Effect of RNA and protein synthesis inhibitors on the release of inflammatory mediators by macrophages responding to phorbol myristate acetate

The interaction of phorbol myristate acetate with resident populations of mouse peritoneal macrophages causes an increased release of arachidonic acid followed by increased synthesis and secretion of prostaglandin E2 and 6-keto-prostaglandin F1 alpha. In addition, phorbol myristate acetate causes the selective release of lysosomal acid hydrolases from resident and elicited macrophages. These effects of phorbol myristate acetate on macrophages do not cause lactate dehydrogenase to leak into the culture media. The phorbol myristate acetate-induced release of arachidonic acid and increased synthesis and secretion of prostaglandins by macrophages can be inhibited by RNA and protein synthesis inhibitors, whereas the release of lysosomal hydrolases is unaffected. 0.1 microgram/ml actinomycin D blocked the increased prostaglandin production due to this inflammatory agent by more than 80%, and 3 microgram/ml cycloheximide blocked prostaglandin production by 78%. Similar results with these metabolic inhibitors were found with another stimulator of prostaglandin production, zymosan. However, these inhibitors do not interfere with lysosomal hydrolase releases caused by zymosan or phorbol myristate acetate. It appears that one of the results of the interaction of macrophages with inflammatory stimuli is the synthesis of a rapidly turning-over protein which regulates the production of prostaglandins. It is also clear that the secretion of prostaglandins and lysosomal hydrolases are independently regulated.     

Relationship of prostaglandin secretion by rabbit alveolar macrophages to phagocytosis and lysosomal enzyme release

The phospholipids of rabbit alveolar macrophages were pulse-labelled with [(14)C]-arachidonic acid, and the subsequent release of labelled prostaglandins was measured. Resting macrophages released measurable amounts of arachidonic acid, the prostaglandins E(2), D(2) and F(2alpha) and 6-oxoprostaglandin F(1alpha). Phagocytosis of zymosan increased the release of arachidonic acid and prostaglandins to 2.5 times the control value. In contrast, phagocytosis of inert latex particles had no effect on prostaglandin release. Indomethacin inhibited the release of prostaglandin, and, at high doses (20mug/ml), increased arachidonic acid release. Analysis of the cellular lipids showed that after zymosan stimulation the proportion of label was decreased in phosphatidylcholine, but not in other phospholipids or neutral lipids. Cytochalasin B, at a dose of 2mug/ml, inhibited the phagocytosis induced by zymosan but increased prostaglandin synthesis to 3.4 times the control. These data suggest that the stimulation of prostaglandin synthesis by zymosan is not dependent on phagocytosis. Exposure to zymosan also resulted in the release of the lysosomal enzyme, acid phosphatase. Furthermore, cytochalasin B augmented the zymosan-stimulated release of acid phosphatase at the same dose that stimulated prostaglandin synthesis. However, indomethacin, at a dose that completely inhibited prostaglandin synthesis, failed to block the lysosomal enzyme release. Thus despite some parallels between the release of prostaglandins and lysosomal enzymes, endogenous prostaglandins do not appear to mediate the release of lysosomal enzymes. The prostaglandins released from the macrophages may function as humoral substances affecting other cells.     

Effect of RNA and protein synthesis inhibitors on the release of inflammatory mediators by macrophages responding to phorbol myristate acetate

The interaction of phorbol myristate acetate with resident populations of mouse peritoneal macrophages causes an increased release of arachidonic acid followed by increased synthesis and secretion of prostaglandin E₂ and 6-keto-prostaglandin F_1α. In addition, phorbol myristate acetate causes the selective release of lysosomal acid hydrolases from resident and elicited macrophages. These effects of phorbol myristate acetate on macrophages do not cause lactate dehydrogenase to leak into the culture media. The phorbol myristate acetate-induced release of arachidonic acid and increased synthesis and secretion of prostaglandins by macrophages can be inhibited by RNA and protein synthesis inhibitors, whereas the release of lysosomal hydrolases is unaffected. 0.1 μg/ml actinomycin D blocked the increased prostaglandin production due to this inflammatory agent by more than 80%, and 3 μg/ml cycloheximide blocked prostaglandin production by 78%. Similar results with these metabolic inhibitors were found with another stimulator of prostaglandin production, zymosan. However, these inhibitors do not interfere with lysosomal hydrolase releases caused by zymosan or phorbol myristate acetate. It appears that one of the results of the interaction of macrophages with inflammatory stimuli is the synthesis of a rapidly turning-over protein which regulates the production of prostaglandins. It is also clear that the secretion of prostaglandins and lysosomal hydrolyses are independently regulated.     

Biosynthesis and transport of lysosomal enzymes in human monocytes and macrophages. Effects of ammonium chloride, zymosan and tunicamycin.

Human monocytes and macrophages synthesize lysosomal enzymes as larger precursors. The polypeptide patterns of several lysosomal-enzyme precursors and their mature forms are similar to those observed in human fibroblasts. Like fibroblasts, the monocytes and macrophages release small amounts of lysosomal-enzyme precursors. The lysosomotropic NH4+ cation enhances this release. In contrast, zymosan, a degranulating agent, causes release of both the mature and the precursor forms of the lysosomal enzymes. Both NH4Cl and zymosan inhibit maturation of the precursors. The fractional amounts of mature cathepsin D and beta-hexosaminidase released in the presence of zymosan are strikingly different. Probably, in the macrophages several lysosomal organelles are packaged with different relative contents of lysosomal enzymes. The transport of the precursors of cathepsin D into lysosomes is inhibited by tunicamycin. Therefore oligosaccharide side chains are likely to function as signals in packaging of lysosomal enzymes in macrophages also.     

Regulation of macrophage eicosanoid production by hydroperoxy-and hydroxy-eicosatetraenoic acids.

Resident mouse peritoneal macrophages when exposed to zymosan during the first day of cell culture synthesize and secrete large amounts of prostaglandin E2 (PGE2) and leukotriene C4 (LTC4), the respective products of cyclo-oxygenase- and 5-lipoxygenase-catalysed oxygenations of arachidonic acid. Under these conditions of cell stimulation only small amounts of hydroxyeicosatetraenoic acids (HETEs) are concomitantly produced. However, exogenously added arachidonic acid is metabolized to large amounts of 12- and 15-HETE and only relatively small amounts of PGE2. No LTC4 is formed under these conditions. In contrast, resident mouse peritoneal macrophages in cell culture for 4 days synthesized less PGE2 and LTC4 when exposed to zymosan. However, these macrophage populations continue to synthesize 12-HETE from exogenously added arachidonic acid. Zymosan induced the secretion of a lysosomal enzyme, N-acetyl-beta-glucosaminidase, equally in both 1- and 4-day cultures. Both 12- and 15-hydroperoxyeicosatetraenoic acids (HPETEs), the precursors of 12- and 15-HETE, were found to be irreversible inhibitors of the cyclo-oxygenase pathway and reversible inhibitors of the 5-lipoxygenase pathway in macrophages. 15-HETE were found to be reversible inhibitors of both pathways. Thus the oxidation of arachidonic oxidation of arachidonic acid to both prostaglandins and leukotrienes may be under intracellular regulation by products of 12- and 15-lipoxygenases.     

Errata

Rat peritoneal macrophages when incubated with leukotriene C exhibited a dose dependent increase of PGE₂ and 6-keto-PGF_1α. Nordihydroguaiaretic acid, an inhibitor of lipoxygenase, prevented the stimulation of prostaglandin release by lipopolysaccharide, but did not affect the release of prostaglandins from non-stimulated cells. Indomethacin, in contrast to nordihydroguaiaretic acid, inhibited the release of prostaglandins from both stimulated and non-stimulated cells. These findings suggest that endogenous leukotrienes are involved in the stimulation of macrophages by lipopolysaccharide.     

Effect of toxins on arachidonic acid metabolism in rat cultured pulmonary alveolar macrophages

The action of a trichothecene (T-2), microcystin-LR and saxitoxin on arachidonic acid metabolism in cultured rat alveolar macrophages was studied. Pulmonary macrophages exposed to T-2 trichothecene were stimulated to synthesize and release large amount of thromboxane B2 (TxB2) and 6-Keto F1 alpha. Microcystin-LR induced significant release of prostaglandins F2 alpha (140%), PGE2 (175%) and TxB2 (169%) compared to controls. Saxitoxin induced TxB2 release by 37%. Arachidonic acid release was stimulated by all three toxins. The release of arachidonic acid and its metabolites in alveolar macrophages exposed to T-2 toxin was partially blocked by fluocinolone (1 microM). These results suggest that macrophages synthesize and release inflammatory mediators in response to toxin exposure, and fluocinolone may protect against T-2 toxicosis.     

Glutathione metabolism in resting and phagocytizing peritoneal macrophages

The steady state GSH content of cultured mouse resident peritoneal macrophages was 34 +/- 5 pmol/microgram of cell protein. Intracellular GSH content decreased concomitantly with zymosan ingestion. The half-life of GSH decreased from 1.9 h in resting cells to 0.58 h during phagocytosis as determined by inhibition of GSH synthesis with buthionine sulfoximine. The decrease in GSH half-life was directly related to the extent of particle uptake. In cytochalasin D-treated cells, attachment of zymosan to the macrophage plasma membrane in the absence of particle interiorization was sufficient to stimulate GSH turnover. Efflux was the major route of GSH loss in [35S]cystine-labeled macrophages, and was enhanced 3-fold by a zymosan challenge. GSH was lost intact since resident macrophages lack gamma-glutamyl transpeptidase (less than 1 pmol of L-gamma-glutamyl-p-nitroanilide/microgram of protein . h). Macrophages obtained from mice challenged in vivo with Corynebacterium parvum maintained higher intracellular GSH levels (50 +/- 5 pmol/microgram of cell protein) than did resident cells. The half-life of GSH in buthionine sulfoximine-treated C. parvum-elicited macrophages was 3.8 +/- 0.2 h while resting and 1.3 +/- 0.2 h during phagocytosis. C. parvum-elicited macrophages, in contrast to resident cells, contained sufficient levels of gamma-glutamyl transpeptidase activity to hydrolyze 55 pmol of L-gamma-glutamyl-p-nitroanilide/microgram of cell protein . h. These studies indicate that phagocytosis and cellular activation have profound effects on GSH metabolism in macrophages.     

Constitutive Secretion of Acid Hydrolases in Tetrahymena thermophila

THE role of lysosomal enzymes in intracellular digestion is now well established [11]. Most often we think of lysosomal hydrolases in catabolism of endogenous or foreign material taken up by endocytosis. There is however, a number of reports dealing with the release of acid hydrolases into the extracellular fluid in a variety of eukaryote cells. These cells range from Saccharomyces cerevisiae [15], Dictyostelium discoideum [10], Leishmania donovani [20], Acanthamoeba castellani [22], Entamoeba histolytica [12, 31], and species of Tetrahymena [1–3, 6] to mammalian cells in culture [49]. Concerning the latter, fibroblasts and hepatocytes in culture release acid hydrolases to the extracellular medium, but only if the synthesis of a specific recognition marker is impaired in the cells. This marker (man-nose-6-phosphate) is used for receptor mediated segregation of lysosomal enzymes into the lysosomal compartments. If the receptor or the marker are lacking, the hydrolases fail to enter the lysosomal compartment, and are secreted in immature form together with molecules belonging to the constitutive secretory pathway of the cells [8, 49]. Such a release of acid hydrolases seems to occur spontaneously from mammalian osteoclasts [4]. Macrophages, on the other hand, need a specific stimulation for their release process [40]. In lower eukaryotes the release may     

Induction of macrophage lysosomal hydrolase synthesis and secretion by beta-1,3-glucan

Studies were undertaken to elucidate the active component in zymosan necessary to induce the delayed-onset synthesis and secretion of representative lysosomal hydrolases, hexosaminidase, and beta-glucuronidase in macrophages. Resident mouse peritoneal macrophages were challenged with zymosan particles and particulate beta-1,3-glucan, the major subcomponent of zymosan. Zymosan was found to induce a rapid secretion of preformed hexosaminidase with maximal release (75%) occurring 6 hr after the addition of zymosan. By contrast, beta-1,3-glucan was totally inactive in this respect. However, both zymosan and beta-1,3-glucan were found to induce the delayed-onset synthesis and secretion of hexosaminidase and beta-glucuronidase while maintaining constant cellular enzyme levels over a 5-day period following the addition of stimulus. These late responses were almost totally blocked by a noncytolytic concentration of cycloheximide, indicating their dependence on de novo protein synthesis. Mannan, the second major subcomponent of zymosan, had no effect on either immediate secretion or delayed-onset synthesis and secretion of hexosaminidase. These results suggest that the induction of the delayed-onset synthesis and secretion of the lysosomal hydrolases by zymosan may be dependent on the glucan subcomponent of zymosan. Moreover, it would also appear that the release of preformed lysosomal enzymes is not the trigger for the delayed-onset synthesis and secretion of hexosaminidase.     

The use of acridine orange cytofluorometry in the study of macrophage lysosomal exocytosis

We present a rather simple cytofluorometric technique for the study of exocytosis of lysosomal contents from individual cultured cells. It is based on the use of the lysosomotropic weak base acridine orange (AO) which, in its stacked form, as it occurs within lysosomes, emits red fluorescence when excited by blue light. Mouse peritoneal macrophages were cultured for 48 h and, after 2 h in serum-free medium, stained with AO. The cells were then exposed to F10-medium with or without newborn calf serum (NCS), zymosan A (Z) or cytochalasin B (CB) for different times at 20 or 37 degrees C. After staining, the macrophages showed no change in red fluorescence intensity, if stored at room temperature in the dark. If, however, the cells were kept in the incubator at 37 degrees C, the cells showed slightly decreasing red fluorescence intensity with time. This decrease was markedly potentiated by the presence of NCS, Z or CB, which are known to induce secretion of lysosomal enzymes from macrophages in vitro. Selective lysosomal enzyme release was confirmed biochemically during treatment with zymosan A. The technique presented here may be of value in further studies on the stimulation of, and the mechanisms behind, lysosomal exocytosis in cultured cells.     

Mechanisms of lysosomal enzyme release from human polymorphonuclear leukocytes. Effects of phorbol myristate acetate

PMA enhanced release of the azurophil granule enzyme, beta-glucuronidase, as well as lysozyme, from cytochalasin B-treated PMN's exposed to either zymosan particles or C5a. PMA was active at nanomolar concentrations, was not toxic to the cells, and was most effective when present for brief durations (0-1 min) before exposure of the cells to the stimuli. Beta-glucuronidase was not released in significant amounts from PMN's exposed to PMA alone, in the absence of stimuli such as zymosan or C5a. In contrast, only the specific granule enzyme, lysozyme, was released from unstimulated cells. Electron micrographs of cells exposed to PMA revealed an increase in the number of visible cytoplasmic microtubules as compared to control cells. Enhancement of lysosomal enzyme (beta-glucuronidase) release by PMA appears to be independent of effects on release of specific granule enzymes (lysozyme), but rather is likely due to PMA-induced elevations of cellular cGMP.     

Effect of Muramyldipeptide,a Synthetic Bacterial Adjuvant,on Enzyme Release from Cultured Mouse Macrophages

Monolayer cultures of macrophages obtained by peritoneal lavage of normal or thioglycollate-stimulated mice spontaneously secreted lysosomal enzymes into the culture medium. When the elicited macrophages were cultured in the presence of muramyldipeptide (MDP), a 20–30% increase in the release of β-glucuro-nidase was consistently observed and the intracellular activity decreased to about 45% of that of control cells after 6–8 days' culture. A stimulatory effect of MDP on lysozyme secretion, though less profound, was also observed. In contrast, release of neither enzyme was stimulated in resident macrophages by the addition of MDP. A neutral α-glucosidase, which has recently been found to localize also in granules of macrophages, remained inside the cells and neither its activity nor its release was affected by the addition of MDP to either type of macrophages. A large amount of lactic dehydrogenase was released only when the resident, not the elicited, macrophages were cultured for 3–4 days and then phagocytosed zymosan.     

Cholesterol delivered to macrophages by oxidized low density lipoprotein is sequestered in lysosomes and fails to efflux normally

Oxidized low density lipoprotein (LDL) has been found to exhibit numerous potentially atherogenic properties, including transformation of macrophages to foam cells. It is believed that high density lipoprotein (HDL) protects against atherosclerosis by removing excess cholesterol from cells of the artery wall, thereby retarding lipid accumulation by macrophages. In the present study, the relative rates of HDL-mediated cholesterol efflux were measured in murine resident peritoneal macrophages that had been loaded with acetylated LDL or oxidized LDL. Total cholesterol content of macrophages incubated for 24 h with either oxidized LDL or acetylated LDL was increased by 3-fold. However, there was no release of cholesterol to HDL from cells loaded with oxidized LDL under conditions in which cells loaded with acetylated LDL released about one-third of their total cholesterol to HDL. Even mild degrees of oxidation were associated with impairment of cholesterol efflux. Macrophages incubated with vortex-aggregated LDL also displayed impaired cholesterol efflux, but aggregation could not account for the entire effect of oxidized LDL. Resistance of apolipoprotein B (apoB) in oxidized LDL to lysosomal hydrolases and inactivation of hydrolases by aldehydes in oxidized LDL were also implicated. The subcellular distribution of cholesterol in oxidized LDL-loaded cells and acetylated LDL-loaded cells was investigated by density gradient fractionation, and this indicated that cholesterol derived from oxidized LDL accumulates within lysosomes. Thus impairment of cholesterol efflux in oxidized LDL-loaded macrophages appears to be due to lysosomal accumulation of oxidized LDL rather than to impaired transport of cholesterol from a cytosolic compartment to the plasma membrane.     

The selective release of phospholipase A2 by resident mouse peritoneal macrophages.

Resident mouse peritoneal macrophages have three phospholipase activities: a phospholipase A2 active at pH 4.5, a Ca2+-dependent phospholipase A2 active at pH 8.5 and a phosphatidylinositol-specific phospholipase C activity. When macrophages are exposed to zymosan in culture, the cellular activity of pH-4.5 phospholipase A2 is diminished in a manner dependent on zymosan concentration and time of exposure, whereas the cellular activities of pH-8.5 phospholipase A2 and phospholipase C remain unchanged. The depletion of pH-4.5 phospholipase A2 activity from the cell is paralleled by a quantitative recovery of this activity in the culture medium in a manner similar to the cellular depletion and extracellular recovery of two lysosomal enzymes. This release is specifically elicited by an inflammatory substance such as zymosan, since macrophages incubated with 6 micrometer latex spheres retain pH-4.5 phospholipase A2 activity and lysosomal enzyme activities intracellularly.     

Transmembrane potential changes during phagocytosis in rat alveolar macrophages

Studies were carried out to measure changes in the transmembrane potential of rat alveolar macrophages during exposure of the cells to zymosan particles or to the membrane perturbant, phorbol-12-myristate-13-acetate (PMA), and to determine if changes in membrane potential are related to superoxide anion release. Exposure of the cells to either zymosan or PMA leads to membrane depolarization, which precedes superoxide anion release. Furthermore, the magnitude of the depolarization is dependent upon the concentration of either zymosan or PMA. During exposure of the alveolar macrophages to increasing levels of zymosan, there is an increase in the amount of superoxide released as well as an increase in the magnitude of the depolarization. Incubation of the cells in medium containing 150 mM K+, a medium which causes membrane depolarization, leads to superoxide release from resting cells and a decrease in the amount of superoxide released from cells exposed to zymosan. These results indicate that release of superoxide anion from rat alveolar macrophages is related to membrane depolarization and suggest that the transmembrane potential change may act as a signal to initiate the phagocytotic responses of the cells.     

Enhanced release of oxygen metabolites by monocyte-derived macrophages exposed to proteolytic enzymes: activity of neutrophil elastase and cathepsin G

Macrophages at sites of inflammation are exposed to proteolytic enzymes derived from neutrophils, platelets, clotting factors, complement, and damaged tissues. To investigate the possible effect of proteases on the plasma membrane-mediated oxidative metabolic response of macrophages in inflammatory sites, cultured human monocyte-derived macrophages were treated in vitro with proteolytic enzymes and were then assayed for their ability to release superoxide anion (O2-) and hydrogen peroxide (H2O2) in response to stimuli. Macrophages pretreated for 1 to 20 min with trypsin, chymotrypsin, pronase, or papain, 0.1 to 200 micrograms/ml, released up to 3.5-times more O2- and H2O2 than did control (untreated) cells. This enhanced production of oxygen metabolites was observed by using either phorbol myristate acetate or opsonized zymosan as the stimulus. Macrophages were also "primed" for enhanced O2- release (2.3-fold) by pretreatment with a subfraction of granules extracted from human neutrophils. This subfraction contained primarily elastase and cathepsin G. Similar enhancement was observed with 60 ng/ml or purified human neutrophil cathepsin G (2.2-fold) and with 20 micrograms/ml of purified neutrophil elastase (3.3-fold). Priming by these neutrophil proteases could be blocked by specific inhibitors of their proteolytic activity. These results suggest that macrophages involved in an inflammatory response might be rapidly primed by proteases released from degranulating neutrophils. Primed macrophages could mount a more effective oxidative metabolic response to microorganisms or tumor cells, but might also cause greater tissue damage.     

Receptor-mediated phagocytosis of zymosan is unaffected by some conditions which reduce macrophage lysosomal enzyme secretion

We have previously shown that several agents which interfere with binding of ligands to the mannose-glycoprotein receptor on macrophages can inhibit zymosan-induced lysosomal enzyme secretion. Here we show that mannose only reduces the association of zymosan with macrophages during the first hour of exposure; after longer periods of uptake no effect is detectable. We have previously shown that mannose reduces surface binding of zymosan, probably by interfering selectively with binding to the mannose receptor. The present inhibition of association of zymosan with macrophages during short exposures can be entirely explained by this reduction of binding. Macrophages must therefore internalize zymosan at sites in addition to the mannose receptor. In contrast to macrophages the murine macrophage-like cell line P388D1 is lacking the mannose-glycoprotein receptor. Accordingly we find that binding of zymosan to P388D1 is much slighter than to macrophages and is unaffected by mannose or mannose-6-phosphate. The spontaneous lysosomal enzyme secretion of P388D1 is also unaffected by mannose. The data on macrophages confirm our previous suggestion that agents interfering with the mannose receptor inhibit the induction of lysosomal enzyme secretion by acting directly on the receptor. The data on P388D1 cells support this assertion by excluding effects at later steps in the secretory pathway.     

Effects of peptidoglycans from periodontal pathogens on selected biological activities of CD-1 murine peritoneal macrophages

Resident CD-1 murine peritoneal macrophages were exposed to various concentrations of purified peptidoglycan isolated from members of the genera Bacteroides, Eikenella, and Actinomyces. Macrophage viability, the release of lysozyme, acid phosphatase, and prostaglandins E1 and E2 were assayed as a function of peptidoglycan concentration and time. Macrophages responded as a function of peptidoglycan concentration with increased release of acid phosphatase and prostaglandins; all cells remained greater than 90% viable during the course of the experiments. However, concentrations of peptidoglycan greater than 50 micrograms/mL were toxic to the macrophages, while the peptidoglycan from B. capillus strain 925.08 and Actinomyces viscosus strain T14AV consumed complement by both the classical and the alternate pathways. Cellular lysozyme activity and phagocytosis of Saccharomyces cerevisiae were significantly reduced in the presence of peptidoglycan. When viewed by scanning electron microscopy, the activated macrophages were rounded, lacked distinct pseudopod extensions, and possessed an increased number of microvilli and plasma membrane associated vesicles. These morphological alterations occurred as early as 3 h. Transmission electron microscopy revealed the purified peptidoglycan to have been taken up into numerous phagosomes; however, even after 24 h incubation, it was only partially degraded.     

Prostaglandin release but not superoxide production by rat Kupffer cells stimulated in vitro depends on Na+/H+ exchange

The release of the prostaglandins E2 and D2, induced by zymosan and phorbol ester in cultured rat Kupffer cells, was found to depend on the extracellular concentration of Na+. Eicosanoid formation following the administration of the Ca2+ ionophore A23187 or of arachidonic acid, however, did not require the presence of sodium ions in the medium. A half-maximal rate of prostaglandin release by zymosan-treated Kupffer cells was obtained between 4 mM and 5 mM Na+; and a Na+ concentration of greater than or equal to 30 mM was required to maximally stimulate prostaglandin E2 and D2 formation in the cultured liver macrophages. In contrast, the superoxide production following the administration of zymosan or of phorbol ester was quite independent of extracellular Na+. The zymosan and phorbol-ester-stimulated release of prostaglandins E2 and D2 was inhibited by amiloride. Artificial intracellular alkalization enhanced the prostanoid production of unstimulated and of zymosan-stimulated cells whereas artificial intracellular acidification inhibited the zymosan-elicited prostaglandin synthesis. In contrast, the superoxide formation was independent of the pH changes. The data presented here suggest that the prostaglandin production elicited by zymosan or phorbol ester in cultured rat Kupffer cells requires an activated Na+/H+ exchange.