Expression of Prostaglandin I2 Synthase, but Not Prostaglandin E ...

2 downloads 0 Views 862KB Size Report
Prostaglandin E Synthase, Changes in Myometrium of. Women at Term Pregnancy ... I and II), but also on the expression and activity of specific. PGE synthases (PGES) (6, ..... PGE synthase converts PGH2, the cyclic endoperoxide in-. FIG. 4.
0013-7227/02/$15.00/0 Printed in U.S.A.

The Journal of Clinical Endocrinology & Metabolism 87(11):5274 –5282 Copyright © 2002 by The Endocrine Society doi: 10.1210/jc.2002-020521

Expression of Prostaglandin I2 Synthase, but Not Prostaglandin E Synthase, Changes in Myometrium of Women at Term Pregnancy D. GIANNOULIAS, N. ALFAIDY, A. C. HOLLOWAY, W. GIBB, M. SUN, S. J. LYE, J. R. G. CHALLIS

AND

Canadian Institutes of Health Research in Human Development, Child and Youth Health, Departments of Physiology and Obstetrics and Gynecology, University of Toronto (D.G., N.A., A.C.H., S.J.L., J.R.G.C.), Toronto, Canada M5S 1A8; Department of Obstetrics and Gynecology, Ottawa Hospital (W.G., M.S.), Ottawa, Canada K1H 8L6; and Samuel Lunenfeld Research Institute, Mount Sinai Hospital (S.J.L.), Toronto, Ontario, Canada M5G 1X5 Prostaglandins (PGs) act as potent uterotonins at the time of labor. Prostaglandin E synthase (PGES) is responsible for the formation of PGE2, a uterotonin. PGI2 is synthesized by the prostaglandin I synthase enzyme (PGIS) and contributes to relaxation in the lower uterine segment. We examined the expression of membrane-bound PGES and PGIS in myometrium from pregnant women during preterm and term labor. Tissues were collected from the lower uterine segment from preterm no labor, preterm labor, term no labor, and term labor patients and used for immunohistochemistry and Western blot analysis using specific antibodies. Immunoreactive (ir-) PGES and PGIS proteins were localized to the cytoplasm of myocytes of the myometrium and vascular smooth muscle cells. Ir-PGES was also detected in vascular endothelial cells.

P

ROSTAGLANDINS (PG) play many key roles in mediating physiological processes during pregnancy and labor, including myometrial relaxation and contractility, regulation of utero-placental blood flow, and cervical ripening (1–3). Therefore, examination of PG synthesis within the pregnant uterus is important to elucidate PG effects locally within myometrium from pregnant women. Prostaglandin E2 (PGE2) is a predominant PG in intrauterine tissues (1–3) and has been shown to play an important role in promoting uterine contractility at the time of labor (4, 5). The synthesis of PGE2 is dependent not only on the actions of PGHS (types I and II), but also on the expression and activity of specific PGE synthases (PGES) (6, 7). Similarly, the synthesis of PGI2 depends upon the action of a specific prostaglandin I synthase enzyme (PGIS), which converts PGH2 to PGI2 (8). PGI2 is the most abundantly produced PG in the myometrium (8 –10) and during pregnancy is important in reducing myometrial contractility and regulating uterine and placental blood flow (11, 12). Therefore, changes in the expression or activity of PGES and PGIS enzymes during pregnancy and labor could lead to altered synthesis of the primary PGs, PGE2 and PGI2. PGES is a member of a protein superfamily consisting of membrane-associated proteins that are involved in eicoAbbreviations: cPGES, Cytosolic prostaglandin E synthase; DTT, dithiothreitol; ir-, immunoreactive; mPGES, membrane-bound prostaglandin E synthase; PG, prostaglandin; PGES, prostaglandin E synthase; PGIS, prostaglandin I synthase enzyme.

Western blot analyses revealed a predominant protein band of 180 kDa, and a second 16-kDa band for ir-PGES and 56-kDa band for ir-PGIS. There was no significant change in ir-PGES protein (180 or 16 kDa) or mRNA levels with preterm or term labor or gestational age. There was a significant decrease in PGIS mRNA and protein with advancing gestational age. We conclude that the gestational age decrease in the inhibitory PGIS is consistent with lessening of its influence in myometrium at the time of labor. The lack of change in PGES indicates that alterations at other points along the pathway of arachidonic acid metabolism may be of greater importance in affecting local changes in PGE2. (J Clin Endocrinol Metab 87: 5274 –5282, 2002)

sanoid and glutathione metabolism (the MAPEG family) (13). There are two distinct forms of PGES that have been identified requiring glutathione as a cofactor: cytosolic PGES (cPGES) and membrane-bound PGES (mPGES). Cytosolic PGES is identical to p23, which is reportedly the weakly bound component of the steroid hormone receptor/hsp90 complex and is constitutively expressed, unaltered by proinflammatory stimuli, in various cells and tissues (14). Jakobsson et al. (6) identified and characterized the human microsomal or membrane-bound PGES enzyme and revealed that mPGES is inducible by IL-1␤ in lung carcinoma A549 cells. In addition, mPGES colocalized with both PGHS isozymes in the perinuclear envelope; however, mPGES was functionally coupled with PGHS-II in marked preference to PGHS-I (15). In sheep placenta, levels of PGES increase progressively through pregnancy (15a). Recent studies have localized mPGES to human fetal membranes and specified strong staining in the epithelial layers of the amnion and chorionic trophoblast layer (16, 17). There was no difference in mPGES protein between the preterm or term groups with or without labor for either amnion or chorion (17). There is no information, however, concerning levels of PGES in myometrium at the time of labor. PGI synthase is a membrane-bound hemoprotein that has been localized by immunohistochemical techniques to endothelial cells and both vascular and nonvascular smooth muscle cells (8, 18). PGI2 inhibits human and sheep myometrial activity in vivo and in vitro (19, 20). Interestingly, PGIS

5274

Giannoulias et al. • PGES and PGIS in Myometrium

is expressed in human myometrial smooth muscle (21), particularly the myofibrils (22), and seems likely to change during pregnancy as myometrial production of PGI2 rises (19). However, it is not known whether the levels of PGIS in myometrium from pregnant women change in association with labor. We hypothesized that labor at term and preterm would be associated with decreased expression of PGIS, thereby lessening the inhibitory influence of PGI2 on myometrium, and with increased expression of PGES, to increase production of the potentially stimulatory PGE2. Therefore, we examined the specific localization pattern of PGES and PGIS in myometrium from pregnant women and determined whether the expression of these enzymes changed with labor at preterm and term. Materials and Methods Tissue collection Patient consent and ethical approval were obtained before the onset of the study and tissue collection, according to the guidelines of Mount Sinai Hospital (Toronto, Canada) and University of Toronto. Tissues were collected from the upper edge of the lower uterine segment during cesarean section deliveries and either snap-frozen in liquid nitrogen for protein extraction or slow-frozen for use in immunohistochemistry or in situ hybridization. Myometrial samples were then separated into four groups: preterm (28 –36 wk), not in labor (n ⫽ 8) and in labor (n ⫽ 4); term (38 – 40 wk) not in labor (n ⫽ 8), and labor (n ⫽ 6). Placenta was collected at term (37– 42 wk) either during spontaneous labor (n ⫽ 5) or at elective cesarean section (no labor; n ⫽ 4; fetal distress). Labor was defined as the presence of regular uterine contractions every 1–2 min and lasting for 1 min, resulting in cervical dilatation of 8 –10 cm. None of the patients in this study had received PG or oxytocin. Cesarean sections were either scheduled (nonlaboring patients) or performed because of fetal or maternal complications (laboring patients). Indications for cesarean section were preterm no labor (fetal anomalies, n ⫽ 3; maternal condition, n ⫽ 3; triplets, n ⫽ 2), preterm labor (fetal anomalies, n ⫽ 1; breech, n ⫽ 3), term no labor (previous cesarean section, n ⫽ 4; breech, n ⫽ 3), and term labor (fetal distress, n ⫽ 4; failure to progress, n ⫽ 2).

Immunohistochemistry Immunoreactive PGES was localized in frozen samples of human myometrium using a rabbit antihuman polyclonal antibody raised against a peptide corresponding to amino acids 59 –75 (CRSDPDVERSLRAHRND) of human mPGES (Cayman Chemical, Ann Arbor, MI) at a dilution of 1:50 in antibody dilution buffer. Immunoreactive (ir-) PGIS was localized in myometrial tissue using a mouse antihuman monoclonal antibody raised against human PGIS (Cayman Chemical) at a dilution of 1:50 in antibody dilution buffer. A monoclonal mouse antismooth muscle ␣-actin antibody (A2547, Sigma, St. Louis, MO) diluted 1:3000 in antibody dilution buffer was raised against the NH2-terminal synthetic decapeptide of ␣-smooth muscle actin and was used to identify smooth muscle cells. Immunohistochemistry was conducted as described previously (23). Negative control sections were treated with PGES and PGIS antibodies that had been preabsorbed with the appropriate antigen overnight at 4 C before application to the tissue sections (23).

Dual immunofluorescence Tissue sections were rehydrated in serial dilutions of alcohol (100%, 90%, 70%, and 50%) and washed in PBS. Nonspecific binding of antibodies was blocked with 1% BSA in PBS for 2 h at room temperature. The samples were then incubated with rabbit antihuman mPGES (1:50), mouse anti-smooth muscle ␣-actin (1:3000), and/or mouse antihuman PGIS (1:50) and placed in a 1% BSA solution containing 0.3% Triton X-100. Tissue sections were incubated overnight (18 –24 h) with the primary antibodies at 4 C. After the incubation period, the sections were washed three times in 0.1 m PBS. The secondary antibodies were added and incubated at room temperature for 2 h. The secondary antibodies

J Clin Endocrinol Metab, November 2002, 87(11):5274 –5282 5275

used were fluorophore- and biotin-labeled Alexa Fluor 568 goat antirabbit IgG (red) and Alexa Fluor 488 goat antimouse IgG (green) (Molecular Probes, Inc., Eugene, OR) at a 1:200 dilution in a 1% BSA solution. Samples were washed again in PBS and then dehydrated in serial dilutions of alcohol (50%, 70%, 90%, and 100%). Anti-fading reagent [pphenylendiamine (1 mg/ml), 50% glycerol, and 50% PBS) was added to the tissue sections, and coverslips were applied before analysis.

Microscopic analysis Tissue sections were analyzed under a fluorescent Optiphot-2 microscope (Nikon, Melville, NY) using a green filter to visualize Alexa Fluor 488 and a red filter to visualize Alexa Fluor 568. A Sensicam 12-bit cooled monochrome imaging camera (Cooke, Inc., Auburn Hills, MI) was used to take a digital photograph of the section using Sensicontrol 4.02 software (Cooke, Inc.), and this was visualized on a computer and then exported into Coreldraw Corel (Eastman Kodak Co., Rochester, NY) to produce generated color images as described previously (23). These images were superimposed to obtain the localization patterns of PGES with ␣-actin and PGIS.

Protein extraction and Western blot analysis Protein was extracted from frozen myometrial and placental samples as previously described (23). Homogenates were transferred to Eppendorf tubes (1–5 ml) and centrifuged at 15,000 ⫻ g at 4 C for 15 min to remove tissue debris. Supernatants were collected and transferred to fresh tubes, and protein concentrations were determined using the Bradford protein assay (24) with BSA diluted in protein assay dye reagent as standard (Bio-Rad Laboratories, Inc., Hercules, CA). Concentrations of protein samples were quantified by linear regression analysis from the standard curve. Myometrial protein (50 ␮g) was solubilized, resolved onto a 10% (PGIS) or 12% (PGES) bis-acrylamide gels/4% stacking gels, and transferred onto a nitrocellulose membrane as described previously (23). The PGES and PGIS antibodies (both from Cayman Chemical) were diluted 1:200 and 1:500, respectively, in blocking solution. The blots were then incubated with either antirabbit IgG or antimouse IgG secondary antisera conjugated to horseradish peroxidase (1:2000; Amersham Pharmacia Biotech, Baie d’Urfe, Canada) in blocking solution. The relative intensities of immunoreactive protein signals were quantified using computerized image analysis (MCID Imaging Research, St. Catharines, Canada). Protein bands were digitized, and the mean pixel density for each band was analyzed to obtain relative OD units for each protein. The blots were stripped (0.1 m glycine, pH 2.7; Sigma) and reprobed with the G␤ antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) as an internal control for protein loading. The G␤ antibody detects a 40-kDa protein that does not change in association with gestational age or labor at term or preterm (25). Because initial studies revealed that the major PGES band in myometrium was 180 kDa, we compared myometrium with placenta, a tissue with high PGES activity (6) previously shown to express predominantly the 16-kDa isoform, using the same extraction techniques.

In situ hybridization for PGES and PGIS Samples of human myometrium were collected and stored in OCT embedding medium (Sakahura Fine Chemical Co., Torrance, CA). These specimens were cut into sections (12 ␮m) on a cryostat and postfixed in 4% paraformaldehyde for 5 min, followed by two washes in PBS, and then dehydrated by immersion for 2 min in 70% ethanol. Solutions for tissue preparation and in situ hybridization were prepared with 0.01 m PBS or ddH2O that had been treated previously with diethyl pyrocarbonate (Sigma) to abolish ribonuclease activity. The 50-mer antisense probe (CTT CTT CCG CAG CCT CAC TTG GCC CGT GAT GAT GGC CAC CAC GTA CAT CT) used for in situ hybridization of PGES was complementary to bases 95–144 of the human PGES gene (26). The PGIS probe (GTG CCA CTG TAG AAA TGA TAT GAC CAA CCC CCG TCC ATA CAG TGG) was complementary to bases 181–226 of the human PGIS gene (GenBank accession no. 030508). The probes were synthesized in the molecular biology facility at University of Ottawa using an Oligo 1000 DNA synthesizer (Beckman, Fullerton, CA) and purified by cartridge purification. Hybridization

5276

J Clin Endocrinol Metab, November 2002, 87(11):5274 –5282

Giannoulias et al. • PGES and PGIS in Myometrium

FIG. 1. Expression of ␣-actin in myometrium from women in late pregnancy. A, Comparison of ␣-actin in myometrium from all four groups of patients determined by Western blotting: PT/NL, preterm no labor (n ⫽ 8); PT/L, preterm labor (n ⫽ 4); T/NL, term no labor (n ⫽ 8); and T/L, term labor (n ⫽ 6). B, Immunohistochemical localization of ir-␣-actin and a representative hematoxylin and eosin (H&E) stain of the myometrium from a term nonlaboring patient (T/NL) and a previous cesarean section patient (PCS). C, Representative Western blot and comparison of ␣-actin in myometrium from term nonlaboring (T/NL) patients and from patients who underwent a previous cesarean section (PCS). The histograms represent the relative OD units of ␣actin protein. Data were analyzed by two-way ANOVA (␣ ⫽ 0.05). There were no significant differences in ␣-actin content between the groups shown. All values are the mean ⫾ SEM.

with corresponding sense probes (prepared in a similar fashion) served as the negative control. Probes were labeled using terminal deoxynucleotide transferase (Life Technologies, Inc., Burlington, Canada) and [␣-35S]deoxy-ATP (12.5 mCi/ml; NEN Life Science Products-DuPont Canada, Inc., Mississauga, Canada). Labeled probe was purified using a SORB 20 column (NEN Life Science Products-DuPont Canada, Inc.) and was used at a concentration of 5000 cpm/␮l. Slides were removed from the ethanol and allowed to air-dry at room temperature in a fumehood. The slides were incubated overnight at 45 C in a moist chamber with the radiolabeled oligonucleotide probe in hybridization buffer. Hybridization buffer was composed of 4⫻ SSC [1⫻ SSC ⫽ 150 mm sodium chloride and 15 mm sodium citrate (Sigma)], 50% deionized formamide (Life Technologies, Inc.), 10% dextran sulfate (Amersham Pharmacia Biotech), 25 mm sodium phosphate (pH 7.0), 1 mm sodium pyrophosphate, 200 ␮g acid-alkali hydrolyzed salmon sperm DNA/ml, 100 ␮g polyadenylic acid/ml, 120 ␮g heparin/ml, 40 mm dithiothreitol (DTT; Sigma), and 5⫻ Denhardt’s solution.

Negative control slides were incubated with sense probes. After the overnight incubation, slides were washed in 1⫻ SSC (containing 10 mm DTT) at room temperature for 20 min, then in 1⫻ SSC (with DTT) at 55 C for 45 min and rinsed in 1⫻ SSC, 0.1⫻ SSC, and ddH2O at room temperature for 10 sec each. Slides were dehydrated in ethanol (70% and 95% for 10 sec each), air-dried for 2 h, and exposed to x-ray film (Biomax, Eastman Kodak Co.) using standard methods. Linearity was established by simultaneous exposure of the film to a 14C-labeled standard (RPA504, Amersham Pharmacia Biotech). The autoradiograms were analyzed using computerized analysis software (Image Research, St. Catherines, Canada). Relative OD values were obtained for slides of three or four sections per animal.

Statistical analysis Results from Western blotting analysis and in situ hybridization are presented as the mean ⫾ sem for each patient group; data were analyzed using a two-way ANOVA. Statistical significance was set at P ⱕ 0.05. A

Giannoulias et al. • PGES and PGIS in Myometrium

J Clin Endocrinol Metab, November 2002, 87(11):5274 –5282 5277

FIG. 2. Ir-PGES and ir-PGIS staining in human myometrium. A, Ir-PGES at term (no labor) cesarean section. B, Ir-PGES at term (no labor) in association with a myometrial blood vessel, viewed at higher magnification. C, Section stained with preabsorbed primary antibody for PGES. D, Ir-PGIS at term (no labor). E, Ir-PGIS at term (no labor) at higher magnification. F, Section stained with preabsorbed primary antibody for PGIS. Brown color indicates positive staining. Photographs were taken at a magnification of ⫻200 and ⫻400. Scale bar, 25 ␮m (⫻200) and 12.5 ␮m (⫻400). SM, Smooth muscle; EN, endothelial cells; VSM, vascular smooth muscle. t test was used if the two-way ANOVA indicated a significant main effect. Calculations were performed using SigmaStat (Jandel Scientific Software, San Rafael, CA).

Results Expression of ␣-actin in myometrium from pregnant women

To compare the proportions of myocytes in different tissue specimens, we measured the content and distribution of ␣-actin in samples of myometrium from different patient groups. There were no significant differences observed in the ␣-actin contents of myometrium from women in all four groups (Fig. 1A). Furthermore, there was no obvious change in the immunohistochemical localization of ␣-actin or in the

proportion of ␣-actin-positive cells in myometrium from women at term in the absence of labor with or without previous cesarean section (Fig. 1B). There was no significant difference in the content of ␣-actin in lower segment myometrium at term in the absence of labor with or without previous cesarean section (Fig. 1C). Expression of PGES in myometrium from pregnant women

Staining for PGES was localized in the cytoplasm of smooth muscle myocytes (Fig. 2A). Ir-PGES was also observed in vascular smooth muscle cells and endothelial cells of the blood vessels (Fig. 2B). Ir-PGES immunostaining was

5278

J Clin Endocrinol Metab, November 2002, 87(11):5274 –5282

Giannoulias et al. • PGES and PGIS in Myometrium

FIG. 3. Dual immunofluorescent staining of PGES, PGIS, and ␣-actin in human myometrium. A, ␣-Actin labeled with Alexa Fluor 488. B, PGES labeled with Alexa Fluor 568. C, The superimposed image of ␣-actin and PGES. D, PGIS labeled with Alexa Fluor 488. E, PGES labeled with Alexa Fluor 568. F, The superimposed image of PGIS and PGES. Alexa Fluor 488 produces a green fluorescence. Alexa Fluor 568 produces a red fluorescence. Yellow labeling represents colocalization. EN, Endothelial cells. All photographs taken at a magnification of ⫻200. Scale bar, 25 ␮m.

removed after preabsorption of the PGES antibody with the PGES antigen (Fig. 2C). The localization pattern of PGES and PGIS by immunohistochemistry was confirmed with dual immunofluorescence. PGES colocalized with ␣-actin in smooth muscle myocytes and was also present in the endothelium of blood vessels that did not immunostain for ␣-actin (Fig. 3C). PGES mRNA was detected by in situ hybridization in all

tissues examined. No signal above background was found when tissues were incubated with the sense probe (Fig. 4, A and B). There was no significant change in PGES mRNA with advancing gestation or with labor at preterm and term (Fig. 4B). Western blot analysis of PGES identified two bands at molecular masses of 180 and 16 kDa (Fig. 5A). The 16-kDa band has been shown previously to be the active form of the enzyme (6). There was no significant effect of gestational age

Giannoulias et al. • PGES and PGIS in Myometrium

FIG. 4. In situ hybridization of PGES in human myometrium. A, Representative in situ of PGES term labor (antisense) and sense control. B, The histogram represents the relative OD of PGES mRNA. Data were analyzed by two-way ANOVA (␣ ⫽ 0.05). PT/NL, Preterm no labor (n ⫽ 4); PT/L, preterm labor (n ⫽ 3); T/NL, term no labor (n ⫽ 5); T/L, term labor (n ⫽ 5). All values are the mean ⫾ SEM.

on PGES protein levels of either the 180- or 16-kDa band (P ⬎ 0.05; Fig. 5B), nor was there any significant change in ir-PGES protein levels with labor at preterm or at term. In human myometrium, we observed both 180- and 16kDa bands for PGES protein; however, the 180-kDa band was predominant. We compared the expression of PGES in myometrium to that in placenta, which has high PGES activity (6). In human placenta the PGES antibody identified two bands at 180 and 16 kDa, but the 16-kDa band was predominant (Fig. 6A). There was significantly more 16-kDa PGES protein in human placenta (with or without labor) than in myometrium from pregnant women (with or without labor; Fig. 6B). There was no significant difference in the 180-kDa band in placenta compared with myometrium, with or without labor (data not shown). Expression of PGIS in myometrium from pregnant women

Ir-PGIS was localized to smooth muscle myocytes in myometrium from all four patient groups and also to vascular smooth muscle cells. There was only very weak staining in the endothelial cells (Fig. 2, D and E). Dual immunofluorescence staining indicated that PGIS colocalized with PGES in smooth muscle cells (Fig. 3E). PGIS immunostaining was abolished after preabsorption of the PGIS antibody with its corresponding antigen (Fig. 2F). PGIS mRNA was detected in myometrium from all four

J Clin Endocrinol Metab, November 2002, 87(11):5274 –5282 5279

FIG. 5. Western blot analysis of PGES protein in human myometrium. A, The first four lanes represent a Western blot of PGES protein (180 and 16 kDa). The next four lanes represent the same samples that were preabsorbed with the primary antibody. G␤ (40 kDa) was used as an internal control for protein loading. B, The histogram represents the relative OD of PGES protein (18 kDa isoform) normalized to G␤ to control for protein loading. Data were analyzed by two-way ANOVA (␣ ⫽ 0.05). PT/NL, Preterm no labor (n ⫽ 8); PT/L, preterm labor (n ⫽ 4); T/NL, term no labor (n ⫽ 8); T/L, term labor (n ⫽ 6). All values are the mean ⫾ SEM.

patient groups, and no signal above background was detected with the sense probe (Fig. 7A). There was a significant decrease in PGIS mRNA levels at term compared with preterm patients (P ⬍ 0.05; Fig. 7B). However, there was no significant change in PGIS mRNA with labor either preterm or at term (Fig. 7C). Immunodetection of PGIS by Western blot resulted in a single band with a molecular mass of 56 kDa (Fig. 8A). There was a significant decrease in ir-PGIS protein at term compared with preterm patients (P ⬍ 0.05; Fig. 8B). However, there was no significant change in ir-PGIS protein levels with preterm or term labor (P ⬎ 0.05; Fig. 8C). Discussion

PGES and PGIS mRNA and protein were expressed in myometrium from pregnant women throughout gestation, suggesting that local production of PGE2 and PGI2 might be important in the regulation of myometrial function. The PGES and PGIS proteins were localized to the smooth muscle myocytes in myometrium and smooth muscle cells of the vasculature. There was no significant change in PGES and PGIS protein levels with labor preterm or at term. However, there was a significant decrease in PGIS mRNA and protein with increasing gestational age consistent with a decreasing influence of PGI2 on the myometrium in later gestation. PGE synthase converts PGH2, the cyclic endoperoxide in-

5280

J Clin Endocrinol Metab, November 2002, 87(11):5274 –5282

Giannoulias et al. • PGES and PGIS in Myometrium

FIG. 7. In situ hybridization of PGIS in human myometrium. A, Representative in situ of PGIS preterm no labor (antisense) and sense control. B, Pooled all preterm patients vs. term patients. *, P ⫽ 0.04. C, PT/NL, Preterm no labor (n ⫽ 4); PT/L, preterm labor (n ⫽ 3); T/NL, term no labor (n ⫽ 5). All values are the mean ⫾ SEM. Data were analyzed by two-way ANOVA (␣ ⫽ 0.05). FIG. 6. Western blot analysis of PGES protein in term placenta and human myometrium (Myo) during cesarean section (labor and no labor). A, The first four lanes show a Western blot of PGES protein in human myometrium (180 and 16 kDa) at term no labor. The next five lanes represent PGES protein in human placenta (term no labor). G␤ (40 kDa) was used as an internal control for protein loading. B, The histogram represents the relative OD of PGES protein (the 16-kDa isoform) normalized to G␤ to control for protein loading. Data were analyzed by two-way ANOVA. *, P ⬍ 0.05, placenta vs. myometrium T/NL, Myometrium term no labor (n ⫽ 5); T/L, myometrium term labor (n ⫽ 3); T/NL, placenta (n ⫽ 4); T/L, placenta (n ⫽ 5). All values are the mean ⫾ SEM.

termediate of PG synthesis, to PGE2. Likewise, prostacyclin synthase converts PGH2 to PGI2. In the present study the antibody we used recognized the membrane-bound form of the PGES enzyme. In human cell lines mPGES is colocalized with both PGHS enzymes in the perinuclear envelope (15). We localized mPGES to the smooth muscle in the myometrium, similar to the localization pattern of PGHS-II as demonstrated previously (23) and consistent with studies that have shown functional coupling of mPGES with either of the PGHS isoforms (15). These results suggest that localization of mPGES and PGHS-II to the smooth muscle of the myometrium may allow for efficient conversion of arachidonic acid to PGE2 at that locale. Previously it had been demonstrated that human vascular smooth muscle cells, but not endothelial cells, expressed PGES (27). However, we localized PGES to the smooth muscle cells and to endothelial cells of the vasculature, consistent with a local effect of PGE2 on vascular function. These differences may reflect different blood vessels since Soler et al. (27) scanned sections of human umbilical vein. Western blot analysis of ir-PGES revealed immunoactive protein bands of 180 and 16 kDa. The cytosolic form of PGES immunoprecipitates as a protein of 23 kDa as described pre-

viously (14) and was not measured in this study. Previous results have demonstrated that purified mPGES immunoprecipitated as two protein bands with molecular masses of 17 and 180 kDa, respectively (28). The 17-kDa protein had a Km for PGH2 of 40 ␮m, and the larger protein had a Km of 150 ␮m. PGH2 is therefore a better substrate for the smaller molecular mass PGES protein, suggesting that the smaller protein is probably the active protein (6, 28). We found no significant change in the levels of PGES mRNA or protein with increasing gestation or with labor at preterm or term. No change in mPGES protein was observed in human amnion and chorion tissue between preterm or term patient groups with or without labor (17). Therefore, changes in the levels of this protein within intrauterine tissues do not appear to contribute to the marked increase in PGE2 output from intrauterine tissues at the time of labor. Our results also showed that whereas the 180-kDa band is the predominant form expressed in myometrium from pregnant women, the 16-kDa band is the predominant band expressed in human placenta and was present at much higher levels in the placenta. If this distribution reflects active enzyme, then the production of PGE2 may be more important in placenta than myometrium, perhaps signifying the importance of PGE2 regulating placental blood flow throughout pregnancy. However, the action of PGE2 also depends on the receptor subtypes present in both placenta and myometrium, and future studies will need to evaluate the distribution and changes in these receptor proteins. Immunostaining and dual immunofluorescence of PGES and PGIS suggested that these enzymes were associated with smooth muscle myocytes, consistent with previous results (23) showing PGHS and PGIS immunostaining in the same cells. In our present study we localized PGIS in myocytes and

Giannoulias et al. • PGES and PGIS in Myometrium

FIG. 8. Western blot analysis of PGIS protein in human myometrium. A, Western blot of PGIS protein. G␤ (40 kDa) was used as an internal control for protein loading. B, PGIS protein, pooled all preterm patients vs. term patients. *, P ⬍ 0.05. The histogram represents the relative OD of PGIS protein normalized to G␤ to control for protein loading. Data were analyzed by two-way ANOVA (␣ ⫽ 0.05). C, PT/NL, Preterm no labor (n ⫽ 8); PT/L, preterm labor (n ⫽ 4); T/NL, term no labor (n ⫽ 8); T/L, term labor (n ⫽ 6). All values are the mean ⫾ SEM.

smooth muscle cells in the vasculature, suggesting that the high concentration of PGIS in uterine smooth muscle cells could represent substantial PGI2 biosynthetic capacity. This is consistent with reports that PGI2 is the main arachidonic acid metabolite in the human myometrium (9), is produced by the smooth muscle myocytes (19, 21), and contributes to smooth muscle relaxation in the lower segment. In addition, we colocalized PGIS with PGES in the same cell types, suggesting that these two enzymes are present in the same cell acting to efficiently convert PGH2 into either PGE2 or PGI2. Although the capacity to produce PGI2 appears to be an inherent activity of the myocytes, the vasculature of the myometrium is also capable of PGI2 production (29). However, immunostaining for PGIS in the smooth muscle of the vasculature was much less than that in myometrial cells, which concurs with the findings of a previous study (21). Western blot analysis of ir-PGIS in myometrium from pregnant women revealed a single band at 56 kDa, which is consistent with previous studies (19, 30, 31). We found no significant change in PGIS protein or mRNA with labor preterm and at term. However, we did find a significant decrease in PGIS mRNA and protein with increasing gestational age, suggesting that PGI2 production may be more important before term, when the myometrium is relatively quiescent, than at term, when the myometrium switches into a relatively contractile state. We appreciate that we have taken lower segment tissues, and there could have been

J Clin Endocrinol Metab, November 2002, 87(11):5274 –5282 5281

variability in the amount of smooth muscle present from one patient to another, especially those patients who had had a previous cesarean section. However, after quantifying the ␣-actin content, we found no significant change in ␣-actin protein during preterm or term labor or in patients who had undergone a previous cesarean section. Our results agree with those of Word et al., (32), who also found no significant change in the actin content between myometrium of nonpregnant and pregnant women. Thus, we submit that the lack of change in mPGES in different patient groups is not a function of altered numbers of myocytes. Furthermore, the different pattern of change in PGES and PGIS, both of which localize predominantly to myocytes, indicates an independent pattern of regulation. Levels of PGES and PGIS protein may not necessarily reflect enzyme activity. Furthermore, the effects of PGE2 and PGI2 may be more dependent on the receptor subtypes expressed in human myometrium. Senior et al. (33, 34) showed that lower uterine segment tissue had more pronounced responses to prostaglandin E2 receptor type 2 and inducing protein receptor activation, consistent with the thesis that the lower segment myometrium remains relatively passive at the time of labor and dilates to allow passage of the fetus. Interestingly, PGE2 induced contraction of strips of myometrium obtained from the upper segment during labor, whereas the lower uterine segment responded with inhibition (33). Lye and Challis (35) showed that in nonpregnant sheep, the infusion of PGI2 significantly reduced the frequency and amplitude of uterine contractions. Similarly, in myometrium from pregnant women, PGI2 significantly lowered the tonus and reduced spontaneous motility (10, 33). Thus, our results are consistent with local production of PGI2 in myometrium, which may play a major role in sustaining uterine quiescence during pregnancy, an effect that may be diminished at the time of labor. In conclusion, the presence of PGES and PGIS in human myometrium supports a potentially important role for locally produced PGI2 and PGE2 in the regulation of myometrial activity. The expression of PGES did not change with gestational age or the presence of labor, perhaps indicating the greater importance of alterations in PGE receptor subtypes in effecting PGE2 action. However, the significant decline in levels of PGIS mRNA and protein in later gestation suggest that local withdrawal of this inhibitory PG in the myometrium may contribute to the labor process. Acknowledgments We thank Ljiana Petkovic, Cristine Botsford, and Lindsay McWhirter (Mount Sinai Hospital, Toronto, Ontario, Canada) for their assistance with collecting tissues for these studies, and Nohjin Kee for help with the dual immunofluorescent studies. Received April 2, 2002. Accepted August 5, 2002. Address all correspondence and requests for reprints to: Ms. Diana Giannoulias, 1 King’s College Circle, Medical Sciences Building, Room 3344, Department of Physiology, University of Toronto, Toronto, Ontario, Canada M5S 1A8. E-mail address: [email protected]. This work was supported by the Canadian Institutes of Health Research in Human Development, Child and Youth Health (MOP 42378).

5282 J Clin Endocrinol Metab, November 2002, 87(11):5274 –5282

References 1. Challis JRG, Lye SJ, Gibb W 1997 Prostaglandins and parturition. Ann NY Acad Sci 828:254 –267 2. Gibb W 1998 The role of prostaglandins in human parturition. Ann Med 30:235–241 3. Hansen WR, Keelan JA, Skinner SJM, Mitchell MD 1999 Key enzymes of prostaglandin biosynthesis and metabolism. Coordinate regulation of expression by cytokines in gestational tissues: a review. Prostaglandins Other Lipid Mediat 57:243–257 4. Granstrom L, Ekman G, Ulmsten U 1990 Myometrial activity after local application of prostaglandin E2 for cervical ripening and term labor induction. Am J Obstet Gynecol 162:691– 694 5. Challis JRG, Matthews SG, Gibb W, Lye SJ 2000 Endocrine and paracrine regulation of birth at term and preterm. Endocr Rev 21:514 –550 6. Jakobsson P, Thoren S, Morgenstern R, Samuelsson B 1999 Identification of human prostaglandin E synthase: a microsomal, glutathione-dependent, inducible enzyme, constituting a potential novel drug target. Proc Natl Acad Sci USA 96:7220 –7225 7. Watanabe K, Kurihara K, Suzuki T 1999 Purification and characterization of membrane-bound prostaglandin E synthase from bovine heart. Biochim Biophys Acta 1439:406 – 414 8. Tanabe T, Ullrich V 1995 Prostacyclin and thromboxane synthases. J Lipid Mediat Cell Signal 12:243–255 9. Christensen N, Green K 1983 Bioconversion of arachidonic acid in human pregnant reproductive tissues. Biochem Med 30:162–180 10. Omini C, Folco GC, Pasargklian R, Fano M, Berti F 1979 Prostacyclin (PGI2) in the pregnant human uterus. Prostaglandins 17:113–120 11. Wilhelmsson L, Wikland M, Wiqvist N 1981 PGH2, TXA2 and PGI2 have potent and differentiated actions on human uterine contractility. Prostaglandins 21:277–286 12. Makila UM, Jouppila P, Kirkinen P, Viinikka L, Ylikorkala O 1986 Placental thromboxane and prostacyclin in the regulation of placental blood flow. Obstet Gynecol 68:537–540 13. Moonen P, Buytenhek M, Nugteren DH 1982 Purification of PGH-PGE isomerase from sheep vesicular glands. Methods Enzymol 86:84 –90 14. Tanioka T, Nakatani Y, Semmyo N, Murakami M, Kudo I 2000 Molecular identification of cytosolic prostaglandin E2 synthase that is functionally coupled with cyclooxygenase-1 in immediate prostaglandin E2 biosynthesis. J Biol Chem 275:32775–32782 15. Murakami M, Naraba H, Tanioka T, Semmyo N, Nakatani Y, Kojima F, Ikeda T, Fueki M, Ueno A, Oh-Shishi S, Kudo I 2000 Regulation of prostaglandin E2 biosynthesis by inducible membrane-associated prostaglandin E2 synthase that acts in concert with cyclooxygenase-2. J Biol Chem 275:32783–32792 15a.Martin RL, Whittle WL, Holloway AC, Gyomorey S, Gibb W, Lye SJ, Challis JRG 2002 Ontogeny and regulation of ovine placental prostaglandin E2 synthase. Biol Reprod 67:868 – 873 16. Marvin KW, Parks B, Keelan JA, Sato, TA, Mitchell MD 2002 Expression of microsomal prostaglandin E synthase in human gestational tissues with labor at term and with infection and labor preterm [Abstract 452]. J Soc Gynecol Invest 9(Suppl 212A) 17. Meadows JW, Eis A, Brockman DE, Myatt L 2002 Expression and localization of membrane-bound prostaglandin E synthase in human amnion and chorion [Abstract 455]. J Soc Gynecol Invest 9(Suppl 212A)

Giannoulias et al. • PGES and PGIS in Myometrium

18. Smith WL 1986 Prostaglandin biosynthesis and its compartmentation in vascular smooth muscle and endothelial cells. Annu Rev Physiol 48:251–262 19. Moonen P, Klok G, Keirse MJ 1984 Increase in concentrations of prostaglandin endoperoxide synthase and prostacyclin synthase in human myometrium in late pregnancy. Prostaglandins 28:309 –321 20. Wu WX, Ma XH, Nathanielsz PW 1998 Changes in prostacyclin synthase in pregnant sheep myometrium, endometrium, and placenta at spontaneous term labor and regulation by estradiol and progesterone. Am J Obstet Gynecol 180:744 –749 21. Moonen P, Klok G, Keirse MJ 1985 Immunohistochemical localisation of prostaglandin endoperoxide synthase and prostacyclin synthase in pregnant human myometrium. Eur J Obstet Gynecol Reprod Biol 19:151–158 22. Moonen P, Klok G, Keirse MJ 1986 Distribution of prostaglandin endoperoxide synthase and prostacyclin synthase in the late pregnant uterus. Br J Obstet Gynaecol 93:255–259 23. Giannoulias D, Patel FA, Lye SJ, Tai HH, Challis JRG 2002 Differential expression of prostaglandin dehydrogenase and prostaglandin H synthase type I and II in pregnant human myometrium. J Clin Endocrinol Metab 87: 1345–1352 24. Bradford M 1976 A rapid and sensitive method for quantification of microgram quantities of protein utilizing the principle of protein dye binding. Anal Biochem 72:248 –254 25. Europe-Finner GN, Phaneuf S, Tolkovsky AM, Watson SP, Lopez Bernal A 1994 Down-regulation of G alpha s in human myometrium in term, and preterm labor: a mechanism for parturition. J Clin Endocrinol Metab 79:1835– 1839 26. Forsberg L, Leeb L, Thoren S, Morgenstern R, Jakobsson P 2000 Human glutathione dependent prostaglandin E synthase: gene structure and regulation. FEBS Lett 471:78 – 82 27. Soler M, Camacho M, Escudero J, Iniguez M, Vila L 2000 Human vascular smooth muscle cells but not endothelial cells express prostaglandin E synthase. Circ Res 87:504 –507 28. Tanaka Y, Ward S, Smith WL 1987 Immunochemical and kinetic evidence for two different prostaglandin H-prostaglandin E isomerases in sheep vesicular gland microsomes. J Biol Chem 262:1374 –1381 29. Wallenburg HCS 1981 Prostaglandins and the maternal placental circulation: review and perspectives. Biol Res Pregnancy 2:15–22 30. Hecker M, Ullrich V 1985 On the mechanism of prostacyclin and thromboxane A2 biosynthesis. J Biol Chem 264:141–150 31. Smith WL, Marnett LJ, Dewitt DL 1991 Prostaglandin and thromboxane biosynthesis. Pharmacol Ther 49:179 32. Word RA, Stull JT, Casey ML, Kamm KE 1993 Contractile elements and myosin light chain phosphorylation in myometrial tissue from nonpregnant and pregnant women. J Clin Invest 92:29 –37 33. Senior J, Marshall R, Sangha R, Clayton JK 1993 In vitro characterization of prostanoid receptors on human myometrium at term pregnancy. Br J Pharmacol 108:501–506 34. Wiqvist N, Bryman L, Lindblom B, Norstrom A, Wikland M 1985 The role of prostaglandins for the coordination of myometrial forces during labour. Acta Physiol Hung 65:313–322 35. Lye SJ, Challis JRG 1982 Inhibition by PGI-2 of myometrial activity in vivo in non-pregnant ovariectomized sheep. J Reprod Fertil 66:311–315