Blocking EP4 down-regulates tumor metabolism and synergizes with anti-PD-1 therapy to activate natural killer cells in a lung adenocarcinoma model.

Prostaglandin E2 (PGE2), a product of the cyclooxygenase (COX) pathway, is produced by tumors and surrounding stromal cells. It stimulates tumor progression, promotes angiogenesis and suppresses the anti-tumor response. Pharmacological inhibition of PGE2 synthesis has been shown to suppress tumor initiation and growth in vivo. In the current study, we demonstrated that the growth of the Ptgs2-deficient 3LL lung adenocarcinoma cell line was down-regulated in vivo through natural killer (NK) cell activation and a reduction in the population of polymorphonuclear leukocyte-myeloid-derived suppressor cells (PMN-MDSCs) and tumor-associated macrophages (TAMs). On the basis of these results, the therapeutic effect of ONO-AE3-208 (EP4i), an inhibitor of EP4 (a PGE2 receptor), combined with anti-PD-1 antibody was evaluated. EP4i, but not anti-PD-1 antibody, decreased tumor metabolism including glycolysis, fatty acid oxidation and oxidative phosphorylation. EP4i induced IFNγ production from only NK cells (not from T cells) and a shift from M2-like to M1-like macrophages in TAMs. These effects were further enhanced by anti-PD-1 antibody treatment. Although CD8 T-cell infiltration was increased, IFNγ production was not significantly altered, even with combination therapy. Tumor hypoxia was ameliorated by either EP4i or anti-PD-1 antibody treatment, which was further affected by the combination. Normalization of tumor vessels was significant only for the combination therapy. The results indicated a novel effect of EP4i for the metabolic reprogramming of tumors and revealed unique features of EP4i that can synergize with anti-PD-1 antibody to promote IFNγ production by NK cells, polarize TAMs into the M1 phenotype, and reduce hypoxia through normalization of the tumor vasculature.

Possible role of LECT2 as an intrinsic regulatory factor in SEA-induced toxicity in d-galactosamine-sensitized mice.

To elucidate whether leukocyte cell-derived chemotaxin 2 (LECT2) controls the progression of staphylococcal enterotoxin A (SEA)-induced toxicity, we examined the role of LECT2 in a mouse model. Almost all the C57BL/6J (B6) mice survived for 72 h after the injection of 0.1 µg of SEA and 20 mg of d-galactosamine (d-GalN). However, the same treatment protocol in LECT2(-/-) mice produced a high lethality (~90%), severe hepatic apoptosis, and massive hepatic and pulmonary hemorrhage, similar to the situation observed in B6 mice treated with 1.0 µg SEA/d-GalN. The plasma LECT2 levels in B6 mice treated with 1.0 µg SEA/d-GalN were inversely correlated with the plasma cytokine levels and were associated with prognosis. LECT2 administration increased the survival of B6 mice and down-regulated TNF-α and IL-6. These results suggest the involvement of LECT2 in the regulation of fatal SEA-induced toxicity in d-GalN-sensitized mice.

Pathogenic importance of cysteinyl leukotrienes in development of gastric lesions induced by ischemia/reperfusion in mice.

We examined the role of cysteinyl leukotrienes (CysLTs) in the gastric ulcerogenic response to ischemia/reperfusion (I/R) in mice. Experiments were performed in male C57BL/6J mice after 18-h fasting. Under urethane anesthesia, the celiac artery was clamped for 30 min, and then reperfusion was achieved by removing the clamp. The stomach was examined for lesions 60 min thereafter. The severity of I/R-induced gastric damage was reduced by prior administration of pranlukast [CysLT receptor type 1 (CysLT(1)R) antagonist] as well as 1-[[5'-(3''-methoxy-4''-ethoxycarbonyl-oxyphenyl)-2',4'-pentadienoyl]aminoethyl]-4-diphenylmethoxypiperidine [TMK688; 5-lipoxygenase (5-LOX) inhibitor]. On the contrary, these lesions were markedly worsened by pretreatment with indomethacin, and this response was abrogated by the coadministration of TMK688 or pranlukast. The gene expression of CysLT(1)R but not 5-LOX was up-regulated in the stomach after I/R, but both expressions were increased under I/R in the presence of indomethacin. I/R slightly increased the mucosal CysLT content of the stomach, yet this increase was markedly enhanced when the animals were pretreated with indomethacin. The increased CysLT biosynthetic response to indomethacin during I/R was attenuated by TMK688. Indomethacin alone caused a slight increase of CysLT(1)R expression and markedly up-regulated 5-LOX expression in the stomach. We concluded that I/R up-regulated the expression of CysLT(1)R in the stomach; CysLTs play a role in the pathogenesis of I/R-induced gastric damage through the activation of CysLT(1)R; and the aggravation by indomethacin of these lesions may be brought about by the increase of CysLT production and the up-regulation of 5-LOX expression, in addition to the decreased prostaglandin production.

Impairment by 5-fluorouracil of the healing of gastric lesions in rats: effect of lafutidine, a histamine H2 receptor antagonist, mediated by capsaicin-sensitive afferent neurons.

We investigated the influence of 5-fluorouracil (5-FU), an anti-tumor agent, on the healing of gastric lesions generated by 0.6 M HCl in rats and the effect of lafutidine, a histamine H(2) receptor antagonist, on the impaired healing. Animals fasted for 18 h were given 1 ml of 0.6 M HCl p.o., fed normally from 1 h later, and killed 1-96 h thereafter. 5-FU was given i.v. twice, 1 h and 24 h after the HCl. The gastric lesions healed spontaneously within 96 h. Although it decreased acid secretion, 5-FU markedly delayed the healing. Lafutidine, but not cimetidine, given p.o. immediately after each dosing of 5-FU significantly reversed the delay in healing by 5-FU, and this effect was attenuated by the chemical ablation of capsaicin-sensitive afferent neurons. Capsaicin also significantly reversed the delay in healing. The mucosal application of 50 mM HCl did not affect gastric mucosal blood flow (GMBF) in the normal stomach, but significantly increased it in the stomach damaged by 0.6 M HCl. The increases in GMBF were attenuated by 5-FU; however, the co-administration of lafutidine significantly restored the response. In addition, 5-FU inhibited both cell proliferation and migration in rat gastric epithelial cells (RGM1) in vitro. These results suggest that 5-FU delayed the healing of gastric lesions generated by 0.6 M HCl, probably through the inhibition of cell migration and proliferation, as well as the impairment of GMBF, and lafutidine reversed the delay in healing, mainly through the amelioration of the GMBF response mediated by capsaicin-sensitive afferent neurons.

Pathogenic importance of pepsin in ischemia/reperfusion-induced gastric injury.

We investigated the role of pepsin in the development of ischemia/reperfusion (I/R)-induced gastric lesions in rats. Under urethane anesthesia, the pylorus was ligated, the celiac artery was clamped, and 1 ml of HCl (50-150 mM) was instilled in the stomach. Then, reperfusion was established 15 min later by removing the clamp, and 2 h later the stomach was assessed for gross mucosal damage. Pepstatin (a specific pepsin inhibitor) or pepsin was given i.g. after the pylorus was ligated while cimetidine, omeprazole, or atropine was given s.c. 30 min before the ligation. I/R produced hemorrhagic gastric injury, with a concomitant increase in the amount of pepsin secreted, and the degree of both these responses was dependent on the concentration of HCl. The formation of lesions by IR in the presence of 100 mM HCl was significantly prevented by atropine or bilateral vagotomy, but neither omeprazole nor cimetidine had any effect. Intragastric administration of pepstatin dose-dependently reduced the severity of the I/R-induced gastric lesions, the effect being significant even at 0.1 mg/kg, while that of pepsin markedly aggravated these lesions. The increased pepsin output during I/R was associated with luminal acid loss and significantly inhibited by bilateral vagotomy or pretreatment with atropine but not cimetidine or omeprazole, while pepstatin significantly inhibited the pepsin activity. In conclusion, we suggest that pepsin plays a pivotal role in the pathogenesis of I/R-induced gastric lesions, and pepsin secretion is increased during I/R, the process being associated with acid back-diffusion and mediated through a vagal-cholinergic pathway.

Dual action of nitric oxide in the pathogenesis of ischemia/reperfusion-induced mucosal injury in mouse stomach.

AIM: We investigated the roles of NO/NOS isoforms in the pathogenesis of ischemia/reperfusion (I/R)-induced gastric injury in mouse stomachs. METHODS: Under urethane anesthesia, the celiac artery was clamped, and then reperfusion was established 30 min later by removal of the clamp. After a 60-min reperfusion, the stomach was examined for macroscopic lesions. RESULTS: Following I/R, hemorrhagic lesions were generated in the mucosa, although ischemia alone caused no visible damage. Prior administration of L-NAME (a nonselective NOS inhibitor) significantly aggravated these lesions, in a L-arginine-inhibitable manner. By contrast, the selective iNOS inhibitor 1400W significantly prevented the occurrence of I/R-induced gastric lesions. The mucosal MPO activity was increased after I/R, and this response was enhanced and attenuated by prior administration of L-NAME and 1400W, respectively. Interestingly, the later treatment with L-NAME, given 10 min before reperfusion, significantly reduced the severity of the I/R-induced gastric damage, in a L-arginine-dependent manner. The expression of iNOS mRNA was up-regulated in the stomach following I/R, with an increase of mucosal NO content, and the NO production was significantly inhibited by both L-NAME and 1400W. CONCLUSION: Endogenous NO plays a dual role in the pathogenesis of IR-induced gastric damage; NO/cNOS is protective while NO/iNOS is proulcerogenic during I/R.

A Rat Model of Ischemic Enteritis: Pathogenic Importance of Enterobacteria, iNOS/NO, and COX-2/PGE2.

BACKGROUND: We introduced a rat model of ischemic enteritis and investigated the roles of enterobacteria, Nitric Oxide (NO), and Prostaglandins (PGs) in its pathogenesis. METHODS: Male rats were used after 18 h of fasting. Ischemic enteritis was induced by partial ligation of the superior mesenteric artery (SMA). Under ether anesthesia, SMA was isolated, and a stenosis was made by placing a needle (23 guage) on the vessel and ligating both the vessel and needle, and then a needle was removed from the ligature. Animals were then fed normally after surgery. Various drugs such as antibiotics, cyclooxygenase (COX) inhibitors, NO synthase (NOS) inhibitors and PGE2 were administered for 2 days after surgery. RESULTS: Stenosis of the SMA caused mucosal ischemia and damaged the small intestine, particularly the ileum, within 3 days. The development of enteritis was accompanied by mucosal invasion of enterobacteria, with the bacterial count being significantly increased 8 h after surgery. The severity of enteritis was prevented by the prior administration of ampicillin, L-NAME, or aminoguanidine, but aggravated by that of indomethacin or rofecoxib. The deleterious effects of indomethacin were antagonized by the co-administration of PGE2; these effects were mimicked by AE1-329, an EP4 agonist, and abrogated by AE3-208, an EP4 antagonist. The expression of iNOS and COX-2 was up-regulated in the small intestine in a time-dependent manner after ischemia caused by stenosis of the SMA, with increases in the mucosal contents of NO and PGE2. CONCLUSION: These results suggest that enterobacteria played a major pathogenic role in this model of ischemic enteritis, and that iNOS/NO was deleterious in the pathogenesis of these lesions, while COX-2/PGE2 prevented the development of ischemic enteritis by activating EP4 receptors.

A novel model of ischemic enteritis induced in rats by stenosis of the superior mesenteric artery.

AIMS: We established a new model of ischemic enteritis in rats and evaluated its usefulness for screening prophylactic drugs. MAIN METHODS: Male SD rats were used after 18 h of fasting. Under ether anesthesia, the superior mesenteric artery (SMA) was exposed, and a calibrated stenosis was produced by placing a needle on a blood vessel, ligating both the vessel and needle, and then removing the needle from the ligature. KEY FINDINGS: The stenosis caused severe damage on the anti-mesenteric side of the small intestine within 3 days; the severity of the damage increased with the gauge of a needle. No damage occurred in the small intestine following the stenosis with a needle of less than 21 gauge. Multiple hemorrhagic lesions occurred at an incidence of 100% when a 23-gauge needle was used. The development of enteritis was accompanied by enterobacterial invasion in the mucosa, with an up- regulation of inducible nitric oxide synthase (iNOS) expression and nitric oxide (NO) production. The ischemia-induced enteritis was significantly prevented by repeated treatment with aminoguanidine (a selective iNOS inhibitor), L-NAME (a nonselective NOS inhibitor), ampicillin, and aztreonam (a gram- negative bacterium antibiotic), but not vancomycin (a gram-positive bacterium antibiotic). SIGNIFICANCE: These results showed that a novel model of ischemic enteritis is induced in rats by stenosis of the SMA, this model may be useful for screening drugs against ischemic enteritis, and gram-negative bacteria as well as iNOS/NO are involved in the pathogenesis of enteritis in this model.

A novel model of ischemic enteritis induced in rats by stenosis of the superior mesenteric artery.

AIMS: We established a new model of ischemic enteritis in rats and evaluated its usefulness for screening prophylactic drugs. MAIN METHODS: Male SD rats were used after 18 h of fasting. Under ether anesthesia, the superior mesenteric artery (SMA) was exposed, and a calibrated stenosis was produced by placing a needle on a blood vessel, ligating both the vessel and needle, and then removing the needle from the ligature. KEY FINDINGS: The stenosis caused severe damage on the anti-mesenteric side of the small intestine within 3 days; the severity of the damage increased with the gauge of a needle. No damage occurred in the small intestine following the stenosis with a needle of less than 21 gauge. Multiple hemorrhagic lesions occurred at an incidence of 100% when a 23-gauge needle was used. The development of enteritis was accompanied by enterobacterial invasion in the mucosa, with an up-regulation of inducible nitric oxide synthase (iNOS) expression and nitric oxide (NO) production. The ischemia-induced enteritis was significantly prevented by repeated treatment with aminoguanidine (a selective iNOS inhibitor), L-NAME (a nonselective NOS inhibitor), ampicillin, and aztreonam (a gram-negative bacterium antibiotic), but not vancomycin (a gram-positive bacterium antibiotic). SIGNIFICANCE: These results showed that a novel model of ischemic enteritis is induced in rats by stenosis of the SMA, this model may be useful for screening drugs against ischemic enteritis, and gram-negative bacteria as well as iNOS/NO are involved in the pathogenesis of enteritis in this model.

Roles of cyclooxygenase-2 and prostacyclin/IP receptors in mucosal defense against ischemia/reperfusion injury in mouse stomach.

We examined the roles of cyclooxygenase (COX) isozymes, prostaglandins (PGs), and their receptors in the mucosal defense against ischemia/reperfusion (I/R)-induced gastric lesions in mice. Male C57BL/6 mice, including wild-type animals and those lacking prostaglandin E(2) (EP)1, EP3, or prostaglandin I(2) (IP) receptors, were used after 18 h of fasting. Under urethane anesthesia, the celiac artery was clamped (ischemia) for 30 min, and then reperfusion was achieved for 60 min through the removal of the clamp, and the stomach was examined for lesions. I/R produced hemorrhagic gastric lesions in wild-type mice. The severity of lesions was significantly increased by pretreatment with indomethacin (a nonselective COX inhibitor) and rofecoxib (a selective COX-2 inhibitor) but not 5-(4-chlorophenyl)-1-(4-methoxyphenyl)-3-(trifluoromethyl)-1H-pyrazole (SC-560; a selective COX-1 inhibitor). The expression of COX-2 mRNA was up-regulated in the stomach following I/R but not by sham operation or ischemia alone. The ulcerogenic response was markedly aggravated in IP receptor knockout mice but not those lacking EP1 or EP3 receptors. I/R increased the levels of 6-keto-PGF(1alpha) and PGE(2) in the stomach of wild-type mice, and this response was attenuated by indomethacin and rofecoxib but not SC-560. Pretreatment of wild-type mice with iloprost, a prostacyclin (PGI(2)) analog, significantly prevented the I/R-induced gastric lesions in the absence and presence of indomethacin or rofecoxib. PGE(2) also reduced the severity of I/R-induced gastric lesions, yet the effect was much less pronounced than that of iloprost. These results suggest that endogenous PGs derived from COX-2 play a crucial role in gastric mucosal defense during I/R, and this action is mainly mediated by PGI(2) through the activation of IP receptors.

A novel gastric lesion model induced in rats by partial gastric vascular ligation.

Previous studies have suggested that histamine treatment after gastric vascular ligation induces mucosal damage in the rat stomach. Although ligation of left gastric artery and vein (L-AV) alone did not cause any damage in the stomach within 4 h, but provoked mild lesions due to ischaemia 24 h later. In the present study we demonstrated a new model of gastric lesions induced by L-AV ligation and examined the effects of various anti-ulcer drugs on this lesion model. The gastric lesions induced by L-AV ligation occurred at the corpus and antrum, especially at the corpus-antrum border, when examined 24 h later, and the severity of damage reached maximum 3 days after L-AV ligation. Repeated treatment with omeprazole or sucralfate for 3 days significantly prevented the development of gastric lesions induced by L-AV ligation, in whole mucosa, including the antrum. By contrast, famotidine given for 3 days showed a significant protection against total lesions in the whole mucosa, but had no effect on the antral lesions. Both omeprazole and famotidine dose-dependently decreased gastric acid output while sucralfate raised the intraluminal pH due to the acid-neutralizing action. These results suggest that the pathogenesis of gastric lesions induced by L-AV ligation differs depending on the region, the corpus and the antrum, and the lesions occurred in the latter area seem to be resistant to acid suppression. It is assumed that this new model of gastric lesions is useful for screening the drugs that affect gastric mucosal defense rather than acid secretion.

Biphasic effect of prostaglandin E2 in a rat model of esophagitis mediated by EP1 receptors: relation to pepsin secretion.

We investigated the roles of prostaglandin (PG) E2 and cyclooxygenase (COX) isoenzymes in the mucosal defense of the esophagus, using subtype-selective EP agonists and antagonists as well as various COX inhibitors, in an acute rat esophagitis model. The animals were used after fasting for 18 h. Acid reflux esophagitis was induced by ligating both the pylorus and the transitional region between the forestomach and the glandular portion under ether anesthesia, and the damage was examined 3 or 4 h later. The esophageal lesions were significantly aggravated by prior administration of indomethacin and SC-560 (a selective COX-1 inhibitor) but not rofecoxib (a selective COX-2 inhibitor). PGE2 prevented these lesions at lower doses, yet the protective effect disappeared at a high dose. This biphasic effect was mimicked by 17-phenyl PGE2 (EP1 agonist) and antagonized by ONO-AE-829 (EP1 antagonist), while neither EP2, EP3, nor EP4 agonists had any effect on the esophageal lesions. PGE2 and 17-phenyl PGE2 had no effect on the acid secretion, but significantly increased the pepsin secretion, in a dose-dependent manner. The development of the esophageal lesions was totally prevented by pepstatin, a specific inhibitor of pepsin, and markedly aggravated by exogenous pepsin. We conclude that endogenous PGs derived from COX-1 are involved in the mucosal defense of the esophagus and that PGE2 has a biphasic influence on esophageal injury, depending on the dose: a protective effect at low doses and a deleterious effect at high doses, both mediated by EP1 receptors--the latter effect of PGE2 may be brought about by stimulation of the pepsin secretion.

Stimulation by capsaicin of duodenal HCO3(-) secretion via afferent neurons and vanilloid receptors in rats: comparison with acid-induced HCO3(-) response.

We compared the HCO3(-) secretory response to capsaicin and mucosal acidification in rat duodenums, especially the relation to vanilloid receptor type 1 (VR1). A proximal duodenal loop was perfused with saline, and the HCO3(-) secretion was measured at pH 7.0 using a pH-stat method and by adding 10 mM HCl. The secretion was stimulated by exposing the loop to capsaicin (0.03-0.3 mg/ml) or 10 mM HCl for 10 min. Indomethacin subcutaneously or ruthenium red intravenously, a nonspecific VR1 antagonist, was given 60 or 10 min, respectively, before exposure to capsaicin or acid, while L-NAME was given intravenously 3 hr before these treatments. Capsazepine, another VR1 antagonist, was coapplied to the loop for 10 min with capsaicin or acid. Luminal application of capsaicin increased the secretion of HCO3(-) in a dose-dependent manner; this effect was markedly attenuated by chemical ablation of capsaicin-sensitive afferent neurons (CSN) as well as pretreatment with ruthenium red or capsazepine, and significantly mitigated by indomethacin or L-NAME (in an L-arginine-sensitive manner). The HCO3(-) secretion was also stimulated by mucosal acidification, and this response was attenuated by both capsaicin pretreatment, indomethacin and L-NAME, but not ruthenium red or capsazepine. Mucosal application of capsaicin as well as acid increased the mucosal PGE2 content, and these effects were both significantly attenuated by indomethacin and L-NAME. These results suggest that both capsaicin and acid cause the CSN-dependent increase in duodenal HCO3(-) secretion mediated by NO and PG, yet the mode of their action differs in terms of the ruthenium red or capsazepine sensitivity. Although luminal H+ plays a modulatory role in duodenal HCO3(-) secretion, it is unlikely that the action results from the interaction of H+ with the ruthenium red- or capsazepine-sensitive site of VR1.