Epigenetic activation of the prostaglandin receptor EP4 promotes resistance to endocrine therapy for breast cancer.

Approximately 75% of breast cancers express estrogen receptor α (ERα) and depend on estrogen signals for continued growth. Aromatase inhibitors (AIs) prevent estrogen production and inhibit ER signaling, resulting in decreased cancer recurrence and mortality. Advanced tumors treated with AIs almost always develop resistance to these drugs via the upregulation of alternative growth signals. The mechanisms that drive this resistance-especially epigenetic events that alter gene expression-are, however, not well understood. Genome-wide DNA methylation and expression analysis of cell line models of acquired AI resistance indicated that prostaglandin E2 receptor 4 (PTGER4) is upregulated after demethylation in resistant cells. Knockdown and inhibitor studies demonstrate that PTGER4 is essential for estrogen-independent growth. Our exploratory analysis of downstream signaling indicates that PTGER4 likely promotes AI resistance via ligand-independent activation of the ERα-cofactor CARM1. We believe that we have discovered a novel epigenetic mechanism for altering cell signaling and acquiring endocrine therapy resistance. Our findings indicate that PTGER4 is a potential drug target in AI-resistant cancers. In addition, the epigenetic component of PTGER4 regulation suggests that further study of PTGER4 may yield valuable insights into how DNA methylation-targeted diagnoses and treatments can improve AI-resistant breast cancer treatment.

Stimulation of glucose transport and glucose transporter phosphorylation by okadaic acid in rat adipocytes.

Okadaic acid, an inhibitor of Type I and IIa protein phosphatases, was recently found to stimulate 2-deoxyglucose uptake in rat adipocytes (Haystead, T. A. J., Sim, A. T. R., Carling, D., Honnor, R. C., Tsukitani, Y., Cohen, P., and Hardie, D. G. (1989) Nature 337, 78-81). In the present experiments the effect of okadaic acid on the phosphorylation and subcellular distribution of the insulin-regulatable glucose transporter (IRGT) was investigated. At maximally effective concentrations, insulin and okadaic acid increased the amount of IRGT in the plasma membrane by 10- and 4-fold, respectively. Thus, the stimulation of glucose transport by okadaic acid was apparently due to an increase in the surface concentration of the IRGT. However, despite its stimulatory actions, okadaic acid partially inhibited the ability of insulin to enhance glucose transport and translocation of the transporter. When cells were incubated with okadaic acid alone or in combination with insulin, phosphorylation of the IRGT in the plasma membrane was increased by approximately 3-fold relative to the intracellular pool of transporters in control cells. Phosphorylation of the IRGT was confined to the presumed cytoplasmic domain at the COOH terminus of the protein. Glucose transporters were dephosphorylated in vitro by Type I or Type IIa protein phosphatases, indicating that inhibition of one or both of these phosphatases could account for the increased phosphorylation produced by okadaic acid. The observation that okadaic acid stimulated translocation of the IRGT implicated a serine/threonine phosphorylation event in triggering movement of the intracellular IRGT-containing vesicles (GTV) to the cell surface. Immunoadsorption of GTV from 32P-labeled adipocytes revealed that the IRGT was the major phosphoprotein in these vesicles. The phosphorylation of at least three other GTV proteins was increased by okadaic acid, and these species would appear to be candidates for regulators of GTV movement to the plasma membrane. It is unlikely that phosphorylation of the IRGT is the signal for translocation because insulin did not increase phosphorylation of the protein. Rather, the inhibitory effect of okadaic acid on insulin-stimulated translocation is consistent with the hypothesis that phosphorylation of the IRGT promotes its internalization.

Phosphorylation of the glucose transporter in rat adipocytes. Identification of the intracellular domain at the carboxyl terminus as a target for phosphorylation in intact-cells and in vitro.

Phosphorylation of the insulin-regulatable glucose transporter (IRGT) is increased by incubating rat adipocytes with isoproterenol or by incubating microsomal membranes with cAMP-dependent protein kinase. To attempt to locate the sites of phosphorylation, the IRGT (apparent Mr = 46,000) was immunoprecipitated from 32P-labeled adipocytes and cleaved with CNBr or trypsin. Essentially all of the 32P could be recovered in a single CNBr fragment, denoted CB-T (Mr = 8,000), which bound a polyclonal antibody (R820) against a peptide having the sequence of the last 12 amino acids in the COOH terminus of the IRGT. 32P-Labeling of the IRGT was also confined to CB-T when membranes were incubated with [gamma-32P]ATP and cAMP-dependent protein kinase. Isoproterenol increased phosphorylation of CB-T, but insulin was without effect. To resolve phosphorylation sites further, IRGT from 32P-labeled cells was subjected to exhaustive proteolysis with trypsin. Samples were applied to a C-18 column, and 32P-labeled fragments were resolved into three peak fractions by elution with an increasing gradient of acetonitrile. [32P]Phosphoserine was the only phosphoamino acid detected in any of the peaks. Peak III contained approximately 80% of the 32P and was increased by isoproterenol. Almost all of the 32P introduced by cAMP-dependent protein kinase in vitro eluted in Peak III. In all cases, the 32P-labeled species in Peak III were quantitatively immunoprecipitated by R820. Digesting the peptide(s) in Peak III with V8 protease generated a single peak of 32P which eluted at lower acetonitrile than Peak III and contained 32P-labeled species that did not interact with R820. Automated Edman degradation indicated that the serine residue in Peak III phosphorylated by cAMP-dependent protein kinase was the 3rd or 4th residue from the NH2 terminus of the peptide. These findings indicate that phosphorylation of the IRGT is restricted to the presumed intracellular domain at the COOH terminus and that Ser488 is a major site phosphorylated both by cAMP-dependent protein kinase in vitro and in response to isoproterenol in vivo.

Insulin stimulates the generation of an adipocyte phosphoprotein that is isolated with a monoclonal antibody against the regulatory subunit of bovine heart cAMP-dependent protein kinase.

Incubating 32P-labeled fat cells with insulin increased by as much as 80-fold the amount of 32Pi in a soluble species of apparent Mr 62,000. This species, designated isp62, was specifically immunoprecipitated from cellular extracts with a monoclonal antibody against the type II regulatory subunit (RII) of cAMP-dependent protein kinase. Fat-cell RII, purified from extracts with cAMP-Sepharose or labeled with 8-azido [32P]cAMP, had an apparent Mr 51,000. Peptide mapping indicated that isp62 and adipocyte RII were different proteins. When cells were metabolically labeled with [35S]methionine, insulin stimulated the appearance of 35S-labeled isp62, indicating that the hormonal effect involves generation of the protein. The insulin-induced increase in isp62 could be observed within 1 min, occurred with physiological concentrations of the hormone, and was rapidly reversible. The increase in isp62 was unaffected by cycloheximide, indicating that insulin stimulates the posttranslational processing of a precursor, rather than de novo synthesis of the protein.

Control of glycogen synthase by insulin and isoproterenol in rat adipocytes. Changes in the distribution of phosphate in the synthase subunit in response to insulin and beta-adrenergic receptor activation.

Rat adipocytes were incubated with [32P]phosphate to label glycogen synthase, which was rapidly immunoprecipitated from cellular extracts and cleaved using either CNBr or trypsin. All of the [32P]phosphate in synthase was recovered in two CNBr fragments, denoted CB-1 and CB-2. Isoproterenol (1 microM) rapidly decreased the synthase activity ratio (-glucose-6-P/+glucose-6-P) and stimulated the phosphorylation of both CB-1 and CB-2 by approximately 30%. Insulin opposed the decrease in activity ratio and blocked the stimulation of phosphorylation by isoproterenol. Incubating cells with insulin alone changed the 32P content of neither CB-1 nor CB-2. Trypsin fragments were separated by reverse phase liquid chromatography and divided into peak fractions, denoted F-I-F-VII in order of increasing hydrophobicity. F-V contained almost half of the [32P]phosphate and was phosphorylated when synthase was immunoprecipitated from unlabeled fat cells and incubated with [gamma-32P]ATP and the cAMP-independent protein kinase, FA/GSK-3. That F-V also had the same retention time as the skeletal muscle synthase fragment containing sites 3(a + b + c) suggests that it contains sites 3. Muscle sites 1a, 5, 1b, and 2 eluted with F-I, F-II, F-VI, and F-VII, respectively. F-V was increased approximately 25% by isoproterenol, but the largest relative increases were observed in F-I (4-fold), F-III (4-fold), and F-VI (2-fold). These results indicate that beta-adrenergic receptor activation results in increased phosphorylation of multiple sites on glycogen synthase. Insulin plus glucose decreased the overall 32P content of synthase by approximately 30%, with the largest decrease (40%) occurring in F-V. Without glucose, insulin decreased the [32P]phosphate in F-V by 17%, an effect which was balanced by increases in F-I, F-II, and F-III so that no net change in the total 32P contents of the fractions was observed. Thus, activation of glycogen synthase by the glucose transport-independent pathway seems to involve a redistribution of phosphate in the synthase subunit.

Phosphorylase kinase isozymes in normal and electrically stimulated skeletal muscles.

Phosphorylase kinase was quantitatively immunoprecipitated from extracts of different rabbit skeletal muscles, subjected to electrophoresis on polyacrylamide gels in the presence of sodium dodecyl sulfate, and stained with silver. Amounts of the two isozymes, enzymes r and w. were determined by the staining intensities of the alpha'- and alpha -subunits, respectively. The kinase in muscles composed primarily of slow oxidative fibers was almost entirely enzyme r, but several times more of this isozyme was found in muscles with a high proportion of fast oxidative-glycolytic fibers. Enzyme w predominated in muscles composed mostly of fast glycolytic fibers. To investigate the possible role of muscle activity in controlling isozyme levels, tibialis anterior muscles were activated by chronic electrical stimulation of the peroneal nerves. After 10 wk of continuous stimulation, the amount of phosphorylase kinase was decreased by approximately 80%, and the ratio of enzyme r to enzyme w was doubled. These results demonstrate that increased muscle activity can decrease levels of phosphorylase kinase and suggest that activity may regulate the expression of the isozymes in different muscle fiber types.

Rat skeletal muscle glycogen synthase: phosphorylation of the purified enzyme by cAMP-dependent and -independent protein kinases.

Glycogen synthase was purified to near homogeneity from rat skeletal muscle, and was found to resemble the rabbit skeletal muscle enzyme in several respects. An apparent molecular weight (Mapp) of 86,000 was estimated from the electrophoretic mobility of the subunit on polyacrylamide gels in the presence of sodium dodecyl sulfate. Limited proteolysis of the rat synthase with trypsin resulted in the formation of species with MappS equal to 75,000, 69,000, and 67,000. The enzyme could be phosphorylated by cAMP-dependent protein kinase, phosphorylase kinase, and the cAMP-independent protein kinases, PC0.7 and FA/GSK-3. Essentially all of the phosphorylation observed occurred on serines located in two cyanogen bromide fragments, denoted CB-1 (Mapp = 13,000) and CB-2 (Mapp = 22,000). FA/GSK-3 and cAMP-dependent protein kinase phosphorylated sites in both fragments. Phosphate introduced by phosphorylase kinase was located exclusively in CB-1, and that incorporated with PC0.7 was found in CB-2. Phosphorylation by FA/GSK-3 reduced the electrophoretic mobility of the subunit, introduced heterogeneity into CB-2, and was synergistic with phosphorylation by PC0.7. To separate phosphorylation sites more completely, samples of glycogen synthase were subjected to extensive proteolysis using trypsin, followed by reverse-phase liquid chromatography. When phosphorylated by the same kinases, the pattern of fragments obtained with rat and rabbit skeletal muscle glycogen synthase were almost identical. The results presented provide strong evidence that the subunit of rat skeletal muscle glycogen synthase has at least five phosphorylation sites that are very similar, if not identical, to sites present on the rabbit muscle enzyme.

Glycogen synthase in rat adipocytes and skeletal muscle is phosphorylated on both serine and threonine.

Glycogen synthase is phosphorylated both in vivo and in vitro on multiple sites per subunit. All phosphorylations of the enzyme thus far identified occur on serines which are found in two cyanogen bromide fragments, denoted CB-1 and CB-2. We have immunoprecipitated [32P]glycogen synthase from rat adipocytes and epitrochlearis muscles incubated with [32P]phosphate. Phosphoamino acid analyses by two-dimensional electrophoresis after acid hydrolysis revealed no [32P]phosphotyrosine, but significant levels of [32P]phosphothreonine (6-14% of the [32P](phosphoserine). The [32P]phosphothreonine was recovered in the large CNBr-fragment (CB-2), indicative of a hitherto unknown phosphorylation site(s).

Hormonal control of glycogen synthase in rat hemidiaphragms. Effects of insulin and epinephrine on the distribution of phosphate between two cyanogen bromide fragments.

The effects of insulin and epinephrine on the phosphorylation of glycogen synthase were investigated using rat hemidiaphragms incubated with [32P]phosphate. Antibodies against rabbit skeletal muscle glycogen synthase were used for the rapid purification of the 32P-labeled enzyme under conditions that prevented changes in its state of phosphorylation. The purified material migrated as a single radioactive species (Mapp = 90,000) when subjected to electrophoresis in sodium dodecyl sulfate. Insulin decreased the [32P]phosphate content of glycogen synthase. This effect occurred rapidly (within 15 min) and was observed with physiological concentrations of insulin (25 microunits/ml). The amount of [32P]phosphate removed from glycogen synthase by either different concentrations of insulin or times of incubation with the hormone was well correlated to the extent to which the enzyme was activated. Epinephrine (10 microM) inactivated glycogen synthase and increased its content of [32P]phosphate by about 50%. Cleavage of the immunoprecipitated enzyme with cyanogen bromide yielded two major 32P-labeled fragments of apparent molecular weights equal to approximately 28,000 and 15,000. The larger fragment (Fragment II) displayed electrophoretic heterogeneity similar to that observed with the corresponding CNBr fragment (CB-2) from purified rabbit skeletal muscle glycogen synthase phosphorylated by different protein kinases. Epinephrine increased [32P]phosphate content of both fragments; however, the increase in the radioactivity of the smaller fragment (Fragment I) was more pronounced. Insulin decreased the amount of [32P] phosphate present in Fragments I and II by about 40%. The results presented provide direct evidence that both insulin and epinephrine control glycogen synthase activity by regulating the phosphate present at multiple sites on the enzyme.