| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Maternal and Fetal Medicine Section, Institute of Medical Genetics, Yorkhill, Glasgow G3 8SJ, U.K.;
Department of Obstetrics and Gynaecology and
Department of Pathology, Royal Victoria Infirmary, Newcastle upon Tyne NE1 4LP, U.K.; and
Department of Obstetrics and Gynecology, University of Cincinnati, College of Medicine, Cincinnati, Ohio 45267, USA
1Correspondence: Maternal and Fetal Medicine Section, Institute of Medical Genetics, Yorkhill, Glasgow G3 8SJ, U.K. E-mail: f.lyall{at}udcf.gla.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: pregnancy • HO-1 • HO-2 • trophoblast • carbon monoxide
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
The maternal uteroplacental circulation adapts to pregnancy via
striking changes in the uterine spiral arteries. During early human
pregnancy, extravillous CTBs from anchoring villi invade the
decidualized endometrium and the myometrium (interstitial trophoblast)
and also migrate in a retrograde direction along the spiral arteries
(endovascular trophoblast), transforming them into large conduit
vessels of low resistance (2)
. This physiological
transformation is characterized by a gradual loss of the normal
musculoelastic structure of the arterial wall and replacement by
amorphous fibrinoid material in which trophoblast cells are embedded
(3
4
5
6
7
8)
. These physiological changes are thought to be
required for a successful pregnancy. However, vasodilation of spiral
arteries has been reported to take place before endovascular CTB
invasion has occurred (2
, 9)
, and Pijnenborg et al.
(2)
have related this to the presence of interstitial CTB,
suggesting that these cells may produce vasoactive mediators. In
support of this idea, we have shown that both endovascular and
interstitial extravillous CTBs within the placental bed express heme
oxygenase (HO) (10)
. These results lend support to the
hypothesis that trophoblast expression of HO could result in carbon
monoxide (CO) release from these cells, which in turn could contribute
to the invasion process itself and to spiral artery transformation.
Failure of spiral artery transformation has been documented in
preeclampsia (PE), one of the leading causes of maternal death. In this
syndrome, hypertension is associated with widespread maternal
endothelial dysfunction and fetal growth restriction (FGR), leading to
significant maternal and perinatal morbidity (11)
. Similar
spiral artery abnormalities have also been reported in the placental
bed of women with FGR in the absence of maternal hypertension (5
, 12
13
14
15
16)
. The factors responsible for the failure of vascular
remodeling are not well understood, but the net result is a reduction
in uteroplacental blood flow (17)
and evidence of
increased oxidative stress within both the maternal and the placental
circulations (18)
.
HO is a microsomal enzyme that oxidatively cleaves heme, a pro-oxidant,
to produce biliverdin, a potent antioxidant, and CO (19)
.
CO, like NO, activates soluble guanylate cyclase to produce cyclic
guanosine monophosphate (20)
. HO consists of three
homologous isoenzymes: HO-1, which is inducible; HO-2, which is
constitutively expressed in tissues (21
, 22)
; and HO-3,
which is a more recently identified isoform with low activity
(23)
. HO-1 is expressed at high concentrations in the
spleen and liver, where it is responsible for the destruction of heme
from red blood cells. HO-1 can be induced by an extraordinary array of
stimuli including hypoxia and hyperoxia (24
25
26
27
28
29
30)
. The
actions of HO-1 rid cells of pro-oxidants, allowing cells to withstand
further exposure to harmful stimuli. HO-2 is not thought to be
inducible and is widely distributed throughout the body. CO acts as a
neurotransmitter (31)
, inhibits platelet aggregation
(32)
, and is a vascular smooth muscle relaxant
(33)
.
We have recently reported that HO-2 is expressed by placental villous
trophoblast and endothelial cells and that expression of HO-2 in these
cells is regulated in a temporal and spatial manner throughout
pregnancy (10)
. In contrast, HO-1 expression was very low
in normal placenta. Furthermore, in placental perfusion studies,
inhibition of HO activity led to a dose-dependent increase in perfusion
pressure, suggesting that CO may also operate as a vasodilator in the
placenta (10)
, and this result is supported by evidence of
HO enzyme activity in the human placenta (34)
. We also
reported that interstitial and endovascular CTBs express both HO-1 and
HO-2 (10)
, implying a possible role for CO in spiral
artery transformation. These findings raise the possibility that
abnormalities in HO expression may contribute to the pathophysiology of
PE and FGR, two conditions characterized by deficient trophoblast
invasion and reduced uteroplacental blood flow.
In this study, we hypothesized that PE and FGR would be associated with alterations in one or both HO isoforms in the placenta and placental bed. To test this hypothesis, we used immunohistochemistry and Western blotting to examine the expression of HO isoforms in the human villous placenta and placental bed in the third trimester in cases of PE and FGR and in matched control pregnancies.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
140/90) and proteinuria (300 mg/24 h) in women who
were normotensive before pregnancy and had no other underlying clinical
problems such as renal disease. FGR was defined ultrasonically as fetal
abdominal circumference (AC) of <10th centile with a decrease in AC
standard deviation score (SDS) of >1.5 SDs
(35)
and an umbilical artery pulsatility index (PI) of
95th centile (36)
. We have previously shown that a fall
in AC SDS of >1.5 SDs is the optimal cut-off to define a
group of fetuses with evidence of wasting at birth and morbidity
associated with FGR (35)
. Birth weight centiles were
obtained from charts of the Northern Region population of England
(37)
.
Sample collection
Placental bed biopsy samples were obtained from women undergoing
elective cesarean section as described previously (10
, 38)
. Briefly, after delivery of the infant, the position of the
placenta was determined by manual palpation. Six placental bed biopsy
samples were then taken under direct vision using biopsy forceps (Wolf,
Wimbledon, U.K.). In three cases, samples were collected after vaginal
delivery. These samples were taken under ultrasound guidance using the
same biopsy forceps as that introduced through the cervix. Placental
bed biopsy samples were included in this study if they contained
decidual and/or myometrial spiral arteries with interstitial
trophoblast. Placental samples of 
1 cm3 were
also collected from all cases. All samples were collected directly into
liquid nitrogen-cooled isopentane and stored sealed at -70°C until
required. Samples were used for subsequent immunohistochemical analysis
and Western blotting experiments, and the numbers in each group are
included in the appropriate section below. Cryosections (7 µm) from
each specimen were stained with hematoxylin and eosin for histological
analysis. In addition, sections of placental bed biopsy samples were
immunostained with antibodies to cytokeratin to detect trophoblast,
desmin to detect muscle, and PECAM (platelet endothelial cell adhesion
molecule) to detect endothelium.
Morphological assessment of spiral arteries
After immunostaining with the above antibodies, we assessed the
integrity of the muscle wall of the spiral artery by the degree of
muscle remaining around the spiral artery (desmin immunostaining). The
muscle was graded as preserved, separated, disrupted, or absent.
Absence of muscle changes was deemed as when the muscle was scored as
preserved or separated, and presence of muscle changes was deemed as
when the muscle was disrupted or absent.
Materials
All chemicals were purchased from Sigma Chemical (Poole, U.K.)
unless stated otherwise.
Antibodies
Desmin (NCL-DES-DERII, 1:100) and cytokeratin (NCL-LP34, 1:200)
monoclonal antibodies were obtained from Novocastra, Newcastle upon
Tyne, U.K. The PECAM antibody (1:10,000) was obtained from R&D Systems,
Abindgon, U.K. The HO-2 (OSA-200) rabbit polyclonal antibody was
obtained from StressGen Biotechnologies, Victoria, Canada, and was
raised against purified native rat testis HO-2. This antibody had been
used in our previous study (10)
. To detect HO-1, we used
the same polyclonal antibody (SPA-895) raised against purified rat
recombinant HO-1 as reported previously (10)
. A different
lot number (904409) was used in the present study, as the original lot
number (708405) was no longer available. This antibody produced
different results in immunohistochemistry compared with the original
lot number. From discussion with the manufacturers of the antibody, it
emerged that the two lots were raised in different rabbits. Because of
these differences, it was important to confirm the specificity of other
commercially available HO-1 antibodies and then compare their ability
to detect HO-1 in the samples from the present study. Thus, three other
HO-1 antibodies were tested: 1) OSA-100, a rabbit polyclonal antibody
raised against purified rat liver HO-1 (StressGen); 2) OSA-110, a
monoclonal antibody raised against a synthetic peptide of human HO-1
(StressGen); and 3) HC 3001, a rabbit polyclonal antibody raised
against a synthetic peptide found in the human sequence (Affiniti,
Exeter, U.K.).
Western blotting
Western blots were used to determine overall expression of HO-2
and HO-1 in placental samples comprising full-thickness blocks from
chorionic plate through to basal plate. Tissue samples were ground to a
fine powder in liquid nitrogen with a mortar and pestle and added to 4
volumes of cold lysis buffer (25 mM Tris/0.25 M sucrose/1 mM EDTA, pH
7.6) and 50 µl/g tissue protease inhibitor mixture (Sigma). With a
Polytron homogenizer at setting 10, the sample containers were
surrounded by ice and homogenized for three 10-s intervals. The
homogenate was spun at 5000g for 10 min at 4°C to remove
debris, and the resultant supernatant was divided into aliquots and
stored at -70°C. Protein concentrations were determined by the
method of Bradford (39)
, using bovine serum albumin as a
standard, and were diluted to the required concentration.
Samples were mixed 1:1 with loading buffer (1.2 ml 1 M Tris, pH 6.8, 2
ml of glycerol, 4 ml of 10% sodium dodecyl sulfate, 2 ml of 1 M
dithiothreitol, and 0.8 ml of distilled water with bromphenol blue
added to give a deep blue color) and were boiled for 5 min before
loading. Samples were separated on 10% sodium dodecyl
sulfate-polyacrylamide resolving gels with a 4% stacking gel using a
Protean II apparatus (Bio-Rad, Hemelhempstead, U.K.) (40)
at a constant current of 30 mA. Each well was loaded with 50 µg of
protein. Molecular mass markers (SDS-7B prestained 33–205 kDa; Sigma)
were loaded beside the samples.
Protein was transferred overnight in buffer containing 25 mM Tris, 190 mM glycine, and 20% methanol at a constant 30 V to BioBlot NC nitrocellulose membranes (Costar, Corning, High Wycombe, U.K.). Filters were blocked for 1 h at room temperature (RT) in TBSTB buffer (20 mM Tris, pH 7.5, 0.5 M NaCl, 0.4% Tween 20, and 0.25% bovine serum albumin) containing 5% normal donkey serum. Both HO-1 and HO-2 antibodies were prepared at a concentration of 1:1000 in TBSTB containing 5% normal human serum and were preabsorbed for 1 h at RT to reduce nonspecific binding before being used for immunodetection. Omission of this step resulted in many nonspecific bands appearing on the autoradiograph. The antibodies were added for 1 h at RT. The filters were rinsed once and then were washed twice for 5 min in TBSTB and were incubated with horseradish peroxidase-conjugated donkey anti-rabbit immunoglobulin G (Diagnostics Scotland, Carluke, U.K.) diluted 1:2000 in TBSTB for 1 h at RT. Blots were then rinsed again and washed twice in TBSTB followed by one 5-min wash in distilled water. Proteins were detected with the Amersham ECL detection system, and filters were exposed to Hyperfilm ECL (Amersham, Buckinghamshire, U.K.). Bands on the exposed films were scanned on a Bio-Rad GS-700 imaging densitometer connected to a Macintosh computer loaded with Bio-Rad imaging software. Exposures on autoradiographs were such that readings were on the linear range of the densitometer. After scanning of the blots, the density of each band was expressed as a ratio of one sample of a control case (internal control), which was added to every gel.
Induction of HO-1 in myometrial cell cultures
Because we previously showed that HO-1 is present at very low to
undetectable levels in normal placenta, an additional positive control
for HO-1 was added to Western blots. This control consisted of
homogenates of primary human myometrial cells stimulated with 100 µM
cadmium chloride (41)
. These cells do not express HO-1
under normal culture conditions, but HO-1 can be induced following
incubation with cadmium chloride (41)
.
Immunohistochemistry
Sections were all stained on the same day (for each antibody) to
eliminate day-to-day variations in immunostaining. Immunohistochemistry
was performed with the Vectastain Universal kit (Vector
Laboratories, Peterborough, U.K.). Sections (7 µm) were cut on a
cryostat and mounted on 3-aminopropyltriethoxysilane-coated glass
slides. Each specimen was stained with hematoxylin and eosin for
histological analysis. In addition, to assist in identification of
spiral arteries and trophoblast, sections from placental bed biopsy
samples were immunostained for cytokeratin (1:200) to detect
trophoblast, desmin (1:100) to detect muscle, and PECAM
(1:10,000) to detect endothelium. For HO immunostaining, sections,
which were used immediately after air-drying, were fixed in 1%
paraformaldehyde for 5 min, dehydrated in 100% ethanol for 5 min, and
then rehydrated in water twice for 5 min. These and all subsequent
steps were performed at RT. Nonspecific binding sites were blocked by
incubation with blocking agent (Biogenex, San Ramon, CA) for 15 min in
a humidified chamber; after washing in TBSTB for 5 min, the sections
were then incubated with either HO-1 or HO-2 antibodies for 45 min.
HO-2 and HO-1 antibodies were used at concentrations of 1:250 and
1:1000, respectively, in TBSTB containing 5% normal human serum and,
as for the Western blots, were preabsorbed for 1 h in this buffer
before immunodetection, to reduce nonspecific binding. Following two
5-min TBSTB washes, the biotinylated secondary antibody was added for
30 min at RT. Two more TBSTB washes were performed, and then endogenous
peroxidase activity was quenched by incubating the sections in 1%
(v/v) hydrogen peroxide in methanol for 15 min. For PECAM, smooth
muscle actin, and cytokeratin antibodies, acetone-fixed sections were
blocked with the blocker supplied with the Universal kit for 30 min at
37°C and were washed in PBS 2 x 5 min, after which the primary
antibody (diluted in blocking buffer) was added for 90 min at 37°C.
Following 2 x 5 min PBS washes, the secondary antibody was added
for 30 min at 37°C. Two more PBS washes were performed, and then
endogenous peroxidase activity was blocked with 1% hydrogen peroxide
in methanol for 15 min at RT. The remaining steps were performed
according to the instructions supplied with the kit. Immunoreactive
proteins were detected with FastTM diaminobenzidine tablets (Sigma).
Sections were counterstained in Harris’s hematoxylin (BDH, Poole,
U.K.) and mounted in synthetic resin. Omission of primary antibody and
substitution of nonimmune serum for the primary antibody were both
included as controls and resulted in no immunostaining. Intensity of
immunostaining was scored 0–3, with 0 representing no staining; 1,
light staining; 2, moderate staining; and 3, dark staining. If the
observer thought that staining was between two scores, a midpoint
score, e.g., 2.5, was allotted. One block of tissue was used from each
placenta. Thirty fields were counted at 10x magnification for each
section. Scoring was performed independently by two observers (FL and
AB) "blinded" to the tissue identity. As there were no significant
differences between the two sets of scores (ANOVA; P>0.05
for all comparisons), the results for one observer (FL) were used for
subsequent analysis.
Statistical analysis
Clinical details were compared by using ANOVA, and post hoc
testing was performed with Fisher’s PLSD test. For Western blot and
immunohistochemical studies, statistical comparisons were performed by
using the Kruskal-Wallis one-way ANOVA test. Comparison between paired
groups was then performed with the Mann-Whitney U test.
Statistical differences were considered to be significant at
P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, which shows the cases used for placental immunohistochemical studies.
Gestational age at delivery was comparable in the three groups. All
fetuses in the control group were appropriately grown for their
gestational age, and four in the PE group were below the 10th centile.
The umbilical artery PI was elevated in all of the FGR fetuses; four
had absent and three had reversed end-diastolic frequencies. Birth
weight was significantly reduced in the FGR group relative to the PE
and control groups; all infants in the FGR group had a birth weight
below the 10th centile, with six below the 5th centile. Figure 1A
shows a representative blot for HO-2 (upper panel) and HO-1
(lower panel). For HO-2, a band of 36 kDa was identified in all of
the samples, which is consistent with our previous observations and
agrees with the molecular mass reported for other human tissues.
Scanning densitometry of the bands revealed no significant overall
differences between the groups (Fig. 1B
). HO-1 was not
detected in any of the samples.
|
|
Immunohistochemistry: placentas
The clinical details for patients used for the placenta
immunohistochemistry studies are shown in Table 1
. Gestational age at
delivery was comparable in the three groups. Umbilical artery PI was
elevated in all of the FGR fetuses; four had absent and three had
reversed end-diastolic frequencies. Birth weight was significantly
reduced in both the FGR and PE groups when compared with the control
group. All infants in the FGR group had a birth weight below the 10th
centile with 6 below the 5th centile. Three of the infants in the PE
group had birth weights less than the 10th centile. Umbilical artery
Doppler data were available for 16 of the PE group: 11 had a normal PI,
3 had a PI > 2 SDs above the mean for gestation, and 1 had
absent end-diastolic frequencies.
Consistent with our previous findings, immunostaining for HO-2 was
strongly expressed on endothelial cells and weakly expressed on
syncytiotrophoblast and CTB cells in control pregnancies in the third
trimester (Fig. 2A
). Endothelial HO-2 expression was reduced in pregnancies
complicated by PE (Fig. 2B
) and FGR (Fig. 2C
). Immunostaining scores showed that HO-2
endothelium immunostaining was significantly reduced in both the
PE group (P<0.01) and the FGR group (P<0.0005)
when compared with the control group (Fig. 3
). These results demonstrate the importance of performing both Western
blots and immunohistochemistry on placental tissue in which, because of
the different ratio of trophoblast to endothelium, changes in
expression of proteins on one particular cell type could be missed by
performing Western blots alone. There was no significant difference in
endothelial HO-2 immunostaining between the PE and FGR group. In
contrast, trophoblast HO-2 immunostaining was comparable in the three
groups (Fig. 3)
. Endothelial and trophoblast immunostaining was
undetectable when the primary antibody was omitted (not shown) or
replaced with rabbit serum (Fig. 2D
), or when the secondary
antibody was omitted (not shown).
|
|
Within the placenta, the immunostaining pattern for HO-1 was different
from that for HO-2. Most samples showed weak or negative staining,
although occasional areas of endothelium, muscle, or
syncytiotrophoblast were stained in some villi (Fig. 2E
F
G
).
These results are consistent with Western blot findings and our
previous observations suggesting that there is little HO-1 in the
normal placenta. Immunostaining for HO-1 in the placentas of PE and FGR
cases was also weak or negative (Fig. 4
). The antibody detected HO-1 present on myometrial cells stimulated
with cadmium chloride (Fig. 2H
).
|
Placental bed biopsies
The clinical details for patients used for the placental bed
immunohistochemistry studies are shown in Table 2
. Gestational age at delivery was comparable in the three groups.
Umbilical artery PI was elevated in all of the FGR fetuses; four had
absent and three had reversed end-diastolic frequencies. Birth
weight was significantly reduced in both the FGR and PE groups when
compared with the control group. All infants in the FGR group had a
birth weight below the 10th centile, with six below the 5th centile.
Four of the infants in the PE group had birth weights of <10th
centile. Cytokeratin-positive extravillous CTBs consistently stained
positive for HO-2 regardless of whether they were interstitial,
endovascular, or in the wall of a spiral artery. The intensity of CTB
HO-2 immunostaining within the placental bed was comparable in PE and
FGR cases relative to the controls (Fig. 5
).
|
|
Figure 5A
B
C
D
shows two separate cases of control
pregnancies. In Fig. 5A
, cytokeratin-positive cells surround
a spiral artery, and these same cells plus the endothelium are also
HO-2 positive (Fig. 5B
). In Fig. 5C
and
D, a spiral artery at the decidual myometrial interface is
shown. The vessel is surrounded by cytokeratin-positive CTBs
(D), and a parallel section shows that these cells are also
HO-2 positive (Fig. 5C
). A case of PE is shown in Fig. 5E
(HO-2) and Fig. 5F
(cytokeratin). Similar
to the control pregnant cases, in PE the CTBs within the placental bed
were also HO-2 positive. Figure 5G
and H shows a
case of FGR. This spiral artery at the decidual myometrial interface
was surrounded by cytokeratin-positive CTBs (Fig. 5H
), which
were also HO-2 positive. CTBs within the lumen or CTBs adjacent to the
lumen but separated from the lumen by endothelium were also HO-2
positive.
In control pregnancies, extravillous CTBs, which stained positive for
cytokeratin (Fig. 6A
), were virtually HO-1 negative (Fig. 6B
). This
finding was quite unexpected, as our previous studies suggested that
this population of CTBs expressed HO-1. HO-1 immunostaining was also
absent on extravillous CTBs from cases of PE and FGR (Fig. 6C
Fig. 6D
Fig. 6E
Fig. 6F
).
|
Analysis of HO-1 expression using different commercially available
antibodies
We had previously demonstrated HO-1 immunostaining on CTBs using
the same antibody (SPA-895, StressGen) albeit a different lot number
(708405) (10)
. After discussion with the company, it
emerged that the two lots were raised in different rabbits.
Furthermore, we were unable to obtain any of the original antibody.
Because of the differences in HO-1 immunostaining, we thought that it
was important to confirm the immunohistochemical findings by first
checking the specificity of other commercially available HO-1
antibodies and then comparing their ability to detect HO-1 in the
samples used in the this study. Figure 7
shows the results of testing available HO-1 antibodies for their
ability to detect HO-1 in Western blots. All antibodies were used at
1:1000.
|
SPA-895 (lot 904409) detected rat positive control HO-1 protein and
HO-1 in human myometrial cells stimulated with cadmium chloride
treatment (Fig. 7A
). HO-1 was undetectable in unstimulated
cells. Antibody OSA-100 detected the rat positive control protein but
not the increased HO-1 in stimulated human myometrial cells (Fig. 7B
). Thus, further immunostaining was not undertaken with
this antibody. Antibody OSA-110 failed to detect the rat positive
control protein but did detect HO-1 in the human stimulated myometrial
cells. This antibody also detected a second species of lower molecular
weight than the HO-1, which was not present in the nonstimulated cells
(Fig. 7C
). Thus, this antibody is probably suitable only for
studies of human tissues, but because of the appearance of the second
band in the stimulated cells, it was deemed less suitable than SPA-895.
Antibody HC 3001 failed to detect the rat positive control protein but
did detect HO-1 in the stimulated human cells (Fig. 7D
).
Next, immunostaining of cytospin preparations of myometrial cells
stimulated with cadmium chloride was undertaken with the three
successful antibodies, i.e., SPA-895, OSA-110, and HC 3001. Each
antibody was tested at a range of dilutions, and the dilution that
produced staining in stimulated cells but no staining in unstimulated
cells was chosen. Immunostaining with SPA-895 (1:1000) was very low in
unstimulated cells (Fig. 2H
left panel) and was up-regulated
in response to incubation with cadmium chloride (right panel). However,
this lot number was not as sensitive as the previous lot number
(41)
. Both OSA-110 (1:2000) and HC 3001 (1:250) detected
HO-1 in stimulated cells, but unstimulated cells were negative (Fig. 5I
J
). The monoclonal antibody OSA-110 produced
the best staining of the cytospin preparations; HC 3001 was the least
sensitive.
Finally, the three antibodies were tested, at the same dilutions, on
parallel sections of placental bed biopsy samples containing
extravillous CTBs (Fig. 6G-J
Fig. 6G
Fig. 6H
Fig. 6I
Fig. 6J
). All three
antibodies failed to detect HO-1 on these cells. Collectively, these
data suggest that extravillous CTBs do not express HO-1.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The
increased resistance may be due to altered vascular anatomy
(43)
and/or altered production or response to vasoactive
agents (44)
. This is the first study to examine the
expression of HO isoforms in placental and placental bed biopsy
material in PE and FGR. All cases were carefully selected. Birth weight
was reduced in both study groups, although this was most marked in the
FGR fetuses. The extent of placental disease in the FGR group was also
reflected in the severity of the umbilical artery Doppler
abnormalities. Although we found no differences in HO-2 expression on
trophoblasts, expression of this isoform in villous endothelial cells
was dramatically reduced in both PE and FGR. These findings suggest
that reduced availability of CO within the placental vasculature may
contribute to the pathophysiology of these disorders.
Vasomotor control within the fetoplacental circulation is regulated by
vasoconstrictors and vasodilators including NO (1)
. Human
placental syncytiotrophoblasts express the endothelial-type NO synthase
(eNOS) (45
, 46)
. NO produced by syncytiotrophoblast has at
least three possible physiological targets: the intervillous space,
autocrine effects on trophoblast function, and paracrine interactions
with villous core components. NO inhibits platelet aggregation and
leukocyte adhesion while modulating the immune response. There appear
to be no overt differences in eNOS immunostaining in the
syncytiotrophoblast (47)
or in eNOS activity in the
villous trophoblast (48)
in PE, although placentas from
cases of PE show a more basal distribution of eNOS in the
syncytiotrophoblast (44)
. We have speculated that, in
keeping with the complementary roles of HO-2 and NOS in other tissues
(49)
, HO-2 on the syncytiotrophoblast may have roles
overlapping with those proposed for eNOS (10)
. CO produced
by the syncytiotrophoblast would benefit both maternal and fetal blood
flow through the intervillous space and possibly avoidance of immune
recognition of the fetoplacental semiallograft. HO-2 expression on the
syncytiotrophoblast is known to be highest in early pregnancy compared
with the third trimester (10)
. In the present study, we
found that HO-2 expression on the syncytiotrophoblast was not
significantly different from that observed in cases of PE and FGR.
The expression of HO-2 on placental villous endothelial cells was
reduced in pregnancies complicated by PE and FGR. We previously showed
that inhibition of HO using zinc protoporphyrin increased resistance in
the perfused placenta in a dose-dependent manner, suggesting that CO is
an endogenous vasodilator within the placental vasculature. Thus, the
reduction in HO-2 expression on endothelial cells in PE and FGR may
well contribute to the reduction of blood flow in these conditions
(17
, 50)
. Although the PE cases were not selected on the
basis of Doppler studies, it was noteworthy that in the FGR group, in
which Doppler results were extremely abnormal, the reduction in HO-2
immunostaining was also more significant. Villous endothelial cells
also express eNOS. Inhibition of NOS also increases placental perfusion
pressure (1)
, suggesting that NO and CO may have a
synergistic role in the maintenance of normal basal tone. However, in
contrast to HO, eNOS expression has been reported to be increased on
the endothelium of stem and terminal villous vessels in placentas from
PE and FGR pregnancies (44)
. This increase in eNOS
expression may be a secondary adaptive response to the increased
resistance and poor perfusion in these pathological disorders. This
suggestion is supported by findings that stable end-products of NO are
increased in the placental but not the maternal circulation in PE and
FGR (51
, 52)
. These findings suggest that the decrease in
HO is not part of a general up-regulation of vasodilators in an attempt
to increase perfusion and that HO has a more fundamental role in the
pathophysiology of PE and FGR.
The Western blotting and immunohistochemical experiments in this study
support our previous findings that HO-1 is expressed at only low levels
in the human placenta. HO-1 can be induced by an extraordinary array of
stimuli, including oxidative stress, hypoxia, and hyperoxia
(24
25
26
27
28
29
30)
. The actions of HO-1 rid cells of pro-oxidants,
allowing cells to withstand further exposure to harmful stimuli. There
is good evidence for increased oxidative stress in PE (18
, 53)
. Kingdom and Kaufmann have suggested that the placenta may
also become relatively hypoxic or hyperoxic in pathological pregnancies
depending on several factors (54)
. They have suggested
that reduced uteroplacental blood flow will result in reduced entry of
oxygen into the intervillous space, leading to intraplacental hypoxia
when fetoplacental blood flow and oxygen extraction by the fetus are
normal (e.g., PE at term); however, when uteroplacental and
fetoplacental blood flow are abnormal, oxygen extraction may be lower
than the entry of oxygen, thus resulting in intraplacental hyperoxia
(severe FGR with absent end-diastolic flow velocity). However, we found
no evidence of alterations in HO-1 expression in FGR and PE, although
this does not mean that these samples were not hypoxic or hyperoxic. It
is also possible that the degree of oxygen imbalance and/or oxidative
stress in these cases was insufficient to stimulate expression of HO-1.
However, this seems unlikely in view of the severity of the placental
disease. A second (unlikely) possibility is that there is sufficient
HO-2 expression to cope with increased oxidative stress. It is also
possible that alternative existing protective mechanisms to deal with
oxidative stress in the placenta are sufficient to cope with the
stress, although the evidence for this is not compelling
(53)
.
A wide spectrum of obstetric disorders including PE, FGR, and
miscarriage has been reported to be associated with reduced
modification of maternal spiral arteries by invasive trophoblast. The
reduced uteroplacental blood flow that accompanies PE and FGR can be
demonstrated using Doppler ultrasound techniques (17
, 50)
.
Thus, understanding the control of trophoblast invasion is an area of
major physiological importance, with potential implications for failed
pregnancy. In PE, normal transformation of the myometrial spiral
arteries fails (15)
. Less is known about FGR in the
absence of maternal hypertension. Uteroplacental blood flow is reduced
in PE and FGR (55)
, and there is extensive evidence that
this reduction is associated with reduced modification of maternal
spiral arteries by invasive CTBs. Abnormal uterine artery Doppler
waveforms predict subsequent PE and FGR, and the finding of absent
physiological change in myometrial vessels is more often noted in PE
and FGR cases with an abnormally high uterine artery PI (56
, 57)
. In PE, complete physiological change is present in <20%
of myometrial spiral arteries (58
, 59)
. This is consistent
with the findings on our own collection of specimens in which 73% of
the myometrial vessels examined from women with PE had either
completely intact or partially disrupted muscle. Less is known about
the placental bed in FGR. Absence of physiological change has been
reported in 45–100% of cases of isolated FGR (7
, 60
, 61)
. These studies have defined FGR according to birth weight,
typically below the 10th centile (59)
. The criteria used
to select growth-restricted fetuses for inclusion in the present study
were much stricter, and the birth weight and extent of umbilical artery
Doppler abnormalities attest to the severity of placental disease in
this group. In our collection of FGR specimens, 45% of myometrial
vessels demonstrated completely intact or partially disrupted muscle
This result is in keeping with the data of Gerretsen et al.
(62)
who showed that absence of physiological change in
myometrial arteries was more likely in severely small infants (birth
weight < the 2.3rd centile), which are more likely to be growth
restricted, than in those with birth weights between the 2.3 and 10th
centiles.
It has been suggested that some maternal vessels undergo morphological
changes without interaction with CTBs (9)
. The possibility
arises that local vasoactive mediators such as NO or CO result in
spiral artery dilatation prior to their invasion. Our previous studies
have shown that extravillous CTBs do not express eNOS or inducible NOS
at any time during invasion (38)
, suggesting that
trophoblast-derived NO is unlikely to contribute to spiral artery
dilatation. In contrast, we have confirmed that extravillous CTBs
within the placental bed express HO-2 (10)
, and CO
produced by these cells could have contributed to the initial spiral
artery changes. However, no changes in expression of HO-2 were found in
PE or FGR. Furthermore, these cells continued to remain negative for
HO-1.
Although we confirmed that the placenta contains relatively little HO-1, we were unable to repeat our previous observation that extravillous CTBs express HO-1. Similar findings were present in cases with PE or FGR. From the comprehensive series of experiments we have performed, we can conclude only that the original lot of antibody obtained from StressGen (SPA-895, lot number 708405), which is now no longer available, contained immunoglobulins not present in the current lot number that were able to bind to extravillous trophoblast proteins that are not HO-1. These findings highlight the importance of undertaking Western blot and immunhistochemical studies in positive controls when using a new antibody. Our original findings, that invasive trophoblast may release CO as part of the mechanisms to dilate spiral arteries, still holds true, because these cells express HO-2.
In summary, this is the first study to examine the expression of HO isoforms in placental and placental bed biopsy material in PE and FGR. Although HO-2 trophoblast expression was not altered in pathological pregnancies, the reduced expression of this isoform in endothelial cells of the placenta in both PE and FGR may contribute to the pathophysiology of these disorders. We speculate that reduced expression of HO-2 on placental endothelial cells in PE and FGR could therefore lead to increased vascular resistance via reduced CO production. Finally, although our data are consistent with a role for HO is trophoblast invasion, there is no evidence of altered HO expression in PE or FGR.
|
ACKNOWLEDGMENTS |
|---|
Received for publication August 16, 2000.
Revision received November 20, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||
