P3-01-07: A Murine Xenograft Model for the Study of Ciclooxygenase-2 Expression and Function in Human Breast Cancer

BACKGROUND: ciclooxygenase-2 (COX-2) has been suggested as a necessary component of the cellular and molecular mechanisms behind breast cancer cell motility and invasion The potential therapeutic benefit of COX-2 inhibitors in a range of cancers is being seen as a great promise however, there have been recent concerns about potential cardiotoxicity. Thus, there is an urgency to develop new inhibitors exhibiting a better risk/benefit ratio.

Acyl-CoA synthetase 4 (ACSL4), belongs to a five-member familyof enzymes that esterifies mainly arachidonic acid into acyl-CoA. We have provided first-time evidence demonstrating that, ACSL4 is the key enzyme that regulates the induction of COX-2, the production of prostaglandin E2 (PGE2) and the proliferation and metastatic potential of breast cancer cells. Therefore, the aim of the study was the development of an in vivo model of human breast tumor xenografts for the study of the regulation of COX-2 expression and action.

RESULTS: stable transfection of MCF-7 cells with ACSL4 cDNA under the control of tetracycline (MCF-7 Tet-off-ACSL4) resulted in a significant increase in the expression of COX-2. The increment in COX-2 expression registered in MCF-7 Tet-off-ACSL4 cells is accompanied with an increase in the production of PGE2, and the proliferation and metastatic potential of cancer cells. Next, we tested whether the injection of MCF-7 Tet-off/ACSL4 cells into nude mice resulted in tumor development.

The results of those experiments demonstrate that MCF-7 Tet-off/ACSL4 cells develop into murine mammary tumors whereas cells transfected with the MCF-7 Tet-off empty vector did not. Interestingly, treatment of nude mice with tetracycline resulted in tumor growth inhibition. The results show that the sole transfection of ACSL4 results in a phenotype change that endows the cells with the capacity to develop into tumors when injected into nude mice. Tumor volume (TV) at day 70 reached 468.4 ± 189.3mm3. Treatment with tetracycline reduced TV to 189.3 ± 65.9mm3 (p<0.05). The growth rate between days 45–70 was 12.30 mm3/day. Tumors were classified as Basal-like subtypes.

We have also observed an inhibition of proliferation and migration of MDA-MB-231 cell cultures when exposed to a combination of ACSL4 and COX-2 inhibitors. Interestingly, the compounds assayed markedly reduced cell proliferation and migration at concentrations that are less effective when used alone. These results point to an effect that has the advantage of exposing the cells to lower drug concentrations.

CONCLUSION: Based on our results, we hypothesize that ACSL4 and COX-2 could constitute potential therapeutic targets for the control of tumor growth. Our animal model of mammary tumors constitutes a proper platform for the study of those therapies.

Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P3-01-07.

Adrenocorticotropin induces heme oxygenase-1 expression in adrenal cells

Heme oxygenase (HO) catalyzes the first and rate-controlling step of heme catabolism into biliverdin, iron and carbon monoxide. Three isoforms of HO have been identified so far: the inducible HO-1 and the constitutive HO-2 and HO-3. Both HO-1 and HO-2 were expressed in zona fasciculata (ZF) adrenal cells and in a mouse adrenocortical cell line (Y1). HO-1 but not HO-2 expression was upregulated by adrenocorticotropic hormone (ACTH) and accumulation of HO-1 protein correlated with an increase in HO activity in Y1 cells. ACTH induced HO-1 expression in a time- and dose-dependent manner with a maximum after 5 h of treatment and a threshold concentration of 0.1 mIU/ml. Actinomycin D and cycloheximide completely blocked the effect of ACTH on HO-1 mRNA expression whereas mRNA stability was not affected by ACTH. Permeable analogs of cAMP mimicked the effect of ACTH on HO-1 expression and ACTH induction was prevented by the protein kinase A (PKA) inhibitor H89. Steroid production was significantly increased when both HO-1 and HO-2 activities were inhibited by Sn-protoporphyrin IX (SnPPIX). The lipid peroxidation and increase in carbonyl content triggered by hydrogen peroxide was prevented by treatment of Y1 cells with bilirubin and ACTH.

Corticotropin increases protein tyrosine phosphatase activity by a cAMP-dependent mechanism in rat adrenal gland

Corticotropin signal transduction pathway involves serine/threonine protein phosphorylation. Recent reports suggest that protein tyrosine dephosphorylation may also be an integral component of that pathway. The present study was performed to investigate the role played by protein tyrosine phosphatases (PTPs) on acute response to corticotropin and the hypothetical regulation of PTPs by this hormone. We have used two powerful cell permeant PTP inhibitors, phenylarsine oxide (PAO) and pervanadate (PV), in order to examine the relevance of PTP activity on hormone-stimulated and 8-bromo-adenosine 3',5'-phosphate (8Br-cAMP is a permeant analogue of adenosine 3',5'-phosphate)-stimulated steroidogenesis in adrenal zona fasciculata (ZF) cells. In both cases, PAO and PV inhibited the steroid production in a dose-dependent fashion, and had no effect on steroidogenesis supported by a permeant analogue of cholesterol. The effect of hormonal stimulation on PTP activity was analyzed in rat adrenal ZF. In vivo corticotropin treatment reduced phosphotyrosine content in endogenous proteins and produced a transient increase of PTP activity in the cytosolic fraction, reaching a maximum (twofold) after 15 min. Incubation of adrenal ZF with 8Br-cAMP also produced PTP activation, suggesting that it can be mediated by cAMP-dependent protein kinase (PKA)-dependent phosphorylation. Detection of PTP activity in an in-gel assay showed three corticotropin-stimulated soluble PTPs with molecular masses of 115, 80 and 50 kDa. In summary, we report for the first time a hormone-dependent PTP activation in a steroidogenic tissue and provide evidence that PTP activity plays an important role in corticotropin signal pathway, acting downstream of PKA activation and upstream of cholesterol transport across the mitochondrial membrane.

Exocytosis of vacuolar apical compartment (VAC) in Madin-Darby canine kidney epithelial cells: cAMP is involved as second messenger

Vacuolar apical compartment (VAC) is a transient organelle originally observed in Madin-Darby canine kidney (MDCK) epithelial cells impaired from forming cell-cell contacts. VACs are large vacuoles which contain microvilli and apical plasma membrane markers (among others, a 184-kDa plasma membrane protein, AP2), but exclude basolateral membrane markers. Upon reestablishment of cell-cell contacts, VACs are rapidly (within 1 h) exocytosed toward intercellular spaces, after which the apical plasma membrane drifts toward its final destination (Vega-Salas, Salas, and Rodriguez-Boulan. 1988. J. Cell Biol. 107, 1717-1728). In this work, we studied the role of cAMP as a mediator for the exocytosis of VACs. We shifted confluent cells from low to normal calcium medium (thus reestablishing cell-cell contacts and causing VAC exocytosis), a shift which resulted in a significant rise of cellular levels of both total intracellular and protein-bound cAMP. The 8-Br analog of cAMP (8-Br-cAMP) (5-50 microM) caused externalization of the intracellular compartment of AP2 as measured by radioimmunoassay. A similar effect was observed with 3-isobutyl-1-methylxanthine. 8-Br-cAMP also caused the appearance of AP2-positive VAC images in nonpermeabilized cells, namely, VACs that become accessible to extracellular antibodies upon fusion with the plasma membrane. Lanthanum, which abolishes the peak of intracellular free calcium during a calcium switch, failed to block the exocytosis. On the other hand, 12-O-tetradecanoylphorbol-13-acetate induced only a modest exocytic response. Finally, 8-Br-cAMP induced VAC exocytosis in sparse MDCK cells grown in normal calcium medium. These data indicate that cAMP is a mediator between the extracellular signal provided by cell-cell contacts and VAC exocytosis.

The cytosol as site of phosphorylation of the cyclic AMP-dependent protein kinase in adrenal steroidogenesis

The mitochondria, the microsomes and the cytosol have been described as possible sites of cAMP-dependent phosphorylation. However, there has been no direct demonstration of a cAMP-dependent kinase associated with the activation of the side-chain cleavage of cholesterol. We have investigated the site of action of the cAMP-dependent kinase using a sensitive cell-free assay. Cytosol derived from cells stimulated with ACTH or cAMP was capable of increasing progesterone synthesis in isolated mitochondria when combined with the microsomal fraction. Cytosol derived from cyclase or kinase of negative mutant cells did not. Cyclic AMP and cAMP-dependent protein kinase stimulated in vitro a cytosol derived from unstimulated adrenal cells. This cytosol was capable of stimulating progesterone synthesis in isolated mitochondria. Inhibitor of cAMP-dependent protein kinase abolished the effect of the cAMP. ACTH stimulation of cytosol factors is a rapid process observable with a half maximal stimulation at about 3 pM ACTH. The effect was also abolished by inhibitor of arachidonic acid release. The function of cytosolic phosphorylation is still unclear. The effect of inhibitors of arachidonic acid release, and the necessity for the microsomal compartment in order to stimulate mitochondrial steroidogenesis, suggest that the factor in the cytosol may play a role in arachidonic acid release.

Stimulation of an individual cell with peptide hormone in a prescribed region of its plasma membrane results in a compartmentalized cyclic AMP-dependent protein kinase response

This work describes the stimulation by a peptide hormone of an individual cell in a prescribed region of its plasma membrane. When Leydig cells were stimulated via a section of membrane tightly sealed to an electrode containing LH, a very localized area exhibited the morphological change known as 'rounding up', which is a cyclic AMP-dependent protein kinase-mediated response. This localized stimulation did not produce a wider response through intracellular, intermembranous or extracellular signals. Each individual cell responded to peptide stimulation gradually, with an increase over time and with dose. In contrast, when the stimulation was accomplished using a non-hydrolysable cyclic AMP analogue in the patch electrode, a general response throughout an individual cell was produced. Locally stimulated peptide hormone receptors, adenylate cyclases and cyclic AMP-dependent protein kinases appear to be closely associated so that second messenger production and the effects it mediates are compartmentalized.

The action of luteinizing hormone on the testis

Luteinizing hormone (LH) and human chorionic gonadotrophin (hCG) receptors are coupled to intracellular effector systems, most notably adenylate cyclase, through guanyl nucleotide-binding proteins or G-proteins. The molecular mechanism involved in the dynamic coupling of the LH/hCG receptor however, are not known. It has been postulated that receptor aggregation at the molecular level plays a critical role in this process. There have been attempts to understand the receptor association and dissociation phenomena at the molecular level. One of them involves the participation of the major histocompatibility complex (MHC) class I antigen in the mechanism of receptor activation and/or expression. One molecular basis for these mechanisms consists of a physical interaction between MHC proteins and receptors to form "compound receptors" able to transfer a hormonal signal to the cell. Using a photo-reactive probe we demonstrated that the LH/hCG receptors and the class I antigens are closely associated in the membrane. Thus, it is possible to form covalent complexes of hCG and class I antigens through the binding of the hormone to specific receptors. These findings imply that LH/hCG receptors and the MHC class I antigens may interact at the level of the plasma membrane in the mechanism of LH action. We also performed experiments using a single cell and limiting stimulation to a patch of membrane. The results stimulating the cell in a localized area suggested that even if all components are entirely free to float there is a constraint in the localization of the receptor, G-protein, and/or the effector, supporting the constraint dissociation model. Within a limited area subunits could dissociate, but they would not be free to diffuse throughout the membrane. Moreover the concept of compartmentalization that has been utilized to explain some inconsistencies in second-messenger action now can be proved by experimental design.

Lipoxygenase products as common intermediates in cyclic AMP-dependent and -independent adrenal steroidogenesis in rats

Aldosterone secretion from adrenal glomerulosa cells can be stimulated by angiotensin II (AII), extracellular potassium and ACTH. Mitochondria from these cells respond to intracellular factors generated by AII (cyclic AMP (cAMP)-independent steroidogenesis) and ACTH (cAMP-dependent steroidogenesis), suggesting that the two-signal-transduction mechanisms are linked by a common intermediate. We have evaluated this hypothesis by stimulating mitochondria from the unstimulated zona glomerulosa with a subcellular post-mitochondrial fraction (PMF) obtained from the zona glomerulosa after stimulation with AII or from the fasciculata gland after stimulation with ACTH; the subcellular fractions were also tested on mitochondria from fasciculata cells. PMFs obtained after incubation of adrenal zona glomerulosa with or without AII (0.1 microM) or ACTH (0.1 nM) were able to increase net progesterone synthesis 4.5-fold in mitochondria isolated from unstimulated rat zona glomerulosa. AII-pretreated PMFs from the zona glomerulosa also stimulated steroidogenesis by mitochondria from zona fasciculata cells. Separate experiments showed that inhibitors of arachidonic acid release and metabolism (bromophenacyl bromide, nordihydroguaiaretic acid, caffeic acid or esculetin) blocked corticosterone production in fasciculata cells stimulated with ACTH, suggesting that arachidonic acid could be the common intermediate in the actions of AII and ACTH on steroid synthesis. Evidence to support this concept was obtained from experiments in which the formation of an activated PMF by treatment of zona fasciculata with ACTH was blocked by the presence of the same inhibitors. Moreover, the inhibitory effects of these substances on PMF activation by ACTH were overcome by exogenous arachidonic acid and, in addition, arachidonic acid release was stimulated by ACTH.(ABSTRACT TRUNCATED AT 250 WORDS)

Rat adrenal cycloheximide-sensitive factors and phospholipids in the control of acute steroidogenesis

ACTH in vivo induces the formation of several steroidogenic factors in both cytosol and extramitochondrial particulate fractions of rat adrenal. Cycloheximide prevents the formation of these factors. Here we show the presence of a cytosolic steroidogenic component (C1) which is cycloheximide-sensitive and not ACTH-dependent. C1 is able to solubilize an ACTH-dependent steroidogenic factor (C2) from particulate fractions resulting in the release of the rat-limiting constraint of mitochondrial steroidogenesis. The thermolabile and trypsin-resistant factor C1 has an apparent mol.wt of 28,000 Daltons. In contrast, the cycloheximide-sensitive factor C2 from extra-mitochondrial fractions of ACTH-treated rats comigrates on Sephadex G-10 with phospholipids. Endogenous phospholipids isolated from particulate adrenal fractions of ACTH-treated rats or exogenous phospholipids will also stimulate steroidogenesis in vitro. Indeed, cytosolic solubilizing factor C1 enhances the exogenous phospholipid effect 3-4-fold. The results taken together suggest that C1 may be very similar to a well defined phospholipid exchange protein and C2 is itself a phospholipid. Both factors seem to be obligatory for the ACTH-induced steroidogenesis.

Steroidogenic action of calcium ions in isolated adrenocortical cells

The corticotropin-induced increase of total intracellular and receptor-bound cyclic AMP in isolated rat adrenocortical cells was strictly dependent on extracellular Ca(2+). A rise in bound cyclic AMP with rising Ca(2+) concentrations was accompanied by a decrease in free cyclic AMP-receptor sites. A Ca(2+)-transport inhibitor abolished the rise in bound cyclic AMP induced by corticotropin. These data suggested that during stimulation by corticotropin some Ca(2+) has to be taken up in order to promote the rise of the relevant cyclic AMP pool. In agreement with this view, adenylate cyclase activity from isolated cells proved also to be dependent on a sub-millimolar Ca(2+) concentration in the presence of corticotropin and GTP. When cells were treated under specific conditions, corticosterone production could be activated by Ca(2+) in the absence of corticotropin (cells primed for Ca(2+)). Ca(2+)-induced steroidogenesis of these cells, in the absence of corticotropin, was also accompanied by an increase in total intracellular and receptor-bound cyclic AMP, as was found previously with corticotropin-induced steroidogenesis in non-primed cells. Calcium ionophores increasing the cell uptake of Ca(2+) were not able, however, to increase the cyclic AMP pools in non-primed cells, unlike corticotropin in nonprimed cells or Ca(2+) in cells primed for Ca(2+). It was concluded that during stimulation by either corticotropin or Ca(2+) a possible cellular uptake of Ca(2+) must be very limited and directed to a specific site which may affect the coupling of the hormone-receptor-adenylate cyclase complex.

Adrenocorticotropin (ACTH) induces phosphorylation of a cytoplasmic protein in intact isolated adrenocortical cells

In 32P incorporation experiments with intact adrenocortical cells, adrenocorticotropin (ACTH) or adenosine 3',5'-cyclic monophosphate (cAMP) induced a rapid and transient increase of approximately 300-500% in the phosphorylation of a 32P-containing cytoplasmic protein of about 150,000 daltons (APS150). Half-maximal stimulation of APS150 phosphorylation was observed with about 3 pM ACTH. Receptor-bound cAMP, corticosterone production, and the appearance of phosphorylated APS150 increased in parallel with respect to both time and ACTH concentration. All three responses were dependent on extracellular calcium. Inhibition of protein synthesis with cycloheximide suggested a half-life of APS150 of about 10 min. The time course of 32P incorporation into ACTH-induced APS150 in the absence and presence of nonradioactive phosphate shows that the phosphorylation of APS150 is under simultaneous control of cAMP-dependent protein kinase and of phosphoatase activity. Thus a rapid ACTH-dependent and cAMP-dependent protein phosphorylation in intact adrenocortical cells within steroidogenic ACTH concentrations has now been demonstrated.

Steroidogenesis in isolated adrenocortical cells. Correlation with receptor-bound adenosine e 3':5'-cyclic monophosphate

Because several groups have recently questioned a mediating role for cyclic AMP in adrenocortical steroidogenesis, we analysed the problem in more detail by measuring three different cyclic AMP pools in cells isolated from decapsulated rat adrenals. Extra-cellular, total intracellular and bound intracellular cyclic AMP were determined by radioimmunoassay in comparison with corticosterone production induced by low corticotropin concentrations. The increase in extracellular and total intracellular cyclic AMP with low corticotropin concentrations was dependent on the presence of a phosphodiesterase inhibitor and short incubation times. Bound intracellular cyclic AMP was less dependent on these two parameters. In unstimulated cells cyclic AMP bound to its receptor represents only a small fraction of the total intracellular cyclic AMP. After stimulation by a concentration of corticotropin around the threshold for corticosterone production, an increase in bound cyclic AMP was observed which correlated very well with steroidogenesis both temporally and with respect to corticotropin concentration. This finding was complemented by measuring a concomitant decrease in free receptor sites. Full occupancy of the receptors was not necessary for maximal steroidogenesis. Binding kinetics of cyclic [(3)H]AMP in concentrations equivalent to the intracellular cyclic AMP concentration suggest the presence of at least three different intracellular cyclic AMP pools. These observations are in agreement with a possible role for cyclic AMP as a mediator of acute steroidogenesis induced by low corticotropin concentrations.

Characterization of two forms of cyclic 3', 5'-adenosine monophosphate-dependent protein kinase in rat testicular interstitial cells

The adenosine 3', 5'-cyclic monophosphate (cyclic AMP)-dependent protein phosphokinase of rat interstitial cells was characterized by ion-exchange chromatography and sucrose density gradient centrifugation. The 0.2 M NaCl fraction from DEAE-Sephadex showed a small 2.9-S peak of basal enzyme activity, and a large 6.5-S peak of cyclic AMP-dependent protein kinase activity; fractions eluted from DEAE-Sephadex with 0.3-0.5 M NaCl contained a major 3.8-S peak of cyclic AMP-dependent enzyme activity. Activation of protein kinase in cell extracts by cyclic AMP, and in intact interstitial cells by trophic hormone, caused a major shift of enzyme activity to the 2.9-S cyclic AMP-dependent form which was eluted from DEAE-Sephadex by 0.2 M NaCl. These results are consistent with the presence of two distinct protein kinase holoenzymes, with a common 2.9-S catalytic subunit. During hormonal activation of protein kinase in dispersed interstitial cells by 10-10 M human chorionic gonadotropin (hCG), conversion to the 2.9-S catalytic subunit was observed between 2 and 30 min of incubation. Protein kinase activity was correlated with cyclic AMP production, and full enzyme activation occurred at the time of maximum intracellular cyclic AMP concentration. The presence of two forms of cyclic AMP-dependent protein kinase in the Leydig cell provides a potential mechanism whereby progressive occupancy of gonadotropin receptors could evoke a series of discrete target cell responses.

Physical characteristics of the gonadotropin receptor-hormone complexes formed in vivo and in vitro

The physical properties of detergent-solubilized gonadotropin receptor-hormone complexes, determined by density gradient centrifugation and gel filtration, were compared after in vivo and in vitro labeling of specific ovarian binding sites with radioiodinated human chorionic gonadotropin (hCG). Following intravenous administration of biologically active 125I-labeled hCG, up to 50% of the gonadotropin tracer was bound to the luteinized ovaries of immature female rats treated with pregnant mare serum/human chorionic gonadotropin. Comparable binding of 125I-labeled hCG was observed after equilibration of ovarian particles with the labeled hormone in vitro. The sedimentation properties of the solubilized receptor-hormone complexes formed in vivo were identical with those derived for the corresponding complexes formed in vitro and extracted with Triton X-100 and Lubrol PX, with sedimentation constants of 8.8 S for the Triton-solubilized complex and 7.0 S for the complex extracted with Lubrol PX. During analytical gel filtration of the Triton-solubilized receptor-hormone complex on Sepharose 6B in 0.1% Triton X-100, the partition coefficient (Kav) of the "in vivo" complex (0.32) was not significantly different from that of the complex formed in vitro (0.29). Gel filtration of the Lubrol-solubilized ovarian particles on Sepharose 6B in 0.5% Lubrol PX gave Kav values for the "in vivo" and "in vitro" labeled complexes of 0.36 and 0.32, respectively. These findings demonstrate that the physical properties of size and shape which determine the partition coefficient and sedimentation characteristics of detergent-solubilized gonadotropin receptor-hormone complexes formed in vitro are not distinguishable from those of the complexes extracted after specific interaction of the ovarian gonadotropin receptors with radioiodinated hCG in vivo.