Molecular Human Reproduction vol.3 no.11 pp. 995–1003, 1997
Changes in expression of the nitric oxide synthase isoforms in rat
uterus and cervix during pregnancy and parturition
Mariam Ali1, Irina Buhimschi1, Kristof Chwalisz2 and Robert E.Garfield1,3
1The University of Texas Medical Branch, Department of Obstetrics and Gynecology, Division of Reproductive Sciences,
301 University Boulevard, Galveston, TX 77555-1062, and 2Research Laboratories, Schering AG, Berlin, Germany
3To
whom correspondence should be addressed
Nitric oxide (NO) is considered to be an important local mediator that suppresses uterine contractility in rats
and rabbits during pregnancy until term. The aim of this study was to investigate the mRNA concentrations
for the three isoforms of nitric oxide synthase (NOS) in rat uterus and cervix and to determine whether
alterations occur in association with labour at term or preterm. RNA was isolated from full thickness uterine
and cervical tissues from pregnant rats at various times during gestation, during labour at term or preterm
and post partum. RNA was analysed using reverse transcription–polymerase chain reaction (RT–PCR) with a
single set of amplimers specifically designed to detect all three isoforms of NOS. Three distinct PCR products
were detected which corresponded to the expected sizes for endothelial (e)NOS, neuronal (b)NOS and
inducible (i)NOS products (805, 521 and 428 bp respectively). In all tissues, the 428 bp product predominated
and sequence analysis revealed this to be iNOS mRNA with a very close homology (97%) to the published
sequence of rat iNOS. Densitometric analysis showed that uterine iNOS mRNA was increased during
pregnancy, decreased on day 22 before labour and decreased further during labour at term. In contrast,
cervical iNOS mRNA was low until delivery (day 22) when it increased and was dramatically elevated during
labour. Similarly, 3 h after injection with the antiprogestin onapristone, iNOS mRNA was significantly
decreased in the uterus (~45%) and increased in the cervix (~245%) when compared with controls. The
mRNAs to bNOS and eNOS (corresponding to the 521 and 805 bp bands) were generally greatly reduced in
quantity compared with the 428 bp product. The changes in these constitutive isoforms during gestation
were minor compared with those in the inducible isoform. We conclude that the iNOS transcript is the most
abundant NOS mRNA in the uterus as well as in the cervix and this probably indicates that the inducible
NOS is the main isoform present in these tissues. The changes in iNOS mRNA at the end of pregnancy may
play a role in the initiation of term labour and cervical ripening. Furthermore, the changes in expression of
iNOS can be mimicked during preterm labour following antiprogesterone treatment, and may suggest that
progesterone differentially controls the expression of iNOS in the uterus and cervix.
Key words: labour/mRNA/nitric oxide/pregnancy/uterus
Introduction
Three different types of nitric oxide synthase (NOS) isoforms
have been identified and cloned to date: neuronal (bNOS),
endothelial (eNOS) and cytokine-inducible NOS (iNOS). The
C-terminal portions of all NOS isoforms show homology to
NADPH cytochrome P-450 reductase (Bredt et al., 1991).
Consensus sequences for NADPH, FAD, and FMN binding
sites were also identified. The sequences are highly conserved
across species for each isoform, but the overall sequence
identity between any two isoforms is modest (Xie et al., 1992).
Major differences in regulation as well as in the subcellular
location have been described between the NOS isoforms. For
instance, eNOS and bNOS, which are constitutive isoforms,
are also calcium-calmodulin-dependant, while iNOS can only
be identified following exposure to inflammatory cytokines or
lipopolysaccharide (LPS) (Griffith and Stuehr, 1995). Moreover, bNOS and iNOS sequences are devoid of membraneassociated elements and are found mainly in the soluble portion
© European Society for Human Reproduction and Embryology
of tissue homogenates. On the other hand eNOS is largely
associated with endothelial cell membranes due to both a
myristoylation (Nishida et al., 1992) and a reversible palmitoylation site (Robinson et al., 1995).
Functional and morphological studies on the nitric oxide
system in the rat uterus to date show that: (i) the substrate and
donors of nitric oxide produce uterine relaxation (Izumi et al.,
1993; Yallampalli et al., 1994); (ii) inhibitors of the NO–
cGMP pathway block the relaxation responses (Yallampalli
et al., 1994); (iii) NO synthase is localized to several uterine
cell types (i.e. myometrial smooth muscle cells, uterine epithelial cells, endometrial stromal cells and mast cells) (Huang
et al., 1995); (iv) nitric oxide is produced by the uterus
during periods when L-arginine is consumed and citrulline
concentrations increase (Sladek et al., 1993; Yallampalli et al.,
1994); (v) effects of NO substrate on relaxation are mimicked
by cGMP (Izumi et al., 1993); (vi) gestational differences in
the response to NO have been described suggesting that during
pregnancy the relaxing effect of NO is enhanced, compared
995
M.Ali et al.
with the uterus during term or preterm labour (Izumi et al.,
1993; Natuzzi et al., 1993; Weiner et al., 1994); (vii) the NO
synthase isoforms are present in the rat uterus and upregulated during pregnancy but decreased when labour begins
(Buhimschi et al., 1996). These provide strong evidence that
NO controls uterine contractility during pregnancy and that a
decrease in NO production or responsiveness to NO in the
uterus at term may facilitate labour (Izumi et al., 1993; Natuzzi
et al., 1993; Buhimschi et al., 1996).
Recent studies from our group using Western blots with
monoclonal antibodies identified the presence of all three
isoforms in the pregnant rat cervix and only iNOS and eNOS
in the pregnant rat uterus. Moreover it was demonstrated that
the down-regulation in the NO production in the uterus during
term and preterm labour is accompanied by an up-regulation
of the NO generation in the cervix which may be involved
in connective tissue remodelling during cervical ripening
(Buhimschi et al., 1996). The cervix is composed mostly of
connective tissue and there is a decrease in the collagen
concentration prior to birth (Granstrom et al., 1989) associated
with elevated tissue collagenase activity (Rajabi et al., 1988).
The exact mechanisms responsible for initiating these changes
in the cervix associated with birth are unknown. However, it
is well known that this process is similar to an inflammatory
response at least in the human cervix (Junqueira et al., 1980).
Since inflammatory cells such as macrophages are also known
to be rich in iNOS, it is possible that in the cervix these cells
regulate NO synthesis. Evidence to support this concept was
found using LPS, a known inducer of NO synthesis in
inflammatory cells, which increased NO production in the
pregnant rat cervix (Buhimschi et al., 1996).
The objectives of the present study were to examine:
(i) differences in mRNA expression of different NOS isoforms
in uterine and cervical tissues from pregnant rats at various
times of pregnancy versus labour; (ii) the relative abundance
of NOS isoforms in these tissues by using a single set of
primers to detect all three isoforms of NOS but not the P-450
sequence; and (iii) whether the mRNA expression of NOS
enzyme can be altered by antiprogesterone, known to induce
preterm labour.
Materials and methods
Animals and treatments
Timed pregnant (weighing 250–300 g) rats received in the animal
care unit from Harlan Sprague–Dawley (Houston, TX, USA), were
housed separately and allowed free access to food and water. The
animals were maintained on a constant 12 h light:12 h dark cycle.
The pregnant rats used in these studies have a gestational period of
22 days (day 0 5 day sperm plug observed). Five to seven pregnant
rats at each time frame were used at different days of gestation (i.e.
days 16, 18, 20, 21) as well as on day 22, the morning of delivery,
during spontaneous term delivery on day 22 and 1 and 3 days postpartum. Animals were killed by CO2 inhalation and all procedures
were approved by the Animal Care and Use Committee of the
University of Texas Medical Branch.
Rats at day 18 of gestation were also injected s.c. under the back
skin with 10 mg of the progesterone receptor antagonist onapristone
(ZK 98,299; Schering, Berlin, Germany) suspended in castor oil and
996
benzyle benzoate (1:1 mixture). Animals were killed at 3, 8, 24 h
after injection of onapristone (four animals at each time point) or
vehicle (controls: three animals at each time point). At 24 h the
animals were either bleeding or had already expelled 1–2 pups.
Tissue preparation
Uterine horns were removed, placed in ice-cold phosphate-buffered
saline (PBS) solution, and opened along the line of mesometrial
attachment. The tissues were rinsed repeatedly to remove all traces
of blood. The cervix, defined as the less vascular tissue with parallel
lumina between the uterine horns and the vagina (Harkness and
Harkness, 1959), was isolated and dissected from the surrounding fat
and large blood vessels. Whole rat brain was also collected, washed
thoroughly in PBS and processed along with the other tissue samples
being used as positive control for all NOS isoforms.
Rat alveolar macrophages were obtained from lungs of anaesthesized animals by lavage using a 15 gauge stub adapter needle inserted
in the trachea. The rats were pretreated 6 h before death with a single
injection of either 400 mg/kg LPS from Escherichia coli (Sigma, St
Louis, MO, USA) or saline (control group). Three alveolar rinses
with 10 ml ice-cold PBS each were performed. The solutions were
mixed together and centrifuged at 1000 g for 15 min at 4°C. The
supernatant was decanted from the pellet which was further collected.
A RAW 264.7 (mouse monocyte macrophage, TIB 71; ATCC,
Rockville, MD, USA) cell line, used as positive control for iNOS,
was grown in 90% minimal essential medium 1 10% fetal bovine
serum (Gibco, Grand Island, NY, USA) in a humidified CO2 incubator
with 95% air:5% CO2 at 37°C in the presence or absence of LPS
(1 mg/ml for 11 h). Bovine pulmonary artery endothelial cells (ATCC
1733; ATCC), an established cell line, were used as positive controls
for eNOS.
All tissues and cells were immediately frozen in liquid nitrogen,
and stored at –70°C until used. Total cellular RNA was isolated and
purified by the acid–guanidinium thiocyanate–phenol–chloroform
method (Chomczynski and Sacchi, 1987) and stored frozen as an
ethanol precipitate at –70°C. All RNA preparations were routinely
tested for 28S and 18S RNA using agarose/formaldehyde gels. DNA
contamination was eliminated by treating the RNA with ribonuclease
(RNase)-free deoxyribonuclease (Stratagene, La Jolla, CA, USA).
Primer design
A set of specific amplimers was designed to recognize a common
region on the different known bNOS (Bredt et al., 1991), eNOS
(Nishida et al., 1992) and iNOS (Woods et al., 1993) sequences but
not on cytochrome P-450. The sense primer nucleotide sequence
corresponded to position 724–744 and the antisense primer complementary to position 1129–1149 of rat iNOS. The primers gave single
products corresponding to the three different isoforms of NOS, (i.e.
iNOS, bNOS and eNOS) (Figure 1).
RT–PCR, gel electrophoresis and quantitative analysis of
the PCR products
To analyse different samples of RNA for mRNA NOS transcripts,
we performed RT–PCR using a Gene Amp RNA PCR Kit (PerkinElmer/Cetus, Norwalk, CT, USA). For the first strand cDNA synthesis,
1 mg total cellular RNA was reverse transcribed according to the
manufacturers recommendations. The first strand was subsequently
amplified in the same tube with a Gene Amp PCR System 9600
(Perkin-Elmer/Cetus). PCR was performed with the following conditions: denaturation at 94°C for 15 s, annealing at 60°C for 30 s, and
polymerization at 72°C for 2 min. In all, 28 cycles of amplification
were performed in all experiments as this was determined as optimal
(see below). A portion (15 ml) of the PCR mixture was electrophoresed
Expression of NOS isoforms during pregnancy and parturition
Figure 1. Polymerase chain reaction (PCR) products detected in
representative samples of uterus on day 18 (Lane 1 5 Ut) and
cervix on day 22 of gestation (Lane 2 5 Cx) Note the alignment of
the products with those detected in the tissues used as positive
controls (Lanes 3–5) Lane 3 5 mRNA from rat brain (B); Lane
4 5 mRNA extracted from RAW 264.7 mouse macrophages grown
in culture (MF); Lane 5 5 mRNA from bovine pulmonary artery
endothelial cells (E); Lane 6 5 PCR control (NRT: no reverse
transcriptase); Lane 7 5 PCR control (PC: no RNA added). M 5
molecular size markers.
on 4% polyacrylamide gel. The gel was stained with ethidium bromide
and photographed and 10 ml of double-stranded DNA marker
(BioMarker Low Standards, BioVenture, Murfreesboro, TN, USA)
was also loaded onto each gel. The optical density (OD) of the bands
obtained from uterus, cervix, rat alveolar or RAW 264.7 macrophages
mRNA were densitometrically scanned and analysed using a Dekmate
III scanner and PDI 1D software package (PDI, Huntingstation, NY,
USA). The amounts were normalized against the optical density of
the 400 bp band from the marker lane within the same gel in order
to make intergel comparisons.
Linear regression analysis was used to establish the efficiency of
the PCR reaction parameters used. The optical density (OD) of the
iNOS band was quantified and expressed as a function of either the
initial mRNA (for control of the RT reaction) (Figure 1A and B) or
the number of cycles (for control of the amplification reaction) (Figure
1A and C) to exclude the plateau effect (Nakayama at al., 1992;
Tezuka et al., 1995).
Direct sequencing of PCR products
To prepare for sequencing of iNOS the corresponding band was
excised from the polyacrylamide gel and re-amplified for 30 cycles
of PCR (above conditions). Products were separated on 1.2% agarose
gel, excised and then purified using a Geneclean II kit (Bio 101 Inc,
La Jolla, CA, USA). Direct sequencing of purified PCR products was
performed using a Taq Dideoxy Terminator cycle sequencing kit
(Applied Biosystems Inc, Foster City, CA, USA).
Statistical analysis
All results were expressed as the mean 6 SEM of between five and
seven rats in each group. Statistical comparisons between groups
were performed using one-way or two-way analysis of variance
followed by post-hoc tests utilizing the Tukey–Kramer multiple
comparison test. A two-tailed P , 0.05 was considered to be
statistically significant.
Results
Three PCR products were detected in all uterine and cervical
samples corresponding to the expected molecular sizes of the
eNOS (805 bp), bNOS (521 bp) and iNOS (428 bp). The
bands (after 28 PCR cycles), detected in uterus on day 18 of
gestation, in cervix on day 22 versus different tissues or cells
used as positive controls are shown in Figure 1. After 28 PCR
amplification cycles the most abundant product was iNOS.
The other two products (i.e. b and eNOS) were present in
lesser amounts. However, further amplification beyond 34
cycles increased the abundance of eNOS and bNOS, but
decreased the abundance of iNOS (see below). Additional
amplification (beyond 34 cycles) masked the relative differences between products due to the plateau effect which was
reached for the most abundant product (iNOS) but not for
those present in reduced amounts (b and eNOS).
Rat alveolar macrophages harvested from animals as well
as RAW 264.7 mouse macrophages grown in culture exhibited
only one band of 428 bp (iNOS). This product was present in
both LPS-treated and untreated cells. However, both in-vivo
injection of LPS and treatment of macrophages in vitro with
LPS markedly increased the density of the iNOS band (data
not shown). The iNOS, eNOS and bNOS bands from uterus
and cervix were positioned in alignment with the bands of
iNOS from macrophages, the iNOS, eNOS and bNOS from
brain and eNOS of endothelial cells. These results confirm
that the products correspond to the various NOS isoforms.
Validation of the PCR conditions
In order to perform quantitative PCR without the use of
internal standards, two conditions had to be met: (i) the
variation in efficiency of the amplification reaction between
tubes and experiments performed in similar conditions must
be minimal and (ii) the reactions should not reach the plateau
phase. In Figure 2A increasing amounts of initial mRNA were
amplified at 26, 28 and 30 cycles. Identical samples were
also amplified in separate tubes to examine the tube-to-tube
variation. This experiment was repeated three times. Both the
tube-to-tube variation within the same PCR reaction and the
inter-experimental variation were found to be ,10%. Figure
2B shows the linear relationship between the number of
amplified target molecules (expressed as the OD of the iNOS
band) and the initial number of targeted molecules (mRNA)
for the three cycle numbers presented. Figure 2C shows the
amplified target molecules as a function of the number of PCR
cycles. The linear and parallel relationship shows that the
plateau phase is not reached in the conditions presented above.
To compare target amounts in different samples, all our initial
PCR mixtures contained 1 mg total mRNA and the data were
collected after 28 PCR cycles. It should be noted that these
conditions have been validated for the 428 bp product (iNOS)
and do not seem to be optimum for the other two isoforms
where further amplification seems necessary for a more abundant product and more intense bands. However, further amplification would result in erroneous conclusions regarding the
relative abundance of the products within each sample.
The above results demonstrate that the PCR conditions as
described in the Materials and Methods section and discussed
above are valid for quantification of the 428 bp product in
different samples and for comparing the relative abundance of
the three products within each sample. However quantitative
determination of the 805 bp (eNOS) and 521 bp (bNOS)
997
M.Ali et al.
partum. Thus the most abundant NOS mRNA in the uterus is
that of the inducible isoform and it is dramatically downregulated prior and during labour at term.
NOS mRNA abundance in rat cervix
All three NOS isoforms were also detected in the rat cervix
(Figure 5A). Low amounts of products were recorded from
days 16–21 followed by an increase in iNOS mRNA in the
cervices from labouring animals (Figure 5B). Intermediate but
still statistically different values were observed at term before
labour (d22NL). bNOS values were low throughout gestation
in the cervix but rose during labour. However, eNOS mRNA
values were very low and generally below the limits of
reasonable detectability by densitometry.
Figure 2. Reverse transcription–polymerase chain reaction (RT–
PCR) signals as a function of the amount of total RNA and number
of PCR cycles. RT–PCR was performed according to conditions
described in the Materials and methods section. One tissue sample
from the uterus of an animal at day 16 of gestation is presented.
(A) Photograph of a gel stained with ethidum bromide after adding
various amounts of total RNA (2, 1 and 0.5 mg) at 26, 28 and 30
cycles. MF 5 mRNA from RAW 264.7 mouse macrophage cells;
M 5 molecular size markers. (B) Relative amounts of iNOS
products versus the number of PCR cycles for different amounts of
RNA added for each RT–PCR reaction. OD 5 optical density.
(C) Relative amounts of iNOS product as a function of the initial
amount of RNA illustrated for different numbers of PCR cycles.
Note the linear relationship between these parameters.
iNOS mRNA abundance in rat uterus and cervix after
administration of antiprogesterone
The time course of effect of the antiprogestin onapristone on
mRNA concentrations for iNOS in the uterus (Panel A) and
cervix (panel B) is shown in Figure 6. In the uterus (Panel C),
a significant decrease was recorded at 3 h (63%) after injection
compared with control animals. This was followed by a gradual
increase reaching 85% of the control animals at 8 h. At
24 h, at the time when most treated animals were delivering
prematurely, iNOS mRNA in the onapristone-injected animals
was not statistically different from the vehicle-injected rats.
In the cervix (Figure 6D) administration of onapristone
resulted in a significant (~3–4-fold) increase compared with
controls at 3 h and sustained induction of iNOS mRNA.
Therefore, the antiprogestin decreased iNOS mRNA in the
uterus while increasing it in the cervix. Low values of
both eNOS and bNOS were detected in all cervical samples
irrespective of treatment and time frames (data not shown).
Discussion
bands may suffer from the limits of the detection techniques
(densitometry) due to low intensity.
Identity of the products with NOS isoforms
Figure 3 represents the alignment window of the purified
428 bp product from the rat uterus (top sequence) compared
with the published sequence of rat iNOS (Woods et al., 1993).
There is 97% nucleotide identity between the two.
NOS mRNA abundance in rat uterus
In all samples at all gestational ages investigated, the most
abundant product in the rat uterus was the iNOS mRNA
(Figure 4A). However, quantitative analysis during gestation
showed that on day 22 both before and during labour iNOS
values were significantly lower compared with those recorded
between days 16 and 21 (Figure 4B). A significant increase
in iNOS mRNA occurred post partum to values not statistically
different from those recorded from days 16–21. The relatively
abrupt decrease in the iNOS mRNA at the end of gestation
contrasts with the gradual decrease observed in bNOS mRNA.
eNOS values, although detectable, were low at all the times
explored and constant throughout gestation, labour and post
998
This study shows that: (i) mRNA for the three isoforms for
nitric oxide synthase (iNOS, eNOS and bNOS) can be detected
in the rat uterus and cervix during pregnancy and parturition;
(ii) in the uterus the predominant NOS mRNA is that coding
for the inducible isoform and this is significantly decreased
during labour at term; (iii) in the cervix NOS expression for
all isoforms is low during pregnancy and iNOS dramatically
increases on day 22 before and during labour; (iv) these
contrasting changes in iNOS expression in the uterus compared
with the cervix can be induced by administration of an
antiprogesterone prior to preterm birth. These results are
consistent with the hypothesis that NO production in the uterus
during pregnancy is elevated to assist in suppressing uterine
contractility and maintaining uterine quiescence while its
withdrawal prior to term or preterm might play an important
role in initiating term or preterm labour. Moreover, these
studies suggest that progesterone and progesterone withdrawal
at term may differentially regulate the expression of iNOS
levels in the uterus and cervix and thereby control NO levels
in these organs.
The L-arginine–NO–cGMP pathway has been identified in
the uterus in several species: rat (Izumi et al., 1993; Yallampalli
Expression of NOS isoforms during pregnancy and parturition
Figure 3. Alignment window of the 428 bp product purified from the rat uterus (lower sequence) compared to the published sequence of rat
iNOS (Woods et al., 1993). The sequence homology is 97%.
et al., 1994), guinea pig (Chwalisz et al., 1994a), rabbit (Sladek
et al., 1993) and humans (Buhimschi et al., 1995a). Gestational
differences in the presence of NOS isoforms (Buhimschi et al.,
1996) and response to NO have also been described suggesting
that during pregnancy the relaxing effect of NO is enhanced,
compared with the uterus during term or preterm labour (Izumi
et al., 1993; Natuzzi et al., 1993; Weiner et al., 1994).
Additionally, iNOS is present in the placenta where, like the
uterus, a prominent withdrawal occurs at the end of pregnancy
(Purcell et al., 1997).
There appear to be contradictions in the literature over what
isoforms of NOS are present in the uterus. However, different
species have been examined. Constitutive NOS activity has
been primarily described in the sheep (Figueroa and Massmann,
1995) and guinea-pig uterus (Weiner et al., 1994), and regulated
by oestrogen and not progesterone. Moreover these studies
suggest that the increase in NOS activity during pregnancy or
after oestrogen treatment is due to increased expression of
eNOS and bNOS and not iNOS (Weiner et al., 1994). Purification of NOS from the non-pregnant rat uterus identified only
a calmodulin-dependent 155 kDa product which was attributed
to bNOS (Jaing et al., 1996). Immunohistochemically the
enzyme was not confined to neurones of the uterus but was
widely distributed to several non-neural cells (Schmidt et al.,
1992). Another study, however, identified iNOS mRNA in the
rat uterine smooth muscle which was increased after exposure
to LPS (Nakaya et al., 1996). Previously we investigated the
presence and gestational differences in NOS isoforms in the
rat uterus versus cervix and concluded that only iNOS and
eNOS were present in the uterus during pregnancy (Buhimschi
et al., 1996). Although we identified bNOS in the non-pregnant
rat uterus (unpublished results), during pregnancy its levels
were virtually absent (Buhimschi et al., 1996). In the cervix
all three isoenzymes were present but only iNOS and/or
bNOS expression was up-regulated during labour. eNOS levels
remained unchanged. However, since these studies used
Western blotting and different monoclonal antibodies with
different affinities, comparisons between abundance of the
isoforms within the same sample were not feasible. For this
purpose, the present study explored mRNA values of all NOS
isoforms with the same set of primers. The results not only
confirm our previous findings on protein levels but also
reinforce the conclusion that the isoform that is most abundant
and probably most important in the production of NO within
the rat uterus during gestation is the inducible isoform (iNOS)
and that this is also the isoform that is abruptly induced in the
cervix at the time of parturition. The advantage of using a set
of primers that detects a common region of the three NOS
isoforms is that it allows for a comparison between abundance
of the isoforms. However due to great differences in abundance
the plateau in the PCR reaction for iNOS is reached before
optimum amplification of the other two products occurs.
Therefore the variation of eNOS and bNOS during different
gestational days, labour and post partum should be interpretated
with some caution.
Interestingly, although we detected an absolute increase in
iNOS mRNA in the uterus during pregnancy, our studies on
iNOS using monoclonal antibodies against iNOS did not show
a statistically significant increase in iNOS levels. However, a
difference in banding pattern was observed between late
gestation and during labour (Buhimschi et al., 1996). Similarly,
a previous study reported a discrepancy between gene transcription and iNOS mRNA levels in cultured macrophages. Moreover this study detected a dual, both transcriptional and
999
M.Ali et al.
Figure 4. Representative reverse transcription–polymerase chain
reaction (RT–PCR) products of NOS isoforms mRNA from rat
uterus during pregnancy, term and post partum. (A) Lanes 1–7
show products from days 16–22 (morning of the delivery day,
22NL), day 22 during labour (22L), 1 and 3 days post partum (pp1,
pp3); Lane 8 5 RNA from RAW 264.7 mouse macrophages (MF);
Lane 9 5 PCR control (PC: no RNA added); Lane 10 5 Molecular
size markers. (B) Quantitative data based on densitometric scanning
of iNOS and bNOS at various times of gestation and postpartum.
There is also a significant difference (P , 0.001) in the abundance
of the two products (–j– 5 iNOS; –d– 5 bNOS) (two-way
analysis of variance). Statistical significance from multiple
comparisons (Tukey–Kramer tests) among different days is shown
for iNOS mRNA and bNOS mRNA. No gestational differences in
bNOS abundance were observed. Means with a common
superscript are not significantly different at a value of P , 0.05
For points lacking a SE bar, error was smaller than or equal to the
size of symbol (n 5 5–7 animals for each gestational point). The
eNOS band was too low to be detected by densitometry.
post-transcriptional, regulation of the enzyme (Weisz et al.,
1994). In our case, one possibility regarding the expression of
iNOS in the uterus is that at term, although iNOS gene
transcription ceases, the protein remains present but its activity
is abruptly decreased by post-translational modifications. Such
changes have been described for iNOS in mouse macrophages
(Forstermann and Kleiert, 1995). The reverse changes in iNOS
mRNA levels from the uterus and the cervix might also be
due to changes in mRNA stability within the cervix since this
phenomenon has been demonstrated in cultured macrophages:
LPS and not g-interferon prolongs the iNOS mRNA half-life
from 1 to 4–6 h (Weisz et al., 1994). Also feasible is that the
1000
Figure 5. Representative reverse transcription–polymerase chain
reaction (RT–PCR) products of NOS mRNA isoforms from rat
cervix during late pregnancy, term and postpartum. (A) Lanes 1–7
show products from days 16–22, (morning of the delivery day
22NL), day 22 during labour (22L), one and three days postpartum
(pp1, pp3); Lane 8 5 RNA from rat brain; Lane 9 5 RNA from
bovine pulmonary artery cell line (CRL); Lane 10 5 RNA from
RAW 264.7 mouse macrophages (MF); Lane 11 5 PCR control (no
RNA); Lane 12 5 molecular size markers. (B) Quantitative data
based on densitometric scanning of bands at various times of
gestation and postpartum. There is a significant difference
(P ,0.001, two-way analysis of variance) in the abundance of
iNOS mRNA compared with the other two products (–j– 5 iNOS;
–d– 5 bNOS) eNOS was not quantified due to the low amount.
Statistical significance from multiple comparisons (Tukey–Kramer
tests) among different days is shown for iNOS mRNA and bNOS
mRNA. No gestational differences in eNOS abundance were
observed. Means with a common superscript are not significantly
different at a value of P , 0.05. For points lacking a SE bar, error
was smaller than or equal to the size of symbol (n 5 5–7 animals
for each gestational point).
uterine iNOS is ‘different’ from the cervical protein, but its
mRNA contains the common region recognized by the primers
used in this study. This hypothesis is supported by the finding
that the band on Western blots in our previous studies was
similar for cervix and cultured macrophages, but different
from the uterus (Buhimschi et al., 1996). These two forms
might exhibit differential regulation. A similar mechanism of
regulation has been suggested for the neuronal isoform, where
alternative splicing produces a novel variant present only in
the striated muscle and heart and not in the nervous system
(bNOSm) (Silvagno et al., 1996).
Expression of NOS isoforms during pregnancy and parturition
Figure 6. (A) Lanes 1–9 show representative reverse transcription–
polymerase chain reaction (RT–PCR) products of the iNOS mRNA
from rat uterus on day 18 of gestation after 3, 8 and 24 h of an
injection of 10 mg onapristone (ZK) or vehicle (C); Lane 10 5
RNA from RAW 264.7 mouse macrophages (MF); Lane 11 5
molecular size markers. (B) Lanes 1–9: Representative RT–PCR
products of the iNOS mRNA from rat cervix on day 19 of
gestation at 3, 8 and 24 h after the injection of 10 mg onapristone
(ZK) or vehicle (C); Lane 10 5 RNA from RAW 264.7 mouse
macrophages (MF); Lane 11 5 molecular size markers.
Quantitative data based on densitometric scanning of the iNOS
mRNA band in the rat uterus (C) and cervix (D) following
antiprogesterone (ZK) treatment. Data is expressed as a percentage
of the average value obtained from the control animals at each
corresponding time (3, 8 or 24 h) Statistical analysis was performed
using one-way analysis of variance followed by multiple
comparisons using the Tukey–Kramer test. Means with different
superscripts (uterus: lowercase; cervix: uppercase) are significantly
different at a value of P , 0.001. For points lacking a SE bar,
error was smaller than or equal to the size of symbol (n 53–4
animals for each point).
It has been long recognized that uterine contractility is
influenced by female sex steroids during pregnancy as well as
in the non-pregnant state. In many species, including rat, the
myometrium is relatively quiescent during pregnancy when
progesterone concentrations are elevated. At term when progesterone values decline and oestrogen values rise, myometrial
activity increases and eventually evolves to labour (Garfield
et al., 1988). It has been shown that sex steroids play an
important role in the process of preparation of the uterus for
parturition by inducing the formation of gap junctions in
the myometrium (Garfield et al., 1977), oxytocin receptors
(Alexandrova and Soloff, 1980) or increasing prostaglandin
metabolism (Olson et al., 1983). It is possible that the changes
seen in the NO system during spontaneous and preterm labour
are also necessary for the labour and delivery process. It is
possible that iNOS mRNA production is controlled by an (as
yet unidentified) paracrine regulator which is decreasing in the
uterus at the same time as it is increasing in the cervix (or
vice versa). The present study confirms the hypothesis that
progesterone, acting also as gene inducer, can modulate the
NO system since administration of antiprogesterone to the
animals resulted in changes in NOS mRNA levels in the uterus
or cervix similar to those seen during term labour (Figures 4
to 6). Previous studies showed that the ability of L-arginine
and cGMP to relax tissues in vitro diminishes after progesterone
withdrawal both spontaneously at term and after onapristone
treatment (Yallampalli et al., 1994). However, whether the
same hormonal milieu can induce opposite changes in the
uterus in comparison with the cervix is still an open question,
although differential regional regulation of the iNOS isoforms
has been also described in the digestive tract (Chen et al.,
1996). Onapristone treatment induced dissociation of collagen
fibres and increased collagenolysis followed by dramatic
softening of the cervix (Chwalisz et al., 1991). These cervical
changes occur prior to the onset of preterm labour suggesting
that progesterone modulates cervical ripening independently
from its actions on the myometrium. The underlying processes
are not yet well understood but prostaglandins do not seem to
be the key mediators (Chwalisz et al., 1991; Radestad and
Bygdeman, 1993). In this study, we only examined the effects
of an antiprogestin and infer that progesterone is involved in
NOS expression. However, other steroids may also play roles
in the uterus and cervix.
Cervical ripening is thought to be similar to an inflammatory
reaction with recruitment of neutrophils and monocytes
(Liggins, 1981; Chwalisz et al., 1994b). This influx of immune
cells that occurs in the cervix gradually toward the end of
gestation, may result not only in increased iNOS from these
imported cells, but also generate intrinsic LPS-like mediators
or cytokines that locally regulate the iNOS mRNA (synthesis
or stability). This might also explain the differential expression
of iNOS in the cervix versus the uterus. The increased NO
production in the cervix may have a role in the softening of
the cervix. Consistent with this hypothesis are observations
that NO modulates connective tissue degrading enzymes,
matrix metalloproteinases (MMPs) (Evans et al., 1995;
Koplakov et al., 1995). In the endometrium progesterone
has been shown to inhibit MMPs, an effect antagonized by
antiprogestins (Rodgers et al., 1994). These studies, together
with our present data, suggest that progesterone may indirectly
regulate cervical softening through control of NO synthesis
and action on MMPs.
Another overall issue that needs to be addressed is the
importance of the NO system in regulating uterine contractility
in rat since it has been demonstrated that competitive inhibition
of NOS with NG-nitro-L-arginine methyl esther (L-NAME) in
rats does not change the duration of gestation (Buhimschi
et al., 1995b). Moreover, homozygous mutant mice with
disrupted NOS genes are fertile and without evident abnormalities in their reproductive function (Wei et al., 1995). However,
1001
M.Ali et al.
several studies indicate the importance of synergistic effects
between NO and progesterone. Our studies in rats show that,
although NO inhibition alone is unable to induce preterm
parturition, it markedly increases the labour-inducing activity
of antiprogesterones at doses which are inactive alone
(Yallampalli et al., 1996). It thus appears that NO and progesterone act synergically to maintain pregnancy and at term the
withdrawal of progesterone and NO triggers the onset of labour.
In summary we conclude that firstly, the iNOS transcript is
the most abundant NOS mRNA in the uterus as well as in
cervix indicating that the inducible NOS is the main isoform
present in these tissues and secondly, the decrease of iNOS in
the uterus during labour at term and preterm is accompanied
by a change in the opposite direction in the cervix. These
alterations in iNOS expression may be progesterone dependent
and have a role in the initiation of labour and cervical ripening.
References
Alexandrova, M. and Soloff, M.S. (1980) Oxytocin receptors and parturition.
I. Control of oxytocin receptor concentration in the rat myometrium at
term. Endocrinology, 106, 730–735.
Bredt, D.S., Hwang, P.M., Glatt, C.E. et al. (1991) Cloned and expressed
nitric oxide synthase structurally resembles cytochrome P-450 reductase.
Nature, 351, 714–718.
Buhimschi, I., Yallampalli, C., Dong, I.L. and Garfield, R.E. (1995a)
Involvement of a nitric oxide-cGMP pathway in control of human uterine
contractility during pregnancy. Am. J. Obstet. Gynecol., 172, 1577–1584.
Buhimschi, I., Yallampalli, C., Chwalisz, K. and Garfield, R.E. (1995b)
Preeclampsia-like conditions induced by nitric oxide inhibition: effects of
L-arginine, D-arginine and steroid hormones. Hum. Reprod., 10, 2723–2730.
Buhimschi, I., Ali, M., Jain, V. et al. (1996) Differential regulation of nitric
oxide in the rat uterus and cervix during pregnancy and labour. Hum.
Reprod., 11, 1755–1766.
Chen, K., Inoue, M. and Okada, A. (1996) Expression of inducible nitric
oxide synthase mRNA in rat digestive tissues after endotoxin and its role
in intestinal mucosal injury. Biochem. Biophys. Res. Commun., 224,
703–708.
Chomczynski, P. and Sacchi, N. (1987) Single-step method of RNA isolation
by acid guanidinium thiocyanate–phenol–chloroform extraction. Anal.
Biochem., 162, 156–159.
Chwalisz, K., Hegele-Hartung, C., Schulz, R. et al. (1991) Progesterone
control of cervical ripening – experimental studies with the progesterone
antagonists onapristone, liloprostone and mifepristone. In Leppert, P. and
Woessner, F. (eds), The Extracellular Matrix of the Uterus, Cervix and
Fetal Membranes: Synthesis, Degradation and Hormonal Regulation.
Perinatology Press, Ithaca, NY, USA, pp. 119–131.
Chwalisz,K., Ciesla,I. and Garfield,R.E. (1994a) Inhibition of nitric oxide
(NO) synthesis induces preterm parturition and preeclampsia-like conditions
in guinea pigs. Communication at the Society for Gynecologic Investigation,
41st Annual Meeting, 26 pp.
Chwalisz, K., Benson, M., Scholz, P. et al. (1994b) Cervical ripening with
cytokines interleukin 8, interleukin 1b and tumor necrosis factor a in
guinea-pigs. Hum. Reprod., 9, 2173–2181.
Evans, C.H., Stevanovic-Racic, M. and Lancaster, J. (1995) Nitric oxide and
its role in orthopaedic disease. Clin. Orthop., 312, 275–294.
Figueroa, J.P. and Massmann, G.A. (1995) Estrogen increases nitric oxide
synthase activity in the uterus of nonpregnant sheep. Am. J. Obstet.
Gynecol., 173, 539–545.
Forstermann, U. and Kleinert, H. (1995) Nitric oxide synthase: expression
and expressional control of the three isoforms. Naunyn Schmiedebergs
Arch. Pharmacol., 352, 351–364.
Garfield, R.E., Sims, S. and Daniel, E.E. (1977) Gap junctions: their presence
and necessity in myometrium during parturition. Science, 198, 958–960.
Garfield, R.E., Blennerhassett, M.G. and Miller, S.M. (1988) Control of
myometrial contractility: Role and regulation of gap junctions. Oxford Rev.
Reprod. Biol., 10, 436–490.
Granstrom, L., Ekman, G., Ulmsten, U. and Malmstrom, A. (1989) Changes
1002
in the connective tissue of corpus and cervix uteri during ripening and
labour in term pregnancy. Br. J. Obstet. Gynaecol., 96, 1198–1202.
Griffith, O.W. and Steuhr, D.J. (1995) Nitric oxide synthases: properties and
catalytic mechanism. Ann. Rev. Physiol., 57, 707–736.
Harkness, M.R.L. and Harkness, R.D. (1959) Changes in the physiological
properties of the uterine cervix of the rat during pregnancy. J. Physiol.,
148, 524–547.
Huang, J., Roby, K.F., Pace, J.L. et al. (1995) Cellular localization and
hormonal regulation of inducible nitric oxide synthase in cycling mouse
uterus. J. Leukocyt. Biol., 57, 27–35.
Izumi, H., Yallampalli, C. and Garfield, R.E. (1993) Gestational changes in
L-arginine-induced relaxation of pregnant rat and human myometrial smooth
muscle. Am. J. Obstet. Gynecol., 169, 1327–1337.
Jaing, Y., Singh, A.K., Kannan, M.S. et al. (1996) Isolation of nitric oxide
synthase in the rat uterus. Life Sci., 58, 1009–1014.
Junqueira, L.C.U., Zugaib, M., Montes, G.S. et al. (1980) Morphologic and
histochemical evidence for the occurrence of collagenolysis and for the
role of neutrophilic polymorphonuclear leukocytes during cervical dilation.
Am. J. Obstet. Gynecol., 138, 273–281.
Kolpakov, V., Gordon, D. and Kulik, T.J. (1995) Nitric oxide-generating
compounds inhibit total protein and collagen synthesis in cultured vascular
smooth muscle cells. Circ. Res., 76, 305–309.
Liggins,G.C. (1981) Cervical ripening as an inflammatory reaction. In Ellwood,
D.A. and Anderson, A.M.B. (eds), The Cervix in Pregnancy and Labour,
Clinical and Biochemical Investigation. Churchill Livingstone, Edinburgh,
UK, pp. 1–9.
Nakaya, Y., Yamamoto, S., Hamada, Y. et al. (1996) Inducible nitric oxide
synthase in uterine smooth muscle. Life Sci., 58, PL249–PL255.
Nakayama, H., Yokoi, H. and Fujita, J. (1992) Quantificiation of mRNA by
non-radioactive RT–PCR and CCD imaging system. Nucleic Acids Res.,
20, 4939.
Natuzzi, E.S., Ursell, P.C., Harrision, M. et al. (1993) Nitric oxide synthase
activity in the pregnant uterus decreases at parturition. Biochem. Biophys.
Res. Commun., 194, 1–8.
Nishida, K., Harrision, D.G., Navas, J.P. et al. (1992) Molecular cloning and
characterization of the constitutive bovine aortic endothelial cell nitric
oxide synthase. J. Clin. Invest., 90, 2092–2096.
Olson, D.M., Skinner, K. and Challis, J.R. (1983) Estradiol-17 beta and
2-hydroxyestradiol-17
beta-induced
differential
production
of
prostaglandins by cells dispersed from human intrauterine tissues at
parturition. Prostaglandins, 25, 639–651.
Purcell, T.L., Buhimschi, I.A., Given, R. et al. (1997) Inducible nitric oxide
synthase is present in the rat placenta at the fetal-maternal interface and
decreases prior to labour. Mol. Hum. Reprod., 3, 485–491.
Radestad, A. and Bygdeman, M. (1993) Are prostaglandins mediator of
Mifepristone (RU 486) – induced cervical softening in early pregnancy?
J. Lipid Med., 6, 503–507.
Rajabi, M.R., Dean, D.D., Beydoun, S.N. and Woessner, J.F Jr. (1988)
Elevated tissue levels of collagenase during dilation of uterine cervix in
human parturition. Am. J. Obstet. Gynecol., 159, 971–976.
Robinson, L.J., Busconi, L. and Michel, T. (1995) Agonist-modulated
palmitoylation of endothelial nitric oxide synthase. J. Biol. Chem., 270,
995–998.
Rodgers, W.H., Matrisian, L.M., Giudice, L.C. et al. (1994) Patterns of matrix
metalloproteinase expression in cycling endometrium imply differential
functions and regulation by steroid hormones. J. Clin. Invest., 94, 946–953.
Schmidt, H.H., Gagne, G.D., Nakane, M. et al. (1992) Mapping of neural
nitric oxide synthase in the rat uterus suggests frequent co-localization with
NADPH diaphorase but not with soluble guanylyl cyclase, and novel
paraneural functions for nitrinergic signal transduction. J. Histochem.
Cytochem., 40, 1439–1456.
Sladek, S.M., Regenstein, A.C., Lykins, D. and Roberts, J.M. (1993) Nitric
oxide synthase activity in pregnant rabbit uterus decreases on the last day
of pregnancy. Am. J. Obstet. Gynecol., 169, 1285–1291.
Silvagno, F., Xia,H.H. and Bredt, D.S. (1996) Neuronal nitric-oxide synthasem, an alternatively spliced isoform expressed in differentiated skeletal
muscle. J. Biol. Chem., 271, 11204–11208.
Tezuka, N., Ali, M., Chwalisz, K. and Garfield, R.E. (1995) Changes in
transcripts encoding calcium channel subunits of rat myometrium during
pregnancy. Am. J. Physiol. 269, C1008–C1017.
Xie, Q.W., Cho, H.J., Calycay, J. et al. (1992) Cloning and characterization
of inducible nitric oxide synthase from mouse macrophages. Science, 256,
225–228.
Wei, X.Q., Charles, I.G., Smith, A. et al. (1995) Altered immune responses
in mice lacking inducible nitric oxide synthase. Nature, 375, 408–411.
Expression of NOS isoforms during pregnancy and parturition
Weiner, C.P., Lizasoain, I., Baylis, S.A. et al. (1994) Induction of calciumdependent nitric oxide synthases by sex hormones. Proc. Natl. Acad. Sci.
USA, 91, 5212–5216.
Weisz, A., Oguchi, S., Cicatiello, L. and Esumi, H. (1994) Dual mechanism for
the control of inducible-type NO synthase gene expression in macrophages
during activation by interferon-gamma and bacterial lipopolysaccharide.
Transcriptional and post-transcriptional regulation. J. Biol. Chem., 269,
8324–8333.
Woods, E.R., Berger, H., Sherman, P.A. and Lapetina, E.G. (1993) Hepatocytes
and macrophages express an identical cytokine inducible nitric oxide
synthase gene. Biochem. Biophys. Res. Comm., 191, 767–774.
Yallampalli, C. and Garfield, R.E. (1993) Inhibition of nitric oxide synthesis
in rats during pregnancy produces symptoms identical to preeclampsia.
Am. J. Obstet. Gynecol., 169, 1316–1320.
Yallampalli, C., Izumi, H., Byam-Smith, M. and Garfield, R.E. (1994) An
L-arginine–nitric oxide–cyclic guanosine monophosphate system exists in
the uterus and inhibits contractility during pregnancy. Am. J. Obstet.
Gynecol., 170, 175–185.
Yallampalli, C., Buhimschi, I., Chwalisz, K. et al. (1996) Preterm birth in rats
produced by the synergistic action of a nitric oxide inhibitor (NG-nitro-Larginine methyl ester) and an antiprogestin (onapristone) Am. J. Obstet.
Gynecol., 175, 207–212.
Received on February 17, 1997; accepted on September 4, 1997
1003