The Onset of Brain Injury and Neurodegeneration Triggers the Synthesis of Docosanoid Neuroprotective Signaling
- First Online:
- Cite this article as:
- Bazan, N.G. Cell Mol Neurobiol (2006) 26: 899. doi:10.1007/s10571-006-9064-6
- 245 Views
Bioactive lipid messengers are formed through phospholipase-mediated cleavage of specific phospholipids from membrane reservoirs. Effectors that activate the synthesis of lipid messengers, include ion channels, neurotransmitters, membrane depolarization, cytokines, and neurotrophic factors. In turn, lipid messengers regulate and interact with multiple pathways, participating in the development, differentiation, function (e.g., long-term potentiation and memory), protection, and repair of cells of the nervous system. Overall, bioactive lipids participate in the regulation of synaptic function and dysfunction. Platelet-activating factor (PAF) and COX-2-synthesized PGE2 modulate synaptic plasticity and memory. Oxidative stress disrupts lipid signaling, fosters lipid peroxidation, and initiates and propagates neurodegeneration. Lipid messengers participate in the interactions among neurons, astrocytes, oligodendrocytes, microglia, cells of the microvasculature, and other cells. A conglomerate of interrelated cells comprises the neurovascular unit. Signaling at the neurovascular unit is clearly altered in the early stages of cerebrovascular disease as well as in neurodegenerations. Here we will provide examples of how signaling by lipids regulates critical events essential for neuronal survival. We will highlight a newly identified, DHA-derived messenger, neuroprotectin D1, which attenuates oxidative stress-induced apoptosis. The specificity and potency of this novel docosanoid (neuroprotectin D1) indicate a potentially important target for therapeutic intervention.
KEY WORDS:docosahexaenoic acidischemia-reperfusionneuroprotectin D1neuroprotectionoxidative stress
It is an honor and a distinct pleasure to contribute to this special issue devoted to Dr. Julius Axelrod. I met Julie in the early 1970's at one of first meetings of the Society for Neuroscience. Julie quickly became not only a colleague but also a life long friend.
I clearly recall his curiosity as though we had just spoken this morning. In the mid-70s, before we knew each other well, Dr. Juan Saavedra, who was working in Julie's NIH section, brought us together. I sat on the bench next to Julie in his lab, and he asked me question after question about lipids. Julie had just published papers on the methylation of phospholipids and their relationship to neurotransmission, and he couldn't get enough of this line of inquiry. He was like a child with his incessant “why?” Julie never tired of honing in on new insights. It made for the beginning of a lifetime relationship. Julie was always a source of wise counsel to me.
Julie was a friend of remarkable character and a scientist of incomparable distinction. May this volume serve to continue to contribute to his abundant legacy.
OXIDATIVE STRESS, PRO-INFLAMMATORY SIGNALING, AND CELL DAMAGE
Oxidative stress and pro-inflammatory signaling are activated during the initiation and propagation of neuropathological conditions including neurodegenerative diseases (Alzheimer's disease), amyotropic lateral sclerosis, spinal cord injury, traumatic brain injury, epilepsy, and brain ischemia/reperfusion (Bazan et al., 2002; Bramlett and Dietrich, 2004; Consilvio et al., 2004; Danton and Dietrich, 2003; Dirnagl et al., 2003; Lo et al., 2003; Iadecola, 2004; Minghetti, 2004). The heterogeneity of the root causes of these conditions displays pro-death signaling pathways. The purpose of this article is to provide an overview of how membrane-derived docosanoid messengers may intervene in these pathways to achieve neuroprotection. Experimental data from our laboratory are presented to illuminate the neuroprotective roles that docosahexaenoic acid (DHA), and derivatives thereof, play in ischemia-reperfusion (IR) injury and in neurodegenerative diseases. Also, to give a more mechanistic understanding of how lipids messengers intervene in death signaling at the pre-mitochondrial level, experiments examining the protective effects of neuroprotectin D1 (NPD1) on retinal pigment epithelial cells and on neural cells exposed to oxidative insults will also discussed.
Host defense against infection or injury depends upon inflammatory responses in order to destroy pathogens and stimulate tissue repair. Unresolved inflammatory responses in the brain are associated with stroke, neurotrauma and neurodegenerations. Inflammation is marked by multiple, concurrent processes including the activation of microglia and astrocytes, adhesion of leukocytes (T-cells, neutrophils, and monocytes/macrophages) to endothelial cells and their subsequent infiltration into the brain parenchyma, as well as release of oxygen- and nitrogen-derived free radicals, cytokines, chemokines, and bioactive lipids by all of these cells. Leukocyte infiltration, in particular, is an important contributing factor to ischemia-reperfusion injury, due to their release of IL-1β, tumor necrosis factor α (TNFα), myeloperoxidase, and pro-inflammatory lipid mediators, such as platelet activating factor and eicosanoids (Basu et al., 2004; Danton and Dietrich, 2003; Hoy et al., 2003). Data from experimental models of ischemia suggest that these acute inflammatory processes are cell damaging (Barone and Feuerstein, 1999). However, the onset of a more delayed inflammation may be reparative (Bethea and Dietrich, 2002; Kerschensteiner et al., 1999.) In IR injury, the inflammatory response occurs hours after the actual ischemic insult is sustained.
Oxygen and glucose deprivation resulting from obstructed blood flow initiates a cascade of distinct pathological events stemming from energy depletion, including the collapse of energy-dependent ion transport, non-specific release of glutamate as well as inhibition of its uptake, intracellular Ca2+ overload, and generation of reactive oxygen species (ROS) (Bramlett and Dietrich, 2004; Choi and Rothman, 1990). Generation of ROS after IR injury can be attributed to an overload of Ca2+ into mitochondria: reduction of proximal electron carriers in the inner membrane leads to partial reduction of molecular oxygen, forming the superoxide anion and its more toxic derivatives H2O2 and•OH (Dykens, 1994; Kowaltowski et al., 1995). Enzymatic sources of free radical species under these conditions include the cyclooxygenases (COX), lipoxygenases, myeloperoxidase, and nitric oxide synthase. Sudden restoration of blood flow in the reperfusion phase of injury places additional demands on already overwhelmed anti-oxidant defenses. ROS can modify both free and protein-bound amino acids leading to alterations in enzyme activity and susceptibility to proteolysis (Jesberger and Richardson, 1991). Nuclear and mitochondrial DNA is also targeted by ROS, resulting in highly mutagenic base modifications, abasic sites, and strand breaks (Seeberg et al., 1995). Among the most damaging effects of ROS is lipid peroxidation. Hydroxyl radical attack on fatty acid side chains generates carbon-centered radicals and subsequently, lipoperoxyl radicals capable of attacking adjacent fatty acids and propagating further damage (Halliwell, 1991). Of note, the high unsaturated lipid content of neuronal membranes renders neurons excellent at propagating free radical species. Accumulation of lipid hydroperoxides alters membrane permeability and fluidity and oxidizes membrane proteins, leading to alterations in ion transport, notably the intracellular flux of Ca2+ (Mattson, 1998).
DOCOSAHEXAENOYL CHAINS OF MEMBRANE PHOSPHOLIPIDS ARE RESERVOIRS FOR LIPID MESSENGERS
Membrane organization has conceptually evolved from the notion of a lipid bilayer with embedded proteins to that of a highly dynamic, heterogeneous patchwork of microdomains containing ion channels, receptors, transporters, and other proteins. In the past, cellular membranes in the nervous system were categorized as being fluid membranes (e.g., those of cells of gray matter) or rigid membranes (e.g., the oligodendrocyte plasma membrane that spirals around the axon to form myelin) based on a higher or lower content of polyunsaturated fatty acids (PUFA) in phospholipids, respectively. We now know that neurons, glia, and endothelial cells of the cerebrovasculature are endowed with phospholipid pools that serve as reservoirs of lipid messengers. Specific lipid messengers are cleaved and released from these reservoir phospholipids by a class of proteins known as the phospholipases in response to signals such as neurotrophic factors, cytokines, membrane depolarization, ion channel activation, and neurotransmitters such as glutamate. These lipid messengers can act intracellularly or in an autocrine and/or paracrine fashion to regulate other signaling cascades, thereby contributing to the development, differentiation, function, protection, and repair of the cells of the nervous system (Bazan, 2003). Our laboratory has devoted considerable effort to sorting out specific signals, mainly those of PUFA and their peroxidation products, generated during IR, and the neuroprotective effects of these compounds.
Phospholipids consist of a glycerol backbone with a hydrophilic phosphate-containing head at sn3. Various saturated and unsaturated fatty acids occupy the sn1 and sn2 positions, respectively. The latter position is targeted by a family of acylhydrolases known as phospholipase A2 (PLA2), which liberate the PUFA situated here (Horrocks and Farooqui, 1994; Sun et al., 2004; Capper and Marshall, 2001). Massive influx of Ca2+ during IR triggers PLA2 activation as reflected by a rapid accumulation of n-6 and n-3 PUFA, specifically arachidonic acid (AA; 20:4n-6) and DHA (22:6n-3) (Horrocks and Farooqui, 1994; Sun et al., 2004; Marcheselli et al., 2003). PLA2 activation can also be triggered by the pro-inflammatory cytokines, IL-1β and TNFα(Anthonsen et al., 2001).
Arachidonic acid, an n-6 PUFA, is synthesized from dietary sources of linoleic acid, whereas the n-3 PUFA (including DHA) are synthesized from α-linoleic acid (Calder and Grimble, 2002). Interconversion of these two fatty acid families is not possible in mammalian tissues.
OMEGA-3 FATTY ACIDS DISPLAY ANTI-INFLAMMATORY ACTIVITIES
Fish oil is the major source of the omega-3 fatty acids eicosapentaenoic acid (EPA; 20:5n-3) and docosahexaenoic acid (DHA; 22:6n-3). In the absence of fish oil, α-linolenic acid is the precursor to EPA and DHA. Increased consumption of oily fish increases EPA- and DHA-containing phospholipids at the expense of decreased AA incorporation into phospholipids (Calder and Grimble, 2002). Dietary supplementation of EPA and DHA has beneficial effects in many diverse disorders including asthma, heart disease, rheumatoid arthritis, cancer, mental depression, and transplant rejection (Billman et al., 1999; Grimminger et al., 1996; Hibbeln, 1998; Iigo et al., 1997; James et al., 2000; McLennan et al., 1996; Marchioli, 1999; Rapp et al., 1991; Stephenson, 2004). In some of these conditions, such as cardiovascular disease, the benefits of fish oil have been attributed solely to DHA (McLennan et al., 1996). For many years, the positive effects of DHA had been attributed to its ability to antagonize the production of AA and its derivatives, the eicosanoids (Calder and Grimble, 2002). However, whether these effects could be due to specific lipid products derived from DHA was unknown.
DHA-DERIVED DOCOSATRIENES ARE ANTI-INFLAMMATORY SIGNALS
Anti-Inflammatory Effects of NPD1 in IR Iinjury
Free DHA generated from membrane phospholipids following reperfusion is met with the appearance of a novel stereospecific DHA-derived messenger, 10, 17S-docosatriene or neuroprotectin D1 (NPD1). This docosatriene was also recently described as being present in human blood, glial cells, and mouse brain (Hong et al., 2003). NPD1 attenuates classical features of inflammation, namely polymorphonuclear neutrophil (PMN) infiltration and pro-inflammatory gene signaling. Administration of exogenous NPD1 inhibits PMN infiltration (Marcheselli et al., 2003). Likewise, continuous infusion of either DHA or NPD1 into the third ventricle during the intial two days of reperfusion inhibits PMN infiltration into the hippocampus (one of the most vulnerable regions to ischemic injury) and the neocortex. IR-induced increases in NF-κB binding activity and up-regulation of COX-2 mRNA was also attenuated by DHA and NPD1 infusion. These results were mirrored by cultured human neural progenitor cells treated with the pro-inflammatory cytokine, IL-1β; this cytokine was chosen, since it increases during IR as a result of PMN infiltration (Marcheselli et al., 2003). The significance NPD1's anti-inflammatory effects is underscored by its ability to reduce the infarct volume by approximately half, demonstrating the considerable neuroprotective effects of this DHA-derived lipid after IR injury. Of note, NPD1 levels were bolstered by exogenous administration of DHA, and infusion of either DHA or NPD1 led to indistinguishable outcomes in the PMN infiltration and pro-inflammatory signaling measures described above. This implies that exogenous DHA is used as a precursor for NPD1 synthesis under these conditions and points to NPD1 as a possible mediator of DHA's observed anti-inflammatory effects in other conditions. The fact that brain responses to IR injury, including synthesis of DHA-derived messengers such as NPD1, were unsuccessful in preventing death suggests that they were counterbalanced by pro-death signals, lipid-derived and otherwise. Depending on its magnitude, injury-induced protective lipid-signaling responses may be overwhelmed. We know that DHA is susceptible to oxidation and yields neuroprostanes during IR (Roberts et al., 1998). Therefore, given its propensity for peroxidation, formation of neuroprotective, DHA-derived messengers such as NPD1 may be thwarted under conditions of severe oxidative stress.
Given the anti-inflammatory effects of DHA and DHA-derived lipids, the arguments for a diet rich in α-linoleic acid or DHA-containing fish oils are apparent. However, altering phospholipids content in the brain may not be so easy. Due to the very high content of brain DHA and due to its tenacious retention, it is difficult to modify brain content by simple dietary manipulation. In addition, there seems to be a specific liver-to-brain (and retina) DHA-supply system that provides DHA for the biogenesis and repair of membranes. It has been postulated that when ischemia removes free DHA from brain, its replenishment may be met through DHA-carrying blood proteins. Which blood proteins may perform this function is unclear, but albumin is a possible candidate. Albumin has been shown to induce mobilization of n-3 PUFA and is thought to replenish these PUFA lost from the membrane after ischemia (Belayev et al., 2005). In addition, human serum albumin, when systemically injected, does cross the blood-brain-barrier, reaching even intraneuronal sites, and elicits neuroprotection in IR in experimental animals (Ginsberg, 2003; Remmers et al., 1999). In any case, the identification of the proteins involved in the maintenance and replenishment of DHA-containing PUFA in the brain may prove to be useful targets for therapeutic interventions following ischemia.
Effect of NPD1 on Mitochondrial Apoptotic Pathways
Oxidative injury, depending on its magnitude, can lead to varied forms of cell death. Necrosis refers to the immediate, uncontrolled cellular disintegration resulting from an acute injury such as would occur in the ischemic core. This type of death elicits an inflammatory response that involves leukocyte invasion and edema with damage to surrounding tissues. Areas surrounding the core, also known as the penumbra, are vulnerable to a more delayed death that retains features of apoptosis, a programmed cell death exemplified by the developing nervous system. Neuronal apoptosis is regarded as a major cause of loss-of-function associated with not only brain ischemia, but also aging, Alzheimer's disease, amyotropic lateral sclerosis, traumatic brain injury, and spinal cord injury. Apoptosis is characterized by cytoplasmic condensation, cell shrinkage, and membrane blebbing. It is consistently associated with fragmentation of nuclear DNA into 120-200 base pair fragments, or multiples thereof, also known as oligonucleosomal fragments, and it is considered by some to be one of the hallmarks of apoptosis (Compton, 1992). Programmed cell death pathways are initiated through engagement of receptors at the surface of the cell, or through the mitochondria, with the latter pathway serving as a feed-forward mechanism for the receptor pathways (Creagh et al., 2003). A family of cysteine proteases known as the caspases (cysteine aspartic acid-specific proteases) orchestrate the orderly breakdown of cells during apoptosis, and the mitochondrial pathway of caspase activation is triggered in response to a variety of cellular stresses such as growth factor withdrawal, heat shock, DNA damage, and oxidative stress (Creagh et al., 2003). TNFα/H2O2 injury of ARPE-19 cells (spontaneously transformed human retinal pigment epithelial cells [RPE]) is a well-established model of apoptosis triggered through the mitochondrial pathway (Mukherjee et al., 2004). Accordingly, we took advantage of this model to expand our investigation of NPD1 by examining its effects on apoptotic mediators (caspases) and regulators (Bcl-2 protein family).
Before discussing findings related to NPD1 and mitochondrial pathways of apoptosis, a discussion of the clinical significance of oxidative stress-induced RPE death is in order. Retinal pigment epithelial cells (RPE) cells are derived from neuroectoderm and form a monolayer above tips of the photoreceptor outer segments. RPE cells perform functions vital to photoreceptor survival including the transport and re-isomerization of bleached visual pigments, the maintenance of the blood-outer retinal barrier, and the recycling of rod outer segments (Hu and Bok, 2001). Photoreceptor outer segments contain rhodopsin and the highest content of DHA of any cell type (Bazan, 1990; Anderson et al., 2002). In a daily cycle, RPE cells engulf and phagocytize the distal tips of photoreceptor outer segments. This process is tightly regulated so that photoreceptor length and phospholipid composition are maintained (Bazan et al., 1985; Chen et al., 1996; Gordon et al., 1992; Stinson et al., 1991). Thus, given photoreceptors’ high DHA content, RPE cells are particularly pertinent to this discussion since they are strategically positioned to be targets of DHA-derived messengers. Importantly, oxidative stress-induced apoptosis of RPE cells compromises photoreceptor survival and impairs vision. These processes are clinically applicable to age-related macular degeneration and Stargardt disease (Sieving et al., 2001; Sparrow et al., 2003).
Experiments examining the effect of NPD1 on apoptosis in ARPE-19 cells revealed that inclusion of this lipid at nanomolar concentrations inhibits apoptosis induced by TNFα/H2O2 injury as measured by decreased nuclear condensation and DNA fragmentation (Mukherjee et al., 2004). This protective effect was accompanied by an attenuation of caspase-3 activation. Next, NPD1's effects on levels of the pro-apoptotic Bcl-2 family members (Bad and Bax), as well as on those of anti-apoptotic members (Bcl-2 and Bcl-xL) were examined. Bcl-2 family proteins participate in the initiation and amplification of pre-mitochondrial events in the apoptotic cascade, primarily through their ability to regulate cytochrome c release from the mitochondrial intermembrane space, which triggers caspase activation. Levels of the anti-apoptotic protein Bcl-xL increased in response to TNFα/H2O2, but those of Bcl-2 did not. Inclusion of NPD1 enhanced levels of both proteins after injury. The pro-apoptotic proteins Bax and Bad were up-regulated by TNFα/H2O2, and this up-regulation was lessened by the inclusion of NPD1. NPD1 was also capable of inhibiting increased promoter activity of the COX-2 gene in ARPE-19 cells treated with IL-1β. NPD1's ability to act upstream of caspase activation by influencing expression of the Bcl-2 proteins suggests that it is part of an immediate, early response to injury that targets pre-mitochondrial events. It is possible that NPD1 modulates signaling at the transcription factor level and regulates promoters of the genes encoding death repressors and effectors of the Bcl-2 family of proteins. This is a plausible explanation, given that many of the Bcl-2 family members contain NF-κB consensus sequences in their promoters (Catz and Johnson, 2001; Dixon et al., 1997; Glasgow et al., 2001). Alternatively, translational or post-translational events may integrate a concerted response to counteract oxidative stress. This newly uncovered “NPD1 regulatory pathway” may aid our understanding of the effects of other neuroprotective mediators in retinal degeneration. For example, a connection between NPD1 and certain growth factors, particularly fibroblast growth factor 2, which is important to photoreceptor survival, should be explored (Brckaert et al., 1999). In any case, the findings related to the effects of NPD1 on RPE survival-signaling pathways have several implications related to the understanding of how NPD1, endogenously synthesized in the brain, modulates IR injury responses.
As discussed earlier, the primary insult arising from IR is oxidative in nature (combined with an excitotoxic component) with a secondary wave of inflammatory processes providing additional injury. As such, our findings related to TNFα/H2O2 injury of ARPE-19 can be extended to IR injury. That is, NPD1-mediated coordinate regulation of Bcl-2 family members and inhibition of caspase activation may be occurring during IR as well. That NPD1 negatively regulates NF-κB activity and, as a consequence, leads to down-regulation of pro-inflammatory gene expression in IR is consistent with the in vitro findings regarding COX-2 promoter activity and suggests that NPD1 acts at or above the level of transcription factor regulation. Other transcription factors, such as AP-1 and p53, have been implicated in IR injury, and NPD1's protective effects may also be explained, at least in part, by its ability to regulate their activity (Halterman et al., 1999; Mattson, 2000). In the same way, NPD1 may have positive effects on pro-survival signaling pathways such as those involving protein kinase B (Akt) (Sugawara et al., 2004). Like other bioactive lipids, such as platelet-activating factor (PAF, 1-O-alkyl-2-acetyl-glycero-3-phosphocholine), NPD1 may achieve its gene-regulatory effects through an intracellular and/or extracellular receptor. The identification of any such receptor(s) will be essential from a therapeutic standpoint and awaits further characterization. Also, given the short-lived nature of lipid signals, another important aspect of NPD1 function worthy of investigation is its spatial distribution, both intra- and inter-cellular, after oxidative injury. Release of NPD1 may elicit or amplify survival signaling by the same cell in an autocrine fashion and/or it may act on neighboring cells, counteracting their production of pro-inflammatory gene expression. Regulation of the signals that control NPD1 synthesis and degradation will also yield insights necessary to take advantage of maximal therapeutic potential of NPD1.
The newly identified NPD1 elicits potent counter-regulatory actions on molecular signaling after oxidative stress-induced retinal cell injury, brain ischemia-reperfusion and beta-amyloid–triggered neural cell damage. Thus, NPD1 elicits protection in experimental paradigms where oxidative stress and neurodegeneration are enhanced. These findings imply that at least some of the DHA's neuroprotective properties are not due to its ability to antagonize AA synthesis and, therefore, prostanoid production; instead, specific lipid messengers, like NPD1, are synthesized from DHA and have counter-inflammatory actions in their own right. Moreover, other members of this family of messengers are also likely to occur. The findings of our laboratory highlight a fundamental property of the CNS when confronted with injury, i.e., a response with a plethora of signals, some of which are harmful and some of which are protective, and the preponderance of signals in either direction will influence the global outcome (neuronal survival or death). Therapeutic strategies targeted to enhance the synthesis and inhibit the degradation of DHA-derived messengers will tip the balance in favor of neuroprotection in IR injury and, very possibly, other CNS disorders with a neuroinflammatory component. The potent bioactivity of NPD1 makes it a particularly important target for therapeutic, neuroprotective interventions in these diseases. Finally, other as yet un-characterized bioactive lipids with neuroprotective actions undoubtedly exist. The discovery of these messengers will further improve our understanding of the therapeutic possibilities of docosanoids for neuroinflammatory diseases.