Objective
The long-term maternal cardiovascular and metabolic implications associated with preeclampsia (PE) include risk of hypertension, heart disease, and metabolic syndrome. The objective of this study was to investigate if a recent history of PE was associated with detectable alterations in the circulating maternal proteome.
Study Design
Six-month postpartum plasma from women with a history of PE (n = 12) and women with uncomplicated obstetrical history (n = 12) were used for analysis. Depleted maternal plasma was analyzed by label-free liquid chromatography-mass spectrometry assay. Identified peptides were searched against the International Protein Index human database version 3.87. Exponentially modified protein abundance indices were used for comparison. Results were analyzed using pathway analysis software.
Results
A total of 126 eligible peptides were identified for analysis; 3 peptides were differentially expressed in the PE proteome, and an additional 5 peptides were unique to control subjects and 7 to PE subjects. PE peptide profiles were more strongly associated with markers of coagulation and complement activation compared to controls and mapped more significantly to cardiovascular disease (CVD) functions. Stratification of subjects by low (<39%) and high (≥39%) lifetime risk of CVD rather than by diagnosis produced similar findings. Comparison of controls (n = 6) to PE subjects (n = 6) without traditional cardiovascular risk factors found that while similar for body mass indices, blood pressure, and fasting lipid profiles at 6 months postpartum, PE peptide profiles continued to display stronger associations for coagulation and CVD functions. Global network analysis found that unique peptides to low-risk PE subjects were associated with cardiac infarction, CVD, and organismal injury and abnormalities.
Conclusion
Markers of CVD risk and progression are evident in the maternal circulating proteome 6 months postpartum after PE. Augmentations in circulating peptide profiles occur in patients with previous PE who otherwise do not have clinically measurable cardiovascular risk factors. Our data highlight the need for the implementation of postpartum prevention programs in the PE population and identifies molecules that may be targeted for screening or therapeutic benefit.
Introduction
Preeclampsia (PE) is a hypertensive disorder of pregnancy that poses significant risk of adverse maternal and fetal outcomes. The pathogenesis underlying PE remains poorly understood although manifestations of impaired endothelial function, oxidative stress, and hypercoagulability bear striking similarity to those observed in states of cardiovascular risk and cardiovascular disease (CVD). Indeed, the long-term maternal cardiovascular and metabolic implications associated with PE are well established and include risk of heart disease, stroke, and the metabolic syndrome.
We postulate that the early postpartum period offers a unique window of opportunity for the identification and education of high-risk women. A lack of programs providing cardiovascular risk screening based on obstetrical history poses a significant hurdle to risk reduction in this population, however. Development of risk factors for CVD soon after PE, including persistent hypertension, dyslipidemia, insulin resistance, as well as a propensity for weight retention allude to early changes in biophysical profiles that could be targeted before the development of CVD itself. In addition to the development of traditional risk factors, evidence exists for impaired flow-mediated dilation, increased artery intima media thickness, and cardiac remodeling after pregnancies complicated by PE.
The use of nontraditional markers of cardiovascular risk to assess states of cardiovascular health in postpartum women remains important in establishing the short- and long-term implications of PE. Recently, induction of PE-like symptoms via overexpression of soluble fms-like tyrosine kinase (sFlt)-1 in CD1 mice was shown to induce long-term alterations in maternal peptide profiles after pregnancy. The translation of knowledge gained from animal models of PE such as these to human studies is critical for the development of preventative strategies aimed at minimizing the burden of PE and CVD on health care systems.
Only a handful of clinical studies have been undertaken in the past 10 years to examine the PE proteome, and the majority have focused on the examination of placental tissue. Of those studies examining maternal biofluids none have attempted analysis of the postpartum circulating proteome. Given the evidence for a predisposition to CVD after PE, we sought to determine if women who had recently developed PE would have detectable alterations in their circulating proteomes. In addition, data on cardiovascular risk profiles were used to examine whether the presence or absence of classic cardiovascular risk factors would impact the difference between control and PE peptide profiles.
Materials and Methods
Sample collection
This study was approved by the Queen’s University Research Ethics Board. Written informed consent was obtained from all subjects. Plasma samples were collected from women attending the Maternal Health Clinic at Kingston General Hospital at 6 months postpartum. The Maternal Health Clinic is designed to provide postpartum cardiovascular risk screening and counseling to all mothers delivering at Kingston General Hospital with select obstetrical complications, including PE. Biophysical measurements and fasting blood work collected by the clinic are used to generate lifetime cardiovascular risk scores for each patient to help guide maternal cardiovascular risk counseling. Plasma samples included for analysis were from women who either had experienced pregnancies complicated by PE (n = 12) or normotensive uncomplicated pregnancies (n = 12). PE was defined as the development of de novo hypertension (≥140/90 mmHg) and proteinuria (>300 mg/24 h or +1 on repeat dipstick). A clinical diagnosis was confirmed by chart review. Individuals with a self-reported history of hypertension, diabetes (including the development of gestational diabetes), kidney disease, CVD, or current smoking were excluded following confirmation by chart review.
Lifetime risk for CVD was calculated for each subject at the Maternal Health Clinic. Calculations for lifetime cardiovascular risk are based on biophysical factors: sex, smoking, total cholesterol fasting glucose, systolic blood pressure, diastolic blood pressure, and antihypertensive usage. Risk estimates are categorical: low risk (<39%) and high risk (≥39%).
Sample preparation for mass spectrometry
Plasma was analyzed for each subject individually. Whole blood samples were collected into ethylenediaminetetraacetic acid and centrifuged at 1000 g for 10 minutes within 2 hours of collection. Plasma was isolated and stored in aliquots at –80°C for a maximum of 3 years prior to analysis.
A total of 20 μL of each plasma sample was diluted with 0.7 μL (×10 dilution) of protease inhibitor (Sigma-Genosys, Spring, TX). Whole plasma was depleted of 14 highly abundant proteins using Agilent Human 14 Multiple Affinity Removal Columns (Agilent Technologies, Santa Clara, CA) according to the manufacturer’s instructions. Flow-through fractions were concentrated and buffer exchanged to 100 mmol/L ammonium bicarbonate by centrifugal filtration through a 5-kDa molecular weight cut-off Agilent spin concentrator (Agilent Technologies). High molecular weight fractions were collected and small aliquots used to perform a Bradford total protein assay (Bio-Rad, Hercules, CA).
Depleted plasma samples were denatured with 6 mol/L urea in 150 mmol/L Tris(hydroxymethyl)aminomethane hydrochloride (Tris HCl), pH 8.0, and reduced with 20 mmol/L dithiothreitol at 37°C for 40 minutes. Samples were then alkylated with 40 mmol/L iodoacetamide in the dark for 30 minutes and diluted 10-fold with 50 mmol/L Tris-HCl pH 8.0 prior to overnight digestion at 37°C with trypsin (Promega, Madison, WI). Digestion was terminated with equal volume 1% formic acid and samples desalted with Waters Oasis C18 cartridges (Waters, Milford, MA).
Tandem liquid chromatography mass spectrometry analysis
An aliquot of the tryptic digest (in 2% acetonitrile/0.1% formic acid in water) was analyzed by tandem liquid chromatography mass spectrometry on an LTQ-Orbitrap-XL mass spectrometer (MS) (Thermo-Fisher Scientific, Bremen, Germany) interfaced with an Eksigent Nano-LC-Ultra-2D plus CHiPLC Nanoflex system (AB SCIEX, Framingham, MA). A total of 0.5 μg of each sample was loaded onto a ChromXP C 18 -CL trap column (200-μm inner diameter, 0.5-mm length, 3 μm; AB SCIEX, Framingham, MA) at flow rate of 3 μλ/min. Reverse-phase C 18 chromatographic separation of peptides was carried on a ChromXP C 18 -CL column (75-μm inner diameter, 15-cm length, 3 μm; AB SCIEX, Framingham, MA) at 300 nL/min; column temperature was controlled at 35°C. Gradient conditions were: 3-8% B, 5 minutes; 8-33% B, 120 minutes; 33-90% B, 10 minutes; 90% B, 10 minutes; 90-3% B, 5 minutes (solvent A, 0.1% formic acid in water; solvent B, 0.1% formic acid in acetonitrile). Total run time was 150 minutes. The LTQ Orbitrap was operated in parallel mode with measurement of the full mass scan at 100,000 resolution in the Orbitrap concurrent with the acquisition of 5 most intense data-dependent MS/MS scans in the ion trap. For each cycle, MS1 was acquired at target value 1E6, and MS2 scanned at 3E4. The spray voltage was 1.35 kV; charge state screening and rejection of singly charged ions were enabled. Ion selection thresholds were 8000 for MS2: 35% normalized collision energy was used, activation Q was 0.25, and dynamic exclusion was employed for 90 seconds. Each sample was analyzed in duplicate.
Data processing and analysis
Raw data files were processed to generate a Mascot Generic Format with Mascot Distiller and searched against the International Protein Index human database version 3.87 (containing 91,464 protein sequences) using the licensed Mascot search engine version 2.3.02 (Matrix Science Inc, Boston, MA) run on an in-house server. The spectra were also searched against decoy database. For trypsin proteolysis, up to 2 missed cleavages were allowed. MS tolerance was set at 10 ppm; MS/MS tolerance at 0.6 Da. Carbamidomethylation on cysteine residues was used as fixed modification; serine, threonine, tyrosine phosphorylation, methionine oxidation, and protein N-terminal acetylation were set as variable modifications. The false discovery rate for both proteins and peptides were set at 0.01.
Ingenuity pathways analysis
Pathway analysis (Ingenuity Pathway Analysis [IPA], version 21901358; Qiagen, Redwood City, CA) was used to determine the long-term effect of PE according to methods described in detail by our coinvestigators elsewhere. Analyses included canonical pathways analysis, biological and toxicity functions analyses, and examination of network associations. Peptides present in <8/12 samples per subject group (control, PE, high risk, or low risk) were excluded from analysis. Remaining peptides were averaged by subject group and exponentially modified protein abundance indices were used for comparison. Molecules with intensity values >0.1 were included in the final analysis. Results from Ingenuity Knowledge Database (Qiagen) were filtered for species (human), confidence (experimentally observed), and relationships considered (both direct and indirect). Analyses were run based first on diagnosis (controls, n = 12, vs PE, n = 12) and second to control for CVD risk factors (low-risk controls, n = 6, vs low-risk PE, n = 6). In addition, PE/control protein ratios were calculated to determine the potential effect of exposure to PE on circulating peptides.
Statistical analysis
Demographic variables are presented as mean ± SD. GraphPad Prism 5 Software (GraphPad Prism Software, La Jolla, CA) was used for statistical comparisons. Unpaired t test, 2-tailed, was used to compare continuously distributed variables and a Fisher exact test was used for categorical measures. Statistical significance was accepted if the null hypothesis could be rejected at P < .05. IPA uses a right-tailed Fisher exact test to calculate P values by considering the number of focus peptides participating in a given biological function or process and comparing it to the total number of peptides known to be associated with that process.
Materials and Methods
Sample collection
This study was approved by the Queen’s University Research Ethics Board. Written informed consent was obtained from all subjects. Plasma samples were collected from women attending the Maternal Health Clinic at Kingston General Hospital at 6 months postpartum. The Maternal Health Clinic is designed to provide postpartum cardiovascular risk screening and counseling to all mothers delivering at Kingston General Hospital with select obstetrical complications, including PE. Biophysical measurements and fasting blood work collected by the clinic are used to generate lifetime cardiovascular risk scores for each patient to help guide maternal cardiovascular risk counseling. Plasma samples included for analysis were from women who either had experienced pregnancies complicated by PE (n = 12) or normotensive uncomplicated pregnancies (n = 12). PE was defined as the development of de novo hypertension (≥140/90 mmHg) and proteinuria (>300 mg/24 h or +1 on repeat dipstick). A clinical diagnosis was confirmed by chart review. Individuals with a self-reported history of hypertension, diabetes (including the development of gestational diabetes), kidney disease, CVD, or current smoking were excluded following confirmation by chart review.
Lifetime risk for CVD was calculated for each subject at the Maternal Health Clinic. Calculations for lifetime cardiovascular risk are based on biophysical factors: sex, smoking, total cholesterol fasting glucose, systolic blood pressure, diastolic blood pressure, and antihypertensive usage. Risk estimates are categorical: low risk (<39%) and high risk (≥39%).
Sample preparation for mass spectrometry
Plasma was analyzed for each subject individually. Whole blood samples were collected into ethylenediaminetetraacetic acid and centrifuged at 1000 g for 10 minutes within 2 hours of collection. Plasma was isolated and stored in aliquots at –80°C for a maximum of 3 years prior to analysis.
A total of 20 μL of each plasma sample was diluted with 0.7 μL (×10 dilution) of protease inhibitor (Sigma-Genosys, Spring, TX). Whole plasma was depleted of 14 highly abundant proteins using Agilent Human 14 Multiple Affinity Removal Columns (Agilent Technologies, Santa Clara, CA) according to the manufacturer’s instructions. Flow-through fractions were concentrated and buffer exchanged to 100 mmol/L ammonium bicarbonate by centrifugal filtration through a 5-kDa molecular weight cut-off Agilent spin concentrator (Agilent Technologies). High molecular weight fractions were collected and small aliquots used to perform a Bradford total protein assay (Bio-Rad, Hercules, CA).
Depleted plasma samples were denatured with 6 mol/L urea in 150 mmol/L Tris(hydroxymethyl)aminomethane hydrochloride (Tris HCl), pH 8.0, and reduced with 20 mmol/L dithiothreitol at 37°C for 40 minutes. Samples were then alkylated with 40 mmol/L iodoacetamide in the dark for 30 minutes and diluted 10-fold with 50 mmol/L Tris-HCl pH 8.0 prior to overnight digestion at 37°C with trypsin (Promega, Madison, WI). Digestion was terminated with equal volume 1% formic acid and samples desalted with Waters Oasis C18 cartridges (Waters, Milford, MA).
Tandem liquid chromatography mass spectrometry analysis
An aliquot of the tryptic digest (in 2% acetonitrile/0.1% formic acid in water) was analyzed by tandem liquid chromatography mass spectrometry on an LTQ-Orbitrap-XL mass spectrometer (MS) (Thermo-Fisher Scientific, Bremen, Germany) interfaced with an Eksigent Nano-LC-Ultra-2D plus CHiPLC Nanoflex system (AB SCIEX, Framingham, MA). A total of 0.5 μg of each sample was loaded onto a ChromXP C 18 -CL trap column (200-μm inner diameter, 0.5-mm length, 3 μm; AB SCIEX, Framingham, MA) at flow rate of 3 μλ/min. Reverse-phase C 18 chromatographic separation of peptides was carried on a ChromXP C 18 -CL column (75-μm inner diameter, 15-cm length, 3 μm; AB SCIEX, Framingham, MA) at 300 nL/min; column temperature was controlled at 35°C. Gradient conditions were: 3-8% B, 5 minutes; 8-33% B, 120 minutes; 33-90% B, 10 minutes; 90% B, 10 minutes; 90-3% B, 5 minutes (solvent A, 0.1% formic acid in water; solvent B, 0.1% formic acid in acetonitrile). Total run time was 150 minutes. The LTQ Orbitrap was operated in parallel mode with measurement of the full mass scan at 100,000 resolution in the Orbitrap concurrent with the acquisition of 5 most intense data-dependent MS/MS scans in the ion trap. For each cycle, MS1 was acquired at target value 1E6, and MS2 scanned at 3E4. The spray voltage was 1.35 kV; charge state screening and rejection of singly charged ions were enabled. Ion selection thresholds were 8000 for MS2: 35% normalized collision energy was used, activation Q was 0.25, and dynamic exclusion was employed for 90 seconds. Each sample was analyzed in duplicate.
Data processing and analysis
Raw data files were processed to generate a Mascot Generic Format with Mascot Distiller and searched against the International Protein Index human database version 3.87 (containing 91,464 protein sequences) using the licensed Mascot search engine version 2.3.02 (Matrix Science Inc, Boston, MA) run on an in-house server. The spectra were also searched against decoy database. For trypsin proteolysis, up to 2 missed cleavages were allowed. MS tolerance was set at 10 ppm; MS/MS tolerance at 0.6 Da. Carbamidomethylation on cysteine residues was used as fixed modification; serine, threonine, tyrosine phosphorylation, methionine oxidation, and protein N-terminal acetylation were set as variable modifications. The false discovery rate for both proteins and peptides were set at 0.01.
Ingenuity pathways analysis
Pathway analysis (Ingenuity Pathway Analysis [IPA], version 21901358; Qiagen, Redwood City, CA) was used to determine the long-term effect of PE according to methods described in detail by our coinvestigators elsewhere. Analyses included canonical pathways analysis, biological and toxicity functions analyses, and examination of network associations. Peptides present in <8/12 samples per subject group (control, PE, high risk, or low risk) were excluded from analysis. Remaining peptides were averaged by subject group and exponentially modified protein abundance indices were used for comparison. Molecules with intensity values >0.1 were included in the final analysis. Results from Ingenuity Knowledge Database (Qiagen) were filtered for species (human), confidence (experimentally observed), and relationships considered (both direct and indirect). Analyses were run based first on diagnosis (controls, n = 12, vs PE, n = 12) and second to control for CVD risk factors (low-risk controls, n = 6, vs low-risk PE, n = 6). In addition, PE/control protein ratios were calculated to determine the potential effect of exposure to PE on circulating peptides.
Statistical analysis
Demographic variables are presented as mean ± SD. GraphPad Prism 5 Software (GraphPad Prism Software, La Jolla, CA) was used for statistical comparisons. Unpaired t test, 2-tailed, was used to compare continuously distributed variables and a Fisher exact test was used for categorical measures. Statistical significance was accepted if the null hypothesis could be rejected at P < .05. IPA uses a right-tailed Fisher exact test to calculate P values by considering the number of focus peptides participating in a given biological function or process and comparing it to the total number of peptides known to be associated with that process.
Results
Comparison group profiles
A complete summary of prepregnancy, pregnancy, and postpartum characteristics of the study participants is presented in Table 1 . Control and PE groups were similar for age, and prepregnancy and postpartum body mass indices, although PE subjects exhibited elevated postpartum blood pressures and fasting triglycerides. Low-risk control and PE groups (<39% lifetime risk of CVD) were similar for age, body mass index, and postpartum blood pressures.
| All control, n = 12 | Low-risk control, n = 6 | All PE, n = 12 | Low-risk PE, n = 6 | |
|---|---|---|---|---|
| Maternal age, y, mean (SD) | 30.9 (3.9) | 31.8 (3.97) | 31.0 (5.46) | 33.0 (5.22) |
| Parity, n (%) a | ||||
| Primiparous | 5 (41.7) | 2 (33.3) | 7 (58.3) | 4 (66.7) |
| Multiparous | 7 (58.3) | 4 (66.7) | 5 (41.7) | 2 (33.3) |
| Gestational age at delivery, wk, mean (SD) | 40.1 (1.14) | 40.1 (1.11) | 35.4 (3.97) b | 36.6 (3.63) |
| Prepregnancy BMI, kg/m 2 , mean (SD) | 24.6 (4.28) | 25.4 (5.44) | 27.8 (7.79) | 25.3 (4.08) |
| Postpartum BMI, kg/m 2 , mean (SD) | 25.7 (5.42) | 27.5 (6.66) | 28.6 (7.57) | 26.6 (6.62) |
| Prepregnancy to postpartum follow-up change in BMI, kg/m 2 , mean (SD) | 1.07 (2.17) | 2.11 (2.08) | 0.86 (2.85) | 1.35 (2.62) |
| Blood pressure, mm Hg, mean (SD) | ||||
| Systolic | 103.6 (10.1) | 103.7 (8.89) | 121.5 (18.3) b | 109.2 (7.88) |
| Diastolic | 68.8 (8.69) | 68.2 (7.05) | 81.2 (13.3) b | 74.0 (5.76) |
| Glucose, mmol/L, mean (SD) | 4.48 (0.39) | 4.72 (0.20) | 4.65 (0.43) | 4.75 (0.45) |
| Total cholesterol, mmol/L, mean (SD) | 4.96 (1.09) | 4.04 (0.51) | 4.59 (1.03) | 4.06 (0.58) |
| Triglycerides, mmol/L, mean (SD) | 0.63 (0.24) | 0.61 (0.20) | 1.17 (0.65) b | 0.93 (0.28) c |
| HDL cholesterol, mmol/L, mean (SD) | 1.63 (0.49) | 1.42 (0.28) | 1.29 (0.38) | 1.37 (0.40) |
| LDL cholesterol, mmol/L, mean (SD) | 3.05 (0.93) | 2.34 (0.33) | 2.77 (0.84) | 2.26 (0.46) |
| Microalbumin/creatinine, mg/mmol, median (lower, upper quartile) | 0.65 (0.42, 1.02) | 0.80 (0.40, 1.1) | 1.20 (0.75, 2.2) | 0.95 (0.65, 1.5) |
| hsCRP, median (lower, upper quartile) | 1.00 (0.98, 4.12) | 1.00 (0.97, 4.6) | 2.00 (0.90, 6.00) | 0.95 (0.80, 3.2) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
