| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Departments of Obstetrics and Gynecology (M.A.J., R.R.M., D.A.H., I.M.B.) and Meat and Animal Science (R.R.M.), University of Wisconsin-Madison Medical School, Perinatal Research Laboratories, Madison, Wisconsin 53715
Address all correspondence and requests for reprints to: Ian M. Bird, Ph.D., University of Wisconsin-Madison Medical School, Department of Obstetrics and Gynecology, Perinatal Research Laboratories, Meriter Hospital/Park-7E, 202 South Park Street, Madison, Wisconsin 53715. E-mail: imbird{at}facstaff.wisc.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During pregnancy the levels of the potent vasodilator prostacyclin (PGI2) are increased in systemic plasma, uterine venous blood, and urine (3, 7). We have shown that during ovine pregnancy, the uteroplacental unit is a major source of the elevated circulating concentrations of PGI2 (1). In addition, the potent cyclooxygenase (COX) inhibitor, indomethacin, increases both the systemic pressor effects (8, 9) and the uterine vasoconstrictor responses induced by iv infusions of AII (1, 9). Thus, the pregnancy-associated refractoriness to the vasoconstrictor effects of infused AII may relate to elevated local uterine and systemic levels of vasodilating prostaglandins observed during gestation (1, 3, 7, 8, 9).
PGI2 is the primary prostaglandin produced by the uterine artery endothelial and vascular smooth muscle (VSM) cells (3, 10, 11, 12) and is principally formed by the endothelial cells, with only 30% of total uterine artery prostacyclin production contributed by the VSM (10). We have shown that the in vitro production of PGI2 by uterine artery endothelium from pregnant sheep is also greater than that observed for nonpregnant ewes (10, 12, 13). The molecular level at which these pregnancy-induced increases in PGI2 occur remains unclear, but it may be regulated, in part, via increases in either the activity or expression of the enzyme cyclooxygenase (COX, also called prostaglandin H synthase). This is supported by the proportionally greater PGI2 production by uterine arteries from pregnant vs. nonpregnant ewes when incubated in the presence of excess arachidonic acid, the substrate for COX (10).
To date, two isoforms of COX have been described (14). COX-1 is constitutively expressed in many tissues including pulmonary endothelial cells of fetuses and lambs and may regulate vascular homeostasis via production of prostanoids (15, 16). COX-2 is not constitutively expressed, but it is induced by mitogens, lipopolysaccharides, and human CG in some tissues and is associated with inflammatory responses (17, 18, 19, 20). In the current study, we propose that COX expression is primarily localized to the endothelium rather than VSM and that the pregnancy-induced increase in PGI2 production by uterine arteries is due, in part, to an increase in COX-1 expression in uterine artery endothelial cells. To achieve this, we have determined the cellular localization of COX-1 in the uterine artery cross-sections by immunohistochemistry as well as quantified the expression of COX-1 at the level of protein (Western analysis) and messenger RNA (mRNA) (RT-PCR) in acutely isolated endothelial cell preparations obtained by collagenase dispersion directly from the vessel lumen.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Immunohistochemistry
Immunohistochemical procedures were based on a method previously
reported (21, 22). At the time of euthanasia, intact uterine artery
segments from three nonpregnant and four pregnant (120–130 days’
gestation) sheep were placed in 4% formaldehyde in sodium cacodylate
buffer (0.1 M, pH 7.4), and fixed overnight. Tissues were
sequentially dehydrated in graded ethanol solutions, embedded in
paraffin, cut into sections (6 µm), and mounted on polylysine coated
glass slides. Following deparaffinization and graded rehydration,
sections were incubated in 3% H2O2 in methanol
for 20 min to quench endogenous peroxidase activity. COX-1 antibody
(mouse monoclonal antibody, Cayman Chemical, Ann Arbor, MI) was then
applied (1:2000 dilution). Staining was detected using a biotinylated
secondary antibody in combination with the avidin-biotin-peroxidase
(ABC) method (ELITE ABC kit, Vector Laboratories, Burlingame, CA) with
3,3'-diaminobenzidine (DAB, Vector) as chromagen. As controls, adjacent
sections were alternatively treated with a control mouse IgG fraction
(2.5 µg/ml, Vector), without primary antisera, or without second
antibody. These control sections were routinely stained in parallel
with each batch of uterine artery slides prepared from both nonpregnant
and pregnant animals.
Isolation of uterine artery endothelial cells
Endothelial cells were isolated by collagenase digestion from
four nonpregnant sheep and six pregnant ewes (110–142 days’
gestation) as previously described (22). Uterine arteries obtained at
the time of euthanasia were quickly flushed free of blood using sterile
M199 medium, before tying off arterial branches, clamping off the
larger diameter end, and inflating with M199 containing 5 mg/ml
collagenase B (Boehringer Mannheim, Indianapolis, IN) and 0.5% BSA
through a luerlock three-way stopcock. Digestion was allowed to proceed
at 37 C for 55 min before flushing the collagenase solution and sheets
of endothelial cell from the tunica intimal lining of the vessel. We
have previously shown that cells isolated in this way are endothelial
in origin and are of high viability and purity (21, 22), based on their
uniform expression of endothelial nitric oxide synthase (eNOS) and
Factor VIII related antigen (both endothelial markers in the uterine
artery), their cobblestone morphology in primary culture, and by uptake
of acetylated low density lipoprotein as visualized under UV excitation
(21, 22, 23).
Western analysis
Endothelial cells isolated by collagenase digestion were
transferred as previously described (21, 22) directly into lysis buffer
[150 mM NaCl, 50 mM Tris-HCl, 10
mM EDTA (pH 7.4), 0.1% Tween 20, 0.1%
ß-mercaptoethanol, 0.1 mM phenylmethylsulfonylfluoride, 5
µg/ml leupeptin, 5 µg/ml aprotinin] and immediately snap frozen in
liquid nitrogen. Endothelial proteins were subsequently solubilized in
lysis buffer by sonication. Solubilized protein was quantified using a
modified Lowry assay procedure (Bio-Rad, Hercules, CA). Proteins (10
µg/lane) were then separated by size on 7.5% polyacrylamide gels
(100 V, 2.5 h, MiniProtean II, Bio-Rad) before transfer to
Immobilon P membrane (100 V, 2 h). The Immobilon P membrane was
then probed for COX-1 using the enhanced chemiluminescence (ECL)
reagent detection system, as described by Amersham (Arlington Heights,
IL), and exposed to hyperfilm (5 min). The COX-1 specific antisera
(Cayman Chemical Company, Ann Arbor, MI) was used at a dilution of
1:3000, and second antibody (rabbit antimouse/horse radish peroxidase
conjugate; Amersham) was used at 1:6000 dilution. Levels of COX-1 were
then quantified by scanning densitometry (Bio-Rad 670 scanning
densitometer) and expressed relative to mean nonpregnant
absorbance.
Isolation of RNA
For extraction of the total cellular RNA, as previously
described (22), freshly isolated endothelial cells were washed free of
collagenase in M199 and pelleted by centrifugation before solubilizing
in 1 ml RNAzolB (Cinna Biotech, Houston, TX). After addition of 150
µl chloroform, and phase separation by centrifugation (12,000 x
g, 20 min) the upper aqueous phase was removed, extracted
twice with phenol/chloroform/isoamyl alcohol using Heavy grade phase
lock gel (5-Prime, 3-Prime, Boulder, CO) and finally mixed with 110%
by volume of isopropanol. RNA was then precipitated by standing at -20
C for 1 h before recovery by centrifugation (12,000 x
g, 30 min), and washing of the pellet in 75% ethanol. RNA
was finally solubilized in molecular biology grade water (5-Prime,
3-Prime) and quantified by spectrophotometry. RNA was extracted by this
procedure from n=4 nonpregnant and n=4 pregnant ewes uterine artery
endothelium.
RT-PCR assay
COX-1 mRNA levels were quantified by coupled RT-PCR
amplification in single tube assays using avian myeloblastosis virus
reverse transcriptase and Taq polymerase, based on the
method of Mallet et al. (24), as described previously (22).
Targets were designed such that the final products, by homology to the
human and bovine sequences, spanned at least two intron sites in each
case; thus, genomic contamination would not result in a false signal
because it would be of considerably greater size. Total cellular RNA (1
µg for COX-1 assay) was incubated in a 50 µl final volume
containing 1x PCR buffer, 2 mM MgCl2, 10 nmol
each deoxy (d)-ATP, dTTP, dCTP, and dGTP, and 30 pmol each of forward
and reverse temperature matched primers, targeting amplification of 313
bases of the ovine COX-1 mRNA (sequence and cDNA courtesy of D. L.
DeWitt, Michigan State University, East Lansing, MI). Amplification was
performed for 27 cycles. Controls of pooled endothelial cell RNA gave
negligible product in the absence of reverse transcriptase. Serial
dilutions of RNA were also accompanied by a proportionate loss of
signal. A standard curve containing known copy numbers of COX-1 cDNA
target sequence was run in each assay. The presence or absence of mRNA
species other than the reverse primer target has little effect on the
standard curve and so was not included with the standards (25). At the
end of the assay, 10 µl of products were separated by size on a 2%
TAE/agarose gel and transferred to MagnaGraph hybridization membrane
(Molecular Separations, Incorporated, available through Fisher,
Pittsburgh, PA) for Southern blotting against a probe encoding the
antisense target sequence, generated by asymmetric PCR (26). After
hybridization, membranes were washed once in 2 x SSC/0.1% SDS
for 30 min and twice in 0.1 x SSC/0.1% SDS (2 x 30 min)
before drying and direct exposure to a phosphorimager (Bio-Rad BI
screen, 5 min) for direct quantification (Molecular Analyst version
version 1.4, Bio-Rad). Data were calculated as copy number COX-1 mRNA
per µg total cellular RNA from the standard curve. All data were
normalized to 28S rRNA content of each sample, determined by slot blot
analysis of 1 µg total RNA.
Statistical analysis
Data were analyzed by Student’s t test. Data are
presented as the mean ± one SEM.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, we show
representative uterine artery cross-sections from nonpregnant and
pregnant ewes immunohistochemically stained for COX-1. Uterine arteries
from four pregnant ewes exhibited intense COX-1 staining in the tunica
intima particularly in the endothelial cells, whereas only faint and
somewhat more diffuse brown staining was observed in the VSM cells of
the tunica media (Fig. 1, upper left). The uterine arteries
obtained from three nonpregnant ewes stained distinctly less in both
the endothelium and the VSM (Fig. 1, lower left). We
observed no immunostaining in the control segments in which either the
primary antibody or secondary antibody were replaced with the
appropriate nonimmune IgG serum (Fig. 1, right panels).
|
). Subsequent
quantitative analysis confirmed that COX-1 (70 kDa) protein in the
uterine artery endothelium of pregnant ewes was 3.5-fold greater than
that from nonpregnant ewes (346.4% ± 28% of nonpregnant control mean
by scanning densitometery, P < 0.0005).
|
RT-PCR quantification
To quantitatively determine if this pregnancy-induced increase in
COX-1 in uterine artery endothelium was also observed at the level of
mRNA, we performed RT-PCR amplification of 1 µg total RNA obtained
from freshly isolated uterine artery endothelial cells. The PCR
products were then separated by size using electrophoresis and
quantified by Southern blot hybridization analysis (Fig. 3). The relative amount of mRNA encoding
COX-1 in the endothelial cells isolated from pregnant ewes was found to
be 51.04-fold ± 7.98 (P < 0.001) greater than
that from nonpregnant ewes, with results normalized for 28S content.
These data clearly illustrate that COX-1 mRNA is greatly elevated in
uterine artery endothelium of pregnant ewes.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have investigated COX expression in the uterine artery because COX is a rate-limiting enzyme in the biosynthetic pathway of PGI2 in other cells (16, 28), and our previous observations (10) showed that there is a pregnancy-induced increase in PGI2 production by uterine arteries when evaluated both under basal conditions, or as maximal PGI2 production under conditions of saturating concentrations of arachidonate. Addition of arachidonate is a useful, albeit indirect, measure of COX activity because this treatment circumvents the initial step in the PGI2 biosynthetic pathway, i.e. the liberation of arachidonate from the cell membrane that is carried out by phospholipases (A and/or C with diacylglycerol lipase). Our data, therefore, suggested that COX expression is higher in the uterine arteries from pregnant vs. nonpregnant ewes. This experimental approach is limited, however, because there are two isoforms of COX known, namely COX-1 and COX-2 (14), and activity measurements neither distinguishes between the COX-1 and COX-2 isoforms nor determines whether the activity of COX or the expression of the COX protein itself were increased.
Although COX-2 is present in uterine tissue (myometrium, endometrium) of many species and has also been reported to be increased during pregnancy (29, 30), we have not detected any COX-2 in uterine arteries from nonpregnant or pregnant ewes using Western analysis (two different antisera, data not shown). In contrast, we clearly detected expression of COX-1 in both VSM and endothelium of uterine arteries from nonpregnant and pregnant ewes using immunohistochemistry. This finding is also consistent with the previous report that there is COX-1, but not COX-2, expression in pulmonary artery endothelial cells from ovine fetuses and newborn lambs (16), and so we chose to focus on the expression of the COX-1 isoform in the regulation of uterine artery PGI2 production during pregnancy.
The immunohistochemical observation that COX-1 is primarily localized to the uterine artery endothelium, rather than the VSM, is consistent with our previous observations that production of PGI2 is greater in uterine arteries with an intact endothelium rather than in arteries denuded of endothelium (10, 13). This also agrees with further studies which showed that bovine aortic endothelial cells contained 20-fold more COX-1 than the VSM (31). Herein, we report for the first time that the expression of this COX-1 protein is also predominant in uterine artery endothelium and that its pregnancy-induced increase is detectable by immunohistochemistry. Although staining for COX-1 in the VSM was much less intense than in the endothelium, it also appeared to be increased by pregnancy. This is consistent with our previous studies in which VSM production of PGI2 was still increased in denuded uterine arteries obtained from pregnant vs. nonpregnant sheep (10). Because, however, the increase in the synthesis of PGI2 by uterine arteries during pregnancy derives predominantly from the endothelium and the immunohistochemical localization of COX-1 was also predominantly in the endothelium, we chose to further quantify the endothelial cell expression of the COX-1 protein and attempted to confirm our findings at the level of mRNA. The increased COX-1 expression in the isolated uterine artery endothelial cells from pregnant compared with nonpregnant ewes was clearly apparent by both the Western analysis and RT-PCR, indicating that the pregnancy-induced elevation in this enzyme occurs at the level of both protein and mRNA. The greater pregnancy-induced fold increase in mRNA as compared with the COX-1 protein could be explained by its extremely short half-life [less than 10 min (32)], so the regulation of COX-1 expression may be at the level of transcription of unstable mRNA. Collectively, however, these data do confirm the hypothesis that increased COX-1 levels in uterine artery endothelium during pregnancy underlie, at least in part, the increased PGI2 levels characteristic of normal pregnancy.
This study parallels our recent observations that expression of both eNOS and the angiotensin II type-1 receptor (AT1-R) increase in sheep uterine artery endothelium during pregnancy (21, 22). Together these findings suggest that multiple proteins (COX-1 and eNOS) are induced in uterine artery endothelium to ensure abundant capacity for vasodilator production (PGI2 and NO, respectively) throughout pregnancy in conjunction with increased receptor expression (AT1-R) to ensure greater endothelial responsiveness to elevated hormones (AII). Each of these changes may help to maintain the dramatically increased uterine blood flow that facilitates ample oxygen and nutrient delivery to the growing fetus.
A remaining issue is which factors present in normal pregnancy cause the increase in COX-1 mRNA, either via increased transcription or decreased breakdown of mRNA. There are several hormones or conditions that are elevated during normal pregnancy, including estrogen (33, 34), AII (4), growth factors (35, 36, 37, 38, 39), and shear stress (3, 40), which have all been shown to up-regulate prostaglandin production in many different tissues. Furthermore, there are reports that these hormones can act at the level of COX-1 because rat aortic VSM cells exhibit an increase in COX-induced PGI2 production upon estrogen treatment (41, 42), and both EGF and PDGF increase COX-1 and -2 expression together with prostaglandin E2 production by a variety of cell-types (36, 37, 38). An interesting alternative, however, is that endothelial cell COX-1 expression is up-regulated in response to locally elevated levels of NO (43). Additional functional and molecular studies will be necessary to determine what, if any, roles these factors have in controlling COX-1 gene expression.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received July 10, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-agonist. III. In vitro fetal-maternal
placental prostaglandins. Am J Physiol 260:E8–E13
This article has been cited by other articles:
 |
|
 |
|
  K. Chang and Lubo Zhang Review Article: Steroid Hormones and Uterine Vascular Adaptation to Pregnancy Reproductive Sciences, April 1, 2008; 15(4): 336 - 348. [Abstract] [PDF]
|
|
|
|
 |
|
 C. Hermenegildo, P.J. Oviedo, M.C. Garcia-Martinez, M.A. Garcia-Perez, J.J. Tarin, and A. Cano Progestogens stimulate prostacyclin production by human endothelial cells Hum. Reprod., June 1, 2005; 20(6): 1554 - 1561. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Jiang, A. S. Weyrich, G. A. Zimmerman, and T. M. McIntyre Endothelial Cell Confluence Regulates Cyclooxygenase-2 and Prostaglandin E2 Production That Modulate Motility J. Biol. Chem., December 31, 2004; 279(53): 55905 - 55913. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. M. Bird, L. Zhang, and R. R. Magness Possible mechanisms underlying pregnancy-induced changes in uterine artery endothelial function Am J Physiol Regulatory Integrative Comp Physiol, February 1, 2003; 284(2): R245 - R258. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Ospina, D. N. Krause, and S. P. Duckles 17{beta}-Estradiol Increases Rat Cerebrovascular Prostacyclin Synthesis by Elevating Cyclooxygenase-1 and Prostacyclin Synthase Stroke, February 1, 2002; 33(2): 600 - 605. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. L. Rupnow, T. M. Phernetton, M. L. Modrick, M. C. Wiltbank, I. M. Bird, and R. R. Magness Endothelial Vasodilator Production by Uterine and Systemic Arteries. VIII. Estrogen and Progesterone Effects on cPLA2, COX-1, and PGIS Protein Expression Biol Reprod, February 1, 2002; 66(2): 468 - 474. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. T. Davidge Prostaglandin H Synthase and Vascular Function Circ. Res., October 12, 2001; 89(8): 650 - 660. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. R. Rosenfeld Mechanisms regulating angiotensin II responsiveness by the uteroplacental circulation Am J Physiol Regulatory Integrative Comp Physiol, October 1, 2001; 281(4): R1025 - R1040. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. R. Magness, J. A. Sullivan, Y. Li, T. M. Phernetton, and I. M. Bird Endothelial vasodilator production by uterine and systemic arteries. VI. Ovarian and pregnancy effects on eNOS and NOx Am J Physiol Heart Circ Physiol, April 1, 2001; 280(4): H1692 - H1698. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.J.D. Lucca, A.S.O. Adeagbo, and N.L. Alsip Oestrous cycle and pregnancy alter the reactivity of the rat uterine vasculature Hum. Reprod., December 1, 2000; 15(12): 2496 - 2503. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Jovanovic, L. Grbovic, and A. Jovanovic Pregnancy is associated with altered response to neuropeptide Y in uterine artery Mol. Hum. Reprod., April 1, 2000; 6(4): 352 - 360. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. M. Bird, J. A. Sullivan, T. Di, J. M. Cale, L. Zhang, J. Zheng, and R. R. Magness Pregnancy-Dependent Changes in Cell Signaling Underlie Changes in Differential Control of Vasodilator Production in Uterine Artery Endothelial Cells Endocrinology, March 1, 2000; 141(3): 1107 - 1117. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Habermehl, M. A. Janowiak, K. E. Vagnoni, I. M. Bird, and R. R. Magness Endothelial Vasodilator Production by Uterine and Systemic Arteries. IV. Cyclooxygenase Isoform Expression During the Ovarian Cycle and Pregnancy in Sheep Biol Reprod, March 1, 2000; 62(3): 781 - 788. [Abstract] [Full Text]
|
|
|
|
|
|
T. Di, J. A. Sullivan, H. L. Rupnow, R. R. Magness, and I. M. Bird Pregnancy Induces Expression of cPLA2 in Ovine Uterine Artery but Not Systemic Artery Endothelium Reproductive Sciences, November 1, 1999; 6(6): 301 - 306. [Abstract] [PDF]
|
|
|
|
|
|
K. E. Vagnoni, N. D. Christiansen, G. R. Holyoak, M. A. Janowiak, and P. H. Martin Cellular Source in Ewes of Prostaglandin-Endoperoxide Synthase-2 in Uterine Arteries Following Stimulation with Lipopolysaccharide Biol Reprod, September 1, 1999; 61(3): 563 - 568. [Abstract] [Full Text]
|
|
|
|
|
|
K. E. Vagnoni and R. R. Magness Estrogen and Lipopolysaccharide Stimulation of Prostacyclin Production and the Levels of Cyclooxygenase and Nitric Oxide Synthase in Ovine Uterine Arteries Biol Reprod, October 1, 1998; 59(4): 1008 - 1015. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
