Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
When the brain is exposed to stress, a number of adaptive responses can act to reestablish homeostasis. However, when the stress is severe and/or the endogenous protective processes are not sufficiently effective, the cell will die. , Cell death can be initiated by numerous organelles (e.g., nucleus, mitochondrion, endoplasmic reticulum/lysosome, cytoskeleton, plasma membrane) and mitochondria, in particular, play a key role in the initiation and execution of cell death mechanisms in the immature brain. , Once triggered, cell death can proceed via a variety of routes, including the well-characterized apoptosis pathway, as well as by lesser-known necroptosis, ferroptosis, parthanatos, and autophagy-dependent cell death mechanisms ( Fig. 125.1 ). Which cell death mechanism predominates will depend on the metabolic state, severity and type of insult, cell type, developmental age, and other factors. , The traditional classification of cell death by morphology is no longer used, as mixed morphologic phenotypes are often detected in many pathologic situations. , Classification now favors the use of molecular markers indicative of particular cell death pathways. Indeed, for effective neuroprotective intervention, understanding the biochemical route(s) to cell death is critical, considering that when one route is inhibited, cell death may occur via an alternative mechanism. , Long-term functional cell recovery following genetic alteration and/or pharmacologic intervention often yields significant information concerning the essential components in a specific route of cell death and any alternative routes triggered. More recently, unbiased proteomics approaches have provided a more comprehensive identification of the major players in a variety of cell death paradigms. , ,
Cell death can be classified into accidental and regulated (see Fig. 125.1 ). Accidental cell death is evoked by severe insults (such as severe trauma, core of an ischemic infarct, high temperature), which cause immediate cellular demise that does not involve a specific molecular mechanism and cannot be prevented or modulated. However, cellular contents released following accidental cell death act as damage-associated molecular patterns (DAMPs) often having direct toxic effects on surrounding cells that survived the initial insult. Exposure to DAMPs may extend the primary injury, as they have immunogenic properties, contributing to an inflammatory response that may cause injury and aggravate the situation further. , Various interventions that attenuate DAMP-induced cellular actions can provide protective effects. So even if accidental cell death cannot be targeted directly, its consequences can be intercepted and bystander injury prevented to some extent. For example, therapeutic hypothermia is used in mitigation of the effects of hypoxic-ischemic injury. In contrast, regulated cell death involves the coordinated molecular machinery of the cell (see Fig. 125.1 ), and its course can be modulated by pharmacologic and genetic means. , Regulated cell death usually occurs with a delay, in situations when endogenous protective mechanisms fail to restore cellular homeostasis.
In the developing brain, cell damage can be induced by a variety of insults, such as hypoxia, hypoxia-ischemia (HI), trauma, and infections. While definitive studies using human neonatal brain tissues are lacking, substantial knowledge of the mechanisms of cell death in the immature central nervous system has emerged from in vitro studies in cell cultures (exposed to toxins or oxygen/glucose deprivation) and in vivo, in models of HI in rodents and to some extent larger animals (e.g., piglets, rabbits, or fetal sheep). Exposure to HI results in an initial depletion of high-energy phosphates, in particular ATP and phosphocreatine. The levels of these phosphates return transiently to the baseline, but this is followed by a second, more prolonged depletion of cellular energy reserves accompanied by progression of brain injury (see Chapter 167 ). , These disturbances in energy metabolism trigger a number of pathophysiologic responses that ultimately lead to cell death, with mechanisms including initial necrosis, and subsequently apoptosis, , necroptosis or autophagy-dependent cell death.
This chapter will describe the mechanisms underlying the most common forms of regulated cell death and then summarize the evidence for their involvement in cell death in the immature brain exposed to hypoxic-ischemic insult.
The apoptosis pathway triggered depends on whether the stimulus is intracellular (intrinsic, mitochondrial) or extracellular (extrinsic), although both converge at the activation of caspases-3 and -7 ( Fig. 125.2 ). The intrinsic pathway relies on mitochondrial outer membrane permeabilization (MOMP) resulting in the release of a number of proapoptotic proteins into the cytosol. This permeabilization is tightly controlled by the pro- and antiapoptotic members of the B cell lymphoma 2 (BCL2) family. MOMP is directly executed by BCL2-associated X protein (BAX) and BCL2-antagonist/killer 1 (BAK1) that oligomerize to form outer membrane pores (see Fig. 125.2 ). The opening of the BAX/BAK1 pore is regulated by the balanced action of antiapoptotic BCL2 family proteins such as BCL2 itself, BCL2-like 1 (BCL-XL), and myeloid cell leukemia 1, and proapoptotic “sensitizers” such as BCL2 binding component 3 (PUMA), BCL2-like 11 (BIM), and BH3-interacting domain death agonist (BID), which can initiate the pathway. Initiation of MOMP can also be controlled by p53, c-Jun N-terminal kinase (JNK), and caspase-2. Once MOMP is initiated, leakage of proapoptotic proteins from the mitochondrial intermembrane space occurs, most notably cytochrome c (Cyt c ). Cyt c forms an apoptosome complex with deoxy-ATP, and apoptotic peptidase-activating factor 1 (APAF-1), which recruits and activates caspase-9, leading to the downstream activation of the executioner caspase-3. , Additional proapoptotic proteins released from the mitochondrion include Smac/DIABLO and Omi/HtrA2, proteases that cleave X-linked inhibitor of apoptosis (XIAP), preventing XIAP-mediated inhibition of caspase activation. Finally, apoptosis-inducing factor (AIF) is also released from the mitochondrion, translocating to the nucleus where it facilitates chromatinolysis and cell death in a caspase-independent manner. MOMP is not only very rapid but also extensive; once initiated in a limited number of mitochondria, MOMP swiftly proceeds through all mitochondria within the cell.
In the extrinsic pathway, binding of extracellular ligands to a death receptor leads to activation of caspase-8 and cleavage of either BID (to crossover with the intrinsic pathway) or directly to caspase-3. Over 20 ligand-receptor pairings are now included in the death receptor ligand tumor necrosis factor (TNF) superfamily. These TNF-receptor (TNFR) and TNFR-like molecules are similar in structure to TNF and function as trimers (both ligands and receptors). Because of the similarity of their structure, multiple ligands are able to bind and induce signaling through one receptor, or a single ligand is able to bind multiple receptors. Receptors containing the so-called death domain in their intracellular domain (e.g., TNFR1, DR4, DR5, Fas) are able to trigger apoptosis when activated by the binding of the corresponding ligand (e.g., TNF-α, TNF-related apoptosis inducing ligand [TRAIL], FasL). This extrinsic pathway of apoptosis continues with the activation of a death-inducing signaling complex (DISC) adjacent to the death domain of the receptor. DISC comprises Fas-associated death domain protein (FADD) and procaspase-8; binding results in caspase-8 homodimerization and activation by proteolytic cleavage (see Fig. 125.2 ). Activated caspase-8 either directly activates caspase-3 or mediates cleavage of BID to truncated BID (tBID), which integrates different death pathways at the mitochondria. Truncated BID translocates to mitochondria, where it interacts with other proapoptotic proteins and triggers the release of apoptogenic factors, leading to caspase-dependent and caspase-independent cell death. Death receptors can also trigger necroptosis, particularly under conditions when caspase-8 is inactive (see the section “Necrosis and Necroptosis”).
Apoptosis is critical for normal brain development and determines the size and shape of the central nervous system. In some regions more than half of all neurons initially formed undergo apoptotic cell death, with data from mouse brain suggesting that cell death peaks just after birth. Similarly, oligodendrocytes are produced in excess and then pruned to needed numbers by apoptosis. Many of the proteins involved in apoptosis, such as caspase-3, APAF-1, and BCL2-family proteins, , are upregulated during brain development. Mice devoid of caspase-3 or caspase-9 exhibit hyperplastic disorganized brains, , whereas other organs such as the thymus (with ongoing apoptosis) develop normally, supporting the concept that caspases are of particular importance in shaping the developing brain. Thus, several components of the intrinsic pathway are already markedly up-regulated in the postnatal brain because of physiologic apoptosis as part of central nervous system development.
Mitochondria in the developing brain are prone to permeabilization in response to HI. , The timing of mitochondrial permeabilization is debated, but most study findings suggest that it happens 3 to 24 hours after HI—that is, starting during the latent phase and proceeding into the secondary phase of injury depending on the severity of insult, the animal model, and the brain region. , These proposed timings are also supported by evidence from interventions that block mitochondrial permeabilization, which are effective if given up to 6 hours after HI.
Proapoptotic proteins (e.g., Cyt c and AIF) are released from mitochondria, the apoptosome forms, and downstream executioner caspases (particularly caspase-3) are activated after hypoxic-ischemic insult. , Pathways dependent on AIF , and caspases seem to be more strongly activated in the immature brain than in the adult brain, , and mitochondrial permeabilization has been proposed to mark the point of no return in hypoxic-ischemic injury of the immature brain. , Furthermore, studies that ablate the effects of BAX-mediated MOMP (e.g., knockout models of BIM and BCL2-associated death promoter ( BAD ), Tat-BCL-XL–mediated neuroprotection, Bcl-XL transgenic mice ) consistently reduce brain injury after neonatal HI. Indeed, BAX-inhibitory peptides , and BAX deficiency substantially protect the immature brain in mice. Following HI insult in neonatal rats, a peak of caspase-3 activity is observed 24 hours after the insult, and caspase-3 activity remains elevated in excess of 6 days. Caspase inhibitors have therefore been shown to be neuroprotective in models of immature brain injury following HI. ,
The molecular mechanisms of mitochondrial permeabilization under pathologic conditions are still not completely understood. Mitochondria can permeabilize through either BAX-BAK–dependent MOMP (see earlier) or opening of the mitochondrial permeability transition (MPT) pore at the junction of the inner and outer mitochondrial membranes. , The pore opens in response to increased matrix calcium concentration (as might be experienced during excitotoxicity) dissipating mitochondrial membrane potential, resulting in depolarization and mitochondrial swelling. Although the nature of the MPT pore components is still debated (candidates include Fo F1 ATP synthase ), cyclophilin D appears to be critical for the regulation of MPT. , Cell death mediated by the MPT pore opening occurs in adult brain ischemic injury, because deficiency of the cyclophilin D gene Ppid and cyclophilin D inhibitors are neuroprotective. However, in the immature brain, Ppid deficiency aggravates rather than lessens HI injury, and cyclophilin D inhibitors do not reduce injury. , These data suggest that BAX-dependent permeabilization (see above), rather than cyclophilin D–mediated opening of the MPT pore, is the more common pathway in the developing brain.
AIF is normally tethered to the inner mitochondrial membrane; on pathologic stimuli AIF is cleaved and migrates through the permeabilized outer mitochondrial membrane where it accumulates in the cytosol. This release from the mitochondrion is dependent on activation and binding of poly(ADP-ribose) polymerase 1 (PARP1) to AIF and as such, this specific route of cell death is often referred to as parthanatos (see Fig. 125.1 ) rather than apoptosis , as it morphologically resembles aspects of necrosis. , AIF can interact with a number of cytosolic proteins including cyclophilin A, and translocates to the nucleus where it facilitates DNA fragmentation and chromatin condensation, resulting in caspase-independent cell death. In the immature brain, AIF also translocates to the nucleus after neonatal HI, and mice with lower expression of AIF are less vulnerable to HI, especially in combination with administration of a caspase inhibitor. Conversely, in models in which AIF expression is increased (e.g., by compensation following ablation of brain-specific AIF isoform, or by AIF knock-in ), there is an increased cell death and infarct size, suggesting that mitochondrial AIF release contributes to neonatal brain injury.
Protein p53 is a tumor suppressor that triggers apoptosis via multiple pathways, including cell cycle arrest and the regulation of autophagy through transactivating proapoptotic genes and repressing antiapoptotic genes. It is highly conserved and regulates cell death resulting from a wide variety of both physiologic and pathologic stimuli. Protein p53 also has cytoplasmic actions at the mitochondrial level and can promote BAX/BAK-dependent mitochondrial permeabilization. , In unstressed neurons, p53 expression is generally low, limited by its association with its negative regulator MDM2, which functions as a ubiquitin ligase, targeting polyubiquitinated p53 for degradation. Cellular stress displaces p53 from MDM2, and subsequently p53 expression is stabilized through substantial posttranslational modification. The classic role of p53 is as an activator of transcription, and on stabilization, it accumulates in the nucleus, where it up-regulates the transcription of proapoptotic genes such as PUMA and NOXA . Indeed, p53 expression is up-regulated and accumulates in the nucleus and mitochondria in an in vivo rat model of neonatal HI. , In consequence, there is an up-regulation of apoptotic pathways, leading to activation of caspase-3. Targeting pathways upstream of p53 with an nuclear factor (NF)κB inhibitor peptide resulted in a decreased accumulation of p53, increasing neuronal survival in response to neonatal HI. However, the therapeutic benefit of reducing p53 in neonatal brain injury is still unclear, , highlighting the complex regulation of signaling required by a p53-BAX–dependent pathway.
JNKs are members of the mitogen-activated protein kinase family and, as such, are activated in response to stress. There are three mammalian JNK genes and 10 expressed isoforms as the result of alternative splicing; however, it is JNK3 that is predominantly active in the brain. In a mouse model in which JNK3 expression is ablated, both adult and neonatal animals were partially protected against hypoxic-ischemic insult, and in newborn animals compared with wild-type animals, levels of c-jun were reduced. , In rat pups exposed to HI, activation of JNK3 is accompanied by translocation of the transcription factor FOXO3a to the nucleus and upregulation of proapoptotic Bim and caspase-3. Pharmacologic inhibition of FOXO3a decreased cell death and infarct in this model. Inhibition of JNK (either by TAT-JBD or by D-JNKi ) in neonatal mice after HI resulted in reduced infarct size, preservation of mitochondrial integrity, and a more favorable behavioral outcome. JNK3 is hypothesized to act upstream of the proapoptotic BCL2 family as JNK3-mediated increases in BIM and PUMA expression were absent in JNK3 gene–knockout mice. In addition, activation of caspase-3 was also decreased, suggesting that activation of JNK3 in response to hypoxic-ischemic insult results in caspase-dependent apoptosis.
Caspase-2 is a member of the initiator subgroup of caspases, and is developmentally regulated. Canonical activation of caspase-2 is dependent on its dimerization and subsequent cleavage, facilitated through interaction with p53-induced death domain–containing protein (PIDD) and RIP-associated ICH-1/CED3 homologous protein with a death domain (RAIDD) in some cellular systems. In addition, caspase-2 activation can be triggered by nuclear DNA damage or endoplasmic reticulum (ER) or Golgi apparatus stress via a mechanism not dependent on PIDD/RAIDD. Once activated, caspase-2 promotes BID cleavage, resulting in BAX translocation and release of Cyt c . Notably, neonatal caspase-2 –null mice are partially protected from excitotoxic and HI injury, , in contrast with adult caspase-2 –knockout mice. High expression of caspase-2 was found in neonatal mice and rats and in postmortem human tissue from neonates ; TRP601, a group II caspase inhibitor targeting caspase-2 and caspase-3, provided significant protection against white and gray matter loss in neonatal animals subjected to excitotoxicity, arterial stroke, or HI.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here