtubules (7,8). In addition, a possible effect of y--lumicolchicine, a stereoisomer of colchicine without an inhibitory effect on microtubules (9), was investigated. METHODSCell Culture. Human hepatocytes were prepared from pieces ofliver not appropriate for transplantation. The pieces were perfused through polyethylene catheters inserted into the main portal veins of the cut surface. The subsequent isolation procedure was according to the method ofBerry and Friend (10) with some modifications (11). Rat hepatocytes were prepared by collagenase treatment according to a method previously de...