Inhibition of prostaglandin E1-induced elevation of cytoplasmic free calcium ion by protein kinase C-activating phorbol esters and diacylglycerol in Swiss 3T3 fibroblasts

Cyclooxygenase and cytochrome P-450 pathways induced by fetal calf serum regulate wound closure in 3T6 fibroblast cultures through the effect of prostaglandin E2 and 12 and 20 hydroxyeicosatetraenoic acids

Wound-induced injury of 3T6 fibroblast cultures initiated a repair process stimulated by fetal calf serum (FCS) that restored the integrity of cell cultures. In these experimental conditions, FCS induced arachidonic acid (AA) release and eicosanoid production. Our results show that the inhibition of the cyclooxygenase (COX) and/or cytochrome P-450 pathways significantly decreases the wound closure, whereas that of the lipoxygenase pathway does not modify the wound repair process. Both EP(1) and EP(4) receptors of prostaglandin E(2) (PGE(2)) mediate PGE(2) stimulated 3T6 fibroblast wound closure. Our data suggest that calcium and cAMP are involved in the signaling event induced by PGE(2) during the 3T6 fibroblast wound repair process. On the other hand, we show that ketoconazole, a cytochrome P-450 inhibitor, hinders the wound closure induced by FCS in wounded 3T6 fibroblast cultures. 12 and 20 Hydroxyeicosatetraenoic acids (HETEs), which are key AA metabolites synthesized by cytochrome P-450, partially revert the effects of ketoconazole on the wound repair process. Thus, the COX and cytochrome P-450 pathways of the arachidonate cascade are involved in 3T6 fibroblast wound closure.

Role of EP(1) and EP(4) PGE(2) subtype receptors in serum-induced 3T6 fibroblast cycle progression and proliferation

Recent studies have suggested that prostaglandin E(2) (PGE(2)) subtype receptors (EP) are involved in cellular proliferation and tumor development. We studied the role of EP(1) and EP(4) PGE(2) subtype receptor antagonists AH-6809 and AH-23848B, respectively, in serum-induced 3T6 fibroblast proliferation. This was significantly reduced in a dose-dependent manner (IC(50) approximately 100 and approximately 30 microM, respectively) to an almost complete inhibition, without any cytotoxic effect. However, the effect of each antagonist on 3T6 cell cycle progression clearly differed. Whereas the EP(1) antagonist increased the G(0)/G(1) population, the EP(4) antagonist brought about an accumulation of cells in early S phase. These effects were associated with a decrease in cyclin D and E levels in AH-6809-treated 3T6 cells and lower cyclin A levels in AH-23848B-treated fibroblasts with respect to control cells. The G(0)/G(1) accumulation caused by AH-6809 seems to be intracellular Ca(2+) concentration ([Ca(2+)](i)) dependent, because a 6-h 1 microM thapsigargin treatment allowed G(0)/G(1)-arrested cells to enter S phase. Similarly, treatment with 20 microM forskolin for 6 h allowed S-phase and G(2)/M progression of AH-23848B-treated cells. This study shows that the inhibitory effect of the EP(1) and EP(4) antagonists on serum-induced 3T6 fibroblast growth is due to their effect at various levels of the cell cycle machinery, suggesting that PGE(2) interaction with its different subtype receptors regulates progression through the cell cycle by modulating cAMP and [Ca(2+)](i).

Protein kinase C plays a key role in the cross-talk between intracellular signalings via prostanoid receptors in a megakaryoblastic cell line, MEG-01s

In a previous study, we characterized prostanoid and thrombin receptors expressed on a megakaryoblastic cell line, MEG-01s (Blood 78, 2328-2336, 1991). In this study, we examines the mechanism of cross-talk between intracellular Ca2+ ([Ca2+]i) and cAMP signalings through prostanoid and thrombin receptors. Addition of a thromboxane (TX)A2 mimetic (U46619 or STA2) or thrombin stimulated the formation of inositol phosphates and dose-dependently augmented a prostaglandin (PG)I2 mimetic (iloprost)- or forskolin-induced cAMP formation. 12-O-tetradecanoylphorbol-13-acetate (TPA) and ionomycin, to lesser extent, also augmented iloprost-induced cAMP formation. The enhancing effect of U46619 or TPA on cAMP formation was inhibited by prolonged pretreatment of the cells with TPA (2.5 microM, 24 h), but not with calmodulin-antagonists; W-7, W-5, or KN-62. The elevation of [Ca2+]i induced by thrombin, STA2 or PGE2 was significantly suppressed by pretreatment of the cells with TPA (100 nM) as well as cAMP mimetics such as dibutyryl cAMP (5 mM), forskolin (5 microM) and iloprost (1 microM). These results suggest the key role of PKC on the cross-talk between [Ca2+]i and cAMP signalings through prostanoid and thrombin receptors; PKC, which is activated with TXA2 or thrombin, concomitantly suppress further [Ca2+]i elevation and enhances the PGI2 receptor-mediated cAMP formation, which, in turn, suppress [Ca2+]i elevation.

Estrogen induces expression of c-fos and c-jun via activation of protein kinase C in an endometrial cancer cell line and fibroblasts derived from human uterine endometrium

Endometrial fibroblasts derived from uterine endometrium as controls and endometrial cancer cell lines (Ishikawa and HHUA cells) were analyzed for the induction manner of c-fos and c-jun transcripts in endometrial cancers, some of which are estrogen-dependent in growth. Estrogen increased c-fos expression and protein kinase C (PKC) activity in fibroblasts and Ishikawa cells, but not in HHUA cells. Progesterone diminished c-fos and c-jun expression and PKC activity induced by estradiol in the fibroblasts, but not in Ishikawa cells, which persistently overexpressed c-fos and c-jun. In these cells, 12-0-tetra-decanoylphorbol-13-acetate (TPA) increased c-fos and c-jun expression as did estradiol. Pretreatment with 1-(5-isoquinolinesulfonyl)-2-methylpiperazine dihydrochloride (H-7) abolished estrogen-inducible over-expression of c-fos and c-jun. The combination of both estradiol and TPA at maximum effective concentration exerted no additive and synergistic effect on induction of c-fos and c-jun expression. In conclusion, persistent activation of PKC might lead to overexpression of c-fos and c-jun in some endometrial cancers with an estrogen predominant milieu, which might be, at least in part, associated with the transformation or growth potential.

Translocation of protein kinase C isoenzymes by elevated extracellular Ca2+ concentration in cells from a human giant cell tumor of bone

In this study we investigated the protein kinase C isoenzymes expressed by human osteoclast-like cells harvested from a giant cell tumor of bone (GCT23 cells), and by freshly isolated rat osteoclasts. Immunoblotting analysis revealed that the -alpha, -delta, and -epsilon, PKC isoforms, but not the -beta isoenzyme, are expressed by GCT23 cells. Immunofluorescence studies demonstrated that PKC-alpha, -delta, and -epsilon are homogeneously expressed by both mononuclear and multinucleated GCT23 cells, as well as by rat osteoclasts. Similar to authentic osteoclasts, GCT23 cells responded to an increase of extracellular Ca2+ concentration ([Ca2+]o) with a dose-dependent elevation of the cytosolic free Ca2+ concentration ([Ca2+]i). An increase of [Ca2+]o stimulated the translocation of PKC-alpha from the cytosolic to the particulate fraction, suggesting the involvement of this isoenzyme in the signal transduction mechanism prompted by stimulation of the [Ca2+]o sensing. By contrast, PKC-delta was not altered by exposure to elevated [Ca2+]o, whereas PKC-epsilon underwent reciprocal translocation, disappearing from the insoluble fraction and increasing in the cytosol. The effects of PKC on GCT23 cell functions were investigated by treatment with phorbol 12-myristate, 13-acetate (PMA). We observed that activation of PKC by PMA failed to affect adhesion onto the substrate, but down-regulated the [Ca2+]o-induced [Ca2+]i increases. The latter effect was specific, since it was reversed by treatment with the PKC inhibitors staurosporine and chelerythrine.

Characterization of the signal transduction of prostaglandin E receptor EP1 subtype in cDNA-transfected Chinese hamster ovary cells

We examined the signal transduction of mouse prostaglandin E receptor EP1 subtype using Chinese hamster ovary cells stably expressing the cloned EP1. Sulprostone, an EP1 agonist, induced a rapid increase in intracellular Ca2+ concentration in the EP1-expressing cells. Most of the increase was abolished by removal of extracellular Ca2+, and was insensitive to U-73122, a phospholipase C inhibitor. Sulprostone stimulated phosphatidylinositol hydrolysis, but this stimulation was abolished by removal of extracellular Ca2+, indicating that EP1-stimulated phosphatidylinositol hydrolysis is the result of extracellular Ca2+ influx. Thus, the signal transduction of EP1 is extracellular Ca2+ entry through a pathway independent of phospholipase C activation. We further examined the regulation of the signal transduction of EP1 having potential phosphorylation sites for either protein kinase C or protein kinase A. Short-term exposure of the cells to 12-O-tetradecanoylphorbol 13-acetate (TPA) completely suppressed the sulprostone-induced increase in intracellular Ca2+ concentration, while forskolin or dibutyryl cAMP did not affect it, suggesting that protein kinase C but not protein kinase A is involved in the regulation of the EP1 signal transduction. Furthermore, long-term exposure to TPA decreased PGE2 protein kinase A is involved in the regulation of the EP1 signal transduction. Furthermore, long-term exposure to TPA decreased PGE2 binding activity of EP1 due to the reduction of the EP1 mRNA level. Protein kinase C induces short- and long-term desensitization of EP1.

Prostaglandin E1 induces c-Fos and Myc proteins and protects rat hippocampal cells against hypoxic injury

We investigated the effects of prostaglandin (PG) E1 on the hypoxic injury of fetal rat hippocampal cells. Primary hippocampal cell cultures (embryonic day 18) were established and maintained. After 72 h in culture, PGE1 was added to the serum-free medium at a final concentration of 10(-5)-10(-9) M. Cultures were divided into two groups: The normoxia group was in culture for another 48 h, and the hypoxia group was exposed to 24 h of hypoxia followed by continuation of culture for another 24 h. As a quantitative measure of cell death, lactate dehydrogenase (LDH) activity was estimated in the culture medium. The LDH activity, released by the hypoxic insult, was significantly smaller with PGE1 treatment at 10(-6), 10(-7), and 10(-8) M (p < 0.01) and 10(-9) M (p < 0.05) compared with the control. No differences in the LDH activities were observed in the normoxia group. Glial culture was not affected by the hypoxia. Western blot analysis showed an increased induction of 62-kDa c-Fos and 58, 60, and 66 kDa Myc proteins in rat hippocampal cells with 10(-7) M PGE1 treatment. We conclude that PGE1 at concentrations of 10(-6)-10(-9) M protects rat hippocampal neurons against hypoxic insult.

Inhibition of prostacyclin-induced Ca2+ mobilization by phorbol esters in Ob1771 preadipocytes

In addition to cAMP production, a transient elevation of intracellular free Ca2+ has been shown to take place in preadipose cells upon stimulation by carbaprostacyclin (cPGI2), both messengers acting in synergy to initiate adipose cell differentiation (Vassaux, G., Gaillard, D., Ailhaud, G., and Négrel, R. (1992) J. Biol. Chem.267, 11092-11097). Further studies reported herein show that this Ca2+ transient is i) elicited by the natural prostaglandin PGI2, ii) independent of the presence of extracellular Ca2+, suggesting a mobilization of Ca2+ from intracellular pools and ii) unaffected by cAMP elevating agents. Moreover, and in contrast to the InsP3-dependent Ca2+ signal evoked by PGF2 alpha, that induced by PGI2 is fully abolished by pretreatment with phorbol esters (EC50: 1-5 nM). Furthermore, experiments designed to empty the Ca2+ pools, using PGI2 or PGF2 alpha as Ca2+ mobilizing agents as well as pretreatments with drugs, allow to conclude that PGI2 mobilizes Ca2+ from an InsP3 sensitive, ryanodine insensitive intracellular pool. Altogether, these results strongly suggest that PGI2 mobilizes Ca2+ from an intracellular store common to that affected by InsP3, by means of a mechanism which remains to be elucidated.

Effects of hormones and cytokines on stimulation of adenylate cyclase and intracellular calcium concentration in human and canine periodontal-ligament fibroblasts

Adenylate cyclase was stimulated by prostaglandin E2 (PGE2) and parathyroid hormone-related protein (PTHrP) in both these types of fibroblast and by calcitonin gene-related protein (CGRP) in the human fibroblasts in vitro. PGE2 (1 microM), CGRP (1 microM), and PTHrP (1 microM) stimulated adenylate cyclase up to 50-fold, 10-fold and 9-fold, respectively. Calcitonin (CT), substance P (SP), interleukin-1 beta (IL-1 beta), and transforming growth factor-beta 1 (TGF beta 1) had no effect on adenylate cyclase in either fibroblast. Intracellular Ca2+ (iCa2+) was measured in individual fibroblasts from the periodontal ligament using Indo-1 and an adherent cell analysis and sorting interactive laser cytometer. Ionomycin (3 microM) caused a transient rise of iCa2+ in all human and canine fibroblasts tested. The mean percentage increase in iCa2+ in response to ionomycin was 820 and 840% for human and canine fibroblasts, respectively. The human fibroblasts responded to PGE2 (1 microM) by an increased iCa2+ concentration; the mean percentage increase in iCa2+ was 187%. SP caused a less pronounced increase in iCa2+ in the human fibroblasts (56%). CGRP and SP caused a similar response in the canine fibroblasts. The mean percentage increase in iCa2+ in response to SP and CGRP was 95 and 78%, respectively. PTH, PTHrP, platelet-activating factor, CT, and IL-1 beta had no effect on iCa2+ in either type of fibroblast. The data indicate that cAMP and calcium have roles as intracellular secondary messengers in the action of PGE2, SP, CGRP, and PTHrP in fibroblasts of human and canine periodontal ligament.

Dissociation between parathyroid hormone-stimulated cAMP and calcium increase in UMR-106-01 cells

We used the osteogenic sarcoma cell line, UMR-106-01, to determine whether the rise in free cytosolic Ca2+ concentration ([Ca2+]i) and cellular cAMP following PTH stimulation are able to be regulated independently. For this purpose, we compared the effect of a PTH antagonist, stimulation of protein kinase C, augmentation by prostaglandins, and the time course of desensitization of the two cellular responses. Two x 10(-7) M of the PTH antagonist 8,18Nle 34Tyr-bPTH(3-34) amide ([Nle,Tyr]bPTH(3-34)A) was required to inhibit 10(-9) M bPTH(1-34)-stimulated cAMP generation by 50%. 10(-7) M bPTH(1-34) completely overcame the inhibition induced by 10(-6) M [Nle,Tyr]bPTH(3-34)A. Only 7 x 10(-8) M and 2.7 x 10(-7) M [Nle,Tyr]bPTH(3-34)A were required to half maximally inhibit the [Ca2+]i increase evoked by 3 x 10(-8) and 10(-7) M bPTH(1-34), respectively. In addition, dissociation between [Ca2+]i and cAMP signals was observed when modulation by protein kinase C and prostaglandins was tested. Preincubation of the cells with 10 nM TPA for 5 minutes markedly inhibited the PTH-evoked [Ca2+]i increase. Short incubation with PGF2 alpha augmented the PTH-evoked [Ca2+]i increase. Similar pretreatments had no effect on the PTH-stimulated cAMP increase. Finally, preincubation with 1.5 x 10(-9) M bPTH(1-34) for 20 minutes almost completely blocked the effect of 10(-7) M bPTH(1-34) on [Ca2+]i, while preincubation with 5 x 10(-9) M bPTH(1-34) for 4 hours was required to inhibit the effect of 10(-8) M bPTH(1-34) on cAMP production by 50%. The differences in the regulation of the two PTH-stimulated cellular signaling systems, in particular, the response to antagonists and the time course of desensitization, could be at the level of the PTH receptor(s) or at a postreceptor domain.

Calcium signals in growth factor signal transduction

There is a substantial amount of information which has been obtained concerning the effects of growth factors on [Ca2+]i in proliferating cells. A number of different mitogens are known to induce elevations in [Ca2+]i and some characterization of the Ca2+ response to different classes of mitogens has been obtained. In addition, much is known about whether the Ca2+ response to a particular growth factor occurs as the result of an influx of external Ca2+ or a mobilization of internal Ca2+ stores. In addition, a considerable amount of information is available on the mechanism by which the Ins(1,4,5)P3-sensitive internal Ca2+ store takes up and releases Ca2+. However, there is still a large deficiency in our information concerning other Ca2+ stores in proliferating cells as well as in our knowledge of the mechanisms for regulating Ca2+ entry pathways. Much more data addressing these issues exists for other types of agonist-stimulated cells, and we have discussed much of it in this review article. While the wealth of data in nonproliferating cells provides some indications of what mechanisms might be involved in the growth factor-induced changes in [Ca2+]i, it is clear that much work must be done in proliferating cells to fully understand how external factors such as growth factors control [Ca2+]i. In addition, much work remains to be done in identifying the mechanisms for the internal control of [Ca2+]i as cells move through the cell cycle and in identifying the role that these changes in [Ca2+]i may play throughout the cell cycle.

Two possibly distinct prostaglandin E1 receptors in N1E-115 clone: one mediating inositol trisphosphate formation, cyclic GMP formation, and intracellular calcium mobilization and the other mediating cyclic AMP formation

Prostaglandin E1 (PGE1)-mediated transmembrane signal control systems were investigated in intact murine neuroblastoma cells (clone N1E-115). PGE1 increased intracellular levels of total inositol phosphates (IP), cyclic GMP, cyclic AMP, and calcium ([Ca2+]i). PGE1 transiently increased inositol 1,4,5-trisphosphate formation, peaking at 20 s. There was more than a 10-fold difference between the ED50 for PGE1 at cyclic AMP formation (70 nM) and its ED50 values at IP accumulation (1 microM), cyclic GMP formation (2 microM), and [Ca2+]i increase (5 microM). PGE1-mediated IP accumulation, cyclic GMP formation, and [Ca2+]i increase depended on both the concentration of PGE1 and extracellular calcium ions. PGE1 had more potent intrinsic activity in cyclic AMP formation, IP accumulation, and cyclic GMP formation than did PGE2, PGF2 alpha, or PGD2. A protein kinase C activator, 4 beta-phorbol 12 beta-myristate 13 alpha-acetate, had opposite effects on PGE1-mediated IP release and cyclic GMP formation (inhibitory) and cyclic AMP formation (stimulatory). These data suggest that there may be subtypes of the PGE1 receptor in this clone: a high-affinity receptor mediating cyclic AMP formation, and a low-affinity receptor mediating IP accumulation, cyclic GMP formation, and intracellular calcium mobilization.

Unitary model of cell activation, growth control, cancer and other diseases: 1. Activated oxygen species and arachidonic acid modulation of solute permeabilities, internal Ca, Na and AOS levels and DNA transcription and synthesis

A comprehensive model of cellular activation and proliferation is developed. The model has arachidonic acid (ARA) produced mainly from PLA2 on both sides of the membrane, and superoxide and other activated oxygen species (AOS) formed from O2 by electrons passing out through membrane NANPH and NADH oxidases, as the immediate stimulants of solute permeability. Both ARA and AOS interact with the various solute channel proteins especially their external thiols and disulfides, to increase influx of metabolic substrates, Na, Ca and O2. PLA2 and NADPH oxidase are turned on by growth factors at their receptors acting through tyrosine kinase phosphorylations of messenger proteins GP and ras p-21, stimulated proteases, and by Ca-calmodulin. The adenylate cyclase system has opposite, deactivating character as it increases efflux of Ca and desensitizes growth factor receptors by phosphorylation to shut down the increased solute permeability. Most cancer types are due to carcinogen binding to cell membrane channel and mitochondrial sites for increased solute influx with excessive AOS production inside the cell from mitochondria and other vesicles. High Ca, Na and AOS stimulate proliferation with extra high levels causing transformation to the autogenic, more embryonic-type cancer cell.

Intracellular coupling of prostaglandin inhibition of acid secretion in isolated rabbit gastric parietal cells

Acid secretion from isolated rabbit gastric parietal cells can be stimulated by gastric secretagogues, histamine (cyclic-AMP pathway) and carbachol (inositol phosphate pathway). Prostaglandins (PG) from E series are potent inhibitors of acid secretion. The intracellular mechanism of this inhibition was examined by using a stable PGE1-analogue, misoprostol. Aminopyrine (AP) accumulations due to histamine, IBMX and forskolin were dose-dependently inhibited by misoprostol, whereas a weak but significant biphasic effect on carbachol-induced AP accumulation was observed. The cyclic-AMP formation induced by histamine and IBMX were also inhibited by misoprostol in a non-competitive way. The potent effect of forskolin on cyclic-AMP levels was not modified by misoprostol in parietal cells, whereas it was potentiated in non-parietal cells. The inhibitory effect of misoprostol on AP accumulation was reduced by incubation of parietal cells with Bordetella pertussis toxin (IAP) but not with Cholera toxin (CT). Pretreatment of the cells with IAP did not alter cyclic-AMP levels of resting and histamine-stimulated parietal cells but abolished the inhibitory effect of misoprostol. Treatment with CT increased basal and histamine-stimulated cyclic-AMP levels and masked the inhibitory effect of misoprostol. The biphasic effect of misoprostol on carbachol-stimulated AP accumulation in parietal cells was confirmed on carbachol-stimulated phospholipase C activity and on [Ca2+]i stimulated by carbachol. These data confirm a direct and specific effect of the prostanoid on the Gi-subunit of the adenylate cyclase coupled to the histamine H2-receptor, and a biphasic effect on the phospholipase C pathway of the parietal cells.

Contribution of external and internal Ca2+ to changes in intracellular free Ca2+ produced by mitogens in Swiss 3T3 fibroblasts: the role of dihydropyridine sensitive Ca2+ channels

Changes in intracellular free Ca2+ concentration [( Ca2+]i) produced by growth factors and mitogens have been studied using aequorin-loaded Swiss 3T3 cells. Decreasing free Ca2+ in the external medium by using EGTA had no significant effect on the increase in [Ca2+]i produced by vasopressin, bradykinin, bombesin or prostaglandin E2, but reduced the increase in [Ca2+]i produced by platelet derived growth factor (PDGF) by 58%, by prostaglandin E1 44% and by prostaglandin F2 alpha 47%. The dihydropyridine Ca2+-channel antagonist nifedipine at 10 microM inhibited the [Ca2+]i response to PDGF by 41% in both the presence of and in the absence of external Ca2+. Methyl-1,4-dihydro-2,6-dimethyl-3-nitro-4-(2-trifluoromethylphenyl) pyridine-5-carboxylate (BAY K8644), a Ca2+-channel agonist, at 10 microM produced an increase in [Ca2+]i and decreased the [Ca2+]i response to PDGF by 39%. Nifedipine did not block 45Ca2+ uptake or release by inositol 1,4,5-trisphosphate in saponin-permeabilized Swiss 3T3 fibroblasts but BAY K8644 inhibited 45Ca2+ release by inositol 1,4,5-trisphosphate. The results suggest that the increase in [Ca2+]i caused by PDGF in Swiss 3T3 fibroblasts is due to the influx of external Ca2+ through dihydropyridine sensitive Ca2+ channels, as well as release of internal Ca2+.

Effects of corticosteroids, prostaglandin E2, and beta-agonists on adenylate cyclase activity in fetal rat lung fibroblasts and type II epithelial cells

The effect of glucocorticoids on the response of adenylate cyclase in fetal rat lung fibroblast and Type II epithelial cell cultures to beta-agonists and prostaglandin E2 (PGE2) was investigated. There was significant stimulation of cyclic AMP synthesis by adrenalin in both fibroblasts and Type II cells, which was potentiated in a dose-dependent manner by cortisol. Stimulation of adenylate cyclase by PGE2 (10-1000 nM) was demonstrated in fibroblasts but not in Type II cells. The response to PGE2 was stimulated by pretreatment with cortisol only in fibroblasts (p less than .01), and no latent response to PGE2 was evident in Type II cells after cortisol treatment. These experiments suggest that both cell types isolated from late gestation fetal lung contain active beta-agonist and glucocorticoid receptors that synergize in raising intracellular cyclic AMP, which has multiple effects, including surfactant secretion from Type II cells. Since the adenylate cyclase response to PGE2 and its enhancement by glucocorticoids occurred only in fibroblasts, it is concluded that the reported effects of PGE2 on surfactant release are not mediated via raised intracellular cyclic AMP in Type II cells.

Prostaglandin E1: a review

Work on the structure of prostaglandin E1 (PGE1), isolated from natural sources, was completed 25 years ago (1). Shortly after, methods for the chemical synthesis of PG with their natural configuration were developed in the laboratories of the UpJohn Company (2) and of E. J. Corey (3) and, by the late sixties, PGE1 became widely available. The information since accumulated about its biological and clinical effects is more substantial than for any other PG. This review will draw together some of this information, focusing on recent studies of its mechanisms of action.