Nitrosative and Oxidative Injury To Premyelinating Oligodendrocytes in Periventricular Leukomalacia
Nitrosative and Oxidative Injury To Premyelinating Oligodendrocytes in Periventricular Leukomalacia
Key Words:
cursors.
INTRODUCTION
In the United States alone, approximately 55,000 live
newborns are born very prematurely (,1500 gram birth
weight) each year (1). Because of modern intensive care,
nearly 90% of these infants survive. Approximately 10%,
however, develop cerebral palsy (motor deficits) and
25%50% develop cognitive and behavioral deficits (2).
Thus, nearly 5,000 cases of cerebral palsy and 10,000 to
20,000 children with serious learning disabilities result
each year. Periventricular leukomalacia (PVL) is the most
common substrate of this enormous amount of neurologic
disability. The pathogenesis of PVL likely involves multiple factors, particularly ischemia/reperfusion in the critically ill premature infant with impaired regulation of cerebral blood flow (3). Cytokine-induced brain injury
associated with maternal or fetal infection may also play
an important contributory role (3).
Our over-riding hypothesis is that PVL is due to cerebral ischemia with the production of excessive reactive
nitrogen and/or oxygen species in the reperfused tissue,
From the Departments of Neurology (RLH, RJK, PAR, JJV, HCK)
and Neuropathology (RDF, HCK), Childrens Hospital and Harvard
Medical School, Boston, Massachusetts; New England Research Institutes (IS), Watertown, Massachusetts; Department of Physiology and
Biophysics (LIS), Case Western Reserve University School of Medicine, Cleveland, Ohio.
Correspondence to: Robin L. Haynes, PhD, Department of Neurology
(Neuroscience), Childrens Hospital, Boston, MA 02115. E-mail: Robin.
[email protected]
This work was supported by grants from the NINDS (PO1 NS38475)
and NICHD (Childrens Hospital Mental Retardation Research Center
[P30-HD18655]).
resulting in nitrosative and/or oxidative damage to immature cerebral white matter; these reactive species preferentially injure vulnerable premyelinating (O4 and O1)
oligodendrocytes, leading to their loss, subsequent decreased numbers of mature oligodendrocytes, and hypomyelination in long-term survivors. During the peak period of PVL (i.e. 2432 gestational weeks), O4 and O1
oligodendrocytes dominate human cerebral white matter,
with the first appearance of mature, myelin-producing
(myelin basic protein [MBP]1) cells around 30 to 35
gestational weeks (4). Recent studies in perinatal rats (5)
and rodent cell culture systems (6) indicate that O4 and
O1 oligodendrocytes are vulnerable to oxidative stress,
suggesting the possibility that the prevention of oxidative
injury may be an important treatment modality in PVL.
The lack of direct evidence of nitrosative and/or oxidative
stress in O4 and O1 cells in human PVL, however, is a
critical missing link of information that is urgently
needed to confirm the relevance of the experimental models to the human disease, and to guide preventive interventions in the premature infant.
The pathology of PVL consists of 2 components, a
focal, necrotic component in the periventricular region, and a diffuse component characterized by reactive gliosis in the surrounding white matter. In this study,
we focused more on the diffuse than the focal, necrotic
component of PVL because the former is thought to result in widespread impairment in myelination, the likely
basis for the global neurological deficits in long-term survivors. Moreover, in modern neonatal intensive care units
the diffuse component of PVL is much more common
than the focal component by neuroimaging (7). While
441
Abstract. Periventricular leukomalacia (PVL), the major substrate of cerebral palsy in survivors of prematurity, is defined
as focal periventricular necrosis and diffuse gliosis in immature cerebral white matter. We propose that nitrosative and/or
oxidative stress to premyelinating oligodendrocytes complicating cerebral ischemia in the sick premature infant is a key
mechanism of injury interfering with maturation of these cells to myelin-producing oligodendrocytes and subsequent myelination. Using immunocytochemical markers in autopsy brain tissue from 17 PVL cases and 28 non-PVL controls, we found
in the PVL cases: 1) selective regionalization of white matter injury, including preferential involvement of the deep compared
to intragyral white matter; 2) prominent activation of microglia diffusely throughout the white matter; 3) protein nitration and
lipid peroxidation in premyelinating oligodendrocytes in the diffuse component; 4) preferential death of premyelinating oligodendrocytes diffusely; and 5) virtual sparing of the overlying cerebral cortex, as demonstrated by markers of activated
astrocytes and microglia. These data establish that PVL is primarily a white matter disease that involves injury to premyelinating oligodendrocytes, potentially through activation of microglia and release of reactive oxygen and nitrogen species.
Agents that prevent nitrosative and oxidative stress may play a key role in ameliorating PVL in premature infants in the
intensive care nursery.
442
HAYNES ET AL
TABLE 1A
Summary of Cerebral White Matter Pathology in PVL Cases
Case
GA
PNA
PCA
Focal
acute
necrosis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
35*
40
30*
27*
34*
28*
36*
38
35*
36*
40
29*
36*
31*
0
4
0
4
0
7
0
1
5
6
4
20
15
64
35
44
30
31
34
35
36
39
40
42
44
49
51
95
1
1
1
1
2
1
2
1
2
2
1
2
2
2
Focal
organizing
necrosis
Focal
scar
w/mineral
Diffuse
gliosis
WM
and/or
Vents
Hypomyelin
1
1
2
1
1
1
2
2
1
1
2
2
1
1
1
1
2
2
1
1
1
2
2
1
2
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
1
1
2
2
2
2
1
1
1
2
2
2
2
2
2
2
2
2
1
2
1
1
2
Single-Labeling Immunocytochemistry in
Formalin-Fixed, Paraffin-Embedded Tissue
Standard methods in deparaffinized tissue sections (4 mm)
were applied (10). Antibodies specific for the following markers
were used: 4-hydroxy-2-nonenal-protein adducts (HNE) (1:100;
laboratory of Dr. Luke I. Szweda, or purchased from Calbiochem, San Diego, CA); malondialdehydeprotein adducts
(MDA) (1:100; Abcam, Cambridge, United Kingdom); nitrotyrosine protein adducts (NT) (1:100; Upstate, Lake Placid,
NY); heme-oxygenase-1 (HO-1) and 22 (HO-2) (both 1:500;
Stressgen, Victoria, Canada), glial fibrillary acidic protein
(GFAP) (1:9,000, Sternberger Monoclonals, Inc., Lutherville,
MD), and CD68 (1:50, Cell Marque, Austin, TX). Stains were
optimized using known Alzheimer and ALS tissue as positive
controls. CD68-positive control was tonsil. Negative controls
omitted the primary antibodies.
Legend: 1, present; 2, absent; GA 5 gestational age in weeks; PNA 5 postnatal age in weeks; PCA 5 postconceptional age
in weeks; mineral 5 mineralization; WM 5 white matter, Vents 5 ventricles. * Prematurity is defined as a gestational age less
than 37 weeks.
443
TABLE 1B
Continued Summary of the Neuropathology of PVL cases
Case
CB
BG
Thal
Hipp
A, B
2
3
4
5
6
A
2
2
2
A
A
2
F
2
2
C
E
NA
G
E, H
A, B
E
2
G
E, H
A, B
2
2
2
2
D
2
2
2
2
1
2
2
2
1
2
2
2
1
2
2
2
1
2
2
9
10
NA
A
J
L
A
A
A
A
A
2
K
A
1
2
2
2
2
2
11
12
A
2
A
2
NA
A
A, M
2
A
2
2
2
2
2
2
2
2
2
13
14
Diagnosis
Potter syndrome; abnormal temporal lobe gyration;
malrotation of hippocampi; duplication of dentate
gyrus
Congenital diaphragmatic hernia; ECMO treatment
Placental abruption
Breech birth; sepsis; necrotizing enterocolitis
Osteogenesis imperfecta
Necrotizing enterocolitis; bowel perforation; coagulopathy; sepsis
CMV infection in utero; ABO incompatibility;
dwarfism/skeletal dysplasia; fetal heart failure
Urea cycle defect (ornithine transcarbamylase deficiency); hyperammonemia
Breech birth; oligohydramnios; renal sclerosis
Intrauterine growth retardation; breech birth; oligohydramnios; dysplastic kidneys; malformed basal
ganglia and hippocampus
Congenital cardiac malformation; ECMO treatment
Pulmonary hypertension; pneumonia; ECMO treatment
Breech birth; oligohydramnios; pneumothorax; respiratory failure
Congenital cardiac malformation; ECMO treatment
Legend: 1, pathology present; 2, pathology absent; A 5 reactive gliosis; B 5 neuronal loss; C 5 ischemic necrosis; D 5
hemorrhagic infarct; E 5 apoptosis; F 5 focal microhemorrhage; G 5 mineralization; H 5 perivascular hemorrhage; I 5 microglial
nodules; J 5 vermis cortical scar; K 5 organizing infarct; L 5 Purkinje cell loss; M 5 microinfarct; N 5 necrotic focus in
hypothalamus; EMCO 5 extracorporeal membrane oxygenation; NA 5 not available; Bst, brainstem; CB, cerebellum; BG, basal
ganglia; Thal, thalamus; Hipp, hippocampus; CCH, cerebral cortex hemorrhage; WMH, white matter hemorrhage; GMH, germinal
matrix hemorrhage.
of case diagnosis and age. The age-adjusted difference was assessed with age as the covariant.
RESULTS
In the PVL and control tissue sections, we assessed
activated microglia/macrophages with an antibody to
CD68, and reactive astrocytes with an antibody to GFAP.
We used an antibody to nitrotyrosine (NT) to indicate
protein nitration, and antibodies to 4-hydroxynoneal
(HNE)-protein adducts and malondialdehyde (MDA)protein adducts to indicate lipid peroxidation, in singleand double-labeling studies in the diffuse component of
PVL. In addition, the enzymes heme oxygenase (HO)-1
and 22 were examined to indicate cellular response to
oxidative stress. We performed single-labeling immunocytochemistry using markers of nitrosative and oxidative
stress and cellular responses in 11 of the 14 PVL cases
(Cases 1, 514) and 23 controls (without PVL), adjusted
for age, in formalin-fixed, paraffin-embedded tissue sections. We performed double labeling studies in brain tissues from 4 PVL and 4 control cases, fixed in 4% paraformaldehyde. We used a semiquantitative grading
system (03) of the density of immunopositive cells/high
J Neuropathol Exp Neurol, Vol 62, May, 2003
Bst
444
HAYNES ET AL
TABLE 2
Comparison of Age-Adjusted Means (SEM) for Markers
of Oxidative Stress and Cellular Damage between PVL
and Control Cases
Age*
PVL
Control
p value
GFAP
PCA
GA
PNA
2.64 (0.17)
2.73 (0.22)
2.68 (0.17)
1.73 (0.14)
1.66 (0.18)
1.70 (0.14)
,0.001
0.002
0.002
CD68
PCA
GA
PNA
2.24 (0.26)
2.50 (0.30)
2.27 (0.26)
1.18 (0.21)
1.53 (0.23)
1.15 (0.21)
0.005
0.017
0.003
HNE
PCA
GA
PNA
2.57 (0.13)
2.66 (0.14)
2.58 (0.13)
0.98 (0.11)
0.92 (0.12)
0.98 (0.11)
,0.001
,0.001
,0.001
MDA
PCA
GA
PNA
2.78 (0.27)
3.04 (0.29)
2.80 (0.27)
1.43 (0.23)
1.23 (0.24)
1.42 (0.23)
0.002
0.001
0.001
NT
PCA
GA
PNA
1.98 (0.26)
2.00 (0.29)
1.99 (0.26)
0.38 (0.22)
0.37 (0.24)
0.38 (0.22)
,0.001
,0.001
,0.001
HO-1
PCA
GA
PNA
1.99 (0.31)
1.80 (0.34)
2.04 (0.31)
1.51 (0.26)
1.64 (0.28)
1.47 (0.26)
0.252
0.741
0.176
HO-2
PCA
GA
PNA
2.43 (0.28)
2.48 (0.31)
2.44 (0.27)
1.05 (0.24)
1.01 (0.26)
1.04 (0.23)
0.001
0.003
,0.001
term deliveries were complicated by congenital diaphragmatic hernia (Case 2), urea cycle defect (Case 8), and
congenital heart malformation (Case 11) (Table 1). In 9
cases, focal cystic cavities were grossly visible in the
periventricular white matter (Fig. 1A). Microscopically,
7 cases had acute coagulative necrosis, characterized by
fragmentation of all tissue elements and pyknotic or absent nuclei, indicative of an acute insult of 24 to 48 hours
duration; 9 cases had necrotic foci with subacute changes
Marker
(Fig. 1E). The difference in CD68 immunostaining between PVL and control white matter, and between PVL
white matter and cortex, was statistically significant (p ,
0.005). In the deep white matter surrounding the foci of
necrosis there was marked (Grades $2) gliosis with reactive (GFAP1) astrocytes, again with striking sparing of
the overlying cerebral cortex (Fig. 1F; Table 2). There
was marked regional heterogeneity in the density of reactive astrocytes, and of glial cells with positive immunostaining for the selected markers of nitrosative and oxidative stress in the different components of the white
matter (Fig. 2AC).
Immunocytochemistry for Protein Nitration in PVL
Morphologically identified macrophages within the
subacute necrotic foci immunostained positively with NT
(data not shown). Nitrotyrosine immunoreactivity was
also localized to cells morphologically consistent with
glial cells in the diffuse component (Fig. 3A) and was
increased in PVL when compared to controls (Fig. 3B;
J Neuropathol Exp Neurol, Vol 62, May, 2003
445
446
HAYNES ET AL
DISCUSSION
Antibodies to HNE, MDA, HO-1, and HO-2 were localized to cells morphologically consistent with macrophages within the necrotic foci of PVL (data not shown).
In the diffuse white matter lesion, reactive astrocytes
were significantly increased in number compared to control cases, adjusted for PCA, GA, or PNA (Table 2). Glial
cells, some of which were morphologically consistent
with reactive astrocytes (Fig. 4A), within the diffuse lesion were immunopositive for all oxidative stress-related
markers analyzed, including HNE (Fig. 4B), MDA (Fig.
2C), HO-1, and HO-2 (Table 2). There was a statistically
significant difference in the age-adjusted means of the
grade of immunopositive white matter cells between the
PVL and control groups for HNE, MDA, and HO-2, regardless of the particular age measure, i.e. whether the
cases were adjusted for PCA, GA, or PNA (Table 2).
With double labeling, the oxidative stress markers HNE
and MDA were present in O4-positive oligodendrocytes
(Fig. 4C; HNE shown) and O1-positive oligodendrocytes
(data not shown), as well as in astrocytes (Fig. 4D). In
control cases, a slight but detectable amount of HNE and
MDA colocalized primarily with GFAP-immunopositive
reactive astrocytes, and occasionally with O4- and/or O1positive cells (data not shown).
TUNEL Staining and Loss of Oligodendrocytes in PVL
Dying oligodendrocytes were identified in acute PVL
by colocalization of O4 immunoreactive cells with TUNEL-positive nuclei (Fig. 5A, Case 4). In contrast, the
GFAP-immunopositive reactive astrocytes, which separately labeled with markers of nitrosative and oxidative
damage (Figs. 3D, 4D), were not dying, given the lack
of the colocalization of GFAP and TUNEL-positivity
(Fig. 5B) and their increase in number, i.e. proliferation.
J Neuropathol Exp Neurol, Vol 62, May, 2003
Table 2). The degree of NT immunostaining was statistically significant in PVL cases compared to control cases, adjusted for postconceptional age (PCA), gestational
age (GA), or postnatal age (PNA) (Table 2), with increased positive cell density in areas immediately surrounding the focal PVL lesion. To define precisely the
specific cell types susceptible to nitrosative stress in PVL,
double labeling with markers to O4, O1, or GFAP and
to NT was done. In the PVL cases, NT was present in
O4-positive oligodendrocytes (Fig. 3C) and O1-positive
oligodendrocytes (data not shown) in the diffuse component. Nitrotyrosine also colocalized with GFAP (Fig.
3D), indicating that nitrosative damage occurs in reactive
astrocytes, as well as in oligodendrocytes. In control cases, a slight but detectable amount of NT colocalized primarily with GFAP-immunopositive reactive astrocytes
and occasionally with O4 and/or O1 cells (data not
shown).
distribution of damage. This regionalization of white matter damage reinforces the idea that PVL is secondary to
ischemia in distal fields of arterial irrigation, i.e. vulnerable watershed territories in the deeper (and not superficial) white matter.
The use of the CD68 marker in this study reveals a
widespread activation of microglia/macrophages in the
diffuse component of PVL, indicating a need for new
emphasis on the role of microglia/macrophages in the
pathogenesis of PVL. CNS microglia/macrophages are
known to be activated by ischemia/hypoxia (14, 15), including up to 1 month after the initial insult (16). Of
particular relevance in PVL is the ability of activated microglia/macrophages to release reactive nitrogen species
(RNS) (17) and reactive oxygen species (ROS) (18) into
the surrounding environment, thus putting oligodendrocytes at risk for nitrosative and oxidative damage (19).
Nitric oxide (NO), in particular, is fundamental to nitrosative injury. While NO has normal functions in the
brain, including synaptic transmission and the regulation
447
448
HAYNES ET AL
including PVL. Astrocytes may protect surrounding neuronal and oligodendroglial cells under conditions of such
stress due to their capacity to release antioxidants into
the extracellular fluid (36, 37), and to clear extracellular
glutamate (38), which in excessive levels results in excitotoxicity to premyelinating oligodendrocytes (39, 40).
We speculate that reactive astrocytes in diffuse PVL proliferate to act, at least in part, as scavengers of excess
oxidants in the environment, thereby protecting O4 and
O1 cells from reactive nitrogen and oxygen species. Alternatively, reactive astrocytes may not be protective, but
instead contribute to tissue damage in PVL. In this regard, astrocytes upregulate inducible NOS in experimental ischemic injury (41), suggesting that, in addition to
activated microglia, they are another source of the free
radical NO.
In addition to markers of cellular damage, we showed
a statistically significant increase in the density of HO-2immunopositive glial cells in PVL. The HO family functions as a defense mechanism against free radical injury
through degradation of heme, a pro-oxidant, and subsequent production of carbon monoxide, free iron, and biliverdin, which is subsequently converted to bilirubin, a
potent antioxidant (42). In contrast to HO-2, HO-1 was
not significantly increased in the PVL cases compared to
the control. Heme oxygenase-1, shown to be induced by
ischemic injury, is also induced by cytokines and prostaglandins, and in response to hemorrhage, sepsis, and
pregnancy (43). The lack of a significant difference in
HO-1 between PVL and controls suggests that, unlike in
other reported settings, HO-1 may not be a specific marker of oxidative stress in developing human brain.
In conclusion, we show a region-specific vulnerability
to nitrosative and oxidative damage within the deep periventricular white matter, and a relative sparing of the
overlying cerebral cortex. We also show widespread activation of microglia, leading us to speculate that these
cells may play a direct role in the white matter damage
via release of NO. It is unknown whether nitrosative or
oxidative damage in PVL is the primary insult to premyelinating oligodendrocytes, or is secondary to other
insults, e.g. excitotoxicity from excessive glutamate release (44), or cytokine toxicity (45) complicating diffuse
microglial/macrophagocytic activation. Irrespective of the
precise sequence of events, the burden of nitrosative and
oxidative damage to the white matter appears enormous
in PVL. Our findings indicate that continued research into
the role of reactive nitrogen and oxygen species in the
pathogenesis of PVL is urgently needed, with the ultimate goal of the development of drug strategies to prevent this devastating disorder.
ACKNOWLEDGMENT
We thank Drs. Geraldine Pinkus, Neil Kowall, and Pamela L. Follett
for assistance in immunocytochemistry, Ms. Diane Brown and Ms. Lena
REFERENCES
1. Hoyert DL, Freedman MA, Strobino DM, Guyer B. Annual summary of vital statistics: 2000. Pediatrics 2001;108:124155
2. Volpe J. Neurology of the newborn. Philadelphia: W. B. Saunders
Company, 2001
3. Volpe JJ. Neurobiology of periventricular leukomalacia in the premature infant. Pediatr Res 2001;50:55362
4. Back SA, Luo NL, Broenstein NS, Levine JM, Volpe JJ, Kinney
HC. Late oligodendrocyte progenitors coincide with the development window of vulnerability for human perinatal white matter
injury. J Neurosci 2001;21:130212
5. Levison SW, Rothstein RP, Romanko MJ, Snyder MJ, Meyers RL,
Vannucci SJ. Hypoxia/ischemia depletes the rat perinatal subventricular zone of oligodendrocyte progenitors and neural stem cells.
Dev Neurosci 2001;23:23447
6. Back SA, Gan X, Li Y, Rosenberg PA, Volpe JJ. Maturation-dependent vulnerability of oligodendrocytes to oxidative stress-induced death caused by glutathione depletion. J Neurosci 1998;18:
624153
7. Maalouf EF, Duggan PJ, Counsell SJ, et al. Comparison of findings
on cranial ultrasound and magnetic resonance imaging in preterm
infants. Pediatrics 2001;107:71927
8. Kinney H, Armstrong D. Perinatal neuropathology. In: Graham D,
Lantos P, eds. Greenfields neuropathology, Vol. 1. London: Arnold,
1997:55255
9. Xie Z, Wei M, Morgan TE, et al. Peroxynitrite mediates neurotoxicity of amyloid beta-peptide1-42- and lipopolysaccharide-activated
microglia. J Neurosci 2002;22:348492
10. Lee BH, Hecht JL, Pinkus JL, Pinkus GS. WT1, estrogen receptor,
and progesterone receptor as markers for breast or ovarian primary
sites in metastatic adenocarcinoma to body fluids. Am J Clin Pathol
2002;117:74550
11. Inder TE, Huppi PS, Warfield S, et al. Periventriculuar white matter
injury in the premature infant is followed by reduced cerebral cortical gray matter volume at term. Ann Neurol 1999;46:75560
12. Deguchi K, Oguchi K, Matsuura N, Armstrong DD, Takashima S.
Periventricular leukomalacia: Relation to gestational age and axonal
injury. Pediatr Neurol 1999;20:37074
13. Ghosh A, Antonini A, McConnell SK, Shatz CJ. Requirement for
subplate neurons in the formation of thalamocortical connections.
Nature 1990;347:17981
14. Cai Z, Pang Y, Xiao F, Rhodes PG. Chronic ischemia preferentially
causes white matter injury in the neonatal rat brain. Brain Res 2001;
898:12635
15. Ivacko JA, Sun R, Silverstein FS. Hypoxic-ischemic brain injury
induces an acute microglial reaction in perinatal rats. Pediatr Res
1996;39:3947
16. Bona E, Andersson AL, Blomgren K, et al. Chemokine and inflammatory cell response to hypoxia-ischemia in immature rats. Pediatr
Res 1999;45:500509
17. Chao CC, Hu S, Molitor TW, Shaskan EG, Peterson PK. Activated
microglia mediate neuronal cell injury via a nitric oxide mechanism. J Immunol 1992;149:273641
18. Colton CA, Gilbert DL. Microglia, an in vivo source of reactive
oxygen species in the brain. Adv Neurol 1993;59:32126
19. Thorburne SK, Juurlink BH. Low glutathione and high iron govern
the susceptibility of oligodendroglial precursors to oxidative stress.
J Neurochem 1996;67:101422
449
450
HAYNES ET AL
40. Follett PL, Rosenberg PA, Volpe JJ, Jensen FE. NBQX attenuates
excitotoxic injury in developing white matter. J Neurosci 2000;20:
923541
41. Endoh M, Maiese K, Wagner J. Expression of the inducible form
of nitric oxide synthase by reacitve astrocytes after transient global
ischemia. Brain Research 1994;651:92100
42. Otterbein LE, Choi AM. Heme oxygenase: Colors of defense
against cellular stress. Am J Physiol Lung Cell Mol Physiol 2000;
279:L102937
43. Bidmon HJ, Emde B, Oermann E, Kubitz R, Witte OW, Zilles K.
Heme oxygenase-1 (HSP-32) and heme oxygenase-2 induction in
neurons and glial cells of cerebral regions and its relation to iron