Significant antitumor effect of a synthetic lipid A analogue,DT-5461, on murine syngeneic tumor models

Summary The antitumor effect of a synthetic lipid A analogue, DT-5461, was investigated using syngeneic tumor models in mice. Intravenous injection of DT-5461 into mice transplanted with solid tumors of MethA fibrosarcoma, MH134 hepatoma, MM46 mammary carcinoma, Lewis lung carcinoma (3LL), and colon adenocarcinomas 26 and 38 resulted in significant reductions in the weight of all tumors except Colon 26, with marked hemorrhagic necrosis of tumor tissues. Efficacy was almost equal to that of anEscherichia coli-type synthetic lipid A (compound 506), and also to those of some chemotherapeutics including Adriamycin, mitomycin C, fluorouracil and cisplatin. Furthermore, DT-5461 was more effective than other immunotherapeutics, including picibanil (OK-432) and lentinan. However, its antitumor effects were inferior to those of Adriamycin or OK-432 against the malignant ascites caused by intraperitoneal inoculation with MethA or with MH134 cells; life span was not prolonged by either intraperitoneal or intravenous administration. In addition, although DT-5461 showed direct inhibitory effects on the in vitro growth of MethA or MH134, these were much weaker than those of Adriamycin. These findings clearly indicated that DT-5461 with systemic administration is a highly effective antitumor agent on solid tumors, and suggest that the antitumor effect of DT-5461 with potent necrotizing activity might derive from indirect mechanisms related to the activation of host immune systems and not to the weak direct cytotoxicity.     

Parallel antitumor,granuloma-forming and tumor-necrosis-factor-priming activities of mycoloyl glycolipids fromNocardia rubra that differ in carbohydrate moiety: Structure—activity relationships

Summary Multiple intravenous injections (30 µg, ten times) in ICR mice of trehalose dimycolate and glucose monomycolate fromNocardia rubra, containing C_36–48 mycolic acids, showed a prominent antitumor effect on a subcutaneously implanted sarcoma-180, an allogeneic sarcoma of mice with a significant granuloma formation in lungs, spleen and liver. On the other hand, mycoloyl glycolipids other than glucose monomycolate and trehalose dimycolate, such as mannose or fructose mycolate, showed no significant activity for tumor regression or granuloma formation in mice.Trehalose dimycolate and glucose monomycolate fromN. rubra, and glucose monomycolate with C_56–60 mycolic acids fromRhodococcus terrae also showed a distinctive priming activity for tumor necrosis factor (TNF), when lipopolysaccharide fromEscherichia coli was administered as an eliciting agent. The TNF activity in the sera of mice was abrogated almost completely by anti-(murine TNF) antibody with protein-A—agarose. Again in contrast, mannose and fructose mycolate fromN. rubra and glucose monomycolate with C_30–34 mycolic acids fromRhodococcus equi did not show such activities in mice.Meth-A, a syngeneic fibrosarcoma of BALB/c mice, was less sensitive to administration of glycolipids than sarcoma-180. These results indicated that the existence of a glucose or trehalose molecule was necessary for the expression of immunomodifying activities among various mycoloyl glycolipids differing in carbohydrate structure. However, since the administration of lipopolysaccharide was essentially required as an eliciting agent for the induction of TNF, while no eliciting agent was required for the antitumor activities, TNF does not seem to contribute directly to the antitumor activities of mycoloyl glycolipids in our systems. There was, however, a parallel structure-activity relationship among granuloma-forming, antitumor and TNF-priming activities, indicating that the structures of both the carbohydrate moiety and the mycoloyl residues influenced an initial step, such as macrophage activation, commonly and profoundly.     

Effects of intracerebroventricular administration of PGE1, E2 and F2α on electrically induced convulsions in mice

Intracerebroventricular administration of prostaglandins E₁ or E₂ was shown to block, while PGF_2α increased the incidence of tonic convulsion due to electroshock in mice. The Prostaglandins were administered intracerebroventricularly (i.c.v.) to conscious mice by a modification of Haley and McCormick's method (1) prior to a transcorneal maximal electroshock (MES) or a transcorneal supra-maximal electroshock (SMES). PGE₁ and PGE₂ i.c.v. blocked the tonic hindlimb extension (THE) and protected the animals from death induced by MES with ED₅₀'s for PGE₁ and PGE₂ for inhibition of the THE of 6.6 (4.3–12.0) μg/mouse i.c.v. and 13.3 (8.9–22.4) μg/mouse i.c.v. respectively. When PGE₂ was administered intraperitoneally (i.p.) in doses as high as 4.0 mg/kg it did not block the THE. However, the duration of the THE as well as the mortality were reduced by doses of 0.5–4.0 mg/kg PGE₂ i.p.. Both PGE₁ and PGE₂ were shown to cause a dose related significant (p<.001) decrease in the duration of the THE with SMES in doses of 1–10 μg/mouse i.c.v. for PGE₁ and 2–40 μg/mouse i.c.v. for PGE₂. PGF_2α, administered i.c.v. prior to a transcorneal electroshock equivalent to a current at the ED₁ level, increased the incidence of the THE as well as the mortality in doses of 20–50 μg/mouse.     

The in vitro effect of indomethacin on basal bone resorption, on prostaglandin production and on the response to added prostaglandins

Mouse calvaria were maintained in organ culture without serum additives. Basal active resorption, as measured by ⁴⁵Ca and hydroxyproline release, was significantly inhibited to 74% control levels by indomethacin (1.4 × 10⁻⁷ M). Prostaglandin F and prostaglandin E₂ production, determined by radioimmunoassay, were both significantly lowered by this concentration of indomethacin. DNA, protein and hydroxyproline synthesis, as indices of cell toxicity, were unaffected by low concentrations of indomethacin, while concentrations of 1.4 × 10⁻⁶M inhibited protein synthesis (p<0.005). In the presence of indomethacin (1.4 × 10⁻⁷M) both PGE₂ and PGF_2α stimulated resorption in a dose-dependent manner, with PGE₂ being the more potent. Neither prostaglandin affected hydroxyproline synthesis at low concentrations, but PGE₂ had a marked inhibitory action at a higher concentration (10⁻⁶M). In combination, the effects of PGE₂ and PGF_2α showed no evidence of synergism or any antagonistic action. The study shows that in vitro calcium and hydroxyproline resorption in the unstimulated mouse calvaria are inhibited by indomethacin at concentrations measured in serum during human therapy. The decreased PGF and PGE₂ production associated with this decreased bone resorption in the presence of non-toxic concentrations of indomethacin would suggest a role for these prostaglandins in maintaining the basal resorption of cultured bone.     

Prostaglandin regulation of macrophage function: Effect of endogenous and exogenous prostaglandins

The expression of macrophage antitumor activity and the production of prostaglandins (PG) by operationally defined macrophage populations differed under varying culture conditions. Culture conditions that caused increased PGE₂ production by activated macrophages resulted in an inhibition of their tumoricidal activity. In contrast, production of high levels of PGE₂ by resident and elicited macrophages was associated with an increase in antitumor activity. The activation of resident or elicited cells by lipopolysaccharide (LPS) could be blocked by indomethacin. Treatment of these macrophages with PGE₂ alone also resulted in their activation and subsequent tumor cell destruction. Activation of resident and elicited macrophages by LPS appears to be mediated by PGE₂.     

Prostaglandin formation in bacteria

We report here that intact Gram positive and Gram negative bacteria, in the presence of exogenous arachidonic acid, produce and release PGE₂ and PGF_2α into the medium as measured by radioimmunoassay. The seven bacterial strains so far studied release 6.5–50.9 ng PGE₂ and less than 0.02–0.51 ng PGF_2α per mg bacterial protein during a 1 hour incubation, quantities of the same order of magnitude as those observed in mammalian systems. PGE₂ and PGF_2α formation in bacteria are inhibited by indomethacin.     

The role of prostaglandins in the excitatory innervation of the rat urinary bladder

The possible role of PGs in hyoscine-resistant nerve mediated responses of the rat urinary bladder was investigated. Responses to electrical stimulation were inhibited by cinchocaine (30 μmol/l) but were only partially inhibited by a high concentration of hyoscine (25 μmol/l) or by the choline uptake inhibitors, hemicholinium-3 (500 μmol/l) and troxypyrrolidinium (500 μmol/l). Indomethacin (50 μmol/l) produced partial blockade (30%) of responses to electrical stimulation without markedly affecting responses to acetylcholine and the degree of blockade was of a similar order in the presence of hyoscine or troxypyrrolidinium. PGE₂ (0.028 – 2.8 μmol/l) or F_2α (0.029 – 2.9 μmol/l) produced a slowly developing increase in tone and spontaneous activity. Responses to electrical stimulation were at most only slightly increased in the presence of either PG. However, the PGs always increased the responses to electrical stimulation after indomethacin, indomethacin plus hyoscine or indomethacin plus troxypyrrolidinium. Responses to acetylcholine in the presence of indomethacin were not increased by PGE₂. It is concluded that PGE₂ and F_2α do not function as transmitters responsible for resistance to anti-muscarinic drugs in the bladder but may exert a modulating effect on nervous transmission.     

Modulation of prostaglandin of the renal vascular action of arginne vasopressin

To determine whether the renal vascular effect of arginine vasopressin (AVP) is modulated by renal prostaglandin E₂ (PGE₂) were determined during the infusion of AVP in dogs during control conditions and after the administration of the inhibitor of prostaglandin synthesis, indomethacin. During control conditions, intrarenal administration for 10 min of a dose of AVP calculated to increase arterial renal plasma AVP concentration by 75 pg/ml produced a slight renal vasodilation (p<0.01) and an increase in renal venous plasma concentration of PGE₂. Renal venous PGE₂ concentration during control and AVP infusion averaged 33 ± 7 and 52 ± 12 pg/ml (p<0.05), respectively. After administration of indomethacin, the same dose of AVP induced renal vasoconstriction (p<0.05) and failed to enhance renal venous PGE₂ concentration (9 ± 1 to 8 ± 1 pg/ml). Intrarenal administration of 20 ng/kg. min of AVP for 10 min induced a marked renal vasoconstriction (p<0.01) and increased renal venous plasma PGE₂. Renal venous PGE₂ during control and AVP infusion averaged 31 ± 10 and 121 ± 31 pg/ml (p<0.01), respectively. Administration of the same dose of AVP following indomethacin produced a significantly greater and longer lasting renal vasoconstriction (p<0.01) and failed to increase renal venous plasma PGE₂ (10 ± 1 to 9 ± 1 pg/ml). These results indicate that a concentration of AVP comparable to that observed in several pathophysiological conditions induces a slight renal vasodilation which is mediated by renal prostaglandins. The results also indicate that higher doses of AVP induce renal vasoconstriction and that prostaglandin synthesis induced by AVP attenautes the renal vasoconstriction produced by this peptide.     

Increased production of tumor necrosis factor and prostaglandin E2 by monocytes in cancer patients and its unique modulation by their plasma

Summary We investigated the role of monocytes in the production of tumor necrosis factor (TNF) and prostaglandin E₂ (PGE₂) in 77 cancer patients with malignancies of the digestive tract, using 30 normal individuals and 18 noncancer patients as controls. Monocytes were incubated with lipopolysaccharide for 20 h, and TNF production and PGE₂ production were analyzed by bioassays. Elevated levels of TNF (>512 U/ml) and PGE₂ (>8 ng/ml) production were demonstrated in many cancer patients when these factors were induced in the medium with 10% fetal bovine serum. The elevated level of TNF was seen to be restricted for the most part to patients with malignancies. Thus, 51 out of 59 cancer patients (86%), consisting of 44 primary cancer patients and 15 recurrent cancer patients, showed an increased level of TNF. In contrast, almost all of 18 postoperative cancer patients showed TNF levels comparable to those of normal individuals. Furthermore, 16 primary cancer patients were also demonstrated to have reduced levels of TNF production by monocytes after curative operation. When 10% cancer-patient plasma was added to the induction culture, TNF production by monocytes was drastically suppressed in the cancer patients. Interestingly, the same addition of plasma induced a prominent enhancement of PGE₂ production in the cancer patients. The plasma of noncancer patients did not modulate production of these factors. No TNF activity was found in the plasma of cancer patients, but such plasma did contain an increased level of PGE₂ (100–300 pg/ml). Although PGE₂ (>2ng/ml) was able to suppress TNF production by monocytes, the addition of 10% plasma PGE₂ was not enough to induce suppression. An unknown factor(s) in the plasma of cancer patients may uniquely modulate the elevated TNF and PGE₂ production in these patients.     

Prostaglandins and pacemaker activity in isolated guinea pig SA node

Prostaglandins PGE₂, PGE₁, PGF_2α, and PGA₁ substantially increase automaticity in SA-nodal, right atrial preparations excised from guinea pigs. This natural pacemaker tissue is sensitive to nanomolar doses of PG with, for example, 10⁻⁸ M PGE₂, increasing SA rate by about 20%. If these preparations are pretreated with 2 μM indomethacin, a blocker of endogenous prostaglandin synthesis, then spontaneous rate drops and subsequent rate increases due to PGE₂ administration can be more easily demonstrated. Guinea pig pacemaker tissue differs from similar rabbit tissue not only in that it is directly responsive to PGE₂, but also in that PGE₂ does not depress the absolute response to transmural stimulation (adrenergically mediated rate increase). The positive chronotropic responses to PGE₂ also occur when the guinea pig tissue is pretreated in 0.6 μM propranolol, which causes blockade of beta-adrenergic receptors.The pacemaker myocardium in the guinea pigs thus appears to be directly stimulated by exogenous PGE₂ at very low doses. The observation that 2 μM indomethacin reduces SA-nodal rate suggests the presence of a very sensitive, functionally important, PGE-like system which modulates heart rate in this mammalian species.     

Effects of indomethacin, prostaglandin E2, prostaglandin F2α and 6-keto-prostaglandin F1α on hatching of mouse blastocysts

The present study has been performed to investigate how PGs would participate the hatching process. Effects of indomethacin, an antagonist to PGs biosynthesis, on the hatching of mouse blastocysts were examined in vitro. Furthermore, it was studied that prostaglandin E₂ (PGE₂), prostaglandin F_2α (PGF_2α) or 6-keto-prostaglandin F_1α (6-keto-PGF_1α) were added to the culture media with indomethacin. (1) The hatching was inhibited by indomethacin yet the inhibition was reversible. (2) In the groups with indomethacin and PGE₂, no improvement was seen in the inhibition of hatching and the inhibition was irreversible. (3) In the groups with indomethacin and PGF_2α, inhibition of hatching was improved in comparison with the group with indomethacin. (4) In the groups with indomethacin and 6-keto-PGF_1α, no improvement was seen. The above results indicated that PGF_2α possibly had an accelerating effect on hatching and a high concentration of PGE₂ would exert cytotoxic effect on blastocysts.     

Prostaglandins VIII: A proposed role for PGE2 in the genesis of stress-induced gastric ulcers

A protective effect on stress-induced gastric ulcers has been d monstrated for PGE₂ and PGF_2α in rats. Pretreatment with indomethacin 5 mg/kg/per os increases the severity of gastric lesions, but this effect is completely antagonized by PGE₂ administration.     

A release of prostaglandins may be responsible for acute tolerance to norepinephrine infusions

A chick isolated rectum pretreated with atropine and indomethacin and superfused with the oxygenated mixed venous blood of anaesthetized cats, was selectively contracted by PGE₁ and PGE₂ at concentrations of <1 ng/ml. Intravenous infusion of norepinephrine (0.2 – 8.0 μg/kg/min) into the cats resulted in a contraction of the blood-bathed chick rectum. This was matched by contractions produced by PGE₂ (0.4 – 7 ng/ml) infused directly over the assay organ. The appearance of a chick rectum contracting substance in the venous blood was paralleled by a decline in the pressor response to norepinephrine. A single injection of indomethacin (3 – 10 mg/kg) prevented both the formation of the prostaglandin-like material and the acute tolerance to the pressor response to norepinephrine. Both effects could then be reproduced by an intra-arterial infusion of PGE₂ at a rate 0.125 – 0.5 μg/kg/min. β-Adrenoceptor blockade had no influence on the response of chick rectum and arterial blood pressure to an infusion of norepine phrine, but α-adrenoceptor blockade abolished both responses. It is postulated that the acute tolerance to norepinephrine infusions is the result of a release of PGE-like material from the contracting vascular bed.     

Effects of prostaglandins E2, I2 and F2α, arachidonic acid and indomethacin on pressor responses to norepinephrine in conscious rats

The effects of prostaglandins E₂ (PGE₂), I₂ (PGI₂) and F₂α (PGF₂α), arachidonic acid and indomethacin on pressor responses to norepinephrine were examined in conscious rats. Intravenously infused PGE₂ (0.3, 1.25 μg/kg/min), PGI₂ (50, 100 ng/kg/min), PGF₂α (1.8, 5.4 μg/kg/min) and arachidonic acid (0.7, 1.4 mg/kg/min) did not change the basal blood pressure. Both PGE₂ and PGI₂ significantly attenuated pressor responses to norepinephrine, whereas PGF₂α significantly potentiated them. Arachidonic acid, a precursor of the prostaglandins (PGs), significantly attenuated pressor responses to norepinephrine. Since the attenuating effect of arachidonic acid was completely abolished by the pretreatment with indomethacin (5 mg/kg), arachidonic acid is thought to exert an effect through its conversion to PGs. On the contrary, intravenously injected indomethacin (0.2–5.0 mg/kg) facilitated pressor responses to norepinephrine in a dose-related manner without any direct effect on the basal blood pressure. These results suggest that endogenous PGs may participate in the regulation of blood pressure by modulating pressor responses to norepinephrine in conscious rats.     

Combinatorial Effects of Nonsteroidal Anti-inflammatory Drugs and Food Constituents on Production of Prostaglandin E2 and Tumor Necrosis Factor-α in RAW264.7 Murine Macrophages

Combinatorial chemopreventive strategies, in contrast to those with individual agents, show potential in terms of potentially lower toxicity and higher efficacy. In this study, we combined several agents and examined their suppressive effects on the combined lipopolysaccharide (LPS)- and interferon(IFN)-γ-induced formation of proinflammatory mediators, including prostaglandin (PG) E₂ and tumor necrosis factor (TNF)-α, in RAW264.7 murine macrophages. The combinatorial effects of indomethacin/genistein (GEN) and aspirin/GEN were found to be synergistic for PGE₂ suppression, while the nimesulide/GEN combination was antagonistic. Further, while (?)-epigallocatechin gallate (EGCG) alone increased LPS/IFM-γ-induced production of PGE₂ and TNF-α as well as cyclooxygenase-2 expression, the EGCG/GEN combination markedly suppressed these parameters. Our results suggest that certain chemopreventive agents act complexly and that, when used in combination, they affect the intracellular signaling pathways of the paired agents to exert additive, synergistic, or antagonistic effects.     

Prostaglandin stimulation of ornithine decarboxylase activity in mammary gland explants from mid-pregnant mice

The effects of various prostaglandins on ornithine decarboxylase (ODC) activity in mammary gland explants from mid-pregnant mice have been tested. PGE₁, E₂ and I₂ elicit a concentration-dependent stimulation of ODC activity. The minimally effective concentrations are 0.5 ug/ml for PGE₁ and E₂, and 50 ug/ml for PGF_2α and 6-keto-PGF_1α. The PGE₁ effect had a time course identical to that of prolactin. The prolactin action on ODC activity was attentuated by indomethacin, an inhibitor of prostaglandin biosynthesis. Arachidonic acid stimulation ODC activity and its effect was abolished by indomethacin. The phosphodiesterase inhibitor, 3-isobutyl-1-methylxanthine, potentiated the PGE₁ effect on ODC activity. The results suggest that the prostaglandins may modulate prolactin's action of ODC activity via a cAMP dependent mechanism.     

Reversal of indomethacin blockade of ovulation in gilts by prostaglandins

Prepubertal gilts given 750 IU pregnant mares′ serum gonadotropin (PMSG) followed 72 h later by 500 IU human chorionic gonadotropin (hCG) to induce follicular growth and ovulation fail to ovulate when 10 mg/kg indomethacin (INDO) is injected 24 h after hCG administration. This study examines the effects of administration of exogenous prostaglandins F_2α and E₂ (PGF_2α and PGE₂) alone or in combination, and at various times prior to the expected time of ovulation, on the INDO blockade of ovulation in PMSG/hCG-treated gilts. Occurrence of ovulation was determined by visual observation at laparotomy 48 h after hCG. When 5 mg or 10 mg PGF_2α was injected at each of 38, 40 and 42 h after hCG injection, 63% and 79%, respectively, of preovulatory follicles ovulated. In contrast, injection of 5 mg PGE₂ or 5 mg PGE₂ plus 5 mg PGF_2α induced ovulation in 0% and 24% of preovulatory follicles, respectively. In control groups, 100% of folicles in PMSG/hCG-treated gilts ovulated whereas none did so in PMSG/hCG/INDO-treated animals. These results indicate that administration of PGF_2α can induce ovulation in the PMSG/hCG/INDO-treated prepubertal gilt and suggest that PGE₂ is ineffective and may be antagonistic to PGF_2α in overcoming the ovulation blocking effect of INDO.     

Indomethacin modulation of Adriamycin-induced effects on multiple cytolytic effector functions

Summary The anticancer agent, Adriamycin (ADM), in addition to being a potent cytotoxic drug has been shown to be an effective immunomodulator. This study was undertaken to determine whether ADM-induced changes in the production of prostaglandins (particularly PGE₂) are involved in ADM-associated modifications of selected host defenses. Spleen cells from normal or ADM-treated (5 mg/kg; day –5) C57BL/6 mice were assessed for the following activities: fresh (day 0) and cultured natural killer (NK), cytotoxic T lymphocyte, lymphokine-activated killer (LAK), Fc-dependent phagocytosis and tumoricidal macrophage. All activities were assessed with and without the addition of indomethacin, an inhibitor of the first step of the cyclo-oxygenase pathway of prostaglandin synthesis. Depending on culture conditions, the cytotoxic T lymphocyte and splenic tumoricidal macrophage activities were either unaffected or were augmented by ADM treatment of the spleen donor mice or by addition of indomethacin to the culture, and these effects were apparently independent of one another. In contrast, ADM treatment generally resulted in reduced NK and LAK activities relative to control and elevated Fc-dependent phagocytosis. The addition of indomethacin to the culture effectively reversed these effects. Furthermore, spleen cells from ADM-treated mice were found to produce twice the amount of PGE₂ in culture compared to cells from untreated mice. Finally, the direct addition of PGE₂ to NK cultures resulted in a dose-dependent inhibition of NK activity and the dose required was comparable to the amount of PGE₂ produced by cultured spleen cells from ADM-treated mice. Taken together, these results indicate that at least some of the immunomodulatory effects of ADM are an indirect result of ADM-induced changes in PGE₂ production.This work was supported in part by USPHS grants CA-28 835, CA-15 142, CA-24 538 and CA-09072 and from Veterans Administration Research Funds     

In vitro stimulation of prostaglandin synthesis in the rat pancreas by carbamylcholine, caerulein and secretin

Rat pancreas pieces spontaneously released PGE₂ (2.3 ng/100 mg × 45 min) and PGF_2α (7.6 ng/100 mg × 45 min). This release corresponds probably to a neo-synthesis since it was abolished by indomethacin. Carbamylcholine (≥ 10 μM), caerulein (≥ 10 nM) and secretin (≥ 10 nM) stimulated the release of PGE₂ and PGF_2α : the concentrations of stimulators required to increase PGs release were thus much higher than those which trigger enzyme secretion. Atropine specifically inhibited the cholinergic stimulation, whereas indomethacin blocked the stimulatory effects of all secretagogues. Stimulation of PGE₂ and PGF_2α release was reduced in a Ca⁺⁺-free medium, abolished by EGTA and mimicked by the ionophore A23187, underscoring the crucial role of Ca⁺⁺ in the regulation of PGs synthesis by the pancreas. Neither PGE₂ nor PGF_2α stimulated enzyme secretion in this system and indomethacin did not inhibit the secretory effect of carbamylcholine. Increased synthesis of prostaglandins in response to pancreatic secretagogues does not appear to be involved in the process of enzyme secretion.     

Direct Melanoma Cell Contact Induces Stromal Cell Autocrine Prostaglandin E2-EP4 Receptor Signaling That Drives Tumor Growth,Angiogenesis, and Metastasis

The stromal cells associated with tumors such as melanoma are significant determinants of tumor growth and metastasis. Using membrane-bound prostaglandin E synthase 1 (mPges1^−/−) mice, we show that prostaglandin E₂ (PGE₂) production by host tissues is critical for B16 melanoma growth, angiogenesis, and metastasis to both bone and soft tissues. Concomitant studies in vitro showed that PGE₂ production by fibroblasts is regulated by direct interaction with B16 cells. Autocrine activity of PGE₂ further regulates the production of angiogenic factors by fibroblasts, which are key to the vascularization of both primary and metastatic tumor growth. Similarly, cell-cell interactions between B16 cells and host osteoblasts modulate mPGES-1 activity and PGE₂ production by the osteoblasts. PGE₂, in turn, acts to stimulate receptor activator of NF-κB ligand expression, leading to osteoclast differentiation and bone erosion. Using eicosanoid receptor antagonists, we show that PGE₂ acts on osteoblasts and fibroblasts in the tumor microenvironment through the EP4 receptor. Metastatic tumor growth and vascularization in soft tissues was abrogated by an EP4 receptor antagonist. EP4-null Ptger4^−/− mice do not support B16 melanoma growth. In vitro, an EP4 receptor antagonist modulated PGE₂ effects on fibroblast production of angiogenic factors. Our data show that B16 melanoma cells directly influence host stromal cells to generate PGE₂ signals governing neoangiogenesis and metastatic growth in bone via osteoclast erosive activity as well as angiogenesis in soft tissue tumors.