PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Am J Obstet Gynecol. Author manuscript; available in PMC 2017 May 1.
Published in final edited form as:
PMCID: PMC4851565
NIHMSID: NIHMS759388

Evaluation of proteomic biomarkers associated with circulating microparticles as an effective means to stratify the risk of spontaneous preterm birth

David E. CANTONWINE, PhD, MPH,(a) Zhen ZHANG, PhD,(b) Kevin ROSENBLATT, MD, PhD,(c) Kevin S. GOUDY, PhD,(d) Robert C. DOSS, PhD,(d) Alan M. EZRIN, PhD,(d) Ms. Gail PAGE,(d) Mr. Brian BROHMAN,(d) and Thomas F. MCELRATH, MD, PhD(a)

Abstract

Background

The analysis of circulating microparticles in pregnancy is of revolutionary potential as it represents an in vivo ‘biopsy’ of active gestational tissues.

Objectives

We hypothesize that circulating microparticle signaling will differ in pregnancies that experience spontaneous preterm birth from those delivering at term and that these differences will be evident many weeks in advance of clinical presentation.

Study Design

Utilizing plasma specimens obtained between 10–12 weeks gestation as part of a prospectively collected birth cohort where pregnancy outcomes are independently validated by two board certified Maternal Fetal Medicine physicians, 25 singleton cases of spontaneous preterm birth ≤34 weeks were matched by maternal age, race, and gestational age of sampling (+/− 2 weeks) to 50 uncomplicated term deliveries. Circulating microparticles s from these first trimester specimens were isolated and analyzed by multiple reaction monitoring mass spectrometry for potential protein biomarkers following from previous studies. Markers with robust univariate performance in correlating spontaneous preterm birth were further evaluated for their biological relevance via a combined functional profiling/pathway analysis, and for multivariate performance.

Results

Among the 132 proteins evaluated, 62 demonstrated robust power of detecting spontaneous preterm birth in bootstrap receiver-operating characteristic curve analysis at a false discovery rate of < 20% estimated via label permutation. Differential dependency network analysis identified spontaneous preterm birth -associated co-expression patterns linked to biological processes of inflammation, wound healing, and the coagulation cascade. Linear modeling of spontaneous preterm birth using a multiplex of the candidate biomarkers with a fixed sensitivity of 80% exhibited a specificity of 83% with median area under the curve of 0.89. These result indicate strong potential of multivariate model development for informative risk stratification.

Conclusion

This project has identified functional proteomic factors with associated biological processes that are already unique in their expression profiles at 10–12 weeks among women who go on to deliver spontaneously ≤34 weeks. These changes, with further validation, will allow the stratification of patients at risk of spontaneous preterm birth before clinical presentation.

Keywords: Differential dependency network analysis, exosomes, pregnancy, nested case/control study

Introduction

Preterm birth is a leading cause of neonatal morbidity and death in children less than 5 years of age, with deliveries at the earlier gestational ages exhibiting a dramatically increased risk.1,2 Compared with infants born after 38 weeks, the composite rate of neonatal morbidity doubles for each earlier gestational week of delivery.3 In 2005, the cumulative annual expense of preterm birth within the Unites States was estimated to be in excess of $26.2 billion.4 Approximately two thirds of preterm births are spontaneous, meaning they are not associated with medical intervention, in nature.5,6 Yet, despite the compelling nature of this condition, there has been little recent advancement in our understanding of the etiology of spontaneous preterm birth (SPTB). While there is an increasing consensus that SPTB represents a syndrome rather than a single pathologic entity, it has been both ethically and physically difficult to study the pathophysiology of the utero-placental interface.7 The evolving field of circulating microparticle (CMP) biology may offer a solution to these difficulties as these particles present a sampling of the utero-placental environment. Additionally, studying the contents of these particles holds the promise of identifying novel blood-based, and possibly clinically useful, biomarkers.

Microparticles are membrane-bound nanovesicles that range in size from 50–300 nm and shed by a wide variety of cell types. Microparticle nomenclature varies, but typically microparticles between 50–100 nm are called exosomes, those >100 nm are termed microvesicles and other terms, such as microaggregates, are often used in literature. Unless otherwise stated, we will use the term microparticle as a general reference to all of these species. Increasingly, microparticles are recognized as important means of intercellular communication in physiologic, pathophysiologic and apoptotic circumstances. While the contents of different types of microparticles vary with cell type and their expression, they can include nuclear, cytosolic and membrane proteins as well as lipids and messenger and micro RNAs.811 Their contents hold information regarding the state of the cell type origin at the time of microparticle expression; thus, they represent an unique ‘window’ in ‘real-time’ into the activities of cells, tissues and organs that might otherwise be remote to sampling.

A high proportion of adverse pregnancy outcomes have their pathophysiologic origins at the utero-placental interface in early pregnancy.7,12,13 The ability to understand protein signaling and state of associated tissue and cell populations may be predictive of impending complications. Our present analysis will demonstrate the ability to capture informative, microparticle-related, protein signaling at the end of the first trimester (mean, 11 weeks). We will further demonstrate that this signaling discriminates between pregnancies delivering at gestational ages marked by considerable neonatal morbidity (≤34 weeks) compared with those delivering at term.

Materials and Methods

Clinical Specimen Collection

Maternal K2-EDTA plasma samples (10–12 weeks gestation) were obtained and stored at −80°C at Brigham and Women’s Hospital (BWH), Boston between 2009–2014 as part of the prospectively collected LIFECODES birth cohort.14 Eligibility criteria included patients who were ≥18yrs of age, initiated their prenatal care at ≤15 weeks of gestation and planned on delivering at the BWH. Exclusion criteria included preexisting medical disorders (preexisting diabetes, gestational diabetes, autoimmune disorders, current cancer diagnosis, HIV, Hepatitis), fetal anomalies, and was restricted to singleton gestations. Gestational age of pregnancy was confirmed by ultrasound scanning ≤12 weeks gestation. If consistent with last menstrual period (LMP) dating, the LMP was used to determine the due date. If not consistent, then the due date was set by the earliest available ultrasound. Full-term birth was defined as ≥37 weeks of gestation, and preterm birth for the purposes of this investigation was defined as SPTB ≤34 weeks. The lower limit of the gestational age was set at 23 weeks. All cases were independently reviewed and validated by two board certified Maternal Fetal Medicine physicians. When disagreement in pregnancy outcome or characteristic arose, the case was re-reviewed and a consensus conference held to determine the final characterization. Twenty-five singleton cases of SPTB ≤34 weeks (N=8 preterm labor, N=12 premature rupture of membranes, N=5 cervical insufficiency) were matched to two control term deliveries by maternal age, race, and gestational age of sampling (+/− 2 weeks). The protocol was approved by institutional review board at BWH, and written informed consent was obtained from all participating women.

CMP Enrichment

Plasma samples were shipped on dry ice to the David H Murdock Research Institute (DHMRI, Kannapolis, NC) and randomized by NX Prenatal to blind laboratory personnel performing sample processing and testing to case/control status. CMP’s were enriched by Size Exclusion Chromatography (SEC) and isocratically eluted using the NeXosome Elution Reagent (in-house proprietary reagent). Briefly, PD-10 columns (GE Healthcare Life Sciences) were packed with 10mL of 2% Agarose Bead Standard (pore size 50 – 150 um) from ABT (Miami, FL), washed with Elution Reagent and stored at 4°C for a minimum of 24 hrs and no longer than 3 days prior to use. On the day of use columns were again washed and 1 mL of thawed neat plasma sample was applied to the column. The circulating microparticles were captured in the column void volume, partially resolved from the high abundant protein peak.15 The samples were processed in batches of 15 to 20 across 4 days to minimize variability between processing individual samples. One aliquot of the pooled CMP column fraction from each clinical specimen, containing 200ug of total protein (determined by BCA) was transferred to a 2 mL microcentrifuge tube (VWR) and shipped on dry ice to Biognosys (Zurich, Switzerland) for proteomic analysis.

Liquid Chromatography-Mass Spectrometry

Quantitative proteomic liquid chromatography-mass spectrometry (LC-MS) analysis was performed by Biognosys AG. Briefly, for each sample a total of 20 ug of protein was lyophilized and then denatured with 8M urea, reduced using dithiothreitol, alkylated with Biognosys alkylation solution, and digested overnight with trypsin (Promega) as previously described.15 Resulting sample peptides were dried using a SpeedVac system and re-dissolved in 45 uL of Biognosys LC solvent and mixed with Biognosys PlasmaDive (extended version 2.0) stable isotope-labeled reference peptide mix containing Biognosys iRT kit.

Then 1 ug of total protein was injected to an in-house packed C18 column (75μm inner diameter and 10 cm column length, New Objective); column material was Magic AQ, 3 μm particle size, 200 Å pore size from Michrom) on a Thermo Scientific Easy nLC nano-liquid chromatography system. LC-MRM assays were measured on a Thermo Scientific TSQ Vantage triple quadrupole mass spectrometer equipped with a standard nano-electrospray source. The LC gradient for LC-MRM was 5–35% solvent B (97% acetonitrile in water with 0.1% FA) in 30 minutes followed by 35–100% solvent B in 2 minutes and 100% solvent B for 8 minutes (total gradient length was 40 minutes). For quantification of the peptides across samples, the TSQ Vantage was operated in scheduled MRM mode with an acquisition window length of 3.25 minutes. The LC eluent was electrosprayed at 1.9 kV and Q1 was operated at unit resolution (0.7 Da). Signal processing and data analysis was carried out using SpectroDive™ - Biognosys’ proprietary software for multiplexed MRM data analysis. A Q-value filter of 1% was applied. Protein concentration was determined based on the normalized 1 ug of protein injected to the LC/MS.

Statistical Analysis

To select informative analytes that differentiate SPTB from term deliveries, the processed protein quantitation data were first subjected to univariate receiver-operating characteristic (ROC) curve analysis.16,17 Bootstrap resampling18 against nulls from sample label permutation was used to control the false-discovery rate (FDR).19 Briefly, for each protein, a ROC analysis was repeated on bootstrap samples from the original data, the mean and standard deviation (SD) of the area-under-curve (AUCs) was estimated. The bootstrap procedure was then applied on the same data again but with sample SPTB status labels randomly permutated. The permutation analysis provided the null results in order to control the FDR and adjust for multiple comparison during the selection of potential protein biomarkers. The Differential Dependency Network (DDN) bioinformatic tool was then applied in order to extract SPTB phenotype-dependent high-order co-expression patterns among the proteins.20 An additional bioinformatic tool, BiNGO, was used to identify gene ontology (GO) categories that were overrepresented in the DDN subnetworks in order to suggest plausible functional links between the observed proteomic dis-regulations and SPTB.21 Finally, to assess the complementary values among the selected proteins and the range of their potential clinically relevant performance, multivariate linear models were derived and evaluated using bootstrap resampling.

Results

The demographic and clinical characteristics of the sample set are presented in Table 1. Maternal age, race, BMI, use of public insurance, smoking during pregnancy, and gestational age at enrollment were similar in both groups. Maternal educational levels were higher in the controls and a greater proportion of the SPTB cases were primaparious and had a prior history of PTB.

Table 1
Baseline characteristics of SPTB vs. term control pregnancies

The 132 proteins evaluated via targeted MRM were individually assessed for ability to differentiate SPTB from term deliveries. By requiring that the mean bootstrap AUCs for each candidate protein be significantly greater than the null (> mean + SD of mean bootstrap AUCs estimated with label permutation; see Figure 1) and excluding proteins with large bootstrap AUCs variances, 62 of the 132 proteins demonstrated robust power for the detection of SPTB. In contrast, using the same criteria with sample label permutation, only 12 proteins would have been selected. The estimated FDR for protein selection was therefore <20% (12/62). These 62 proteins were considered potential candidates for further multivariate analysis. Individually, 25 of the 62 proteins had a p-values <0.10 and an AUC > 0.618 for differentiating SPTB from term controls (Table 2).

Figure 1
Bootstrap ROC analysis to select proteins for detection of SPTBs from term cases. Each protein was plotted as a blue-colored point with mean and SD of the AUCs from bootstrap ROC analysis as x- and y-axis values, correspondingly. Results from the same ...
Table 2
Discriminating single analytes with greatest AUC and lowest p-value (SPTB vs. term controls)

Differential dependency network analysis among the 62 selected proteins identified a number of SPTB phenotype-associated co-expression patterns (Figure 2). A number of GO categories, such as inflammation, wound healing, the coagulation cascade, and steroid metabolism were overrepresented among the DDN analysis co-expression subnetworks. Table 3 provides a listing of the top discriminating pairwise correlations (p-values < 0.001 – 0.069). There were a total of 20 unique proteins that formed the DDN subnetworks. Several of the pairwise correlations (CBPN-TRFE, CPN2-TRFE, A1AG1-MBL2) were markers for inclusion in the term controls rather than the SPTB cases, which suggest a protection against SPTB (Table 3).

Figure 2
Differential Dependency Network (DDN) analysis on the selected proteins identified co-expression patterns associated with STPB. In plot, red or green colors of connections indicate whether the co-expression between the pairs of proteins were observed ...
Table 3
Significance (p-values) of pair-wise connections between nodes (proteins) in DDN analysis identified SPTB phenotype-associated co-expression patterns (subnetworks)

Based on the available sample size, and in order to avoid overtraining, only linear models were evaluated to assess the clinically relevant performance and the variables were limited to all possible combinations of 2 or 3 proteins out of the 20 proteins in Table 3 (1330 models). Each model was derived and evaluated using 200 bootstrap resampled data in order to estimate the median (90%CI) and specificity for ROC AUCs with a fixed sensitivity of 80%. The top 20 models in terms of the lower-bound of 90% CI of AUCs and specificities were listed in Tables 4 and and5,5, respectively. Given limitations imposed by the sample size, the model could not be tested on an independent sample set. To compensate for this the CIs for the panel’s performances in the training dataset were estimated through iterative bootstrap analysis.

Table 4
Median and 90% confidence interval of top 20 models based lower-bound of 90% CI of AUC from ROC analysis (SPTB vs. term controls).
Table 5
Median and 90% confidence interval of top 20 models based lower-bound of 90% CI of specificity at fixed 80% sensitivity (SPTB vs. term controls).

Figure 3 shows the frequency of individual proteins from the DDN analysis being included in the top 20 model’s variable panels. The most frequently “included” proteins were HEMO, KLKB1, and TRFE. In Figure 4 the ROC curve, AUC, and a pair of sensitivity and specificity were plotted for the linear model using a 3 protein panel: HEMO, IC1, and PGRP.

Figure 3
Frequency of DDN-selected proteins from Figure 2 in top 20 multivariate models based on AUC in Table 4 (top) or specificity at fixed sensitivity of 80% in Table 5 (bottom).
Figure 4
ROC curve of example linear models combining 3 proteins: A) using A2MG, HEMO, and MBL2; B) using KLKB1, IC1, and TRFE. With the limited data, the model was not tested on independent samples. However, ROC analysis with bootstrap resampling (See Tables ...

Discussion

Principle findings

This study identifies numerous protein mediators associated with several clinically relevant biological processes that exhibit unique expression profiles by 10–12 weeks gestation among SPTB cases. The protein biomarkers identified are primarily involved in inter-related biological networks linked to coagulation, fibrinolysis, immune modulation and the complement system (Table 6). These systems, in turn, are believed to have an interaction with adaptive immunity and the mediation of inflammatory processes necessary to sustain a successful pregnancy; therefore, we infer that the functioning of these essential processes is mediated, in part, by CMP interactions between various cells and tissues. The potential biological and clinical significance of this approach is in the non-invasive detection and monitoring of protein dysregulation in preterm births and possibly other obstetrical syndromes and conditions.

Table 6
Biological Pathways of Differentiating Microparticle Associated Protein Biomarkers Identified in this Study

Findings in relation to other studies

During pregnancy, there is a complex and intimate interaction between the fetus and mother that includes the direct bathing of the placenta in maternal blood and the incursion of fetal tissue into the maternal decidua in order to facilitate the exchange of metabolic substrates and waste products. There is extensive cross-talk between the two individuals, though the mechanisms are poorly understood. For example, the close contact between maternal and fetal tissues presents an immunological challenge to the mother, but immune tolerance is sustained throughout a healthy pregnancy.22,23 Cross-talk between maternal and fetal tissues must therefore exist.22 CMPs are candidates for communication between maternal and fetal sources as these pregnancy-associated microparticles have already been shown to carry immunomodulatory and regulatory proteins within the maternal circulation.24

Cell-derived microparticles have been described for numerous cells types and bodily fluids.25 Possible cell sources of microparticle-associated proteins described here include vascular endothelium, syncytiotrophoblasts, decidua, immune cells, as well as more distant systemic sources (see Figure 5). Endothelial-derived microparticles have been described to carry numerous adhesion, angiogenesis, and tumor growth molecules.26 Within the context of the present study, we have identified microparticle-associated biomarkers that have actions within the coagulation, inflammation and immune modulation pathways. Several of the top classification models presented include HEMO, KLKB1, A2MG, IC1 and TRFE, with the most frequent components being HEMO, KLKB1 and TRFE. Among these, IC1 is a protease inhibitor that belongs to the Serpin (serum protease inhibitors) superfamily of protease inhibitors. One of its prime functions is the inhibition of spontaneous activation of the complement system; deficiencies of this protein activity cause a rare genetic disorder called hereditary angioedema.

Figure 5
Schematic of microparticle interaction in the setting of gestational tissues.The different potential sources of microparticles from systemic and gestational tissues are indicated, along with the classifications of biomarkers relevant to disease or homeostatic ...

Several complications, including preterm birth, are associated with aberrant complement activation in women with inherited or acquired complement system disorders, while plasma and urine biomarker studies implicate complement proteins and abnormal cascade activation that lead to adverse pregnancy outcomes.27 Complement Component 9 (CO9), a discriminating single analyte ranking in the top 10 proteins with the greatest AUC and lowest p-value, is a component of the membrane attack complex (MAC) that plays a key role in the innate and adaptive immune responses. CO9 works along with complement components C7 and C8 to form pores that insert into target membranes to kill cells. CO7 and CO8 alpha chain are two additional proteins that have been discovered in CMPs in our studies, and it is interesting that the MAC has been localized to trophoblastic basement membranes at sites of villous injury (Figure 5, red arrow.)28 Plasma kallikrein (KLKB1) has been described before as part of a proteomic biomarker panel, used in conjunction with IL-6, to predict preterm labor in patients with intact membranes.29 The protein participates in surface-dependent activation of coagulation and fibrinolysis and is a component of kinin generation. As with many of the CMP markers described here, it is consistent with prevalent theories about the role of inflammation in preterm births and other adverse outcomes.30 Hemopexin (HEMO) acts as a scavenger protein for heme groups released from heme-containg proteins, such as hemoglobin. Free heme groups can generate free radicals and induce significant oxidative damage. HEMO, thus, acts as an anti-oxidant while preserving the body’s iron stores. Oxidative stress is yet another pathologic factor associated with spontaneous preterm births, though it is also considered part of normal term parturition mechanisms.31 Taken together, the top protein discriminators in the classifier are consistent with known mechanisms of preterm birth and adverse outcomes in pregnancy and parturition, but future in vitro and in vivo experiments will be necessary to elucidate their role. Trophoblastic-derived microparticles may play a major role in providing a protective effect to modulate inflammation at the maternal-fetal interface. Uncontrolled complement activation is prevented in normal pregnancies by molecules localized to plasma membranes of trophoblasts, while unrestricted complement activation is associated with complications of pregnancy and fetal death.32 If validated, the coagulation-related, microparticle proteins identified in this study could serve as a nexus for initiating the coagulation cascade, or, as suggested by identified negative regulator factors, CMP’s transporting regulatory proteins within the maternal-fetal environment may inhibit coagulation and complement proteins (see Table 6), thereby also limiting a local inflammatory response.33

It is increasingly understood that immune dysregulation, aberrant coagulation and intrauterine inflammation are common to a large proportion of cases of SPTB.7 Considering the breadth of microparticle protein markers described here that span coagulation, inflammation, and immunomodulatory pathways, while recalling the important role of intrauterine inflammation in SPTB,7, 30,34 it is not surprising that we are observing evidence of microparticle mediated dysfunction as early as the end of the first trimester. Contemporary theory suggests that, if not all, at least a high proportion of adverse pregnancy outcomes have their pathophysiologic origins in early pregnancy.7,12,13 Abnormalities of early placentation and trophoblast function have been observed not only in pregnancies complicated by hypertension, but also in approximately 30% of those experiencing SPTB.35 Our findings suggest that the state, condition, and function of these cells at the maternal-fetal interface during this critical period have already predisposed the pregnancy to adverse outcomes. Others have observed that the concentration of placental-specific microparticles increases significantly with advancing gestation,36 this suggests that early perturbations in microparticle-mediated signaling may gradually become magnified as the pregnancy progresses. Ultimately, these anomalies in the maternal fetal cross-talk may become sufficiently great enough to cause a ‘network crash’ of the systems that were facilitating tolerance resulting in a spontaneous preterm birth.

Clinical and research implications

One of the traditional hindrances to a greater understanding of the underlying causes of SPTB is the difficulty of investigating the maternal-fetal interface itself and the unique nature of human placentation. The intrauterine space is both physically and ethically remote. As such, this is perhaps why, with the possible exception of the measurement of cervical length by ultrasound, little recent progress has been made in the development of useful biomarkers to stratify patients according to risk of SPTB.3739 We have demonstrated here a potential means to move this area of inquiry forward. Differences in the protein content of microparticles likely represent an untapped source of information regarding biology of the maternal-fetal interface. We have further demonstrated that improved specificity (as indicated by increased AUCs) can be obtained with the simultaneous consideration of multiple protein biomarkers associated with a CMP-enriched blood plasma fraction. This is not surprising, given the present conceptualization of maternal-fetal tolerance as a complicated multivariate network with many inputs rather than a more simple bivariate relationship.40 Fears of over-fitting limited the number of markers we could reasonably test, although, increased sample size in future studies would allow the testing of more simultaneous markers and improve the ability to characterize the overall state of the network.

Strengths and limitations

This analysis should be interpreted in the context of the study design. First, it suffers from the limitations common to all observational studies, namely that observation cannot discern causation. Second, we chose to compare SPTB ≤34 weeks given the increased neonatal morbidity in that gestational age range and our belief that the homogeneity of the pathology of SPTB in these earlier gestational age ranges are likely to be greater. For these very reasons, however, applicability to gestational ages after 34 weeks may not be as clear. Finally, we also acknowledge that for reasons of cost, balanced against sample size, we are generalizing from a single data set. While we have taken great statistical and methodological pains to reduce the risks of false discovery and over-fitting, ideally these findings should be replicated in a larger validation cohort set. We are currently undertaking these additional investigations to validate this preliminary analysis.

Conclusions

This study has identified statistically significant CMP associated protein biomarker candidates and multiplex panels associated with biological processes relevant to pregnancy that are already unique in their expression profiles at 10–12 weeks among women who go on to deliver spontaneously at ≤34 weeks. Some may suggest that we are premature in attempting the development of a predictive test without a universally agreed upon therapeutic modality. However, we contend that we are laying the groundwork for the clinical stratification of patients at risk of SPTB well before clinical presentation. Such identification would allow the application of increased observation and the possible application of prophylactic therapies such as progesterone, which together may significantly improve the management of these patients. We acknowledge, however, that additional work will need to be done to demonstrate the effectiveness of such therapies in women stratified using microparticle-based tests, well before clinical presentation.

Acknowledgments

We would like to thank the participants and field staff at BWH, the staff at David H. Murdock Research Institute in Kannapolis, NC (Qi Jiang, Lisa Dewey, and Emily Jackson) for CMP isolation/enrichment, and Claudia Escher at Biognosys AG in Zurich, Switzerland for the quantitative proteomic LC-MS analysis.

Funding: Primary funding for sample and statistical analysis was provided by NX Prenatal Inc. Funding was additionally provided by the National Institute of Environmental Health Sciences, National Institutes of Health (R01ES018872) to provide salary support for D.E.C. and T.F.M.

Footnotes

K.S.G, R.C.D., A.M.E., G.P, and B.B are employed by NX Prenatal Inc. K.R. and Z.Z. are paid consultants to NX Prenatal Inc. The remaining authors report no conflict of interest.

Presented, in part, as an oral presentation at the 36th annual meeting of the Society of Maternal-Fetal Medicine, Atlanta, Georgia, February 1–6, 2016.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

1. Liu L, Oza S, Hogan D, et al. Global, regional, and national causes of child mortality in 2000–13, with projections to inform post-2015 priorities: an updated systematic analysis. Lancet. 2015;385:61698–6. [PubMed]
2. Katz J, Lee AC, Kozuki N, et al. Mortality risk in preterm and small-for-gestational-age infants in low-income and middle-income countries: a pooled country analysis. Lancet. 2013;382:417–25. [PMC free article] [PubMed]
3. March of Dimes. [Accessed November 16, 2015];Preterm Infants: Transition to home and follow-up. http://www.marchofdimes.org/chapterassets/files/nv_abdulla_preterm_infant.pdf.
4. Behrman R, Butlereds A. IOM. Preterm birth: causes, consequences, and prevention. Insitute of Medicine, National Academies Press; Washington, DC: 2006. https://www.iom.edu/Reports/2006/Preterm-Birth-Causes-Consequences-and-Prevention.aspx.
5. Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology and causes of preterm birth. Lancet. 2008;371:75–84. [PubMed]
6. McElrath TF, Hecht JL, Dammann O, et al. Pregnancy disorders that lead to delivery before the 28th week of gestation: an epidemiologic approach to classification. Am J Epidemiol. 2008;168:980–9. [PMC free article] [PubMed]
7. Romero R, Dey SK, Fisher SJ. Preterm labor: one syndrome, many causes. Science. 2014;345:760–5. [PMC free article] [PubMed]
8. Gupta AK, Holzgreve W, Hahn S. Decrease in lipid levels of syncytiotrophoblast micro-particles reduced their potential to inhibit endothelial cell proliferation. Arch Gynecol Obstet. 2008;277:115–9. [PubMed]
9. Redman CW, Tannetta DS, Dragovic RA, Gardiner C, Southcombe JH, Collett GP, Sargent IL. Review: Does size matter? Placental debris and the pathophysiology of pre-eclampsia. Placenta. 2012;33:S48–54. [PubMed]
10. Salomon C, Kobayashi M, Ashman K, Sobrevia L, Mitchell MD, Rice GE. Hypoxia-induced changes in the bioactivity of cytotrophoblast-derived exosomes. PLoS One. 2013;8:e79636. [PMC free article] [PubMed]
11. Rajakumar A, Cerdeira AS, Rana S, et al. Transcriptionally active syncytial aggregates in the maternal circulation may contribute to circulating soluble fms-like tyrosine kinase 1 in preeclampsia. Hypertension. 2012 Feb;59(2):256–64. [PMC free article] [PubMed]
12. Gagnon R. Placental insufficiency and its consequences. Eur J Obstet Gynecol Reprod Biol. 2003;110:S99–107. [PubMed]
13. Masoura S, Kalogiannidis IA, Gitas G, et al. Biomarkers in pre-eclampsia: a novel approach to early detection of the disease. J Obstet Gynaecol. 2012;32:609–16. [PubMed]
14. McElrath TF, Lim KH, Pare E, et al. Longitudinal evaluation of predictive value for preeclampsia of circulating angiogenic factors through pregnancy. Am J Obstet Gynecol. 2012;207:407–14. [PubMed]
15. Ezrin AM, Brohman B, Willmot J, et al. Circulating Serum-Derived Microparticles Provide Novel Proteomic Biomarkers of Spontaneous Preterm Birth. AM J Perinatol. 2015;32:605–14. [PubMed]
16. Fawcett T. An introduction to ROC analysis. Pattern Recognition Letters. 2006;27:861–74.
17. Robin X, Turck N, Hainard AN, et al. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics. 2011;12:77. [PMC free article] [PubMed]
18. Carpenter J, Bithell J. Bootstrap confidence intervals: when, which, what? A practical guide for medical statisticians. Statistics in Medicine. 200(19):1141–64. [PubMed]
19. Xie Y, Pan W, Khodursky AB. A note on using permutation-based false discovery rate estimates to compare different analysis methods for microarray data. Bioinformatics. 2005;21:4280–88. [PubMed]
20. Tian Y, Zhang B, Hoffman EP, et al. KDDN: an open-source Cytoscape app for constructing differential dependency networks with significant rewiring. Bioinformatics. 2015;32:287–89. [PMC free article] [PubMed]
21. Maere S, Heymans K, Kuiper M. BiNGO: a Cytoscape plugin to assess overrepresentation of gene ontology categories in biological networks. Bioinformatics. 2005;21:3448–49. [PubMed]
22. Lash GE. Molecular Cross-Talk at the Feto-Maternal Interface. Cold Spring Harb Perspect Med. 2015 Sep 18; doi: 10.1101/cshperspect.a023010. pii: a023010. [PubMed] [Cross Ref]
23. Mor G, Kwon JY. Trophoblast-microbiome interaction: a new paradigm on immune regulation. Am J Obstet Gynecol. 2015;213:S131–7. [PubMed]
24. Sabapatha A, Gercel-Taylor C, Taylor DD. Specific isolation of placenta-derived exosomes from the circulation of pregnant women and their immunoregulatory consequences. Am J Reprod Immunol. 2006;56:345–55. [PubMed]
25. van der Pol E, Böing AN, Harrison P, Sturk A, Nieuwland R. Classification, functions, and clinical relevance of extracellular vesicles. Pharmacol Rev. 2012;64:676–705. [PubMed]
26. Dignat-George F, Boulanger CM. The many faces of endothelial microparticles. Arterioscler Thromb Vasc Biol. 2011;31:27–33. [PubMed]
27. Regal JF, Gilbert JS, Burwick RM. The complement system and adverse pregnancy outcomes. Mol Immunol. 2015;67:56–70. [PMC free article] [PubMed]
28. Rampersad R, Barton A, Sadovsky Y, Nelson DM. The C5b-9 membrane attack complex of complement activation localizes to villous trophoblast injury in vivo and modulates human trophoblast function in vitro. Placenta. 2008;29:855–61. [PMC free article] [PubMed]
29. Cobo T, Palacio M, Navarro-Sastre A, et al. Predictive value of combined amniotic fluid proteomic biomarkers and interleukin-6 in preterm labor with intact membranes. Am J Obstet Gynecol. 2009;200:499.e1–6. [PubMed]
30. Murtha AP, Menon R. Regulation of fetal membrane inflammation: a critical step in reducing adverse pregnancy outcome. Am J Obstet Gynecol. 2015;213:447–8. [PubMed]
31. Menon R. Oxidative stress damage as a detrimental factor in preterm birth pathology. Front Immunol. 2014;5:567. [PMC free article] [PubMed]
32. Girardi G, Bulla R, Salmon JE, Tedesco F. The complement system in the pathophysiology of pregnancy. Mol Immunol. 2006;43:68–77. [PubMed]
33. Oikonomopoulou K, Ricklin D, Ward PA, Lambris JD. Interactions between coagulation and complement-their role in inflammation. Semin Immunopathol. 2012;34:151–65. [PMC free article] [PubMed]
34. Rey G, Pereyra S, Velazquez T, et al. The Effect of Inflammation on Preterm. In: Morrison J, editor. Preterm Birth – Mother and Child. Croatia: InTech; 2012. pp. 183–200.
35. Kim YM, Bujold E, Chaiworapongsa T, et al. Failure of physiologic transformation of the spiral arteries in patients with preterm labor and intact membranes. Am J Obstet Gynecol. 2003;189:1063–9. [PubMed]
36. Sarker S, Scholz-Romero K, Perez A, et al. Placenta-derived exosomes continuously increase in maternal circulation over the first trimester of pregnancy. J Transl Med. 2014;12:204. [PMC free article] [PubMed]
37. Conde-Agudelo A, Papageorghiou AT, Kennedy SH, Villar J. Novel biomarkers for the prediction of the spontaneous preterm birth phenotype: a systematic review and meta-analysis. BJOG. 2011;118:1042–54. [PubMed]
38. Brou L, Almli LM, Pearce BD, Bhat G, Drobek CO, Fortunato S, Menon R. Dysregulated biomarkers induce distinct pathways in preterm birth. BJOG. 2012;119:458–73. [PubMed]
39. Sananès N, Langer B, Gaudineau A, Kutnahorsky R, Aissi G, Fritz G, Boudier E, Viville B, Nisand I, Favre R. Prediction of spontaneous preterm delivery in singleton pregnancies: where are we and where are we going? A review of literature. J Obstet Gynaecol. 2014 Aug;34(6):457–61. [PubMed]
40. Huppertz B, Berghold VM, Kawaguchi R, Gauster M. A variety of opportunities for immune interactions during trophoblast development and invasion. Am J Reprod Immunol. 2012;67:349–57. [PubMed]