Objective
We sought to determine whether haptoglobin (Hp) phenotype is related to preeclampsia risk, or to plasma concentrations of soluble endoglin (sEng), soluble fms-like tyrosine kinase 1 (sFlt-1), and placental growth factor (PlGF).
Study Design
Hp phenotype was retrospectively determined in primiparous women with uncomplicated pregnancies (n = 309), gestational hypertension (n = 215), and preeclampsia (n = 249). Phenotype was assessed by peroxidase staining following native polyacrylamide gel electrophoresis of hemoglobin-supplemented serum.
Results
Compared with Hp 1-1, Hp 2-1 was associated with a significantly increased risk of preeclampsia (odds ratio, 2.11; 95% confidence interval, 1.07–4.18) and term preeclampsia (odds ratio, 2.45; 95% confidence interval, 1.07–5.83) in Caucasian women. Hp phenotype was not associated with preeclampsia risk in African Americans. Preeclamptic women had higher plasma sEng and sFlt-1, and lower PlGF, than control subjects. sEng, sFlt-1, and PlGF did not differ among women of different Hp phenotypes.
Conclusion
Hp 2-1 is associated with higher preeclampsia risk in primiparous Caucasian women.
Preeclampsia affects 2-7% of pregnancies, and remains a substantial cause of maternal and fetal morbidity and mortality. Oxidative stress; high concentrations of the antiangiogenic proteins soluble endoglin (sEng) and soluble fms-like tyrosine kinase 1 (sFlt-1); and low concentrations of the proangiogenic placental growth factor (PlGF), are proposed to contribute to the vascular dysfunction leading to maternal pathophysiology.
Haptoglobin (Hp) is an acute phase protein that acts as an antioxidant by binding free hemoglobin. Hp also stimulates angiogenesis in vitro and in vivo. Hp has 3 common polymorphisms (1-1, 2-1, and 2-2), which are composed of a common beta allele, and 2 different alpha alleles (1 and 2). Phenotype can be used to infer genotype, except in cases where Hp is not produced (ie, fetal/early life), or has been depleted (ie, severe infection). Structural differences between the polymorphisms affect protein function. Antioxidant activity is strongest for Hp 1-1. Proangiogenic activity is strongest for Hp 2-2. Hp phenotype predicts cardiovascular disease risk, and treatment response to antioxidant supplementation, in other populations.
Small European studies examining the relationship between Hp phenotype and preeclampsia risk have produced conflicting results. Larger studies, and North American studies that include African American women, are needed. When we initiated this investigation, the largest published study showed that Hp 1-1 was associated with increased preeclampsia risk, higher blood pressure, and greater proteinuria. We hypothesized that the reduced angiogenic capacity of Hp 1-1 might explain these effects. The mechanism through which Hp promotes angiogenesis is not known. One possibility is that Hp may affect concentrations of proangiogenic and antiangiogenic factors in a phenotype-dependant manner.
The study objectives were to determine whether Hp phenotype is associated with altered risk of preeclampsia or gestational hypertension in primiparous North American women, and to assess the relationship between Hp phenotype and plasma concentrations of sEng, sFlt-1, and PlGF in normotensive pregnant and preeclamptic women. We hypothesized that Hp 1-1 would be associated with higher preeclampsia risk, lower plasma concentrations of the proangiogenic factor PlGF, and higher concentrations of the antiangiogenic factors sEng and sFlt-1.
Materials and Methods
Study population
We used banked serum from women enrolled in the ongoing Pregnancy Exposures and Preeclampsia Prevention Study (1999 through 2010), which was approved by the University of Pittsburgh Institutional Review Board. All subjects provided written informed consent. This analysis was restricted to primiparous women with a singleton pregnancy. We included all African American women with an uncomplicated pregnancy (n = 106), gestational hypertension (n = 60), or preeclampsia (n = 60); all non-Hispanic Caucasian women with gestational hypertension (n = 155); and a subset of non-Hispanic Caucasian women with an uncomplicated pregnancy (n = 203) or preeclampsia (n = 189). Gestational hypertension was defined as persistent, de novo hypertension (systolic ≥140 mm Hg and/or diastolic ≥90 mm Hg) appearing >20 weeks’ gestation. Preeclampsia was defined as gestational hypertension and proteinuria. Proteinuria was defined as the excretion of ≥300 mg of protein/24 hours, a dipstick of 2+, a catheterized sample of 1+, or protein:creatinine ≥0.3. Hyperuricemia was defined as a plasma uric acid ≥1 SD above the mean value at the gestational age when the sample was obtained. Among preeclamptic women, 80.3% of Caucasians and 83.3% of African Americans had hyperuricemia. Women with gestational hypertension did not have hyperuricemia. Birthweight centiles were based on growth curves derived from Magee-Womens Hospital, which account for race, gestational age at delivery, and fetal sex.
Hp phenotyping
Hp phenotype was determined by native polyacrylamide gel electrophoresis using a modification of previously described methods. A total of 5 μL of serum was supplemented with 3 μL of 25 μmol/L human hemoglobin (Sigma-Aldrich, St Louis, MO). The peroxidase activity of hemoglobin in the Hp-hemoglobin complex can be detected directly, eliminating the need for primary and secondary antibodies. Samples were run on 6% tris-glycine gels (Invitrogen, Carlsbad, CA) for 2 hours at 120 V. Proteins were transferred to polyvinylidene flouride (Millipore, Billerica, MA).
Hemolyzed samples or samples with low Hp concentrations were phenotyped by sodium dodecyl sulfate polyacrylamide gel electrophoresis of 1.5 μL of serum, using a modification of previously described methods. Serum was combined with 2 μL of β-mercaptoethanol and heated for 7 minutes at 82°C. Samples were run on 12% tris-glycine gels (Invitrogen) at 120 V for 75 minutes. Proteins were transferred to a polyvinylidene flouride membrane, and incubated with blocking solution (tris-buffered saline containing 5% nonfat milk, 0.1% Tween 20), primary antibody (1:5000, polyclonal rabbit antihuman Hp; DakoCytomation, Carpinteria, CA), and secondary antibody (1:25,000, goat antirabbit IgG horseradish peroxidase; Millipore) for 1 hour each at room temperature. Antibodies were dissolved in blocking solution. Membranes were washed between incubations in tris-buffered saline containing 0.1% Tween 20.
Membranes were stained for peroxidase activity (SuperSignal West Pico Chemiluminescent Substrate; Fisher Scientific, Pittsburgh, PA) and imaged (FluorChem Q System; Cell Biosciences, Santa Clara, CA). Hp phenotypes were identified by characteristic banding patterns ( Figure 1 ).
Plasma concentrations of angiogenic factors
In a subset of control subjects (n = 247) and women with preeclampsia (n = 77), EDTA plasma was collected, aliquoted, and stored at –80°C. Samples from preeclamptic women were obtained <48 hours before delivery. Control samples were matched for gestational age at sample collection. sFlt-1 and PlGF concentrations were measured on the Architect i2000SR (Abbott Laboratories, Abbott Park, IL) using prototype reagents. sEng was quantified by enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, MN). The interassay variability was: sFlt-1, 3.1-4.5%; PlGF, 1.4-5.5%; and sEng, 9%.
Statistical analysis
Subject characteristics and indicators of disease severity were compared between women of different phenotypes. Continuous data were compared using the Student t test or Wilcoxon rank sum test. Categorical data, expressed as proportions, were compared using the χ 2 test or Fisher exact test. Logistic regression or the generalized logit model was used to evaluate the relationship between Hp phenotype and pregnancy outcome, stratified by race, after adjusting for age. A linear regression model was used to investigate the relationship among Hp phenotype, pregnancy outcome, and the log-transformed concentrations of PlGF, sFlt-1, and sEng, after adjusting for confounding variables. Analyses were performed using PASW 18 (IBM, New York, NY) or SAS 9.2 (SAS Institute Inc, Cary, NC) at a statistical significance level of .05.
Results
Subject characteristics
The rare 2-1M phenotype was found in 3 African American women (control, n = 1; gestational hypertension, n = 2) and 2 Caucasian women (preeclampsia, n = 2). These women were excluded from statistical analyses due to small sample size.
Prepregnancy body mass index (BMI), average systolic and diastolic blood pressure <20 weeks, the percentage of smokers in the third trimester, and infant sex did not differ between women of different phenotypes in the entire sample ( Table 1 ), or when the analysis was stratified by pregnancy outcome (data not shown). Women with the Hp 2-2 phenotype were significantly older than women with the 1-1 and 2-1 phenotypes ( P = .001). When the analysis was stratified by pregnancy outcome, this difference remained significant in the control group ( P = .047), but not in women with gestational hypertension ( P = .094) or preeclampsia ( P = .075). The effect of age was due to the significantly higher proportion of Caucasians in the Hp 2-2 group, as Caucasian women were older than African American women (25.5 ± 5.8 vs 21.3 ± 4.0 years, P < .001). When stratified by race, age did not differ between women of different Hp phenotypes (Caucasian, P = .32; African American, P = .86).
