- Email: [email protected]
Do white cells matter in white matter damage?
320
Opinion
TRENDS in Neurosciences Vol.24 No.6 June 2001
Do white cells matter in white matter damage? Olaf Dammann, Scott Durum and Alan Leviton Support is provided for the hypothesis that activated leukocytes, especially monocytes/macrophages, contribute to cerebral white matter damage in extremely low gestational age newborns. Much of the evidence is indirect and comes from analogies to brain diseases in adults, and from models of brain damage in adult and newborn animals. If the recruitment of circulating cells to the brain contributes to white matter damage in extremely low gestational age newborns, then minimizing the transendothelial migration of circulating cells by pharmacological manipulation might prevent or reduce the occurrence of neonatal white matter damage and the disabilities that follow.
Olaf Dammann Alan Leviton* Children’s Hospital and Harvard Medical School, Boston, MA, USA. e-mail: leviton@ Hub.TCH.Harvard.edu Scott Durum National Cancer Institute, Bethesda, MD, USA.
Neonatal white matter damage, which is observed in ~10% of extremely low gestational age newborns1, can result in motor, cognitive, perceptual, visual, behavior and attention limitations2–4. Usually defined on a cranial ultrasound scan as a cerebral white matter echolucency (because the reduced echo implies loss of brain substance), neonatal white matter damage is often identified as periventricular leukomalacia. This name is technically inappropriate because malacia means softening, which cannot be ascertained by ultrasound examination. Alternative names, such as preterm anoxic injury, ischemic lesions, and ischemic and reperfusion brain injury imply what some consider the most probable explanation for these images5. Excitotoxic damage to neonatal white matter6,7, and damage caused by oxidative stress are other possible explanations8. However, recent publications favor the hypothesis that an inflammatory process precedes or accompanies neonatal white matter damage and its motor consequences9–21. Support for this hypothesis is based on measurements of inflammatory cytokines. In this review, we move beyond the contribution of cytokines in neonatal white matter damage and suggest a role for activated leukocytes, especially the cell identified as a monocyte in the blood and as a macrophage in the brain. The strongest support for this hypothesis is that the monocyte/macrophage is seen in neonatal white matter damage22,23. Previously, the presence of this cell in white matter damage was viewed as support for the hypothesis that its main role is to clean up debris24. Now, however, monocytes/macrophages, and perhaps other leukocytes in the brain, appear to contribute to the damage25–27. Each of the components of the inflammatory process that precede or accompany preterm birth, and
neonatal white matter damage and its functional consequences, can facilitate the transendothelial migration of leukocytes. This leads us to postulate that the inflammatory response and other stimuli that stimulate leukocyte adherence, such as hypoxia and reactive oxygen metabolites28, enable the transendothelial migration of leukocytes into the brain, which then contribute to perpetuating white matter damage. In the absence of evidence that leukocytes other than the macrophage or monocyte are involved in neonatal white matter damage, we review the clinical and related neurobiology literatures to support our hypothesis that additional evidence for leukocyte involvement in neonatal white matter damage will be forthcoming. Specifically, we address three hypotheses: (1) some of the inflammatory cells in the area of the damage are derived from cells outside the brain; (2) these cells contribute to brain damage and (3) interfering with leukocyte migration into the brain will reduce brain damage. The fetal inflammatory response and transendothelial migration of leukocytes
Markers of inflammation in and around the fetus predict second trimester labor29 and/or rupture of membranes30. In the amniotic fluid of pregnancies, characterized by both preterm labor and evidence of infection, most of the neutrophils are of a fetal origin31. Thus, a sizable component of the inflammatory process that presumably contributes to delivery much before term is probably of fetal origin.
…we move beyond the contribution ‘… of cytokines in neonatal white matter damage and suggest a role for activated leukocytes,’ The fetal inflammatory response, which often appears to be triggered by bacteria in the amniotic fluid, is identified histologically by the presence of inflammatory cells in the walls of the vessels of the umbilical cord (i.e. funisitis), and by increased expression and/or higher concentrations of inflammatory cytokines in the amniotic fluid or blood32. These inflammatory phenomena also predict neonatal white matter damage and its consequences9–13. This has led to the hypothesis that a fetal inflammatory response contributes to brain damage in the extremely low gestational age newborn33. The maternal and fetal inflammatory responses include the involvement of cytokines34, chemokines35, adhesion molecules36, extracellular matrix proteins with adhesive properties37, and matrix metalloproteinases38. All five of these groups facilitate the transendothelial migration of cells out of the circulation and into adjacent organ parenchyma. We
http://tins.trends.com 0166-2236/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0166-2236(00)01811-7
Opinion
TRENDS in Neurosciences Vol.24 No.6 June 2001
Fetal endothelium
Fetal systemic circulation
321
BBB
Fetal brain Oligodendrocyte Astrocyte
Activate Activate Activated leukocyte Pre-OL Cytokines Activate Activated leukocyte
Microglia TRENDS in Neurosciences
Fig. 1. Hypothetical model of neonatal white matter damage. Proinflammatory cytokines in the fetal systemic circulation activate white blood cells (left), which render the fetal blood–brain barrier (BBB) ineffective and migrate through the BBB into the fetal brain. Here, they activate resident astrocytes and microglia, which, in turn, adversely affect the development of oligodendrocyte precursor cells (pre-OL) and the well-being of mature oligodendrocytes. Astrocytes and microglia are also engaged in a bilateral sustained activation (right).
reason that if all the components needed for the transendothelial migration of leukocytes are in place when neonatal brain damage begins, then transendothelial migration of leukocytes into the brain is possible. To date, only two of the five protein groups that constitute the chemically defined fetal inflammatory response, cytokines and chemokines, have been associated with neonatal white matter damage or cerebral palsy, the most prominent motor consequence of neonatal white matter damage10,11,14–16. The presence of a third group, adhesion molecules28 is suggested by the associations of funisitis with neonatal white matter damage9 and neurologic impairment11,12. These associations support the possibility that the fetal inflammatory response aids the transendothelial migration of leukocytes into the brain, where they contribute to brain damage. The process of transendothelial migration probably begins with inflammatory cytokines39 (Fig. 1). In addition to activating endothelial cells, these cytokines activate leukocytes in the circulation, diminish the effectiveness of the blood–brain barrier40, which is relatively ineffective in the preterm newborn41, and gain access to and activate microglia and astrocytes, which in turn produce chemokines and more cytokines. The chemokines function as chemo-attractants for activated inflammatory cells, including those in the circulation. Some adhesion molecules facilitate the initial tethering of leukocytes to the vessel wall, followed by their rolling along the endothelium until they adhere tightly to the endothelial surface. Other adhesion molecules then facilitate the release of the leukocytes from the vessel wall, allowing movement through the intercellular junctions into the underlying tissue. The http://tins.trends.com
matrix metalloproteinases further disrupt the blood–brain barrier and facilitate cellular migration through the extracellular matrix42. If leukocytes are commonly involved in other CNS disorders, and can be shown to damage the brain, then perhaps they also contribute to neonatal white matter damage. The role of leukocytes in various forms of brain injury is reviewed below. Table 1 shows which cells are involved in each of these disorders. Presence of leukocytes in areas of brain damage Neonatal white matter damage in human newborns
Most of the infants who are at risk of white matter damage and who die, tend to do so within days of birth, and weeks before the white matter damage would be fully developed23, thereby limiting opportunities to study the development of white matter damage in the most vulnerable humans. The characteristic development of white matter damage usually begins with hypertrophic astrocytes, and not with cells that are unquestionably derived from the blood22. Nevertheless, macrophages are a constituent of neonatal white matter damage22,23. Although some monocytes, macrophages and microglia in the brain are derived from resident cells, others are derived from the circulation27. Some of these cell members of the leukocyte family are considered here. Cerebral palsy
Newborns who later develop cerebral palsy have elevated blood concentrations of cytokines and chemokines10,15,16. These can activate leukocytes and guide their movement out of the blood vessel and into the brain. Laboratory models
Neutrophils and macrophages are early features of white matter damage in neonatal cats and dogs given endotoxin43,44. In neonatal rats, neutrophil and T-cell infiltration are also seen in the brain following cerebral ischemia and reperfusion45. In newborn mice, systemically administered interleukin-9 (IL-9) can exacerbate white matter damage attributed to excitotoxicity6,7. Activated T cells are presently the
322
Opinion
TRENDS in Neurosciences Vol.24 No.6 June 2001
Table 1. Cells implicated in different biological entitiesa Biological entity
Neutrophils
Monocytes
T lymphocytes
Infection Trauma Hemorrhage Ischemia Cerebral palsy Neonatal white matter damage Human Laboratory animal
+ + + + Chemokines
+ + + + Chemokines
+ + + + Chemokines
+
+ +
+ sign indicates the presence of the leukocyte. ‘Chemokines’ appears on the cerebral palsy line because the cell listed at the top of each column has not been associated with cerebral palsy, but the chemokines that serve as a chemo-attractant for that leukocyte have been associated with cerebral palsy.
aA
only known source of IL-9 (Ref. 46). Does this mean that activated T cells are involved in neonatal white matter damage? Following CNS trauma in adult rodents, chemokines are expressed47, adhesion molecules are upregulated on endothelium48, activated neutrophils adhere to endothelial cells49, and neutrophils, monocyte/macrophages, and T cells invade the nervous system47. Similarly, cerebral ischemia in adult rodents followed by reperfusion rapidly leads to the expression of inflammatory cytokines50, metalloproteinases51 and chemokines52. Focal CNS damage is accompanied by distant damage
Newborn cerebral white matter damage, which probably includes damage to oligodendrocytes and their precursors, has both focal and diffuse effects53. The diffuse white matter damage, which often accompanies focal white matter damage, appears to explain a large proportion of the developmental disorders seen in children who were born at an extremely low gestational age54. This concern with diffuse white matter damage prompts us to pay special attention to diffuse and distant injury. We suggest that the penumbra of ischemic damage and the distant damage that is sometimes seen following trauma are examples of a process that might be a model for the diffuse form of white matter damage seen in newborns. Following cerebral ischemia, ‘multiple molecular penumbras’ are found at some distance from the infarct55. The secondary, delayed, and either adjacent or distant injury that sometimes follows CNS injury in rodents is characterized by a series of deleterious, often inflammatory, events that promote progressive tissue damage56. Some of the ‘secondary’ brain damage attributed to meningitis appears to be caused by products of inflammation57. Bacterial components and their products stimulate the production of large amounts of cell-activating cytokines and chemokines58. Because the products of inflammation that appear to account for the secondary damage following meningitis are the same as those that characterize the fetal inflammatory response, the mechanisms of brain damage might also be similar. http://tins.trends.com
Sustained activation of leukocytes and resident brain cells
Cell activation, initiated by the binding of a ligand to its receptor, enhances the ability of the cell to respond to other stimuli59. Sustained cell activation should lead to more damage than a short-lived inflammatory cytokine cascade10. Recruiting activated leukocytes to cerebral white matter appears to result in sustained activation of microglia and astrocytes27, which, in turn, produce a prolonged supply of inflammatory cytokines locally, where they do their damage (Fig. 1). This ‘inflammation-boosting loop’ has been suggested as an explanation for sustained inflammation60. Support for the sustained activation hypothesis is derived from multiple sources. In a neonatal rat model of cerebral ischemia and hypoxia, the activation of lymphocytes, microglia, macrophages, and astroglia persists for ~5 weeks61. Ischemia in the adult rat cerebrum increases the expression of chemokines, which are viewed as stimuli for ‘prolonged leukocyte recruitment, and astrocyte migration and activation’52. Following the placement of extravascular blood in rat brain, the microglial reaction and infiltration by neutrophils and lymphocytes, and the programmed death of brain cells continue for more than three weeks62. In a model of spinal cord trauma in rats, reactive astrocytes remain prominent in the epicenter of the damage for up to 28 days63. Leukocytes contribute to brain damage
All the leukocyte forms contribute to tissue damage25–27. Macrophages are among the cells that appear to contribute to the ‘expansion of infarction’ in the penumbra following focal cerebral ischemia in rats64. In several different laboratory animal species, the brain damaging consequences of ischemia can be diminished by depleting the pool of available leukocytes or by inhibiting leukocyte activation and function65,66. Knockout mice deficient in adhesion molecules also show reduced injury following focal cerebral ischemia67. Both the adherence of activated leukocytes to endothelial cells and the motor consequences of cord compression can be attenuated by leukocytopenia (a decrease in the number of white blood cells) and by interventions that reduce the availability of activated leukocytes to endothelial cells and cord parenchyma49. The evidence that activated leukocytes are involved in this secondary injury has prompted the search for therapies that reduce the probability of leukocyte activation68. Systemic responses predict, accompany and perhaps contribute to CNS damage
We postulate that unlike cerebral ischemia and trauma, in which the inflammation begins in the brain, white matter damage in newborns begins with a systemic inflammatory response that is initiated
Opinion
TRENDS in Neurosciences Vol.24 No.6 June 2001
outside, and then gains access to, the brain33. Support for this hypothesis is also derived from the stroke literature. Following stroke, adults have elevated numbers of peripheral blood cells expressing inflammatory cytokines and other markers of cell activation, sometimes for as long as one month69,70. A component of some of these elevations might be a marker of damage. Alternatively, their duration and ability to predict deterioration suggests that a component contributes to brain damage. References 1 Baud, O. et al. (1999) Antenatal glucocorticoid treatment and cystic periventricular leukomalacia in very premature infants. New Engl. J. Med. 341, 1190–1196 2 Pinto-Martin, J.A. et al. (1995) Cranial ultrasound prediction of disabling and nondisabling cerebral palsy at age two in a low birth weight population. Pediatrics 95, 249–254 3 Whitaker, A.H. et al. (1996) Neonatal cranial ultrasound abnormalities in low birth weight infants: relation to cognitive outcomes at six years of age. Pediatrics 98, 719–729 4 Whitaker, A.H. et al. (1997) Psychiatric outcomes in low-birth-weight children at age 6 years: relation to neonatal cranial ultrasound abnormalities. Arch. Gen. Psychiatry 54, 847–856 5 Inder, T.E. and Volpe, J.J. (2000) Mechanisms of perinatal brain injury. Semin. Neonatol. 5, 3–16 6 Dommergues, M.A. et al. (2000) Proinflammatory cytokines and interleukin-9 exacerbate excitotoxic lesions of the newborn murine neopallium. Ann. Neurol. 47, 54–63 7 Tahraoui, S.L. et al. (2001) Central role of microglia in neonatal excitotoxic lesions of the murine periventricular white matter. Brain Pathol. 11, 56–71 8 Said, S.I. et al. (2000) NMDA receptor activation: critical role in oxidant tissue injury. Free Radic. Biol. Med. 28, 1300–1302 9 Leviton, A. et al. (1999) Maternal infection, fetal inflammatory response, and brain damage in very low birth weight infants. Developmental Epidemiology Network Investigators. Pediatr. Res. 46, 566–575 10 Nelson, K.B. et al. (1998) Neonatal cytokines and coagulation factors in children with cerebral palsy. Ann. Neurol. 44, 665–675 11 Yoon, B.H. et al. (2000) Fetal exposure to an intraamniotic inflammation and the development of cerebral palsy at the age of three years. Am. J. Obstet. Gynecol. 182, 675–681 12 Redline, R.W. et al. (2000) The relationship between placental and other perinatal risk factors for neurologic impairment in very low birth weight children. Pediatr. Res. 47, 721–726 13 Wu, Y.W. and Colford, J.M. Jr (2000) Chorioamnionitis as a risk factor for cerebral palsy: a meta-analysis. J. Am. Med. Assoc. 284, 1417–1424 14 Yoon, B.H. et al. (1996) Interleukin-6 concentrations in umbilical cord plasma are elevated in neonates with white matter lesions associated with periventricular leukomalacia. Am. J. Obstet. Gynecol. 174, 1433–1440 15 Yoon, B.H. et al. (1997) Amniotic fluid inflammatory cytokines (interleukin-6, interleukin-1beta, and tumor necrosis factoralpha), neonatal brain white matter lesions, and http://tins.trends.com
323
The future
To determine if activated leukocytes contribute to the occurrence of cerebral white matter damage in extremely low gestational age newborns will require measurement of phenomena that increase the probability of cell activation and movement of activated cells from the blood to the white matter parenchyma. If activated leukocytes contribute to brain damage in newborns, then perhaps strategies to minimize their activation or their migration into the brain might prevent or reduce the occurrence of neurologic disabilities.
cerebral palsy. Am. J. Obstet. Gynecol. 177, 19–26 16 Grether, J.K. et al. (1999) Interferons and cerebral palsy. J. Pediatr. 134, 324–332 17 Yoon, B.H. et al. (1996) Interleukin-6 concentrations in umbilical cord plasma are elevated in neonates with white matter lesions associated with periventricular leukomalacia. Am. J. Obstet. Gynecol. 174, 1433–1440 18 Martinez, E. et al. (1998) Elevated amniotic fluid interleukin-6 as a predictor of neonatal periventricular leukomalacia and intraventricular hemorrhage. J. Maternal–Fetal Invest. 8, 101–107 19 Duggan, P. et al. (1999) Relation of fetal inflammation to abnormal magnetic resonance images at birth in preterm infants. Pediatr. Res. 45, 908 20 Deguch, I.K. et al. (1997) Characteristic neuropathology of leukomalacia in extremely low birth weight infants. Pediatr. Neurol. 16, 296–300 21 Yoon, B.H. et al. (1997) High expression of tumor necrosis factor-alpha and interleukin-6 in periventricular leukomalacia. Am. J. Obstet. Gynecol. 177, 406–411 22 Gilles, F.H. et al. (1983) Developing Human Brain: Growth and Epidemiologic Neuropathology, John Wright–PSG Publishing Co. 23 Golden, J.A., et al. (1997) Frequency of neuropathological abnormalities in very low birth weight infants. J. Neuropathol. Exp. Neurol. 56, 472–478 24 Hickey, W.F. (1999) Leukocyte traffic in the central nervous system: the participants and their roles. Semin. Immunol. 11, 125–137 25 Nussler, A.K. et al. (1999) Leukocytes, the Janus cells in inflammatory disease. Langenbecks Arch. Surg. 384, 222–232 26 Yrjanheikki, J. et al. (1999) A tetracycline derivative, minocycline, reduces inflammation and protects against focal cerebral ischemia with a wide therapeutic window. Proc. Natl. Acad. Sci. U. S. A. 96, 13496–13500 27 Stoll, G. and Jander, S. (1999) The role of microglia and macrophages in the pathophysiology of the CNS. Prog. Neurobiol. 58, 233–247 28 Kevil, C.G. and Bullard, D.C. (1999) Roles of leukocyte/endothelial cell adhesion molecules in the pathogenesis of vasculitis. Am. J. Med. 106, 677–687 29 Mazor, M. et al. (1998) Cytokines in preterm parturition. Gynecol. Endocrinol. 12, 421–427 30 Fortunato, S.J. et al. (2000) Programmed cell death (apoptosis) as a possible pathway to metalloproteinase activation and fetal membrane degradation in premature rupture of membranes. Am. J. Obstet. Gynecol. 182, 1468–1476 31 Sampson, J.E. et al. (1997) Fetal origin of amniotic fluid polymorphonuclear leukocytes.
Am. J. Obstet. Gynecol. 176, 77–81 32 Romero, R. et al. (1998) A fetal systemic inflammatory response is followed by the spontaneous onset of preterm parturition. Am. J. Obstet. Gynecol. 179, 186–193 33 Dammann, O. and Leviton, A. (2000) Role of the fetus in perinatal infection and neonatal brain damage. Curr. Opin. Pediatr. 12, 99–104 34 Romero, R. et al. (2000) Further observations on the fetal inflammatory response syndrome: a potential homeostatic role for the soluble receptors of tumor necrosis factor alpha. Am. J. Obstet. Gynecol. 183, 1070–1077 35 Athayde, N. et al. (1999) A role for the novel cytokine RANTES in pregnancy and parturition. Am. J. Obstet. Gynecol. 181, 989–994 36 Steinborn, A. et al. (2000) Serum intercellular adhesion molecule-1 levels and histologic chorioamnionitis. Obstet. Gynecol. 95, 671–676 37 Goldenberg, R.L. et al. (2000) Vaginal fetal fibronectin measurements from 8 to 22 weeks gestation and subsequent spontaneous preterm birth. Am. J. Obstet. Gynecol. 183, 469–475 38 Maymon, E. et al. (2000) Matrilysin (matrix metalloproteinase 7) in parturition, premature rupture of membranes, and intrauterine infection. Am. J. Obstet. Gynecol. 182, 1545–1553 39 Weerasinghe, D. et al. (1998) A role for the alphavbeta3 integrin in the transmigration of monocytes. J. Cell Biol. 142, 595–607 40 de Boer, A.G. and Breimer, D.D. (1998) Cytokines and blood–brain barrier permeability. Prog. Brain Res. 115, 425–451 41 Adinolfi, M. (1985) The development of the human blood–CSF–brain barrier. Dev. Med. Child Neurol. 27, 532–537 42 Lukes, A. et al. (1999) Extracellular matrix degradation by metalloproteinases and central nervous system diseases. Mol. Neurobiol. 19, 267–284 43 Gilles, F.H. et al. (1977) Neonatal endotoxin encephalopathy. Ann. Neurol. 2, 49–56 44 Young, R.S. et al. (1983) Systemic and neuropathologic effects of E. coli endotoxin in neonatal dogs. Pediatr. Res. 17, 349–353 45 Benjelloun, N. et al. (1999) Inflammatory responses in the cerebral cortex after ischemia in the P7 neonatal Rat. Stroke 30, 1916–1924 46 Durum, S. (2000) Interleukins: overview. In Principles and Practice of the Biological Therapy Of Cancer (Rosenberg, S. ed.), Lippincott 47 McTigue, D.M. et al. (1998) Selective chemokine mRNA accumulation in the rat spinal cord after contusion injury. J. Neurosci. Res. 53, 368–376 48 Carlos, T.M. et al. (1997) Expression of endothelial adhesion molecules and recruitment of
324
49
50
51
52
53
54
55
Opinion
neutrophils after traumatic brain injury in rats. J. Leukocyte Biol. 61, 279–285 Taoka, Y. et al. (1998) Activated protein C reduces the severity of compression-induced spinal cord injury in rats by inhibiting activation of leukocytes. J. Neurosci. 18, 1393–1398 Stoll, G. et al. (1998) Inflammation and glial responses in ischemic brain lesions. Prog. Neurobiol. 56, 149–171 Gasche, Y. et al. (1999) Early appearance of activated matrix metalloproteinase-9 after focal cerebral ischemia in mice: a possible role in blood–brain barrier dysfunction. J. Cereb. Blood Flow Metab. 19, 1020–1028 Wang, X. et al. (1998) Prolonged expression of interferon-inducible protein-10 in ischemic cortex after permanent occlusion of the middle cerebral artery in rat. J. Neurochem. 71, 1194–1204 Leviton, A. and Gilles, F. (1996) Ventriculomegaly, delayed myelination, white matter hypoplasia, and ‘periventricular’ leukomalacia: how are they related? Pediatr. Neurol. 15, 127–136 Holling, E.E. and Leviton, A. (1999) Characteristics of cranial ultrasound whitematter echolucencies that predict disability: a review. Dev. Med. Child Neurol. 41, 136–139 Sharp, F.R. et al. (2000) Multiple molecular penumbras after focal cerebral ischemia. J. Cereb. Blood Flow Metab. 20, 1011–1032
TRENDS in Neurosciences Vol.24 No.6 June 2001
56 Amar, A.P. and Levy, M.L. (1999) Pathogenesis and pharmacological strategies for mitigating secondary damage in acute spinal cord injury. Neurosurgery 44, 1027–1040 57 Leib, S.L. et al. (2000) Matrix metalloproteinases contribute to brain damage in experimental pneumococcal meningitis. Infect. Immun. 68, 615–620 58 Plata-Salaman, C.R. et al. (1998) Gram-negative and gram-positive bacterial products induce differential cytokine profiles in the brain: analysis using an integrative molecular-behavioral in vivo model. Int. J. Mol. Med. 1, 387–397 59 Ibelgauft, H. (2000) Cytokines Online Pathfinder Encyclopaedia (http://www.copewithcytokines.de/cope.cgi?1166) 60 Terui, T. et al. (2000) Role of neutrophils in induction of acute inflammation in T-cellmediated immune dermatosis, psoriasis: a neutrophil-associated inflammation-boosting loop. Exp. Dermatol. 9, 1–10 61 Bona, E. et al. (1999) Chemokine and inflammatory cell response to hypoxia– ischemia in immature rats. Pediatr. Res. 45, 500–509 62 Xue, M. and Del Bigio, M.R. (2000) Intracortical hemorrhage injury in rats: relationship between blood fractions and brain cell death. Stroke 31, 1721–1727
Multiple brainmemory systems: the whole does not equal the sum of its parts Jeansok J. Kim and Mark G. Baxter Most contemporary theories of memory are based on the assumption that memory can be divided into multiple psychological systems that are subserved by different neural substrates and that contribute to performance in a relatively independent manner. Although the study of individual memory systems has proved to be enormously useful, recent data increasingly point towards complex interactions between memory systems during performance of any given memory task. Three basic classes of interactions between different memory systems (competition, synergism and independence) are presented that appear to be congruent with the findings of many behavioral studies. Consideration of interactions among multiple memory systems will enhance our current understanding of memory by encouraging the view that memory systems are dynamic interactive units, rather than independent modules that act in isolation.
Memory can be defined in many ways, as a function of many variables and at many levels of analysis. Modern definitions typically attribute memory to processes of information acquisition, storage and
63 Popovich, P.G. et al. (1997) Cellular inflammatory response after spinal cord injury in Sprague-Dawley and Lewis rats. J. Comp. Neurol. 377, 443–464 64 Mabuchi, T. et al. (2000) Contribution of microglia/macrophages to expansion of infarction and response of oligodendrocytes after focal cerebral ischemia in rats. Stroke 31, 1735–1743 65 Barone, F.C. and Feuerstein, G.Z. (1999) Inflammatory mediators and stroke: new opportunities for novel therapeutics. J. Cereb. Blood Flow Metab. 19, 819–834 66 Ritter, L.S. et al. (2000) Leukocyte accumulation and hemodynamic changes in the cerebral microcirculation during early reperfusion after stroke. Stroke 31, 1153–1161 67 Soriano, S.G. et al. (1999) Mice deficient in Mac-1 (CD11b/CD18) are less susceptible to cerebral ischemia/reperfusion injury. Stroke 30, 134–139 68 Taoka, Y. and Okajima, K. (2000) Role of leukocytes in spinal cord injury in rats. J. Neurotrauma 17, 219–229 69 Kostulas, N. et al. (1998) Ischemic stroke is associated with a systemic increase of blood mononuclear cells expressing interleukin-8 mRNA. Stroke 29, 462–466 70 Ferrarese, C. et al. (1999) Increased cytokine release from peripheral blood cells after acute stroke. J. Cereb. Blood Flow Metab. 19, 1004–1009
retrieval1–4. In operational terms, memory is inferred from alterations in behavior that are caused by some prior experience that cannot be attributed to factors such as modifications of the responsivity of sensoryeffector organs5. For many, the inclusion of diverse phenomena within one general term – memory – is neither intuitive nor satisfactory. The memory involved in learning to ride a bicycle, for example, seems qualitatively different from that involved in learning to solve mathematical equations or recall what one had for breakfast. The concept of separable types of memory originated with the discovery of some preserved learning and memory capabilities in the profoundly amnesic patient H.M. (Refs 1,6), and has received widespread acceptance in spite of continuing discussion of how memory systems should be categorized7–15. Current status of multiple memory systems theory
Memory systems, as currently defined, are at the same time psychological and biological entities. In psychological terms, memory systems are considered to be specialized modules16 that process particular kinds of information (e.g. facts and events)8, perform particular operations (e.g. working and reference)10 or store information for a particular period of time (e.g. short term and long term)7. On a biological level, a memory system is usually defined as a neural structure (or network of structures17–19) and its interconnections, which together operate on a particular type of information and then participate in the storage of that information, either within the structure itself or elsewhere1,2. For example, an
http://tins.trends.com 0166-2236/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0166-2236(00)01818-X
Opinion
TRENDS in Neurosciences Vol.24 No.6 June 2001
Do white cells matter in white matter damage? Olaf Dammann, Scott Durum and Alan Leviton Support is provided for the hypothesis that activated leukocytes, especially monocytes/macrophages, contribute to cerebral white matter damage in extremely low gestational age newborns. Much of the evidence is indirect and comes from analogies to brain diseases in adults, and from models of brain damage in adult and newborn animals. If the recruitment of circulating cells to the brain contributes to white matter damage in extremely low gestational age newborns, then minimizing the transendothelial migration of circulating cells by pharmacological manipulation might prevent or reduce the occurrence of neonatal white matter damage and the disabilities that follow.
Olaf Dammann Alan Leviton* Children’s Hospital and Harvard Medical School, Boston, MA, USA. e-mail: leviton@ Hub.TCH.Harvard.edu Scott Durum National Cancer Institute, Bethesda, MD, USA.
Neonatal white matter damage, which is observed in ~10% of extremely low gestational age newborns1, can result in motor, cognitive, perceptual, visual, behavior and attention limitations2–4. Usually defined on a cranial ultrasound scan as a cerebral white matter echolucency (because the reduced echo implies loss of brain substance), neonatal white matter damage is often identified as periventricular leukomalacia. This name is technically inappropriate because malacia means softening, which cannot be ascertained by ultrasound examination. Alternative names, such as preterm anoxic injury, ischemic lesions, and ischemic and reperfusion brain injury imply what some consider the most probable explanation for these images5. Excitotoxic damage to neonatal white matter6,7, and damage caused by oxidative stress are other possible explanations8. However, recent publications favor the hypothesis that an inflammatory process precedes or accompanies neonatal white matter damage and its motor consequences9–21. Support for this hypothesis is based on measurements of inflammatory cytokines. In this review, we move beyond the contribution of cytokines in neonatal white matter damage and suggest a role for activated leukocytes, especially the cell identified as a monocyte in the blood and as a macrophage in the brain. The strongest support for this hypothesis is that the monocyte/macrophage is seen in neonatal white matter damage22,23. Previously, the presence of this cell in white matter damage was viewed as support for the hypothesis that its main role is to clean up debris24. Now, however, monocytes/macrophages, and perhaps other leukocytes in the brain, appear to contribute to the damage25–27. Each of the components of the inflammatory process that precede or accompany preterm birth, and
neonatal white matter damage and its functional consequences, can facilitate the transendothelial migration of leukocytes. This leads us to postulate that the inflammatory response and other stimuli that stimulate leukocyte adherence, such as hypoxia and reactive oxygen metabolites28, enable the transendothelial migration of leukocytes into the brain, which then contribute to perpetuating white matter damage. In the absence of evidence that leukocytes other than the macrophage or monocyte are involved in neonatal white matter damage, we review the clinical and related neurobiology literatures to support our hypothesis that additional evidence for leukocyte involvement in neonatal white matter damage will be forthcoming. Specifically, we address three hypotheses: (1) some of the inflammatory cells in the area of the damage are derived from cells outside the brain; (2) these cells contribute to brain damage and (3) interfering with leukocyte migration into the brain will reduce brain damage. The fetal inflammatory response and transendothelial migration of leukocytes
Markers of inflammation in and around the fetus predict second trimester labor29 and/or rupture of membranes30. In the amniotic fluid of pregnancies, characterized by both preterm labor and evidence of infection, most of the neutrophils are of a fetal origin31. Thus, a sizable component of the inflammatory process that presumably contributes to delivery much before term is probably of fetal origin.
…we move beyond the contribution ‘… of cytokines in neonatal white matter damage and suggest a role for activated leukocytes,’ The fetal inflammatory response, which often appears to be triggered by bacteria in the amniotic fluid, is identified histologically by the presence of inflammatory cells in the walls of the vessels of the umbilical cord (i.e. funisitis), and by increased expression and/or higher concentrations of inflammatory cytokines in the amniotic fluid or blood32. These inflammatory phenomena also predict neonatal white matter damage and its consequences9–13. This has led to the hypothesis that a fetal inflammatory response contributes to brain damage in the extremely low gestational age newborn33. The maternal and fetal inflammatory responses include the involvement of cytokines34, chemokines35, adhesion molecules36, extracellular matrix proteins with adhesive properties37, and matrix metalloproteinases38. All five of these groups facilitate the transendothelial migration of cells out of the circulation and into adjacent organ parenchyma. We
http://tins.trends.com 0166-2236/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0166-2236(00)01811-7
Opinion
TRENDS in Neurosciences Vol.24 No.6 June 2001
Fetal endothelium
Fetal systemic circulation
321
BBB
Fetal brain Oligodendrocyte Astrocyte
Activate Activate Activated leukocyte Pre-OL Cytokines Activate Activated leukocyte
Microglia TRENDS in Neurosciences
Fig. 1. Hypothetical model of neonatal white matter damage. Proinflammatory cytokines in the fetal systemic circulation activate white blood cells (left), which render the fetal blood–brain barrier (BBB) ineffective and migrate through the BBB into the fetal brain. Here, they activate resident astrocytes and microglia, which, in turn, adversely affect the development of oligodendrocyte precursor cells (pre-OL) and the well-being of mature oligodendrocytes. Astrocytes and microglia are also engaged in a bilateral sustained activation (right).
reason that if all the components needed for the transendothelial migration of leukocytes are in place when neonatal brain damage begins, then transendothelial migration of leukocytes into the brain is possible. To date, only two of the five protein groups that constitute the chemically defined fetal inflammatory response, cytokines and chemokines, have been associated with neonatal white matter damage or cerebral palsy, the most prominent motor consequence of neonatal white matter damage10,11,14–16. The presence of a third group, adhesion molecules28 is suggested by the associations of funisitis with neonatal white matter damage9 and neurologic impairment11,12. These associations support the possibility that the fetal inflammatory response aids the transendothelial migration of leukocytes into the brain, where they contribute to brain damage. The process of transendothelial migration probably begins with inflammatory cytokines39 (Fig. 1). In addition to activating endothelial cells, these cytokines activate leukocytes in the circulation, diminish the effectiveness of the blood–brain barrier40, which is relatively ineffective in the preterm newborn41, and gain access to and activate microglia and astrocytes, which in turn produce chemokines and more cytokines. The chemokines function as chemo-attractants for activated inflammatory cells, including those in the circulation. Some adhesion molecules facilitate the initial tethering of leukocytes to the vessel wall, followed by their rolling along the endothelium until they adhere tightly to the endothelial surface. Other adhesion molecules then facilitate the release of the leukocytes from the vessel wall, allowing movement through the intercellular junctions into the underlying tissue. The http://tins.trends.com
matrix metalloproteinases further disrupt the blood–brain barrier and facilitate cellular migration through the extracellular matrix42. If leukocytes are commonly involved in other CNS disorders, and can be shown to damage the brain, then perhaps they also contribute to neonatal white matter damage. The role of leukocytes in various forms of brain injury is reviewed below. Table 1 shows which cells are involved in each of these disorders. Presence of leukocytes in areas of brain damage Neonatal white matter damage in human newborns
Most of the infants who are at risk of white matter damage and who die, tend to do so within days of birth, and weeks before the white matter damage would be fully developed23, thereby limiting opportunities to study the development of white matter damage in the most vulnerable humans. The characteristic development of white matter damage usually begins with hypertrophic astrocytes, and not with cells that are unquestionably derived from the blood22. Nevertheless, macrophages are a constituent of neonatal white matter damage22,23. Although some monocytes, macrophages and microglia in the brain are derived from resident cells, others are derived from the circulation27. Some of these cell members of the leukocyte family are considered here. Cerebral palsy
Newborns who later develop cerebral palsy have elevated blood concentrations of cytokines and chemokines10,15,16. These can activate leukocytes and guide their movement out of the blood vessel and into the brain. Laboratory models
Neutrophils and macrophages are early features of white matter damage in neonatal cats and dogs given endotoxin43,44. In neonatal rats, neutrophil and T-cell infiltration are also seen in the brain following cerebral ischemia and reperfusion45. In newborn mice, systemically administered interleukin-9 (IL-9) can exacerbate white matter damage attributed to excitotoxicity6,7. Activated T cells are presently the
322
Opinion
TRENDS in Neurosciences Vol.24 No.6 June 2001
Table 1. Cells implicated in different biological entitiesa Biological entity
Neutrophils
Monocytes
T lymphocytes
Infection Trauma Hemorrhage Ischemia Cerebral palsy Neonatal white matter damage Human Laboratory animal
+ + + + Chemokines
+ + + + Chemokines
+ + + + Chemokines
+
+ +
+ sign indicates the presence of the leukocyte. ‘Chemokines’ appears on the cerebral palsy line because the cell listed at the top of each column has not been associated with cerebral palsy, but the chemokines that serve as a chemo-attractant for that leukocyte have been associated with cerebral palsy.
aA
only known source of IL-9 (Ref. 46). Does this mean that activated T cells are involved in neonatal white matter damage? Following CNS trauma in adult rodents, chemokines are expressed47, adhesion molecules are upregulated on endothelium48, activated neutrophils adhere to endothelial cells49, and neutrophils, monocyte/macrophages, and T cells invade the nervous system47. Similarly, cerebral ischemia in adult rodents followed by reperfusion rapidly leads to the expression of inflammatory cytokines50, metalloproteinases51 and chemokines52. Focal CNS damage is accompanied by distant damage
Newborn cerebral white matter damage, which probably includes damage to oligodendrocytes and their precursors, has both focal and diffuse effects53. The diffuse white matter damage, which often accompanies focal white matter damage, appears to explain a large proportion of the developmental disorders seen in children who were born at an extremely low gestational age54. This concern with diffuse white matter damage prompts us to pay special attention to diffuse and distant injury. We suggest that the penumbra of ischemic damage and the distant damage that is sometimes seen following trauma are examples of a process that might be a model for the diffuse form of white matter damage seen in newborns. Following cerebral ischemia, ‘multiple molecular penumbras’ are found at some distance from the infarct55. The secondary, delayed, and either adjacent or distant injury that sometimes follows CNS injury in rodents is characterized by a series of deleterious, often inflammatory, events that promote progressive tissue damage56. Some of the ‘secondary’ brain damage attributed to meningitis appears to be caused by products of inflammation57. Bacterial components and their products stimulate the production of large amounts of cell-activating cytokines and chemokines58. Because the products of inflammation that appear to account for the secondary damage following meningitis are the same as those that characterize the fetal inflammatory response, the mechanisms of brain damage might also be similar. http://tins.trends.com
Sustained activation of leukocytes and resident brain cells
Cell activation, initiated by the binding of a ligand to its receptor, enhances the ability of the cell to respond to other stimuli59. Sustained cell activation should lead to more damage than a short-lived inflammatory cytokine cascade10. Recruiting activated leukocytes to cerebral white matter appears to result in sustained activation of microglia and astrocytes27, which, in turn, produce a prolonged supply of inflammatory cytokines locally, where they do their damage (Fig. 1). This ‘inflammation-boosting loop’ has been suggested as an explanation for sustained inflammation60. Support for the sustained activation hypothesis is derived from multiple sources. In a neonatal rat model of cerebral ischemia and hypoxia, the activation of lymphocytes, microglia, macrophages, and astroglia persists for ~5 weeks61. Ischemia in the adult rat cerebrum increases the expression of chemokines, which are viewed as stimuli for ‘prolonged leukocyte recruitment, and astrocyte migration and activation’52. Following the placement of extravascular blood in rat brain, the microglial reaction and infiltration by neutrophils and lymphocytes, and the programmed death of brain cells continue for more than three weeks62. In a model of spinal cord trauma in rats, reactive astrocytes remain prominent in the epicenter of the damage for up to 28 days63. Leukocytes contribute to brain damage
All the leukocyte forms contribute to tissue damage25–27. Macrophages are among the cells that appear to contribute to the ‘expansion of infarction’ in the penumbra following focal cerebral ischemia in rats64. In several different laboratory animal species, the brain damaging consequences of ischemia can be diminished by depleting the pool of available leukocytes or by inhibiting leukocyte activation and function65,66. Knockout mice deficient in adhesion molecules also show reduced injury following focal cerebral ischemia67. Both the adherence of activated leukocytes to endothelial cells and the motor consequences of cord compression can be attenuated by leukocytopenia (a decrease in the number of white blood cells) and by interventions that reduce the availability of activated leukocytes to endothelial cells and cord parenchyma49. The evidence that activated leukocytes are involved in this secondary injury has prompted the search for therapies that reduce the probability of leukocyte activation68. Systemic responses predict, accompany and perhaps contribute to CNS damage
We postulate that unlike cerebral ischemia and trauma, in which the inflammation begins in the brain, white matter damage in newborns begins with a systemic inflammatory response that is initiated
Opinion
TRENDS in Neurosciences Vol.24 No.6 June 2001
outside, and then gains access to, the brain33. Support for this hypothesis is also derived from the stroke literature. Following stroke, adults have elevated numbers of peripheral blood cells expressing inflammatory cytokines and other markers of cell activation, sometimes for as long as one month69,70. A component of some of these elevations might be a marker of damage. Alternatively, their duration and ability to predict deterioration suggests that a component contributes to brain damage. References 1 Baud, O. et al. (1999) Antenatal glucocorticoid treatment and cystic periventricular leukomalacia in very premature infants. New Engl. J. Med. 341, 1190–1196 2 Pinto-Martin, J.A. et al. (1995) Cranial ultrasound prediction of disabling and nondisabling cerebral palsy at age two in a low birth weight population. Pediatrics 95, 249–254 3 Whitaker, A.H. et al. (1996) Neonatal cranial ultrasound abnormalities in low birth weight infants: relation to cognitive outcomes at six years of age. Pediatrics 98, 719–729 4 Whitaker, A.H. et al. (1997) Psychiatric outcomes in low-birth-weight children at age 6 years: relation to neonatal cranial ultrasound abnormalities. Arch. Gen. Psychiatry 54, 847–856 5 Inder, T.E. and Volpe, J.J. (2000) Mechanisms of perinatal brain injury. Semin. Neonatol. 5, 3–16 6 Dommergues, M.A. et al. (2000) Proinflammatory cytokines and interleukin-9 exacerbate excitotoxic lesions of the newborn murine neopallium. Ann. Neurol. 47, 54–63 7 Tahraoui, S.L. et al. (2001) Central role of microglia in neonatal excitotoxic lesions of the murine periventricular white matter. Brain Pathol. 11, 56–71 8 Said, S.I. et al. (2000) NMDA receptor activation: critical role in oxidant tissue injury. Free Radic. Biol. Med. 28, 1300–1302 9 Leviton, A. et al. (1999) Maternal infection, fetal inflammatory response, and brain damage in very low birth weight infants. Developmental Epidemiology Network Investigators. Pediatr. Res. 46, 566–575 10 Nelson, K.B. et al. (1998) Neonatal cytokines and coagulation factors in children with cerebral palsy. Ann. Neurol. 44, 665–675 11 Yoon, B.H. et al. (2000) Fetal exposure to an intraamniotic inflammation and the development of cerebral palsy at the age of three years. Am. J. Obstet. Gynecol. 182, 675–681 12 Redline, R.W. et al. (2000) The relationship between placental and other perinatal risk factors for neurologic impairment in very low birth weight children. Pediatr. Res. 47, 721–726 13 Wu, Y.W. and Colford, J.M. Jr (2000) Chorioamnionitis as a risk factor for cerebral palsy: a meta-analysis. J. Am. Med. Assoc. 284, 1417–1424 14 Yoon, B.H. et al. (1996) Interleukin-6 concentrations in umbilical cord plasma are elevated in neonates with white matter lesions associated with periventricular leukomalacia. Am. J. Obstet. Gynecol. 174, 1433–1440 15 Yoon, B.H. et al. (1997) Amniotic fluid inflammatory cytokines (interleukin-6, interleukin-1beta, and tumor necrosis factoralpha), neonatal brain white matter lesions, and http://tins.trends.com
323
The future
To determine if activated leukocytes contribute to the occurrence of cerebral white matter damage in extremely low gestational age newborns will require measurement of phenomena that increase the probability of cell activation and movement of activated cells from the blood to the white matter parenchyma. If activated leukocytes contribute to brain damage in newborns, then perhaps strategies to minimize their activation or their migration into the brain might prevent or reduce the occurrence of neurologic disabilities.
cerebral palsy. Am. J. Obstet. Gynecol. 177, 19–26 16 Grether, J.K. et al. (1999) Interferons and cerebral palsy. J. Pediatr. 134, 324–332 17 Yoon, B.H. et al. (1996) Interleukin-6 concentrations in umbilical cord plasma are elevated in neonates with white matter lesions associated with periventricular leukomalacia. Am. J. Obstet. Gynecol. 174, 1433–1440 18 Martinez, E. et al. (1998) Elevated amniotic fluid interleukin-6 as a predictor of neonatal periventricular leukomalacia and intraventricular hemorrhage. J. Maternal–Fetal Invest. 8, 101–107 19 Duggan, P. et al. (1999) Relation of fetal inflammation to abnormal magnetic resonance images at birth in preterm infants. Pediatr. Res. 45, 908 20 Deguch, I.K. et al. (1997) Characteristic neuropathology of leukomalacia in extremely low birth weight infants. Pediatr. Neurol. 16, 296–300 21 Yoon, B.H. et al. (1997) High expression of tumor necrosis factor-alpha and interleukin-6 in periventricular leukomalacia. Am. J. Obstet. Gynecol. 177, 406–411 22 Gilles, F.H. et al. (1983) Developing Human Brain: Growth and Epidemiologic Neuropathology, John Wright–PSG Publishing Co. 23 Golden, J.A., et al. (1997) Frequency of neuropathological abnormalities in very low birth weight infants. J. Neuropathol. Exp. Neurol. 56, 472–478 24 Hickey, W.F. (1999) Leukocyte traffic in the central nervous system: the participants and their roles. Semin. Immunol. 11, 125–137 25 Nussler, A.K. et al. (1999) Leukocytes, the Janus cells in inflammatory disease. Langenbecks Arch. Surg. 384, 222–232 26 Yrjanheikki, J. et al. (1999) A tetracycline derivative, minocycline, reduces inflammation and protects against focal cerebral ischemia with a wide therapeutic window. Proc. Natl. Acad. Sci. U. S. A. 96, 13496–13500 27 Stoll, G. and Jander, S. (1999) The role of microglia and macrophages in the pathophysiology of the CNS. Prog. Neurobiol. 58, 233–247 28 Kevil, C.G. and Bullard, D.C. (1999) Roles of leukocyte/endothelial cell adhesion molecules in the pathogenesis of vasculitis. Am. J. Med. 106, 677–687 29 Mazor, M. et al. (1998) Cytokines in preterm parturition. Gynecol. Endocrinol. 12, 421–427 30 Fortunato, S.J. et al. (2000) Programmed cell death (apoptosis) as a possible pathway to metalloproteinase activation and fetal membrane degradation in premature rupture of membranes. Am. J. Obstet. Gynecol. 182, 1468–1476 31 Sampson, J.E. et al. (1997) Fetal origin of amniotic fluid polymorphonuclear leukocytes.
Am. J. Obstet. Gynecol. 176, 77–81 32 Romero, R. et al. (1998) A fetal systemic inflammatory response is followed by the spontaneous onset of preterm parturition. Am. J. Obstet. Gynecol. 179, 186–193 33 Dammann, O. and Leviton, A. (2000) Role of the fetus in perinatal infection and neonatal brain damage. Curr. Opin. Pediatr. 12, 99–104 34 Romero, R. et al. (2000) Further observations on the fetal inflammatory response syndrome: a potential homeostatic role for the soluble receptors of tumor necrosis factor alpha. Am. J. Obstet. Gynecol. 183, 1070–1077 35 Athayde, N. et al. (1999) A role for the novel cytokine RANTES in pregnancy and parturition. Am. J. Obstet. Gynecol. 181, 989–994 36 Steinborn, A. et al. (2000) Serum intercellular adhesion molecule-1 levels and histologic chorioamnionitis. Obstet. Gynecol. 95, 671–676 37 Goldenberg, R.L. et al. (2000) Vaginal fetal fibronectin measurements from 8 to 22 weeks gestation and subsequent spontaneous preterm birth. Am. J. Obstet. Gynecol. 183, 469–475 38 Maymon, E. et al. (2000) Matrilysin (matrix metalloproteinase 7) in parturition, premature rupture of membranes, and intrauterine infection. Am. J. Obstet. Gynecol. 182, 1545–1553 39 Weerasinghe, D. et al. (1998) A role for the alphavbeta3 integrin in the transmigration of monocytes. J. Cell Biol. 142, 595–607 40 de Boer, A.G. and Breimer, D.D. (1998) Cytokines and blood–brain barrier permeability. Prog. Brain Res. 115, 425–451 41 Adinolfi, M. (1985) The development of the human blood–CSF–brain barrier. Dev. Med. Child Neurol. 27, 532–537 42 Lukes, A. et al. (1999) Extracellular matrix degradation by metalloproteinases and central nervous system diseases. Mol. Neurobiol. 19, 267–284 43 Gilles, F.H. et al. (1977) Neonatal endotoxin encephalopathy. Ann. Neurol. 2, 49–56 44 Young, R.S. et al. (1983) Systemic and neuropathologic effects of E. coli endotoxin in neonatal dogs. Pediatr. Res. 17, 349–353 45 Benjelloun, N. et al. (1999) Inflammatory responses in the cerebral cortex after ischemia in the P7 neonatal Rat. Stroke 30, 1916–1924 46 Durum, S. (2000) Interleukins: overview. In Principles and Practice of the Biological Therapy Of Cancer (Rosenberg, S. ed.), Lippincott 47 McTigue, D.M. et al. (1998) Selective chemokine mRNA accumulation in the rat spinal cord after contusion injury. J. Neurosci. Res. 53, 368–376 48 Carlos, T.M. et al. (1997) Expression of endothelial adhesion molecules and recruitment of
324
49
50
51
52
53
54
55
Opinion
neutrophils after traumatic brain injury in rats. J. Leukocyte Biol. 61, 279–285 Taoka, Y. et al. (1998) Activated protein C reduces the severity of compression-induced spinal cord injury in rats by inhibiting activation of leukocytes. J. Neurosci. 18, 1393–1398 Stoll, G. et al. (1998) Inflammation and glial responses in ischemic brain lesions. Prog. Neurobiol. 56, 149–171 Gasche, Y. et al. (1999) Early appearance of activated matrix metalloproteinase-9 after focal cerebral ischemia in mice: a possible role in blood–brain barrier dysfunction. J. Cereb. Blood Flow Metab. 19, 1020–1028 Wang, X. et al. (1998) Prolonged expression of interferon-inducible protein-10 in ischemic cortex after permanent occlusion of the middle cerebral artery in rat. J. Neurochem. 71, 1194–1204 Leviton, A. and Gilles, F. (1996) Ventriculomegaly, delayed myelination, white matter hypoplasia, and ‘periventricular’ leukomalacia: how are they related? Pediatr. Neurol. 15, 127–136 Holling, E.E. and Leviton, A. (1999) Characteristics of cranial ultrasound whitematter echolucencies that predict disability: a review. Dev. Med. Child Neurol. 41, 136–139 Sharp, F.R. et al. (2000) Multiple molecular penumbras after focal cerebral ischemia. J. Cereb. Blood Flow Metab. 20, 1011–1032
TRENDS in Neurosciences Vol.24 No.6 June 2001
56 Amar, A.P. and Levy, M.L. (1999) Pathogenesis and pharmacological strategies for mitigating secondary damage in acute spinal cord injury. Neurosurgery 44, 1027–1040 57 Leib, S.L. et al. (2000) Matrix metalloproteinases contribute to brain damage in experimental pneumococcal meningitis. Infect. Immun. 68, 615–620 58 Plata-Salaman, C.R. et al. (1998) Gram-negative and gram-positive bacterial products induce differential cytokine profiles in the brain: analysis using an integrative molecular-behavioral in vivo model. Int. J. Mol. Med. 1, 387–397 59 Ibelgauft, H. (2000) Cytokines Online Pathfinder Encyclopaedia (http://www.copewithcytokines.de/cope.cgi?1166) 60 Terui, T. et al. (2000) Role of neutrophils in induction of acute inflammation in T-cellmediated immune dermatosis, psoriasis: a neutrophil-associated inflammation-boosting loop. Exp. Dermatol. 9, 1–10 61 Bona, E. et al. (1999) Chemokine and inflammatory cell response to hypoxia– ischemia in immature rats. Pediatr. Res. 45, 500–509 62 Xue, M. and Del Bigio, M.R. (2000) Intracortical hemorrhage injury in rats: relationship between blood fractions and brain cell death. Stroke 31, 1721–1727
Multiple brainmemory systems: the whole does not equal the sum of its parts Jeansok J. Kim and Mark G. Baxter Most contemporary theories of memory are based on the assumption that memory can be divided into multiple psychological systems that are subserved by different neural substrates and that contribute to performance in a relatively independent manner. Although the study of individual memory systems has proved to be enormously useful, recent data increasingly point towards complex interactions between memory systems during performance of any given memory task. Three basic classes of interactions between different memory systems (competition, synergism and independence) are presented that appear to be congruent with the findings of many behavioral studies. Consideration of interactions among multiple memory systems will enhance our current understanding of memory by encouraging the view that memory systems are dynamic interactive units, rather than independent modules that act in isolation.
Memory can be defined in many ways, as a function of many variables and at many levels of analysis. Modern definitions typically attribute memory to processes of information acquisition, storage and
63 Popovich, P.G. et al. (1997) Cellular inflammatory response after spinal cord injury in Sprague-Dawley and Lewis rats. J. Comp. Neurol. 377, 443–464 64 Mabuchi, T. et al. (2000) Contribution of microglia/macrophages to expansion of infarction and response of oligodendrocytes after focal cerebral ischemia in rats. Stroke 31, 1735–1743 65 Barone, F.C. and Feuerstein, G.Z. (1999) Inflammatory mediators and stroke: new opportunities for novel therapeutics. J. Cereb. Blood Flow Metab. 19, 819–834 66 Ritter, L.S. et al. (2000) Leukocyte accumulation and hemodynamic changes in the cerebral microcirculation during early reperfusion after stroke. Stroke 31, 1153–1161 67 Soriano, S.G. et al. (1999) Mice deficient in Mac-1 (CD11b/CD18) are less susceptible to cerebral ischemia/reperfusion injury. Stroke 30, 134–139 68 Taoka, Y. and Okajima, K. (2000) Role of leukocytes in spinal cord injury in rats. J. Neurotrauma 17, 219–229 69 Kostulas, N. et al. (1998) Ischemic stroke is associated with a systemic increase of blood mononuclear cells expressing interleukin-8 mRNA. Stroke 29, 462–466 70 Ferrarese, C. et al. (1999) Increased cytokine release from peripheral blood cells after acute stroke. J. Cereb. Blood Flow Metab. 19, 1004–1009
retrieval1–4. In operational terms, memory is inferred from alterations in behavior that are caused by some prior experience that cannot be attributed to factors such as modifications of the responsivity of sensoryeffector organs5. For many, the inclusion of diverse phenomena within one general term – memory – is neither intuitive nor satisfactory. The memory involved in learning to ride a bicycle, for example, seems qualitatively different from that involved in learning to solve mathematical equations or recall what one had for breakfast. The concept of separable types of memory originated with the discovery of some preserved learning and memory capabilities in the profoundly amnesic patient H.M. (Refs 1,6), and has received widespread acceptance in spite of continuing discussion of how memory systems should be categorized7–15. Current status of multiple memory systems theory
Memory systems, as currently defined, are at the same time psychological and biological entities. In psychological terms, memory systems are considered to be specialized modules16 that process particular kinds of information (e.g. facts and events)8, perform particular operations (e.g. working and reference)10 or store information for a particular period of time (e.g. short term and long term)7. On a biological level, a memory system is usually defined as a neural structure (or network of structures17–19) and its interconnections, which together operate on a particular type of information and then participate in the storage of that information, either within the structure itself or elsewhere1,2. For example, an
http://tins.trends.com 0166-2236/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0166-2236(00)01818-X