Mechanisms of hypoxic—ischemic injury in the term infant

Mechanisms of Hypoxic–Ischemic Injury in the Term Infant Claire McLean, MD, and Donna Ferriero, MD The pathogenesis of hypoxic–ischemic brain injury in the term infant is multifactorial and complex. Over the past decade the investigative emphasis has turned to cellular and molecular mechanisms of injury, and it has been increasingly recognized that the neonatal brain differs vastly from the adult brain in terms of response to hypoxia–ischemia. This review will discuss the initiation and evolution of brain injury in the term neonate, and the inherent biochemical and physiologic qualities of the neonatal brain that make its response to hypoxia–ischemia unique. Attention will be given to specific areas of investigation including excitotoxicity, oxidative stress, and inflammation. The coalescence of these entities to a final common pathway of hypoxic–ischemic brain injury will be emphasized. Semin Perinatol 28:425-432 © 2004 Elsevier Inc. All rights reserved.

A

lthough there are multiple etiologies for neonatal encephalopathy (NE) in the newborn period, hypoxia– ischemia (HI) predominates as the major cause of subsequent neurologic disability.1 Increased understanding in the past three decades of the pathogenesis of neonatal encephalopathy indicate that it is much more complex than originally thought. It is clear that avoidance or early recognition and intervention for intrapartum “fetal distress” will not eradicate the problem.2 Further, it is becoming more evident that hypoxia–ischemia (HI) is the final common endpoint for a complex convergence of events, some genetically determined and some triggered by an in utero (but not necessarily intrapartum) stressor. The knowledge that HI is a process that evolves over hours to days provides a “window of opportunity” for intervention. However, numerous clinical and laboratory observations have demonstrated that the response of the neonatal brain is vastly different than that of the adult brain. Although the potential for effective therapies is perhaps greater in the immature organism due to plasticity, so is the potential for harm. Molecules that play important roles in HI (for example, iron, glutamate, calcium, and nitric oxide) are often mediators of crucial developmental processes. The final pattern that HI assumes, both radiographically and neurodevelopmentally, depends on the gestational age at which the insult occurs. The concept of “selective vulnerability” of different Division of Neonatology, Department of Pediatrics, University of California, Neonatal Brain Disorders Center, San Francisco, CA. Address reprint requests to: Dr. Donna Ferriero, Neonatal Brain Disorders Center, 521 Parnassus Avenue, Room C215, San Francisco, CA 941430663. E-mail: [email protected]

0146-0005/04/$-see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1053/j.semperi.2004.10.005

brain structures and neural cells at different stages of development determines the immature brain’s response to hypoxia– ischemia. This review will discuss the various components of cellular/molecular response including excitotoxicity, oxidative injury, and inflammation as they relate to newborn brain injury.

The Neonatal Brain Is Different The fact that the newborn brain is very different from the adult brain has been frequently overlooked. Perhaps the aspect of the immature brain that has traditionally received the most controversial attention is the blood brain barrier (BBB). The common belief that the neonate, especially preterm but term as well, has a less effective BBB is evident in the medical approach to indirect hyperbilirubinemia, systemic infection, and administration of medication. However, “less effective” may be inaccurate. Studies of the BBB in the immature organism are often limited by paradigms using huge volumes and/or osmotic loads; this in addition to the fact that immature blood vessels are indeed weaker leads to these erroneous conclusions. The fetal BBB possesses mechanisms not found in the adult, and the high protein concentration seen in fetal and newborn cerebrospinal fluid (CSF) is not necessarily evidence of barrier immaturity.3 There is a mechanism in choroid plexus endothelial cells that promotes passage of proteins into the CSF that declines in effectiveness as the brain matures.3 This protein permeability is by no means indiscriminate; proteins of different species are recognized and excluded. The purpose of this mechanism is not fully understood but it clearly serves a critical developmental pur425

C. McLean and D. Ferriero

426 pose. Tight junctions, the foundation of the blood brain barrier, are present as soon as embryonic vessels invade the brain and go through progressive changes during development that may correlate with a decline in passive permeability.4-6 Sequential development of different ion gradients between CSF and plasma also occurs over the course of fetal development, and the rate of transport of heavy metals (specifically iron) and amino acids into CSF changes during development and is highly regulated and specific.3 Cerebrovascular autoregulation is another factor in HI, particularly in the preterm infant. The concept that preterm infants have a “pressure passive” cerebral circulation is widely accepted. However, sick term infants have also demonstrated impaired cerebrovascular autoregulation.7 Physiologically increased expression of inducible and neuronal isoforms of nitric oxide synthase (iNOS and nNOS) in the newborn period may narrow the autoregulatory window,8 making newborns more susceptible to brain injury with fluctuations in systemic pressure. In response to high circulating levels of prostaglandins in the newborn period, prostaglandin receptors are down-regulated. This may blunt prostaglandin-mediated vasoconstrictive response to systemic hypertension, leading to pathologically increased cerebral blood flow.9 The range of systemic blood pressure over which cerebrovascular autoregulation is functional expands with increasing maturity.10,11 Newborn rabbits have a very narrow window and are more susceptible to hyper- or hypotensive injury.11 Severely encephalopathic human newborns with evidence of cerebral damage on amplitude-integrated EEG display impaired autoregulation.12 Taken together, this evidence suggests that the neonatal brain is more vulnerable to large fluctuations in systemic pressure. Once an ischemic insult has occurred the neonate may be even more likely than an adult to sustain further damage in the acute recovery phase, as an encephalopathic infant is at risk for both systemic hypotension and hypertension. Glucose transport and metabolism in the immature brain are processes that also display differences from the adult. During periods of rapid brain growth, particularly the perinatal period, metabolic need increases.13 The pattern of injury after HI can be explained, in part, on the basis of this metabolic demand, as areas of high oxygen and glucose utilization correspond to areas of subcortical white matter injury.14 After HI in rats, glucose transport from plasma and into neurons increases and is compensatory.15,16 Normal brain development is associated with a switch in energy substrate from both glucose and ketones in the immature organism to an absolute requirement for glucose in the adult. These findings illustrate the importance of understanding the utilization of glucose, lactate, and ketones in the newborn brain under both normal and pathologic conditions. Evidence of susceptibility to cell death programs in the newborn period is an important concept for understanding HI.17 Multiple studies have suggested a more prominent role for neuronal apoptosis after HI in the neonate than in the adult, where necrosis seems to be more primary.18-20 Immature neurons in vitro are also more susceptible to apoptotic death than mature neurons.21 Perhaps apoptotic neuronal

death is more prominent in the neonate because of the physiologic role of apoptosis in normal brain development. Caspase-3, an executor of apoptosis, is expressed at much higher constitutive levels in immature versus mature brain.20 The basis for the evolution of injury after neonatal HI may be an apoptotic–necrotic continuum, with some cells exhibiting evidence of both.22 This prolonged evolution of injury in the neonate presents an opportunity for therapeutic intervention.23

Excitotoxicity and Energy Failure Excitotoxicity as a mechanism for neuronal injury after hypoxic–ischemic insult has received a great deal of attention during the last decade, and much has been elucidated regarding this process in the immature brain. Most of what we know comes from animal studies (mainly rodent) but studies in the human neonate have provided some corroboration. Neuronal glutamate metabolism has recently been shown by magnetic resonance spectroscopy to be tightly coupled to cerebral glucose oxidation,24,25 indicating that the cycle of glutamate release, reuptake, and resynthesis is a major metabolic pathway in the brain.26 When glutamate is released from presynaptic vesicles into the synapse, it can stimulate postsynaptic receptors (NMDA, AMPA, or kainate). Removal of glutamate from the synapse is dependent on glutamate transporters present mainly on glial cells. The glia convert glutamate to glutamine, glutamine is transported out of the glia and into neurons, and the neurons convert glutamine back to glutamate26 (Fig. 1). This process requires intact cellular energy machinery and function, and can be disrupted by any process that causes energy failure27-29 including glucose deprivation or HI.30 Since the fetus is adapted to a low oxygen tension and has a low cerebral basal energy consumption compared with the mature organism, this could explain the more delayed energy failure that occurs in immature brain. The NMDA receptor is of primary importance in excitatory neurotransmission. The structure and function of this receptor vary between species, brain region, and stage of development.31 Structurally it is composed of four heteromeric subunits, the combinations of which create different functional states.32 The receptor possesses multiple functional sites within it that recognize glutamate, coagonists, modulatory molecules such as glycine, dissociative anesthetics (PCP, ketamine), redox agents, steroids, histamine, zinc, magnesium (which blocks permeability to calcium), and a cation-selective ion channel which admits Na⫹, K⫹, and Ca2⫹. When the neurotransmitter recognition site is activated by glutamate, the ion channel allows influx of calcium and sodium. The increase in intracellular calcium that results is the stimulus for a multitude of downstream events, including regulation of transcription factors, cell cycle regulation, and DNA replication.33 The NMDA receptor is relatively over-expressed in the developing brain compared with the adult brain34 (Fig. 2). In postnatal day 6 to 14 rats (which approximates

Hypoxic–ischemic injury in the term infant

Figure 1 The glutaminergic synapse. (A) Glutamate is released by presynaptic neurons via a synaptosomal mechanism on excitation. Synaptically, glutamate acts at postsynaptic receptors until removed by energy-requiring astrocytic pumps. The astrocyte converts glutamate to glutamine which can be taken up by the presynaptic neuron and converted back to glutamate. (B) The astrocyte requires ATP and glucose to perform these functions. During HI, the mechanism of glutamate reuptake and recycling is impaired, glutamate remains in the synapse, resulting in excitotoxicity. Reprinted with permission.26 (Color version of figure is available online.)

the term human neonate), the NMDA receptor is expressed at 150 to 200% of adult levels.35,36 In humans, receptor expression follows a bell-shaped curve that peaks at 23 to 27 weeks gestation and is significantly higher at term than in the adult.37 In the fetal guinea pig, the affinity of the receptor for glutamate peaks in the perinatal period.38 The particular combination of subunits determine the NMDA receptor’s functional state, and the predominating combination in the perinatal period seems to favor a more prolonged and pronounced calcium influx for a given excitation.39 From a developmental perspective, this enhancement of the excitatory neurotransmitter network corresponds with a surge in brain growth and differentiation and highlights the role of glutamate and NMDA receptors in early learning40,41 and “physiologic neuronal death” or morphologic “sculpting.”42 The higher density and activity of glutamate receptors in the perinatal period creates a potential for a more devastating effect when energy failure does occur. The same NMDA receptor activation which allows for plasticity and synaptogenesis can, in the setting of HI, lead to massive Na⫹ and water influx, cellular swelling and necrosis, pathologically elevated intracellular calcium and its associated mitochondrial dysfunction, energy failure, and apoptosis. As neurons die, degradative enzymes are released, leading to a “spiral of death”27 potentiated further by inhibition or even reversal of glutamate reuptake by glia. Since glycolytic metabolism provides the ATP that powers the glial glutamate transporter,24 depri-

427 vation of glucose (which occurs in ischemia) will lead to high synaptic glutamate levels. HI in rat hippocampal neurons leads to marked reduction in the activity of the pumps which remove glutamate from the synapse,43 and the presynaptic glial glutamate transporter in immature rats subjected to unilateral carotid artery ligation and hypoxia is severely affected.44 To strengthen the theory that the immature brain is uniquely susceptible to excitotoxicity, it has been shown that injection of NMDA into rat brain produces more extensive cell death in the neonate than in the adult.45 HI in the rat follows the same age-dependent pattern (peak at postnatal day 6 with a decline toward adulthood).46 Furthermore, changes in NMDA receptor expression during early development could explain the different patterns of injury seen in the preterm versus term infant. A rodent model using intracerebral injection of glutamate receptor agonist had selective white matter injury at P7 and severe cortical infarction with no white matter selectivity at P10. The fact that both HI and direct injection of excitotoxin selectively damage postsynaptic neurons (namely those with NMDA receptors)47 and create a setting in which energy failure decreases reuptake of EAAs by glia27-29 highlight the importance of excitotoxicity in hypoxic–ischemic brain injury. This hypothesis is further strengthened by animal studies using unilateral carotid artery ligation and hypoxia that have shown near complete protection by the NMDA receptor antagonist MK-801.48 Adenosine receptors are expressed on excitatory neurons, and the involvement of this system in excitotoxicity and subsequent brain injury is becoming increasingly clear. The nucleoside adenosine, present in all tissues including the brain, is coreleased (in the form of ATP) with glutamate into the synaptic cleft on neuronal depolarization.49 Levels of adenosine can increase exponentially during ischemia,50 resulting in increased adenosine receptor activation. Recent evidence

NMDA receptor nNOS

P0

P5

P10

P15

P20

P25

P30 adult

Figure 2 Graphic depiction of NMDA receptor (adapted from reference 39) and nNOS maturation (adapted from reference 89) in rat brain. NMDA receptor expression and function and nNOS expression peak at approximately postnatal day 6 to 7, a developmental age that corresponds to the term human (adapted from references 89 and 113).

428 indicates that this receptor activation may inhibit axonal growth and white matter formation.51 From a clinical perspective, elevated glutamate has been documented (by proton magnetic resonance spectroscopy) in the CSF of infants who have suffered severe HI injury.52 In fact, CSF levels of excitatory amino acids are directly proportional to the severity of NE.53,54 NE in term neonates is characterized by excessive neuronal excitation that manifests as seizures and a burst-suppression EEG pattern.55 The neonatal brain is much more prone to seizure activity than the mature brain,56 and seizures in asphyxiated term infants can be atypical in appearance, difficult to diagnose, and quite resistant to traditional therapeutic measures.57-59 This suggests an age-specific mechanism for seizures that is not targeted by interventions that show efficacy in older patients (phenobarbital, lorazepam, phenytoin). Morphologically, MRI has allowed for classification of different types of hypoxic injury in term infants.60,61 The fact that seizures are common to all of them suggests a prominent role for neuronal hyperexcitability and excitotoxicity. Further, the pattern of excitatory neuronal circuitry can be used to explain the pattern of injury after severe HI in which injury to the excitatory regions of the brain (putamen, thalamus, and perirolandic cerebral cortex) is foremost.17,62 Although white matter damage has been thought of as predominantly a preterm injury, myelination is a process which is not complete at term and therefore the term infant is also vulnerable. Imaging and postmortem data from infants who sustained an HI insult between 35 and 42 weeks show extensive damage to the corpus callosum and internal capsule as well as the basal ganglia and thalamus.63 In human infants, myelination of the posterior limb of the internal capsule and the central part of the corona radiata appears from MRI evaluation to occur between 34 and 46 weeks.64 Therefore the developing oligodendrocyte is clearly a target in term HI. The immature oligodendrocyte (O4⫹/galactoceramide⫺) is uniquely susceptible to hypoxia–ischemia.65 A “feedback loop” mechanism has been invoked whereby glutamate released from oligodendrocytes could stimulate receptors on the same cell, leading to calcium influx and cell death.65

Oxidative Stress and Energy Failure The concepts of excitotoxicity and oxidative stress are inextricably linked, and many of the nuances of this complex relationship are still being clarified. Oxidative stress is a general term for the increase in free radical production as a result of oxidative metabolism under pathologic conditions. At a physiologic level in a cell with normally functioning mitochondria, more than 80% of oxygen in the cell is reduced to energy equivalents (ATP) by cytochrome oxidase. The rest is converted to superoxide anions that will, in a normally functioning cell, be reduced to water by enzymatic and nonenzymatic antioxidant mechanisms. On a simplistic level, any damage to the energy-producing machinery of the mitochondria will result in an accumulation of superoxide, and any

C. McLean and D. Ferriero process that results in depletion of antioxidant defenses will result in the default conversion of superoxide to even more reactive species such as the hydroxyl radical.66 The concept of ischemia/reperfusion whereby glucose and oxygen deprivation lead to primary cell death and reperfusion and reoxygenation lead to secondary cell loss67 is fundamental to the understanding of oxidative stress. When oxygen floods the microenvironment of cells that have been damaged by hypoxia, mitochondrial oxidative phosphorylation is overwhelmed and reactive oxygen species accumulate.66 Antioxidant defenses are depleted and free radicals damage the cell by peroxidation of lipid membranes, alteration of membrane potentials, activation of pro-apoptotic mediators, and direct DNA and protein damage. Excitotoxicity causes energy depletion, mitochondrial dysfunction, and cytosolic calcium accumulation, leading to the generation of free radicals such as superoxide, nitric oxide derivatives, and the highly reactive hydroxyl radical. Free radicals in turn alter membrane and pump function, allowing for more glutamate release and NMDA receptor activation and leading to more excitotoxicity. Because of its high lipid (specifically, polyunsaturated fatty acid) content, the brain is particularly susceptible to free radical attack and lipid peroxidation.33 This heightened vulnerability is magnified in the term newborn brain for several reasons. First, the polyunsaturated fatty acid content of the brain increases during gestation.68 There is a basal level of lipid peroxidation under normal conditions that is higher in term than preterm brain.69 Lipid peroxidation leads to the activation of phospholipases that increase free radical production. Under hypoxic conditions, free radical accumulation in brain occurs,70 and hypoxic tissue undergoes peroxidation much faster than normoxic tissue.69 Second, the immature brain has immature antioxidant defenses. Specifically, the antioxidant enzyme systems superoxide dismutase, catalase, and glutathione peroxidase display less activity in the immature than the mature rat brain.71 Third, the newborn brain is rich in free iron relative to the adult brain.66 Developmentally this is advantageous because iron is a cofactor in many enzymatic reactions that correspond to neuronal growth and differentiation. However, free iron can catalyze the production of various reactive oxygen species. Increased free iron is detectable in the plasma72 and CSF of asphyxiated newborns.73 In the rat, brain regions with high iron content are more vulnerable to injury74 and deferoxamine, an iron chelator, is protective against HI in animal models.75,76 The damaging potential of abundant iron and immaturity of the enzymatic oxidant defenses of the immature brain are tightly interrelated. Copper-zinc superoxide dismutase (SOD-1) is the cytosolic enzyme responsible for conversion of superoxide to hydrogen peroxide. Hydrogen peroxide is further reduced to water by glutathione peroxidase or catalase, or alternatively it can be converted to the hydroxyl radical in the presence of ferrous iron (Fig. 3). Enhanced glutathione peroxidase is protective when immature neurons in vitro are exposed to hydrogen peroxide.77 Therefore, an imbalance of enzymatic maturity can be invoked to explain the maturational differences; lack of sufficient glutathione peroxidase

Hypoxic–ischemic injury in the term infant

429

Detoxification of Reactive Oxygen Species

superoxide dismutase 2O2. + 2H+

glutathione peroxidase H2O2 + O2

Fe++

H2O

Reduced glutathione oxidized glutathione

Fe+++

OH. Figure 3 Illustration of the reduction of superoxide (OH.) to hydrogen peroxide (H2O2) by superoxide dismutase and the further reduction of hydrogen peroxide to water by glutathione peroxidase. In the setting of abundant free iron and inadequate glutathione peroxidase activity (or inadequate reduced glutathione stores), hydrogen peroxide is converted to the extremely reactive hydroxyl radical (OH.). The immature brain has immature antioxidant enzyme systems, relatively decreased stores of reduced glutathione, and abundant free iron, making it extremely vulnerable to oxidative attack.

activity in the immature brain can lead to accumulation of hydrogen peroxide78 and lipid peroxidation products. Both free radical scavengers (PBN, a nitrone spin-trap that converts free radicals to stable adducts) and metal chelators (deferoxamine and TPEN) have been shown to protect neurons from injury mediated by hydrogen peroxide in vitro75,79 and in vivo.80 These interventions also protected neurons from NMDA-induced toxicity, strengthening the link between excitotoxicity and oxidative stress.75 Adequate stores of antioxidants themselves are necessary to protect the CNS from oxidative injury. Specifically, depletion of neuronal reduced glutathione (GSH) has been shown to exacerbate oxidative injury.81,82 When glutathione peroxidase converts hydrogen peroxide to water, GSH is oxidized. Conversion back to the reduced form is catalyzed by glutathione reductase at the expense of NADPH, which itself is supplied by intact oxidative metabolism. Therefore intact mitochondrial machinery is essential for maintenance of an adequate GSH pool, and energy failure secondary to excitotoxicity will expose the cell to free radical damage. Conversely, when glutathione reductase activity and the pentose phosphate pathway are supported by exogeneous fructose 1,6 bisphosphate (FBP) (which acts endogenously a glycolytic intermediate), neurons in culture are protected from both oxygen-glucose deprivation and nitric oxide-mediated injury, even in the setting of GSH depletion.83 FBP also attenuates cortical neuronal loss in vivo after HI in neonatal rat brain.84 The link between excitotoxicity and free radical injury is well exemplified in studies of the role of nitric oxide (NO), another reactive oxygen species, in HI brain injury. NO can

function both physiologically and pathologically. Produced constitutively in endothelial cells, astrocytes, and neurons in response to an increase in intracellular calcium, it has a role in pulmonary, systemic, and cerebral vasodilation, and is thought to exert a compensatory vascular effect after ischemia during reperfusion. iNOS is produced in macrophages, endothelial cells, neurons, and astrocytes in response to stress. Hypoxia induces generation of NO in the cortex of newborn guinea pigs,85 and NO can modify the glycine-binding site of the NMDA receptor of cortical neurons during hypoxia, facilitating calcium entry and enhancing excitotoxicity.86 Neurons that express neuronal NOS (nNOS) in the striatum are selectively resistant to HI injury.87 Neuronal NOS expression corresponds anatomically to immature NMDA receptor expression, especially in the basal ganglia88 and disruption of the nNOS gene89 and pharmacologic inhibition of nNOS90 both ameliorate HI injury. In addition to their participation in oxidative injury and in the excitotoxic cascade, NO and NOS have been implicated in the programmed cell death that results from HI injury. It has been clearly shown that HI injury in the P7 rat is multiphasic, with immediate cytotoxic death that occurs within hours of the injury followed by death that occurs days to weeks later.91 This multiphasic quality of injury is supported by studies in human neonates.92 Inhibition of nNOS in newborn piglets prevents the increase in caspase-3 (the so-called “death effector”) activity and subsequent DNA fragmentation.93,94 In a separate study of hypoxia in newborn piglets NOS inhibition can block activation of ERK and JNK,95 two of the mitogen-activated protein kinase (MAPK) family that mediates signal transduction from cell surface to nucleus and thereby regulates programmed cell death.96,97

Inflammation It has been proposed that cytokines may be the final common mediators of brain injury that is initiated by hypoxia–ischemia, reperfusion, and infection.98 Cytokines are polypeptides that act either systemically or in a local fashion to guide the cellular response to inflammation, HI, infection, and a variety of other stressors. Their cellular targets are myriad and located throughout the body, including astrocytes, neurons, microglia, and endothelium of the CNS. The cytokines IL-1beta, TNF␣, IL-6, and IL-8 have been implicated clinically in the pathologic effects of brain inflammation.99,100 In addition, mediators such as platelet activating factor (PAF), arachidonic acid, and their metabolites (prostaglandins, leukotrienes, thromboxanes, cyclo-oxygenase) are involved in the inflammatory response during the evolution of brain injury after ischemia/reperfusion. Animal models of inflammation have drawn clear lines of causation between inflammation, excitotoxicity, and brain injury. For example, Gressens and colleagues performed intracerebral injections of ibotenate (a glutaminergic agonist of NMDA receptors) in mice to produce a model of excitotoxic brain injury.101 In P5 mice, this produces a pattern of white matter injury analogous to the preterm infant, and in P10 mice a pattern of cortical injury analogous to the term infant

C. McLean and D. Ferriero

430 results. Pretreatment with systemically administered IL1beta, IL-6, IL-9, or TNF-alpha leads to a statistically significant increase in the extent of the lesion produced by ibotenate, in a dose-dependent manner.102 This demonstrates that systemically circulating inflammatory mediators could in fact potentiate excitotoxic CNS injury, lending credence to the clinical paradigm of maternal infection leading to “cytokinemia” in the fetus and eventually to brain injury. Further, intracisternal injection of endotoxin before an HI insult in neonatal rats leads to potentiation of brain injury, and endotoxin injection is accompanied by increased TNF-alpha immunoreactivity.103 The cellular origin of inflammatory mediators that appear to exacerbate HI brain injury is still unclear. There is a role for mediators that are produced systemically (by the mother or by the fetus itself) and affect the CNS either through vascular mechanisms or by entry across the BBB and direct action on brain parenchyma. However, microglia, the resident macrophages of the CNS, are activated by HI and can release glutamate, free radicals, and nitric oxide.104 In fact, microglia are activated experimentally by ibotenate, an exogeneous excitotoxin.105 Drugs that block resident microglial and blood-derived monocyte activation (such as minocycline or chloroquine) protect the newborn brain from this excitotoxin.106 The wide variability in the effect of HI on the newborn brain highlights the probability that genetic factors play a significant role. Rodent models of HI show wide interstrain variability in the severity of injury after an identical insult.107,108 Possibilities have emerged regarding genetic predisposition to brain injury,109 and once again the perinatal period is of particular interest and represents one of the most likely time periods for a genetic modifier to present itself. For instance, genetic abnormalities causing a hypercoagulable tendency seem to increase risk for stroke in adults only if they are present in combination.110 However, the same mutation in a single gene in the neonate can be associated with increased risk for ischemic stroke.111 Gender differences with respect to the response to HI have also been observed.112 It is reasonable to assume that there are a variety of genetic factors that may achieve their highest potential to manifest during the perinatal period. Taken as a whole, the initiation and evolution of brain injury in the term infant after a hypoxic–ischemic insult is a vastly complex process, with contributing mechanisms densely interwoven to create a picture in which it is challenging to find a common thread. Recent investigations have focused on how the different yet related processes of excitotoxicity, oxidative stress, and inflammation come together in the neonate to produce a picture that is unique from that of the adult. It is by continuing to focus on this synthesis that we will arrive at an understanding of how neonatal encephalopathy occurs, and what we can do to prevent it.

Acknowledgment This work is supported in part by Grants T32 HD 07162 and NS33997.

References 1. Volpe JJ: Perinatal brain injury: From pathogenesis to neuroprotection. Ment Retard Dev Disabil Res Rev 7:56-64, 2001 2. Nelson KB, Dambrosia JM, Ting T, et al: Uncertain value of electronic fetal monitoring in predicting cerebral palsy. New Engl J Med 334: 613-618, 1996 3. Saunders NR, Habgood MD, Dziegielewska KM: Barrier mechanisms in the brain, II. Immature brain. Clin Exp Pharmacol Physiol 26:8591, 1999 4. Stewart PA, Hayakawa K: Early ultrastructural changes in blood-brain barrier vessels of the rat embryo. Dev Brain Res 78:25-34, 1994 5. Schultze C, Firth JA: Inter-endothelial junctions during blood-brain barrier development in the rat: Morphological changes at the level of individual tight junctional contacts. Dev Brain Res 69:85-95, 1992 6. Kniesel U, Risau W, Wolburg H: Development of the blood-brain barrier tight junctions in the rat cortex. Dev Brain Res 96:229-240, 1996 7. Boylan G, Young K, Panerai R, et al: Dynamic cerebral autoregulation in sick newborn infants. Pediatr Res 48:12-17, 2000 8. Hardy P, Nuyt AM, Dumont II, et al: Developmentally increased cerebrovascular NO in newborn pigs curtails cerebral blood flow autoregulation. Pediatr Res 46:375-382, 1999 9. Chemtob S, Li DY, Abran D, et al: The role of prostaglandin receptors in regulating cerebral blood flow in the perinatal period. Acta Paediatr 85:517-524, 1996 10. Verma PK, Panerai RB, Rennie JM, et al: Grading of cerebral autoregulation in preterm and term neonates. Pediatr Neurol 23:236-242, 2000 11. Tuor UI, Grewal D: Autoregulation of cerebral blood flow: Influence of local brain development and postnatal age. Am J Physiol 267: H2220-H2228, 1994 12. Pryds O, Greisen G, Lou H, et al: Vasoparalysis associated with brain damage in asphyxiated term infants. J Pediatr 117:119-125, 1990 13. Vannucci RC, Vannucci SJ: Glucose metabolism in the developing brain. Semin Perinatol 24:107-115, 2000 14. Azzarelli B, Caldemeyer K, Phillips J, et al: Hypoxic–ischemic encephalopathy in areas of primary myelination: A neuroimaging and PET study. Pediatr Neurol 14:108-116, 1996 15. Zovein A, Flowers-Ziegler J, Thamotharan S, et al: Postnatal hypoxic– ischemic brain injury alters mechanisms mediating neuronal glucose transport. Am J Physiol Regul Integr Comp Physiol 286:R273-R282, 2004 16. Vannucci SJ, Reinhart R, Maher F, et al: Alterations in GLUT1 and GLUT3 glucose transporter gene expression following unilateral hypoxia–ischemia in the immature rat brain. Dev Brain Res 107:255264, 1998 17. McQuillen P, Ferriero DM: Selective vulnerability in the developing central nervous system. Pediatr Neurol 30:227-235, 2004 18. Sidhu RS, Tuor UI, Del Bigio MR: Nuclear condensation and fragmentation following cerebral hypoxia–ischemia occurs more frequently in immature than older rats. Neurosci Lett 223:129-132, 1997 19. Pulera MR, Adams LM, Liu H, et al: Apoptosis in a neonatal rat model of cerebral hypoxia–ischemia. Stroke 29:2622-2630, 1998 20. Hu BR, Liu CL, Ouyang Y, et al: Involvement of caspase-3 in cell death after hypoxia–ischemia declines during brain maturation. J Cereb Blood Flow Metab 20:1294-1300, 2000 21. McDonald JW, Behrens MI, Chung C, et al: Susceptibility to apoptosis is enhanced in immature cortical neurons. Brain Res 759:228-232, 1997 22. Martin LJ, Al-Abdulla NA, Brambrink AM, et al: Neurodegeneration in excitotoxicity, global cerebral ischemia, and target deprivation: A perspective on the contributions of apoptosis and necrosis. Brain Res Bull 46:281-309, 1998 23. Nakajima W, Ishida A, Lange M, et al: Apoptosis has a prolonged role in the neurodegeneration after hypoxic ischemia in the newborn rat. J Neurosci 20:7994-8004, 2000 24. Sibson NR, Shen J, Mason GF, et al: Functional energy metabolism: In vivo 13C-NMR spectroscopy evidence for coupling of cerebral glu-

Hypoxic–ischemic injury in the term infant

25.

26. 27. 28.

29.

30. 31.

32.

33.

34.

35.

36.

37.

38.

39.

40. 41. 42.

43.

44.

45.

46.

47.

48.

cose consumption and glutamatergic neuronal activity. Dev Neurosci 20:321-330, 1998 Sibson NR, Dhankhar A, Mason GF, et al: Stoichiometric coupling of brain glucose metabolism and glutamatergic neuronal activity. Proc Natl Acad Sci U S A 95:316-321, 1998 Magistretti PJ, Pellerin L, Rothman DL, et al: Energy on demand. Science 283:496-497, 1999 Choi DW: Glutamate neurotoxicity and diseases of the nervous system. Neuron 1:623-634, 1988 Minc-Golomb D, Levy Y, Kleinberger N, et al: D-[3H]aspartate release from hippocampus slices studied in a multiwell system: Controlling factors and postnatal development of release. Brain Res 402:255-263, 1987 Hagberg H, Andersson P, Kjellmer I, et al: Extracellular overflow of glutamate, aspartate, GABA and taurine in the cortex and basal ganglia of fetal lambs during hypoxia–ischemia. Neurosci Lett 78:311-317, 1987 Danbolt NC: Glutamate uptake. Prog Neurobiol 65:1-105, 2001 Haberny KA, Paule MG, Scallet AC, et al: Ontogeny of the N-methylD-aspartate (NMDA) receptor system and susceptibility to neurotoxicity. Toxicol Sci 68:9-17, 2002 Danysz W, Parsons CG: Glycine and NMDA receptors: Physiologic significance and possible therapeutic applications. Pharmacol Rev 50: 597-664, 1998 Mishra OP, Delivoria-Papadopoulos M: Cellular mechanisms of hypoxic injury in the developing brain. Brain Res Bull 48:233-248, 1999 Fox K, Schlaggar BL, Glazewski S, et al: Glutamate receptor blockade at cortical synapses disrupts development of thalamocortical and columnar organization in somatosensory cortex. Proc Natl Acad Sci U S A 93:5584-5589, 1996 Tremblay E, Roisin MP, Represa A, et al: Transient increased density of NMDA binding sites in the developing rat hippocampus. Brain Res 461:393-396, 1988 McDonald JW, Johnston MV, Young AB: Ontogeny of the receptors comprising the NMDA receptor complex. Soc Neurosci Abstr 15:198, 1989 Represa A, Tremblay E, Ben-Ari Y: Transient increase of NMDA-binding sites in human hippocampus during development. Neurosci Lett 99:61-66, 1989 Mishra OP, Delivoria-Papadopoulos M: Modification of modulatory sites of NMDA receptor in the fetal guinea pig brain during development. Neurochem Res 17:1223-1228, 1992 Danysz W, Parsons CG: Glycine and N-methyl-D-aspartate receptors: Physiological significance and possible therapeutic applications. Pharmacol Rev 50:598-664, 1998 Lincoln J, Coopersmith R, Harris EW, et al: NMDA receptor activation and early olfactory learning. Dev Brain Res 39:309-312, 1988 Riedel G, Platt B, Micheau J: Glutamate receptor function in learning and memory. Behav Brain Res 140:1-47, 2003 Hattori H, Wasterlain CG: Excitatory amino acids in the developing brain: Ontogeny, plasticity, and excitotoxicity. Pediatr Neurol 6:219228, 1990 Jabaudon D, Scanziani M, Gahwiler BH, et al: Acute decrease in net glutamate uptake during energy deprivation. Proc Natl Acad Sci U S A 97:5610-5615, 2000 Tao F, Lu SD, Zhang LM, et al: Role of excitatory amino acid transporter 1 in neonatal rat neuronal damage induced by hypoxia–ischemia. Neurosci 102:508-513, 2001 McDonald JW, Silverstein FS, Johnston MV: Neurotoxicity of N-methyl-D-aspartate is markedly enhanced in developing rat central nervous system. Brain Res 459:200-203, 1988 Ikonomidou C, Mosinger JL, Salles KS, et al: Sensitivity of the developing rat brain to hypobaric/ischemic damage parallels sensitivity to N-methyl-D-aspartate neurotoxicity. J Neurosci 9:2809-2818, 1989 Ikonomidou C, Price MT, Mosinger JL, et al: Hypobaric-ischemic conditions produce glutamate-like cytopathology in infant rat brain. J Neurosci 9:1693-1700, 1989 Hagberg H, Gilland E, Diemer NH, et al: Hypoxia–ischemia in the

431

49.

50.

51.

52.

53. 54.

55. 56. 57.

58.

59. 60.

61.

62. 63.

64.

65.

66. 67. 68.

69.

70.

71. 72.

73.

neonatal rat brain: Histopathology after post-treatment with NMDA and non-NMDA receptor antagonists. Biol Neonate 66:206-213, 1994 Cunha RA, Vizi ES, Ribeiro JA, et al: Preferential release of ATP and its extracellular catabolism as a source of adenosine upon high-but not low-frequency stimulation of rat hippocampal slices. J Neurochem 67:2180-2187, 1996 Dunwiddie TV, Fredholm B: Adenosine neuromodulation, in Jacobson KA, Jarvis M (eds): Purinergic Approaches in Experimental Therapeutics. New York, NY, Wiley–Liss, 1998, pp 359-382 Turner CP, Seli M, Ment L, et al: A1 adenosine receptors mediate hypoxia-induced ventriculomegaly. Proc Natl Acad Sci U S A 100: 11718-11722, 2003 Pu Y, Li Q-F, Zeng C-M, et al: Increased detectability of alpha brain glutamate/glutamine in neonatal hypoxic–ischemic encephalopathy. AJNR Am J Neuroradiol 21:203-212, 2000 Riikonen RS, Kero PO, Simell OG: Excitatory amino acids in cerebrospinal fluid in neonatal asphyxia. Pediatr Neurol 8:37-40, 1992 Hagberg H, Thornberg E, Blennow M, et al: Excitatory amino acids in the cerebral spinal fluid of asphyxiated infants: Relationship to hypoxic–ischemic encephalopathy. Acta Paediatr 82:925-929, 1993 Johnston MV: Excitotoxicity in neonatal hypoxia. Ment Retard Dev Disabil Res Rev 7:229-234, 2001 Holmes GL, Ben-Ari Y: The neurobiology and consequences of epilepsy in the developing brain. Pediatr Res 49:320-325, 2001 Painter MJ, Scher MS, Stein AD, et al: Phenobarbital compared with phenytoin for the treatment of neonatal seizures. N Engl J Med 341: 485-489, 1999 Volpe JJ: Hypoxic/ischemic encephalopathy: clinical aspects, in Neurology of the Newborn. Philadelphia, PA, W.B. Saunders, 2001, pp. 331-394 Volpe JJ: Neurologic evaluation: neonatal seizures, in Neurology of the Newborn. Philadelphia, PA, W.B. Saunders, 2001, pp. 178-214 Rivkin MJ: Hypoxic–ischemic brain injury in the term newborn: Neuropathology, clinical aspects and neuroimaging. Clin Perinatol 24: 607-626, 1997 Barkovich AJ, Hajinal BL, Vigneron D, et al: Prediction of neuromotor outcome in perinatal asphyxia: Evaluation of MR scoring systems. AJNR Am J Neuroradiol 19:143-149, 1998 Pasternak JF, Gorey MT: The syndrome of acute near-total intrauterine asphyxia in the term infant. Pediatr Neurol 18:391-398, 1998 Azzarelli B, Caldemeyer K, Phillips J, et al: Hypoxic–ischemic encephalopathy in areas of primary myelination: A neuroimaging and PET study. Pediatr Neurol 14:108-116, 1996 Sie LT, van der Knaap MS, van Wezel-Meijler G, et al: MRI assessment of myelination of motor and sensory pathways in the brain of preterm and term-born infants. Neuropediatrics 28:97-105, 1997 Fern R, Moller T: Rapid ischemic cell death in immature oligodendrocytes: A fatal glutamate release feedback loop. J Neurosci 20:34-42, 2000 Ferriero DM: Oxidant mechanisms in neonatal hypoxia–ischemia. Dev Neurosci 23:198-202, 2001 Inder T, Volpe J: Mechanisms of perinatal brain injury. Semin Neonatol 5:3-16, 2000 Crawford MA, Sinclair AJ: Nutritional influences in the evolution of mammalian brain. In: Lipids, malnutrition and the developing brain. Ciba Found Symp 267-292, 1971 Mishra OP, Delivoria-Papadopoulos M: Lipid peroxidation in developing fetal guinea pig brain during normoxia and hypoxia. Dev Brain Res 45:129-135, 1989 Maulik D, Numagami Y, Ohnishi T, et al: Direct detection of oxygen free radical generation during in-utero hypoxia in the fetal guinea pig brain. Brain Res 798:166-172, 1998 Khan JY, Black SM: Developmental changes in murine brain antioxidant enzymes. Pediatr Res 54:77-82, 2003 van Bel F, Dorrepaal C, Benders MJNL, et al: Neurologic abnormalities in the first 24 hours following birth asphyxia are associated with increasing plasma levels of free iron and TBA-reactive species. Pediatr Res 35:388A, 1994 Ogihara T, Hirano K, Ogihara H, et al: Non-protein-bound transition

C. McLean and D. Ferriero

432

74. 75.

76.

77.

78.

79.

80.

81.

82. 83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

metals and hydroxyl radical generation in cerebrospinal fluid of newborn infants with hypoxic ischemic encephalopathy. Pediatr Res 53:594-599, 2003 Subbarao KV, Richardson JS: Iron-dependent peroxidation of rat brain: A regional study. J Neurosci Res 26:224, 1990 Almli LM, Hamrick SEG, Koshy AA, et al: Multiple pathways of neuroprotection against oxidative stress and excitotoxic injury in immature primary hippocampal neurons. Dev Brain Res 132:121-129, 2001 Sarco D, Becker J, Palmer C, et al: The neuroprotective effect of deferoxamine in the hypoxic–ischemic immature mouse brain. Neurosci Lett 282:113-116, 2000 McLean CW, Mirochnitchenko O, Ferriero DM: Immature murine cortical neurons overexpressing human glutathione peroxidase are protected from oxidative stress. Pediatr Res 53:347A, 2003 (abstr) Fullerton HJ, Ditelberg JS, Chen FS, et al: Copper/zinc superoxide dismutase transgenic brain accumulates hydrogen peroxide after perinatal hypoxia ischemia. Ann Neurol 44:357-364, 1998 Yue TL, Gu JL, Lysko PG, et al: Neuroprotective effects of phenyl-tbutyl-nitrone in gerbil global brain ischemia and in cultured rat cerebellar neurons. Brain Res 574:193-197, 1992 Cao X, Phillis JW: alpha-phenyl-tert-butyl-nitrone reduces cortical infarct and edema in rats subjected to focal ischemia. Brain Res 644: 267-272, 1994 White AR, Cappai R: Neurotoxicity from glutathione depletion is dependent on extracellular trace copper. J Neurosci Res 71:889-897, 2003 Chen CJ, Liao SL: Zinc toxicity on neonatal cortical neurons: Involvement of glutathione chelation. J Neurochem 85:443-453, 2003 Vexler ZS, Wong A, Francisco C, et al: Fructose-1,6-bisphosphate preserves intracellular glutathione and protects cortical neurons against oxidative stress. Brain Res 960:90-98, 2003 Rogido M, Husson I, Bonnier C, et al: Fructose-1,6-biphosphate prevents excitotoxic neuronal cell death in the neonatal mouse brain. Dev Brain Res 140:287-297, 2003 Mishra OP, Zanelli S, Ohnishi ST, et al: Hypoxia-induced generation of nitric oxide free radicals in cerebral cortex of newborn guinea pigs. Neurochem Res 25:1559-1565, 2000 Sorrentino DF, Fritz KI, Haider SH, et al: Nitric oxide-mediated modification of the glycine binding site of the NMDA receptor during hypoxia in the cerebral cortex of the newborn piglet. Neurochem Res 29:455-459, 2004 Ferriero DM, Arcavi LJ, Sagar SM, et al: Selective sparing of NADPHdiaphorase neurons in neonatal hypoxia–ischemia. Ann Neurol 24: 670-676, 1988 Black SM, Bedolli MA, Martinex S, et al: Expression of neuronal nitric oxide synthase corresponds to regions of selective vulnerability to hypoxia–ischaemia in the developing rat brain. Neurobiol Dis 2:145155, 1995 Ferriero DM, Holtzman DM, Black SM, et al: Neonatal mice lacking neuronal nitric oxide synthase are less vulnerable to hypoxic–ischemic injury. Neurobiol Dis 3:64-71, 1996 Ishida A, Trescher WH, Lange MS, et al: Prolonged suppression of brain nitric oxide synthase activity by 7-nitroindazole protects against cerebral hypoxic–ischemic injury in neonatal rat. Brain Dev 23:349354, 2001 Northington FJ, Ferriero DM, Graham EM, et al: Early neurodegeneration after hypoxia–ischemia in neonatal rat is necrosis while delayed neuronal death is apoptosis. Neurobiol Dis 8:207-219, 2001 Roland EH, Poskitt K, Rodriguez E, et al: Perinatal hypoxic–ischemic thalamic injury: Clinical features and neuroimaging. Ann Neurol 44: 161-166, 1998

93. Peeters-Scholte C, Koster J, van den Tweel E, et al: Effects of selective nitric oxide synthase inhibition on IGF-1, caspases and cytokines in a newborn piglet model of perinatal hypoxia–ischaemia. Dev Neurosci 24:396-404, 2002 94. Parikh NA, Katsetos CD, Ashraf QM, et al: Hypoxia-induced caspase-3 activation and DNA fragmentation in cortical neurons of newborn piglets: role of nitric oxide. Neurochem Res 28:1351-1357, 2003 95. Mishra OP, Zubrow AB, Ashraf QM: Nitric oxide-mediated activation of extracellular signal-regulated kinase (ERK) and c-jun N-terminal kinase (JNK) during hypoxia in cerebral cortical nuclei of newborn piglets. Neuroscience 123:179-186, 2004 96. Harris C, Maroney AC, Johnson EM Jr: Identification of JNK-dependent and -independent components of cerebellar granule neuron apoptosis. J Neurochem 83:992-1001, 2002 97. Wang X, Zhu C, Qiu L, et al: Activation of ERK 1/2 after neonatal rat cerebral hypoxia–ishcaemia. J Neurochem 86:351-362, 2003 98. Dammann O, Leviton A: Role of the fetus in perinatal infection and neonatal brain damage. Curr Opin Pediatr 12:99-104, 2000 99. Sävman K, Blennow M, Gustafson K, et al: Cytokine response in cerebrospinal fluid after birth asphyxia. Pediatr Res 43:746-751, 1998 100. Foster-Barber A, Ferriero DM: Neonatal encephalopathy in the term infant: Neuroimaging and inflammatory cytokines. Ment Retard Dev Disabil Res Rev 8:20-24, 2002 101. Marret S, Mukendi R, Gadisseux JF, et al: Effect of ibotentate on brain development: An excitotoxic mouse model of microgyria and posthypoxic-like lesions. J Neuropathol Exp Neurol 54:358-370, 1995 102. Dommergues M, Patkai J, Renauld J-C, et al: Pro-inflammatory cytokines and interleukin-9 exacerbate excitotoxic lesions of the newborn murine neopallium. Ann Neurol 47:54-63, 2000 103. Coumans A, Middelanis J, Garnier Y, et al: Intracisternal application of endotoxin enhances the susceptibility to subsequent hypoxic–ischemic brain damage in neonatal rats. Pediatr Res 53:770-775, 2003 104. Wood PL: Microglia as a unique cellular target in the treatment of stroke: Potential neurotoxic mediators produced by activated microglia. Neurol Res 17:242-248, 1995 105. Tahraoui SL, Marret S, Bodenant C, et al: Central role of microglia in neonatal excitotoxic lesions of the murine periventricular white matter. Brain Pathol 11:56-71, 2001 106. Dommergues MA, Plaisant F, Verney C, et al: Early microglial activation following neonatal excitotoxic brain damage in mice: A potential target for neuroprotection. Neuroscience 121:619-628, 2003 107. Sheldon RA, Sedik C, Ferriero DM: Strain-related brain injury in neonatal mice subjected to hypoxia–ischemia. Brain Res 810:114122, 1998 108. Yonekura I, Kawahara N, Nakatomi H, et al: A model of global cerebral ischemia in C57 BL/6 mice. J Cerebral Blood Flow Metab 24:151158, 2004 109. Banasiak KJ, Xia Y, Haddad G: Mechanisms underlying hypoxia-induced neuronal apoptosis. Prog Neurobiol 62:215-249, 2000 110. Szolnoki Z, Somogyvari F, Kondacs A, et al: Evaluation of the interactions of common genetic mutations in stroke subtypes. J Neurol 249:1391-1397, 2002 111. Gunther G, Junker R, Strater R, et al: Childhood Stroke Study Group: Symptomatic ischemic stroke in full-term neonates: Role of acquired and genetic prothrombotic risk factors. Stroke 31:2437-2441, 2000 112. Nunez JL, McCarthy NM: Sex differences and hormonal effects in a model of preterm infant brain injury. Ann N Y Acad Sci 1008:281284, 2003 113. Jensen FE: The role of glutamate receptor maturation in perinatal seizures and brain injury. Int J Devl Neuroscience 20:339-347, 2002