H. Anne Leaver

Brain endothelial cell death: modes, signaling pathways, and relevance to neural development, homeostasis, and disease.

Emerging evidence indicates that brain microvascular endothelial cells play a critical role in brain development, maturation, and homeostasis. Acute or chronic insults, including oxidative stress, oxygen–glucose deprivation, trauma, infections, inflammatory cytokines, DNA damaging agents, β-amyloid deposition, and endoplasmic reticulum stress induce brain endothelial cell dysfunction and damage, which can result in cell death. The homeostatic balance between endothelial cell survival and endothelial cell death is critical for brain development, remodeling, and repair. On the other hand, dysregulation of brain endothelial cell death exacerbates, or even initiates, several inflammatory, ischemic, and degenerative disorders of the central nervous system. In here, the morphological, biochemical, and functional characteristics of the brain endothelium and its contribution to brain homeostasis will be reviewed. Recent insights into modalities and regulatory pathways involved in brain endothelial cell death will be described. The effects of regulated and dysregulated endothelial cell death leading to angiogenesis will be outlined. The relevance of brain endothelial cell dysfunction and death to disease processes will be discussed with special reference to recent findings that could help translate current knowledge on brain endothelial cell apoptosis into new therapeutic strategies for the treatment of certain neurological disorders.

Microsomal prostaglandin E synthase-1 regulates human glioma cell growth via prostaglandin E2–dependent activation of type II protein kinase A

Dysregulation of enzymes involved in prostaglandin biosynthesis plays a critical role in influencing the biological behavior and clinical outcome of several tumors. In human gliomas, overexpression of cyclooxygenase-2 has been linked to increased aggressiveness and poor prognosis. In contrast, the role of prostaglandin E synthase in influencing the biological behavior of human gliomas has not been established. We report that constitutive expression of the microsomal prostaglandin E synthase-1 (mPGES-1) is associated with increased prostaglandin E2 (PGE2) production and stimulation of growth in the human astroglioma cell line U87-MG compared with human primary astrocytes. Consistently, pharmacologic and genetic inhibition of mPGES-1 activity and expression blocked the release of PGE2 from U87-MG cells and decreased their proliferation. Conversely, exogenous PGE2 partially overcame the antiproliferative effects of mPGES-1 inhibition and stimulated U87-MG cell proliferation in the absence of mPGES-1 inhibitors. The EP2/EP4 subtype PGE2 receptors, which are linked to stimulation of adenylate cyclase, were expressed in U87-MG cells to a greater extent than in human astrocytes. PGE2 increased cyclic AMP levels and stimulated protein kinase A (PKA) activity in U87-MG cells. Treatment with a selective type II PKA inhibitor decreased PGE2-induced U87-MG cell proliferation, whereas a selective type I PKA inhibitor had no effect. Taken together, these results are consistent with the hypothesis that mPGES-1 plays a critical role in promoting astroglioma cell growth via PGE2-dependent activation of type II PKA. [Mol Cancer Ther 2006;5(7):1817–26]

Arachidonic acid uptake into and release from guinea-pig endometrium in vitro on days 7 and 15 of the oestrous cycle

Endometrium from guinea-pigs on Days 7 and 15 of the oestrous cycle (days of low and high endometrial prostaglandin F2 alpha production, respectively) was maintained in tissue culture for periods up to 24 h (uptake experiments) or 48 h (release experiments). Tritiated arachidonic acid (3H-AA) was incorporated into endometrial phospholipids and neutral lipids in a time-dependent manner. After 24 h of culture, phosphatidylcholine (PC) and phosphatidylethanolamine (PE) were the major phospholipids, and triglyceride (TG) was the major neutral lipid which had incorporated 3H-AA. PC, PE and phosphatidylserine/phosphatidylinositol (PS/PI) incorporated significantly more 3H-AA on Day 15 than on Day 7. TG also incorporated more 3H-AA on Day 15 than on Day 7, but the increase was not statistically significant. Tritiated oleic acid (3H-OA) was incorporated into endometrial phospholipids and neutral lipids in a time-dependent manner. No increase in uptake of 3H-OA occurred on Day 15 compared to Day 7. There appears to be a specific stimulation of the mechanisms involved in the uptake of arachidonic acid into guinea-pig endometrium (particularly into the phospholipids) at the end of the oestrous cycle. There was little apparent release of 3H-AA from any endometrial lipid class, except diglyceride (DG) and monoglyceride (MG), on Day 7. In contrast, there was an apparent 50 to 80% decrease in the 3H-AA content of several endometrial lipid classes, particularly PC, PE and TG, on Day 15. Overall, the uptake and release studies suggest that PC, PE and possibly TG form the source of free arachidonic acid for PGF2 alpha synthesis by the guinea-pig endometrium.

Prostaglandin E2 activates cAMP response element-binding protein in glioma cells via a signaling pathway involving PKA-dependent inhibition of ERK

Prostaglandin E(2) (PGE(2)) plays a critical role in influencing the biological behavior of tumor cells. We previously demonstrated that PGE(2) stimulates human glioma cell growth via activation of protein kinase A (PKA) type II. This study was undertaken to further elucidate the intracellular pathways activated by PGE(2) downstream to PKA. Stimulation of U87-MG glioma cells with PGE(2) increased phosphorylation of the cyclic-AMP response element (CRE) binding protein CREB at Ser-133 and CREB-driven transcription in a dose- and time-dependent manner. Expression of dominant CREB constructs that interfere with CREB phosphorylation at Ser-133 or with its binding to the CRE site markedly decreased PGE(2)-induced CREB activation. Inhibition of PKA by H-89 or expression of a dominant negative PKA construct attenuated PGE(2)-induced CREB activation. Moreover, inhibition of PKA type II decreased PGE(2)-induced CREB-dependent transcription by 45% compared to vehicle-treated cells. To investigate the involvement of additional signaling pathways, U87-MG cells were pretreated with wortmannin or LY294002 to inhibit the PI3-kinase/AKT pathway. Both inhibitors had no effect on PGE(2)-induced CREB phosphorylation and transcriptional activity, suggesting that PGE(2) activates CREB in a PI3-kinase/AKT independent manner. Challenge of U87-MG cells with PGE(2), at concentrations that induced maximal CREB activation, or with forskolin inhibited extracellular signal-regulated kinase (ERK) phosphorylation. Pretreatment of U87-MG cells with the ERK inhibitor PD98059, accentuated ERK inhibition and increased CREB phosphorylation at Ser-133 and CREB-driven transcription stimulated by PGE(2), suggesting that inhibition of ERK contributes to PGE(2)-induced CREB activation. Inhibition of ERK by PGE(2) or by forskolin was rescued by treatment of cells with H-89 or by the dominant negative PKA construct. Moreover, PGE(2) or forskolin inhibited phosphorylation of Raf-1 phosphorylation at Ser-338. Challenge of U87-MG cells with 11-deoxy-PGE(1) increased CREB-driven transcription and stimulated cell growth, while other PGE(2) analogues had no effect. Together our results reveal a novel signaling pathway whereby PGE(2) signals through PKA to inhibit ERK and increase CREB transcriptional activity.

Arachidonic acid induces brain endothelial cell apoptosis via p38-MAPK and intracellular calcium signaling

Arachidonic acid (AA), a bioactive fatty acid whose levels increase during neuroinflammation, contributes to cerebral vascular damage and dysfunction. However, the mode of injury and underlying signaling mechanisms remain unknown. Challenge of primary human brain endothelial cells (HBECs) with AA activated a stress response resulting in caspase-3 activation, poly(ADP-ribose) polymerase cleavage, and disruption of monolayer integrity. AA also induced loss of mitochondrial membrane potential and cytochrome c release consistent with activation of intrinsic apoptosis. HBEC stimulation with AA resulted in sustained p38-MAPK activation and subsequent phosphorylation of mitogen-activated protein kinase activated protein-2 (MAPKAP-2) kinase and heat shock protein-27 (Hsp27). Conversely, other unsaturated and saturated fatty acids had no effect. Pharmacological and RNA interference-mediated p38α or p38β suppression abrogated AA signaling to caspase-3 and Hsp27, suggesting involvement of both p38 isoforms in AA-induced HBEC apoptosis. Hsp27 silencing also blocked caspase-3 activation. AA stimulated intracellular calcium release, which was attenuated by inositol 1,4,5-trisphosphate (IP3) receptor antagonists. Blockade of intracellular calcium release decreased caspase-3 activation, but had no effect on AA-induced p38-MAPK activation. However, inhibition of p38-MAPK or blockade of intracellular calcium mobilization abrogated AA-induced cytochrome c release. AA-induced caspase-3 activation was abrogated by pharmacological inhibition of lipooxygenases. These findings support a previously unrecognized signaling cooperation between p38-MAPK/MAPKAP-2/Hsp27 and intracellular calcium release in AA-induced HBEC apoptosis and suggest its relevance to neurological disorders associated with vascular inflammation.

Glioma Cell Death: Cell-Cell Interactions and Signalling Networks

The prognosis for patients with malignant gliomas is poor, but improvements may emerge from a better understanding of the pathophysiology of glioma signalling. Recent therapeutic developments have implicated lipid signalling in glioma cell death. Stress signalling in glioma cell death involves mitochondria and endoplasmic reticulum. Lipid mediators also signal via extrinsic pathways in glioma cell proliferation, migration and interaction with endothelial and microglial cells. Glioma cell death and tumour regression have been reported using polyunsaturated fatty acids in animal models, human ex vivo explants, glioma cell preparations and in clinical case reports involving intratumoral infusion. Cell death signalling was associated with generation of reactive oxygen intermediates and mitochondrial and other signalling pathways. In this review, evidence for mitochondrial responses to stress signals, including polyunsaturated fatty acids, peroxidising agents and calcium is presented. Additionally, evidence for interaction of glioma cells with primary brain endothelial cells is described, modulating human glioma peroxidative signalling. Glioma responses to potential therapeutic agents should be analysed in systems reflecting tumour connectivity and CNS structural and functional integrity. Future insights may also be derived from studies of signalling in glioma-derived tumour stem cells.

Evidence for two particulate cholesterol ester hydrolase enzymes in rat corpus luteum

Abstract Two cholesterol ester hydrolases (CEH) were detected in the particulate fraction of the rat corpus luteum cell, one sedimenting at low centrifugal forces (between 760 and 25,410g) and the other at higher centrifugal forces (25,410–124,000g). The CEH in the fractions sedimenting at low centrifugal forces had an activity of 20–70 pmol/min/mg protein. This enzyme showed a similar sedimentation profile to the lysosomal acid phosphatase enzyme. A second, more active (200–350 pmol/min/mg protein) CEH activity had a different pH optimum and sedimented over higher centrifugal forces. These findings suggest that two CEH can provide cholesterol for steroidogenesis, using as substrate either cholesterol esters stored in the cell, or cholesterol esters derived from plasma lipoprotein.

To Die or Not To Die? Current Questions in Cell Death Signalling

A philosopher once stated that ‘over every good laboratory, there hangs a question mark’ [George MacLeod 1939, The Coracle], and it is certainly the case that questions surrounding the mechanism and role of cell death have hung over many laboratories over recent years. Cell death signalling has provided a fruitful area of enquiry, and it is now possible to carry out functional analysis of linked systems of signals, receptors, membranes, organelles and mediators involved in cell death signalling. But in such analysis, a central question, discussed in this edition of Molecular Neurobiology, relates to decision points: Under which circumstances and at what point do cells commit to die? This is an important question, not least because, in cases of pathological cell loss or gain, it marks the point after which therapeutic intervention may be futile, while, during development, it marks the point at which external cues are integrated to instruct a cell to live or to die. The diverse cell death signalling pathways leading to these decision points are discussed [Wyllie; Lukiw and Bazan] including oxidative and non-oxidative signalling pathways, inflammatory and non-inflammatory signals; responses of differently dividing and differentiated cell types; and sensitivity to environmental and therapeutic conditions. Further, the key molecular determinants of this decision involve both a lack of trophic stimuli [Brady and Morfini] as well as the active effects of co-ordinated cell death signals [Colquhoun]. Key effectors and integrators of the decision process, the intrinsic and extrinsic apoptotic signalling pathways, are discussed by Wyllie, Whittle, Rizzo, Colqhoun and Leaver. Andrew Wyllie describes the diversity of early signals and events leading to cell death, and how cross-talk between sub-lethal signals may cumulatively tip a cell or system towards cell death. He proposes that signalling processes involved in normal physiological functions can become involved in death under pathological conditions. Various aspects of brain-specific microenvironments are considered in this series of Cell Death Signalling reviews. The unique properties of brain endothelial cells and their response to stress signalling are described by Maria Teresa Rizzo. Brain endothelial cells may exacerbate, or even initiate, inflammatory, ischemic and degenerative disorders of the central nervous system. Both acute and chronic signals which activate endothelial cell death signalling are discussed. Important responses include stress kinase activation and changes in gene expression, but these responses depend on cellular contexts such as differentiation status and microenvironment. Rizzo proposes that the limited success of recent neuroprotective therapies relate to failure to consider the integrated response of the neurovascular unit. Glioma mitochondrial metabolism is discussed by Whittle et al., including mitochondrial-specific ketone metabolism in the Warburg hypothesis, and recent evidence for Akt oncogene expression and associated anaerobic glycolysis in malignant gliomas. Also, abnormalities in cardiolipin structure and distribution in glioma mitochondria are associated with disorders of mitochondrial respiratory complex function. Proteomic analysis of gliomas has revealed abnormalities in the expression of key components of the intrinsic apoptosis pathway. Further support for specific mitochondrial changes leading to cell death in H. A. Leaver (*) Department of Clinical Neurosciences, University of Edinburgh, Edinburgh, UK e-mail: hanne.leaver@googlemail.com