Benign and malignant breast disease: initial study results of serum and breast fluid analyses of endogenous estrogens.

Design, methods, and study population of a long-term multidisciplinary investigation of benign and malignant breast disease were reported. This initial report focused on the relation of menstrual, reproductive, and other factors to serum and breast fluid estrogen measures [estradiol (E2), estrone (E1), percent free estrogen, and sex hormone binding globulin] among control women. After adjustment for the factors found to be related to the various estrogen measures, estrogen levels in women with benign and malignant disease were compared to those of controls. Findings were as follows: a) little evidence of any relation of most breast cancer risk factors with the various serum estrogen parameters studied; b) differences in breast fluid estrogen levels that may be relevant to the protective effect of parity on breast cancer risk; c) markedly higher levels of E2 and E1 in breast fluid than in serum and no evidence of a correlation of serum with breast fluid measures; d) no support for the hypothesis that breast cancer patients have higher serum percent free E2 than controls or women with benign breast disease; and e) higher breast fluid E2 and E1 levels in women with biopsied benign breast disease than in controls.

Control of cell proliferation: evidence for negative control on estrogen-sensitive T47D human breast cancer cells.

The human breast tumor cloned cell lines T47D-A8 and All are estrogen dependent for cell proliferation in the nude mouse model. In contrast, these cells multiplied at similar rates when grown in serum-free cultures, regardless of the presence of 17 beta-estradiol (3 X 10(-11) to 3 X 10(-8) M estradiol). Addition of 10% charcoal-dextran stripped human female serum to the culture medium resulted in a marked inhibition of cell proliferation. The addition of 3 X 10(-11) M estradiol overcame the inhibitory effect of serum. Similar results were obtained with the human breast tumor C7MCF7 cell line. Both cell lines contain similar estrophilin levels. The Kd of the estrophilin-estradiol complex was 0.39 X 10(-10) M for C7MCF7 cells and 4.4 X 10(-10) M for T47D-A11 cells. Maximal cell yields were achieved at 5 X 10(-12) M free estradiol levels in 10% charcoal-dextran stripped serum supplemented medium. These data are compatible with the following interpretation: (a) estradiol-sensitive cells are inhibited from proliferating by a serum-borne factor; and (b) estradiol neutralizes this inhibitory effect. This mechanism seems not to be mediated by estradiol binding to the cellular estrophilins because (a) the free estradiol levels needed for maximal response are significantly lower than the estrophilin Kds, and (b) maximal proliferation rates occur at similar estradiol concentrations for these three cell lines, regardless of the binding properties of their estrophilins.

Altered distribution and excretion of N-1-methylnicotinamide in rats with Walker 256 carcinosarcoma.

Levels of nicotinamide and N-1-methylnicotinamide in serum, liver, and kidney as well as renal clearances and 24-hr urine levels of N-1-methylnicotinamide were compared in normal rats and rats bearing Walker 256 tumors. There was no significant difference between normal and tumor-bearing rats with regard to nicotinamide levels. With regard to N-1-methylnicotinamide, tumor-bearing rats had significantly lower serum and liver levels and significantly higher 24-hr urine levels and renal clearances. Walker 256 tumor tissue and liver and kidney from a normal and a tumor-bearing rat were separately examined for S-adenosylmethionine:nicotinamide methyltransferase activity. The specific activity in tumor tissue extract was greater than that in each liver extract, which, in turn, was much greater than the specific activity in each tissue (liver and kidney) from the tumor-bearing rat was equal to the specific activity in the corresponding tissue of the normal rat. S-adenosylmethionine:nicotinamide methyltransferase was obtained with 18-fold purification from a tissue extract of Walker 256 tumor. The enzyme activity required activation by thiols, and maximal activity was observed at pH 8.6. The Km's for the substrates, S-adenosylmethionine and nicotinamide, were 7.0 x 10--3 mM and 0.50 mM respectively. The Ki's for the products, S-adenosylhomocysteine and N-1-methylnicotinamide, were respectively, 25 x 10--3 mM and greater than 5 mM.

Regulation of the cell cycle by diethylstilbestrol and progesterone in cultured endometrial cells.

Cultures of primary explanted rabbit endometrium epithelial cells were found to be a mixture of dividing or cycling (20%) and non-dividing or GO (80%) cells. The addition of diethylstilbestrol had the effect of switching most of the GO cells into the cell cycle and to shorten the length of the cell cycle through a shortening of G1 and S. Progesterone had the opposite effects.

Cortisol levels, binding, and properties of corticosteroid-binding globulin in the serum of primates.

New World primates have exceptionally high plasma levels of cortisol and other steroid hormones when compared with humans and other primates. It has been suggested that this difference can be explained by either low affinity or concentration of cellular steroid receptors. We have assessed cortisol availability in serum from several species of New and Old World primates under physiological conditions (whole serum at 37 degrees C). Measurements were made of total and free cortisol, corticosteroid-binding globulin (CBG) binding capacity and affinity for cortisol, distribution of cortisol in serum, and its binding to albumin. In agreement with earlier reports, plasma free cortisol levels in Old World primates, prosimians, and humans range from 10-300 nM. However, very high total plasma cortisol together with low CBG binding capacity and affinity result in free cortisol concentrations of 1-4 microM in some New World primates (squirrel monkey and marmosets) but not in others such as the titi and capuchin. In squirrel monkeys, free cortisol levels are far greater than might be predicted from the affinity of the glucocorticoid receptor estimated in cultured skin fibroblasts. In addition to low affinity, CBG from squirrel monkeys and other New World primates exhibits differences in electrophoretic mobility and sedimentation behavior in sucrose density ultracentrifugation, suggestive of a molecular weight that is approximately twice that of CBG from other species. Together with other data these results indicate that the apparent glucocorticoid resistance found in New World primates is a complex phenomenon that is not easily explained by present concepts of glucocorticoid action.

Structure and function of the hepatic form of 11 beta-hydroxysteroid dehydrogenase in the squirrel monkey, an animal model of glucocorticoid resistance.

Both cortisol and aldosterone bind to and activate the mineralocorticoid receptor. Cortisol concentrations are generally 100- to 200-fold higher than aldosterone concentrations, yet mineralocorticoids clearly exert effects different from glucocorticoids. One hypothesis is that 11 beta-hydroxysteroid dehydrogenase (11 beta-HSD), which converts cortisol to biologically inactive cortisone, protects the mineralocorticoid receptor from cortisol. The circulating concentrations of cortisol in the squirrel monkey are 20- to 50-fold higher than human cortisol concentrations, yet this animal has no evidence of glucocorticoid or mineralocorticoid excess. We used this experiment of nature to test the hypotheses that the known (hepatic) form of 11 beta-HSD protects renal mineralocorticoid receptors from the action of cortisol and that it modulates glucocorticoid concentrations in target tissues. Using a long oligonucleotide based on the rat sequence, we cloned the squirrel monkey 11 beta-HSD complementary DNA and gene. The encoded monkey amino acid sequence is 75% and 91% identical to the corresponding rat and human sequences, respectively. The tissue abundance of the messenger RNA for the monkey enzyme was similar to or less than that seen for the rat and human enzymes. Both the monkey and human 11 beta-HSD complementary DNAs were cloned into an expression vector and used to transfect cultures of Chinese hamster ovary cells. Both vectors were transcribed and translated into equivalent amounts of 11 beta-HSD enzyme. The monkey enzyme was slightly more efficient than the human enzyme in converting [3H]cortisol to cortisone, and estimates of the Michaelis-Menten constant and maximum velocity of both enzymes are similar. These data indicate that the abundance and activity of the hepatic form of 11 beta-HSD are insufficient to inactivate the very high concentrations of cortisol in the squirrel monkey, suggesting that this form of 11 beta-HSD does not defend the mineralocorticoid receptor or protect tissues from high cortisol concentrations. Rather, this enzyme appears to favor conversion of cortisone to cortisol, thus maximizing tissue concentrations of cortisol to overcome glucocorticoid resistance associated with a 50% reduction in glucococorticoid receptors.

Conservation of free but not total or non-sex-hormone-binding-globulin-bound testosterone in serum from Nagase analbuminemic rats.

Nagase analbuminemic rats have normal reproductive capacity, normal apparent libido, and normal serum concentrations of LH and FSH. Therefore, it is reasonable to assume that intracellular sex steroid hormone concentrations are normal or at least adequate to maintain normal reproductive function in these rats. To test whether intracellular testosterone concentrations in these rats are maintained by the circulating concentration of free or free-plus-weakly-bound testosterone, we measured the concentrations of total testosterone, free testosterone, and non-sex-hormone-binding-globulin-bound testosterone in sera from five adult male Nagase analbuminemic rats and from five age- and sex-matched controls. We found that the analbuminemic rats had markedly decreased serum concentrations of total and non-sex-hormone-binding-globulin-bound testosterone, but normal serum concentrations of free testosterone. These results suggest that intracellular concentrations of testosterone in biologically relevant organs of the rat are maintained by the concentration of free rather than free-plus-weakly-bound testosterone in plasma, in accord with the free hormone hypothesis.

Rapid cortisol and corticosteroid-binding globulin responses during pregnancy and after estrogen administration in the squirrel monkey.

Plasma cortisol and corticosteroid-binding globulin (CBG) levels were assessed in pregnant squirrel monkeys and in intact and castrated males after estrogen administration. Pregnant females showed a rapid and dramatic rise in cortisol and CBG during the first 8 weeks after conception. Estrogen treatment also caused marked elevations in cortisol and CBG. Cortisol levels increased significantly by 24 h after estrogen injection and remained elevated for 6 weeks of treatment, but a relatively greater rise in CBG resulted in a higher CBG/cortisol ratio. The data support prior research indicating that estrogen can simultaneously stimulate adrenal output and the compensatory binding of circulating cortisol by increased CBG synthesis. In addition, it appears that even in the absence of exogenous treatment, the pituitary-adrenal axis of male squirrel monkeys is stimulated by estrogen derived either from the testes or by the peripheral conversion of testosterone to estrogen.

Uptake of cortisol by the perfused rat liver: validity of the free hormone hypothesis applied to cortisol.

The mechanism by which cortisol in plasma enters hepatic cells was investigated using the isolated perfused rat liver. To determine whether hepatic uptake of cortisol from serum can be accounted for entirely by the pool of unbound (free) cortisol, we compared observed uptake rates with the equilibrium-free fraction of cortisol in serum and the rates of dissociation of cortisol from its serum binding proteins (determined using a rapid filtration assay based on transfer of [3H] cortisol to dextran-coated charcoal). More than 95% of the cortisol in both human and rat serum dissociated spontaneously from its binding proteins within 5 sec at 37 C. The fractional unidirectional hepatic uptakes of cortisol from pooled human serum and pooled rat serum were 59.4 +/- 5.4% and 59.5 +/- 1.0% (mean +/- SE), respectively, at the physiological flow rate of 1 ml/min.g liver. The corresponding free cortisol fractions in these sera were 4.53 +/- 0.15% and 8.16 +/- 0.23%, respectively. The fractional unidirectional hepatic uptake of cortisol from protein-free buffer averaged 99.9% (n = 5) at a flow rate of 3 ml/min.g liver. By calculating the appropriate rate constants and applying the Kety-Renkin-Crone equation to the above data, it can be shown that all of the cortisol taken up from serum by the perfused rat liver can be accounted for by the pool of free cortisol, which turns over very rapidly. The physiological significance of this finding is discussed in terms of a general mathematical model of hormone transport that delineates the conditions under which the free hormone hypothesis is and is not valid.

Squirrel monkey corticosteroid-binding globulin: primary structure and comparison with the human protein.

Squirrel monkey (Saimiri sciureus) corticosteroid-binding globulin (CBG) is the product of a 1.6-kilobase mRNA in the liver. Analyses of two overlapping cDNAs revealed that the squirrel monkey CBG precursor comprises 406 amino acids, the first 22 residues of which exhibit 91% identity with the human CBG leader sequence. The mature form of squirrel monkey CBG, therefore, very likely comprises 384 amino acids and has a polypeptide mol wt of 42,854. Compared to human CBG, the squirrel monkey protein contains an additional residue (threonine) at position 144, and the two proteins exhibit 86% sequence identity if this is taken into account. Squirrel monkey CBG contains five consensus sites for N-glycosylation, four of which are located in analogous positions in human CBG, and has two cysteine residues in the same relative positions as the cysteines in human CBG. Unlike CBG in most other species, squirrel monkey CBG appears to circulate as a dimer, and its affinity for glucocorticoids is remarkably low. We, therefore, expressed cDNAs for human and squirrel monkey CBGs in Chinese hamster ovary (CHO) cells and compared the physico-chemical properties of the products with those of the corresponding serum proteins. Squirrel monkey CBG is produced by CHO cells as a dimer, and its subunit size heterogeneity is similar to that associated with CBG in serum. In addition, the cortisol-binding affinity of squirrel monkey CBG produced by CHO cells is similar to that of the natural protein and is 5- to 8-fold lower than that of natural or recombinant human CBG. Mutants in which a threonine at position 144 was either added to human CBG or subtracted from squirrel monkey CBG were also expressed in CHO cells. This demonstrated that this additional amino acid in the squirrel monkey CBG sequence may actively contribute to its propensity for spontaneous dimerization, but does not account for its relatively low steroid-binding affinity.

Free fatty acids do not influence the concentrations of free steroid hormones in serum under physiological conditions.

In several recent studies it has been suggested that FFA may influence the concentrations of unbound steroid hormones in serum, but the experimental design of these studies has been questioned. We have reexamined the effects of oleic acid on the unbound concentrations of several steroid hormones in serum, including cortisol, testosterone, and estradiol. The results demonstrate that under physiological conditions, oleic acid does not affect the unbound concentrations of these hormones when assays are carried out with whole serum.

Interleukin-6 secretion in vitro is up-regulated in ectopic and eutopic endometrial stromal cells from women with endometriosis.

An in vitro model developed to compare human endometrial and endometriosis stromal cells was used to examine basal and stimulated expression of interleukin (IL-6). Stromal cells isolated from normal endometrium (NE) exhibited the lowest level of IL-6 secretion (84 pg/10(6) cells-48 h), whereas those cells isolated from endometriosis implants (EI) secreted the highest concentration of this inflammatory cytokine (46,284 pg/10(5) cells-48 h; P < 0.01). Eutopic endometrial stromal cells from women with endometriosis (EE) expressed an intermediate concentration of IL-6 (831 pg/10(6) cells-48 h). Stimulation of the various cultures with IL-1 beta dramatically augmented stromal cell production of IL-6. The mean concentrations of stimulated IL-6 secretion were 16,257, 37,800, and 264,290 pg/10(5) cells-48 h for NE, EE, and EI cells, respectively (P < 0.03). Exposure of the cell cultures to 10 nmol/L estradiol had little direct effect on IL-6 production. The results indicate that endometrial stromal cells isolated from tissues of women with and without endometriosis express IL-6 under basal and cytokine-stimulated conditions. Differential responsiveness among the three cell sources indicates that NE, EE, and EI cells have intrinsic quantitative differences in cytokine regulation.

Elevated nonesterified fatty acid concentrations in severe preeclampsia shift the isoelectric characteristics of plasma albumin.

We previously hypothesized that the endothelial cell dysfunction observed in women with preeclampsia might be caused by an imbalance between circulating very low density lipoproteins and a cytoprotective pI 5.6 isoform of albumin, referred to as toxicity preventing albumin (TxPA). An accurate simplified method was developed to quantify TxPA in small volumes of pregnancy plasma by gel electrofocusing. This assay revealed that circulating TxPA concentrations in women with severe preeclampsia were significantly reduced compared to those in normal pregnant women and women with benign transient hypertension of pregnancy. Nonesterified fatty acids (NEFA) and triglycerides were elevated in plasma from women with severe preeclampsia compared to those in plasma from the two control groups. The inverse correlation between TxPA and NEFA values led us to analyze the NEFA bound to plasma albumin. Gas chromatography and mass spectrometry demonstrated no qualitative differences in the specific fatty acids bound to plasma albumin in severe preeclamptic and normal pregnant women. However, the quantity of NEFA bound to albumin was greater in preeclampsia plasma (2.5 mol NEFA/mol albumin) compared to that in normal pregnancy plasma (0.8 mol NEFA/mol albumin), accounting for the acidic pI shift observed in albumin from the former patients. Functional assays demonstrated that human very low density lipoprotein particles were toxic to human umbilical vein endothelial cells in vitro, but this toxicity was prevented by the addition of TxPA albumin to the culture medium.

Isolation and characterization of the three Waxy genes encoding the granule-bound starch synthase in hexaploid wheat.

Complete genomic DNA sequences of three homoeologous Waxy structural genes, located on the chromosomes 7A, 4A, and 7D in hexaploid wheat (Triticum aestivum L. cv. Chinese Spring), were separately determined and analyzed. Those structural genes in lengths from start to stop codon were 2781bp in Wx-7A, 2794bp in Wx-4A, and 2862bp in Wx-7D, each of which consisted of 11 exons and ten introns. They were closely similar to one another in the nucleotide sequences, with 95.6-96.3% homology in mature protein regions, 88. 7-93.0% in transit-peptide regions, and 70.5-75.2% in the introns. These wheat Waxy genes were GC-rich when compared with standard values for plant genomes reported so far. This was reflected in the extremely high G/C occupation frequency at the third position of the codons in the coding regions. The sequence divergence in the exon regions was mostly due to the substitution of nucleotides, whereas that found in the introns was attributed to substitution, insertion and/or deletion of nucleotides. Only the Wx-4A gene contained a trinucleotide insertion (CAA) in the region encoding the transit peptide. Most of the substitutions observed in the exon regions were categorized as synonymous, and higher sequence similarities (96.5-97. 4%) were conserved at the protein level. The phylogenetic tree obtained in terms of the amino acid sequence variations showed a well-resolved phylogenetic relationship among wheat Waxy genes and those from other plants.

Downregulation of protein kinase C by phorbol ester increases expression of epidermal growth factor receptors in transformed trophoblasts and amplifies human chorionic gonadotropin production.

Epidermal growth factor (EGF) and its homologue, transforming growth factor-alpha (TGF-alpha), regulate human chorionic gonadotropin (hCG) synthesis in the human placenta. The current study was designed to investigate the involvement of the protein kinase C pathway in EGF-mediated hCG-beta production by JAr choriocarcinoma cells. Downregulation of protein kinase C activity by chronic exposure to the phorbol ester, phorbol 12,13-dibutyrate (PDB), produced a greater increase in hCG-beta secretion than did activation of protein kinase C activity by short-term exposure to PDB. Pretreatment with the protein kinase C inhibitors calphostin and chelerythrine also resulted in enhanced basal and EGF-stimulated hCG-beta production. Individual concentrations (5 nM EGF and 500 nM PDB) that maximally stimulated hCG production, were additive in combination. The additive effect of PDB on EGF-induced hCG-beta secretion was mediated in part by increased JAr cell EGF-receptor concentrations detected by Western blot and Scatchard analyses. The results suggest that EGF and PDB stimulate hCG production in JAr cells by different but interactive mechanisms. It is speculated that downregulation of protein kinase C stimulates basal and EGF-mediated hCG-beta production by uninhibiting other signalling pathways that regulate hCG-beta secretion in trophoblasts.

Rates of dissociation of steroid and thyroid hormones from human serum albumin.

A rapid filtration assay employing dextran-coated charcoal as acceptor particles for free hormone was used to measure rates of dissociation of steroid and thyroid hormones from human serum albumin. Modification of a previously described assay allowed measurements at 1-s intervals. Nevertheless, this still permitted only minimum estimates of the dissociation rate constants. The hormones studied were thyroxine, 3,5,3'-triiodothyronine, cortisol, corticosterone, testosterone, dihydrotestosterone, estradiol, progesterone, and aldosterone. The apparent dissociation rate constant of the thyroxine-albumin complex at 37 degrees C was 1.3 +/- 0.2 s-1 (t 1/2, 0.5 s). The apparent dissociation rate constants of the other hormone-albumin complexes at 37 degrees C generally exceeded 2 s-1 (t 1/2 less than 0.35 s). Apparent dissociation rate constants at 4 degrees C were only slightly lower. These findings indicate that steroid and thyroid hormones dissociate from albumin rapidly compared with the 1-s capillary transit times that characterize many tissues.

Aromatase gene expression and its exon I usage in human breast tumors. Detection of aromatase messenger RNA by reverse transcription-polymerase chain reaction.

The expression of aromatase in human breast tumors has been studied by the reverse-transcription polymerase chain reaction (RT-PCR) method on 70 breast tissue specimens. An RT-PCR analysis using two oligonucleotide primers derived from the exon II of the human aromatase gene revealed that aromatase mRNA was detected in all but three tissue specimens. Furthermore, primer-directed RT-PCR was performed to determine the exon I usage in aromatase mRNA in these breast tumor specimens. The analysis has revealed that exons I.3 and PII are the two major exon Is present in aromatase mRNA isolated from breast tumors, suggesting that promoters I.3 and II are the major promoters driving aromatase expression in breast cancer and surrounding adipose stromal cells. The RT-PCR analysis also detected two products, I.3A (334 bp in length) and I.3B (222 bp in length), when it was carried out using a primer derived from exon I.3 and a reverse primer derived from exon II. The nucleotide sequences of these products have been determined and indicate that I.3A contains a region which was previously thought to be an intron. In addition, RT-PCR analyses of RNA isolated from eight pairs of breast tumor and neighboring normal tissue specimens were performed to evaluate the exon I usage and the distribution of I.3A- and I.3B-containing aromatase RNA messages in breast tumor and neighboring normal tissues. The results suggest that I.3B- and I.3A-containing messages are mainly present in breast tumor and neighboring normal tissues, respectively. Finally, the exon I/promoter usage for aromatase expression in eight cell lines (skin fibroblast, MCF-7, MDA-MB-231, T-47D, SK-BR-3, JAR, OVCAR-3, and human adipose stromal cells) was examined by primer-directed RT-PCR analyses. These studies provide a basis for further evaluation of the control mechanism of aromatase expression and estrogen biosynthesis in breast tumors.

Aromatase gene is amplified in MCF-7 human breast cancer cells.

The levels of the aromatase gene and its expression in MCF-7 human breast cancer cells and seven additional cultured cells were investigated. Using normal human foreskin fibroblasts as the control, the aromatase gene appeared to be amplified in MCF-7 cells as shown by Southern and DNA slot blot analyses utilizing human placental aromatase cDNA as the probe. However, the promoter I.1 and the first exon of the aromatase gene were not amplified in MCF-7 cells based on results obtained from DNA slot blot analysis using oligonucleotide probes having sequences derived from those regions of human aromatase gene. Aromatase was expressed at a very low level in this cell line as indicated by Northern blot analysis to measure the level of aromatase mRNA, immunoprecipitation analysis to measure the level of aromatase protein, and aromatase activity measurement. Furthermore, nucleotide sequence analysis of the aromatase cDNA obtained from MCF-7 cells by PCR techniques, revealed no sequence difference from that of the enzyme expressed in placenta. These results lead us to conclude that the expression of aromatase in MCF-7 cells is under the control of an unusual promoter and aromatase gene expression is repressed at the transcriptional level in these cells.

Maternal and fetal modulators of lipid metabolism correlate with the development of preeclampsia.

The pregnancy syndrome preeclampsia is associated with placental dysfunction, dyslipidemia, and endothelial cell activation, and is a major cause of maternal and fetal morbidity and mortality. In this report, a nested case-control study of matched preeclamptic and normal pregnant women was used to investigate the association of maternal and fetal modulators of lipid metabolism with pregnancy outcome. Maternal body mass index (BMI), triglyceride levels, and nonesterified fatty acid (NEFA) concentrations were all significantly increased in women who developed preeclampsia (P < .01). Human placental lactogen (hPL), which is secreted by the syncytiotrophoblast layer of the fetal placenta and reportedly has lipolytic activity, also was found to be elevated in women with preeclampsia (P < .01). By contrast, hemoglobin levels were not found to be statistically different between the two groups of women, indicating that the increased plasma lipids and hPL were not a result of hemoconcentration in preeclamptic patients. The results suggest a multihit hypothesis for the pathophysiology of preeclampsia in which maternal obesity and a placental lipolytic hormone (hPL) converge to adversely affect free fatty acid concentrations in the maternal circulation.

Eicosanoid secretion by human endothelial cells exposed to normal pregnancy and preeclampsia plasma in vitro.

Eicosanoids play an important role in the pathogenesis of preeclampsia. The major eicosanoid metabolite reported to be secreted by endothelial cells, the vasodilator prostacyclin, is generally reduced in preeclampsia. By contrast, it was shown previously that prostacyclin secretion by cultured human umbilical vein endothelial (HUVE) cells is increased significantly after exposure to blood from preeclamptic women. In the current study, eicosanoid profiles in conditioned media from HUVE cells incubated with pregnancy plasma were analyzed by high-performance liquid chromatography, thin layer chromatography and quantitative radio- and enzyme immunoassays. More prostaglandin F2alpha, prostacyclin and 8-isoprostane were secreted after exposure to plasma from preeclamptic women than plasma from matched, normal pregnant patients. Predominant secretion of the vasoconstrictor prostaglandin F2alpha by HUVE cell cultures and a stimulatory effect of preeclampsia plasma on eicosanoid biosynthesis underscore the importance of bioactive lipids in the vasospasm associated with clinical preeclampsia.