|Home | About | Journals | Submit | Contact Us | Français|
Inflammation appears to play a role in the pathogenesis of perinatal brain damage in fetuses/infants born much before term. We raise the possibility that non-inflammatory phenomena induce endoplasmic reticulum (ER) stress, which, in turn, leads to the unfolded protein response (UPR), which is followed by apoptosis-promoting processes and inflammation. Perhaps by these events, non-inflammatory stimuli lead to perinatal brain damage.
Perinatal brain white matter damage, especially in the preterm newborn, is likely to have multiple causes (1). White-matter cells that play a key role in the pathophysiology of periventricular white matter damage (PWMD) in animal models include preoligodendrocytes (2–4) and microglia/macrophages (5–7). Preoligodendrocytes, the precursors of myelinating oligodendrocytes (OLs), which constitute a major glial population in the white matter, are highly vulnerable to various insults. Antenatal infection and subsequent inflammation comprise one etiologic pathway that links preterm delivery and brain/lung damage in preterm newborns (8). Some of the damage probably reflects extra-cellular signalling (with such proteins as cytokines and growth factors) (9), as well as inter-cellular signalling, (e.g., microglial-oligodendrocyte communication). In this paper, we go inside the cell in search of intra-cellular mechanisms. The central topic of this paper is the inter-relationships between inflammation and endoplasmic reticulum (ER) stress (Figure 1). We raise the possibility that ER stress contributes to perinatal brain damage in the infant born much before term.
The ER is an intracellular organelle consisting of a membranous network that extends throughout the cytoplasm of the cell and is contiguous with the nuclear envelope. In eukaryotic cells, the ER is the site of three major physiological functions: protein folding, intracellular Ca2+ storage, and synthesis of unsaturated fatty acids, sterols, and phospholipids (10). Under various pathological conditions, the perturbation of any of the three physiological functions of the ER results in so-called ER stress and can lead to an accumulation of the newly synthesized unfolded and misfolded proteins in the ER. Among the stimuli that promote ER stress are ischemia, inflammation, and disorders that impair the ER’s ability to properly fold polypeptides or eliminate unfolded proteins (11).
The ER responds to the presence of misfolded proteins in two ways that influence one another. In the first, called the unfolded protein response (UPR) and a focus of this paper, homeostasis is achieved by reducing protein synthesis, restoring the protein-folding capacity of the ER, and clearing away the debris (11). The second, called ER-associated degradation (ERAD), also known as the misfolded protein response (12), specifically recognizes terminally misfolded proteins and moves them across the ER membrane into the cytosol, where they can be degraded (13).
These two systems are intimately linked: UPR induction increases ERAD capacity, loss of ERAD induces UPR, and simultaneous impairments of ERAD and the UPR greatly decrease cell viability.
The next 18 sections deal with pathways identified in the figure, which illustrates the interrelationships between inflammation and ER stress, the central topic of this paper. The figure remotely overlaps with, and expands upon, an illustration from a paper that focused on a role for mitochondria in damage to the immature brain (14).
Events like disorders or insults can activate the unfolded protein response (UPR) signal transduction pathways.
The UPR has at least 3 main pathways, each named for a membrane-anchored protein whose cytoplasmic domain is released from the ER when it is stressed:
Professional secretory cells of endocrine and exocrine organs are well equipped to synthesize large amounts of protein. Nevertheless, a general axiom is that the higher the protein synthesis rate, the higher the probability of ER stress.
In the adult, oligodendrocytes synthesize approximately 100,000 myelin-specific proteins per minute (19). Mature oligodendrocytes are very sensitive to ER stress (20,21). The protein-synthesis demands made on immature oligodendrocytes or their precursors are currently unknown. We raise the possibility that the neonatal brain is especially vulnerable to ER stress because of the abundant protein synthesis by preoligodendrocytes and maturing oligodendrocytes.
Efficient folding of many newly synthesized proteins depends on assistance from molecular chaperones, which serve to prevent protein misfolding and aggregation in the crowded environment of the cell. Nascent chain-binding chaperones, including trigger factor, Hsp70, and prefoldin, stabilize elongating chains on ribosomes in a nonaggregated state. Folding in the cytosol is achieved by these chaperones or after transfer of newly synthesized proteins to downstream chaperones, such as the chaperonins (22).
It is still unclear how the ER recognizes misfolded proteins. Current evidence suggests that all three major membrane-anchored proteins (PERK, ATF6, and IRE1) function as sensors of stress. The existence of multiple UPR sensors in higher eukaryotes allows for a more nuanced response to misfolded proteins.
New studies suggest a model by which distinct combinations of individual UPR signalling pathways determine a cell’s fate after ER stress. The switch between cytoprotective and also pro-apoptotic output seems to lie in part in the duration of individual UPR branch activity (23). The initial combined activation of IRE1, PERK, and ATF6 produces cytoprotective outputs such as reduced translation, enhanced ER protein folding capacity, and clearance of misfolded ER proteins, along with pro-apoptotic outputs such as CHOP (C/EBP homologous protein) production (23). Cytoprotective outputs would outweigh pro-apoptotic factors at this point, which would be helped by the relatively longer mRNA and protein half-lives of factors such as immunoglobulin heavy chain binding protein (BiP) (24). This phase of predominantly beneficial UPR output would thus provide a “window of opportunity” for cells to readjust their ER to cope with stress. If these steps fail to re-establish homeostasis, IRE1 and ATF6 signalling are attenuated, creating an imbalance in which unchecked pro-apoptotic output guides the cell toward its demise (23).
Low–level activation of the UPR is accompanied by changes in the patterns of protein expressions that are qualitatively distinct from the UPR induced by severe acute stress (24). Adaptation to chronic stress is a consequence not of selective activation of proximal ER stress sensors, but of preferential stabilization of mRNAs and proteins that facilitate adaptation.
Inflammation is a major component of perinatal damage to the lung (25) and brain (1,26), and might be a common final pathway in both (8). In the following sections, we discuss the bidirectional links between inflammation and the UPR.
Lipopolysaccharide can have numerous, sometimes opposite, effects on the ER. First, it can increase the expression level of chaperones in the ER lumen (29). This should be expected to decrease ER stress. Second, it can cause ER stress as documented by the activation of PERK, one of the proteins in the UPR, and even ER disintegration (30). In mice, systemic lipopolysaccharide activates lung CHOP, another gene involved in the UPR (31). Third, it can prevent ER stress and resulting apoptosis induced by some toxins (32).
CHOP, as well as the 3 proteins with pathway names (PERK, ATF6, IRE1), are not the only ones that mediate the UPR. CREBH (cAMP-responsive element binding protein H), a new member of the membrane bound transcription factor family that is structurally similar to ATF6, is liver specific and can be detected as early as mid-gestation (33). After being cleaved in response to ER stress, the active component transits to the nucleus, where it activates expression of acute phase response genes, such as those encoding C-reactive protein (CRP).
Observations that non-inflammatory phenomena contribute to ER stress have led to the “intriguing notion that many physiological and pathological processes that induce ER stress … can induce an inflammatory response through CREBH cleavage and CRP production” (33). The expression of CREBH is induced by proinflammatory cytokines and ER stress induces cleavage of CREBH. The CREBH does not contribute to the classical UPR induction but it is required for the acute phase response by regulating transcription of CRP and other acute phase reactants. CREBH and ATF6 bind to a conserved promoter element in the specific acute phase response genes where they interact and synergistically activate transcription of target genes upon ER stress (33). Knowing that proinflammatory cytokines induce cleavage of CREBH and activate the acute phase response and the UPR in the liver in vivo, we consider the following sequence and inference:
Under conditions of prolonged ER stress, the UPR, unable to maintain ER homeostasis, triggers apoptosis. Multiple pathways may be involved in ER-stress mediated apoptosis, including direct activation of caspases in the ER, transcriptional activation of the apoptotic program through the modulation of BCL-2 (B-cell lymphoma-2) members or death receptors, and the mitochondrial pathway involving the upregulation of the pro-apoptotic and down-regulation of the anti-apoptotic BCL-2 members of proteins (38,39). These apoptotic responses are activated via the important Ca2+-mediated “ER-mitochondrial crosstalk” (39), which is facilitated/regulated by mitochondria (40,41). The BCL-2 proteins are important regulators of apoptosis at both the ER, and the mitochondria, and may have similar functions in both organelles (38).
The complex UPR is a pro-survival, cytoprotective response to reduce the accumulation of unfolded and misfolded proteins, and to restore the ER to its normal state. However, when a stressor is so strong or persistent that ER function cannot be restored, apoptosis occurs, leading to targeted removal of cells that are irreparably damaged (42). Given the central role of the three ER transmembrane receptors (Pathway 1) in UPR signalling, it is likely that these mediators are also fundamental to ER stress-induced apoptosis (43,44). Activation of PERK is initially protective. However, the activation of PERK also leads to the induction of CHOP, an important element of the switch from pro-survival to pro-death signalling (44). Although ATF6 can induce CHOP mRNA expression, it seems that ATF6-mediated signals are purely pro-survival (11), while IRE1 reveals both pro- and anti-apoptotic functions (44). Although these crucial mediators can trigger pro-apoptotic signals, they do not directly cause cell death, but rather initiate downstream molecules such as CHOP, JNK, or BCL-2, which further push the cell down the path of death. Interestingly, recent data suggest that Bax inhibitor-1 (BI-1), an ER resident protein interacting with the pro-apoptotic protein BCL-2 associated X protein (Bax) and known to neutralize Bax’s pro-apoptotic activity, can inhibit the IRE1α-dependent branch of the UPR (45). The fact that the UPR is suppressed by BI-1 under stress conditions seems to contradict its well-documented anti-apoptotic activity (45).
Hypoxia-ischemia, a purported mechanism of perinatal brain damage (14), results in depletion of cellular energy stores, depolarization of neurons and glia, and the release of excitatory amino acids into the extracellular space. Energy-dependent reuptake mechanisms become compromised allowing glutamate to accumulate to excitotoxic levels (46–48). Glutamate is capable of inducing maturation-dependent death of premyelinating oligodendrocytes (pre-OLs). The loss of pre-OLs characterizes the cerebral white matter injury, which is the most common form of injury to the preterm brain and is associated with high risk of neurodevelopmental impairment (49). Excitotoxic cell death can also occur in all neurons possessing glutamate receptors as well as in glial cells including astrocytes and oligodendrocytes. Oligodendrocytes appear to be highly vulnerable to excitotoxic overactivation and a physiologic functioning of glutamate homeostasis seems to be crucial to prevent glutamate-induced damage to oligodendrocytes (50).
In the developing brain, hypoxia increases anaerobic metabolism, leading to a rapid rise in levels of lactic acid and reactive oxygen species (ROS) production (51), and/or impaired function of the antioxidant defense mechanism (52,53). Preterm newborns are at high risk of oxidative stress and are very susceptible to free radical oxidative damage (54). In the very immature brain, oxidative stress is an important factor in the response of the immature brain to hypoxic-ischemic injury (55). Additionally, oligodendrocyte progenitor cells and oligodendrocyte precursors are selectively vulnerable to the depletion of antioxidants or exposure to free radicals (56).
The pattern of brain oxidative injury is dependent on the developmental stage of the brain. Increased release of glutamate and glutamate action has been implicated in rodent studies of preterm white matter injury (3,57,58). Glutamate-induced injury to pre-OLs is mediated by ionotropic glutamate receptors (iGluRs) of the AMPA/kainate type (59–62). Activation of mGluRs attenuates oligodendrocyte excitotoxicity by controlling downstream oxidative stress after iGluR overactivation (57). Excessive extracellular glutamate causes glutathione depletion in pre-OLs with consequent free radical generation (63). Reactive oxygen and nitrogen species are found in damaged white matter of human newborns (64).
Perinatal insults can disrupt synaptic function leading to accumulation of extracellular glutamate and excessive stimulation of the several groups of glutamate receptors (NMDA, AMPA, kainate, and metabotropic receptors (mGluRs)). Activation of NMDA receptors does not appear to significantly modulate ER stress signalling pathways (65). On the other hand, another excitatory transmitter, kainic acid (KA), activates ER sensors that are part of the PERK, ATF6, and IRE1 pathways (66). The mode of cell death caused by KA includes the activation of caspases as well as changes in expression of various pro- and anti-apoptotic molecules (67). KA induces fragmentation of the ER membrane in neurons, which in turn activates the ER proteins Bip and CHOP.
Although controversial, one view is that the resulting cleavage of caspase-12 is the death-driving force in ER stress (66,68). The apoptosome, a large ternary protein structure formed in the process of apoptosis, acts upstream of caspase-12 cleavage resulting in cell death that does not require caspase-12 activity (69,70).The execution of cell death prompted by ER stress can also be achieved via other caspases, e.g. caspase-2 (71), caspase-3 (72), or caspase-9 (73).
Glutamate, a major excitatory neurotransmitter in the brain, can cause excitotoxic injury to the nervous system under many pathological circumstances. Emerging evidence indicates that oligodendrocytes share with neurons a high vulnerability to excitotoxic injury (74–76).Developing oligodendrocytes (pre-OLs) appear to be more sensitive to excitotoxic death than mature OLs (3,74–76), and pre-OL excitotoxicity is implicated in the pathogenesis of periventricular leukomalacia (PVL), a leading cause of cerebral palsy in premature infants (2,3,57,74). Additionally, the glutamate-receptor agonist kainic acid triggers a delayed type of excitotoxic cell death in vulnerable brain regions (67,77,78) that results in an ER stress response with the activation of the PERK, ATF6, and IRE1 pathways (66).
The formation of reactive oxygen species might be part of early events leading to mitochondrial impairment, excitotoxicity, and apoptosis (14). Immature brains are much more likely than adults to release pro-apoptotic proteins from mitochondria in response to oxidative stress (14).
Increasing evidence supports the view that mitochondria are involved in the apoptosis that follows unresolved ER stress (48,79–81). Some of the apoptosis is due to the release of cytochrome c from the mitochondria, which is regulated by the BCL-2 family of proteins, including Bak (BCL2 homologous antagonist/killer) and Bax (39,79). Although the mitochondrion remains the principle subcellular target of the BCL-2 protein family, the recent finding that certain members of the BCL-2 family are present on the ER where they seem to have a much broader and comprehensive function could lead to the speculation that activities of the proteins at the ER might be more relevant to sustaining cellular function and possibly setting the threshold of sensitivity to apoptotic stimuli (39). While it is generally accepted that mitochondrial expression of these proteins suffices for apoptosis, it was only discovered very recently that exclusive reticular expression of Bak is sufficient to fully trigger mitochondrial apoptosis (79). These findings support the involvement of an ER-to-mitochondria apoptotic pathway and reveal that, by not relying on mitochondrial Bak and Bax, ER-localized multidomain effectors might function as autonomous integration sites for apoptotic signals, thus adding a new layer of complexity to the regulation of apoptosis (79).
The efficient functioning of the ER is essential for most cellular activities and survival. The accumulation of unfolded proteins in the ER represents a cellular stress induced by multiple stimuli and pathological conditions. If the ER stress is persistent and the UPR cannot suppress the ER stress, signaling switches from pro-survival to pro-apoptotic. A number of pathways are involved in ER-stress induced apoptosis, including direct activation of initiator caspases at the ER, transcriptional activation of the apoptotic program through the modulation of BCL-2 family members or death receptors, and apoptosis through ER-mitochondrial crosstalk. Increasing evidence of an essential role for the three functional groups of proteins comprising the BCL-2 family members at the ER is becoming apparent (38,45,82,83). BCL-2 family proteins at the ER regulate both direct modulation of signalling pathways leading to caspase activation and through the modulation of ER-mitochondria Ca2+ crosstalk signalling (84). Additionally, the BCL-2 family appears to play essential roles in ER apoptosis and in the regulation of autophagy, a degradative survival that limits metabolic stress, genomic instability, and tumorigenesis (38).
Two major pathways of apoptosis have been recognized as the “intrinsic” mitochondrial pathway and the “extrinsic” death receptor pathway. At the core of the mitochondrial pathway is the process of mitochondrial outer membrane permeabilization (MOMP) mediated predominantly by the pro-apoptotic BCL-2 family members, Bax and Bak, and inhibited by the anti-apoptotic BCL-2 family proteins (85). The release of cytochrome c from the mitochondrial membrane space triggers the activation apoptosis protease-activating factor 1 (APAF-1) within the apoptosome, the large protein structure formed in the process of apoptosis, leading to a cascade of caspases that achieve apoptosis by cleaving key substrates (86). Caspase-independent pathways can also involve pro-apoptotic proteins such as apoptosis inducing factor (AIF), endonucleases G, and BNIP3 (87,88).
By contrast, the extrinsic cell-death pathway can function independently of mitochondria and is activated by cell-surface receptors. These cell surface death receptors belong to the tumor necrosis factor receptor (TNFR) superfamily, and include TNF-R1, Fas (also called CD95), death receptor 3 (DR3), and two tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) receptors (89). A shared feature of death receptors is the death domain in the cytoplasmic tail of these molecules. Ligand binding to the death domain leads to the creation of a death-inducing signalling complex (DISC) directly activating the caspase cascade via an initiator caspase (caspase-8) resulting in the cleavage of caspase-3 in the execution phase of apoptosis (90).
Apoptosis, needed for eliminating non-functional cells during normal brain development with a minimum of inflammation, plays a prominent role in neonatal brain injury (91–93). Immature neurons appear to be more prone to apoptotic death than terminally differentiated neurons (94). Additionally, signals from TNF/Fas receptors result in apoptosis when endogenous inhibitors of apoptosis are deficient. Fas-dependent signalling pathways can contribute to apoptosis in the immature brain in response to hypoxia-ischemia (95,96) and head trauma (97,98). In addition, Fas and its ligand FasL have been detected in neonatal cerebrospinal fluid (99,100).
The upregulation of pro-inflammatory cytokines (e.g. TNF-α, interferon-gamma, IL-1, IL-6, and IL- 10) in response to a strong pro-inflammatory challenge can lead to deleterious effects in the developing brain (101). TNF-α receptors have been identified in human brains with PVL, raising the possibility that activation of these TNF-α receptors contributes to inflammatory reactions and apoptosis (102). Indeed, apoptosis is prominent in the brains of infants who died with white matter injury (103). Exposing developing oligodendrocytes to TNF increases apoptosis (104) and reduced staining for myelin basic protein (MBP) (105), which might help explain the reduced myelination that is considered a hallmark of inflammation-associated white matter damage in preterm infants (106).
Additional essential regulators of inflammation are enzymes termed “the inflammatory caspases”. They are activated by cellular sensors of danger signals, the inflammasomes, and subsequently convert pro-inflammatory cytokines into their mature active forms (107). By acting as key regulators of inflammation, energy metabolism and cell death, inflammatory caspases and inflammasomes exert profound influences on innate immunity and infectious and non-infectious inflammatory diseases (107).
Hypoxia-ischemia has been the experimental paradigm for perinatal brain damage over the past decades. The reader is referred to Hagberg, Peebles & Mallard (108) for a comprehensive overview.
The relationships between hypoxia-ischemia and inflammation are highly complex. Hypoxia-ischemia and inflammation each alone (109), as well as together (preconditioning or sensitizing) (110,111), can contribute to brain damage. Preconditioning, also known as tolerance, constitutes the phenomenon where exposure to a sublethal insult, prior to a severe insult, attenuates the pathology. Protective in a number of organs and in both adult and immature animals, preconditioning has also followed many different kinds of sublethal insults, including hypoxia, ischemia, inflammatory cytokines, and endotoxins (110). It is not necessary for the sublethal event to be the same type as the secondary insult (so-called cross-tolerance). The underlying mechanisms of preconditioning remain largely unknown.
Sensitization, also known as negative preconditioning, occurs when a sublethal insult exacerbates the damage of a subsequent severe insult. Since both, inflammation and hypoxia-ischemia have the ability to contribute to ER stress and therefore to activate the UPR pathways, these propensities might be linked.
Preterm infants exposed to chorioamnionitis (36), fetal vasculitis (35), high concentrations of proinflammatory cytokines in the amniotic fluid (37) or fetal blood (112) are at increased risk of white matter injury and/or cerebral palsy (CP). Intracerebral and systemic lipopolysaccharide injections induce inflammatory responses and prominent white matter lesions (for an overview see (26)). Taken together, the evidence from both observational and experimental research suggests that the association between inflammation and brain injury is causal (1).
Endoplasmic stress and the UPR have been implicated in a number of neurologic disorders, including Alzheimer’s disease (113), amyotrophic lateral sclerosis (114) in adults, the leukodystrophy Pelizaeus-Merzbacher disease (115) and the vanishing white matter disease (116) in children.
Vanishing white matter disease is one of the most prevalent autosomal-recessive childhood leucoencephalopathies in which febrile infections or minor head trauma may provoke major neurologic deterioration leading to cerebellar ataxia and spasticitiy. All three main UPR pathways mentioned in pathway 1 are activated in patients with vanishing white matter disease (117). The activation seems to be restricted to the white matter and enhanced expression of UPR markers (e.g. ATF6, BiP) is almost exclusively observed in the oligodendrocytes and astrocytes. These cells are known to be primarily involved in the pathology of vanishing white matter disease (117).
Pelizaeus-Merzbacher leucodystrophy is a neurodegenerative disease causing diffuse hypomyelination of the central nervous system. The severity and onset of this X-linked recessive pediatric disorder ranges widely and extends from the mild, adult-onset spastic paraplegia to the severe form with onset at infancy and death in early childhood. One possible pathogenic mechanism could be the UPR activation and signaling through the UPR downstream effector molecule CHOP (118). This CHOP protein, widely known as a pro-apoptotic transcription factor, modulates the pathogenesis in different mouse models of Pelizaeus-Merzbacher disease, indicating that this protein might be a key regulatory component (118).
We extend the possibility of UPR-induced apoptosis to the brain of the most vulnerable humans, those born much before term. Oligodendrocytes produce a vast amount of myelin as an extension of their plasma membranes (119). If the immature oligodendrocytes that characterize the cerebral white matter near the end of the second trimester synthesize proteins, they might be prone to ER stress during their active phase, and thereby highly vulnerable to apoptosis via UPR failure leading to brain injury. Thus, it could be that the outcome of ER stress is determined by the developmental status of the cell.
Although infection can lead to a fetal inflammatory response, non-inflammatory stimuli can initiate/promote inflammatory phenomena via ER stress leading to the UPR, which might then lead to brain damage. While recent advances in our understanding intracellular stress responses add yet another level of complexity to this evolving story, it might also open new avenues for intervention design (43). On the one hand, ER stress promotes cell survival in fully myelinated mature oligodendrocytes (120). On the other hand, ER stress leads actively myelinating / remyelinating oligodendrocytes to cell death (121). How the UPR in response to ER stress selectively promotes apoptotic and adaptive pathways remains unknown (42). Perhaps the fate of the (pre-) oligodendrocyte depends on the developmental status of the cell or the amount of protein production. The neonatal brain might be especially vulnerable to ER stress because of the abundant protein synthesis by preoligodendrocytes and myelinating oligodendrocytes. Perhaps acute interventions to limit the UPR can minimize this neonatal brain damage (43). Among the challenges posed by interfering with the UPR are how best to achieve this goal in light of the multiple pathways involved in the apoptosis that follows unresolved UPR, and minimizing unintended disturbances to processes needed for normal development (120). Continued research in this exciting field is needed to improve our understanding of how these pathways are linked and how brain damage that follows activation of these pathways might be minimized.
Financial Support: While preparing this article, authors were supported by grants from the Wilhelm Hirte Stiftung (Hannover), NIH (1 U01 NS 40069-01A2), and the European Union (LSHM-CT-2006-036534).