Keywords

Introduction

The close anatomical and functional relationship between brain endothelial cells and the remaining neural cells within the neurovascular unit underscores the importance of the cerebral endothelium in brain homeostasis and disease (Park et al. 2003). Brain endothelial cells release neurotrophic factors essential for neuronal survival and function, while their physical interactions with astrocytes and pericytes regulate endothelium vascular tone and BBB function, ensuring the proper metabolic exchange between blood and the brain parenchyma (Ward and Lamanna 2004; Hawkins and Davis 2005; Iadecola and Nedergaard 2007; Hamel 2006; Abbott et al. 2006). Emerging evidence also indicates a crucial role for brain endothelial cells in regulating neural stem cell functions. In the vascular niches of the brain, the direct contact between neural stem cells and brain endothelial cells, together with a continuous supply of endothelial-derived growth factors, is essential for neural stem cell self-renewal and differentiation (Shen et al. 2004; Kim et al. 2004). Given that the finely tuned communication between endothelial cells and other components of the neurovascular unit ensures the anatomical and functional integrity of the brain, injuries to brain endothelial cells can have dramatic clinical consequences and contribute to the pathogenesis of diverse neurological disorders (Rizzo and Leaver 2010). Thus, acute disruption of the anatomical integrity of the BBB, due to endothelial cell damage or death, results in vasogenic edema and tissue damage (Hatashita and Hoff 1990), while chronic injuries to the endothelium appear to have a critical role in the pathogenesis of neurodegeneration (Zipfel et al. 2009; Bell and Zlokovic 2009). Moreover, an injured and dysfunctional brain endothelium can interfere with the formation of new blood vessels and therefore delay the recovery of neurological functions after cerebral ischemia or traumatic brain injury (Beck and Plate 2009; Xiong et al. 2010). On the other hand, stimulation of brain endothelial cell angiogenic responses is a peculiar and essential component of malignant brain tumor development and progression (Plate et al. 1992).

Oxidative stress, a condition induced when ROS production overcomes antioxidant cellular defenses, plays a critical role in the pathogenesis of brain endothelial cell injury and vascular damage; although when appropriately regulated and controlled, ROS production contributes to brain endothelial cell homeostatic functions (Chrissobolis et al. 2011; Dröge 2001). Brain endothelial cells are especially susceptible to oxidative stress-induced injuries because of intrinsic features as well as environmental factors. Compared to endothelial cells from the periphery, brain endothelial cells are enriched in mitochondria and therefore are a potent source of ROS, which in turn can act upon the surrounding brain cellular environment inducing damage or regulating homeostatic functions (Oldenford et al. 1977; Adam-Vizi 2005). On the other hand, the high number of mitochondria renders brain endothelial cells an easy target of endogenously produced ROS (Adam-Vizi 2005). Additionally, brain endothelial cells lie in a microenvironment especially favorable for ROS-dependent cell damage. The brain parenchyma is enriched in highly reactive transition metal ions, such as iron and copper, which catalyze several oxidation-reduction reactions involved in generation of ROS (Jomova and Valko 2011). An additional factor is the availability of polyunsaturated fatty acids, in which neural cell membranes are especially enriched (Hazel and Williams 1990). While some polyunsaturated fatty acids (PUFA) such as docosahexaenoic acid (DHA) can have specific protective effects on the brain, others such as arachidonic acid are not only targets of ROS attack but also substrates for enzymatic generation of ROS and therefore contribute, potentiate, and exacerbate ROS-induced vascular damage or alternatively regulate, in a context-specific manner, endothelial cell homeostatic functions (Zhang and Bazan 2010; Calandria and Bazan 2010; Rapoport 2008).

Increased production of arachidonic acid has been linked to neuroinflammation (Rapoport 2008; Rao et al. 2011). Activation of the phospholipase A2 enzyme family by proinflammatory cytokines, hypoxia, or ROS results in the release of arachidonic acid, which either directly or via metabolic oxidation to eicosanoids, induce endothelial cell injury and deterioration of vascular functions (Rosa and Rapoport 2009; Adibhatla and Hatcher 2007; Phillis et al. 2006). Interestingly, while in the peripheral vascular system, the transformation of arachidonic acid into prostacyclin is the main mechanism maintaining physiological vasodilation and ensuring delivery of nutrients and oxygen to tissues, brain microvessels produce mainly PGE2, and both COX-1 and COX-2 are expressed in neonatal brain microvascular endothelial cells under basal conditions and after cytokine challenge (Dusting et al. 1977; Ellis et al. 1979; Parfenova et al. 2002). Thus, while certain arachidonic acid metabolites are involved in regulation of brain microvasculature physiological functions, such as control of vascular tone and blood flow (Niwa et al. 2001; Medhora et al. 2007), arachidonic acid-derived ROS are mostly relevant to cerebral vascular pathology (Armstead 2003).

Evidence obtained from other organs points to a significant degree of cross-talk and cross-regulation between oxidative stress and inflammation (Spychalowicz et al. 2012). Thus, oxidative stress exacerbates endothelial dysfunction caused by inflammation and can contribute to the production of arachidonic acid via several mechanisms, including a direct effect on cPLA2 phosphorylation and sPLA2 release (Goldman et al. 1992; Buschbeck et al. 1999; Jensen et al. 2009). On the other hand, inflammation not only increases the production of ROS but also impairs the antioxidant system further damaging the endothelium (Spychalowicz et al. 2012). Given the relevance of the brain endothelium to brain homeostasis and disease processes and the lack of effective treatments for vascular inflammation, a clearer understanding of the signaling integration and cross-talk between mediators of oxidative stress and inflammation, such as ROS and arachidonic acid, may improve the identification of new signaling targets that can be exploited to prevent or control the clinical consequences of vascular damage in a wide range of neurological disorders.

Detection and Source of ROS in Brain Endothelial Cells

Due to their transient nature and complex cellular distribution and pathophysiology, ROS participating in oxidative stress in the brain microvasculature, as well as in other cellular systems, are difficult to quantitate, and most studies employ downstream markers to monitor oxidative stress (Halliwell and Chirico 1993; Morrow et al. 1990). However, advances in imaging and analytical techniques, including spin trap techniques and mass spectrometry and the development of novel animal models, have improved identification of ROS and their metabolites in brain endothelial cells in situ (Yamato et al. 2003; Sparvero et al. 2010; Maier et al. 2006). ROS generated by or targeting brain microvascular endothelial cells include oxygen radicals such as superoxide anion, hydroxyl and peroxyl radicals, and non-radicals such as hydrogen peroxide (Miller et al. 2006). Moreover, superoxide reaction with NO generates peroxynitrites, which have direct cytotoxic effects while also inducing cellular damage by decreasing NO availability (Miller et al. 2006).

While production of ROS was initially viewed as an inevitable consequence of cell metabolic activity, the discovery of several enzyme complexes involved in generation of ROS led to the current notion that ROS production is a highly regulated process. In the brain endothelium, NADPH oxidase, the mitochondrial respiratory chain, and xanthine oxidase play a major role in production of ROS (Chrissobolis and Faraci 2008; Zhang and Gutterman 2007; Terada et al. 1991). It should, however, be emphasized that the damaging vascular consequences of ROS are a function of several factors, including their concentrations, their cellular compartmentalization, the activity of the cell detoxification systems, and the surrounding microenvironment (Sies 1993). When defenses against ROS, including superoxide dismutase (which converts superoxide into hydrogen peroxide) and catalase (which converts hydrogen peroxide to water and molecular oxygen), together with other intracellular and extracellular antioxidants such as glutathione and ascorbic acid are functioning, the free radical attack is limited, and resistance and repair mechanisms against oxidative damage are enhanced (Sies 1993). Recent evidence also indicates that endogenously expressed heme oxygenase-2 plays an important role in protecting brain endothelial cells against oxidative stress-induced injuries (Parfenova and Leffler 2008). Therefore, in the presence of a functioning detoxification system, ROS are maintained at low concentrations and as such serve as physiological regulators of endothelial cell functions. On the other hand, when the amount of ROS overcomes the antioxidant system, as during inflammation, ischemia-reperfusion, tumor growth, and γ-radiation-induced cellular stress, endothelial cell injury can occur leading ultimately to tissue damage and organ failure.

Contribution of Arachidonic Acid to ROS Production and ROS-Dependent Cellular Targets

During inflammation, oxidation of arachidonic acid via cyclooxygenases and lipoxygenases contributes to ROS production by the endothelium or by inflammatory and immune cells (Virdis et al. 2005; Cho et al. 2011; Leaver et al. 1995). In the cyclooxygenase reaction cycle, the derived alkoxyl radical intermediate PGG2 can oxidize a large number of substrates including NADPH to generate superoxide, which can be rapidly transformed by tissue dismutases to hydrogen peroxide (Kukreja et al. 1986; Marnett et al. 1999). Moreover, in a reaction catalyzed by iron, superoxide reacts to produce hydroxyl radicals (Halliwell and Chirico 1993). Importantly, ROS can also be generated by COX-2 via epoxidation reactions that generate cytotoxic peroxyl radicals without the requirement of cyclooxygenase activity and PGG2 formation (Im et al. 2006). Generation of superoxide from arachidonic acid oxidation into prostaglandins constitutes a relevant source of ROS in the brain during pathological processes, as shown by the increase of PGE2 in experimental models of brain injury (Easton and Fraser 1998; Easton and Abbott 2002) or during age-induced impairment of vascular functions (Armstead 2003). Products of 5-lipooxygenase are also involved in mediating arachidonic acid-induced cerebral edema (Yen and Lee 1987).

The involvement of other enzyme complexes in arachidonic acid-induced ROS production in the cerebral vasculature is not well characterized. However, in phagocytes, arachidonic acid is an essential cofactor of superoxide production due to its ability to directly activate NADPH oxidase (Shiose and Sumimoto 2000; Leaver et al. 1995). Interestingly, emerging evidence indicates that oxidative products of arachidonic acid metabolism are direct activators of NADPH oxidase, thus providing additional possibilities for arachidonic acid-dependent contribution and amplification of oxidative stress (Cho et al. 2011). Given the emerging role of endothelial NADPH oxidase in generation of ROS, activation of NADPH oxidase by arachidonic acid could constitute a relevant source of ROS in the brain microvasculature and a point of cross-talk between ROS- and arachidonic acid-dependent cellular responses.

In addition to its contribution to ROS generation, arachidonic acid is a critical target of ROS. Generation of lipid peroxides from ROS attack on arachidonic acid constitutes a major mechanism of ROS-induced cytotoxicity (Halliwell and Chirico 1993; Fraser 2011). Brain endothelial cells are particularly sensitive to lipid peroxides. 4-Hydroxynonenal, a lipid peroxidation product whose levels increase during hypoxia-reperfusion, increases brain endothelial cell vascular permeability and induces endothelial cell apoptosis (Mertsch et al. 2001; Kunstmann et al. 1996). These highly cytotoxic lipid peroxides are produced enzymatically or via auto-oxidation of arachidonic acid (Jian et al. 2005). This latter process is enhanced in the presence of iron and could constitute an important mechanism of lipid peroxidation-induced damage in vivo during hemorrhagic stroke, for example, because of the availability of hemoglobin-derived iron (Wang and Lo 2003).

ROS-mediated peroxidation of arachidonic acid generates iso-eicosanoids, including isoprostanoids and isoleukotrienes. Among isoprostanoids, F2-isoprostane, which is generated independently of cyclooxygenase, is not only the best characterized product of lipid peroxidation but also a useful marker of lipid peroxidation in vivo (Morrow et al. 1990). The isoprostane 8-iso-PGF2α has been shown to cause endothelial cell injury and impairment of neovascularization (Benndorf et al. 2008). Additional isoprostanes, such as 8-iso-PGE2, which is derived from PGE2, can also mediate oxidative stress-induced vascular damage during inflammation (Huber et al. 2003). Isomers of leukotrienes, such as B4-isoleukotrienes, are also generated, at least in vitro, via free radical-induced oxidation of arachidonic acid or arachidonate-containing glycerophospholipids (Harrison and Murphy 1995). While these complex molecules act as potent agonists of calcium mobilization in neutrophils (Harrison and Murphy 1995), their relevance to cerebral vascular damage remains to be determined.

Additional free radicals that have harmful consequences on the brain microvasculature are generated from isomerization of arachidonic acid by NO2. These compounds, known as trans-arachidonic acid (TAA) metabolites, are produced in vivo under a variety of pathological conditions (Jiang et al. 1999). There is substantial evidence in support of a relevant role of TAA metabolites in inducing cytotoxicity of the brain vasculature (Kooli et al. 2008). In a recent study, exposure of newborn pups to hypercapnia was associated with production of TAA metabolites and subsequent degeneration and death of the cerebral microvasculature. Importantly, damage to the cerebral microvasculature resulted in impairment of brain growth (Honoré et al. 2010). The damaging-inducing effects of TAA products are mediated by the production of the proapoptotic and antiangiogenic factor thrombospondin-1 (Kermorvant-Duchemin et al. 2005) (Fig. 86.1).

Fig. 86.1
figure 1

ROS- and arachidonic acid-sensitive signaling interactions and feedback loops in regulation of brain endothelial cell responses. Exogenously or endogenously generated by a variety of agonists, ROS target the brain endothelium either directly or through their interactions with arachidonic acid (AA) and activate signaling pathways that ultimately mediate their consequences on brain endothelial cell function and dysfunction. ROS-induced regulation of ion channels, activation of AKT and ERK, and heme oxygenase–2 (HO–2)–dependent release of carbon monoxide (CO) transmit signals that regulate vascular tone, survival, and cytoprotection. ROS-dependent modulation of angiogenic growth factor receptors, such as Tie-2 and VEGFR, activates downstream effectors responsible for stimulation of angiogenesis. ROS-induced activation of JNK, ASK-1, and p38-MAPK and mitochondrial perturbations result in endothelial cell death. Multiple signaling pathways, including P-glycoprotein upregulation, phospholipase C (PLC)-dependent intracellular calcium mobilization, tyrosine phosphorylation of the tight junction (TJ) complexes and subsequent activation of PI3 kinase and Rho, and activation of matrix metalloproteases (MMP-9, MMP-2) mediate ROS-induced changes in cell shape, loss of monolayer integrity, and TJ opening resulting in BBB breakdown and increases in vascular permeability. ROS can interact with AA at several levels. ROS can activate PLA2 and release AA, which in turn can serve as a target of ROS attack and generate lipid peroxides (LP), isoprostanes (F2-iso), and trans-arachidonic acid (TAA). These ROS-AA derivatives can then affect downstream signaling effectors responsible for cell death and disruption of monolayer integrity. On the other hand, AA can function as a supply of ROS by activating NADPH oxidase either directly, at least in phagocytes, or via COX- or LOX-dependent mechanisms. These signaling interactions potentiate ROS and AA-deleterious cellular consequences and contribute to brain endothelial cell dysfunction

Full size image

ROS and Arachidonic Acid in Regulation of Brain Endothelial Cell Physiological Functions

Under normal physiological conditions, ROS are constantly produced and detoxified in the brain environment. There is substantial evidence that ROS produced at low concentrations are required for normal physiological functions of the brain microvasculature acting as signaling molecules and impinging upon homeostatic, prosurvival, and cytoprotective pathways (Luczak et al. 2004; Chan et al. 2009; Lassègue and Griendling 2010) (Fig. 86.1). Studies performed in vivo or in isolated cerebral vessels indicate that exogenously generated or endogenously produced ROS play a critical role as regulators of vascular tone (Wei et al. 1996). This cellular response is of particular importance in the brain where the lack of glucose storage makes this organ uniquely vulnerable and dependent on blood flow for nutrients. The mechanisms involved in physiological regulation of cerebral vessel vascular tone by ROS such as hydrogen peroxide and peroxynitrite include regulation of calcium-sensitive or ATP-sensitive potassium channel opening (Sobey et al. 1997; Fraser 2011) (Fig. 86.1).

Studies of Kontos et al. provided the first evidence that ROS produced endogenously by cerebral vessels are important mediators of physiological vasodilation. Endogenous production of ROS appears also to mediate the vasodilation effects of arachidonic acid and bradykinin in the cerebral arteries in a manner dependent on arachidonic acid metabolism (Kontos et al. 1984). Consistent with this possibility, other studies have shown a role for COX-mediated arachidonic acid metabolism in cerebral vessel vasodilation (Kanu et al. 2006). Moreover, arachidonic acid-induced vasodilation was reduced in a COX-1 knockout mouse model, further indicating that oxidative metabolites of arachidonic acid play a regulatory role in physiological vasodilation (Niwa et al. 2001). The mechanisms of arachidonic acid and PGE2-induced vasodilation include the release of carbon monoxide in a manner independent of ROS generation (Kanu et al. 2006; Kanu and Leffler 2011). It is important, however, to emphasize that ROS and arachidonic acid-induced responses in the cerebral vasculature are under multicellular control, as the underlying pericytes and associated astrocytes critically contribute to modulate capillary vasoactivity either directly or via the production of soluble factors (Abbott et al. 2006; Hamel 2006; Shimizu et al. 2012).

In addition to vasodilation, ROS participate in regulation of endothelial cell survival acting as signaling molecules that impact prosurvival or antiapoptotic signaling networks with the outcome of protecting and maintaining the anatomical and functional integrity of the brain endothelium against inflammatory insults. Consistent with this possibility, Basuroy et al. recently demonstrated that production of ROS during TNF-α-induced brain endothelial cell injury resulted in activation of heme oxygenase-2 and subsequent production of the antioxidant carbon monoxide, which in turn protected brain endothelial cell from ROS-induced death by inhibiting Nox4 via AKT-dependent suppression of ERK and p38-MAPK (Basuroy et al. 2011). These intriguing findings underscore a novel role of ROS as signaling effectors that exert not only cytotoxic effects but activate also feedback systems able to restore redox homeostasis and possibly to initiate repair responses.

The intracellular source of ROS responsible for stimulating endothelial cell functions under physiological conditions is not well defined, possibly because it is under the dynamic control and regulation of several factors, including stimuli and environmental conditions. However, the NADPH oxidase system appears to constitute a major source of ROS under physiological conditions in endothelial cells (Chan et al. 2009), while ROS derived from the mitochondrial respiratory chain and generated during inflammation and other cellular insults are mostly linked to endothelial cell damage (Babior 2000). Consistent with a homeostatic role of NADPH, oxidase-derived ROS are the findings showing that expression of NOX2 and NOX4 in brain endothelial cells has been linked to endothelial cell proliferation (Van Buul et al. 2005; Luczak et al. 2004). However, there is evidence that NADPH oxidase also mediates ROS-induced vascular damage, suggesting that the environmental context under which the injury occurs dictates the involvement and cellular outcome of certain ROS-producing enzymes (Lassègue and Griendling 2010).

ROS and Arachidonic Acid in Brain Endothelial Cell Injury

Among the deleterious consequences of oxidative stress in the cerebral vascular system is disruption of BBB endothelial monolayer integrity with subsequent increase of vascular permeability, vasogenic edema, and tissue damage (Yang and Rosenberg 2011). Furthermore, enhanced permeability of the BBB allows the transmigration of leukocytes thus exacerbating the inflammatory process (Van der Goes et al. 2001). The signaling pathways mediating the deleterious effects of ROS and arachidonic acid-derived ROS in brain endothelial cells are not fully characterized. However, several pathways that target cellular effectors implicated in regulating the passage of molecules through the BBB are affected by ROS in cooperation with arachidonic acid, as shown in Fig. 86.1. Studies of Haorah et al. (2007) demonstrated that ROS-induced tyrosine phosphorylation of the BBB tight junctional complexes results in intracellular gap formation and increased vascular permeability. Alterations of the tight junctional complexes responsible for disruption of the BBB are also elicited by ROS-dependent phospholipase C activation and changes in the intracellular levels of calcium (Van der Goes et al. 2001). Moreover, using an in vitro model of BBB, Scheibert et al. (2007) demonstrated that ROS-induced stress fiber formation and increases in transmigration of monocytes were mediated by a signaling pathway involving activation of Rho kinase, AKT, and PI3 kinase. An additional mechanism by which ROS affect the BBB function is via the multidrug transporter P-glycoprotein, which was found to be upregulated following exposure of cultured brain endothelial cells to oxidative stress (Felix and Barrand 2002).

Loss of adherence to the matrix substratum and subsequent disruption of endothelial monolayer integrity also result in increases of BBB permeability. Recent studies have shown that matrix metalloproteases (MMP) are significant targets of ROS signaling and ROS attack on arachidonic acid (Lehner et al. 2011). Both mediators cooperate in targeting MMPs as indicated by recent studies in which ROS and arachidonic acid-derived lipid peroxides mediated ammonia-induced disruption of the BBB via activation of MMP-9 and MMP-2 in an in vivo model of liver failure (Skowrońska et al. 2012). Additionally, isoprostanes exert potent cytotoxic effects on the brain microvasculature that result in vasoconstriction, increased vascular permeability, and monocyte adhesion to the endothelium (Brault et al. 2003; Huber et al. 2003). Thus, disruption of the BBB constitutes an example of cooperation between ROS and arachidonic acid as arachidonic acid-derived lipid peroxide and ROS generated by the COX or LOX pathways are especially relevant to increase in vascular permeability and cerebral edema (Yen and Lee 1987; Fraser 2011).

Additional mechanisms mediating ROS-induced increases of BBB vascular permeability involve the release of angiogenic growth factors, such as VEGF. In a model of ischemia-reperfusion, ROS stimulated the release of VEGF, which in turn induced alteration of BBB integrity and increased vascular permeability via activation of ERK (Narasimhan et al. 2009). How ERK mediates VEGF-induced vascular permeability is not fully understood. However, changes in cytoskeletal or junctional proteins are possibly involved.

Endothelial cell survival is essential to the maintenance of the BBB monolayer integrity. Increased production of ROS, derived from the mitochondrial respiratory chain, can negatively affect endothelial cell survival via activation of proapoptotic signaling pathways (Zhang and Gutterman 2007). High levels of free fatty acids have been linked to brain endothelial cell apoptosis via a ROS-mediated mechanism (Zhou et al. 2009). Stress kinases, including JNK and ASK-1, play an important role in mediating ROS and arachidonic acid death-inducing signals (Hsu et al. 2007; Shen and Liu 2006; Rizzo and Carlo-Stella 1996). ASK-1-dependent activation of p38-MAPK has been implicated in apoptosis of brain endothelial cells induced by Aβ-fragments (Hsu et al. 2007). The involvement of p38-MAPK in brain endothelial cell death responses was also reported in a model of ischemia-reperfusion or following in vitro exposure of brain microvascular endothelial cells to free arachidonic acid (Lee and Lo 2003; Rizzo et al. unpublished observations).

While ROS involved in brain endothelial cell injury are produced within the brain, it should be noted that ROS or arachidonic acid and its metabolites produced by the peripheral vasculature can target brain endothelial cells, especially in the presence of a permissive BBB as it occurs, for example, during the acute phase of inflammation. Thus, peripherally generated ROS or ROS sources such as arachidonic acid and its metabolites can target the brain endothelium and exacerbate brain endothelial cell dysfunction during system inflammation (Davidson et al. 2001; Skowrońska et al. 2012).

Regulation of Brain Endothelial Cell Angiogenic Responses by ROS and Arachidonic Acid

Angiogenesis, the formation of new blood vessels from preexisting capillaries, occurs in the brain during embryonic development in concert with vasculogenesis, the formation of new blood vessels from hemangioblasts (Mancuso et al. 2008; Conway et al. 2001). In adult brain, where growth and development are complete, endothelial cells become quiescent being mainly involved in preserving the BBB integrity (Plate 1999). However, in response to inflammatory, traumatic, neoplastic, or degenerative insults to the brain, the balance between inhibitors and stimulators of angiogenesis, which under normal conditions prevent angiogenesis, is shifted in favor of neovascularization (Lee et al. 2009; Rizzo and Leaver 2010). Under these conditions, the release of proangiogenic protein and lipid growth factors, together with increased metabolic demands of the brain, results in stimulation of endothelial cell angiogenic responses and neovascularization (Plate et al. 1992; Rush et al. 2007; Rizzo and Leaver 2010).

Several lines of evidence indicate that ROS regulate brain endothelial cell angiogenic responses. Several angiogenic growth factors, such as VEGF and angiopoietin-1, in addition to hypoxia and products of arachidonic acid metabolism, activate NADPH oxidase. Work from the laboratory of Ushio-Fukai has elegantly characterized ROS-dependent signaling networks that regulate endothelial cell angiogenic responses. ROS-dependent stimulation of VEGF receptor autophosphorylation and downstream activation of AKT mediates VEGF-induced neovascularization, while ROS production through activation of Rac-1 subunit of NADPH oxidase mediates angiopoietin-1-induced angiogenesis (Ushio-Fukai 2006). While information on the cellular source of ROS production during brain angiogenesis is scanty, the relevance of NADPH oxidase in ROS-induced angiogenesis has been supported by studies demonstrating upregulation of Nox4 expression in new brain capillaries following ischemia-reperfusion (Vallet et al. 2005). PLA2 and products of arachidonic acid metabolism also contribute to endothelial cell angiogenic responses in vitro and in vivo (Rizzo et al. 2000; Nie et al. 2000). The extent of ROS involvement in these responses is not known. Moreover, these studies were performed in endothelial cells from peripheral organs. Therefore, their relevance to the brain microvasculature remains to be established.

Superoxide and hydrogen peroxide, produced during ischemia-reperfusion or hypoxia, can promote angiogenesis via upregulation of hypoxia- and redox-sensitive transcription factors and their target genes (Ushio-Fukai 2006). On the other hand, they can stimulate apoptotic pathways and thus impair endothelial cell angiogenic responses. Similarly, isoprostanes have been shown to selectively promote necrosis of brain microvascular endothelial cells during ischemia (Brault et al. 2003).

Rationale for Targeting ROS and Arachidonic Acid-Sensitive Signaling Pathways in the Brain Endothelium

Brain endothelial cell dysfunction contributes to several neurological disorders, while interfering with brain endothelial cell angiogenic responses can affect response to therapy in brain tumors. Substantial evidence indicates that ROS contribute to the progression of malignant gliomas by targeting the endothelium and triggering the development of structurally disorganized and highly permeable neovascular structures, the presence of which correlates with a more aggressive tumor phenotype (Lee et al. 2009). Whether these tumor-associated angiogenic responses are stimulated by ROS produced endogenously by the endothelial cells or by ROS derived from tumor cells or from other components of the tumor microenvironment is not entirely clear. However, it is clear that intratumoral hypoxia, which is a potent stimulus for the angiogenic switch in gliomas, is also a potent stimulator of ROS production, suggesting that tumor-derived ROS are involved in targeting the endothelium (Richard et al. 1999). Importantly, enzymes participating in the oxidation of arachidonic acid during tumorigenesis, such as COX-2, are also upregulated by hypoxia in glioma cells and contribute to stimulation of tumor-associated angiogenesis (Rizzo 2011). Thus, ROS generation and upregulation of COX-2 during hypoxia constitute an important point of functional cross-talk influencing several aspects of tumor biological behavior. Understanding the signaling integration between ROS and arachidonic acid metabolism during glioma-associated neovascularization may provide new therapeutic opportunities for these very aggressive and clinically challenging tumors.

The interplay between ROS and arachidonic acid may be significant also in nonneoplastic disorders of the brain, including ischemia, traumatic brain injuries, and neurodegeneration. Emerging evidence indicates a crucial contribution of ROS and arachidonic acid-derived lipid peroxides to neurodegeneration. Deposition of oligomeric Aβ fragments in cerebral vessel walls induces ROS-mediated apoptosis (Hsu et al. 2007). On the other hand, activation of the PLA2 family and subsequent release of arachidonic acid by Aβ peptides could initiate ROS-dependent cytotoxic responses leading to cerebrovascular dysfunction (Shelat et al. 2008; Moses et al. 2006). Intriguingly, patients with Alzheimer’s disease have increased amount of arachidonic acid and lipid peroxidation (Esposito et al. 2008). Studies in cPLA2 knockout animals have provided evidence in support for a role of PLA2, arachidonic acid, and ROS in cerebral ischemia and associated edema (Tabuchi et al. 2003). Additional studies in these areas of brain pathology should provide useful insights of the interplay between ROS and arachidonic acid that could translate into clinical benefits.

Concluding Remarks

In summary, cooperation and cross-talk between oxidative stress and inflammation play an important role in regulating brain endothelial cell responses. While physiological fluctuations in certain ROS play a role in homeostasis, oxidative stress can also have drastic effects on the cerebral vasculature, promoting and/or exacerbating pathologies ranging from stroke to neurodegenerative disease and oncogenesis. Similarly, while arachidonic acid contributes to vascular homeostasis, generation of ROS via arachidonic acid metabolism or generation of lipid peroxides may result in endothelial cell injury and vascular damage.

The signaling pathways of ROS-induced damage of the brain microvasculature and the contribution of arachidonic acid to ROS-dependent networks are, however, not well characterized, partly because of limitations to the detection of transiently formed ROS and arachidonic acid oxidative products in complex cellular systems. Moreover, the majority of the information on ROS-dependent signaling pathways in the vasculature derives mostly from studies performed with endothelial cells isolated from peripheral organs. Such knowledge cannot readily apply to the brain microvasculature because of its unique morphological and biochemical characteristics. Therefore, studies using brain endothelial cells and new experimental models that better represent the multicellular nature of the brain vasculature together with improved in vivo imaging techniques are needed to provide further insights into the functional and signaling interactions between ROS and arachidonic acid and allow the development of better therapeutic strategies to modulate ROS production and enhance mechanisms of repair and cytoprotection. Identification of new oxidative and inflammatory cellular targets and molecular profiling within the brain microvasculature should also facilitate the development of new noninvasive biomarkers than can be introduced into clinical practice.

Given that inflammatory, traumatic, degenerative, and neoplastic disorders of the brain constitute a major threat and burden to public health worldwide, understanding the mechanisms that control brain endothelial cell function and dysfunction may aid the development of targeted vascular therapies aimed at restoring the anatomical and functional integrity of the neurovascular unit.