Altered inflammatory responses in preterm children with cerebral palsy.
ABSTRACT Perinatal inflammatory responses contribute to periventricular leukomalacia (PVL) and cerebral palsy (CP) in preterm infants. Here, we examined whether preterm children with CP had altered inflammatory responses when school-aged.
Thirty-two preterm children with PVL-induced CP (mean [+/-standard deviation] age, 7.2 +/- 3.6 years) and 32 control preterm children with normal neurodevelopment (6.2 +/- 2.2 years) and matched for gestational age were recruited. We measured tumor necrosis factor (TNF)-alpha levels in the plasma and the supernatants of peripheral blood mononuclear cells (PBMCs) before and after lipopolysaccharide (LPS) stimulation, and proinflammatory gene expression in the PBMCs.
TNF-alpha expression was significantly higher in the plasma (p < 0.001) and supernatants of LPS-stimulated PBMCs (p = 0.003) in the CP group than in the control group. After LPS stimulation, the intracellular TNF-alpha level in the PBMCs was significantly lower in the control group (p = 0.016) and significantly higher in the CP group (p = 0.01). The CP group also had, in their nonstimulated PBMCs, significantly higher mRNA levels of inflammatory molecules: toll-like receptor 4 (TLR-4) (p = 0.0023), TNF-alpha (p = 0.0016), transforming growth factor-beta-activated kinase 1 (p = 0.038), IkappaB kinase-gamma (p = 0.029), and c-Jun N-terminal kinase (p = 0.045). The TLR-4 mRNA levels in the PBMCs were highly correlated with TNF-alpha levels in LPS-stimulated PBMCs (Spearman rank correlation = 0.38, p = 0.03).
The finding that preterm children with PVL-induced CP have altered inflammatory responses indicates the possibility of programming effect of PVL or inflammation-related events during early life.
- SourceAvailable from: Bobbi Fleiss[show abstract] [hide abstract]
ABSTRACT: Fetal/neonatal brain injury is an important cause of neurological disability. Hypoxia-ischemia and excitotoxicity are considered important insults, and, in spite of their acute nature, brain injury develops over a protracted time period during the primary, secondary, and tertiary phases. The concept that most of the injury develops with a delay after the insult makes it possible to provide effective neuroprotective treatment after the insult. Indeed, hypothermia applied within 6 hours after birth in neonatal encephalopathy reduces neurological disability in clinical trials. In order to develop the next generation of treatment, we need to know more about the pathophysiological mechanism during the secondary and tertiary phases of injury. We review some of the critical molecular events related to mitochondrial dysfunction and apoptosis during the secondary phase and report some recent evidence that intervention may be feasible also days-weeks after the insult.Neurology research international. 01/2012; 2012:506320.
- [show abstract] [hide abstract]
ABSTRACT: Perinatal brain injury frequently complicates preterm birth and leads to significant long-term morbidity. Cytokines and inflammatory cells are mediators in the common pathways associated with perinatal brain injury induced by a variety of insults, such as hypoxic-ischemic injury, reperfusion injury, toxin-mediated injury, and infection. This paper examines our current knowledge regarding cytokine-related perinatal brain injury and specifically discusses strategies for attenuating cytokine-mediated brain damage.Neurology research international. 01/2012; 2012:561494.
Altered Inflammatory Responses in
Preterm Children with Cerebral Palsy
Chang-Yi Lin, BS,1Ying-Chao Chang, MD, PhD,2Shan-Tair Wang, PhD,3
Ting-Yang Lee, MS,4Chiou-Feng Lin, PhD,5
and Chao-Ching Huang, MD5,6
Objective: Perinatal inflammatory responses contribute to periventricular leukomalacia (PVL) and cerebral palsy (CP)
in preterm infants. Here, we examined whether preterm children with CP had altered inflammatory responses when
Methods: Thirty-two preterm children with PVL-induced CP (mean [?standard deviation] age, 7.2 ? 3.6 years) and
32 control preterm children with normal neurodevelopment (6.2 ? 2.2 years) and matched for gestational age
were recruited. We measured tumor necrosis factor (TNF)-? levels in the plasma and the supernatants of peripheral
blood mononuclear cells (PBMCs) before and after lipopolysaccharide (LPS) stimulation, and proinflammatory gene
expression in the PBMCs.
Results: TNF-? expression was significantly higher in the plasma (p ? 0.001) and supernatants of LPS-stimulated
PBMCs (p ? 0.003) in the CP group than in the control group. After LPS stimulation, the intracellular TNF-? level
in the PBMCs was significantly lower in the control group (p ? 0.016) and significantly higher in the CP group
(p ? 0.01). The CP group also had, in their nonstimulated PBMCs, significantly higher mRNA levels of inflamma-
tory molecules: toll-like receptor 4 (TLR-4) (p ? 0.0023), TNF-? (p ? 0.0016), transforming growth factor-?–
activated kinase 1 (p ? 0.038), I?B kinase-? (p ? 0.029), and c-Jun N-terminal kinase (p ? 0.045). The TLR-4
mRNA levels in the PBMCs were highly correlated with TNF-? levels in LPS-stimulated PBMCs (Spearman rank
correlation ? 0.38, p ? 0.03).
Interpretation: The finding that preterm children with PVL-induced CP have altered inflammatory responses indi-
cates the possibility of programming effect of PVL or inflammation-related events during early life.
ANN NEUROL 2010;68:204–212
ronmental stressors may have a permanent rather than a
transient effect on the organism. Newborns are sensitive
to stress, steroid use, and inflammation, all of which per-
manently affect development.1–3Injuries during critical
periods of development may have long-term effects on
growth, metabolism, and inflammatory responses.4
Follow-up study5reports an increasing rate of cere-
bral palsy (CP) in preterm infants who survive. Currently,
spastic CP develops in 5 to 10% of preterm survivors,6,7
eonatal events have a powerful effect on subsequent
susceptibility to certain chronic diseases. Early envi-
and neonatal periventricular leukomalacia (PVL) is the
major form of brain injury and also the leading cause of
spastic CP in preterm infants.6The neuropathologic hall-
mark of PVL is a multitude of activated microglia and
macrophages that produce proinflammatory cytokines,
such as tumor necrosis factor (TNF)-?, which directly
cause oligodendroglial cell death or exacerbate white mat-
ter lesions.6,8Perinatal inflammatory responses contribute
to white matter injury and CP in preterm infants.9,10Be-
cause inflammation is an etiologic factor in the white
matter damage that gives rise to CP,3,6white matter dam-
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ana.22049
Received Jan 8, 2010, and in revised form Mar 5. Accepted for publication Apr 2, 2010.
Address correspondence to Dr Huang, MD, Department of Pediatrics, National Cheng Kung University Hospital, 138 Sheng-Li Road, Tainan City
70428, Taiwan. Phone: 886-6-235-3535 ext 5289; Fax: 886-6-275-3083, E-mail: email@example.com
From the1Institute of Basic Medical Sciences, National Cheng Kung University College of Medicine, Tainan;2Department of Pediatrics, Chang
Gung Memorial Hospital-Kaohsiung Medical Center, Chang Gung University College of Medicine, Kaohsiung;3Institute of Public Health, National
Cheng Kung University College of Medicine, Tainan;4Department of Biochemistry and Molecular Biology, National Cheng Kung University
College of Medicine, Tainan;5Institute of Clinical Medicine, National Cheng Kung University College of Medicine, Tainan; and6Department of
Pediatrics, National Cheng Kung University College of Medicine, Tainan, Taiwan.
Additional Supporting Information can be found in the online version of this article.
204© 2010 American Neurological Association
age may arise from the recruitment of peripheral leuko-
cytes passing through the immature blood-brain barrier,
which intensifies the interactions between innate and
adaptive immune responses in injured cerebral white mat-
ter.10,11Macrophages and microglia are found where
white matter injury is ongoing, which indicates that acti-
vated leukocytes are involved in white matter damage.10
Exposing prenatal animals to lipopolysaccharide
(LPS) worsens hypoxic-ischemic white matter injury
weeks or months after birth.12Brief systemic exposure to
LPS may cause long-lasting cerebral and peripheral vul-
nerability, including immune responses, into adult-
hood.3,13If activated white blood cells in preterm infants
contribute to neurological dysfunction, developmental
plasticity altered by neonatal injury, such as inflammation
or PVL or even subsequent phenomena related to inflam-
mation or PVL or both, may have a long-term effect on
the inflammatory responses of circulating leukocytes.14
Previous studies6,8,9,11have shown that perinatal in-
flammatory responses contribute to white matter injury
and CP in preterm infants. In this study, we compared
LPS sensitivity in the peripheral blood mononuclear cells
(PBMCs) of 2 groups of school-age preterm children—
gestational age-matched preterm, neurodevelopmentally
normal controls and preterm children with neonatal PVL-
induced CP—to examine whether the preterm children
with CP had altered inflammatory responses when school-
Subjects and Methods
The study was approved by the institutional review board of
National Cheng Kung University Hospital, and written in-
formed consent was obtained from the parents of the participat-
ing children. The children were recruited from the follow-up
clinic for children born prematurely in the Tainan Area.15This
follow-up clinic prospectively enrolled and followed up preterm
babies discharged from neonatal intensive care units of the 5
major teaching hospitals in the Tainan County in southern Tai-
wan.2At each visit, a neurodevelopmental examination, visual
and hearing testing, and neuromotor and cognitive evaluations
were done. Cognitive function was assessed using the Bayley
Scales of Infant Development II for children younger than 42
months, and 2 Wechsler intelligence tests (the Wechsler Pre-
school and Primary Scale of Intelligence, 3rd edition and the
Wechsler Intelligence Scale for Children, 3rd edition), which
give verbal intelligence quotient (IQ), performance IQ, and full-
scale IQ scores. We classified children with CP on the basis of
abilities and limitations in gross motor function using the Gross
Motor Function Classification Scale.16
Between January 2006 and June 2008, we studied 34 con-
secutive preterm children with spastic CP and a gestational age
?34 weeks. Children with spastic CP had abnormally increased
muscle tonicity and deep tendon reflexes in the limbs, and they
had a Gross Motor Function score of at least 2. Magnetic res-
onance imaging (MRI) examinations showed PVL lesions in 32;
these children were designated the CP group. Another 32 ges-
tational age-matched preterm children with normal muscle to-
nicity, normal deep tendon reflexes, normal Gross Motor Func-
tion Classification Scale, and normal MRI findings at follow-up
were designated the control group.
Magnetic Resonance Imaging
MRI examinations (Signa Infinity 1.5T; GE Healthcare,
Waukesha, WI) were done at a mean age (?standard deviation
[SD]) of 18.2 ? 5.4 months for the CP group and 60.5 ? 4.6
months for the control group. All children had routine clinical
pulse sequences, including sagittal and axial T1-weighted, fat-
saturated axial T2-weighted, and fast fluid-attenuated inversion
recovery sequences. The diagnosis of PVL was made based on
the following findings: the extent of white-matter signal abnor-
mality and ventricular dilatation, the loss in volume of periven-
tricular white matter, and the thinning of the corpus callosum.17
Isolating Peripheral Blood Mononuclear Cells
Plasma and PBMCs were collected from afebrile preterm chil-
dren who showed no signs of fever or infection during the past
5 months. Venous blood samples were collected in pyrogen-free
tubes (Vacutainer K3 EDTA, 10ml; Becton Dickinson, Bedford,
MA). The blood samples were numbered, and the laboratory
examiners were blinded to the study groups. After it had been
centrifuged, the plasma was dispensed into aliquots for analysis,
and the cell components were immediately diluted to a total
volume of 10ml using RPMI-1640, layered on 10ml Ficoll-
Hypaque density gradient (Sigma-Aldrich, St. Louis, MO), and
centrifuged at 2,000 rpm for 30 minutes. After they had been
centrifuged, the PBMCs were recovered from the interphase,
washed twice with RPMI-1640, resuspended in RPMI-1640
containing 10% fetal bovine serum medium, and seeded on
3.5cm dishes at a density of 2 ? 106cells/dish (1 ? 106cells/
ml). The PBMCs were stimulated with or without LPS (10ng/
ml) (Sigma-Aldrich) for 3 hours, and the supernatants were col-
lected.18TNF-? and interleukin-6 (IL-6) concentrations in the
plasma and the supernatants were measured using enzyme-
linked immunosorbent assay kits (DuoSet DY210 and DY225;
R&D Systems, Minneapolis, MN).
Samples of PBMCs (2 ? 106cells), with or without LPS (10ng/
ml) stimulation for 3 hours, were washed with phosphate-
buffered saline (PBS) (without Ca2?and Mg2?) and fixed with
1% formaldehyde in PBS for 15 minutes at room tempera-
ture.19After they had been washed with PBS, the fixed cells
were resuspended with permeabilization buffer (eBioscience, San
Diego, CA) and incubated at 4°C for 15 minutes. An equal
volume of permeabilization buffer (CAS-Block; Zymed Labora-
tories, San Francisco, CA) was added for 30 minutes to reduce
nonspecific background staining. The PBMCs were then resus-
pended in PBS that contained unlabeled anti-cytokine 1st Ab
Lin et al: Proinflammation in CP
(1?g/106cells) for 60 minutes at 4°C. Fluorescence-labeled 2nd
Ab (BD Biosciences, San Jose, CA) was diluted 200? and incu-
bated at 4°C for 60 minutes. List mode data were acquired using
flow cytometry (FACSCalibur and WinMDI 2.8; BD Bio-
Microarray Experiments and Real-Time
Polymerase Chain Reaction
The quantity and purity of extracted RNA (RNeasy Mini Kit;
Qiagen, Valencia, CA) were estimated by measuring A260and
A280. For microarray experiments, total RNA (0.5?g) was reverse-
transcribed (RT2First Strand Kit; SuperArray Bioscience, Freder-
ick, MD), and the expression of 84 genes involved in the TNF
ligand and TNF receptor signaling pathways was profiled (RT2
Profiler PCR Array; SuperArray Bioscience). cDNA (5ng) was
added to each well of a polymerase chain reaction (PCR) array for
quantitative PCR. For real-time quantitative PCR, cDNA was
synthesized with Moloney murine leukemia virus (MMLV) re-
verse transcriptase (Promega, Madison, WI). Real-time PCR
(LightCycler 2.0 Real-Time PCR System and software; Roche Di-
agnostics, Mannheim, Germany) was done using a 10?l reaction
mixture that contained 8?l of SYBR Green (Roche) and 0.5?M
of primers (toll-like receptor-4 [TLR-4]: CAGCCTCCTCA-
GAAACAG and CAGTGAAGAAGGGTTCCAA; transforming
growth factor-?–activated kinase 1 (TAK1): AGCAGCAGAAAC-
GACAAG and ATAGGCAGT TGGCATTCAG; IkappaB
kinase-gamma (IKK-?: CAAGAATACGACAACCACATCAA
and AACGGTCTCCATCACAAT; TNF-? TCTCCAGAT-
GTTTCCAGACTTC and CCCGGTCTCCCAAATAAATA C;
c-Jun N-terminal kinase (JNK): AGAAATACTGTTGTAGTTT-
GTGAG and CTAATGACAGAATGGTGT TTGG; ?-actin:
CATTGCCGACAGGATGC and TGGAAGGTGGACAG-
CGA), and run in the Light Cycler real-time detection system
(Roche Diagnostics). The hot-start enzyme was activated (95°C
for 5 minutes), and cDNA was then amplified for 40 cycles con-
sisting of denaturation at 95°C for 30 seconds and annealing/
extension at 60°C for 30 seconds. A melt curve was then per-
formed after amplification. ?-Actin used as an internal control
was amplified with the primers. Amplicon size and reaction spec-
ificity were confirmed using 1% agarose gel electrophoresis. Data
were analyzed using the Light Cycler software (Roche Diagnos-
tics). The cycle threshold (Ct) value of fluorescence units was used
for analysis, and quantification was done using the Ct of target
genes cDNA relative to that of ?-actin cDNA in the same sample.
Continuous data are mean ? the standard error of the mean
(SEM), unless otherwise indicated. The CP and control groups
were matched by gestational age. The matching and other po-
tential confounding variables such as mother’s age and sex were
considered in multivariate analysis of biochemical data using
generalized estimating equations that take into account intercor-
relation between the data in the matched sets.20Statistical sig-
nificance was set at p ? 0.05.
Clinical Characteristics and Outcomes
The CP group had a significantly higher percentage of
males, surfactant use, patent ductus arteriosus, and reti-
nopathy of prematurity than did the control group (Table
1). There were no significant differences in gestational
age, birth body weight, Apgar scores, the percentage of
clinical chorioamnionitis, prenatal dexamethasone use, re-
spiratory distress syndrome, sepsis or bacteremia, or
chronic lung disease between the 2 groups. There was also
no significant difference in the mean ages of the 2 groups
at the follow-up examination (mean [?SD], CP group:
7.2 ? 3.6 years; control group: 6.2 ? 2.2 years). Of the
32 preterm children with CP, 24 (75%) had spastic di-
plegia, 3 (9%) had spastic triplegia, and 5 (16%) had
spastic quadriplegia; 14 patients (44%) had Gross Motor
Function Classification Scale scores of 2 or 3, and 18
(56%) had scores of 4 or 5. The CP group also had sig-
nificantly lower full IQ, verbal IQ, and performance IQ
scores than did the control group.
Plasma Levels of TNF-?
Plasma levels of TNF-? were significantly higher in CP
group children than in control group children (means ?
SEM: CP, 53 ? 16pg/ml vs control, 10 ? 4pg/ml; p ?
.001) (Fig 1); IL-6 levels, however, were not significantly
higher (CP, 1.0 ? 0.2pg/ml vs control, 0.2 ? 0.1pg/ml;
p ? 0.6). The median number [range] of admissions for
infections was significantly higher in the CP group chil-
dren (CP, 1 [0-21] vs control, 0 [0-2]; p ? 0.038). How-
ever, after excluding 3 outliers—CP group children with
frequent hospital admissions because of infections—the
difference was no longer significant (CP, 0 [0-2] vs con-
trol, 0 [0-2]; p ? 0.104). The CP group, on the other
hand, continued to have significantly higher plasma levels
of TNF-? (CP, 53 ? 17pg/ml vs control, 10 ? 4pg/ml;
p ? 0.0032).
TNF-? Expression in LPS-Stimulated PBMCs
An in vitro model of LPS-stimulated PBMCs was estab-
lished to examine the extracellular and intracellular levels of
TNF-?. The levels of TNF-? in the supernatant of non-
stimulated PBMCs from both groups were comparable
(CP, 50 ? 19pg/ml vs control, 45 ? 16pg/ml). LPS stim-
ulation of PBMCs resulted in significant increases of
TNF-? in the supernatants in both groups of children
(both p ? 0.001). The CP group had significantly higher
levels of secreted TNF-? from LPS-stimulated PBMCs
than the control group (CP, 1,736 ? 252pg/ml vs control,
1,031 ? 135pg/ml; p ? 0.003) (Fig 2A). After 3 outli-
ers—CP group children with frequent hospital admissions
because of infections—had been excluded, the CP group
ANNALS of Neurology
206Volume 68, No. 2
still had significantly higher levels of TNF-? released from
the LPS-stimulated PBMCs (CP, 1,755 ? 272pg/ml vs
control, 1,031 ? 135 pg/ml; p ? 0.006) than the control
group. Flow cytometry analysis showed that after LPS stim-
ulation, the levels of intracellular TNF-? were significantly
decreased in the control group (n ? 13; geometric mean
[Gm] before LPS, 1.9 ? 0.1; after LPS, 1.7 ? 0.1; p ?
0.016); in contrast, the levels were significantly elevated in
the CP group (n ? 15; Gm before LPS, 1.6 ? 0.1; after
LPS, 2.1 ? 0.2; p ? 0.01) (see Fig 2B, C). After adjust-
ment for mother’s age and sex, the conclusions regarding
the comparisons of the data between the CP and control
groups in Figures 1 and 2 remained valid.
Inflammatory Signaling Molecule Expression in
We then examined whether the increased response of
TNF-? in the LPS-stimulated PBMCs of the CP children
TABLE 1: Perinatal Demographics, and Cognitive and Motor Outcome of the Control Preterm Children and
Preterm Children with Periventricular Leukomalacia-Induced Cerebral Palsy
Variables Control Group,
n ? 32
n ? 32
Mother’s age, yr (years)
Clinical chorioamnionitis (%)
Gestational age, wk
Birth body weight, g
1-minute Apgar score
5-minute Apgar score
Multiple birth (%)
Cesarean section (%)
Prenatal dexamethasone (%)
Premature rupture of membrane (%)
Respiratory distress syndrome (%)
Surfactant use (%)
Patent ductus arteriosus (%)
Chronic lung disease (%)
Retinopathy of prematurity (%)
Mean age at examination, yr
Gross Motor Function Classification Scale
6.2 (2.2)7.2 (3.6)0.46
Data are means (?standard deviation) unless otherwise indicated.
CP ? cerebral palsy; WPPSI ? Wechsler Preschool and Primary Scale of Intelligence, 3rd edition; WISC? Wechsler
Intelligence Scale for Children, 3rd edition; IQ ? intelligence quotient.
Lin et al: Proinflammation in CP
was related to the increased expression of inflammatory
signaling molecules in their PBMCs. In the nonstimu-
lated PBMCs, we did a gene profile of 84 TNF super-
family genes (SuperArray Bioscience), the expression of
which is involved in the TNF-ligand and TNF-receptor
signaling pathways; mRNA expression differences between
the randomly selected age-matched CP (n ? 5) and con-
trol group (n ? 5) were compared (Table 2). To focus on
particular pathways, the threshold was adjusted to at least
a 2-fold difference. The expression of 10 genes (CAD,
CASP2, CRADD, EDA2R, IKBKG, TAK1/TGF1a, JNK/
JNK1, 4-1BB/CD137, TL1/TL1A, and TRAF3) was
higher in the CP group than in the control group. The
genes were distributed in the inflammation and apoptosis
signaling pathways. Of particular interest were TAK1 and
JNK, which were expressed 2? (p ? 0.05) and 4.6?
(p ? 0.05) higher in the CP group. TAK1, JNK, and
IKK-? are downstream molecules of the TLR-4 mediated
pathway that triggers inflammatory responses. Real-time
PCR measurement of TLR-4, TAK1, IKK-?, JNK, and
TNF-? mRNA expression in the nonstimulated PBMCs
confirmed that mRNA levels of these genes were signifi-
cantly higher in the CP group (n ? 17) than in the con-
trol group (n ? 13) (TLR-4: CP, 13.8 ? 3.4-fold vs
control, 3.2 ? 0.6-fold; p ? 0.0023; TNF-?: CP, 53 ?
13.7-fold vs control, 8.5 ? 1.3-fold; p ? 0.0016; TAK1:
CP, 3.7 ? 1.3-fold vs control, 1.6 ? 0.2-fold; p ?
0.038; IKK-?: CP, 7.6 ? 2.8-fold vs. control, 1.6 ? 0.5-
fold; p ? 0.029; and JNK: CP, 6.1 ? 2.6-fold vs control,
0.5 ? 0.1-fold; p ? 0.045) (Fig 3A–E). In addition, the
TLR-4 mRNA levels of the nonstimulated PBMCs were
highly correlated with the secreted protein levels of
TNF-? from the LPS-stimulated PBMCs (see Fig 3F)
(Spearman rank correlation ? 0.38, p ? 0.03).
We provide the first evidence that plasma levels of
TNF-?, levels of TNF-? released from LPS-stimulated
showed that preterm school-age children with periventricu-
lar leukomalacia-induced cerebral palsy (CP) (n ? 32) had
significantly higher plasma levels of tumor necrosis factor
(TNF)-? compared with preterm term school-age control
group children (n ? 32). Data are means ? standard error
of the mean, ***p < 0.001.
1: Anenzyme-linkedimmunosorbent assay
had significantly higher levels of tumor necrosis factor
(TNF)-? released from lipopolysaccharide (LPS)-stimulated
peripheral blood mononuclear cells (PBMCs) than did pre-
term control group children. PBMCs (2 ? 106cells/dish)
were exposed to LPS (10ng/ml) for 3 hours. There were no
significant differences in the supernatant levels of TNF-? in
the nonstimulated PBMCs between the CP group (n ? 32)
and the control group (n ? 32). LPS stimulation signifi-
cantly upregulated TNF-? in the supernatants in both
groups. The CP group had significantly higher TNF-? levels
than the control group did. Data are means ? standard
error of the mean (SEM), **p < 0.01, ***p < 0.001. (B, C)
Flow cytometry showed that after LPS stimulation, intra-
cellular TNF-? levels were significantly lower in the control
group (n ? 13), but significantly higher in the CP group
(n ? 15). Gm: geometric mean. FL1-H: fluorescence inten-
sity. Data are means ? SEM. *p < 0.05.
(A) Preterm children with cerebral palsy (CP)
ANNALS of Neurology
208 Volume 68, No. 2
PBMCs, and mRNA levels of inflammatory signaling
molecules (TLR-4, TAK1, JNK, IKK-?, and TNF-?) in
PBMCs are significantly higher in school-age preterm
children with PVL-induced CP than in normal control
school-age preterm children. In addition, after LPS stim-
ulation, the intracellular TNF-? level of the PBMCs was
significantly higher in the CP group, but significantly
lower in the control group. This finding suggests that pre-
term children with a remote PVL injury-induced CP have
altered inflammatory responses.
Numerous studies21–24have reported that higher
levels of proinflammatory cytokines, such as TNF-?, in
the amniotic fluid, plasma, and umbilical cord blood are
associated with PVL or CP in preterm infants. Our study
shows that preterm children with CP had significantly
higher plasma levels of TNF-? than normal preterm chil-
dren did. The mechanism for the elevated plasma levels of
TNF-? in the preterm children with CP may in part be
related to the increased TNF-? secretion from PBMCs.
White blood cells may contribute to several neurological
diseases, such as PVL, autistic spectrum disorders, and
stroke.10,25,26Proinflammatory responses to LPS in the
monocytes of preterm sheep 7 to 14 days after exposure
to intra-amniotic endotoxin tended to exceed those of
adults and preterm control animals,27which indicates
augmented immune function by prior exposure to inflam-
mation. Because activated white blood cells contribute to
neurological dysfunction in preterm infants,10we hypoth-
esized that school-age preterm children with PVL-induced
CP have an altered reactivity to LPS in their PBMCs.
Our CP group children had significantly higher than nor-
mal levels of TNF-? secretion and production that con-
tinued to be significant even after we excluded partici-
pants with frequent hospital admissions because of
TNF-? promotes white matter injury through path-
ways, such as inhibiting the differentiation of oligoden-
drocyte progenitor cells.28TNF-? is produced by several
types of cells, but especially by macrophages in peripheral
blood and by microglia in the nervous system. TNF-? is
Preterm Children and Preterm Children with Periventricular Leukomalacia-Induced CP
mRNA Expression Levels of TNF Superfamily Genes with >2-Fold Changes between Control
UniGene SymbolAvg ?Cta
(n ? 5)
(n ? 5)
Data are means (?standard deviation.); p values were calculated using a t test.
aCt ? Avg Ct (housekeeping gene).
cp ? 0.05.
TNF ? tumor necrosis factor; CP ? cerebral palsy; Avg ? average; Ct ? cycle threshold.
Lin et al: Proinflammation in CP
August, 2010 209
an important mediator of PVL in preterm infants.29,30
We found that the levels of secreted TNF-? from the
LPS-stimulated PBMCs were higher in CP group children
than in control-group children. In addition, intracellular
TNF-? expression in PBMCs increased after LPS-
stimulation in the CP group, but decreased in the control
group. These findings suggest an augmented function of
immune responses in CP group children.
LPS stimulation of TLR-4 in immune cells activates
mitogen-activated kinases and nuclear factor (NF)-?B,
and causes the synthesis and release of TNF-?.31Our
findings that TLR-4, JNK, TAK1, IKK-?, and TNF-?
mRNA expression levels were significantly higher in the
CP group children than in the control group children are
evidence that JNK and NF-?B signaling pathways are as-
sociated with the increased sensitivity of PBMCs to LPS
in CP group children. Indeed, the JNK pathway is asso-
ciated with LPS-induced upregulation of TNF-? in mac-
rophages.32JNK and IKK-? are downstream molecules of
the TLR-4 signaling pathway that triggers inflammatory
responses. The quantity of TLR-4 limits the intensity of
the LPS response in mouse macrophages,33and variation
in the TLR-4 gene controls LPS responses.34We found
that the TLR-4 mRNA levels in nonstimulated PBMCs
were correlated with the levels of secreted TNF-? from
LPS-stimulated PBMCs. Taken together, these findings led
us to hypothesize that TNF-? expression, upregulated
through the TLR-4/JNK and TLR-4/NF-?B signaling
pathways, is related with increased LPS sensitivity in the
PBMCs of preterm children with PVL-induced CP (Fig 4).
The underlying causes of altered inflammatory re-
sponses in the preterm children with remote PVL injury-
induced CP remain unknown. In addition to the possi-
bility of the effect of genetic susceptibility, the increased
LPS sensitivity of PBMCs in CP group children suggested
the hypothesis that PVL or related inflammation events or
from cerebral palsy (CP) group children had significantly
higher levels of proinflammatory signaling molecules. (A–E)
Real-time polymerase chain reaction was used to measure
the mRNA expression of toll-like receptor-4 (TLR-4), trans-
forming growth factor-?–activated
(JNK), and tumor necrosis factor (TNF)-? in nonstimulated
PBMCs. mRNA expression of TNF-?, TLR-4, TAK1, IKK-?,
and JNK was significantly higher in the CP group (n ? 17)
than in the control group (n ? 13). (F) The TLR-4 mRNA
levels of nonstimulated PBMCs were highly correlated with
the protein levels of TNF-? released by the lipopolysac-
charide (LPS)-stimulated PBMCs (Spearman rank correla-
tion ? 0.38, p ? 0.03). mRNA expression is relative to
?-actin. Data are means ? standard error of the mean.
*p < 0.05, **p < 0.01.
Peripheral blood mononuclear cells (PBMCs)
ANNALS of Neurology
210Volume 68, No. 2
both during the perinatal and postnatal period may have a
programming effect, causing altered inflammatory re-
sponses in preterm children with CP. The other possibil-
ity may relate with the finding that the nervous and im-
mune systems communicate with each other.35It may be
possible to use peripheral immune cells to assess neuro-
logical function, stress response, and post-traumatic stress
disorder.36,37Crosstalk between neurons and immune cells
has been reported in several neurological disorders, includ-
ing Alzheimer disease, stroke, and autism spectrum disor-
ders,25,26,38,39which indicates that brain dysfunction can
be detected by aberrant gene expression in immune cells.
Neonates are sensitive to programming that perma-
nently affects development. Plasticity altered after perina-
tal programming may have long-term effects on the in-
flammatory responses of circulating leukocytes. It remains
to be determined whether the altered inflammatory re-
sponses in preterm children with PVL-induced CP reflects
persistent neuroinflammation. Our study provides sup-
port for a recent hypothesis14that the inflammatory pro-
cess that starts during the perinatal period may continue
into the childhood period. More studies are needed to
determine the functional consequence of altered inflam-
matory responses in preterm children with CP. A longi-
tudinal study, in which blood samples from preterm in-
fants are collected from the perinatal period until school
age, is needed to determine whether the increased TNF-?
expression in school-age preterm children with CP indi-
cates a persistently proinflammatory state that began dur-
ing the perinatal period.
This study was supported by grants from the Taiwan Na-
tional Health Research Institute (NHRI-EX 95,96,97-
9414NI), the National Science Counsel (NSC 97-2811-
Center for Gene Regulation and Signal Transduction Re-
search, National Cheng Kung University (NCKU96-99).
We thank C.-J. Ho and J.-H. Chao for skillful as-
sistance in the patient recruitment and laboratory work,
K.-J. Chen for statistical assistance, and Dr W. J. Rogan
for comments on the manuscript during its preparation.
Potential Conflicts of Interest
Nothing to report.
1.Shanks N, Lightman SL. The maternal-neonatal neuro-immune
interface: are there long-term implications for inflammatory or
stress-related disease? J Clin Invest 2001;108:1567–1573.
2.Yeh TF, Lin YJ, Lin HC, et al. Outcomes at school age after
postnatal dexamethasone therapy for lung disease of prematu-
rity. N Engl J Med 2004;350:1304–1313.
3.Hagberg H, Mallard C. Effect of inflammation on central nervous
system development and vulnerability. Curr Opin Neurol 2005;
4. Gluckman PD, Hanson MA, Cooper C, Thornburg KL. Effect of in
utero and early-life conditions on adult health and disease.
N Engl J Med 2008;359:61–73.
5.Vincer MJ, Allen AC, Joseph KS, et al. Increasing prevalence of
cerebral palsy among very preterm infants: a population-based
study. Pediatrics 2006;118:e1621–e1626.
6.Khwaja O, Volpe JJ. Pathogenesis of cerebral white matter injury
of prematurity. Arch Dis Child 2008;93:F153–F161.
7.Saigal S, Doyle LW. An overview of mortality and sequelae of
preterm birth from infancy to adulthood. Lancet 2008;371:
8.Rezaie P, Dean A. Periventricular leukomalacia, inflammation and
white matter lesions within the developing nervous system. Neu-
9.Saliba E, Marret S. Cerebral white matter damage in the preterm
infant: pathophysiology and risk factors. Semin Neonatol 2001;6:
10. Dammann O, Durum S, Leviton A. Do white cells matter in white
matter damage? Trends Neurosci 2001;24:320–324.
11.Leviton A, Dammann O, Durum SK. The adaptive immune re-
sponse in neonatal cerebral white matter damage. Ann Neurol
12. Wang X, Hagberg H, Nie C, et al. Dual role of intrauterine
immune challenge on neonatal and adult brain vulnerability to
hypoxia-ischemia. J Neuropathol Exp Neurol 2007;66:552–561.
13.Galic MA, Riazi K, Heida JG, et al. Postnatal inflammation in-
creases seizure susceptibility in adult rats. J Neurosci 2008;28:
receptor-4 (TLR-4)/c-Jun N-terminal kinase (JNK) and TLR-
4/nuclear factor (NF)-?B signaling pathways may be in-
volved in increases of lipopolysaccharide (LPS)-mediated
tumor necrosis factor (TNF)-? expression in the peripheral
blood mononuclear cells of preterm children with periven-
tricular leukomalacia (PVL)-induced cerebral palsy (CP).
TAK1 ? transforming growth factor-?–activated kinase 1;
IKK ? IkappaB kinase; TAB1 ? TAK1 binding protein 1;
TAB2 ? TAK1 binding protein 2; AP1 ? activator protein 1.
[Color figure can be viewed in the online issue, which is avail-
able at www.interscience.wiley.com.]
Schema shows that upregulation of toll-like
Lin et al: Proinflammation in CP
August, 2010 211
14.Dammann O. Persistent neuro-inflammation in cerebral palsy: a
therapeutic window of opportunity? Acta Paediatr 2007;96:6–7.
15. Wang LW, Wang ST, Huang CC. Preterm infants of educated
mothers have better outcome. Acta Paediatr 2008;97:568–573.
16.Palisano R, Rosenbaum P, Walter S, et al. Development and
reliability of a system to classify gross motor function in children
with cerebral palsy. Dev Med Child Neurol 1997;39:214–223.
17.Volpe JJ. Hypoxic-ischemic encephalopathy: clinical aspects. In
Volpe JJ, eds. Neurology of the newborn. 5th ed. Philadelphia,
PA: WB Saunders, 2009:400–480.
18.Kowalski ML, Wolska A, Grzegorczyk J, et al. Increased respon-
siveness to toll-like receptor 4 stimulation in peripheral blood
mononuclear cells from patients with recent onset rheumatoid
arthritis. Mediators Inflamm 2008;2008:1–7.
19. Lin CF, Lei HY, Shiau AL, et al. Antibodies from dengue patient
sera cross-react with endothelial cells and induce damage.
J Med Virol 2003;69:82–90.
20.Diggle PJ, Heagerty P, Liang KY, Zeger SL. Analysis of longitu-
dinal data. Oxford, UK: Oxford University Press, 2002.
21.Tsukimori K, Komatsu H, Yoshimura T, et al. Increased inflam-
matory markers are associated with early periventricular leukoma-
lacia. Dev Med Child Neurol 2007;49:587–590.
22.Yoon BH, Jun JK, Romero R, et al. Amniotic fluid inflammatory
cytokines (interleukin-6, interleukin-1beta, and tumor necrosis
factor-alpha), neonatal brain white matter lesions, and cerebral
palsy. Am J Obstet Gynecol 1997;177:19–26.
23.Kaukola T, Satyaraj E, Patel DD, et al. Cerebral palsy is charac-
terized by protein mediators in cord serum. Ann Neurol 2004;
24.Hansen-Pupp I, Hallin AL, Hellstrom-Westas L, et al. Inflammation
at birth is associated with subnormal development in very pre-
term infants. Pediatr Res 2008;64:183–188.
25.Jyonouchi H, Sun S, Le H. Proinflammatory and regulatory cyto-
kine production associated with innate and adaptive immune re-
sponses in children with autism spectrum disorders and devel-
opmental regression. J Neuroimmunol 2001;120:170–179.
26.Moore DF, Li H, Jeffries N, et al. Using peripheral blood mononu-
clear cells to determine a gene expression profile of acute ischemic
stroke: a pilot investigation. Circulation 2005;111:212–221.
27.Kramer BW, Ikegami M, Moss TJ, et al. Endotoxin-induced cho-
rioamnionitis modulates innate immunity of monocytes in pre-
term sheep. Am J Respir Crit Care Med 2005;171:73–77.
28.Feldhaus B, Dietzel ID, Heumann R, Berger R. Effects of
interferon-gamma and tumor necrosis factor-alpha on survival
and differentiation of oligodendrocyte progenitors. J Soc Gy-
necol Investig 2004;11:89–96.
29. Kadhim H, Khalifa M, Deltenre P, et al. Molecular mechanisms of
cell death in periventricular leukomalacia. Neurology 2006;67:
30.Kadhim H, Tabarki B, Verellen G, et al. Inflammatory cytokines in
Lipopolysaccharide-induced tumor necrosis factor alpha produc-
tion by human monocytes involves the raf-1/MEK1-MEK2/ERK1-
ERK2 pathway. Infect Immun 1999;67:3824–3829.
derBruggen T,Nijenhuis S,vanRaaij E,et al.
32.Alexander M, Daniel T, Chaudry IH, Schwacha MG. MAP kinases
differentially regulate the expression of macrophage hyperactivity
after thermal injury. J Cell Physiol 2004;201:35–44.
33.Takeuchi O, Hoshino K, Kawai T, et al. Differential roles of TLR2
and TLR4 in recognition of gram-negative and gram-positive bac-
terial cell wall components. Immunity 1999;11:443–451.
34.Mathison JC, Virca GD, Wolfson E, et al. Adaptation to bacterial
lipopolysaccharide controls lipopolysaccharide-induced tumor ne-
crosis factor production in rabbit macrophages. J Clin Invest
35.Boulanger LM, Huh GS, Shatz CJ. Neuronal plasticity and cellular
immunity: shared molecular mechanisms. Curr Opin Neurobiol
36.Bierhaus A, Wolf J, Andrassy M, et al. A mechanism converting
psychosocial stress into mononuclear cell activation. Proc Natl
Acad Sci U S A 2003;100:1920–1925.
37.Segman RH, Shefi N, Goltser-Dubner T, et al. Peripheral blood
mononuclear cell gene expression profiles identify emergent
post-traumatic stress disorder among trauma survivors. Mol Psy-
38. Maes OC, Xu S, Yu B, et al. Transcriptional profiling of Alzheimer
blood mononuclear cells by microarray. Neurobiol Aging 2007;
39.Vargas DL, Nascimbene C, Krishnan C, et al. Neuroglial activa-
tion and neuroinflammation in the brain of patients with autism.
Ann Neurol 2005;57:67–81.
ANNALS of Neurology
212 Volume 68, No. 2