To systematically review the literature on human gene expression data of placental tissue in pre-eclampsia and to characterize a meta-signature of differentially expressed genes in order to identify novel putative diagnostic markers.
Medline through 11 February 2011 using MeSH terms and keywords related to placenta, gene expression and gene expression arrays; GEO database using the term “placent*”; and reference lists of eligible primary studies, without constraints.
From 1068 studies retrieved from the search, we included original publications that had performed gene expression array analyses of placental tissue in the third trimester and that reported on differentially expressed genes in pre-eclampsia versus normotensive controls. Two reviewers independently identified eligible studies, extracted descriptive and gene expression data and assessed study quality. Using a vote-counting method based on a comparative meta-profiling algorithm, we determined a meta-signature that characterizes the significant intersection of differentially expressed genes from the collection of independent gene signatures.
We identified 33 eligible gene expression array studies of placental tissue in the 3rd trimester comprising 30 datasets on mRNA expression and 4 datasets on microRNA expression. The pre-eclamptic placental meta-signature consisted of 40 annotated gene transcripts and 17 microRNAs. At least half of the mRNA transcripts encode a protein that is secreted from the cell and could potentially serve as a biomarker.
In addition to well-known and validated genes, we identified 14 transcripts not reported previously in relation to pre-eclampsia of which the majority is also expressed in the 1st trimester placenta, and three encode a secreted protein.
Citation: Kleinrouweler CE, van Uitert M, Moerland PD, Ris-Stalpers C, van der Post JAM, et al. (2013) Differentially Expressed Genes in the Pre-Eclamptic Placenta: A Systematic Review and Meta-Analysis. PLoS ONE 8(7): e68991. doi:10.1371/journal.pone.0068991
Editor: Cees Oudejans, VU University Medical Center, The Netherlands
Received: March 26, 2013; Accepted: June 4, 2013; Published: July 12, 2013
Copyright: © 2013 Kleinrouweler et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: C. Emily Kleinrouweler is supported by a PhD Scholarship from the AMC Graduate School. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Pre-eclampsia, a pregnancy-specific disorder characterized by the development of de novo hypertension and proteinuria during the second half of gestation, is a major obstetric problem that contributes substantially to maternal and perinatal morbidity and mortality worldwide .
The exact mechanisms involved in the development of pre-eclampsia are not well defined, but it is generally accepted that the pathogenesis of pre-eclampsia initiates at the time of trophoblast invasion and remodeling of the spiral arteries during the first and early second trimester of pregnancy. Impaired invasion of myometrial spiral arteries relates to abnormal placentation, which in turn results in the release of various components from the intervillous space into the maternal circulation. Consequently, exaggerated maternal endothelial activation and a generalized hyper-inflammatory state precede the onset of clinical symptoms of pre-eclampsia –.
Gene expression array analysis is an important and powerful tool to identify genes differentially expressed between tissues from normotensive and pre-eclamptic pregnancies. Some key targets involved in the pathogenesis of pre-eclampsia, such as Fms-like tyrosine kinase 1 (FLT1) and Endoglin (ENG) were discovered using this approach , . Although many studies with the aim to identify placenta-specific gene expression patterns in pre-eclampsia have been published, no clear pre-eclampsia-specific gene expression consensus signature (a list of differentially expressed genes) has been defined. This is probably due to relatively small study sample sizes, and variability in platforms, patient characteristics and statistical methods used .
The aim of this study was to systematically review the literature on array-based gene-expression studies of placental tissues in pre-eclampsia in order to determine a common expression signature and identify novel putative diagnostic leads.
This systematic review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) and Meta-analysis Of Observational Studies in Epidemiology (MOOSE) guidelines. ,  A protocol was not specified before undertaking this systematic review.
We performed an electronic search on 11 February, 2011 in Medline (from inception) through PubMed and GEO (Gene Expression Omnibus) without language or publication date restrictions to identify all articles reporting on differentially expressed genes in human placental samples relating to pre-eclampsia.
The electronic search strategy for Medline was based on MeSH terms and keywords related to placenta, gene expression and gene expression arrays. Document S1 lists the full search strategy. Reference lists of eligible primary studies were checked to identify cited articles not captured by the electronic search. The search strategy for GEO consisted of the term “placent*”. The identified records were scanned for references not identified by the search in Medline.
Eligible were those original publications that had performed gene expression array analyses of human placental tissue in the third trimester and that reported on differentially expressed genes in pregnancies that had pre-eclampsia compared to normotensive controls. Only studies presenting their results as a gene signature (defined as a list of differentially expressed genes, selected according to criteria determined by the authors of the paper) were included.
Studies were selected in a staged process.
First, two reviewers (MM and GA) independently scrutinized titles and abstracts of all retrieved references to select potentially eligible articles. Full text papers of references selected by at least one reviewer were obtained. Second, two or three reviewers (EK, MM and GA) independently examined these full text papers to see whether they met the inclusion criteria. In case of multiple studies of the same set of women we included only the most recent or most complete one. Disagreements about inclusion were resolved by consensus.
For each study, data on clinical characteristics of included women (age, obstetric history, parity), diagnostic criteria for pre-eclampsia, characteristics of the gene expression array experiment (tissue, platform, details on sampling and preparation including extraction of RNA), details of the statistical analysis, the gene signature and other experimental results were extracted independently by two experienced reviewers (EK and MM) using standardized data extraction forms.
As there are no formal quality assessment approaches for this type of study, we assessed study quality by discussing reported details of sampling, testing and analysis processes among four of the authors (EK, MM, CR and GA). Attention was paid to the results presented and whether the reported list of differentially expressed genes had been subject to additional selection, such that not the whole list was reported.
Acceptable diagnostic criteria for pre-eclampsia were persistent high systolic (≥140 mm Hg) or diastolic (≥90 mm Hg) blood pressure combined with proteinuria (≥0.3 g/24 hours or a dipstick result of ≥1+, equivalent to 30 mg/dL in a single urine sample or spot urine protein/creatinine ratio ≥30 mg protein/mmol creatinine) of new onset after 20 weeks of gestation, according to the International Society for the Study of Hypertension in Pregnancy criteria .
From each included paper we extracted the published gene signature and selected the official gene symbol as gene identifier. If the gene symbol was not given, we mapped the other gene information (accession number or GenBank number, gene name, probe ID and UniGene ID) to the official gene symbol using the NCBI Gene database. We checked for each symbol whether this was an official symbol or an alias or both. If the symbol was not an official symbol but only an alias, we used the other gene information to obtain the official symbol.
On two occasions, multiple papers were published on the same set of placental samples. In four papers by Pang and Xing – various gene expression platforms targeting different, but relatively small subsets of the whole genome were used on the same samples to report on differentially expressed genes from different biological processes (for example: metabolism-related genes, apoptosis-related genes, trophoblast-invasion-associated genes). Järvenpää et al. ,  presented partial data of the same gene expression analysis in two separate papers. In both cases we created one signature per set of samples consisting of the combined results from the different papers.
Using all gene signatures presented in the included papers, we determined a significant meta-signature using a vote counting approach based on a modified version of the so-called comparative meta-profiling method . This method identifies the significant intersection of multiple gene expression signatures from a collection of datasets and consists of four steps: Step 1: For each gene mentioned, we counted how often it appeared in a signature and ordered the genes by number of appearance. This way we obtained (N1, …, NS), with Nx: the number of genes that appear x times in a signature, and S: the total number of signatures. Step 2: We created a random signature as follows. For each signature, the number of genes remained the same as in the original signature, but the gene symbols were replaced by symbols randomly drawn from a (selected) list of 7000 official gene symbols, representing the intersection of all genes targeted by the whole genome platforms included in the meta-analysis of which the NCBI GeneID is publicly available. For the microRNA data we took a set of 854 publicly available annotated microRNAs. For these random signatures we repeated Step 1, and obtained (E1, …, ES), with Ex: the number of genes that appear x times in a random signature. Step 3: Step 2 was repeated 1000 times. Step 4: For each i = 1, …, S, we computed the empirical probability distribution function for Ei+…+ES. Then we determined the significance of the observed number of genes present in at least i signatures, Ni+…+NS. The associated p-value is the probability of Ei+…+ES ≥ Ni+…+NS, reflecting the probability that the number of genes present in at least i random signatures is at least as large as the observed number. If there was an i such that the p-value was <0.05, then a significant meta-signature was found. If there were multiple i for which the p-value was <0.05, we took the smallest i. The genes forming the meta-signature were listed.
We performed the analyses twice: once on those studies that presented the complete list of all genes identified as differentially expressed, and once on all studies, thereby also including studies that reported an incomplete (partial) gene signature. All analyses were performed using R version 2.14.0 (The R Foundation for Statistical Computing, Vienna, Austria).
Functional interaction networks of proteins encoded by the genes in the meta-signature were visualized using STRING 9.05 .
The results of the search and the selection process are shown in Figure 1: The search in Medline provided a total of 1030 citations. In addition, the search in GEO yielded 67 eligible datasets, 65 with a publication, of which 38 publications had not been identified by the Medline search. Of the total of 1068 unique citations, 1023 studies were discarded after reviewing titles and abstracts. The full text of the remaining 45 papers was examined in more detail. Of these, 14 studies did not meet the inclusion criteria as described (see Document S2). Two new studies where identified from reference lists of eligible studies and by keeping up with the literature. Finally, 33 studies met the inclusion criteria and were included in this systematic review. All studies reported on third trimester placental samples of which 30 studies determined the mRNA expression profile –, – and 4 studies determined the microRNA expression profile , –. All studies were published between 2002 and 2011. Study sample size ranged from 3 to 62 included women. We provide an overview of all studies in Table S1.
Figure 1. PRISMA flowchart summarizing the identification and selection of studies for inclusion in this systematic review.doi:10.1371/journal.pone.0068991.g001
Tables S1 and S2 summarize the details of the study characteristics, as described in the original papers, that we obtained for quality assessment. Overall, the impression of study quality was fair and we did not exclude any studies for poor quality. Most studies had adequately reported an acceptable definition of pre-eclampsia used in the study, the type and location of tissue sampling, the storage and RNA isolation conditions, and the array platform used. Most samples were stored in liquid nitrogen or RNAlater (Qiagen, Valencia, USA or Ambion, Austin, USA), and most studies used either RNeasy (Qiagen, Valencia, USA) or Trizol (Invitrogen, Carlsbad, USA) for RNA isolation. Gene expression array platforms were very diverse, ranging from whole genome arrays to nylon membrane hybridizations, most of them commercially available. Most studies had a case-control design with an appropriate control group and adequately reported characteristics of the case and control women (age, parity, pregnancy details and gestational age at delivery). The major obstacles in the quality assessment were the biased listing of incomplete gene signatures and the diversity in statistical procedures used (normalization and analysis methods).
Fourteen third-trimester placental tissue RNA expression studies that reported a complete gene signature were included in the meta-analysis. These studies represented 159 samples of pre-eclamptic placenta and 188 control samples. The case where a gene was present in 3 or more of 14 complete gene signatures turned out to be significant (p<0.001). According to this criterion, 17 of the 840 genes present in at least one signature were incorporated in the meta-signature. Most genes were up-regulated in pre-eclampsia compared to normotensive placentas (Table 1, Table S3). The meta-analysis with complete and incomplete signatures combined covered a total of 254 pre-eclampsia samples and 278 control samples from 30 studies. As the four studies from Pang and Xing were considered to form one signature, as were the two studies by and Järvenpää et al., in total 26 gene signatures were included in the analysis. A gene had to be present in 3 or more of these 26 signatures in order to be part of the significant meta-signature (p<0.001). This way, 40 of 1343 genes were incorporated in the meta-signature, in which a large majority of genes was up-regulated in pre-eclampsia (Table 1, Table S3).
Table 1. Meta signature of third-trimester placental tissue mRNA expression.doi:10.1371/journal.pone.0068991.t001
The 4 studies on third-trimester placental tissue microRNA expression represented 39 pre-eclampsia samples and 38 control samples. A gene had to be present in 2 or more signatures in order to be part of the significant meta-signature (p = 0.001) resulting in a meta-signature consisting of 17 out of 127 miRNAs (Table 2).
Table 2. Meta signature of 4 studies on third-trimester placental tissue microRNA expression.doi:10.1371/journal.pone.0068991.t002
Interaction Network Analysis
Using high confidence scores based on experimental data and associations in curated protein databases and text mining, 15 of the 40 signature genes could be mapped into an interaction network (Figure 2). Subsequent clustering identified at least four different functional clusters. In addition, 60% of the gene products are secreted by cells, according to the DAVID functional annotation tool , and therefore potentially detectable in the maternal circulation (Table 1, Table S3).
Figure 2. Functional networks within the meta-signature.
To build a functional network map the 40 meta-signature genes were analyzed using STRING software. Interactions are based on information from experimental and curated protein databases, and text mining with a high confidence setting of 0.8; the resulting networks were clustered using Markov clustering.doi:10.1371/journal.pone.0068991.g002
The placenta is the key organ involved in pre-eclampsia. Identification of genes differentially expressed between normotensive and pre-eclamptic placentas is essential to understand the molecular mechanisms involved in the etiology of the disease and to develop novel biomarkers or identify therapeutic targets. This systematic review provides an overview of placental tissue gene expression array studies in pre-eclampsia. It is the first study that extracts a meta-signature from these individual studies consisting of up-regulated and down-regulated genes in pre-eclampsia as identified in multiple publications.
When all 30 studies generating signatures on placental tissue mRNA expression are included, the pre-eclamptic placental meta-signature consists of 40 genes. As authors reporting an incomplete list of genes may be driven by bias towards the genes they list (for instance only specific genes of the authors’ particular interest), we also generated a meta-signature from the 14 complete gene-lists by which we obtained a meta-signature of 17 genes. As the ratio of the number of times a gene is present in a signature compared to the total number of datasets is similar for both meta-signatures we conclude that the inclusion bias does not affect the overall meta-signature very strongly.
Both meta-signatures have identical well-known and validated genes on top of their list. The relation of the angiogenesis-related cluster FLT1, VEGFA, and PlGF to pre-eclampsia has been extensively documented. Members of this cluster are used as biomarkers and have been shown to play a functional role in pre-eclampsia –. Also the clusters around INH(B)A, CGA/CGB, and LEP have previously been reported to be affected in the pre-eclamptic placenta and some cluster members have been evaluated in the circulation of affected women –.
In both the meta-signature derived from complete and incomplete signatures, the most frequently occurring genes are only mentioned in approximately one third of the individual signatures. Variation in technical aspects, as reported in Tables S1 and S2, may contribute to inconsistencies. As studies were published between 2002 and 2011, the data also reflect the development of more sophisticated array techniques in this decade, including an increased number of genes interrogated and improved gene annotations. A major shortcoming is the variability between studies in statistical methods used for detecting differentially expressed genes. We repeatedly encountered inadequate reporting of statistical procedures and their results, and some studies used inappropriate or not stringent enough statistical criteria. However, as our comparative meta-profiling method extracts only results that are relatively consistent across studies, the contribution of individual biased results is limited.
There was good overall agreement on the direction of differential gene expression among studies. An exception seems to be the data of Mayor-Lynn et al.  that show a deviant direction for CRH, EBI3, PAPPA2, and CYP11A1 compared to the other studies. For CRH, the authors report a more than 10-fold down-regulation of CRH mRNA by microarray analysis in pre-eclamptic placentas while subsequent RT-QPCR in the same tissue shows a 5-fold up-regulation of CRH mRNA. As the raw data are not publically available, the reason for this discrepancy is untraceable.
Apart from transcripts previously related to pre-eclampsia, the meta-signature includes 14 genes for which no information with regard to pre-eclampsia has been reported. They encode structural proteins (KRT19, PLEC, SPAG4, VIM), plasma membrane proteins with a channel/transporter or receptor function (AQP1, PVRL4, SEMA4C, SLCO2A1), proteins involved in immunomodulation (EBI3, TNFSF10), a mitochondrial protein (RDH13), a nuclear protein (BHLHE40) and cytosolic proteins (PHYHIP, SASH1). With the exception of VIM, the expression of all these genes is up-regulated in pre-eclamptic placenta.
Interestingly, 3 of these 14 genes encode a secreted protein that might be detectable in the maternal circulation. However, as the placenta deports large amounts of cellular material into the maternal circulation, non-secreted proteins should not be excluded as candidate biomarkers , . Therefore, placental expression in early pregnancy may be more relevant, but available information on this condition is very limited. There are two studies reporting mRNA expression data of chorionic villous samples from women that developed pre-eclampsia later during their pregnancy. Although one study  reports two differentially expressed genes (FSTL3 and IGFBP1) that are also included in our meta-signature, these genes are not reported in the other study . Analysis of the data from the first study  combined with data from 1st trimester placental mRNA expression  shows that 10 of the 14 genes (that are novel in relation to pre-eclampsia) included in our meta-signature are expressed in the early placenta (Table S3).
Due to the limited number of available studies, meta-analysis of the microRNA expression studies resulted in a less robust meta-signature. In particular the direction (up/down) of the microRNA expression showed a high level of ambiguity between studies. As microRNA expression is considered a regulatory layer for silencing gene expression predominantly by degradation of their target mRNAs , our finding of overall down-regulation of microRNA expression in pre-eclamptic placenta supports our results that most gene transcripts in the meta-signature are up-regulated.
Meta-analysis of raw whole genome expression data would partly overcome the methodological variation encountered in this paper and should be undertaken to validate and extend the novel leads with respect to diagnosis and putative therapeutic targets identified here.
In addition, the combined analysis of microRNA and mRNA meta-signatures is a further necessary step to comprehend the overall deviating transcription profile of the pre-eclamptic placenta.
Clinical data: characteristics of all studies and women included in this systematic review and meta-analysis, as they were reported in the original papers. Data on maternal age, parity and gestational age at delivery are reported as mean (standard deviation) or median (range), unless indicated otherwise. FGR: fetal growth restriction; IUGR: intra uterine growth restriction; PE: pre-eclampsia.
Experimental data: details of the sampling, testing and analysis processes of all studies, as they were reported in the original papers. ANOVA: analysis of variance; PE: pre-eclampsia. a: A complete list is defined as the complete list of genes identified as differentially expressed according to the studies’ selection criteria.
Meta signature of third-trimester placental tissue mRNA expression. Genes present in 3 or more complete signatures, or 3 or more signatures in total, are part of the meta-signature. Genes incorporated in the meta-signature are ordered by (1) the total number of publications that reported the gene as differentially expressed and (2) alphabetically. References printed in bold indicate publications that reported a complete signature. Enquobahrie 2008, Kang 2011, Pang (combined) and Vaiman 2005 did not report the direction of differential expression (upregulation or downregulation). For comparison, the number of times a gene is present in complete signatures below the significance threshold is shown in grey. *: Results were inconsistent among studies, the majority of studies reported the result presented here. –: No consensus among studies: equal numbers of studies reported upregulated expression and downregulated expression in pre-eclampsia. 1st trimester expression was scored using microarray data from control chorionic villous sampling (NCBI Gene Expression Omnibus GSE12767) and 1st trimester placenta samples (NCBI Gene Expression Omnibus GSE9984). Microarrays were analyzed for expression using The Gene Expression Barcode 2.0 (http://barcode.luhs.org/). ‘+’ indicates that in >90% of the samples expression of the gene was detected, ‘–’ indicates that in <10% of the samples expression of the gene was detected. The remaining samples showed a variable degree of expression (‘+/−’).
Electronic search strategy PubMed.
Studies that were excluded after screening of full text and new studies that were included, as described in Figure 1.
PRISMA 2009 Checklist.
Conceived and designed the experiments: JAMvdP GBA. Performed the experiments: CEK MvU GBA. Analyzed the data: MvU PM CRS GBA. Contributed reagents/materials/analysis tools: MvU PM. Wrote the paper: CEK MvU PM CRS JAMvdP GBA.
- 1. Steegers EA, von Dadelszen P, Duvekot JJ, Pijnenborg R (2010) Pre-eclampsia. Lancet 376: 631–644.
- 2. Roberts JM, Hubel CA (2009) The Two Stage Model of Preeclampsia: Variations on the Theme. Placenta 30: 32–37.
- 3. Huppertz B (2008) Placental origins of preeclampsia: challenging the current hypothesis. Hypertension 51: 970–975.
- 4. Maynard SE, Min JY, Merchan J, Lim KH, Li J, et al. (2003) Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest 111: 649–658.
- 5. Venkatesha S, Toporsian M, Lam C, Hanai J, Mammoto T, et al. (2006) Soluble endoglin contributes to the pathogenesis of preeclampsia. Nat Med 12: 642–649.
- 6. Ein-Dor L, Kela I, Getz G, Givol D, Domany E (2005) Outcome signature genes in breast cancer: is there a unique set? Bioinformatics 21: 171–178.
- 7. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, et al. (2009) The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ 339: b2700.
- 8. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, et al. (2000) Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA 283: 2008–2012.
- 9. Brown MA, Lindheimer MD, de Swiet M, Van Assche A, Moutquin JM (2001) The classification and diagnosis of the hypertensive disorders of pregnancy: statement from the International Society for the Study of Hypertension in Pregnancy (ISSHP). Hypertens Pregnancy 20: IX–XIV.
- 10. Pang ZJ, Xing FQ (2004) DNA microarrays detect the expression of apoptosis-related genes in preeclamptic placentas. J Perinat Med 32: 25–30.
- 11. Pang ZJ, Xing FQ (2003) Comparative study on the expression of cytokine–receptor genes in normal and preeclamptic human placentas using DNA microarrays. J Perinat Med 31: 153–162.
- 12. Pang ZJ, Xing FQ (2004) Comparative profiling of metabolism-related gene expression in pre-eclamptic and normal pregnancies. Arch Gynecol Obstet 269: 91–95.
- 13. Pang ZJ, Xing FQ (2003) Expression profile of trophoblast invasion-associated genes in the pre-eclamptic placenta. Br J Biomed Sci 60: 97–101.
- 14. Jarvenpaa J, Vuoristo JT, Santaniemi M, Ukkola O, Savolainen ER, et al. (2009) Adiponectin induced placental cell apoptosis could be mediated via the ADIPOR1-receptor in pre-eclampsia with IUGR. J Perinat Med 37: 257–262.
- 15. Jarvenpaa J, Vuoristo JT, Savolainen ER, Ukkola O, Vaskivuo T, et al. (2007) Altered expression of angiogenesis-related placental genes in pre-eclampsia associated with intrauterine growth restriction. Gynecol Endocrinol 23: 351–355.
- 16. Rhodes DR, Yu J, Shanker K, Deshpande N, Varambally R, et al. (2004) Large-scale meta-analysis of cancer microarray data identifies common transcriptional profiles of neoplastic transformation and progression. Proc Natl Acad Sci U S A 101: 9309–9314.
- 17. Szklarczyk D, Franceschini A, Kuhn M, Simonovic M, Roth A, et al. (2011) The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored. Nucleic Acids Res 39: D561–D568.
- 18. Ahr A, Holtrich U, Karn T, Reitter A, Rody A, et al. (2006) [Identification of preeclampsia by cDNA-gene expression profiling in human placentas and serum – a pilot study]. Zentralbl Gynakol 128: 138–142.
- 19. Centlow M, Wingren C, Borrebaeck C, Brownstein MJ, Hansson SR (2011) Differential gene expression analysis of placentas with increased vascular resistance and pre-eclampsia using whole-genome microarrays. J Pregnancy 2011: 472354.
- 20. Enquobahrie DA, Meller M, Rice K, Psaty BM, Siscovick DS, et al. (2008) Differential placental gene expression in preeclampsia. Am J Obstet Gynecol 199: 566.e1–11.
- 21. Gack S, Marme A, Marme F, Wrobel G, Vonderstrass B, et al. (2005) Preeclampsia: increased expression of soluble ADAM 12. J Mol Med (Berl) 83: 887–896.
- 22. Han JY, Kim YS, Cho GJ, Roh GS, Kim HJ, et al. (2006) Altered gene expression of caspase-10, death receptor-3 and IGFBP-3 in preeclamptic placentas. Mol Cells 22: 168–174.
- 23. Hansson SR, Chen Y, Brodszki J, Chen M, Hernandez-Andrade E, et al. (2006) Gene expression profiling of human placentas from preeclamptic and normotensive pregnancies. Mol Hum Reprod 12: 169–179.
- 24. Heikkila A, Tuomisto T, Hakkinen SK, Keski-Nisula L, Heinonen S, et al. (2005) Tumor suppressor and growth regulatory genes are overexpressed in severe early-onset preeclampsia–an array study on case-specific human preeclamptic placental tissue. Acta Obstet Gynecol Scand 84: 679–689.
- 25. Herse F, Dechend R, Harsem NK, Wallukat G, Janke J, et al. (2007) Dysregulation of the circulating and tissue-based renin-angiotensin system in preeclampsia. Hypertension 49: 604–611.
- 26. Hoegh AM, Borup R, Nielsen FC, Sorensen S, Hviid TV (2010) Gene expression profiling of placentas affected by pre-eclampsia. J Biomed Biotechnol 2010: 787545.
- 27. Kang JH, Song H, Yoon JA, Park DY, Kim SH, et al. (2011) Preeclampsia leads to dysregulation of various signaling pathways in placenta. J Hypertens 29: 928–936.
- 28. Lee GS, Joe YS, Kim SJ, Shin JC (2010) Cytokine-related genes and oxidation-related genes detected in preeclamptic placentas. Arch Gynecol Obstet 282: 363–369.
- 29. Liu Y, Li N, You L, Liu X, Li H, et al. (2008) HSP70 is associated with endothelial activation in placental vascular diseases. Mol Med 14: 561–566.
- 30. Mayor-Lynn K, Toloubeydokhti T, Cruz AC, Chegini N (2011) Expression profile of microRNAs and mRNAs in human placentas from pregnancies complicated by preeclampsia and preterm labor. Reprod Sci 18: 46–56.
- 31. Nishizawa H, Ota S, Suzuki M, Kato T, Sekiya T, et al. (2011) Comparative gene expression profiling of placentas from patients with severe pre-eclampsia and unexplained fetal growth restriction. Reprod Biol Endocrinol 9: 107.
- 32. Nishizawa H, Pryor-Koishi K, Kato T, Kowa H, Kurahashi H, et al. (2007) Microarray analysis of differentially expressed fetal genes in placental tissue derived from early and late onset severe pre-eclampsia. Placenta 28: 487–497.
- 33. Reimer T, Koczan D, Gerber B, Richter D, Thiesen HJ, et al. (2002) Microarray analysis of differentially expressed genes in placental tissue of pre-eclampsia: up-regulation of obesity-related genes. Mol Hum Reprod 8: 674–680.
- 34. Sitras V, Paulssen RH, Gronaas H, Leirvik J, Hanssen TA, et al. (2009) Differential placental gene expression in severe preeclampsia. Placenta 30: 424–433.
- 35. Soleymanlou N, Jurisica I, Nevo O, Ietta F, Zhang X, et al. (2005) Molecular evidence of placental hypoxia in preeclampsia. J Clin Endocrinol Metab 90: 4299–4308.
- 36. Tsai S, Hardison NE, James AH, Motsinger-Reif AA, Bischoff SR, et al. (2011) Transcriptional profiling of human placentas from pregnancies complicated by preeclampsia reveals disregulation of sialic acid acetylesterase and immune signalling pathways. Placenta 32: 175–182.
- 37. Tsoi SC, Cale JM, Bird IM, Kay HH (2003) cDNA microarray analysis of gene expression profiles in human placenta: up-regulation of the transcript encoding muscle subunit of glycogen phosphorylase in preeclampsia. J Soc Gynecol Investig 10: 496–502.
- 38. Vaiman D, Mondon F, Garces-Duran A, Mignot TM, Robert B, et al. (2005) Hypoxia-activated genes from early placenta are elevated in preeclampsia, but not in Intra-Uterine Growth Retardation. BMC Genomics 6: 111.
- 39. Varkonyi T, Nagy B, Fule T, Tarca AL, Karaszi K, et al. (2011) Microarray profiling reveals that placental transcriptomes of early-onset HELLP syndrome and preeclampsia are similar. Placenta 32 Suppl: S21–S29
- 40. Winn VD, Gormley M, Paquet AC, Kjaer-Sorensen K, Kramer A, et al. (2009) Severe preeclampsia-related changes in gene expression at the maternal-fetal interface include sialic acid-binding immunoglobulin-like lectin-6 and pappalysin-2. Endocrinology 150: 452–462.
- 41. Zhou R, Zhu Q, Wang Y, Ren Y, Zhang L, et al. (2006) Genomewide oligonucleotide microarray analysis on placentae of pre-eclamptic pregnancies. Gynecol Obstet Invest 62: 108–114.
- 42. Enquobahrie DA, Abetew DF, Sorensen TK, Willoughby D, Chidambaram K, et al. (2011) Placental microRNA expression in pregnancies complicated by preeclampsia. Am J Obstet Gynecol 204: 178.e12–21.
- 43. Hu Y, Li P, Hao S, Liu L, Zhao J, et al. (2009) Differential expression of microRNAs in the placentae of Chinese patients with severe pre-eclampsia. Clin Chem Lab Med 47: 923–929.
- 44. Zhu XM, Han T, Sargent IL, Yin GW, Yao YQ (2009) Differential expression profile of microRNAs in human placentas from preeclamptic pregnancies vs normal pregnancies. Am J Obstet Gynecol 200: 661–667.
- 45. Huang da W, Sherman BT, Lempicki RA (2009) Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 4: 44–57.
- 46. Maynard SE, Karumanchi SA (2011) Angiogenic factors and preeclampsia. Semin Nephrol 31: 33–46.
- 47. Foidart JM, Schaaps JP, Chantraine F, Munaut C, Lorquet S (2009) Dysregulation of anti-angiogenic agents (sFlt-1, PLGF, and sEndoglin) in preeclampsia–a step forward but not the definitive answer. J Reprod Immunol 82: 106–111.
- 48. Verlohren S, Stepan H, Dechend R (2012) Angiogenic growth factors in the diagnosis and prediction of pre-eclampsia. Clin Sci (Lond) 122: 43–52.
- 49. El-Gharib MN, Morad M (2011) Maternal serum inhibin-A for predicting preeclampsia. J Matern Fetal Neonatal Med 24: 595–599.
- 50. Fitzgerald B, Levytska K, Kingdom J, Walker M, Baczyk D, et al. (2011) Villous trophoblast abnormalities in extremely preterm deliveries with elevated second trimester maternal serum hCG or inhibin-A. Placenta 32: 339–345.
- 51. Guo J, Tian T, Lu D, Xia G, Wang H, et al. (2012) Alterations of maternal serum and placental follistatin-like 3 and myostatin in pre-eclampsia. J Obstet Gynaecol Res 38: 988–996.
- 52. Morris RK, Cnossen JS, Langejans M, Robson SC, Kleijnen J, et al. (2008) Serum screening with Down’s syndrome markers to predict pre-eclampsia and small for gestational age: systematic review and meta-analysis. BMC Pregnancy Childbirth 8: 33.
- 53. Olsen RN, Woelkers D, Dunsmoor-Su R, Lacoursiere DY (2012) Abnormal second-trimester serum analytes are more predictive of preterm preeclampsia. Am J Obstet Gynecol 207: 228–7.
- 54. Perkins AV, Linton EA, Eben F, Simpson J, Wolfe CD, et al. (1995) Corticotrophin-releasing hormone and corticotrophin-releasing hormone binding protein in normal and pre-eclamptic human pregnancies. Br J Obstet Gynaecol 102: 118–122.
- 55. Shen Z, Cai LY, Suprapto IS, Shenoy P, Zhou X (2011) Placental and maternal serum inhibin A in patients with preeclampsia and small-for-gestational-age. J Obstet Gynaecol Res 37: 1290–1296.
- 56. Rumer KK, Uyenishi J, Hoffman MC, Fisher BM, Winn VD (2013) Siglec-6 expression is increased in placentas from pregnancies complicated by preterm preeclampsia. Reprod Sci 20: 646–653.
- 57. Askelund KJ, Chamley LW (2011) Trophoblast deportation part I: review of the evidence demonstrating trophoblast shedding and deportation during human pregnancy. Placenta 32: 716–723.
- 58. Rajakumar A, Cerdeira AS, Rana S, Zsengeller Z, Edmunds L, et al. (2012) Transcriptionally active syncytial aggregates in the maternal circulation may contribute to circulating soluble fms-like tyrosine kinase 1 in preeclampsia. Hypertension 59: 256–264.
- 59. Founds SA, Conley YP, Lyons-Weiler JF, Jeyabalan A, Hogge WA, et al. (2009) Altered global gene expression in first trimester placentas of women destined to develop preeclampsia. Placenta 30: 15–24.
- 60. Farina A, Morano D, Arcelli D, De Sanctis P, Sekizawa A, et al. (2009) Gene expression in chorionic villous samples at 11 weeks of gestation in women who develop preeclampsia later in pregnancy: implications for screening. Prenat Diagn 29: 1038–1044.
- 61. Mikheev AM, Nabekura T, Kaddoumi A, Bammler TK, Govindarajan R, et al. (2008) Profiling gene expression in human placentae of different gestational ages: an OPRU Network and UW SCOR Study. Reprod Sci 15: 866–877.
- 62. Huntzinger E, Izaurralde E (2011) Gene silencing by microRNAs: contributions of translational repression and mRNA decay. Nat Rev Genet 12: 99–110.