- Email: [email protected]
Vasoactive mediator release by fetal endothelial cells in intrauterine growth restriction and preeclampsia
Vasoactive mediator release by fetal endothelial cells in intrauterine growth restriction and preeclampsia Mauro C. Parra, MD,a, b Christoph Lees, MD,a Giovanni E. Mann, PhD,b Jeremy D. Pearson, PhD,b and Kypros H. Nicolaides, MDa London, United Kingdom OBJECTIVE: Preeclampsia and fetal growth restriction are associated with poor placental perfusion, which may be accompanied by a compensatory release of vasoactive substances in the fetoplacental circuit. This study examines the effects of preeclampsia and fetal growth restriction on nitric oxide and prostacyclin signaling pathways in fetal endothelial cells. STUDY DESIGN: Human umbilical vein endothelial cells from 30 control pregnancies, 18 pregnancies with preeclampsia, and 9 pregnancies with intrauterine growth restriction were cultured. Intracellular cyclic guanosine monophosphate accumulation and 6-keto-prostaglandin F1α production were determined. RESULTS: Intracellular accumulation of cyclic guanosine monophosphate was significantly higher in the preeclampsia group and lower in the growth restriction group than in the control group (9.8, 1.8, and 3.9 pmol/µg protein for 5 minutes, respectively), whereas 6-keto-prostaglandin F1α production was not significantly different in the 3 groups. CONCLUSION: The data suggest that the fetoplacental vascular response to preeclampsia is to increase production of cyclic guanosine monophosphate, perhaps to maintain vessel dilatation and maximum flow through placental villi. In fetal growth restriction the umbilical vein endothelial cells do not or cannot respond to chronic hypoxia by increasing cyclic guanosine monophosphate, which may lead to fetoplacental vasoconstriction. (Am J Obstet Gynecol 2001;184:497-502.)
Key words: Fetal growth restriction, human endothelial cells, nitric oxide, preeclampsia, prostacyclin
In normal pregnancy progressive trophoblast invasion transforms the high-resistance uteroplacental spiral arteries into a low-resistance circulation. This vascular transformation facilitates the 10-fold increase in uteroplacental blood flow that occurs between conception and term. Histologic studies have shown that the process of spiral artery vascular transformation is incomplete in pregnancies affected by preeclampsia and fetal growth restriction.1 The resulting uteroplacental circulation is characterized by high resistance to flow in the maternal uterine arteries, which can be detected noninvasively by the use of Doppler ultrasonography.2 Preeclampsia and fetal growth restriction are therefore both associated with relative placental hypoperfusion. This may stimulate a compensatory increase in prostacyclin and nitric oxide, which
From the Harris Birthright Research Centre for Fetal Medicine, King’s College Hospital, Denmark Hill,a and the Centre for Cardiovascular Biology and Medicine, King’s College London, Guy’s Campus.b Supported by The Fetal Medicine Foundation (registered United Kingdom charity 1037116) and Mideplan (Chile). Received for publication December 31, 1999; revised February 29, 2000; accepted July 20, 2000. Reprint requests: Kypros H. Nicolaides, MD, Harris Birthright Research Centre for Fetal Medicine, King’s College Hospital, Denmark Hill, London, United Kingdom SE5 8RX. Copyright © 2001 by Mosby, Inc. 0002-9378/2001 $35.00 + 0 6/1/110311 doi:10.1067/mob.2001.110311
are the main vasodilators of the human fetal-placental circulation.3, 4 Nitric oxide is synthesized from the amino acid L-arginine via the endothelial isoform of nitric oxide synthase and modulates vascular tone by activating soluble guanylate cyclase and elevating intracellular cyclic guanosine 3´,5´-monophosphate (cGMP) levels in target smooth muscle cells.5 There is disagreement about placental nitric oxide production in preeclampsia and fetal growth restriction, with studies reporting a decrease, no change, or an increase, which may reflect differences in the methods of assessment, such as levels of nitric oxide synthase messenger ribonucleic acid, immunoreactivity of nitric oxide synthase, conversion from arginine to citrulline, and levels of basal or agonist-stimulated cGMP.6-10 Study results may additionally be confounded because placental nitric oxide production is highest early in normal pregnancy and falls with increasing gestational age.8 It is also possible that alterations in cGMP levels reflect the action of atrial natriuretic peptide rather than nitric oxide.11 For isolated fetal vessels or endothelial cells derived from them and cultured in vitro, there are fewer data on nitric oxide production, but they are again conflicting. Pinto et al12 reported a decrease in the release of endothelium-derived vasodilator substances from perfused vessels in preeclampsia, whereas Akar et al13 used a similar technique and found raised basal nitric oxide re497
498 Parra et al
lease (but unaltered agonist-stimulated release). Boccardo et al14 found no change in levels of nitric oxide synthase messenger ribonucleic acid or in the conversion of citrulline to arginine. However, higher circulating levels of fetal nitrite and nitrate and of amniotic fluid have been reported in both preeclampsia and fetal growth restriction.15, 16 Prostacyclin synthesis is initiated by the enzyme phospholipase A2, which liberates arachidonic acid from membrane phospholipids. Arachidonic acid is converted by cyclooxygenase into prostaglandin endoperoxides, and prostacyclin synthase subsequently forms prostacyclin (prostaglandin I2) from the endoperoxide prostaglandin H2. Goeschen et al17 showed that the maternal plasma concentration of prostacyclin in severe preeclampsia, in comparison with the concentration in normal pregnancies, was significantly lower, whereas the concentration of the vasoconstrictor eicosanoid thromboxane A2 was significantly higher. Fetal blood vessels and placenta synthesize prostacyclin, and there is a consensus view that, as in the mother, the production of prostacyclin is lower and the production of thromboxane A2 is higher in fetal tissues from women with preeclampsia.18 In this study we took for culture the early-passage human umbilical vein endothelial cells during a range of gestational ages in the third trimester of healthy pregnancies and of pregnancies affected by preeclampsia and fetal growth restriction. To assess the relative contributions of these vasodilators, we compared rates of transport of the nitric oxide precursor arginine and levels of basal and agonist-stimulated nitric oxide and prostacyclin synthesis in these cells. Material and methods Subjects and definitions. We studied cells cultured from 57 human umbilical cords taken from women delivered in our institution during a 20-month period in 1997 and 1998. Collection of the cords was according to the Polkinghorne recommendations19 on the research use of fetal material. Cords were from 18 pregnancies complicated by preeclampsia, from 9 pregnancies with fetal growth restriction, from 14 otherwise normal pregnancies with spontaneous preterm delivery, and from 16 normal pregnancies with delivery at term. Preeclampsia was defined as indicated by blood pressure >140/90 mm Hg measured on 2 occasions at least 6 hours apart and a urinary protein concentration >300 mg/24 h or ≥+1 on dipstick test, with no known history of hypertension or renal disease before pregnancy. Fetal growth restriction was defined as birth weight at <10th percentile for gestational age and sex. None of the fetal growth restriction cases in this study were complicated by preeclampsia, and neither were preeclampsia cases complicated by fetal growth restriction. No women were taking aspirin or were nitric oxide donors at the time of delivery.
February 2001 Am J Obstet Gynecol
Umbilical cord cells were cultured, and depending on the quantity of endothelial cells obtained, they were studied for at least 2 of the 3 experiments—transport of arginine, basal and stimulated cGMP release, and basal and stimulated prostacyclin release. Isolation and primary culture of endothelial cells. Cells were isolated from the umbilical cord veins by a collagenase dispersion method within 24 to 48 hours of delivery. In brief, after the veins were cannulated and rinsed with phosphate-buffered saline solution, the endothelial cells were loosened by incubation with 0.1% collagenase for 10 minutes at 37°C. The cells were then washed out of the cord, collected by centrifugation, washed, and resuspended in 5 mL of culture medium, composed of medium 199 containing 10% (vol/vol) fetal calf serum, 10% (vol/vol) newborn calf serum, L-glutamine (3.2 mmol/L), 0.016% penicillin, 0.008% streptomycin, and sodium bicarbonate (20 mmol/L). The cells were incubated in a T-25 culture flask at 37°C in a 5% carbon dioxide–95% air atmosphere. Confluent primary cultures were then dispersed, subcultured with 0.05% trypsin plus 0.025% ethylenediaminetetraacetic acid, and split 1:3. Cells from passage number 2 or 3 were used in all experiments. Cell number and cell protein were determined in confluent monolayers as previously described.20 Transport of arginine. Confluent third-passage monolayers grown in 24-well culture trays were rinsed with warmed (37°C) Krebs solution (sodium chloride, 131 mmol/L; potassium chloride, 5.6 mmol/L; sodium bicarbonate, 25 mmol/L; sodium biphosphate, 1 mmol/L; Dglucose, 5 mmol/L; N-[2-hydroxyethyl]piperazine-N´-[2ethanesulfonic acid] [HEPES], 20 mmol/L; calcium chloride, 2.5 mmol/L; and magnesium chloride, 1 mmol/L [pH 7.4]) and preincubated for 60 minutes at 37°C in Krebs solution containing L-arginine, 100 µmol/L. Kinetics of arginine transport were measured for 30 seconds in endothelial cells incubated with increasing concentrations of [3H]arginine (0.015-1 mmol/L) as previously described.20 Kinetic constants were derived by fitting a Michaelis-Menten curve plus a linear component to the raw data. In all experiments uptake was terminated by removal of the uptake medium immediately before monolayers were rinsed 3 times with 200 µL of icecold stop solution (Krebs buffer solution containing unlabeled substrate, 10 mmol/L). Protein was determined by the bicinchoninic acid reagent, and radioactivity in formic acid cell digests was determined by liquid scintillation counting. Basal and histamine-stimulated cGMP and prostacyclin levels. Confluent monolayers in 24-well plates were preincubated for 15 minutes with Krebs solution (1 mL, 37°C) containing L-arginine, 100 µmol/L, and the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine, 0.5 mmol/L. After this period 500-µL aliquots of cell-free supernatant were removed from each well, and histamine
Parra et al 499
Volume 184, Number 3 Am J Obstet Gynecol
Table I. Clinical characteristics of patients with cultured umbilical vein endothelial cells Study group Variable Maternal age (y) Mean ± SD Range Primiparous (%) Gestational age at delivery (wk) Mean ± SD Range Birth weight (kg) Mean ± SD Range
Control (n = 30)*
Preeclampsia (n = 18)
Fetal growth restriction (n = 9)
29.7 ± 6.8 (17-39) 23
31.6 ± 6.1 (17-40) 28
29.3 ± 6.1 (21-38) 11
37.0 ± 3.6 (29-42)
37.1 ± 2.0 (32-40)
36.6 ± 2.2 (33-40)
2.97 ± 0.77 (1.61-4.34)
2.82 ± 0.56 (1.63-4.02)
2.06 ± 0.55 (1.27-2.86)†
*In control group, 14 deliveries were preterm. †P < .05, compared with preeclampsia and control groups.
solution (5 µL, concentrated 100 times; final concentration, 10 µmol/L) was added to the wells for a further 5 minutes at 37°C. The supernatants from before and after stimulation by histamine were used for analysis of basal and stimulated rates of prostacyclin release by radioimmunoassay of its stable metabolite, 6-keto-prostaglandin F1α. Cells were then placed on ice and incubated with 0.1N hydrochloric acid (500 µL per well for 60 minutes). Afterward the extracts were stored at –20°C for radioimmunoassay of cGMP after acetylation. To confirm that elevated cGMP levels were a result of the production of nitric oxide, we treated cells with the nitric oxide synthase inhibitor Nω-nitro-L-arginine methyl ester (L-NAME; 15 minutes, 100 µmol/L).20, 21 Protein was assayed with the bicinchoninic acid reagent. Statistical analysis. Results were analyzed by unpaired nonparametric methods (Mann-Whitney U and Wilcoxon tests), as appropriate; differences were considered significant if the P value was < .05. Trends were analyzed by linear regression. Results Patients’ baseline clinical details are shown in Table I. The groups were matched for maternal age and gestational age at delivery. As expected, the mean birth weight was significantly lower in the fetal growth restriction group than in the control and preeclampsia groups. The kinetics of L-arginine transport in cells from 9 control and 9 preeclampsia pregnancies, matched in range of gestational age, were compared. Monolayers were incubated with L-[3H]arginine, 0.15 to 1 mmol/L at 37°C, and initial rates of influx were determined for 30 seconds. A Michaelis-Menten hyperbola, together with a linear nonsaturable component, was fitted to mean influx values weighted for the reciprocal of the respective standard error. There was no significant difference in the characteristics of arginine uptake between the 2 groups. For the control group the Michaelis-Menten constant (Km) and
Table II. Arginine transport by umbilical vein endothelial cells from control, preeclampsia, and fetal growth restriction pregnancies Type of pregnancy
L-Arginine (240 µmol/L) (pmol [µg protein]–1 min–1)
Control (n = 14) Preeclampsia (n = 8) Fetal growth restriction (n = 4)
5.7 ± 0.8 5.1 ± 0.6 4.1 ± 0.9
Values are mean ± SEM.
maximum velocity values were 79 ± 8 µmol/L and 6.2 ± 0.3 pmol (µg protein)–1 min–1, respectively. For the preeclampsia group the values were 111 ± 4 µmol/L and 6.5 ± 0.1 pmol (µg protein)–1 min–1, respectively. When uptake of arginine was compared at a fixed concentration (240 µmol/L) in a series of experiments, no significant differences were observed between control, fetal growth restriction, and preeclampsia groups (Table II). The mean basal level of cGMP accumulation was significantly higher in cells from the preeclampsia group than from the control group, whereas, in contrast, it was significantly lower in cells from the fetal growth restriction group (Table III). When these data were examined more closely, it became evident that the highest levels in preeclampsia were associated with the lowest gestational ages, with a negative trend that just failed to reach statistical significance (n = 17; r = –0.44; P = .07; Fig 1). In contrast, basal rates of cGMP accumulation in control cells were positively correlated with gestational age (n = 29; r = 0.42; P = .02; Fig 1). Basal cGMP levels from fetal growth restriction–derived cells were lower than normal at all gestational ages (Fig 1). When basal cGMP levels were related to growth percentiles, levels in control cells were unrelated to newborn size, but a negative correlation was observed for the preeclampsia cases (n = 17; r = –0.46; P = .05; Fig 2). Histamine significantly increased cGMP production in
500 Parra et al
February 2001 Am J Obstet Gynecol
Table III. Basal and histamine-stimulated cGMP production in human umbilical vein endothelial cells from control, preeclampsia, and fetal growth restriction pregnancies Type of pregnancy Control (n = 29) Preeclampsia (n = 17) Fetal growth restriction (n = 8)
Basal cGMP (pmol [µg protein]–1 [5 min]–1)
Histamine- stimulated cGMP (pmol [µg protein]–1 [5 min]–1)
3.9 ± 0.5 9.8 ± 2.1* 1.8 ± 0.2†
10.4 ± 2.8 23.4 ± 4.7* 2.9 ± 1.0†
Fold stimulation by histamine 2.31 ± 0.34 2.95 ± 0.46 1.43 ± 0.14†
Values are mean ± SEM. *P < .05, compared with fetal growth restriction and control groups. †P < .05, compared with control and preeclampsia groups.
all cells (Table III). However, the increase was lowest in fetal growth restriction, where the basal levels were lowest, and highest in preeclampsia, where the basal levels were highest. Histamine-induced increases in cGMP were blocked in the presence of L-NAME, 100 µmol/L, in all 3 groups, confirming that the increases were a result of nitric oxide production (data not shown). The fold increase stimulated by histamine was similar in the control and preeclampsia groups but significantly lower in the fetal growth restriction group (Table III). The mean levels of basal and histamine-stimulated prostacyclin production were not significantly different between groups (Fig 3). In contrast to basal cGMP production, there was no relationship between basal prostacyclin synthesis and either gestational age or newborn size in cells from the control and preeclampsia groups. Moreover, no relationship between maximal histaminestimulated prostacyclin release and gestational age was found for any group. Comment We examined the ability of umbilical vein endothelial cells derived from healthy (control) pregnancies, pregnancies with preeclampsia, and pregnancies with fetal growth restriction in the third trimester to synthesize 2 important vasodilators, nitric oxide and prostacyclin. We measured basal rates of nitric oxide–mediated cGMP production and prostacyclin synthesis and the rates stimulated by a maximally effective dose of histamine. Differences in basal rates are likely to reflect intrinsic differences in the control of vasoactive mediator synthesis that are maintained for at least 2 or 3 passages in cell culture. The differences might be caused by alterations in the intracellular levels of important cofactors for enzyme activity (eg, cytosolic ionized calcium, which regulates both nitric oxide and prostacyclin synthesis) or substrates (eg, arginine for nitric oxide synthesis, which is acquired by uptake from the medium). Differences in agonist-stimulated rates might reflect alterations in histamine receptor levels or, more likely, might reflect differences in the total amount of nitric oxide synthase or the rate-limiting enzyme in the prostacyclin synthetic pathway (phospholi-
pase A2). The lower ability to synthesize cGMP in the fetal growth restriction group could also be caused by the reduced activity of soluble guanylate cyclase. It is now well established that cells in culture can maintain an altered phenotype, reflecting their differences in vivo, for several passages. A pertinent example from our own work is the finding that human umbilical vein endothelial cells cultured from umbilical cords after diagnosis of maternal gestational diabetes maintain significant alterations in nutrient transport and nitric oxide–mediated cGMP production for at least 3 passages.20, 22, 23 We found no impairment of arginine uptake in cells from the preeclampsia and fetal growth restriction groups compared with control group–derived cells. Basal cGMP levels in cells from the fetal growth restriction group were significantly lower than in cells from the control group and were elevated in cells from the preeclampsia group. This elevation was most marked in cells from cords of babies born early, which may also reflect a response to more severe preeclampsia. Because arginine transport was not elevated in cells from the preeclampsia group, this factor cannot contribute to the raised nitric oxide–mediated basal and histamine-stimulated levels of cGMP that we found in this group. These raised levels could be the result of either an increase in nitric oxide synthase levels or an increase in nitric oxide synthase activity, or both, but not an overall change in the level of the enzyme in the umbilical vein. Consistent with this idea, we recently showed that umbilical vein endothelial cells derived from the preeclampsia group had an elevated level of basal cytosolic ionized calcium, but there was no alteration in the maximum levels achieved by internal store release in response to histamine.24 This could be reflected in vivo as an enhanced vasodilator response to flow. It is widely accepted that basal production of nitric oxide contributes to the maintenance of low vascular resistance in the fetoplacental circulation and that the umbilical cord vessels have the capacity to increase nitric oxide synthesis in response to a variety of endotheliumdependent agonists.9 Our results are therefore consistent with previous reports that, in pregnancies affected by pre-
Volume 184, Number 3 Am J Obstet Gynecol
Fig 1. Basal rate of cGMP production in umbilical vein endothelial cells in relation to gestational age. There is a negative correlation between cGMP levels and gestational age in preeclampsia (filled triangles) and a positive correlation between control (open squares) and fetal growth restriction (filled circles) pregnancies. Basal cGMP levels in fetal growth restriction were significantly lower than normal.
eclampsia, nitrite levels in fetal serum and amniotic fluid are increased and basal nitric oxide production is higher in the umbilical artery.9, 16 Immunostaining of endothelium-derived nitric oxide synthase in stem villous vessels and endothelium of the small vessels of the villi from the preeclampsia group has also been shown to be increased compared with that in the control group.25 In vitro studies have shown that nitric oxide production by normal endothelial cells can be increased more by exposure to maternal plasma from patients with preeclampsia than by exposure to plasma from normotensive pregnant women.26 Assuming that the as yet uncharacterized factors responsible for this effect are generated in the placenta or cross it, they would be candidates for causing the increased fetal nitric oxide production. We found no significant difference in overall mean basal or stimulated production of prostacyclin, between cells derived from the control, preeclampsia, and fetal growth restriction groups, which indicates that alterations in prostacyclin synthesis alone are unlikely to contribute to the abnormal fetoplacental blood flow found in preeclampsia and fetal growth restriction. In contrast to the findings in basal cGMP production, there was no trend toward altered prostacyclin production with either gestational age or newborn size in all groups. Preeclampsia and most cases of fetal growth restriction arise from the same pathologic change, impaired placentation, but what determines the subsequent course of a pregnancy is not understood. Pregnancies with evidence of impaired placentation (derived noninvasively by Doppler ultrasonography) may be complicated by either preeclampsia or delivery of a baby with fetal growth re-
Parra et al 501
Fig 2. Basal rate of cGMP production in umbilical vein endothelial cells in relation to size of newborn infant. In human umbilical vein endothelial cells from pregnancies with preeclampsia (filled triangles), basal cGMP accumulation was negatively correlated with fetal growth percentiles, whereas in normotensive pregnancies (open squares) there was no significant correlation. Basal cGMP production from fetal growth restriction–derived cells (filled circles) is also shown.
Fig 3. Basal prostacyclin accumulation in human umbilical vein endothelial cells from preeclampsia, fetal growth restriction, and control pregnancies at different gestational ages. SGA, Small for gestational age.
striction or by a combination of these conditions. Our finding of lower basal nitric oxide production by umbilical vein endothelial cells from babies with fetal growth restriction is consistent with reduced fetoplacental blood flow and a consequent inability to adapt to impaired placentation. In contrast, there was increased nitric oxide production in cells from the preeclampsia group. It has long been postulated that preeclampsia is a method of maternal adaptation in an attempt to improve uteropla-
502 Parra et al
cental perfusion. The implication from our data is that, in preeclampsia, the fetal circulation also adapts, trying to improve fetoplacental perfusion and the delivery of nutrients to the fetus. In summary, our results show that the contribution of nitric oxide production to vasomotor control of the fetal circulation depends on both the size and the gestational age of the fetus, as well as on the level of maternal hypertension, and they underline the importance of delineating preeclampsia and fetal growth restriction separately when one is trying to understand the changes that take place. Our findings support the hypothesis that increased nitric oxide production in the fetal circulation in preeclampsia is a compensatory response to fetoplacental ischemia or is caused by an increasing concentration of unknown circulating factors, whereas failure to increase nitric oxide production leads to impaired fetoplacental vasodilatation and fetal growth restriction. We thank Professor David Yudilevich for his invaluable support during this research. REFERENCES
1. Khong TY, De Wolf F, Brosens I. Inadequate material vascular response to placentation in pregnancies complicated by preeclampsia and by small-for-gestational age infants. BJOG 1986;93:1049-59. 2. Campbell S, Diaz-Recasens J, Griffin DR, Cohen-Overbeek TE, Pearce JM, Willson K, et al. New Doppler technique for assessing uteroplacental blood flow. Lancet 1983;1:675-7. 3. Remuzzi G, Misiani R, Muratore D, Marchesi D, Livio M, Schieppati A, et al. Prostacyclin and human foetal circulation. Prostaglandins 1979;18:341-8. 4. Gude NM, King RG, Brennecke SP. Role of endothelium-derived nitric oxide in maintenance of low fetal vascular resistance in placenta [letter; comment]. Lancet 1990;336:1589-90. 5. Schmidt HHHW, Lohmann SM, Walter U. The nitric oxide and cGMP signal transduction system: regulation and mechanism of action. Biochim Biophys Acta 1993;1178:153-75. 6. Brennecke SP, Gude NM, DiIulio JL, King RG. Reduction of placental nitric oxide synthase activity in pre-eclampsia. Clin Sci 1997;93:51-5. 7. Conrad KP, Davis AK. Nitric oxide synthase activity in placentae from women with pre-eclampsia. Placenta 1995;16:691-9. 8. Rutherford RAD, McCarthy A, Sullivan MHF, Elder MG, Polak JM, Wharton J. Nitric oxide synthase in human placenta and umbilical cord from normal, intrauterine growth-retarded and preeclamptic pregnancies. Br J Pharmacol 1995;116:3099-109. 9. Sladek SM, Magness RR, Conrad KP. Nitric oxide and pregnancy. Am J Physiol 1997;272:R441-63.
February 2001 Am J Obstet Gynecol
10. Buhimschi IA, Saade GR, Chwalisz K, Garfield RE. The nitric oxide pathway in pre-eclampsia: pathophysiological implications. Hum Reprod Update 1998;4:25-42. 11. Sala C, Campise M, Ambrose G, Motta T, Zanchetti A, Morganti A. Atrial-natriuretic-peptide and hemodynamic-changes during normal human-pregnancy. Hypertension 1995;25:631-6. 12. Pinto A, Sorrentino R, Sorrentino P, Guerritore T, Miranda L, Biondi A, et al. Endothelial-derived relaxing factor released by endothelial cells of human umbilical vessels and its impairment in pregnancy-induced hypertension. Am J Obstet Gynecol 1991;164:507-13. 13. Akar F, Ark M, Uydes BS, Soysal ME, Saracoglu F, Abacioglu N, et al. Nitride oxide production by human umbilical vessels in severe pre-eclampsia. J Hypertens 1994;12:1235-41. 14. Boccardo P, Soregaroli M, Aiello S, Noris M, Donadelli R, Lojacono A, et al. Systemic and fetal-maternal nitric oxide synthesis in normal pregnancy and pre-eclampsia. BJOG 1996;103:879-86. 15. Lyall F, Young A, Greer IA. Nitric-oxide concentrations are increased in the fetoplacental circulation in preeclampsia. Am J Obstet Gynecol 1995;173:714-8. 16. Lyall F, Greer IA, Young A, Myatt L. Nitric oxide concentrations are increased in the feto-placental circulation in intrauterine growth restriction. Placenta 1996;17:165-8. 17. Goeschen K, Henkel E, Behrens O. Plasma prostacyclin and thromboxane concentrations in 160 normotensive, hypotensive, and preeclamptic patients during pregnancy, delivery, and the postpartum period. J Perinat Med 1993;21:481-9. 18. Friedman SA. Preeclampsia: a review of the role of prostaglandins. Obstet Gynecol 1988;71:122-37. 19. Polkinghorne J. Review of the guidance on the research use of fetuses and fetal material. London: Her Majesty’s Stationery Office; 1989. p. Cm762. 20. Sobrevia L, Cesare P, Yudilevich DL, Mann GE. Diabetes-induced activation of system y+ and nitric oxide sythase in human endothelial cells: association with membrane hyperpolarization. J Physiol 1995;489:183-92. 21. Toothill VJ, Needham L, Gordon JL, Pearson JD. Desensitization of agonist-stimulated prostacyclin release in human umbilical vein endothelial cells. Eur J Pharmacol 1988;157:189-96. 22. Sobrevia L, Yudilevich DL, Mann GE. Elevated D-glucose induces insulin insensitivity in human umbilical endothelial cells isolated from gestational diabetic pregnancies. J Physiol (Lond) 1998;506(Pt 1):219-30. 23. Sobrevia L, Jarvis SM, Yudilevich DL. Adenosine transport in cultured human umbilical vein endothelial cells is reduced in diabetes. Am J Physiol 1994;267(1 Pt 1):C39-47. 24. Steinert JR, Jacob R, Poston L, Mann GE. Effects of arachidonic acid on intracellular calcium ([Ca2+]i) metabolism in human endothelial and smooth muscle cells from normal and preeclamptic pregnancies. J Soc Gynecol Investig 1999;195P:583P. 25. Myatt L, Eis ALW, Brockman DE, Greer IA, Lyall F. Endothelial nitric oxide synthase in placental villous tissue from normal, preeclamptic and intrauterine growth restricted pregnancies. Hum Reprod 1997;12:167-72. 26. Davidge ST, Baker PN, Roberts JM. NOS expression is increased in endothelial cells exposed to plasma from women with preeclampsia. Am J Physiol 1995;269:H1106-12.
Key words: Fetal growth restriction, human endothelial cells, nitric oxide, preeclampsia, prostacyclin
In normal pregnancy progressive trophoblast invasion transforms the high-resistance uteroplacental spiral arteries into a low-resistance circulation. This vascular transformation facilitates the 10-fold increase in uteroplacental blood flow that occurs between conception and term. Histologic studies have shown that the process of spiral artery vascular transformation is incomplete in pregnancies affected by preeclampsia and fetal growth restriction.1 The resulting uteroplacental circulation is characterized by high resistance to flow in the maternal uterine arteries, which can be detected noninvasively by the use of Doppler ultrasonography.2 Preeclampsia and fetal growth restriction are therefore both associated with relative placental hypoperfusion. This may stimulate a compensatory increase in prostacyclin and nitric oxide, which
From the Harris Birthright Research Centre for Fetal Medicine, King’s College Hospital, Denmark Hill,a and the Centre for Cardiovascular Biology and Medicine, King’s College London, Guy’s Campus.b Supported by The Fetal Medicine Foundation (registered United Kingdom charity 1037116) and Mideplan (Chile). Received for publication December 31, 1999; revised February 29, 2000; accepted July 20, 2000. Reprint requests: Kypros H. Nicolaides, MD, Harris Birthright Research Centre for Fetal Medicine, King’s College Hospital, Denmark Hill, London, United Kingdom SE5 8RX. Copyright © 2001 by Mosby, Inc. 0002-9378/2001 $35.00 + 0 6/1/110311 doi:10.1067/mob.2001.110311
are the main vasodilators of the human fetal-placental circulation.3, 4 Nitric oxide is synthesized from the amino acid L-arginine via the endothelial isoform of nitric oxide synthase and modulates vascular tone by activating soluble guanylate cyclase and elevating intracellular cyclic guanosine 3´,5´-monophosphate (cGMP) levels in target smooth muscle cells.5 There is disagreement about placental nitric oxide production in preeclampsia and fetal growth restriction, with studies reporting a decrease, no change, or an increase, which may reflect differences in the methods of assessment, such as levels of nitric oxide synthase messenger ribonucleic acid, immunoreactivity of nitric oxide synthase, conversion from arginine to citrulline, and levels of basal or agonist-stimulated cGMP.6-10 Study results may additionally be confounded because placental nitric oxide production is highest early in normal pregnancy and falls with increasing gestational age.8 It is also possible that alterations in cGMP levels reflect the action of atrial natriuretic peptide rather than nitric oxide.11 For isolated fetal vessels or endothelial cells derived from them and cultured in vitro, there are fewer data on nitric oxide production, but they are again conflicting. Pinto et al12 reported a decrease in the release of endothelium-derived vasodilator substances from perfused vessels in preeclampsia, whereas Akar et al13 used a similar technique and found raised basal nitric oxide re497
498 Parra et al
lease (but unaltered agonist-stimulated release). Boccardo et al14 found no change in levels of nitric oxide synthase messenger ribonucleic acid or in the conversion of citrulline to arginine. However, higher circulating levels of fetal nitrite and nitrate and of amniotic fluid have been reported in both preeclampsia and fetal growth restriction.15, 16 Prostacyclin synthesis is initiated by the enzyme phospholipase A2, which liberates arachidonic acid from membrane phospholipids. Arachidonic acid is converted by cyclooxygenase into prostaglandin endoperoxides, and prostacyclin synthase subsequently forms prostacyclin (prostaglandin I2) from the endoperoxide prostaglandin H2. Goeschen et al17 showed that the maternal plasma concentration of prostacyclin in severe preeclampsia, in comparison with the concentration in normal pregnancies, was significantly lower, whereas the concentration of the vasoconstrictor eicosanoid thromboxane A2 was significantly higher. Fetal blood vessels and placenta synthesize prostacyclin, and there is a consensus view that, as in the mother, the production of prostacyclin is lower and the production of thromboxane A2 is higher in fetal tissues from women with preeclampsia.18 In this study we took for culture the early-passage human umbilical vein endothelial cells during a range of gestational ages in the third trimester of healthy pregnancies and of pregnancies affected by preeclampsia and fetal growth restriction. To assess the relative contributions of these vasodilators, we compared rates of transport of the nitric oxide precursor arginine and levels of basal and agonist-stimulated nitric oxide and prostacyclin synthesis in these cells. Material and methods Subjects and definitions. We studied cells cultured from 57 human umbilical cords taken from women delivered in our institution during a 20-month period in 1997 and 1998. Collection of the cords was according to the Polkinghorne recommendations19 on the research use of fetal material. Cords were from 18 pregnancies complicated by preeclampsia, from 9 pregnancies with fetal growth restriction, from 14 otherwise normal pregnancies with spontaneous preterm delivery, and from 16 normal pregnancies with delivery at term. Preeclampsia was defined as indicated by blood pressure >140/90 mm Hg measured on 2 occasions at least 6 hours apart and a urinary protein concentration >300 mg/24 h or ≥+1 on dipstick test, with no known history of hypertension or renal disease before pregnancy. Fetal growth restriction was defined as birth weight at <10th percentile for gestational age and sex. None of the fetal growth restriction cases in this study were complicated by preeclampsia, and neither were preeclampsia cases complicated by fetal growth restriction. No women were taking aspirin or were nitric oxide donors at the time of delivery.
February 2001 Am J Obstet Gynecol
Umbilical cord cells were cultured, and depending on the quantity of endothelial cells obtained, they were studied for at least 2 of the 3 experiments—transport of arginine, basal and stimulated cGMP release, and basal and stimulated prostacyclin release. Isolation and primary culture of endothelial cells. Cells were isolated from the umbilical cord veins by a collagenase dispersion method within 24 to 48 hours of delivery. In brief, after the veins were cannulated and rinsed with phosphate-buffered saline solution, the endothelial cells were loosened by incubation with 0.1% collagenase for 10 minutes at 37°C. The cells were then washed out of the cord, collected by centrifugation, washed, and resuspended in 5 mL of culture medium, composed of medium 199 containing 10% (vol/vol) fetal calf serum, 10% (vol/vol) newborn calf serum, L-glutamine (3.2 mmol/L), 0.016% penicillin, 0.008% streptomycin, and sodium bicarbonate (20 mmol/L). The cells were incubated in a T-25 culture flask at 37°C in a 5% carbon dioxide–95% air atmosphere. Confluent primary cultures were then dispersed, subcultured with 0.05% trypsin plus 0.025% ethylenediaminetetraacetic acid, and split 1:3. Cells from passage number 2 or 3 were used in all experiments. Cell number and cell protein were determined in confluent monolayers as previously described.20 Transport of arginine. Confluent third-passage monolayers grown in 24-well culture trays were rinsed with warmed (37°C) Krebs solution (sodium chloride, 131 mmol/L; potassium chloride, 5.6 mmol/L; sodium bicarbonate, 25 mmol/L; sodium biphosphate, 1 mmol/L; Dglucose, 5 mmol/L; N-[2-hydroxyethyl]piperazine-N´-[2ethanesulfonic acid] [HEPES], 20 mmol/L; calcium chloride, 2.5 mmol/L; and magnesium chloride, 1 mmol/L [pH 7.4]) and preincubated for 60 minutes at 37°C in Krebs solution containing L-arginine, 100 µmol/L. Kinetics of arginine transport were measured for 30 seconds in endothelial cells incubated with increasing concentrations of [3H]arginine (0.015-1 mmol/L) as previously described.20 Kinetic constants were derived by fitting a Michaelis-Menten curve plus a linear component to the raw data. In all experiments uptake was terminated by removal of the uptake medium immediately before monolayers were rinsed 3 times with 200 µL of icecold stop solution (Krebs buffer solution containing unlabeled substrate, 10 mmol/L). Protein was determined by the bicinchoninic acid reagent, and radioactivity in formic acid cell digests was determined by liquid scintillation counting. Basal and histamine-stimulated cGMP and prostacyclin levels. Confluent monolayers in 24-well plates were preincubated for 15 minutes with Krebs solution (1 mL, 37°C) containing L-arginine, 100 µmol/L, and the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine, 0.5 mmol/L. After this period 500-µL aliquots of cell-free supernatant were removed from each well, and histamine
Parra et al 499
Volume 184, Number 3 Am J Obstet Gynecol
Table I. Clinical characteristics of patients with cultured umbilical vein endothelial cells Study group Variable Maternal age (y) Mean ± SD Range Primiparous (%) Gestational age at delivery (wk) Mean ± SD Range Birth weight (kg) Mean ± SD Range
Control (n = 30)*
Preeclampsia (n = 18)
Fetal growth restriction (n = 9)
29.7 ± 6.8 (17-39) 23
31.6 ± 6.1 (17-40) 28
29.3 ± 6.1 (21-38) 11
37.0 ± 3.6 (29-42)
37.1 ± 2.0 (32-40)
36.6 ± 2.2 (33-40)
2.97 ± 0.77 (1.61-4.34)
2.82 ± 0.56 (1.63-4.02)
2.06 ± 0.55 (1.27-2.86)†
*In control group, 14 deliveries were preterm. †P < .05, compared with preeclampsia and control groups.
solution (5 µL, concentrated 100 times; final concentration, 10 µmol/L) was added to the wells for a further 5 minutes at 37°C. The supernatants from before and after stimulation by histamine were used for analysis of basal and stimulated rates of prostacyclin release by radioimmunoassay of its stable metabolite, 6-keto-prostaglandin F1α. Cells were then placed on ice and incubated with 0.1N hydrochloric acid (500 µL per well for 60 minutes). Afterward the extracts were stored at –20°C for radioimmunoassay of cGMP after acetylation. To confirm that elevated cGMP levels were a result of the production of nitric oxide, we treated cells with the nitric oxide synthase inhibitor Nω-nitro-L-arginine methyl ester (L-NAME; 15 minutes, 100 µmol/L).20, 21 Protein was assayed with the bicinchoninic acid reagent. Statistical analysis. Results were analyzed by unpaired nonparametric methods (Mann-Whitney U and Wilcoxon tests), as appropriate; differences were considered significant if the P value was < .05. Trends were analyzed by linear regression. Results Patients’ baseline clinical details are shown in Table I. The groups were matched for maternal age and gestational age at delivery. As expected, the mean birth weight was significantly lower in the fetal growth restriction group than in the control and preeclampsia groups. The kinetics of L-arginine transport in cells from 9 control and 9 preeclampsia pregnancies, matched in range of gestational age, were compared. Monolayers were incubated with L-[3H]arginine, 0.15 to 1 mmol/L at 37°C, and initial rates of influx were determined for 30 seconds. A Michaelis-Menten hyperbola, together with a linear nonsaturable component, was fitted to mean influx values weighted for the reciprocal of the respective standard error. There was no significant difference in the characteristics of arginine uptake between the 2 groups. For the control group the Michaelis-Menten constant (Km) and
Table II. Arginine transport by umbilical vein endothelial cells from control, preeclampsia, and fetal growth restriction pregnancies Type of pregnancy
L-Arginine (240 µmol/L) (pmol [µg protein]–1 min–1)
Control (n = 14) Preeclampsia (n = 8) Fetal growth restriction (n = 4)
5.7 ± 0.8 5.1 ± 0.6 4.1 ± 0.9
Values are mean ± SEM.
maximum velocity values were 79 ± 8 µmol/L and 6.2 ± 0.3 pmol (µg protein)–1 min–1, respectively. For the preeclampsia group the values were 111 ± 4 µmol/L and 6.5 ± 0.1 pmol (µg protein)–1 min–1, respectively. When uptake of arginine was compared at a fixed concentration (240 µmol/L) in a series of experiments, no significant differences were observed between control, fetal growth restriction, and preeclampsia groups (Table II). The mean basal level of cGMP accumulation was significantly higher in cells from the preeclampsia group than from the control group, whereas, in contrast, it was significantly lower in cells from the fetal growth restriction group (Table III). When these data were examined more closely, it became evident that the highest levels in preeclampsia were associated with the lowest gestational ages, with a negative trend that just failed to reach statistical significance (n = 17; r = –0.44; P = .07; Fig 1). In contrast, basal rates of cGMP accumulation in control cells were positively correlated with gestational age (n = 29; r = 0.42; P = .02; Fig 1). Basal cGMP levels from fetal growth restriction–derived cells were lower than normal at all gestational ages (Fig 1). When basal cGMP levels were related to growth percentiles, levels in control cells were unrelated to newborn size, but a negative correlation was observed for the preeclampsia cases (n = 17; r = –0.46; P = .05; Fig 2). Histamine significantly increased cGMP production in
500 Parra et al
February 2001 Am J Obstet Gynecol
Table III. Basal and histamine-stimulated cGMP production in human umbilical vein endothelial cells from control, preeclampsia, and fetal growth restriction pregnancies Type of pregnancy Control (n = 29) Preeclampsia (n = 17) Fetal growth restriction (n = 8)
Basal cGMP (pmol [µg protein]–1 [5 min]–1)
Histamine- stimulated cGMP (pmol [µg protein]–1 [5 min]–1)
3.9 ± 0.5 9.8 ± 2.1* 1.8 ± 0.2†
10.4 ± 2.8 23.4 ± 4.7* 2.9 ± 1.0†
Fold stimulation by histamine 2.31 ± 0.34 2.95 ± 0.46 1.43 ± 0.14†
Values are mean ± SEM. *P < .05, compared with fetal growth restriction and control groups. †P < .05, compared with control and preeclampsia groups.
all cells (Table III). However, the increase was lowest in fetal growth restriction, where the basal levels were lowest, and highest in preeclampsia, where the basal levels were highest. Histamine-induced increases in cGMP were blocked in the presence of L-NAME, 100 µmol/L, in all 3 groups, confirming that the increases were a result of nitric oxide production (data not shown). The fold increase stimulated by histamine was similar in the control and preeclampsia groups but significantly lower in the fetal growth restriction group (Table III). The mean levels of basal and histamine-stimulated prostacyclin production were not significantly different between groups (Fig 3). In contrast to basal cGMP production, there was no relationship between basal prostacyclin synthesis and either gestational age or newborn size in cells from the control and preeclampsia groups. Moreover, no relationship between maximal histaminestimulated prostacyclin release and gestational age was found for any group. Comment We examined the ability of umbilical vein endothelial cells derived from healthy (control) pregnancies, pregnancies with preeclampsia, and pregnancies with fetal growth restriction in the third trimester to synthesize 2 important vasodilators, nitric oxide and prostacyclin. We measured basal rates of nitric oxide–mediated cGMP production and prostacyclin synthesis and the rates stimulated by a maximally effective dose of histamine. Differences in basal rates are likely to reflect intrinsic differences in the control of vasoactive mediator synthesis that are maintained for at least 2 or 3 passages in cell culture. The differences might be caused by alterations in the intracellular levels of important cofactors for enzyme activity (eg, cytosolic ionized calcium, which regulates both nitric oxide and prostacyclin synthesis) or substrates (eg, arginine for nitric oxide synthesis, which is acquired by uptake from the medium). Differences in agonist-stimulated rates might reflect alterations in histamine receptor levels or, more likely, might reflect differences in the total amount of nitric oxide synthase or the rate-limiting enzyme in the prostacyclin synthetic pathway (phospholi-
pase A2). The lower ability to synthesize cGMP in the fetal growth restriction group could also be caused by the reduced activity of soluble guanylate cyclase. It is now well established that cells in culture can maintain an altered phenotype, reflecting their differences in vivo, for several passages. A pertinent example from our own work is the finding that human umbilical vein endothelial cells cultured from umbilical cords after diagnosis of maternal gestational diabetes maintain significant alterations in nutrient transport and nitric oxide–mediated cGMP production for at least 3 passages.20, 22, 23 We found no impairment of arginine uptake in cells from the preeclampsia and fetal growth restriction groups compared with control group–derived cells. Basal cGMP levels in cells from the fetal growth restriction group were significantly lower than in cells from the control group and were elevated in cells from the preeclampsia group. This elevation was most marked in cells from cords of babies born early, which may also reflect a response to more severe preeclampsia. Because arginine transport was not elevated in cells from the preeclampsia group, this factor cannot contribute to the raised nitric oxide–mediated basal and histamine-stimulated levels of cGMP that we found in this group. These raised levels could be the result of either an increase in nitric oxide synthase levels or an increase in nitric oxide synthase activity, or both, but not an overall change in the level of the enzyme in the umbilical vein. Consistent with this idea, we recently showed that umbilical vein endothelial cells derived from the preeclampsia group had an elevated level of basal cytosolic ionized calcium, but there was no alteration in the maximum levels achieved by internal store release in response to histamine.24 This could be reflected in vivo as an enhanced vasodilator response to flow. It is widely accepted that basal production of nitric oxide contributes to the maintenance of low vascular resistance in the fetoplacental circulation and that the umbilical cord vessels have the capacity to increase nitric oxide synthesis in response to a variety of endotheliumdependent agonists.9 Our results are therefore consistent with previous reports that, in pregnancies affected by pre-
Volume 184, Number 3 Am J Obstet Gynecol
Fig 1. Basal rate of cGMP production in umbilical vein endothelial cells in relation to gestational age. There is a negative correlation between cGMP levels and gestational age in preeclampsia (filled triangles) and a positive correlation between control (open squares) and fetal growth restriction (filled circles) pregnancies. Basal cGMP levels in fetal growth restriction were significantly lower than normal.
eclampsia, nitrite levels in fetal serum and amniotic fluid are increased and basal nitric oxide production is higher in the umbilical artery.9, 16 Immunostaining of endothelium-derived nitric oxide synthase in stem villous vessels and endothelium of the small vessels of the villi from the preeclampsia group has also been shown to be increased compared with that in the control group.25 In vitro studies have shown that nitric oxide production by normal endothelial cells can be increased more by exposure to maternal plasma from patients with preeclampsia than by exposure to plasma from normotensive pregnant women.26 Assuming that the as yet uncharacterized factors responsible for this effect are generated in the placenta or cross it, they would be candidates for causing the increased fetal nitric oxide production. We found no significant difference in overall mean basal or stimulated production of prostacyclin, between cells derived from the control, preeclampsia, and fetal growth restriction groups, which indicates that alterations in prostacyclin synthesis alone are unlikely to contribute to the abnormal fetoplacental blood flow found in preeclampsia and fetal growth restriction. In contrast to the findings in basal cGMP production, there was no trend toward altered prostacyclin production with either gestational age or newborn size in all groups. Preeclampsia and most cases of fetal growth restriction arise from the same pathologic change, impaired placentation, but what determines the subsequent course of a pregnancy is not understood. Pregnancies with evidence of impaired placentation (derived noninvasively by Doppler ultrasonography) may be complicated by either preeclampsia or delivery of a baby with fetal growth re-
Parra et al 501
Fig 2. Basal rate of cGMP production in umbilical vein endothelial cells in relation to size of newborn infant. In human umbilical vein endothelial cells from pregnancies with preeclampsia (filled triangles), basal cGMP accumulation was negatively correlated with fetal growth percentiles, whereas in normotensive pregnancies (open squares) there was no significant correlation. Basal cGMP production from fetal growth restriction–derived cells (filled circles) is also shown.
Fig 3. Basal prostacyclin accumulation in human umbilical vein endothelial cells from preeclampsia, fetal growth restriction, and control pregnancies at different gestational ages. SGA, Small for gestational age.
striction or by a combination of these conditions. Our finding of lower basal nitric oxide production by umbilical vein endothelial cells from babies with fetal growth restriction is consistent with reduced fetoplacental blood flow and a consequent inability to adapt to impaired placentation. In contrast, there was increased nitric oxide production in cells from the preeclampsia group. It has long been postulated that preeclampsia is a method of maternal adaptation in an attempt to improve uteropla-
502 Parra et al
cental perfusion. The implication from our data is that, in preeclampsia, the fetal circulation also adapts, trying to improve fetoplacental perfusion and the delivery of nutrients to the fetus. In summary, our results show that the contribution of nitric oxide production to vasomotor control of the fetal circulation depends on both the size and the gestational age of the fetus, as well as on the level of maternal hypertension, and they underline the importance of delineating preeclampsia and fetal growth restriction separately when one is trying to understand the changes that take place. Our findings support the hypothesis that increased nitric oxide production in the fetal circulation in preeclampsia is a compensatory response to fetoplacental ischemia or is caused by an increasing concentration of unknown circulating factors, whereas failure to increase nitric oxide production leads to impaired fetoplacental vasodilatation and fetal growth restriction. We thank Professor David Yudilevich for his invaluable support during this research. REFERENCES
1. Khong TY, De Wolf F, Brosens I. Inadequate material vascular response to placentation in pregnancies complicated by preeclampsia and by small-for-gestational age infants. BJOG 1986;93:1049-59. 2. Campbell S, Diaz-Recasens J, Griffin DR, Cohen-Overbeek TE, Pearce JM, Willson K, et al. New Doppler technique for assessing uteroplacental blood flow. Lancet 1983;1:675-7. 3. Remuzzi G, Misiani R, Muratore D, Marchesi D, Livio M, Schieppati A, et al. Prostacyclin and human foetal circulation. Prostaglandins 1979;18:341-8. 4. Gude NM, King RG, Brennecke SP. Role of endothelium-derived nitric oxide in maintenance of low fetal vascular resistance in placenta [letter; comment]. Lancet 1990;336:1589-90. 5. Schmidt HHHW, Lohmann SM, Walter U. The nitric oxide and cGMP signal transduction system: regulation and mechanism of action. Biochim Biophys Acta 1993;1178:153-75. 6. Brennecke SP, Gude NM, DiIulio JL, King RG. Reduction of placental nitric oxide synthase activity in pre-eclampsia. Clin Sci 1997;93:51-5. 7. Conrad KP, Davis AK. Nitric oxide synthase activity in placentae from women with pre-eclampsia. Placenta 1995;16:691-9. 8. Rutherford RAD, McCarthy A, Sullivan MHF, Elder MG, Polak JM, Wharton J. Nitric oxide synthase in human placenta and umbilical cord from normal, intrauterine growth-retarded and preeclamptic pregnancies. Br J Pharmacol 1995;116:3099-109. 9. Sladek SM, Magness RR, Conrad KP. Nitric oxide and pregnancy. Am J Physiol 1997;272:R441-63.
February 2001 Am J Obstet Gynecol
10. Buhimschi IA, Saade GR, Chwalisz K, Garfield RE. The nitric oxide pathway in pre-eclampsia: pathophysiological implications. Hum Reprod Update 1998;4:25-42. 11. Sala C, Campise M, Ambrose G, Motta T, Zanchetti A, Morganti A. Atrial-natriuretic-peptide and hemodynamic-changes during normal human-pregnancy. Hypertension 1995;25:631-6. 12. Pinto A, Sorrentino R, Sorrentino P, Guerritore T, Miranda L, Biondi A, et al. Endothelial-derived relaxing factor released by endothelial cells of human umbilical vessels and its impairment in pregnancy-induced hypertension. Am J Obstet Gynecol 1991;164:507-13. 13. Akar F, Ark M, Uydes BS, Soysal ME, Saracoglu F, Abacioglu N, et al. Nitride oxide production by human umbilical vessels in severe pre-eclampsia. J Hypertens 1994;12:1235-41. 14. Boccardo P, Soregaroli M, Aiello S, Noris M, Donadelli R, Lojacono A, et al. Systemic and fetal-maternal nitric oxide synthesis in normal pregnancy and pre-eclampsia. BJOG 1996;103:879-86. 15. Lyall F, Young A, Greer IA. Nitric-oxide concentrations are increased in the fetoplacental circulation in preeclampsia. Am J Obstet Gynecol 1995;173:714-8. 16. Lyall F, Greer IA, Young A, Myatt L. Nitric oxide concentrations are increased in the feto-placental circulation in intrauterine growth restriction. Placenta 1996;17:165-8. 17. Goeschen K, Henkel E, Behrens O. Plasma prostacyclin and thromboxane concentrations in 160 normotensive, hypotensive, and preeclamptic patients during pregnancy, delivery, and the postpartum period. J Perinat Med 1993;21:481-9. 18. Friedman SA. Preeclampsia: a review of the role of prostaglandins. Obstet Gynecol 1988;71:122-37. 19. Polkinghorne J. Review of the guidance on the research use of fetuses and fetal material. London: Her Majesty’s Stationery Office; 1989. p. Cm762. 20. Sobrevia L, Cesare P, Yudilevich DL, Mann GE. Diabetes-induced activation of system y+ and nitric oxide sythase in human endothelial cells: association with membrane hyperpolarization. J Physiol 1995;489:183-92. 21. Toothill VJ, Needham L, Gordon JL, Pearson JD. Desensitization of agonist-stimulated prostacyclin release in human umbilical vein endothelial cells. Eur J Pharmacol 1988;157:189-96. 22. Sobrevia L, Yudilevich DL, Mann GE. Elevated D-glucose induces insulin insensitivity in human umbilical endothelial cells isolated from gestational diabetic pregnancies. J Physiol (Lond) 1998;506(Pt 1):219-30. 23. Sobrevia L, Jarvis SM, Yudilevich DL. Adenosine transport in cultured human umbilical vein endothelial cells is reduced in diabetes. Am J Physiol 1994;267(1 Pt 1):C39-47. 24. Steinert JR, Jacob R, Poston L, Mann GE. Effects of arachidonic acid on intracellular calcium ([Ca2+]i) metabolism in human endothelial and smooth muscle cells from normal and preeclamptic pregnancies. J Soc Gynecol Investig 1999;195P:583P. 25. Myatt L, Eis ALW, Brockman DE, Greer IA, Lyall F. Endothelial nitric oxide synthase in placental villous tissue from normal, preeclamptic and intrauterine growth restricted pregnancies. Hum Reprod 1997;12:167-72. 26. Davidge ST, Baker PN, Roberts JM. NOS expression is increased in endothelial cells exposed to plasma from women with preeclampsia. Am J Physiol 1995;269:H1106-12.