| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the National Center for Toxicological Research, U.S. Food & Drug Administration, Jefferson, AR.
Address correspondence and reprint requests to Cheng Wang, MD, PhD, Division of Neurotoxicology, National Center for Toxicological Research/FDA, 3900 NCTR Road, Jefferson, AR 72079-9502. Address e-mail to cheng.wang{at}fda.hhs.gov.
Abstract
Advances in pediatric and obstetric surgery have resulted in an increase in the duration and complexity of procedures requiring anesthesia. It has been reported that anesthetic drugs cause widespread and dose-dependent apoptosis in the developing rat brain. The similarity of the physiology, pharmacology, metabolism, and reproductive systems of the nonhuman primate to that of the human, especially during pregnancy, make the monkey an exceptionally good animal model for assessing potential neurotoxic effects of anesthetics. The window of vulnerability to these neuronal effects of pediatric anesthetics is restricted to the period of rapid synaptogenesis, also known as the brain growth spurt period. To minimize the risks to children resulting from the use of anesthesia, the following questions should be addressed:
Pharmacogeneomic/system biology approaches have great potential for helping to advance the understanding of brain-related biological processes, including neuronal plasticity and neurotoxicity. Because of the complexity and temporal features of how developmental neurotoxicity is manifested, pharmacogenomic/systems biology approaches may prove to be useful tools for enhancing our understanding of the biological processes induced by anesthetics. Therefore, the main purpose of this review is to describe the application of these approaches and models, as well as protection strategies, especially as regards the issue of anesthetic-induced neuronal cell death during development.
Much of the discussion that follows is based on experiments conducted with ketamine. This is due in part to the use of ketamine in the early studies and the volume of preclinical experimental work performed with this drug, as well as its use in anesthetic studies in developing rodents and nonhuman primates. Although ketamine use in pediatric anesthesia is relatively limited, the findings of the studies are sufficiently strong to merit concern about the N-methyl-d-aspartate antagonist drugs as a class. Our focus on ketamine should not be construed as implying that the risk of neurodegeneration with ketamine is greater, or less, than with other anesthetics. We are simply describing the effects where we have the most preclinical data.
Various anesthetic protocols have been used in pediatric medicine for many decades without systematic assessment concerning drug exposure and possible adverse effects. Most of the currently used general anesthetic drugs have either N-methyl-d-aspartate (NMDA) receptor blocking or 
-aminobutyric acid (GABA) receptor enhancing properties. These receptors mediate their actions by the activation of ionotropic (ligand-gated ion channels) and metabotropic (G protein-coupled) receptors, and act to influence earlier developmental events, including synapse formation, neuroplasticity, and survival processes during neural development.
The amino acid, l-glutamate is generally recognized as the major excitatory neurotransmitter of the mammalian central nervous system (CNS) and glutamate receptors play a major role in fast excitatory synaptic transmission. NMDA-type glutamate receptors are widely distributed throughout the CNS and operate ligand-activated ion channels primarily composed of three families of NMDA receptor subunits: NR1 with eight known splice variants, NR2 (A-D),1–3 and NR3A and B.4,5 The NR1 subunit is essential for receptor/channel function. The functional properties of the NMDA receptor vary throughout the CNS and the binding affinities of various ligands for recombinant NMDA receptors depend on subunit composition.6 NMDA receptors are involved in a variety of physiological and pathological processes, including memory and learning,7 neuronal development,8 epileptiform seizures, synaptic plasticity, and acute neuropathologies associated with stroke and traumatic injury.9 During the brain growth spurt, blockade of the NMDA receptor for a period of hours triggers widespread apoptotic neurodegeneration in the rodent brain.10
GABA, the principal inhibitory transmitter in the adult CNS, acts as an excitatory transmitter in the early postnatal stage.11 Functional GABAA receptors are expressed in neurons as early as embryonic stages and investigations by different groups have led to the conclusion that a transient excitatory action of GABA, via GABAA receptors is a general feature of the developing neurons. Activation of GABAA receptors depolarizes neuroblasts and immature neurons in all regions of the CNS studied, including spinal cord,12–14 hypothalamus,15 cerebellum,16 cortex,17 hippocampus,18,19 and olfactory bulb.13 This depolarization is not due to unusual properties of neonatal GABAA channels but rather to an elevated intracellular Cl– concentration, probably from developmental changes in [Cl–]i homeostasis systems.13,20,21 Postsynaptic GABAB receptor-mediated responses, that is, the activation of K+ and inhibition of Ca2+ currents, are absent from embryonic and neonatal rat hippocampus and neocortical neurons until the end of the first postnatal week of life.22,23 The reasons for this delayed maturation of postsynaptic GABAB receptor-mediated inhibition are not yet well understood, but may be due to a lack of coupling between receptors, G proteins, and K+ or Ca2+ channels,22 rather than to late development of receptors.23
It has been hypothesized that exposure of the developing brain to NMDA antagonists induces neuronal cell death, most likely through compensatory mechanisms. An important working hypothesis is that exposure of developing brains to individual anesthetics (such as ketamine), with continuous blockade of NMDA receptors, causes a compensatory up-regulation of these receptors. This up-regulation makes neurons bearing these receptors more vulnerable, after ketamine washout, to the excitototoxic effects of glutamate, because these up-regulated NMDA receptors allow for the accumulation of toxic levels of intracellular free calcium under normal physiological conditions. In addition, prolonged supra-physiologic stimulation of immature neurons by GABA agonists enhances the excitatory component in the action of GABA and may contribute to increased excitability during early development.24 This increased excitability, along with NMDA antagonist-induced alteration of NMDA receptors, could lead to neuronal cell death.
Modifications of synaptic efficacy are believed to play an important role in information processing and storage by neuronal networks. It has been suggested that synaptic abnormalities are an important component of the anesthetic-induced neurotoxicity. Synaptophysin is a synaptic vesicle-associated protein involved in synaptogenesis. The sialic acid polymer on neural cell adhesion molecules (PSA-NCAM) is an important regulator of cell surface interactions.25 PSA-NCAM is also a neuron-specific marker known to be an NMDA-regulated molecule important in synaptogenesis during development.26 Some experiments27,28 have been performed to determine the correlation between anesthetics and PSA-NCAM expression, because quantifying the levels of PSA-NCAM after anesthetic exposure would validate the activity states of neuronal synaptic plasticity.
Neuronal susceptibility to neurotoxic insult varies with stage of development. Both in vitro and in vivo approaches have been used to assess the neurotoxicity associated with a wide range of these drugs at a variety of doses and exposure durations. Although comprehensive gene expression/proteomic studies and long-term behavioral assessments remain to be accomplished, in vivo and in vitro models and analysis strategies have been developed to study the biological pathways and behavioral outcomes of anesthetic-induced cell death in the developing nonhuman primate and rodent.
NEUROTRANSMISSION, SYNAPTOGENESIS, AND ANESTHETIC-INDUCED NEURONAL CELL DEATH
Glutamate promotes certain aspects of neuronal development, including migration, differentiation, and plasticity.29 The NMDA-type glutamate receptor NR1 subunit is widely distributed throughout the brain and is the fundamental subunit necessary for NMDA channel function. NMDA receptor density has been shown to increase in cultured cortical neurons after exposure to the NMDA receptor antagonists D-AP5, CGS-19755, and MK-801, but not after exposure to the AMPA/kainate receptor antagonist CNQX.30 Over-activation of NMDA receptors kills neurons (Fig. 1) via a necrotic mechanism characterized by excessive sodium and calcium entry, accompanied by chloride and water entry that leads to cell swelling and death.31 More recently, it has been shown that NMDA receptor activation can also lead to apoptotic cell death.32–34
|
Of particular interest are the possible mechanisms by which NMDA antagonists such as ketamine enhance neuronal cell death as a result of ketamine-induced compensatory up-regulation of NMDA receptors (by continuously blocking the NMDA receptor in the developing brain). This up-regulation then makes neurons bearing these receptors more vulnerable, after ketamine washout, to the excitotoxic effects of endogenous glutamate. This compensatory hypothesis is supported by the following observations: (1) NR1 subunit mRNA (Fig. 2; in situ hybridization) is up-regulated in ketamine-treated monkey fetuses (gestation day 122) and infants (postnatal day (PND) 5)35, (2) increased expression of NMDA receptor NR1 protein is accompanied by enhanced cell death27, and (3) co-administration of NR1 antisense oligonucleotide (targeted to NR1 NMDA receptor subunit mRNA) is able to block the neuronal cell death induced by ketamine in rat and monkey cortical culture.26,27 Given the key role of the NR1 subunit, it is not surprising that up-regulated NR1 expression along with alterations in other NMDA receptor subunits (such as the NR2 family) and the composition of receptor subunits play an important role in determining the pharmacological properties of the receptor. In addition, it has been reported28 that even low concentrations of ketamine could interfere with dendritic arbor development of immature GABAergic neurons and could potentially interfere with the development of neural networks. Further studies are needed to determine receptor specificity of those effects induced by anesthetics such as ketamine.
|
On the other hand, studies in vivo on the protective effects of NMDA antagonists, such as ketamine, have given inconsistent results. Both no (or minimal) and substantial protective effects have been found against the lesions produced in vivo by NMDA agonists36–39 and by neuronal ischemia.40–42 Ketamine has a very short half-life in the brain,43,44 and hence some of the inconsistencies could be due to the dose used and the length of time for which neuroprotective concentrations remain.
Prolonged or repetitive pain may occur during critical periods of brain development in hospitalized neonates.45 Rapid brain growth, synaptogenesis, expression of excitatory receptors,46 and developmentally regulated neuronal cell death47 also occur at this time, which may explain why repetitive neonatal pain persistently alters pain processing in rats, mice, and humans.48–52 Very few animal experiments (rodents or nonhuman primates) have studied surgical or other noxious stimuli during exposure to anesthetics. It is important, therefore, to study the mechanisms by which repetitive pain alters development in the neonatal brain through factors altering cell survival, neuronal activity, and plasticity, and the relationship between pain and the analgesic and antiinflammatory effects of anesthetics.
It has been shown in previous studies53 that peak vulnerability to the apoptogenic action of anesthetics is during synaptogenesis, also known as the brain growth spurt. This is a period during which the brain grows at an accelerated rate because newly differentiated neurons throughout the brain are rapidly expanding their dendritic arbors to provide the required surface area to accommodate new synaptic connections. NCAM is believed to be an important regulator of developmental and functional neuroplasticity. In particular, embryonic PSA-NCAM has a vital role in forming connections between neurons.54 Synaptophysin is a synaptic vesicle-associated protein that is involved in synaptogenesis. Interestingly, our data show that PSA-NCAM (Fig. 3A) is partially co-localized with synaptophysin (Fig. 3B) in the neuronal cell membrane in the organotypic slice cultures (control) during development (Fig. 3). Therefore, PSA-NCAM activity seems critical to the process controlling the trafficking and targeting of vesicular proteins to the synapse.
|
The sialylation state of PSA-NCAM is controlled by developmentally regulated Golgi sialyltransferase activity.55 This transferase activity is Ca2+ dependent,56 which may account for its regulation by NMDA receptors.54,57 The regulation of PSA-NCAM expression by NMDAergic activity plays a critical role in neuroplasticity during development, particularly in NCAM-mediated cell-cell interactions and synapse formation.58 In our previous study,27 treatment of frontal cortical cultures from the developing monkey with ketamine caused a substantial decrease in mitochondrial metabolism of [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide], along with a concomitant decrease in PSA-NCAM protein expression (Fig. 4). The decrease in PSA-NCAM corresponded to an approximately 40% decrease in PSA-NCAM immunoreactivity, which decrease could be the direct result of local NMDA receptor blockade (subsequent reduction in Ca2+-regulated polysialyl transferase activity) or the indirect result of neuronal loss.27,56 The fact that SN-50 (a peptide inhibitor of NF-kB transport) dose-dependently blocked the ketamine-induced cortical neuronal cell death, as well as the loss PSA-NCAM immunoreactivity in culture, argues for the latter mechanism. Future experiments using N-butanoyl-mannosamine to inhibit polysialyl transferase or endoneuramididase N to selectively cleave PSA chains may be able to specifically address this hypothesis.
|
IN VIVO AND IN VITRO NONHUMAN PRIMATE AND RODENT MODELS
Ketamine-Induced Neuronal Cell Death in the Perinatal Rhesus Monkey (In Vivo)
Although it is clear that anesthetics cause neuronal cell death in the rodent model when given repeatedly during the brain growth spurt period,26,53 it is not yet known whether a similar phenomenon also occurs in primates. To better determine if ketamine-induced neurodegeneration in the developing rat has clinical relevance, ketamine was examined in a nonhuman primate model that more closely mimics the developing pediatric population.27,59 The similarity of the physiology, pharmacology, metabolism, and reproductive systems of the nonhuman primate to that of the human, especially during pregnancy, allow the monkey to be an exceptionally good animal model to detect the potential neurodegenerative effects of ketamine. In our monkey study,35 important maternal and infant physiological variables, including percent oxygen saturation, exhaled carbon dioxide, body temperature, heart rate, arterial blood pressure, glucose, and hematocrit were all monitored and maintained within normal ranges. This is an essential component of any animal model, and is far easier in a primate than a rodent. Because prolonged hypoperfusion can lead to cerebral hypoperfusion and ischemic-related cell death, it is particularly important to ensure normal blood pressure and oxygen saturation.60
Placental transfer of ketamine occurs rapidly. Fetal blood levels have been reported to be comparable to maternal blood levels within 2 min of maternal administration.61 Ketamine is distributed rapidly after IV administration and plasma levels decrease rapidly thereafter.62,63 In our recent publication,35 we reported that the ketamine levels required to maintain anesthesia were 10 to 25 µg/mL, which is 5 to 10 times higher than those observed in humans (2–3 µg/mL). It is important to note that the plasma concentrations of ketamine were highest in the PND 35 monkeys, even though no evidence of neuronal cell death was observed compared with control animals of the same age. In the PND 5 monkeys, where neuronal cell death was evident, plasma levels averaged approximately 10 µg/mL (Fig. 5), which is only three to five times the plasma levels observed in humans.
|
The window of vulnerability to the neurotoxic effect of ketamine in the primate is restricted to the period of rapid synaptogenesis, which occurs during the perinatal stages, occurring at least by 75% of gestation and lasting to early postnatal life, ending sometime before PND 35.35 However, in the rat, the window of vulnerability begins 1 day after birth and ends approximately 14 days later.53 This is a period in the rat before the complete development of motor and other relevant systems during intense synaptic remodeling in multiple brain regions. This is also a period of rapid myelin formation during which most afferent pathways are already present in their target areas, although their distribution and synaptic targets are still immature. During these sensitive stages, altered glutamatergic neurotransmission and receptor expression induced by NMDA antagonists (such as ketamine) could affect neuroplasticity and cause neuronal toxicity. Thus, abnormal neuronal development, abnormal synaptic plasticity, and neurodegeneration have been proposed as causal or contributing factors in anesthetic-induced neurotoxicity.27 The period of vulnerability of the immature rat brain to ketamine-induced neuronal cell death coincides with the critical stages of monkey brain development. GD 122 fetuses and PND 5 infants are undergoing synaptogenesis and appear much more vulnerable than PND 35 monkeys that are undergoing much less synaptogenesis. Although the precise correlation between stages of brain development in humans and monkeys and/or species-specific vulnerabilities to anesthesia-induced neuronal damage cannot be specified, clearly, matching such events between human and nonhuman primates is less problematic than matching these phenomena between primates and rodents. For example, according to a recent study, the gestational day 122 monkey fetus is equivalent to the 199 gestational day human fetus as determined by cortical development, and both are in the range of 75% to 80% of normal term.64
It should be mentioned that the pattern of topography, characteristics, and neuronal susceptibility to neurotoxic insult that leads to apoptosis or necrosis are not clear, and may depend on the concentration, the duration of exposure, the receptor subtype activated and the cell type, as well as neuronal stage of development or maturity.32,33,35 Among these factors, the concentration and duration are thought to be the most significant contributors. How do concentration and duration relate to neuronal toxicity?
The dosing paradigms used in animal studies typically do not reflect doses used for pediatric patients. The doses and durations of exposure used in the animal studies typically exceed those used clinically, sometimes substantially. While this is an important observation, it is a key component to preclinical research. Once a toxic dose and effect have been identified, it is possible to evaluate lower doses and/or shorter durations to determine an exposure level associated with no adverse effects, which can than be compared with clinically relevant dosing paradigms to establish the existence, or lack thereof, of a safety margin. This approach is used for drug development. In general, we cannot interpret duration without considering the concentration that was maintained, nor can we consider concentration without consideration of the duration of the exposure.
Another goal is to determine whether there is an anesthetic duration below which no significant ketamine-induced neuronal cell death can be detected. In our study, PND 5 monkeys were evaluated after 3 and 24 h of ketamine anesthesia (IV infusion at a rate of 20 to 50 mg · kg–1 · h–1 to maintain a light surgical plane of anesthesia). The 24-h duration was selected as a relatively long duration, while the 3-h duration was applied as a relatively short duration. No significant neurotoxic effects were observed if the anesthesia duration was 3 h. By contrast, the 24-h anesthetic sessions produced a significant increase in the number of caspase 3-positive neurons, Silver-stained cells, and Fluoro-Jade C-positive cells in layers II and III of the cortex (Fig. 6). These data are consistent with those observed in our time-course studies using a primary cortical culture system established from developing rats and PND 3 monkeys.26,27 In these time-course studies, the cultures were exposed to ketamine for 2, 4, 6, 12, and 24 h. The data indicated that the addition of 10 µM ketamine results in about a 30% loss in cell viability at 6 h, and an approximately 50% to 70% loss after 12 to 24 h of exposure. However, there were no significant differences between control and ketamine-exposed cultures at 2 h. After 4 h of exposure, a slight but nonsignificant increase in neuronal cell death was observed. The data suggest that continuous activation of up-regulated NMDA receptors (compensatory) over 6 to 24 h is critical for ketamine-induced cell death in developing neurons in monkey frontal cortical cultures.27 Ketamine-induced cell death was also exposure-time dependent in vivo.
|
Application of Rodent and Nonhuman Primate In Vitro Models in the Evaluation of Anesthetics During Development
Both in vitro and in vivo approaches have been used to assess the neurotoxicity associated with a wide range of drugs at a variety of doses and exposure durations. We have used in vitro systems (primary culture26,27,33 and organotypic slice culture34,65) that parallel our in vivo studies28,35 to assess the effects of anesthetic exposure on the developing nonhuman primate and rodent models. Primary frontal cortical culture systems (Fig. 7) and organotypic slice cultures (Fig. 8), established using tissues from monkey and rodent, provide parallel in vitro models that assist in evaluating the neurotoxicity of various anesthetics at a variety of doses using a minimal number of animals in a short period of time.
|
|
These in vitro preparations are useful for the rapid evaluation of the neurotoxic effects of anesthetic drugs and enable direct study of the brain at various stages of development. Primary (Fig. 7) and organotypic (Fig. 8) cultures maintain important anatomical relationships and synaptic connectivities, allow for direct assessment of cell death, and are reliable models for screening and evaluating the neurotoxicity of different anesthetic drugs. In addition, these preparations allow for the direct application of antisense oligodeoxynucleotides that target specific receptor genes, as well as direct enzymatic and therapeutic drug treatment. This approach allows for the collection of a large amount of data from a minimal number of subjects and allows for the investigation of cellular mechanisms associated with anesthetic-induced cell damage in simplified primate or rodent systems.
PHARMACOGENOMIC/SYSTEM BIOLOGY APPROACHES
Application of Pharmacogenomic/System Biology Approaches at mRNA/DNA Levels (Genomics)
Microarrays
Gene microarray techniques have been used to determine which of several candidate death genes might be associated with anesthetic drug-induced apoptosis. Microarrays facilitate analysis of the relationship between anesthetic drug-induced neurotoxicity and changes in gene expression. Therefore, gene expression profiles for different anesthetic drugs at different time points were compared.
Tissue (In Vivo and In Vitro).
Frozen tissue was sectioned using a cryostat before laser capture microdissection (LCM). Briefly, the tissue was embedded in a mold containing Optical Cutting Temperature compound. This tissue-containing cryomold was allowed to equilibrate on dry ice and was then frozen onto the cutting stage of the cryostat. Approximately 7 to 10 µm sections were cut and immediately adhered to chilled microscope slides which were kept on dry ice until ‘same-day’ processing for LCM or stored at –80°C for future processing. Keeping the tissue frozen before LCM ensures that high quality mRNA is obtained from the cells collected. Cells from the same brain region were collected from each subject to be used as part of the systems biology study. Total RNA was isolated from these cells and an RNA amplification kit was used to generate labeled cRNA for microarray experiments. Briefly, double-stranded (ds) cDNA was synthesized from the isolated RNA before in vitro transcription to generate antisense RNA, which was then used in a second round of amplification to generate additional ds cDNA. This ds cDNA was amplified along with cyanine 3-CTP or cyanine 5-CTP dye to generate labeled antisense cRNA.
Oligonucleotide Microarrays (In Vivo and In Vitro).
The labeled RNA, along with labeled reference RNA, was co-hybridized onto a rat oligo microarray containing over 22,000 sequences. The microarray slides were scanned and the data analyzed using ArrayTrack in-house software (Figs. 9 and 10).
|
|
Application of Pharmacogenomic/System Biology Approaches at Protein Levels
Proteomics
To identify specific proteins that are correlated with NMDA antagonist/GABA agonist-induced apoptosis, 2-dimensional polyacrylamide gel electrophoresis or a Phosphoprotein Isotope-coded Solid-phase Tag method can be used to separate proteins from tissue samples. Excision or digestion followed by affinity isolation can also be used to separate specific proteins. Quantification and identification can be performed by matrix-assisted laser desorption/ionization-mass spectrometry.35
Western Blot Analysis (In Vitro and In Vivo)
Antisense oligonucleotides were used to determine the effect of co-administration of NMDA receptor antisense oligonucleotides on anesthetic drug-induced neurotoxicity and to determine whether the administration of antisense oligonucleotides targeted to the NR1 NMDA receptor subunit blocked steady-state protein levels. To determine whether decreased PSA-NCAM expression is associated with ketamine-induced loss of cortical neurons or to local NMDA receptor blockade (as determined by polysialyl transferase activity), Western blot analyses of PSA-NCAM/NR1 protein/Actin ratios and the protective effects of the antisense oligonucleotides were examined.26 Co-administration of NR1 antisense oligonucleotide was able to almost completely block the neuronal cell death induced by ketamine (Fig. 11). Our data indicate that ketamine significantly up-regulated NMDA receptor NR1 subunit protein. Co-administration of antisense oligonucleotide specifically prevented NR1 up-regulation and blocked the reduction of PSA-NCAM expression induced by ketamine.
|
PERINATAL ANESTHETIC ADMINISTRATION AND LONG-TERM BEHAVIORAL DEFICITS
The loss and maintenance of synapses during development are believed to be controlled by an energy-dependent mechanism involving the NMDA receptor, similar to those involved in long-term potentiation.66,67 The administration of NMDA receptor antagonists, such as ketamine, phencyclidine (PCP), and MK-801, to rats during a critical time of perinatal development results in neuronal cell death in several major brain areas.10,53 In 1999, Ikonomidou et al.53 demonstrated severe widespread apoptotic degeneration throughout the rapidly developing brain of the 7-day-old rat pup after ketamine administration. It has been reported that exposure to a drug combination (midazolam, nitrous oxide, and isoflurane) causes widespread neurodegeneration in the developing rat brain and persistent learning deficits.68 Our previous studies have also demonstrated that repeated administration of PCP (a non-competitive NMDA antagonist) results in a sensitized locomotor response in rats subjected to later PCP challenge.69 This sensitization is associated with apoptotic cell death and an increase in NMDA receptor NR1 subunit mRNA and immunoreactivity of the NMDA receptor in rat forebrain.70,71 The neurodegeneration and associated deficits in prepulse inhibition were prevented by treatment with a superoxide dismutase mimetic, M40403,72 suggesting an important role of superoxide anions in NMDA antagonist-induced apoptosis and behavioral alterations.
Previous in vivo studies have also demonstrated that perinatal PCP administration results in profound behavioral abnormalities in the adolescent rat that may be related to enhanced apoptotic cell death of neurons in the frontal cortex.73 These cortical deficits may have a significant impact on the function of subcortical structures such as the nucleus accumbens, which serves as an important regulatory center by integrating the functions of the basal ganglia and the limbic system. The nucleus accumbens receives glutamatergic afferents from several brain regions, in particular, the frontal cortex.74–77 Medium spiny neurons account for the majority of the neostriatal cell population and are a major synaptic target of dopaminergic input to the striatum.78–81 Thus, alteration of the cortical input to these neurons during development may play a significant role in mediating the behavioral effects of perinatal PCP/ketamine treatment later in life (Fig. 12).
|
CLINICAL CORRELATION OF PRESENT DATA
Sedatives and general anesthetics have been used for years in pediatric patients without overt clinical evidence of CNS sequelae. Although the doses and durations of ketamine exposure that resulted in neurodegeneration were substantially larger than those used in the clinical setting, those associated with isoflurane were in the same range as the human.68 There are insufficient human data to either support or refute the clinical applicability of rodents and nonhuman primate findings. However, moderate adverse effects related to CNS function in pediatric populations may be difficult, if not impossible, to detect.
Drugs that block the NMDA receptor subtype of glutamate and/or positively modulate or gate the GABAA receptor have been associated with apoptotic neuronal cell death in developing rodents.35,65,68,82 To induce or maintain a surgical plane of anesthesia, it is a common practice in pediatric or obstetrical medicine to use a combination of drugs from these two classes.
NMDA antagonists and GABA agonists are often used in combination during general anesthesia. For example, the anesthetic gas nitrous oxide, an NMDA receptor antagonist, and isoflurane, a volatile anesthetic that acts on multiple receptors including postsynaptic GABA receptor, are commonly used together. Thus, another important goal of our anesthetic studies was to determine if a combination of NMDA antagonists and GABA agonists would prevent or enhance each other’s effects (including potential neuronal cell death). In PND 7 rat pups, an enhancement in brain damage was noted when nitrous oxide (75%) was combined with isoflurane (0.55%). Maximal neuronal cell death was observed after 6 to 8 h of exposure. There were no significant effects from 2 h of exposure (Fig. 13) in PND 7 rat brain. These findings are consistent with previous studies performed using the same animal model (68). Our data28 indicate that a low dose of isoflurane (0.55%) showed no significant enhancement of apoptotic cell death in the brain. In contrast, relatively high doses of isoflurane (0.75, 1.0, and 1.5%)68 may produce neurodegeneration in multiple brain regions.
|
Immature GABA receptors are excitatory early in development and convert to an inhibitory role in mature neurons.83 It can be postulated that prolonged supra-physiologic stimulation of immature neurons by GABA agonists enhances the excitatory component in the action of GABA and may contribute to increased excitability during early development.24 This increased excitability, along with NMDA antagonist-induced alteration of NMDA receptors, could lead to neuronal cell death. Our data28 indicate that a significant effect was observed in the frontal cortex when this low dose of isoflurane (0.55%) was combined with nitrous oxide (75%). Therefore, apoptosis is found in many neuronal populations after a higher dose of isoflurane (0.75%, 1.0%, and 1.5%) combined with nitrous oxide (75%), but is limited to the cortex after a low dose of isoflurane (0.55%) combined with nitrous oxide (75%).
Additive toxicity between a nontoxic concentration of an NMDA antagonist and a GABAmimetic drug is also seen with combinations such as ketamine and midazolam. Additional experimental models (in vitro and in vivo studies of nonhuman primate and rodent models) will be necessary to confirm this conclusion.
POTENTIAL NEUROPROTECTION
l-carnitine plays an integral role in attenuating neurological brain injury associated with mitochondria-related degenerative disorders. l-carnitine is an l-lysine derivative and its main role lies in the transport of long chain fatty acids into mitochondria to enter the β-oxidation cycle.84 Another important property of this agent is the neutralization of toxic acylCoA production in the mitochondria,85 which correlates with various pathological processes, including organic aciduria86 and numerous diseases of the CNS such as neurodegenerative diseases,87–89 ornithine transcarbamylase deficiency,90,91 and other mitochondrial diseases.85 l-carnitine administration may offer a straightforward approach to mitigating neurotoxic effects. Such studies are underway.
Another group of molecules regulating mitochondrial function and stability is BCL-2. Although the precise mechanism by which BCL-2 family members act remains unclear, it has been established that they play a key role in the mitochondrial apoptotic pathway.92 Among the potential downstream regulators, the effect of inhaled anesthetics on Bax and BCL-XL has been measured. Bax is a proapoptotic protein, a pore-forming cytoplasmic protein that translocates to the outer mitochondrial membrane, influencing its permeability and inducing cytochrome c release from the intermembrane space of the mitochondria into the cytosol, subsequently leading to cell death.93 The anesthetic combination [nitrous oxide (75%) with isoflurane (0.55%)] resulted in a significant up-regulation of Bax protein compared with control (Fig. 14), and this effect was blocked by the co-administration of l-carnitine (300 or 500 mg/kg). These data suggested that increased superoxide anion activates an apoptotic pathway involving regulation of many genes, including Bax and BCL-XL.
|
Melatonin is produced at night by the pineal gland and promotes sleep. Melatonin functions as a direct free oxygen radical scavenger and indirect antioxidant, reducing the toxicity of a large number of drugs.94 It has been reported that melatonin suppresses apoptosis in cultured pineal cells by up-regulating Bcl-XL which in turn inhibits cytochrome c release and caspase-3 activation, thus blocking the apoptotic cascade activation.95 Yon et al.96 reported that co-administration of an anesthesia cocktail (midazolam, isoflurane, and nitrous oxide) with melatonin reduces the severity of anesthesia-induced damage in the developing rat brain. This study has demonstrated that melatonin provides significant protection against anesthesia-induced neuroapoptotic damage in the developing brain of the immature rats. Although the exact mechanism by which melatonin protects against apoptotic cell death is not known, this study has show that the neuroprotective effect is mediated, at least in part, via mitochondria-dependent apoptotic cascade. It appears that key elements involve the stabilization of the inner mitochondrial membrane, which in turn control cytochrome c release and apoptotic cascade activation. Several mechanisms involved in inner membrane stabilization by melatonin have been suggested. One involves a decrease in mitochondrial protein and DNA damage and the improvement of adenosine triphosphate synthesis by scavenging oxygen97 and nitrogen-based reactants generated in mitochondria, which in turn control the loss of the intra-mitochondrial glutathione.98 Melatonin may also stabilize the inner membrane by increasing the efficiency of the electron transport chain and by controlling the reduction potential.99 In addition, direct action of melatonin on the control of currents through the mitochondrial transition pores100 has been demonstrated. Jevtovic-Todorovic and her colleagues95 have shown that melatonin stabilizes the inner mitochondrial membrane by increasing the protein levels of Bcl-XL. It is most likely that multiple, rather than a single, mechanisms contribute to melatonin-induced restoration of mitochondrial function.
CONCLUSION
Exposures of developing mammals to anesthetics, including those that block NMDA-type glutamate receptors and those that activate GABA receptors, affect the endogenous neuronal transmission systems and enhance neuronal cell death in a dose and developmental stage-dependent manner. This review emphasizes the incorporation of in vivo models with more circumscribed in vitro preparations in the developing rodent and nonhuman primate. These combined models provide the opportunity for the rapid evaluation of anesthetics over a wide range of doses and exposure durations. This combination of models enables the collection of a large amount of data from a minimal number of subjects and allows for the investigation of cellular mechanisms associated with anesthetic-induced cell loss in simplified nonhuman primate or rodent systems.
In comparing the rodent and the nonhuman primate data, it is clear that the same principles apply: drug exposure (measured as dose, concentration, duration of exposure, or area under the concentration versus time curve) and neuronal developmental stage. The duration of anesthetic exposure to induce cell death, as measured by minimal exposure requirements, is similar (4–6 h) for nonhuman primate and rodent brain cells in culture. The duration of anesthetic exposure is also similar between rodents in vivo (4–6 h) and nonhuman primates (more than 3 h, although less well defined). The susceptible stage or period of development has not been completely described but begins somewhere before the last quarter of pregnancy and continues to shortly after birth in the nonhuman primate. This developmental stage is similar to the period of rapid synaptogenesis in the rodent nervous system that occurs between 1 and 14 days postgestational age.
Application of pharmacogenomic and systems biology approaches has great potential for helping advance the understanding of brain-related biological processes, including neuronal plasticity and neurotoxicity. These approaches may also allow for monitoring efficacy of treatment regimens. In addition, by using in vivo and in vitro nonhuman primate and rodent models, these approaches may enhance our understanding of complex biological processes such as neuronal cell death (apoptosis and/or necrosis) induced by anesthetics in the developing brain. Understanding these complex biological processes will elucidate pathways to predict anesthetic-induced brain cell death, and help discover treatments to ameliorate the consequences (if any) of anesthetic toxicity in pediatric patients.
Footnotes
Accepted for publication March 10, 2008.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  A. W. Loepke, G. K. Istaphanous, J. J. McAuliffe III, L. Miles, E. A. Hughes, J. C. McCann, K. E. Harlow, C. D. Kurth, M. T. Williams, C. V. Vorhees, et al. The Effects of Neonatal Isoflurane Exposure in Mice on Brain Cell Viability, Adult Behavior, Learning, and Memory Anesth. Analg., January 1, 2009; 108(1): 90 - 104. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. X. McGowan Jr and P. J. Davis Anesthetic-Related Neurotoxicity in the Developing Infant: Of Mice, Rats, Monkeys and, Possibly, Humans Anesth. Analg., June 1, 2008; 106(6): 1599 - 1602. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
