| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Departments of Obstetrics, Gynecology, and Reproductive Sciences (J.L.V., S.J.F., R.N.T.), Stomatology (S.J.F.), and the Mass Spectrometry Facility (W.P.J., S.J.F.), University of California, San Francisco, California 94143; Geron Corp., Inc. (J.T.M.), Menlo Park, California 94025; and Arbogast Pharmaceuticals, Inc. (B.W.A.), Johnson City, Tennessee 37601
Address all correspondence and requests for reprints to: Robert N. Taylor, M.D., Ph.D., Reproductive Endocrinology Center, HSE 1679, University of California School of Medicine, San Francisco, California 94143-0556.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methanolic-HCl was purchased from Supelco (Bellefonte, PA). Free fatty acids, esterified fatty acid standards, human plasma VLDL (L 2264), fatty acid-free human albumin (A 3782), and 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) kits were obtained from Sigma Chemical Co. (St. Louis, MO). Affi-Blue Sepharose columns were obtained from Pharmacia Biotech (Piscataway, NJ) and used according to the manufacturer’s recommendations.
Patient selection
A nested case-control study was designed by selecting pregnant women from a prospectively collected cohort of pregnant women receiving obstetrical care at the University of California-San Francisco (UCSF). Twelve nulliparous women with severe preeclampsia [American College of Obstetricians and Gynecologists’ criteria (7)] were identified. These women were matched for maternal age (±3 yr), nulliparity, race, and body mass index (BMI) to 12 pregnant women with transient hypertension of pregnancy and to 12 normal pregnant controls. Because of the known association of NEFA concentrations with insulin resistance, women with abnormal glucose tolerance screening tests were excluded from the study. Written informed consent was provided by all participating patients under a protocol approved by the UCSF committee on human research.
The subjects were assigned to each of the three study groups by previously established strict criteria reported by Chesley (8) as recommended by the NIH consensus conference for investigations of pregnancy hypertension (9). Preeclamptic women were nulliparous, with pregnancy-onset hypertension, proteinuria, hyperuricemia, and reversal of hypertension and proteinuria within 12 weeks after delivery. Pregnancy-onset hypertension was defined as an increase of 30 mm Hg systolic or 15 mm Hg diastolic compared to values obtained before 20 weeks’ gestation or an absolute blood pressure of 140/90 mm Hg or higher. Proteinuria was designated as 100 mg/dL or more in a voided specimen or 30 mg/dL in a catheterized specimen. Hyperuricemia was defined as 1 SD or more above the normal mean concentration corrected for gestational age (10). Transient hypertensive patients also were nulliparous and met the same criteria for hypertension as those described for the preeclampsia group, with resolution of the elevated blood pressure by the 12th postpartum week. However, these women had neither proteinuria nor hyperuricemia (11). The duration of elevated blood pressure in the two hypertensive groups (6–10 days) did not differ.
Measurements of blood pressure and BMI
Blood pressures were calculated from the average of at least three blood pressure readings (Korotkov V, seated position) taken before 20 completed weeks gestation and again during the early intrapartum period, before iv fluid or pharmacological (including anesthetic) therapy was administered. BMI, defined as weight (kilograms)/height2 (meters), was selected as the best practical index of obesity and body composition (12). The heights and weights of each patient were obtained from the prenatal record.
Plasma collection and determinations
To avoid potential fetal effects of maternal ketosis, venipuncture was performed in nonfasting gravidas. Plasma samples were collected between 0900–1300 h in 5 mmol/L ethylenediamine tetraacetate approximately 1 week before delivery and frozen at -70 C. The 36 matched plasma specimens were thawed and analyzed for TxPA, NEFA, and triglyceride concentrations as described below.
TxPA determinations
A slab-gel modification of the electrofocusing assay of Arbogast et al. (6) was developed to quantify TxPA, the pI 5.6 isoform of plasma albumin. Isoelectric focusing (IEF) was carried out in 10% polyacrylamide gels using the Bio-Rad 111 mini-gel system (Richmond, CA). In a series of comparative experiments, this simplified IEF-PAGE assay was shown to correlate accurately with the sucrose gradient electrofocusing assay of Arbogast et al. (6). Albumin isoforms were stained selectively with 0.1% bromocresol green (6), and the gels were fixed and scanned. Bromocresol green-positive bands in the IEF-PAGE assays that comigrated with fatty acid free albumin (pI 5.6) were integrated by laser densitometry and referred to as TxPA. Quantification of TxPA bands in the IEF-PAGE assay was linear for both exogenous and endogenous NEFA-free albumin. Endogenous TxPA was examined over a range of 0.2–2.0 µL pregnancy plasma and was found to be linear up to 1.2 µL plasma. The TxPA bands were normalized to an internal standard included in each experiment.
NEFA and triglyceride determinations
Plasma NEFA were quantified using a microtiter plate modification of the NEFA C diagnostic kit (Wako Pure Chemicals, Richmond, VA). In our laboratory this assay had a sensitivity of 0.02 mmol/L, with inter- and intraassay coefficients of variation of 12% and 3%, respectively. Plasma triglycerides were quantified in matched patients using a microtiter plate modification of the Sigma triglycerides diagnostic kit. The colorimetric assay had a sensitivity of 1 mg/dL, with inter- and intraassay coefficients of variation of 3% and 1%, respectively.
Lipid extraction from plasma albumin
Albumin was purified from pooled plasma samples by affinity chromatography. Briefly, 1 mL plasma was loaded onto a 1-mL Affi-Blue Sepharose column, and after washing, the retained proteins were eluted by increasing the sodium concentration to 1 mol/L. The fractions containing the eluted proteins were pooled and dialyzed against 20 mmol/L phosphate buffer (pH 7.4). Protein concentrations were determined using the Bradford method (13), and the purity of the samples was assessed by SDS-PAGE (14), which demonstrated a single band of 68 kDa. Pooled plasma albumin fractions were delipidated according to the chloroform/methanol procedure of Bligh and Dyer (15). Lipids obtained from the extraction were solubilized in hexanes, evaporated to dryness, and stored under nitrogen at -20 C before analysis.
GC/MS
A qualitative assessment of the fatty acids extracted from
pregnancy plasma albumin was afforded by negative liquid secondary ion
mass spectrometry (LSIMS) using an MS50 mass spectrometer (Kratos,
Manchester, UK). The spectra generated from each peak allowed
definitive identification of the fatty acids bound to plasma albumin.
To determine the concentration of fatty acids extracted from plasma
albumin, these samples were methylated by the methanolic-HCl method,
according to the manufacturer’s protocol (Supelco, Bellefonte, PA),
and subjected to GC/MS analysis. Methyl fatty acid peaks were
normalized using an internal standard (methyl erucic acid, 27:2). The
GC/MS was performed on a Hewlett-Packard model 5890 gas chromatograph
(Palo Alto, CA) coupled to a VG-70SE mass spectrometer (Micromass,
Manchester, UK). The GC was equipped with a 25-m BPX5 column (SGE) with
a 1-µm film thickness of fused silica. The conditions of elution were
as follows: 60 C for 0.5 min with increasing temperature increments of
15 C/min until 100 C was reached, then the temperature was ramped by
increments of 20 C/min until a final temperature of 280 C was reached
and sustained for 20 min. Total ions were quantified. The GC/MS method
was calibrated using test mixtures of 1:1 and 1:2 molar ratios of oleic
and palmitic fatty acid standards. The ratio of the integrated surface
areas matched the ratio of masses of the fatty acids injected, yielding
response factors of 
1.
Endothelial cell cultures
HUVE cells were isolated and cultured using a modification of the method of Jaffe et al. (16) as described by us previously (17). Briefly, endothelial cells were flushed from neonatal umbilical cords after collagenase digestion (1 mg/mL; 15 min at 37 C) and cultured in medium 199 supplemented with 20% FBS and antibiotics. At confluence, the medium was changed to serum-free medium 199 containing 20 mmol/L HEPES (pH 7.4), 500 µg/mL BSA, 5 µg/mL transferrin, and 1 µg/mL insulin. Human VLDL (Sigma; 100 µg/mL protein), human fatty-acid free albumin, and increasing concentrations of oleic acid were added to triplicate wells of HUVE cells in 96-well plates (Becton Dickinson, Lincoln Park, NJ). Cell viability was determined using the colorimetric MTT assay (Sigma).
Statistical analyses
Data are presented as the mean ± SD of each study group. ANOVA and Scheffe’s post-hoc tests were used to make comparisons among the normal, preeclamptic, and transient hypertensive pregnant patients. Linear regression analyses were used to compare matched results from the IEF-PAGE assay with the sucrose gradient IEF assay and to assess the correlation among clinical biochemistry indicators. For all comparisons, two-tailed tests were accepted as significant when P < 0.01.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
demonstrate the successful matching of
severely preeclamptic women and their controls with respect to maternal
age, parity, and BMI. The mean duration of gestation was significantly
shorter in those patients with severe preeclampsia than in the other
two groups. The mean (±SD) durations of gestation for each
of the three groups were as follows: normal, 39 ± 2 weeks;
preeclampsia, 34 ± 4 weeks; and transient hypertension, 40
± 2 weeks. ANOVA (P < 0.01) and Scheffe’s
post-hoc tests indicated that the women with severe
preeclampsia had significantly shorter gestations than the other two
patient groups (P < 0.01). Likewise, the neonatal
birth weights and placental weights were lowest in the women with
severe preeclampsia (P < 0.01). As expected from the
diagnostic definitions, the mean ± SD systolic and
diastolic blood pressures were higher in preeclamptic and transient
hypertensive women than in the normal controls (P <
0.01), and proteinuria and serum uric acid levels were highest in the
severe preeclampsia group (P < 0.01). Plasma
hemoglobin levels showed subtle differences among the three groups, but
these did not reach the assigned level of statistical significance.
|
shows the results of a typical
IEF-PAGE analysis, with standards in the left lanes and
patient samples on the right. Lane 1 contains 50 µg fatty
acid-free human albumin and shows a major protein species migrating at
pI 5.6 (TxPA). Lane 2 contains 50 µg fatty acid-free human albumin
preincubated with a 10-fold molar excess of oleic acid, and a shift to
the pI 4.8 isoform is observed. Lane 3 contains a mixture of the two
previous standards and demonstrates that both isoforms are resolved by
IEF-PAGE. At the right, 1-µL plasma samples from a normal
pregnant woman and a matched severely preeclamptic patient demonstrate
the shift to a more acidic albumin phenotype in the latter case.
|
shows the linear relationship, up to 1.2 µL plasma, between
the detection of TxPA and the volume of plasma subjected to IEF-PAGE.
The reproducibility of the IEF-PAGE quantification also was evaluated.
The inter- and intraassay coefficients of variation were both 12% in
10 independent determinations of plasma TxPA at a mean concentration of
2.0 g/dL. Direct comparisons of the IEF-PAGE assay with the previously
standardized sucrose density gradient IEF assay (6) were performed in
22 independent specimens. A high degree of correlation between the 2
assays (r = 0.56; P < 0.01) was noted.
The TxPA bands in each of 36 plasma samples were quantified by
densitometry and normalized to an internal pI 5.6 albumin standard run
in parallel. The results, grouped according to clinical category, are
shown as a scattergram in Fig. 2. The
mean ± SD for each of the 3 groups are as follows:
normal, 2.30 ± 0.54 g/dL; severe preeclampsia, 1.57 ± 0.47
g/dL; and transient hypertension, 2.75 ± 0.40 g/dL. ANOVA
(P < 0.01) and Scheffe’s post-hoc tests
indicated that the women with severe preeclampsia had significantly
lower TxPA than the other 2 patient groups (P <
0.01).
|
.
Statistical analyses indicated that the plasma NEFA concentrations were
significantly different among the groups (by ANOVA, P
< 0.01) and that preeclamptic women had higher levels than the other
two patient groups (P < 0.01). Triglyceride
concentrations in normal, preeclamptic, and transient hypertensive
pregnant women were 166 ± 56, 246 ± 92, and 224 ± 99
mg/dL, respectively. By ANOVA, these values did not reach a significant
level of statistical difference (P = 0.07).
|
Albumin was isolated by Affi-Blue Sepharose affinity chromatography
from pooled plasma obtained from the normal and preeclamptic patient
groups. Equal amounts of albumin (30 nmol each) were extracted
according to the method of Bligh and Dyer (15). The extracted lipids
were solubilized in hexanes and analyzed by mass spectrometry. The
moieties detected by negative ion LSIMS arose as de-protonated
molecules, (M-H)-1. The ions were separated in the mass
analyzer based on their mass to charge (m/z) ratios and were detected
by an electron multiplier. The mass spectra showed prominent peaks at
m/z 255, 279, 281, and 283, which corresponded to palmitic, linoleic,
oleic, and stearic acids, respectively (see Fig. 4A).
|
, B and C, shows representative chromatograms
from normal and preeclampsia plasma pools, respectively. As in the
LSIMS analyses, the electron impact spectra of these peaks showed that
palmitic, linoleic, oleic, and stearic acids were the predominant fatty
acids bound to normal plasma albumin. In the preeclamptic samples (Fig. 4C), five major peaks were detectable. Based upon their mass
determinations, these were identified as the methyl esters of palmitic,
linoleic, oleic, and stearic acids. Two isomeric forms of linoleic acid
(eluting at 59.5 and 60.5 min) were observed. Based on the recovery of
the methyl erucic acid standard, we calculated the total recovery of
albumin-bound fatty acids after extraction and GC/MS to be 40%.
Integration of the peak areas revealed that 1 nmol plasma albumin
contained 0.8 and 2.5 nmol total fatty acids in normal and preeclampsia
plasma samples, respectively. These findings are not the result of
decreased albumin mass in the severe preeclampsia patients, as the
total plasma albumin concentrations did not differ between the 2
patient groups (3.3 ± 0.7 and 3.4 ± 0.9 g/dL in normal
pregnancy and severe preeclampsia, respectively; P =
0.89). Only 1 of 12 severe preeclamptics had elevated liver enzymes,
and this subject had neither an unusually low TxPA nor total albumin
level.
To investigate the possible biological effect of the increased
NEFA/albumin ratio observed in plasma from women with severe
preeclampsia, we used an in vitro model of human endothelial
cell toxicity. HUVE cells grown to confluence in 96-well plates were
rinsed and incubated for 24 h in fresh, serum-free medium. To
mimic the conditions present in preeclampsia, the cell cultures were
exposed to increasing molar ratios of oleic acid/albumin (from 0–10).
The MTT assay was used to assess cell viability. No direct effect of
this treatment on HUVE cell viability was observed (Fig. 5). However, when the cells were
incubated with human VLDL (100 µg/mL protein) in the absence of human
albumin, more than 95% of the cells did not survive. Addition of 2.5
g/dL fatty acid-free human albumin (pI 5.6) prevented the VLDL-induced
cytotoxicity. This concentration of albumin was chosen to approximate
the mean level of TxPA detected in normal pregnancy plasma (see Fig. 2). The protective effect of exogenous human albumin was lost with
increasing molar ratios of oleic acid/albumin. The cytotoxicity
threshold molar ratio of 1.5 (oleic acid/albumin) observed in
vitro (Fig. 5) approximates the ratio observed in vivo
in the plasma of women with severe preeclampsia (NEFA/albumin
ratio = 2.5).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The finding that plasma NEFA levels are inversely proportional to TxPA concentrations is consistent with the hypothesis that fatty acid binding to plasma albumin is responsible for its shift in pI (5, 6). We directly tested this hypothesis by quantifying the amount of fatty acid bound to plasma albumin in normal and preeclamptic women. The results of our GC/MS analyses indicate that there is quantitatively more NEFA bound to albumin in preeclampsia than in normal pregnancy. The conversion of pI 5.6 albumin to pI 4.8 albumin not only results from the neutralization of basic charges by bound fatty acids, but also reflects a conformational change in the albumin molecule (22). Albumin with an average of 1.3 mol NEFA bound focuses at pI 5.6, whereas albumin with 2.5–10 mol NEFA focuses at pI 4.8 (23). This conformational change accounts for the abrupt loss of cytoprotective activity of TxPA when the molar ratio of oleic acid/albumin exceeds 1.5. In severe preeclampsia, elevated NEFA levels appear to saturate the buffering capacity of TxPA.
The 3-fold increase in albumin-bound NEFA is consistent with the 3-fold increase in total detectable NEFA circulating in these women. The NEFA/albumin ratios of 2.5 and 0.8 that we have calculated directly in severely preeclamptic and normal pregnant women are very similar to the 1.6 and 0.9 ratios estimated by Endresen and colleagues (4) in preeclamptic and normal pregnant women, respectively. Although preeclampsia is a relatively proteinuric state, the high NEFA/albumin ratio in severe preeclamptics is not simply a consequence of decreased total plasma albumin in these women. Indeed, their total circulating albumin concentrations were not different from those in normal pregnant women. This observation is similar to our findings in mildly preeclamptic women and their matched controls (6) and probably reflects the sampling of these patients before florid signs of the syndrome were manifested.
It is not surprising that women with severe preeclampsia were delivered after a shorter mean gestation than the two control groups. Even at the current time, expedient delivery is the most effective clinical management of severe preeclampsia (21). The shorter pregnancy duration in the preeclampsia group is not a confounder for the TxPA and lipid measurements. In fact, TxPA levels tend to fall during gestation, whereas NEFA levels increase during the course of gestation (6). Thus, the effects of pregnancy duration would mitigate against, rather than accentuate, differences in these parameters among the three groups. The hemoglobin results indicate that differences among the patient groups were not due to differences in hemoconcentration of the preeclamptic women.
Our in vitro studies suggest that the NEFA/albumin ratios
observed in severely preeclamptic women may approach levels at which
human endothelial cell viability is compromised. Previous studies by
ourselves (24, 25) and others (18) failed to detect evidence of
endothelial cell death after exposure to low (
30%) concentrations of
serum from preeclamptic women. However, when cultured endothelial cells
were further challenged by the addition of VLDL particles, the presence
of TxPA correlated with the prevention of cytotoxicity in HUVE cells.
Similar effects were observed previously in porcine aortic endothelial
cells (26). The biological characteristics of toxic principles
associated with VLDL remain unknown (27). Endotoxin has been suggested
as a possible candidate, but this can be difficult to detect
biochemically in lipoproteins (28).
In summary, our findings support the concept that endothelial cell dysfunction in preeclampsia may be a manifestation of lipid-induced injury to those cells. Patients with severe preeclampsia have 2- to 3-fold higher concentrations of NEFA than those with transient hypertension or normal pregnancies. The acidic shift in the pI of plasma albumin in preeclamptic patients appears to be the result of increased NEFA binding to albumin molecules. We postulate that elevated concentrations of NEFA either directly, by alteration of albumin function, or via intracellular incorporation into triglyceride (4, 29) perturb maternal endothelial cell function, leading to systemic manifestations of preeclampsia (30). Increased endothelial cell uptake of NEFA, particularly linoleic and arachidonic acids, may lead to abnormalities of PG metabolism associated with clinical preeclampsia (31) and in vitro models of this disorder (32, 33). Future therapeutic strategies to prevent excessive cellular NEFA uptake may reduce the development or progression of preeclampsia.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received May 29, 1997.
Revised July 22, 1997.
Accepted August 1, 1997.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  L. L. Waite, R. E. Louie, and R. N. Taylor Circulating Activators of Peroxisome Proliferator-Activated Receptors Are Reduced in Preeclamptic Pregnancy J. Clin. Endocrinol. Metab., February 1, 2005; 90(2): 620 - 626. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. E. Gandley, V. A. Tyurin, W. Huang, A. Arroyo, A. Daftary, G. Harger, J. Jiang, B. Pitt, R. N. Taylor, C. A. Hubel, et al. S-Nitrosoalbumin-Mediated Relaxation Is Enhanced by Ascorbate and Copper: Effects in Pregnancy and Preeclampsia Plasma Hypertension, January 1, 2005; 45(1): 21 - 27. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Kauma, P. Takacs, C. Scordalakes, S. Walsh, K. Green, and T. Peng Increased Endothelial Monocyte Chemoattractant Protein-1 and Interleukin-8 in Preeclampsia Obstet. Gynecol., October 1, 2002; 100(4): 706 - 714. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. C. Person, L. L. Waite, R. N. Taylor, and T. S. Scanlan Albumin Regulates Induction of Peroxisome Proliferator-Activated Receptor-{{gamma}} (PPAR{{gamma}}) by 15-Deoxy-{{Delta}}12-14-Prostaglandin J2 in Vitro and May Be an Important Regulator of PPAR{{gamma}} Function in Vivo Endocrinology, February 1, 2001; 142(2): 551 - 556. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. THADHANI, M. J. STAMPFER, D. J. HUNTER, J. E. MANSON, C. G. SOLOMON, and G. C. CURHAN High Body Mass Index and Hypercholesterolemia: Risk of Hypertensive Disorders of Pregnancy Obstet. Gynecol., October 1, 1999; 94(4): 543 - 550. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
