Plasma From Women With Preeclampsia Increases Endothelial Cell Nitric Oxide Production
Abstract In preeclampsia, a factor in the maternal circulation alters endothelial function via a reduction in nitric oxide synthesis. We measured the in vitro effects of 2% plasma from women with preeclampsia, compared with 2% plasma from normotensive pregnant women, on cultured endothelial cell nitrite production and nitric oxide synthase activity. On finding differential effects, we measured the effects on cellular viability (assessed by lactate dehydrogenase levels) and performed a time course study. Endothelial cell nitrite production was found to be higher after exposure to plasma from the preeclamptic group than the normotensive pregnant group. The effects of long-term exposure (120 hours) were similar to those of short-term exposure (24 hours). In addition, nitric oxide synthase activity was significantly greater after exposure to preeclamptic plasma than after exposure to normotensive pregnant plasma. No differential effect on cellular viability was found. Contrary to our hypothesis, exposure of endothelial cells to preeclamptic plasma resulted in increased nitric oxide production and nitric oxide synthase activity.
One of the major features of preeclampsia is reduced perfusion of many organs.1 Reduced organ perfusion appears to be caused by vasoconstriction secondary to increased sensitivity to endogenous pressors2 and occlusion of small vessels by microthrombi produced by activation of the coagulation cascade. Activation or injury of vascular endothelial cells, consequent upon circulating agents that probably arise from the fetus or placenta, may well be responsible for these changes.3 Intact endothelium has anticoagulant properties and modifies the response of vascular smooth muscle to agonists; however, when activated, endothelial cells promote coagulation and vasoconstriction.4
Our hypothesis was that the blunted response to pressor agents in normotensive pregnancy and the augmented response in preeclampsia could be due to increased release of the vasodilator nitric oxide (NO) by endothelial cells during pregnancy and to reduced release in women with preeclampsia. Increased NO release is felt to be secondary to augmented synthesis by NO synthase (NOS), although it is possible that stable nitrosocompounds may function as an NO reservoir.5 6 7 The isoform of NOS in endothelial cells is expressed constitutively and is membrane bound and calcium sensitive.5 6 7 In addition to vasodilator effects, NO mediates other functions relevant to preeclampsia, including a reduction in platelet sensitivity to proaggregatory agents.5 In animal models suppression of NO synthesis results in thrombosis and infarction,8 and long-term blockade in pregnancy produces hypertension, proteinuria, and a suppression of plasma volume expansion.9 At present, there is little evidence from human studies to support the involvement of NO in preeclampsia. However, umbilical vessels from pregnancies complicated by preeclampsia produce less NO by bioassay than vessels obtained from infants of normotensive pregnant women.10 Moreover, in a comparison of three subjects with preeclampsia and three normotensive pregnant subjects, serum from subjects with preeclampsia contained a fraction that inhibited the ability of acetylcholine to induce relaxation of precontracted rabbit aortic rings.11
We therefore tested the hypothesis that in preeclampsia a factor or factors in the maternal circulation alter endothelial function via a reduction in NO synthesis. We studied the in vitro effects of plasma from preeclamptic subjects on nitrite production by cultured endothelial cells. We report that contrary to our hypothesis, exposure to plasma from subjects with preeclampsia increased NO production by increasing NOS activity.
We studied plasma to avoid the confounding effects of cellular products released into serum during blood coagulation. Samples were collected from 25 nulliparous pregnant women before delivery. Subjects were recruited by the same investigators from the Obstetric Service of the Medical Centre at the University of California–San Francisco (J.M.R.) and from the Magee-Womens Hospital, Pittsburgh, Pa (J.M.R., S.T.D., P.N.B.), using protocols approved by the respective hospital Ethics Committees. All subjects gave informed consent, and the procedures followed were in accordance with institutional guidelines. Fifteen subjects had preeclampsia, defined using the criteria of hypertension, proteinuria, hyperuricemia, and reversal of hypertension and proteinuria after pregnancy.12 Hypertension was defined as an increase of 30 mm Hg systolic or 15 mm Hg diastolic pressure compared with values obtained before 20 weeks of gestation, or an absolute blood pressure greater than or equal to 140/90 mm Hg after 20 weeks (if blood pressure recordings in the first half of pregnancy were unknown). All subjects were normotensive when reviewed in the puerperium. Proteinuria was defined as greater than 500 mg/24-hour collection, or greater than or equal to 2+ on a voided or greater than or equal to 1+ on a catheterized random urine specimen. Hyperuricemia was defined as greater than or equal to 1 SD above the normal mean concentration at term (≥0.33 mmol/L at term). Ten normotensive pregnant subjects were also recruited and matched for age (within 5 years) and race.
No subject was known to have chronic hypertension or renal or metabolic disease. All subjects gave informed consent for their inclusion in the study after explanation of the nature of the research. The characteristics of each group are detailed in the Table⇓.
Since samples were collected by the same investigative team, collection, preparation, and storage were similar in San Francisco and Pittsburgh. Briefly, samples were maintained at 26°C for 2 to 10 hours before centrifugation at 2000g for 20 minutes and then aliquoted under sterile conditions and stored at −80°C. We previously measured plasma thromboxane levels to check for platelet removal using this processing protocol and found barely detectable levels. There was no alteration in nitrite production consequent upon exposure to plasma with (1) time from venipuncture to centrifugation, (2) time of storage (up to 30 months), and (3) episodes of freezing and thawing (up to five episodes). Before the experiments below, all samples were frozen and thawed on two occasions. There were no differences between subjects from the San Francisco and Pittsburgh populations when any of the measured variables were compared (P>.2).
The endothelial cell culture was an endothelial cell line from a bovine coronary microvessel (B88; Gensia, Inc).13 B88 cells were grown on uncoated plastic culture dishes in α-minimum essential medium supplemented by 10% heat-inactivated horse serum, 2 mmol/L l-glutamine, 10 000 U/mL nystatin, 7 μmol/L gentamicin, and 34 μmol/L kanamycin at 37°C in 5% CO2. Cell cultures were dispersed with 0.05% trypsin and 0.53 mmol/L EDTA, plated in six-well dishes, and grown as confluent monolayers. Preliminary experiments, in which cell number was measured by a hemacytometer (Fisher), demonstrated that each well contained 106 cells when a confluent monolayer had been established. The consistency of the cell number of each well was confirmed by measurement of protein content with the Bradford technique.14 Experimental results are thus expressed per milligram of protein.
The cell monolayers were then made quiescent in 1 mL serum-free medium for 24 hours. This serum-free medium was replaced with an equal volume of heparinized plasma. Heparinized plasma (500 000 U/L plasma) from each subject was added to duplicate wells. It was necessary to heparinize the EDTA-prepared plasma to prevent the diluted samples from clotting when they were added to the cells. A preliminary experiment indicated that at this concentration heparin did not affect B88 NO production. In further preliminary experiments a 2% plasma concentration minimized the cytotoxic effects of plasma (as assessed by lactate dehydrogenase [LDH] levels) while maximizing relative differences between subject groups. This concentration therefore was used in the remainder of the study.
Nitrite production was determined by a colorimetric assay.15 An aliquot of medium (180 μL) from each culture well was mixed with 20 μL Greiss reagent (1% sulfanilamide and 0.1% naphthylethylenediamine dihydrochloride in 2% phosphoric acid). The mixture was incubated for 10 minutes at room temperature and the absorbance (optical density, 550 nm) measured in a Vmax kinetic microplate reader (Molecular Devices). Concentrations were determined by comparison with a standard solution of sodium nitrite in plasma-free medium. The reaction was linear from 0.25 to 64 μmol/L. The lower limit of detection (+3 SD of the zero standard) was 0.24 μmol/L. Results are expressed as nitrite production per 24 hours. In preliminary experiments nitrate levels were determined by the conversion of nitrate to nitrite with the addition of the enzyme nitrate reductase (50 U/mL, Boehringer Mannheim Biochemica).
Assay of NOS Activity
The conversion of [14C]arginine to [14C]citrulline was used as evidence of the formation of NO from l-arginine.16 Briefly, the B88 cells were harvested in homogenizing buffer containing 50 mmol/L Tris, 0.1 mmol/L EDTA, 0.1 mmol/L EGTA, 3 μmol/L leupeptin (Boehringer Mannheim Biochemica), 1 μmol/L pepstatin A (Sigma Chemical Co), 1 mmol/L phenylmethylsulfonyl fluoride (Sigma), and 12.5 mmol/L 2-mercaptoethanol. Particulate or cytosolic preparations (80 μL) were added to tubes prewarmed to 37°C containing 20 μL of a cocktail containing 2.04 μmol/L [14C]arginine, 5 μmol/L “cold” arginine, 1 mmol/L NADPH, 10 μmol/L tetrahydrobiopterin, 2 μmol/L flavin adenine dinucleotide, and either 2 mmol/L CaCl2 and 50 000 U/L calmodulin or 1 mmol/L EGTA. Incubations were at 37°C for 30 minutes before the reaction was terminated by the addition of 2 mL of stop buffer containing 40 mmol/L HEPES (pH 5.5) and 4 mmol/L EDTA at 4°C. The solution was then passed over a Dowex AG 50W-X8 column (1-mL), and 2 mL of water was used for removal of the remaining [14C]citrulline. The fall through of [14C]arginine was 2%, and the recovery of [14C]citrulline was 95%. The reaction was linear from 0.1 to 10 mg/mL of protein.
Cellular viability was assessed by measurement of LDH levels. The assay was based on the measurement of NADH formed from NAD plus lactate. Media was added to 500 μL reagent (Sigma diagnostic kit No. 228-UV). LDH was measured by the change in absorbance at 340 nm at 30°C with the use of a Spectronic Genesys 5 spectrophotometer (Milton Roy Co). For assessment of maximal LDH release, LDH concentrations were also measured after cells had been exposed to 0.05% Triton X-100 for 24 hours.
Demographic, clinical, and experimental data are presented as group mean±SEM and were compared with the Student’s t test, ANOVA with repeated measures, or the nonparametric Mann-Whitney U test. The correlation between nitrite concentration and NOS activity was assessed with Fisher’s r to z transformation. A value of P<.05 was taken as the level of statistical significance.
Clinical and demographic data from the two groups are summarized in the Table⇑. As anticipated from the definition criteria, systolic and diastolic pressures at term were significantly greater in the preeclamptic group than in the normotensive pregnant group (P<.001). Gestational ages at delivery and infant birth weights were significantly lower in the preeclamptic group compared with the normotensive pregnant group (P<.01). None of the infant birth weights in the normotensive pregnant group were below the 10th percentile for gestational age (small for gestational age); in contrast, five of the infants in the preeclamptic group were small for gestational age. No other differences between the groups were found.
Nitrite levels could not be detected in 2% plasma–containing media per se. In preliminary studies with nitrate reductase, nitrate levels were not detectable after incubation with several plasma samples, and in no case was the nitrate level greater than 10% of the total level of nitrite plus nitrate. There were no differences in nitrate levels between the two subject groups. In subsequent studies nitrate levels were not measured.
Nitrite production was measured in the cell media after exposure to 2% plasma from either subjects with preeclampsia (n=10) or normotensive pregnant women (n=10) for 24 hours. Nitrite production was significantly greater in cells exposed to plasma from subjects with preeclampsia (97.3±9.6 nmol/mg protein per 24 hours) than in cells exposed to plasma from normotensive pregnant women (71.9±4.3 nmol/mg protein per 24 hours, P<.05, as illustrated in Fig 1⇓). Within the preeclamptic subject group, there were no correlations between nitrite production and blood pressure, plasma urate concentration, platelet count, hematocrit, or infant birth weight.
Enzyme activity was found to be calcium sensitive, absolutely NADPH dependent, and in the particulate fraction; no activity was found in the cytosol. The levels of enzyme activity determined after the 24-hour incubation are illustrated in Fig 2⇓. NOS activity was significantly greater after exposure to plasma from subjects with preeclampsia (0.053±0.014 pmol/mg per minute) than after exposure to plasma from normotensive pregnant women (0.022±0.008 pmol/mg per minute, P<.05, Student’s t test or Mann-Whitney U test). There was a significant correlation between levels of nitrite production and NOS activity (r2=.60, P<.05).
Determination of Protein Content and Assessment of Cellular Viability With the Use of LDH Levels
LDH concentrations in 2% plasma in media per se were greater than 2.0 U/L. There was no significant difference in the protein content of the wells containing cells exposed to plasma from subjects with preeclampsia (0.049±0.005 mg) compared with cells exposed to plasma from normotensive pregnant women (0.053±0.004 mg).
There was also no significant difference in LDH levels of media from cells exposed to plasma from subjects with preeclampsia (1.37±0.12 U/μg protein) compared with media from cells exposed to plasma from normotensive pregnant women (1.56±0.11 U/μg protein). Total cellular LDH content, as indicated by LDH levels after cells had been exposed to 0.05% Triton X-100, was approximately 5.60 U/μg protein.
Time Course Study
We have previously reported that endothelial cell prostacyclin production is increased over 24 hours of exposure to plasma from subjects with preeclampsia but that long-term exposure (over 72 hours) to plasma from subjects with preeclampsia results in diminished production.13 To determine whether there was a difference between short- and long-term exposure when nitrite production was studied, we performed a time course experiment using plasma from all 25 subjects (15 subjects with preeclampsia, 10 normotensive pregnant women). Medium (180 μL) was removed for measurement of nitrite concentration after 24, 48, 72, and 120 hours, and an equal volume of 2% plasma was replaced on each occasion. Nitrite production was calculated by subtracting the amount of nitrite present in the culture well before the time interval from the total produced at the end of the time interval. The levels of nitrite production determined in the time course experiments are illustrated in Fig 3⇓. Nitrite production by cells exposed to plasma from subjects with preeclampsia remained greater than or equal to that of cells exposed to plasma from normotensive pregnant women and was found to be higher in the preeclamptic group when the overall time course data were analyzed with repeated-measures ANOVA (P<.05).
LDH levels after 120 hours of incubation at the end of the time course experiment were 80.7±1.8 U/L in the preeclamptic group and 88.6±3.5 U/L in the normotensive pregnant group.
Contrary to our initial hypothesis, exposure of bovine microvessel endothelial cells to plasma from subjects with preeclampsia increased NO production compared with the production of cells exposed to plasma from normotensive pregnant women. The greater variability of the data from the preeclamptic group typifies measurement of different variables in the condition.3 17 18 NOS activity was found to be calcium sensitive and in the particulate fraction, which are characteristics of the constitutive isoform of the enzyme. NOS activity was also greater in cells exposed to plasma from subjects with preeclampsia than in cells exposed to plasma from normotensive pregnant subjects. There are differences in endothelial cell prostacyclin production after long-term compared with short-term exposure to plasma from subjects with preeclampsia.13 Short-term exposure results in increased prostacyclin production compared with exposure to plasma from normotensive pregnant women, whereas prostacyclin production is reduced after long-term exposure.13 No such differences were found when endothelial cell nitrite production was considered; production remained elevated in cells exposed to plasma from subjects with preeclampsia throughout the time course experiment (although the possibility was not excluded that the proportion of nitrate to total nitrite plus nitrate production may vary with incubation time).
The working hypothesis of many groups, including our own, has been that the pathogenesis of preeclampsia may involve diminished NO production. However, the data regarding levels of plasma and urinary nitrites are conflicting. A small study of women with preeclampsia found decreased urinary and plasma nitrites compared with normotensive pregnant women.19 Two studies reported similar levels of urinary nitrites and nitrates in normotensive pregnant women compared with women with pregnancy-induced hypertension20 or preeclampsia,21 although in one of the studies an inverse relationship between urinary nitrite levels and the rise in systolic pressure was demonstrated.20 Differences in subject populations, disease definitions, and methods used, combined with a lack of dietary control and the use of random rather than 24-hour urine collections, presumably account for these discrepancies, which prevent any conclusions being drawn from these studies.
Our results do not suggest a role for reduced NO production in preeclampsia. However, the in vitro effect found in this study may be irrelevant to a consideration of the in vivo situation in preeclampsia. Extrapolation of in vitro findings to the in vivo situation is hazardous, particularly because these in vitro studies were performed in the absence of shear stress, a known stimulator of constitutive NOS in endothelial cells. Moreover, NO production by other cell types may contribute to the disease process. Marked increases in NO release by vascular myocytes are seen after cytokine or endotoxin administration.7 22 Differences in vascular myocyte NO production may contribute to the disease because of either a differential effect of preeclamptic compared with normotensive pregnant plasma or increased exposure of myocytes to plasma products after the endothelial disruption of preeclampsia.23 Alternatively, we may be merely presenting the results of a bioassay for a potent stimulator of enzyme expression such as estradiol,7 although with 2% plasma as the stimulant this seems unlikely. The hemoconcentration that occurs in preeclampsia could provide a further explanation for there being an increased effect if the plasma was from subjects with preeclampsia. However, a contribution of hemoconcentration does not seem likely in our study, in which hematocrit did not differ between the groups.
Although there were no differences in white blood cell counts between the two subject groups, it is also possible that disparities in absolute monocyte counts may have affected the results. Monocytes in EDTA for periods of 6 to 10 hours undergo morphological changes, with vacuolization that may result in the production or release of products such as cytokines. In future studies we will measure the differential white cell count. However, we have been unable to demonstrate any cytokine induction of NOS in B88 endothelial cells (unpublished data, 1994), suggesting that this mechanism is unlikely to account for the differences in nitrite production.
The increase in NO production and NOS activity did not seem to be secondary to nonspecific cytotoxicity, as LDH levels indicated that there were no differences in cellular viability after exposure to plasma from the two groups. However, plasma from subjects with preeclampsia has been found to increase in vitro endothelial cell prostaglandin production.24 25 Considerations of cytotoxicity are thus complicated by the known cytoprotective effect of prostaglandin26 that might mask any initial cytotoxic effect.
It is possible that the increased NO production and NOS activity after exposure to plasma from subjects with preeclampsia reflect the in vivo situation. NO reacts rapidly with oxygen and oxygen radicals to form peroxynitrite, which damages cellular components in a manner similar to other chemical oxidants,5 7 22 and increased levels of free radicals are generated in preeclampsia.3 Palmer et al27 suggested that NO may mediate the vascular endothelial damage that occurs in immunologically based conditions. However, cytotoxic levels of NO typically result from inducible NOS activity,27 and the bovine endothelial NOS was calcium sensitive and in the particulate fraction, which are characteristics of the constitutive isoform of the enzyme5 and in keeping with our inability to demonstrate any cytokine induction of NOS in these cells. Alterations in intracellular calcium may account for in vivo differences in enzyme activity. Elevated intracellular free calcium concentrations in platelets (used as a model of other cell types) have been reported in preeclampsia.18 However, at least part of the differences we have demonstrated were secondary to changes in NOS activity, as the arginine-to-citrulline NOS assay was performed in the presence of a calcium excess. Increased constitutive NOS activity could indicate increased NOS mass or, alternatively, the reversible agonist-induced phosphorylation of NOS,28 as has been described for bradykinin (implicated in preeclampsia29 ).
In summary, this study demonstrated increased endothelial NO production and increased NOS activity on exposure to plasma from subjects with preeclampsia. This finding was not dependent on incubation time. The mechanism for this finding and whether there are any implications for the in vivo situation are unclear. However, the possibility that excessive NO production rather than deficient production may account for altered endothelial function in preeclampsia merits further study.
This study was supported by National Institutes of Health grant HD-30367 and the Irene McLenahan Young Investigators’ Research Fund of the Magee-Womens Health Foundation. P.N.B. was the 1994 Wellbeing (Birthright)/American College of Obstetricians and Gynecologists Exchange Research Fellow. We gratefully acknowledge the help of Dr Jerzy Barankiewicz, Gensia, Inc, San Diego, Calif. We also thank the Clinical Data Core and the nursing staff of both the Magee-Womens Hospital, Pittsburgh, Pa, and the Medical Center at the University of California–San Francisco for invaluable help in sample collection. This article is dedicated to the late Zbigniew Rymaszewski, Department of Endocrinology, University of Cincinnati (Ohio), without whose collaboration this study would not have been possible.
- Received December 9, 1994.
- Revision received February 8, 1995.
- Accepted May 3, 1995.
Friedman SA, Taylor RN, Roberts JM. Pathophysiology of pre-eclampsia. Clin Perinatol. 1991;4:661-682.
Talledo OE, Chesley LC, Zuspan FP. Renin-angiotensin system in normal and toxemic pregnancies. Am J Obstet Gynecol. 1968;100:218-221.
Rodgers GM. Hemostatic properties of normal and perturbed vascular cells. FASEB J. 1988;2:116-123.
Knowles RG, Moncada S. Nitric oxide synthases in mammals. Biochem J. 1994;298:249-258.
Billiar TR, Curran RD, Harbrecht BG, Steuhr DJ, Demetris AJ, Simmons RL. Modulation of nitrogen oxide in vivo. J Leukoc Biol. 1990;48:565-569.
Baylis C, Engels K. Adverse interactions between pregnancy and a new model of systemic hypertension produced by chronic blockade of endothelial derived relaxing factor (EDRF) in the rat. Clin Exp Hypertens Preg. 1992;B11:117-129.
Pinto A, Sorrentino R, Sautebin L, Miranda L, Sorrentino L. Identification and partial purification of an endothelium-derived relaxing factor-inhibitor factor(s) in serum of hypertensive pre-eclamptic parturients. In: Biology of Nitric Oxide. London, UK: Portland Press Proceedings; 1992:373-376.
Baker PN, Davidge ST, Roberts JM. Plasma from preeclamptic women initially stimulates and then inhibits endothelial prostacyclin production. Hypertension. 1994;24:383. Abstract.
Bredt DS, Snyder SH. Isolation of nitric oxide synthase: a calmodulin-requiring enzyme. Biochemistry. 1990;87:682-685.
Baker PN, Broughton Pipkin F. Screening tests for pregnancy-induced hypertension. In: Studd J, ed. Progress in Obstetrics and Gynaecology. Edinburgh, UK: Churchill Livingstone; 1993;10:69-86.
Lopez-Jaramillo P, Narvaez M, Felix C, Sosa C, Calle A, Nava E, Palmer RMJ, Moncada S. Decreased nitric oxide synthesis in Andean Ecuadorian women with pregnancy-induced hypertension. In: Biology of Nitric Oxide. London, UK: Portland Press Proceedings; 1992:376.
Buchi W, Ignarro LJ, Chaudhuri G. The role of nitric oxide (NO) in the pathogenesis of pre-eclampsia. In: Scientific Program for the Society for Gynecologic Investigation; 1993:169. Abstract S202.
Nathan C. Nitric oxide as a secretory product of mammalian cells. FASEB J. 1992;6:3051-3064.
Branch DW, Dudley DJ, LaMarche S, Mitchell MD. Sera from preeclamptic patients contain factor(s) that stimulate prostacyclin production by human endothelial cells. Prostaglandins Leukot Essent Fatty Acids. 1992;45:1991-1995.
Michel T, Li GK, Busconi L. Phosphorylation and subcellular translocation of endothelial nitric oxide synthase. Proc Natl Acad Sci U S A. 1993;90:6252-6256.