Increased cyclic AMP content directly correlated with morphological transformation of cells infected with a temperature-sensitive mutant of mouse sarcoma virus

Summary Normal rat kidney cells infected with a cold-sensitive mutant of mouse sarcoma virus [NRK(MSV-1b)] morphologically transform when exposed to adenosine 3′∶5′ cyclic monophosphate (cAMP) at the restrictive temperature. The cAMP-induced morphological changes occur rapidly and are reversible. Agents capable of elevating endogenous levels of cAMP [prostaglandin E₁ (PGE₁) and cholera toxin (CT)] induced morphological transformation of NRK(MSV-1b) cells at the restrictive temperature that was concentration dependent, potentiated by cAMP phosphodiesterase inhibitors, and not prevented by inhibitors of DNA, RNA, and protein synthesis. Prostaglandin E₁ stimulated a transient increase in the intracellular level of cAMP with a concomitant morphological transformation and reversion of cells as cAMP levels decline. The maximum increase is reached by 10 min, followed by a decline to near basal level by 80 min. In contrast, incubation of cells with CT resulted in irreversible morphological transformation and increased levels of cAMP first detectable by 1 hr with maximum levels reached by 24 hr. Heated CT (100°C, 20 min) was without effect. Addition of CT to reverted PGE₁-treated cells resulted in morphological transformation suggesting the existence of discrete receptors in NRK (MSV-1b) cells. This research was supported by Grant BC-207 from the American Cancer Society and Cancer Research Emphasis Grant R01 CA 19714 within the Virus Cancer Program of the National Cancer Institute.     

Somatocrinin (growth hormone releasing factor) in vitro bioactivity; Ca++ involvement, cAMP mediated action and additivity of effect with PGE2

The release of GH induced by purified hypothalamic GRF or native or synthetic tumor-derived GRF is antagonized by the presence of CoCl²; it is simulated by 8Br .cAMP, IBMX, cholera toxin, forskolin, with identical maximal effects (E_max). Somatocrinin (GRF) stimulates the efflux of cAMP by the pituitary cells in parallel to the release of GH. Addition of either 8Br .cAMP, IBMX, cholera toxin or forskolin to a maximally stimulating dose of GRF does not increase the response which remains GRF-E_max. In contradistinction with these results PGE₂ releases GH with a dose-response curve different from that of GRF, and the combination of PGE₂ + GRF produces an E_max far greater than that due to either agonist alone; showing a true additivity. The name $\text{somatocrinin}$ is proposed to replace the acronym GRF.     

The mechanism of action of soluble lymphocyte mediators. V. Mechanism of refractoriness to adenylate cyclase stimulators induced by migration inhibitory factor (MIF).

Exposure of purified guinea pig macrophages to lymphocyte culture supernatants, containing migration inhibitory factor (MIF), for a minimum of 1 hr results in reduced cyclic 3′,5′-adenosine monophosphate (cAMP) accumulation in response to adenylate cyclase stimulators, which persists for at least 6 hr. MIF-induced refractoriness is not due to: inactivation of the stimulating agents, destruction of membrane receptors, excessive leakage of cAMP from the cells, or activation of cAMP phosphodiesterase. Refractoriness develops normally in the absence of protein synthesis. It is concluded that MIF-induced refractoriness is an expression of diminished cAMP synthesis. However, MIF does not directly inhibit basal or stimulated adenylate cyclase, nor are prostaglandin synthesis intermediates responsible for the reduced cAMP production. Evidence is presented in support of the proposal that MIF interferes with the signal transfer from membrane receptors to the enzyme by an effect on cytoplasmic microtubules.     

Cell surface receptors for lymphokines. I. The possible role of glycolipids as receptors for macrophage migration inhibitory factor (MIF) and macrophage activation factor (MAF).

Incubation of culture supernatants from concanavalin A-stimulated guinea pig and rat lymphocytes with protein-free preparations of bovine brain gangliosides abolished their macrophage migration inhibitory factor (MIF) and macrophage activation factor (MAF) activity. The identity of the MIF/MAF-binding component(s) present in these glycolipid mixtures has yet to be established, but adsorption experiments using purified preparations of mono- (GM₁, GM₂, and GM₃), di- (GD_1a), and trisialogangliosides (GT₁) were negative. Since these gangliosides account for over 90% of the glycolipid content in brain ganglioside mixtures it appears that the MIF-binding component(s) is present only in very small amounts. Treatment of guinea pig peritoneal macrophages with liposomes containing similar brain gangliosides or water-soluble glycolipids extracted from guinea pig macrophages enhanced their responsiveness to MIF. The enhanced response to MIF of liposome-treated macrophages was abolished by incubation of the treated macrophages with fucose-binding lectins (Lotus agglutinin and Ulex europaeus agglutinin I) before exposure to MIF, suggesting that the MIF-binding component donated by the liposomes may be a fucose-containing glycolipid. The possible role of glycolipids as surface receptors for MIF and MAF is discussed.     

Facilitation of adenylate cyclase stimulation in macrophages by lectins.

Preincubation of guinea pig peritoneal macrophages with concanavalin A (Con A) markedly enhanced the accumulation of 3′,5′-cyclic-adenosine monophosphate (cAMP) in response to the adenylate cyclase (AC) stimulators prostaglandin E₁ (PGE₁) and isoproterenol (IP). Basal cAMP levels were not altered. Maximal enhancement of cAMP accumulation was induced by preincubation with 50–100 μg/ml Con A for 10 min at 37 °C. Con A-induced facilitation of macrophage responsiveness was prevented by α-methyl-d-mannoside (αMM). No facilitation was induced by the divalent derivative, succinyl-Con A or by Con A immobilized on Sepharose beads. Con A-induced facilitation developed normally in macrophages treated with the microfilament blocking agent, cytochalasin B. The responsiveness of macrophages to PGE₁ and IP was also augmented by phytohemagglutinin (PHA) but wheat germ agglutinin (WGA), soy bean agglutinin (SBA), pokeweed mitogen (PWM), and Lotus tetragonolobus lectin (LL) showed no enhancing effect. The effect of Con A on cAMP levels was the result of augmented cAMP synthesis and not of reduced degradation or a block in cAMP egress from the cells. Lectin-induced facilitation of AC stimulation could be mediated via one of the following mechanisms: (i) induction of receptor clustering; (ii) causing a conformational change in the receptors; (iii) inhibition of negative cooperativity; (iv) causing an increase in membrane fluidity; (v) disruption of microtubules by acting as a Ca²⁺ ionophore; or (vi) inactivation of a sugar-containing inhibitor of AC.     

Interactions of prostaglandins with subpopulations of ovine luteal cells. I. Stimulatory effects of prostaglandins E1, E2 and I21

Preparations of small and large steroidogenic cells from enzymatically dispersed ovine corpora lutea were utilized to study the $\text{in}$$\text{vitro}$ effects of luteinizing hormone (LH) and prostaglandins (PG) E₁, E₂ and I₂. Cells were allowed to attach to culture dishes overnight and were incubated with either LH (100 ng/ml), PGE₂, PGE₂, or PGI₂ (250 ng/ml each). The secretion of progesterone by large cells was stimulated by all prostaglandins tested (P < 0.05) while the moderate stimulation observed after LH treatment was attributable to contamination of the large cell population with small cells. Prostaglandins E₁ and E₂ had no effect on progesterone secretion by small cells, while LH was stimulatory at all times (0.5 to 4 hr) and PGI₂ was stimulatory by 4 hr. Additional studies were conducted to determine if the effects of PGE₂ upon steroidogenesis in large cells were correlated with stimulated activity of adenylate cyclase. In both plated and suspended cells PGE₂ caused an increase (P < 0.05) in the rate of progesterone secretion but had no effect upon the activity of adenylate cyclase or cAMP concentrations within cells or in the incubation media. Exposure of luteal cells to forskolin, a nonhormonal stimulator of adenylate cyclase, resulted in marked increases in all parameters of cyclase activity but had no effect on progesterone secretion. These data suggest that the actions of prostaglandins E₁, E₂ and I₂ are directed primarily toward the large cells of the ovine corpus luteum and cast doubt upon the role of adenylate cyclase as the sole intermediary in regulation of progesterone secretion in this cell type.     

Interactions of prostaglandins with subpopulations of ovine luteal cells. I. Stimulatory effects of prostaglandins E1, E2 and I2

Preparations of small and large steroidogenic cells from enzymatically dispersed ovine corpora lutea were utilized to study the effects of luteinizing hormone (LH) and prostaglandins (PG) E₁, E₂ and I₂. Cells were allowed to attach to culture dishes overnight and were incubated with either LH (100 ng/ml), PGE₂, PGE₂, or PGI₂ (250 ng/ml each). The secretion of progesterone by large cells was stimulated by all prostaglandins tested (P < 0.05) while the moderate stimulation observed after LH treatment was attributable to contamination of the large cell population with small cells. Prostaglandins E₁ and E₂ had no effect on progesterone secretion by small cells, while LH was stimulatory at all times (0.5 to 4 hr) and PGI₂ was stimulatory by 4 hr. Additional studies were conducted to determine if the effects of PGE₂ upon steroidogenesis in large cells were correlated with stimulated activity of adenylate cyclase. In both plated and suspended cells PGE₂ caused an increase (P < 0.05) in the rate of progesterone secretion but had no effect upon the activity of adenylate cyclase or cAMP concentrations within cells or in the incubation media. Exposure of luteal cells to forskolin, a nonhormonal stimulator of adenylate cyclase, resulted in marked increases in all parameters of cyclase activity but had no effect on progesterone secretion. These data suggest that the actions of prostaglandins E₁, E₂ and I₂ are directed primarily toward the large cells of the ovine corpus luteum and cast doubt upon the role of adenylate cyclase as the sole intermediary in regulation of progesterone secretion in this cell type.     

Suppression of lymphokine production: I. Macrophage-mediated inhibition of MIF production

Macrophages have been found to suppress the in vitro production by stimulated T lymphocytes of a lymphokine, migration inhibitory factor. When macrophages isolated from primary MSV-induced tumors were added to antigen-stimulated MSV-immune spleen cells, a complete suppression of MIF production was observed. This suppression was nonspecific, since MIF production by antigen-stimulated alloimmune spleen cells and by PHA-stimulated normal spleen cells was also inhibited. Suppressor macrophages could also be induced by inoculation with Corynebacterium parvum, whereas light mineral oil-induced peritoneal macrophages had no detectable effect on MIF production. The failure to detect MIF in the supernatants of stimulated cultures containing activated macrophages appeared to be due to inhibition of lymphokine production rather than to absorption or inactivation of MIF or to interference with the assay for detection of MIF. Macrophages were able to suppress MIF production only when added during the first 4–5 hr of culture and they had no effect when added later. These data show that activated macrophages can nonspecifically suppress lymphokine production and that this appears to be due to inhibition of an early step in lymphocyte stimulation.     

Alpha-adrenergic receptors on human platelets.

[³H] dihydroergocyrptine, an α-adrenergic antagonist, binds specifically to sites on human platelet membranes. Prostaglandin E₁ (PGE₁) stimulates the production of cyclic AMP (cAMP) in human platelets. Alpha-adrenergic agonists, 1-epinephrine and 1-norepinephrine, and antagonists, phentolamine, phenoxybenzamine, and dihydroergocyrptine inhibit the binding of [³H] dihydroergocryptine. The α-adrenergic agonists inhibit PGE₁-stimulated cAMP production and the α-adrenergic antagonists phentolamine and dihydroergocryptine reverse this inhibition. The β-adrenergic agonist 1-isoproterenol and the β-adrenergic antagonists d1-propranolol and 1-alprenolol do not significantly alter binding or PGE₁-stimulated cAMP production. Clonidine, dopamine, and serotonin inhibit binding, but clonidine and dopamine are weak inhibitors of PGE₁-stimulated cAMP production, and serotonin is without effect. Tyramine, an amine without direct adrenergic activity fails to inhibit binding. Alpha-adrenergic agonists decrease the apparent affinity of a PGE₁-receptor activating cAMP production. The inhibition of PGE₁-stimulated cAMP production is a physiological measure of α-adrenergic agonist binding to the α-receptor.     

Effects of crude and pure cholera toxin on prostaglandin release from the rabbit ileum

Investigations were made into the effects of crude and pure preparations of cholera toxin on the release of prostaglandin-like substances (PLS) from rabbit ileum. Perfusion of ileal loops with buffer containing crude toxin was followed by a release of PLS into the perfusate, in amounts up to 37.5 ng/30 min (PGE₂ equivalents). In contrast, no detectable PLS was released when ileal loops were perfused with pure toxin. Similarly, pieces of ileum opened longitudinally released PLS in amounts up to 107 ng PGE₂/g tissue when incubated with crude toxin for 1–4 hr, but no release of PLS was detected in the presence of pure toxin under comparable conditions.Treatment of rabbits with indomethacin, 1.6 mg/kg p.o., had no effect on the accumulation of fluid in ileal sacs injected with crude or pure cholera toxin. These results support the view that prostaglandins do not play an essential role in the action of cholera toxin.     

Inhibition of cholera toxin-stimulated intestinal epithelial cell adenylate cyclase by adenosine analogs.

The ability of various adenosine analogs to inhibit cholera toxin activation of the intestinal epithelial cell adenylate cyclase-cyclic AMP system was investigated. After incubation of cells with cholera toxin for 6 hr, large increases in cellular cyclic AMP content were observed. Addition of 2', 5'-dideoxyadenosine during the last 30 min of this 6-hr incubation resulted in 70% reduction in elevated cyclic AMP content. Other analogs were not effective inhibitors. 2', 5'-Dideoxyadenosine was also a potent inhibitor of cholera toxin-activated intestinal cell adenylate cyclase activity with half-maximal inhibition occuring at 16 muM. NaF-stimulated cyclase was less susceptible to inhibition. The data suggest that inhibition by 2', 5'-dideoxyadenosine is due at least in part to direct inhibition of the cholera toxin-activated intestinal adenylate cyclase activity.     

The effect of PGE2 on the activation of quiescent lung fibroblasts

The effect of prostaglandin E₂ (PGE₂) on fibroblast proliferation was examined. The presence of PGE₂ for 24 h inhibited the growth of quiescent cells stimulated with serum, platelet-derived growth factor and macrophage-derived factors. Maximal inhibition of nuclear labeling with [³H]thymidine occurred at concentrations greater than 10⁻⁷ M. The inhibitory effect of PGE₂ was less potent in exponentially growing cells and was not the result of conversion of PGE₂ to PGA₂ during incubation in growth medium. The G₁ phase was determined to be 12–14 h in untreated cultures. The extent of growth inhibition by PGE₂ was similar with addition of PGE₂ at 0, 3, 6, or 9 h following restimulation of quiescent cell cultures. Approximately 25% of the cells that enter S phase are refractory to PGE₂-induced growth inhibition. Short-term exposure to PGE₂ (5 min and 30 min) caused substantial growth inhibition. The serum-induced proliferation was also inhibited by the cAMP analogue, dibutyrl cAMP. Our results suggest that PGE₂ affects a distinct subpopulation of cells. Restimulation of quiescent cells treated with PGE₂ for 24 h, indicated that release from PGE₂ exposure is associated with prolongation of the G₁ phase of the cell cycle.     

Effects of crude and pure cholera toxin on prostaglandin release from the rabbit ileum

Investigations were made into the effects of crude and pure preparations of cholera toxin on the release of prostaglandin-like substances (PLS) from rabbit ileum. Perfusion of ileal loops in vivo with buffer containing crude toxin was followed by a release of PLS into the perfusate, in amounts up to 37.5 ng/30 min (PGE₂ equivalents). In contrast, no detectable PLS was released when ileal loops were perfused with pure toxin. Similarly, pieces of ileum opened longitudinally released PLS in amounts up to 107 ng PGE₂/g tissue when incubated with crude toxin for 1–4 hr, but no release of PLS was detected in the presence of pure toxin under comparable conditions.Treatment of rabbits with indomethacin, 1.6 mg/kg p.o., had no effect on the accumulation of fluid in ileal sacs injected with crude or pure cholera toxin. These results support the view that prostaglandins do not play an essential role in the action of cholera toxin.     

Production of prostaglandin E2 by bladder tumor cells in tissue culture and a possible mechanism of lymphocyte inhibition.

Human bladder tumor cell lines were found to produce and secrete prostaglandin E₂ (PGE₂) in tissue culture and to be inhibited in this activity by prostaglandin synthetase inhibitors. The addition of purified peripheral blood lymphocytes from normal human donors to the tumor cells in confluent monolayers enhanced the production of PGE₂ by the tumor cells. At concentrations similar to those produced by the tumor cells, PGE₂ induced the intracellular accumulation of cyclic adenosine 3′, 5′-monophosphate by the lymphocytes. Both natural and antibody-dependent lymphocyte cytotoxicities by purified peripheral human blood lymphocytes directed against the same tumor target cells were inhibited by prostaglandins E₁ and E₂ and enhanced by the presence of indomethacin during incubation. Taken together, these phenomena suggest the possible existence of a mechanism whereby tumor cells, by the production of prostaglandins, may subvert the cellular immune response mounted against them and defend themselves from lymphocyte attack.     

Comparison between the effect of luteinizing hormone and prostaglandin E1 on ovarian cyclic AMP

Isolated whole ovaries from 23–24 day-old rats were studied in order to compare the effects of prostaglandin E₁ (PGE₁) and luteinizing hormone (LH) on ovarian cyclic adenosine 3′,5′-monophosphate (cAMP) production. Both substances produced a dose-dependent accumulation of cAMP in the ovarian tissue as well as in the incubation medium. The release of cAMP to the incubation medium was considerable after long periods of incubation (60–120 min). Time-relationships for LH- and PGE₁-effects were different. Maximal cAMP content in the tissue after addition of PGE₁ was seen already after 5–15 min of incubation whereas LH gave a maximal response after around 60 min. Accumulation of cAMP in the medium was approximately linear with time for both LH and PGE₁. Addition of theophylline potentiated the action of PGE₁ and LH but did not change the time-courses of the effects. It is concluded that the accumulation of cAMP in the medium should be considered in studies with various in vitro types of ovarian preparations. It is also pointed out that the different time-courses of the LH- and PGE₁-effects make the interpretation of additivity experiments difficult.     

Inhibition of ornithine decarboxylase and S-adenosylmethionine decarboxylase activities of S49 lymphoma cells by agents increasing cyclic AMP

We have examined the regulation of two key enzymes that control polyamine biosynthesis-L-ornithine decarboxylase (ODC) and S-adenosylmethionine decarboxylase (SAMDC) - by agents increasing cAMP in S49 lymphoma cells. Incubation of wild type S49 cells with beta-adrenergic agonists (terbutaline or isoproterenol) inhibited ODC and SAMDC activities rapidly (less than 2 hr). more quickly than these agents arrested the cells in the G1 phase of the cell cycle. The beta-adrenergic antagonist propranolol blocked inhibition of ODC activity produced by isoproterenol, but only if added simultaneously or less than 4 hr after the agonist. Incubation of wild type S49 cells with cholera toxin or PGE1 also inhibited ODC activity. Decreases in ODC activity produced by beta-adrenergic agonists, cholera toxin, PGE1 or dibutyryl cAMP were all enhanced by the phosphodiesterase inhibitor Ro 20-1724. Results of studies of ODC and SAMDC activity in S49 variants having lesions in the pathway of cAMP generation and action were as follows: kin- cells (which lack cAMP-dependent protein kinase activity) showed no inhibition of ODC by any agent; AC- cells (which have absent nucleotide coupling units in their adenylate cyclase system) only demonstrated inhibition in response to dibutyryl cAMP; UNC cells (which have deficient coupling of hormone receptors and adenylate cyclase) only demonstrated inhibition in response to dibutyryl cAMP and cholera toxin, and beta-depleted cells (which have a decreased number of beta-adrenergic receptors) responded as did wild type cells except for absent response to isoproterenol. We conclude that inhibition of ODC and SAMDC activity in S49 cells is an early response to agents that increase cAMP and that this action occurs via the "classical" pathways of activation of adenylate cyclase and protein kinase. These results in S49 cells contrast with evidence in other systems in which cAMP has been suggested to enhance polyamine biosynthesis, perhaps through alternative mechanisms.     

Studies of microbial toxins in Xenopus laevis oocytes

Xenopus laevis oocytes have been incubated or microinjected with cholera and diphtheria holotoxins or their respective isolated fragments A and B. Effects on progesterone-induced maturation, protein synthesis and cAMP levels were observed. Xenopus laevis oocytes were highly susceptible to cholera toxin upon incubation as evidenced by the increase of cAMP (two-fold increase in cAMP with 0.1 nM cholera toxin) and the blockade of progesterone-induced maturation. When isolated cholera toxin fragments A or B were incubated with oocytes, no activity could be detected. However, microinjection of cholera toxin fragment A into oocyte was able to mimic the effects of incubated holotoxin. Microinjection of cholera toxin B fragment was only effective at very high concentrations, probably due to trace contaminations by the A fragment. On the other hand, Xenopus laevis oocytes were very resistant to diphtheria toxin action upon incubation, a result attributable to lack of specific membrane receptors since, after microinjection of diphtheria toxin A fragment into oocytes, inhibition of protein synthesis was demonstrated. By simultaneous microinjection of highly radioactive adenine-labelled NAD and diphtheria toxin fragment A into oocytes, radioactive ADP ribosylation of the elongation factor 2 (EF₂) was observed. It is proposed that Xenopus laevis oocytes provide a new experimental approach for studying the mechanisms of action of microbial toxins.     

Prostaglandin E2 Receptors of Monocytes/Macrophages: Regulation by Insulin and Interleukin-1α

The regulation of receptors for prostaglandin E₂ (PGE₂) in monocyte/macrophage-like cells, P388D₁, by interleukin-1α (IL-1α) and insulin has been investigated. Many of the effects of IL-1, such as fever and other inflammatory activities, are linked to the stimulation of PGE₂ synthesis. On the other hand, PGE₂ exhibits suppressive effects on many steps in the immune response, including IL-1 production. The binding of PGE₂ to monocytes is reported to be essential for the inhibition of IL-1 production and activity. This inhibition occurs through the stimulation of cyclic AMP synthesis by the activation of PGE₂ receptor-linked adenylate cyclase. Although IL-1α stimulates PGE₂ synthesis in monocytes/macrophages during immunoactivation, it inhibits the binding of PGE₂ to these cells and may thereby exert a countervailing effect on the immunosuppressive action of this prostanoid. By contrast, insulin at physiological concentrations enhances the PGE₂ binding to these cells. This suggests that insulin at physiological concentrations may enhance the immunosuppressive action of PGE₂. Since the stimulation of cAMP synthesis in cells is regulated by PGE₂ binding, it is possible that these hormonal factors may control the immune response by modulating the PGE₂ receptor activity of monocytes/macrophages. This article focuses on the interactions of insulin and IL-1 with PGE₂ receptors of monocytes/macrophages.     

Augmentation of prostaglandin and thromboxane production in vitro by monocytes exposed to histamine-induced suppressor factor (HSF)

A possible mechanism to explain the suppression of mitogen-induced lymphocyte proliferation in vitro by histamine-stimulated mononuclear cells was investigated. In initial experiments, the inhibitory action of histamine-induced suppressor factor (HSF) on lymphocyte proliferation was documented to be reduced by the addition of indomethacin (1 μg/ml). Moreover, the addition of exogeneous PGE₂ (10^?7-10^?8 M) to mononuclear cell cultures reconstituted HSF activity in the presence of indomethacin. In order to ascertain the nature of the target cell responding to HSF, control and suppressor supernatants were incubated with human lymphocytes or monocytes (5 × 10⁶ cells/ml) for 24 hr. Following incubation, the supernatants were assayed for their content of prostaglandin E₂, F_2α, and thromboxane B₂. Monocytes (but not lymphocytes) incubated with supernatants containing HSF increased their production of prostaglandin E₂, F_2α, and thromboxane B₂ by 169, 53, and 49%, respectively. Suppressor supernatants were generated with histamine or an H-2 agonist (dimaprit) and chromatographed by gel filtration on Sephadex G-100. The elution profiles for the factor(s) inducing suppression of lymphocyte proliferation (25–40,000 daltons) and augmenting PGE₂ production (25,000 daltons) overlapped but were not identical. Collectively, these data suggest that HSF-mediated inhibition of lymphocyte proliferation may occur in part through the augmented production of prostaglandins and/or thromboxane B₂ by human monocytes.     

Regulation by PGE2 of the production of oxygen intermediates by LPS-activated macrophages

The regulation by prostaglandin E2 (PGE2) of production of oxygen radicals by bacterial lipopolysaccharide-(LPS) activated macrophages was studied in vitro. A 48-hr incubation of murine thioglycollate-elicited macrophages with LPS (0.1 micrograms/ml) resulted in an enhanced ability of these cells to produce oxygen radicals when challenged with phorbol myristate acetate (PMA). Macrophages incubated for 48 hr without LPS did not produce measurable amounts of oxygen radicals when exposed to this triggering stimulus. Thus, PMA-triggered production of oxygen radicals was the result of macrophage activation by LPS. The PMA-triggered production of oxygen radicals by the LPS-activated macrophages was inhibited when PGE2 (10(-5) to 10(-9) M) was present during the incubation with LPS. Inhibition by PGE2 occurred during the early stages of macrophage activation, since the addition of PGE2 24 hr after LPS no longer inhibited the production of oxygen radicals by the macrophages. This inhibitory effect of PGE2 on the LPS-induced activation of macrophages could be reproduced by cyclic-adenosine-monophosphate (cAMP) agonists, such as isoproterenol and cholera toxin as well as by the cAMP analog dibutyryl-cAMP, suggesting a cAMP-mediated mechanism for the inhibitory effect of PGE2 on macrophage activation by LPS. Previous reports have implicated prostaglandins as mediators of destructive processes associated with chronic inflammation. Our findings suggest that PGE2 may, on the other hand, reduce tissue damage in a chronic inflammatory site by inhibiting the production of oxygen radicals by macrophages activated in the sera.