- Email: [email protected]
Distinct expression profiles of lncRNAs between early-onset preeclampsia and preterm controls
Clinica Chimica Acta 463 (2016) 193–199
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Distinct expression profiles of lncRNAs between early-onset preeclampsia and preterm controls Wei Long a, Can Rui a, Xuejing Song a, Xiaonan Dai a, Xuan Xue a, Yuanqing Lu a, Rong Shen a, Jun Li b, Jingyun Li b,⁎, Hongjuan Ding a,⁎ a b
Department of Obstetrics, Obstetrics and Gynecology Hospital Affiliated to Nanjing Medical University, China Maternal and Child Health Medical Institute, Department of Plastic & Cosmetic Surgery, Obstetrics and Gynecology Hospital Affiliated to Nanjing Medical University, China
a r t i c l e
i n f o
Article history: Received 9 May 2016 Received in revised form 31 October 2016 Accepted 31 October 2016 Available online 02 November 2016 Keywords: Early-onset preeclampsia LncRNA Microarray Real-time PCR MMP9
a b s t r a c t Early-onset preeclampsia (EOPE), which is the most severe form of the syndrome, confers a high risk of neonatal morbidity and perinatal death. We aim to study the roles of long non-coding RNAs (lncRNAs) in the pathogenesis of early-onset preeclampsia (EOPE). Therefore, we examined the expression profiles of lncRNAs between earlyonset preeclampsia and preterm controls using microarray analysis. Quantitative real-time PCR (qRT-PCR) was performed to verify the selected differentially expressed lncRNAs. In total, we identified 15,646 upregulated and 13,178 downregulated lncRNAs in the placenta of EOPE patients compared to the preterm controls. Gene ontology and pathway analysis revealed that compared to the preterm controls, many of the processes overrepresented in the EOPE patients were related to cell migration and cell motility. A selection of the differentially expressed lncRNA transcripts was confirmed using qRT-PCR, particularly RP11-465L10.10, which is associated with the MMP9 gene. These data may offer a background/reference resource for future functional studies of lncRNAs related to EOPE. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Preeclampsia (PE) is a major health issue in pregnant women and their infants [1]. Recently, the relationship between the gestational week of onset and the severity of PE has attracted more attention [2,3]. Early-onset preeclampsia that develops prior to 34 weeks of gestation causes significant effects on maternal-foetal morbidity and mortality, particularly when it occurs early on in the pregnancy [4,5]. Following the delivery of the placenta, preeclampsia begins to resolve. However, neonates face significant problems secondary to the prematurity that results from the induced early delivery. The genetic basis for preeclampsia is complex. A number of biological molecules are differentially expressed in early- and late-onset PE. However, these molecules are protein-centric, and the main problems concerning the treatment, prediction and prevention of early-onset preeclampsia have not been elucidated [6]. Accumulating evidence suggests that poor placental development is particularly associated with earlyonset preeclampsia [7–9]. Therefore, it is necessary to further study the mechanism of placental development association with early-onset preeclampsia from the placenta. Long non-coding RNAs (lncRNAs), which are defined as non-coding RNAs N 200 nucleotides in length, were believed to have no functional ⁎ Corresponding authors at: Department of Obstetrics, Maternal and Child Health Medical Institute, Department of Plastic&Cosmetic Surgery, Obstetrics and Gynecology Hospital Affiliated to Nanjing Medical University, 123rd Tianfei Street, Mochou Road, Nanjing 210004, China. E-mail addresses: [email protected] (J. Li), [email protected] (H. Ding).
http://dx.doi.org/10.1016/j.cca.2016.10.036 0009-8981/© 2016 Elsevier B.V. All rights reserved.
significance when they were first discovered [10]. However, increasingly lncRNAs have been found to play critical roles in various disorders [11–14], including cardiovascular diseases, cancers and neurodegenerative diseases. Recent studies have reported that lncRNAs may be involved in the pathophysiological mechanisms of preeclampsia [15], and some lncRNAs such as H19, SPRY4-IT1 and MEG3 might have a biological function in regulating the behaviour of trophoblast cells [16–18]. However, the lncRNA profile of early-onset preeclampsia is still unclear. Increased understanding of the lncRNAs involved in early-onset preeclampsia could have a profound impact on our understanding of placental development and therapies for pregnancy complications, which may provide opportunities for interventions in early-onset preeclampsia. In this study, microarray experiments were performed to determine the expression of lncRNAs in the placenta of individuals with earlyonset preeclampsia. We further explored the expression of candidate lncRNAs between individuals with early-onset preeclampsia and controls. The results of this study may provide new insights into the pathophysiology of early-onset preeclampsia.
2. Methods and materials 2.1. Ethics statement This study was approved by the Medical Ethics Committee of the Obstetrics and Gynecology Hospital Affiliated to Nanjing Medical University (project No. NJFY-201253). Patients attending our hospital for a
194
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199
caesarean section read information about the purpose of the study, and written informed consent was obtained from each participant.
The quantity and quality of the RNA was evaluated using the Nanodrop and Agilent 2100 Bioanalyzer (Agilent Technologies, CA, USA).
2.2. Patient information and sample collection
2.4. Microarray analysis
Placental biopsies used for the microarray analysis and quantitative PCR were obtained during caesarean sections from preterm control patients (n = 32) and early-onset preeclampsia patients (n = 32). Preeclampsia was diagnosed following the criteria of the American College of Obstetricians and Gynecologists (2013), which was defined as new onset hypertension (systolic blood pressure ≥ 140 mm Hg or diastolic blood pressure ≥ 90 mm Hg on two occasions at least 4 h apart) and proteinuria (the excretion of 300 mg or more of protein in a 24-hour urine collection) after 20 weeks of gestation in women who previously had normal blood pressure. Early-onset preeclampsia was defined as preeclampsia that developed prior to 34 weeks of gestation. A total of 32 early onset preeclampsia samples were obtained from pregnant women (30–36 weeks) at the time of an elective caesarean section. The indications for the caesarean sections in the control group of pregnant women (30–36 weeks) were either the presence of breech and transverse presentation with premature rupture of the membranes or placenta previa with bleeding. We selected patients who underwent preterm delivery of the placenta as the control group to exclude the possible influence of the number of gestational weeks. The exclusion criteria were as follows: multiple gestations, maternal infections, smoking, chemical dependency, known major foetal or chromosomal anomalies, intrahepatic cholestasis during pregnancy, assisted reproductive technology (ART) treatments and any other confounding pathology (diabetes mellitus, renal disease, chronic hypertension, hyperthyroidism and hypothyroidism). The patient characteristics, such as the maternal reproductive data, deliveries and infant's outcomes, are presented in Table 1. The placentas were prepared within 10 min of delivery. The dissected tissue was immediately snap frozen in liquid nitrogen and stored at − 80 °C until RNA extraction.
Finally, 12 randomly and blindly selected samples (6 early-onset preeclampsia patients and 6 preterm controls) were hybridized and two biological replicates for each condition were used. The information for these patients is shown in Table 2. The expression profiles of the placental LncRNAs were detected using the Arraystar Human LncRNA Microarray v3.0, which was simultaneously used to detect the expression profiles of human genome-wide protein-coding transcripts (Glover et al. 2015). The Arraystar Human LncRNA Microarray v3.0 is designed for the global profiling of human lncRNAs (Kangcheng, Shanghai, China). Quantile normalization and subsequent data processing were performed using the GeneSpring GX v11.5.1 software package (Agilent Technologies, CA, USA). The statistically significant differentially expressed lncRNAs were identified using Volcano Plot filtering. The threshold value used to screen the differentially expressed lncRNAs was a fold change ≥2.0 or ≤−2.0 (P b 0.05). 2.5. Gene ontology (GO) and pathway analysis of the dysregulated lncRNAs A gene ontology analysis was designed to determine the functional trends that were associated with the lncRNAs differentially expressed between the early-onset PE and control groups. These differentially expressed lncRNAs exhibited a significance value of P b 0.05. The candidate genes were mapped to GO terms in the database (http://www. geneontology.org/), and then the number of genes was calculated for each term potential molecular functions term of these target genes in the pathways that were identified using the KEGG (Kyoto Encyclopedia of Genes and Genomes) database (http://www.genome.ad.jp/kegg/) and BioCarta (http://www.biocarta.com). The recommended P-value cut-off was b 0.05. 2.6. Validation of the microarray findings
2.3. Total RNA extraction from the placental tissue To the RNA extraction, the ground, frozen tissues were resuspended in TRIzol reagent (Invitrogen, Carlsbad, CA, USA). RNA purification was performed with the RNeasy minikit (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. After DNase digestion, the total RNA was then eluted using RNase-free water and treated with turbo DNase (Invitrogen, Carlsbad, CA, USA) to remove DNA contaminants.
Table 1 Clinical characteristics of EOPE and control groups. Clinical features
C (n = 32)
EOPE (n = 32)
P value
Maternal age (mean ± SD) Maternal BMI (mean ± SD) Week gestation (mean ± SD) Mode of delivery CS (%) Birth weight (g) (mean ± SD) BP systolic (mm Hg) BP diastolic (mm Hg) Proteinuria (g/24 h) FGR HELLP Apgar score 1 min Apgar score 5 min
29.2 ± 3.7 26.0 ± 2.2 33.8 ± 1.7 100% 2253.4 ± 395.8 N/A N/A N/A 0 0 9.9 ± 0.2 10 ± 0
30.7 ± 5.6 28.9 ± 3.8 33.1 ± 1.7 100% 1624.3 ± 290.1 167.7 ± 16.7 108.4 ± 12.6 6.0 ± 2.9 21 (65.6%) 2 (6.2%) 9.7 ± 0.8 9.9 ± 0.3
0.222 b0.001 0.123 b0.001
b0.001 0.151 0.211 0.321
C, preterm control; EOPE, early-onset preeclampsia; BMI, body mass index; BP, blood pressure; FGR, foetal growth restriction; HELLP, hemolysis, elevated liver enzymes, and low platelets syndrome; P value, early-onset preeclampsia compared with control group.
A total of 1 μg of RNA from each sample was reverse transcribed to cDNA using a random hexamer primer with the Thermo Scientific™ RevertAid First Strand cDNA Synthesis kit (Thermo Scientific, MA, USA). Primers for each lncRNA were designed according to Primer 3 and confirmed using the Basic Local Alignment Search Tool (BLAST) of NCBI to ensure a unique amplification product. Real-time PCR was performed on an Applied Biosystems ViiA™ 7 Dx (Life Technologies, MA, USA) using the SYBR green method according to the manufacturer's instructions. The PCR reaction conditions were as follows: a denaturation step at 95 °C for 10 min, followed by 40 PCR cycles at 95 °C for 15 s and 60 °C for 1 min. The relative gene expression levels were quantified based on the cycle threshold (Ct) values and normalized to the internal control housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH). The 2−ΔΔCt method was used to comparatively quantify the levels of mRNA. 2.7. Statistical analysis The samples were randomized to control for the pre-analytical variables, and an individual blinded to the experimental groups performed the sample preparation and microarray experiments. The statistical significance of the lncRNA datasets was performed using a one-way ANOVA corrected for multiple comparisons followed by the Student's t-test with a significance of P b 0.05. Data from the quantitative realtime PCR experiments were analysed with Student's unpaired or paired t-tests where appropriate. The values are presented as the mean ± standard deviation (SD).
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199
195
Table 2 Clinical characteristics of EOPE and controls for microarray.
C1 C2 C3 C4 C5 C6 P1 P2 P3 P4 P5 P6
Age
BMI (Kg/m2)
Week gestation
BP systolic (mm Hg)
BP diastolic (mm Hg)
Proteinuria (g/24 h)
Birth weight (g)
Apgar score 1 min
Apgar score 5 min
29 30 36 26 27 34 26 33 35 29 26 30
26.7 26.4 26.8 27 30.4 22 27.1 27.8 27 31.8 21.9 27.7
33.1 33 31.4 32.2 34.7 35.2 32 33 34.5 33.2 35.1 31.4
N/A N/A N/A N/A N/A N/A 170 180 170 180 180 205
N/A N/A N/A N/A N/A N/A 120 120 110 120 120 130
N/A N/A N/A N/A N/A N/A 6559 5243 7239 5088 6400 9214
2250 1950 1550 1900 2500 2400 1450 1540 1900 1750 1320 1350
10 10 10 10 10 10 8 6 10 9 10 10
10 10 10 10 10 10 10 8 10 10 10 10
C, preterm control; P, early-onset preeclampsia.
3. Results 3.1. General characteristics of the pattern of lncRNAs in early-onset PE The expression variation of lncRNAs between the early-onset preeclampsia (EOPE) placental tissues and preterm controls are shown in the scatterplot and hierarchical clustering in Fig. 1A and C, respectively.
In total, 15,646 lncRNAs were up regulated and 13,178 were down regulated in EOPE (fold change ≥2.0 or ≤−2.0, P b 0.05). Among the dysregulated lncRNAs, there were 12,195 intergenic, 5182 natural antisense, 4407 intron antisense, 3684 exon sense, 1545 bidirectional and 1352 intron sense-overlapping sequences (Fig. 1B). These lncRNAs were constructed using quality-controlled, public transcriptome databases such as RefSeq, Gencode, and UCSC Knowngenes (Fig. 1D). The lncRNAs
Fig. 1. Expression profiles of the lncRNAs between the EOPE placental tissues and gestational age-matched control samples. A: a scatterplot is a convenient method to quickly visualize the distributions of the lncRNAs profiles. The values of the x- and y-axis are the averaged normalized signal values of the groups of samples (log2 scaled). B: classification of the dysregulated lncRNAs. C: heat map presentation of the expression profiles of the lncRNAs. Hierarchical clustering to analyse the differential expression of the lncRNAs. Each column represents a sample and each row represents a gene. High relative expression is indicated by the yellow colour and low relative expression is indicated by the blue colour. D: the lncRNAs were collected from the most authoritative databases such as Genecode, RefSeq and UCSC Known Genes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
196
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199
Fig. 2. The lncRNAs are mainly between 200 bp and 3000 bp in length. The lncRNAs are distributed throughout the genome and cover all chromosomes (panels A and B).
were mainly between 200 and 3000 bp in length (Fig. 2A). Further analyses showed that these lncRNAs were distributed throughout the genome and covered all chromosomes (Fig. 2B).
3.2. GO and pathway analysis The Gene Ontology project provides a description of gene and gene product attributes in humans and other organisms (http://www. geneontology.org). It covers three domains: cellular component, biological process, and molecular function. In our existing data, the neighbour protein-coding gene function of the downregulated lncRNAs involved the following significant biological processes: (1) neutrophil chemotaxis; (2) response to chemical stimuli; (3) regulation of muscle system processes; (4) regulation of muscle contraction; (5) cell motility; and (6) cell migration (Fig. 3A). In contrast, the neighbour protein-coding gene function of the upregulated lncRNAs mainly involved (1) regulation of histone acetylation; (2) regulation of peptidyl-lysine acetylation; (3) regulation of chromosome organization; (4) progesterone metabolic processes; (5) substrate-dependent cell migration; and (6) cell motility (Fig. 3B). Through pathway analysis, we found that the neighbour gene function of the downregulated lncRNAs involved the following pathways: (1) complement and coagulation cascade; (2) glycine, serine and threonine metabolism; (3) focal adhesion; (4) influenza A; (5) vascular smooth muscle contraction; (6) ECM-receptor interaction; and (7) the
cytosolic DNA-sensing pathway (Fig. 3C). The neighbour gene function of the upregulated lncRNAs mainly involved the following pathways: (1) renal cell carcinoma; (2) chronic myeloid leukaemia; (3) the ErbB signalling pathway; (4) cell adhesion molecules; and (5) hepatitis B (Fig. 3D). Some of the signalling pathways were associated with both placental development and preeclampsia. 3.3. Candidate lncRNAs in early-onset preeclampsia This analysis showed that lncRNAs might play a key role in regulating cell motility and cell migration, which might control the process of placental development in early-onset preeclampsia. For practical purposes, to validate the microarray data and to investigate the lncRNAs potentially involved in early-onset PE, we first selected candidate lncRNAs that expressed fold-change ≥2. The selected lncRNAs and annotated protein-coding genes that were associated with placental development are listed in Table 3, according to the scientific literature and the GO/pathway analysis. 3.4. Validation of the candidate lncRNAs Eight differentially expressed lncRNAs from the microarray analysis data were selected for validation using real-time PCR (Fig. 4, Control n = 20, EOPE n = 20). The primers used in this study are shown in Table 4. The real-time PCR results and microarray data were consistent.
Fig. 3. GO analysis of the coding genes in which one coding gene is connected to at least three lncRNAs and one lncRNA is connected to at least three mRNAs. The top six significantly enriched molecular functions along with their scores for the downregulated and upregulated lncRNAs are listed as the x-axis and the y-axis, respectively, in both panels A and B. Pathway analysis is a functional analysis that maps genes to KEGG pathways (panels C and D). Cell motility, cell migration, cell adhesion, vascular smooth muscle contractions, and the ErbB signalling pathway were modulated during placental development and were associated with preeclampsia.
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199 Table 3 LncRNAs implicated in EOPE development.
LncRNA
Length Source
ENST00000418954 717 TAC3 958
Gencode RefSeq
ENST00000448857 ENST00000453528 ENST00000434181 ENST00000512786 ENST00000414797 uc003mvs.1 ENST00000429228 ENST00000560178 UC.294 NR_046268 ENST00000520055 ENST00000417135 ENST00000558334 NR_027683 NR_024284 ENST00000421891 NR_038380 ENST00000569037 ENST00000535913 ENST00000553732
Gencode Gencode Gencode Gencode Gencode UCSC_knowngene Gencode Gencode UCR RefSeq Gencode Gencode Gencode RefSeq RefSeq Gencode RefSeq Gencode Gencode Gencode
1072 1027 655 690 1072 2580 374 444 444 557 548 743 1168 1567 2232 831 4671 553 1955 580
197
Table 4 Primers used in this study. Associated gene Relationship TAC3 HSD17B3 CSMD2 SLC7A11 MAGI2 SLC22A23 MNX1 INO80 HIF1AN TGFB2 INTS9 ZNF710 C1QBP ZEB1 HDAC2 SLC7A11 MMP2 MMP9 TGFB3
Intergenic Exon sense-overlapping Natural antisense Intergenic Intronic antisense Natural antisense Bidirectional Natural antisense Intronic antisense Natural antisense Intergenic Natural antisense Intronic antisense Intergenic Natural antisense Natural antisense Natural antisense Intronic antisense Natural antisense Intronic antisense Natural antisense Natural antisense
Thus, our data provides a comprehensive profile and analysis of lncRNAs transcripts in the human placenta with early-onset PE. In addition, we found that the dysregulated lncRNAs may be associated with trophoblast invasion and may influence placental development. 3.5. Validation of the lncRNA RP11-465L10.10 and its potential target-gene MMP9 Previous research reported that MMP9 was a key regulator of trophoblast invasion and influenced vascular transformation and uteroplacental blood flow, which was specifically related with preeclampsia [19]. Using bioinformatics analysis methods, RP11465L10.10 was found to be located downstream of the natural antisense of the MMP9 gene in our microarray data (Fig. 5A). The downregulation of RP11-465L10.10 was validated using real-time PCR in additional samples (Fig. 5B, Control n = 20, EOPE n = 20). Furthermore, the expression of MMP9 in placenta from patients in the EOPE group was lower than in the control group (Fig. 5B, Control n = 20, EOPE n = 20). This trend was consistent with the variation tendency of RP11-465L10.10.
Gene name
Primer
Sequence (5′ to 3′)
RP11-90D4.2
Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse
CGGGCAGAAGCTATCAAAAG TCCTGAGGTGCTCTGGAACT ACACACCCTTTCCAGTTTGC CTTTGCCTTCCACCATGACT TGCTGCTATTCACAGCCATC TCTCGGGAGATGTTGATTCC GGTTCAAAAGGCAAGAGCAG AGTTCCTCACTGGGTGTTGG TTCTCTGCTTGGGCAGAACT TTGATGCCTTGGTTCCTTTC GCCCCACTCCACTCACTAAA ACCAACAAGAACCACGAAGG AGTGGATGGGCACAATTCTC CTTTTGCCCACTTCATGGTT GAGTCTGGGGATGTTGGAGA CCCAAACTTCAAGGGAGACA AACCTCCACCTCCCAGTTCAA GTCATCAAGTGCCGATAATGTGCC TGTACCGCTATGGTTACACTCG GGCAGGGACAGTTGCTTCT
AF127936.7 TAC3 AL121980.1 RP11-662B19.2 ATP5SL RP11-240L7.4 RP11-85L21.4 RP11-465L10.10 MMP9
4. Discussion Preeclampsia is one of the great obstetrical syndromes and is complicated by small-for-gestational-age (SGA) babies and preterm delivery [20]. However, the mechanisms of preeclampsia are unknown. Preeclampsia (PE) has been separated into early-onset and late-onset sub-phenotypes. There is evidence that the severity of the maternal and foetal compromise is mainly related to the gestational age at the onset of preeclampsia [5]. Therefore, how to prevent and cure early onset preeclampsia has always been a difficult clinical problem that urgently needs to be solved in obstetrics. Early-onset PE is associated with a significantly higher frequency of placental lesions and abnormal placental morphology than late-onset PE [8,21]. Studies initially revealed the molecular basis of abnormal trophoblastic invasion in early onset preeclampsia [22,23]. However, there is still a lack of effective therapeutic targets to prevent and predict earlyonset preeclampsia. The existing mechanistic research is far from explaining the pathogenesis of early onset preeclampsia. The rapid development of genetic information technologies has given researchers a new understanding of the complex gene regulatory network in early onset preeclampsia. Research on the disease has gradually expanded from the proteomics to epigenetics, such as methylation and miRNAs [24,25]. Long non-coding RNAs compose the largest proportion of
Fig. 4. The differentially expressed lncRNAs were validated using quantitative real-time PCR and normalized to GAPDH mRNA levels. Control n = 20, EOPE n = 20. EOPE: the early-onset preeclampsia group; control: normal placental tissue before 36 weeks of gestation. **P b 0.01.
198
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199
Fig. 5. The lncRNA RP11-465L10.10 was downregulated in early-onset preeclampsia placental tissues and was shown to regulate the MMP9 gene. Control n = 20, EOPE n = 20. **P b 0.01.
non-coding RNAs throughout the human genome and have been identified to be one of the most important factors controlling gene expression [26,27]. Recently, He et al. reported the expression profiles of human lncRNAs in the placenta of term patients with preeclampsia [15]. Here, we first selected early-onset PE placental tissues (gestational weeks 30–36) for lncRNA microarray analysis and identified several dysregulated lncRNAs. Our results are not entirely in agreement with a previous study that identified 738 differentially expressed lncRNAs in the human placentas from patients with early-onset preeclampsia compared to in preterm controls. This may be because the placental samples used in the previous study were all from term placenta (gestational weeks 37–40). Because early-onset PE is the most severe sub-group of PE and is associated with a greater prevalence of placental lesions [7,8, 21], it is usually difficult to maintain pregnancy to term. Therefore, the distinct expression profiles of lncRNAs between early-onset PE and gestational age-matched preterm controls may be more meaningful. Further, we verified the microarray profiles using real-time PCR. The real-time PCR results and microarray data were consistent. A total of 15,646 upregulated and 13,178 downregulated lncRNAs were identified in the placental tissues of EOPE patients compared with the placental tissues from gestational age-matched controls by setting a filter of log fold-change N2. GO and pathway analysis were applied to deeply understand the function of the target genes. In our data, the lncRNAs were mainly associated with biological processes involved in placental development and the initiation of EOPE, such as cell motility [28], cell migration [29], cell adhesion [30] and vascular smooth muscle contractions [3]. We next focused on the lncRNAs confirmed to be differentially expressed in the placenta, specifically the lncRNAs potentially involved in the initiation of early-onset preeclampsia. Several of these lncRNAs are predicted to target the expression of many genes, including matrix metalloproteinase (MMP) [31], hypoxia-inducible factor 1-alpha inhibitor (HIF1AN) [32], transforming growth factor beta-3 (TGFβ3) [33] and complement component 1 Q subcomponent-binding protein (C1QBP) [34]. Functionally, many of these genes serve as mediators of trophoblastic invasion, inflammation, angiogenesis, and oxidative stress, which are events considered to participate in both normal placental development and the pathology of pregnancy disorders. By analysing the associated coding genes of the candidate lncRNAs that we randomly chose to verify our microarray data, we identified one of the downregulated lncRNA RP11-465L10.10 that was of interest to us. RP11-465L10.10 was located near the MMP9 gene, which was previously reported to facilitate trophoblast invasion [19]. During the past ten years, the role of MMP9 in trophoblast invasion and placental development has been well documented [35]. A threshold level of MMP9 at the maternal–foetal interface is required for normal placentation. Severe PE and its early onset phenotype are strongly associated with an MMP9 variant, which causes shallow decidual EVT invasion
that leads to incomplete vascular transformation and reduced uteroplacental blood flow [19]. The co-expression and correlation between RP11-465L10.10 and MMP9 implies their possible regulatory function in sophisticated placental cellular activities. To our knowledge, the current data are the first to indicate that lncRNAs are closely associated with trophoblast invasion and may influence the severity and onset of preeclampsia. In this study, we reported differentially expressed lncRNA profile between EOPE placental tissues and their gestational age-matched preterm control placental tissues, and showed that the differentially expressed lncRNAs were associated with trophoblast invasion. Although direct evidence for the biological relevance of lncRNAs and their alterations in expression in placental cells during normal and complicated pregnancies is difficult to achieve, our data and those from the other studies cited above suggest that validating the true target genes of the differentially expressed lncRNAs may help to establish the regulatory functions of lncRNAs in placental cellular activities. Taken together, our results provide several directions for future research and provide new insights into the pathogenesis of early-onset severe preeclampsia. Disclosure of interests The authors declare that they have no conflicts of interest. Contribution to authorship In this paper, Wei Long, Can Rui and Xuejing Song performed the experiments. Xiaonan Dai and Xuan Xue organized the statistical analysis of the data. Jun Li, Yuanqing Lu and Rong Shen contributed the reagents and materials. Jingyun Li and Hongjuan Ding analysed the data. Details of the ethics approval This study was approved by the Ethics Committee of the Obstetrics and Gynecology Hospital Affiliated to Nanjing Medical University (project no. NJFY-201253). Acknowledgements This work was financially supported by grants from the National Natural Science Foundation of China (81501257, 81501672) and the Natural Science Foundation of Jiangsu Province (BK20140082, BK20140083). References [1] J.O. Lo, J.F. Mission, A.B. Caughey, Hypertensive disease of pregnancy and maternal mortality, Curr. Opin. Obstet. Gynecol. 25 (2) (2013) 124–132.
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199 [2] B.M. Sibai, J.R. Barton, Expectant management of severe preeclampsia remote from term: patient selection, treatment, and delivery indications, Am. J. Obstet. Gynecol. 196 (6) (2007), 514.e1-9. [3] S.J. Sheppard, R.A. Khalil, Risk factors and mediators of the vascular dysfunction associated with hypertension in pregnancy, Cardiovasc. Hematol. Disord. Drug Targets 10 (1) (2010) 33–52. [4] I.P. Gaugler-Senden, A.G. Huijssoon, W. Visser, E.A. Steegers, C.J. de Groot, Maternal and perinatal outcome of preeclampsia with an onset before 24 weeks' gestation. Audit in a tertiary referral center, Eur. J. Obstet. Gynecol. Reprod. Biol. 128 (1–2) (2006) 216–221. [5] Z.S. Khodzhaeva, Y.A. Kogan, R.G. Shmakov, N.I. Klimenchenko, A.S. Akatyeva, O.V. Vavina, A.M. Kholin, K.T. Muminova, G.T. Sukhikh, Clinical and pathogenetic features of early- and late-onset pre-eclampsia, J. Matern. Fetal Neonatal Med.: the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet 1-7 (2015). [6] J.M. Roberts, M.J. Bell, If we know so much about preeclampsia, why haven't we cured the disease? J. Reprod. Immunol. 99 (1–2) (2013) 1–9. [7] M. Egbor, T. Ansari, N. Morris, C.J. Green, P.D. Sibbons, Morphometric placental villous and vascular abnormalities in early- and late-onset pre-eclampsia with and without fetal growth restriction, BJOG: an international journal of obstetrics and gynaecology 113 (5) (2006) 580–589. [8] G. Ogge, T. Chaiworapongsa, R. Romero, Y. Hussein, J.P. Kusanovic, L. Yeo, C.J. Kim, S.S. Hassan, Placental lesions associated with maternal underperfusion are more frequent in early-onset than in late-onset preeclampsia, J. Perinat. Med. 39 (6) (2011) 641–652. [9] L.A. Magee, A. Pels, M. Helewa, E. Rey, P. von Dadelszen, S.H.G. Committee, Diagnosis, evaluation, and management of the hypertensive disorders of pregnancy: executive summary, J. Obstet. Gynaecol. Can.: JOGC = Journal d'obstetrique et gynecologie du Canada: JOGC 36 (7) (2014) 575–576. [10] A. Troy, N.E. Sharpless, Genetic "lnc"-age of noncoding RNAs to human disease, J. Clin. Invest. 122 (11) (2012) 3837–3840. [11] P.J. Batista, H.Y. Chang, Long noncoding RNAs: cellular address codes in development and disease, Cell 152 (6) (2013) 1298–1307. [12] J.T. Kung, D. Colognori, J.T. Lee, Long noncoding RNAs: past, present, and future, Genetics 193 (3) (2013) 651–669. [13] U.A. Orom, R. Shiekhattar, Long noncoding RNAs usher in a new era in the biology of enhancers, Cell 154 (6) (2013) 1190–1193. [14] P.J. McKiernan, K. Molloy, S.A. Cryan, N.G. McElvaney, C.M. Greene, Long noncoding RNA are aberrantly expressed in vivo in the cystic fibrosis bronchial epithelium, Int. J. Biochem. Cell Biol. 52 (2014) 184–191. [15] X. He, Y. He, B. Xi, J. Zheng, X. Zeng, Q. Cai, Y. Ouyang, C. Wang, X. Zhou, H. Huang, W. Deng, S. Xin, Q. Huang, H. Liu, LncRNAs expression in preeclampsia placenta reveals the potential role of LncRNAs contributing to preeclampsia pathogenesis, PLoS One 8 (11) (2013), e81437. [16] Y. Zou, Z. Jiang, X. Yu, M. Sun, Y. Zhang, Q. Zuo, J. Zhou, N. Yang, P. Han, Z. Ge, W. De, L. Sun, Upregulation of long noncoding RNA SPRY4-IT1 modulates proliferation, migration, apoptosis, and network formation in trophoblast cells HTR-8SV/neo, PLoS One 8 (11) (2013), e79598. [17] S. Buckberry, T. Bianco-Miotto, C.T. Roberts, Imprinted and X-linked non-coding RNAs as potential regulators of human placental function, Epigenetics 9 (1) (2014) 81–89. [18] Y. Zhang, Y. Zou, W. Wang, Q. Zuo, Z. Jiang, M. Sun, W. De, L. Sun, Down-regulated long non-coding RNA MEG3 and its effect on promoting apoptosis and suppressing migration of trophoblast cells, J. Cell. Biochem. 116 (4) (2015) 542–550. [19] V. Plaks, J. Rinkenberger, J. Dai, M. Flannery, M. Sund, K. Kanasaki, W. Ni, R. Kalluri, Z. Werb, Matrix metalloproteinase-9 deficiency phenocopies features of preeclampsia
199
and intrauterine growth restriction, Proc. Natl. Acad. Sci. U. S. A. 110 (27) (2013) 11109–11114. [20] J.M. Roberts, C.A. Hubel, The two stage model of preeclampsia: variations on the theme, Placenta 30 (Suppl A) (2009) S32–S37. [21] D. Raymond, E. Peterson, A critical review of early-onset and late-onset preeclampsia, Obstet. Gynecol. Surv. 66 (8) (2011) 497–506. [22] C. Qiao, C. Wang, J. Zhao, C. Liu, T. Shang, Elevated expression of KiSS-1 in placenta of Chinese women with early-onset preeclampsia, PLoS One 7 (11) (2012), e48937. [23] Y. Yang, J. Zhang, Y. Gong, X. Liu, Y. Bai, W. Xu, R. Zhou, Increased expression of prostasin contributes to early-onset severe preeclampsia through inhibiting trophoblast invasion, J. Perinatol.: official journal of the California Perinatal Association 35 (1) (2015) 16–22. [24] A. Kulkarni, P. Chavan-Gautam, S. Mehendale, H. Yadav, S. Joshi, Global DNA methylation patterns in placenta and its association with maternal hypertension in preeclampsia, DNA Cell Biol. 30 (2) (2011) 79–84. [25] M. van Dijk, C. Oudejans, (Epi)genetic control of human trophoblast invasion, Front. Genet. 5 (2014) 38. [26] M. Guttman, I. Amit, M. Garber, C. French, M.F. Lin, D. Feldser, M. Huarte, O. Zuk, B.W. Carey, J.P. Cassady, M.N. Cabili, R. Jaenisch, T.S. Mikkelsen, T. Jacks, N. Hacohen, B.E. Bernstein, M. Kellis, A. Regev, J.L. Rinn, E.S. Lander, Chromatin signature reveals over a thousand highly conserved large non-coding RNAs in mammals, Nature 458 (7235) (2009) 223–227. [27] M. Guttman, J. Donaghey, B.W. Carey, M. Garber, J.K. Grenier, G. Munson, G. Young, A.B. Lucas, R. Ach, L. Bruhn, X. Yang, I. Amit, A. Meissner, A. Regev, J.L. Rinn, D.E. Root, E.S. Lander, lincRNAs act in the circuitry controlling pluripotency and differentiation, Nature 477 (7364) (2011) 295–300. [28] S.K. Gupta, S.S. Malhotra, A. Malik, S. Verma, P. Chaudhary, Cell signaling pathways involved during invasion and syncytialization of trophoblast cells, Am. J. Reprod. Immunol. (2015). [29] K.R. Palmer, T.J. Kaitu'u-Lino, R. Hastie, N.J. Hannan, L. Ye, N. Binder, P. Cannon, L. Tuohey, T.G. Johns, A. Shub, S. Tong, Placental-specific sFLT-1 e15a protein is increased in preeclampsia, antagonizes vascular endothelial growth factor signaling, and has antiangiogenic activity, Hypertension 66 (6) (2015) 1251–1259. [30] L. Anton, A.G. Brown, M.S. Bartolomei, M.A. Elovitz, Differential methylation of genes associated with cell adhesion in preeclamptic placentas, PLoS One 9 (6) (2014), e100148. [31] W. Li, K.M. Mata, M.Q. Mazzuca, R.A. Khalil, Altered matrix metalloproteinase-2 and -9 expression/activity links placental ischemia and anti-angiogenic sFlt-1 to uteroplacental and vascular remodeling and collagen deposition in hypertensive pregnancy, Biochem. Pharmacol. 89 (3) (2014) 370–385. [32] K. Fukushima, M. Murata, M. Hachisuga, K. Tsukimori, H. Seki, S. Takeda, K. Asanoma, N. Wake, Hypoxia inducible factor 1 alpha regulates matrigel-induced endovascular differentiation under normoxia in a human extravillous trophoblast cell line, Placenta 29 (4) (2008) 324–331. [33] R. Tal, The role of hypoxia and hypoxia-inducible factor-1alpha in preeclampsia pathogenesis, Biol. Reprod. 87 (6) (2012) 134. [34] C. Agostinis, R. Bulla, C. Tripodo, A. Gismondi, H. Stabile, F. Bossi, C. Guarnotta, C. Garlanda, F. De Seta, P. Spessotto, A. Santoni, B. Ghebrehiwet, G. Girardi, F. Tedesco, An alternative role of C1q in cell migration and tissue remodeling: contribution to trophoblast invasion and placental development, J. Immunol. 185 (7) (2010) 4420–4429. [35] C.J. Lockwood, M. Basar, U.A. Kayisli, O. Guzeloglu-Kayisli, W. Murk, J. Wang, N. De Paz, J.P. Shapiro, R.J. Masch, N. Semerci, S.J. Huang, F. Schatz, Interferon-gamma protects first-trimester decidual cells against aberrant matrix metalloproteinases 1, 3, and 9 expression in preeclampsia, Am. J. Pathol. 184 (9) (2014) 2549–2559.
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Distinct expression profiles of lncRNAs between early-onset preeclampsia and preterm controls Wei Long a, Can Rui a, Xuejing Song a, Xiaonan Dai a, Xuan Xue a, Yuanqing Lu a, Rong Shen a, Jun Li b, Jingyun Li b,⁎, Hongjuan Ding a,⁎ a b
Department of Obstetrics, Obstetrics and Gynecology Hospital Affiliated to Nanjing Medical University, China Maternal and Child Health Medical Institute, Department of Plastic & Cosmetic Surgery, Obstetrics and Gynecology Hospital Affiliated to Nanjing Medical University, China
a r t i c l e
i n f o
Article history: Received 9 May 2016 Received in revised form 31 October 2016 Accepted 31 October 2016 Available online 02 November 2016 Keywords: Early-onset preeclampsia LncRNA Microarray Real-time PCR MMP9
a b s t r a c t Early-onset preeclampsia (EOPE), which is the most severe form of the syndrome, confers a high risk of neonatal morbidity and perinatal death. We aim to study the roles of long non-coding RNAs (lncRNAs) in the pathogenesis of early-onset preeclampsia (EOPE). Therefore, we examined the expression profiles of lncRNAs between earlyonset preeclampsia and preterm controls using microarray analysis. Quantitative real-time PCR (qRT-PCR) was performed to verify the selected differentially expressed lncRNAs. In total, we identified 15,646 upregulated and 13,178 downregulated lncRNAs in the placenta of EOPE patients compared to the preterm controls. Gene ontology and pathway analysis revealed that compared to the preterm controls, many of the processes overrepresented in the EOPE patients were related to cell migration and cell motility. A selection of the differentially expressed lncRNA transcripts was confirmed using qRT-PCR, particularly RP11-465L10.10, which is associated with the MMP9 gene. These data may offer a background/reference resource for future functional studies of lncRNAs related to EOPE. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Preeclampsia (PE) is a major health issue in pregnant women and their infants [1]. Recently, the relationship between the gestational week of onset and the severity of PE has attracted more attention [2,3]. Early-onset preeclampsia that develops prior to 34 weeks of gestation causes significant effects on maternal-foetal morbidity and mortality, particularly when it occurs early on in the pregnancy [4,5]. Following the delivery of the placenta, preeclampsia begins to resolve. However, neonates face significant problems secondary to the prematurity that results from the induced early delivery. The genetic basis for preeclampsia is complex. A number of biological molecules are differentially expressed in early- and late-onset PE. However, these molecules are protein-centric, and the main problems concerning the treatment, prediction and prevention of early-onset preeclampsia have not been elucidated [6]. Accumulating evidence suggests that poor placental development is particularly associated with earlyonset preeclampsia [7–9]. Therefore, it is necessary to further study the mechanism of placental development association with early-onset preeclampsia from the placenta. Long non-coding RNAs (lncRNAs), which are defined as non-coding RNAs N 200 nucleotides in length, were believed to have no functional ⁎ Corresponding authors at: Department of Obstetrics, Maternal and Child Health Medical Institute, Department of Plastic&Cosmetic Surgery, Obstetrics and Gynecology Hospital Affiliated to Nanjing Medical University, 123rd Tianfei Street, Mochou Road, Nanjing 210004, China. E-mail addresses: [email protected] (J. Li), [email protected] (H. Ding).
http://dx.doi.org/10.1016/j.cca.2016.10.036 0009-8981/© 2016 Elsevier B.V. All rights reserved.
significance when they were first discovered [10]. However, increasingly lncRNAs have been found to play critical roles in various disorders [11–14], including cardiovascular diseases, cancers and neurodegenerative diseases. Recent studies have reported that lncRNAs may be involved in the pathophysiological mechanisms of preeclampsia [15], and some lncRNAs such as H19, SPRY4-IT1 and MEG3 might have a biological function in regulating the behaviour of trophoblast cells [16–18]. However, the lncRNA profile of early-onset preeclampsia is still unclear. Increased understanding of the lncRNAs involved in early-onset preeclampsia could have a profound impact on our understanding of placental development and therapies for pregnancy complications, which may provide opportunities for interventions in early-onset preeclampsia. In this study, microarray experiments were performed to determine the expression of lncRNAs in the placenta of individuals with earlyonset preeclampsia. We further explored the expression of candidate lncRNAs between individuals with early-onset preeclampsia and controls. The results of this study may provide new insights into the pathophysiology of early-onset preeclampsia.
2. Methods and materials 2.1. Ethics statement This study was approved by the Medical Ethics Committee of the Obstetrics and Gynecology Hospital Affiliated to Nanjing Medical University (project No. NJFY-201253). Patients attending our hospital for a
194
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199
caesarean section read information about the purpose of the study, and written informed consent was obtained from each participant.
The quantity and quality of the RNA was evaluated using the Nanodrop and Agilent 2100 Bioanalyzer (Agilent Technologies, CA, USA).
2.2. Patient information and sample collection
2.4. Microarray analysis
Placental biopsies used for the microarray analysis and quantitative PCR were obtained during caesarean sections from preterm control patients (n = 32) and early-onset preeclampsia patients (n = 32). Preeclampsia was diagnosed following the criteria of the American College of Obstetricians and Gynecologists (2013), which was defined as new onset hypertension (systolic blood pressure ≥ 140 mm Hg or diastolic blood pressure ≥ 90 mm Hg on two occasions at least 4 h apart) and proteinuria (the excretion of 300 mg or more of protein in a 24-hour urine collection) after 20 weeks of gestation in women who previously had normal blood pressure. Early-onset preeclampsia was defined as preeclampsia that developed prior to 34 weeks of gestation. A total of 32 early onset preeclampsia samples were obtained from pregnant women (30–36 weeks) at the time of an elective caesarean section. The indications for the caesarean sections in the control group of pregnant women (30–36 weeks) were either the presence of breech and transverse presentation with premature rupture of the membranes or placenta previa with bleeding. We selected patients who underwent preterm delivery of the placenta as the control group to exclude the possible influence of the number of gestational weeks. The exclusion criteria were as follows: multiple gestations, maternal infections, smoking, chemical dependency, known major foetal or chromosomal anomalies, intrahepatic cholestasis during pregnancy, assisted reproductive technology (ART) treatments and any other confounding pathology (diabetes mellitus, renal disease, chronic hypertension, hyperthyroidism and hypothyroidism). The patient characteristics, such as the maternal reproductive data, deliveries and infant's outcomes, are presented in Table 1. The placentas were prepared within 10 min of delivery. The dissected tissue was immediately snap frozen in liquid nitrogen and stored at − 80 °C until RNA extraction.
Finally, 12 randomly and blindly selected samples (6 early-onset preeclampsia patients and 6 preterm controls) were hybridized and two biological replicates for each condition were used. The information for these patients is shown in Table 2. The expression profiles of the placental LncRNAs were detected using the Arraystar Human LncRNA Microarray v3.0, which was simultaneously used to detect the expression profiles of human genome-wide protein-coding transcripts (Glover et al. 2015). The Arraystar Human LncRNA Microarray v3.0 is designed for the global profiling of human lncRNAs (Kangcheng, Shanghai, China). Quantile normalization and subsequent data processing were performed using the GeneSpring GX v11.5.1 software package (Agilent Technologies, CA, USA). The statistically significant differentially expressed lncRNAs were identified using Volcano Plot filtering. The threshold value used to screen the differentially expressed lncRNAs was a fold change ≥2.0 or ≤−2.0 (P b 0.05). 2.5. Gene ontology (GO) and pathway analysis of the dysregulated lncRNAs A gene ontology analysis was designed to determine the functional trends that were associated with the lncRNAs differentially expressed between the early-onset PE and control groups. These differentially expressed lncRNAs exhibited a significance value of P b 0.05. The candidate genes were mapped to GO terms in the database (http://www. geneontology.org/), and then the number of genes was calculated for each term potential molecular functions term of these target genes in the pathways that were identified using the KEGG (Kyoto Encyclopedia of Genes and Genomes) database (http://www.genome.ad.jp/kegg/) and BioCarta (http://www.biocarta.com). The recommended P-value cut-off was b 0.05. 2.6. Validation of the microarray findings
2.3. Total RNA extraction from the placental tissue To the RNA extraction, the ground, frozen tissues were resuspended in TRIzol reagent (Invitrogen, Carlsbad, CA, USA). RNA purification was performed with the RNeasy minikit (Qiagen, Valencia, CA, USA) according to the manufacturer's instructions. After DNase digestion, the total RNA was then eluted using RNase-free water and treated with turbo DNase (Invitrogen, Carlsbad, CA, USA) to remove DNA contaminants.
Table 1 Clinical characteristics of EOPE and control groups. Clinical features
C (n = 32)
EOPE (n = 32)
P value
Maternal age (mean ± SD) Maternal BMI (mean ± SD) Week gestation (mean ± SD) Mode of delivery CS (%) Birth weight (g) (mean ± SD) BP systolic (mm Hg) BP diastolic (mm Hg) Proteinuria (g/24 h) FGR HELLP Apgar score 1 min Apgar score 5 min
29.2 ± 3.7 26.0 ± 2.2 33.8 ± 1.7 100% 2253.4 ± 395.8 N/A N/A N/A 0 0 9.9 ± 0.2 10 ± 0
30.7 ± 5.6 28.9 ± 3.8 33.1 ± 1.7 100% 1624.3 ± 290.1 167.7 ± 16.7 108.4 ± 12.6 6.0 ± 2.9 21 (65.6%) 2 (6.2%) 9.7 ± 0.8 9.9 ± 0.3
0.222 b0.001 0.123 b0.001
b0.001 0.151 0.211 0.321
C, preterm control; EOPE, early-onset preeclampsia; BMI, body mass index; BP, blood pressure; FGR, foetal growth restriction; HELLP, hemolysis, elevated liver enzymes, and low platelets syndrome; P value, early-onset preeclampsia compared with control group.
A total of 1 μg of RNA from each sample was reverse transcribed to cDNA using a random hexamer primer with the Thermo Scientific™ RevertAid First Strand cDNA Synthesis kit (Thermo Scientific, MA, USA). Primers for each lncRNA were designed according to Primer 3 and confirmed using the Basic Local Alignment Search Tool (BLAST) of NCBI to ensure a unique amplification product. Real-time PCR was performed on an Applied Biosystems ViiA™ 7 Dx (Life Technologies, MA, USA) using the SYBR green method according to the manufacturer's instructions. The PCR reaction conditions were as follows: a denaturation step at 95 °C for 10 min, followed by 40 PCR cycles at 95 °C for 15 s and 60 °C for 1 min. The relative gene expression levels were quantified based on the cycle threshold (Ct) values and normalized to the internal control housekeeping gene glyceraldehyde 3-phosphate dehydrogenase (GAPDH). The 2−ΔΔCt method was used to comparatively quantify the levels of mRNA. 2.7. Statistical analysis The samples were randomized to control for the pre-analytical variables, and an individual blinded to the experimental groups performed the sample preparation and microarray experiments. The statistical significance of the lncRNA datasets was performed using a one-way ANOVA corrected for multiple comparisons followed by the Student's t-test with a significance of P b 0.05. Data from the quantitative realtime PCR experiments were analysed with Student's unpaired or paired t-tests where appropriate. The values are presented as the mean ± standard deviation (SD).
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199
195
Table 2 Clinical characteristics of EOPE and controls for microarray.
C1 C2 C3 C4 C5 C6 P1 P2 P3 P4 P5 P6
Age
BMI (Kg/m2)
Week gestation
BP systolic (mm Hg)
BP diastolic (mm Hg)
Proteinuria (g/24 h)
Birth weight (g)
Apgar score 1 min
Apgar score 5 min
29 30 36 26 27 34 26 33 35 29 26 30
26.7 26.4 26.8 27 30.4 22 27.1 27.8 27 31.8 21.9 27.7
33.1 33 31.4 32.2 34.7 35.2 32 33 34.5 33.2 35.1 31.4
N/A N/A N/A N/A N/A N/A 170 180 170 180 180 205
N/A N/A N/A N/A N/A N/A 120 120 110 120 120 130
N/A N/A N/A N/A N/A N/A 6559 5243 7239 5088 6400 9214
2250 1950 1550 1900 2500 2400 1450 1540 1900 1750 1320 1350
10 10 10 10 10 10 8 6 10 9 10 10
10 10 10 10 10 10 10 8 10 10 10 10
C, preterm control; P, early-onset preeclampsia.
3. Results 3.1. General characteristics of the pattern of lncRNAs in early-onset PE The expression variation of lncRNAs between the early-onset preeclampsia (EOPE) placental tissues and preterm controls are shown in the scatterplot and hierarchical clustering in Fig. 1A and C, respectively.
In total, 15,646 lncRNAs were up regulated and 13,178 were down regulated in EOPE (fold change ≥2.0 or ≤−2.0, P b 0.05). Among the dysregulated lncRNAs, there were 12,195 intergenic, 5182 natural antisense, 4407 intron antisense, 3684 exon sense, 1545 bidirectional and 1352 intron sense-overlapping sequences (Fig. 1B). These lncRNAs were constructed using quality-controlled, public transcriptome databases such as RefSeq, Gencode, and UCSC Knowngenes (Fig. 1D). The lncRNAs
Fig. 1. Expression profiles of the lncRNAs between the EOPE placental tissues and gestational age-matched control samples. A: a scatterplot is a convenient method to quickly visualize the distributions of the lncRNAs profiles. The values of the x- and y-axis are the averaged normalized signal values of the groups of samples (log2 scaled). B: classification of the dysregulated lncRNAs. C: heat map presentation of the expression profiles of the lncRNAs. Hierarchical clustering to analyse the differential expression of the lncRNAs. Each column represents a sample and each row represents a gene. High relative expression is indicated by the yellow colour and low relative expression is indicated by the blue colour. D: the lncRNAs were collected from the most authoritative databases such as Genecode, RefSeq and UCSC Known Genes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
196
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199
Fig. 2. The lncRNAs are mainly between 200 bp and 3000 bp in length. The lncRNAs are distributed throughout the genome and cover all chromosomes (panels A and B).
were mainly between 200 and 3000 bp in length (Fig. 2A). Further analyses showed that these lncRNAs were distributed throughout the genome and covered all chromosomes (Fig. 2B).
3.2. GO and pathway analysis The Gene Ontology project provides a description of gene and gene product attributes in humans and other organisms (http://www. geneontology.org). It covers three domains: cellular component, biological process, and molecular function. In our existing data, the neighbour protein-coding gene function of the downregulated lncRNAs involved the following significant biological processes: (1) neutrophil chemotaxis; (2) response to chemical stimuli; (3) regulation of muscle system processes; (4) regulation of muscle contraction; (5) cell motility; and (6) cell migration (Fig. 3A). In contrast, the neighbour protein-coding gene function of the upregulated lncRNAs mainly involved (1) regulation of histone acetylation; (2) regulation of peptidyl-lysine acetylation; (3) regulation of chromosome organization; (4) progesterone metabolic processes; (5) substrate-dependent cell migration; and (6) cell motility (Fig. 3B). Through pathway analysis, we found that the neighbour gene function of the downregulated lncRNAs involved the following pathways: (1) complement and coagulation cascade; (2) glycine, serine and threonine metabolism; (3) focal adhesion; (4) influenza A; (5) vascular smooth muscle contraction; (6) ECM-receptor interaction; and (7) the
cytosolic DNA-sensing pathway (Fig. 3C). The neighbour gene function of the upregulated lncRNAs mainly involved the following pathways: (1) renal cell carcinoma; (2) chronic myeloid leukaemia; (3) the ErbB signalling pathway; (4) cell adhesion molecules; and (5) hepatitis B (Fig. 3D). Some of the signalling pathways were associated with both placental development and preeclampsia. 3.3. Candidate lncRNAs in early-onset preeclampsia This analysis showed that lncRNAs might play a key role in regulating cell motility and cell migration, which might control the process of placental development in early-onset preeclampsia. For practical purposes, to validate the microarray data and to investigate the lncRNAs potentially involved in early-onset PE, we first selected candidate lncRNAs that expressed fold-change ≥2. The selected lncRNAs and annotated protein-coding genes that were associated with placental development are listed in Table 3, according to the scientific literature and the GO/pathway analysis. 3.4. Validation of the candidate lncRNAs Eight differentially expressed lncRNAs from the microarray analysis data were selected for validation using real-time PCR (Fig. 4, Control n = 20, EOPE n = 20). The primers used in this study are shown in Table 4. The real-time PCR results and microarray data were consistent.
Fig. 3. GO analysis of the coding genes in which one coding gene is connected to at least three lncRNAs and one lncRNA is connected to at least three mRNAs. The top six significantly enriched molecular functions along with their scores for the downregulated and upregulated lncRNAs are listed as the x-axis and the y-axis, respectively, in both panels A and B. Pathway analysis is a functional analysis that maps genes to KEGG pathways (panels C and D). Cell motility, cell migration, cell adhesion, vascular smooth muscle contractions, and the ErbB signalling pathway were modulated during placental development and were associated with preeclampsia.
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199 Table 3 LncRNAs implicated in EOPE development.
LncRNA
Length Source
ENST00000418954 717 TAC3 958
Gencode RefSeq
ENST00000448857 ENST00000453528 ENST00000434181 ENST00000512786 ENST00000414797 uc003mvs.1 ENST00000429228 ENST00000560178 UC.294 NR_046268 ENST00000520055 ENST00000417135 ENST00000558334 NR_027683 NR_024284 ENST00000421891 NR_038380 ENST00000569037 ENST00000535913 ENST00000553732
Gencode Gencode Gencode Gencode Gencode UCSC_knowngene Gencode Gencode UCR RefSeq Gencode Gencode Gencode RefSeq RefSeq Gencode RefSeq Gencode Gencode Gencode
1072 1027 655 690 1072 2580 374 444 444 557 548 743 1168 1567 2232 831 4671 553 1955 580
197
Table 4 Primers used in this study. Associated gene Relationship TAC3 HSD17B3 CSMD2 SLC7A11 MAGI2 SLC22A23 MNX1 INO80 HIF1AN TGFB2 INTS9 ZNF710 C1QBP ZEB1 HDAC2 SLC7A11 MMP2 MMP9 TGFB3
Intergenic Exon sense-overlapping Natural antisense Intergenic Intronic antisense Natural antisense Bidirectional Natural antisense Intronic antisense Natural antisense Intergenic Natural antisense Intronic antisense Intergenic Natural antisense Natural antisense Natural antisense Intronic antisense Natural antisense Intronic antisense Natural antisense Natural antisense
Thus, our data provides a comprehensive profile and analysis of lncRNAs transcripts in the human placenta with early-onset PE. In addition, we found that the dysregulated lncRNAs may be associated with trophoblast invasion and may influence placental development. 3.5. Validation of the lncRNA RP11-465L10.10 and its potential target-gene MMP9 Previous research reported that MMP9 was a key regulator of trophoblast invasion and influenced vascular transformation and uteroplacental blood flow, which was specifically related with preeclampsia [19]. Using bioinformatics analysis methods, RP11465L10.10 was found to be located downstream of the natural antisense of the MMP9 gene in our microarray data (Fig. 5A). The downregulation of RP11-465L10.10 was validated using real-time PCR in additional samples (Fig. 5B, Control n = 20, EOPE n = 20). Furthermore, the expression of MMP9 in placenta from patients in the EOPE group was lower than in the control group (Fig. 5B, Control n = 20, EOPE n = 20). This trend was consistent with the variation tendency of RP11-465L10.10.
Gene name
Primer
Sequence (5′ to 3′)
RP11-90D4.2
Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse
CGGGCAGAAGCTATCAAAAG TCCTGAGGTGCTCTGGAACT ACACACCCTTTCCAGTTTGC CTTTGCCTTCCACCATGACT TGCTGCTATTCACAGCCATC TCTCGGGAGATGTTGATTCC GGTTCAAAAGGCAAGAGCAG AGTTCCTCACTGGGTGTTGG TTCTCTGCTTGGGCAGAACT TTGATGCCTTGGTTCCTTTC GCCCCACTCCACTCACTAAA ACCAACAAGAACCACGAAGG AGTGGATGGGCACAATTCTC CTTTTGCCCACTTCATGGTT GAGTCTGGGGATGTTGGAGA CCCAAACTTCAAGGGAGACA AACCTCCACCTCCCAGTTCAA GTCATCAAGTGCCGATAATGTGCC TGTACCGCTATGGTTACACTCG GGCAGGGACAGTTGCTTCT
AF127936.7 TAC3 AL121980.1 RP11-662B19.2 ATP5SL RP11-240L7.4 RP11-85L21.4 RP11-465L10.10 MMP9
4. Discussion Preeclampsia is one of the great obstetrical syndromes and is complicated by small-for-gestational-age (SGA) babies and preterm delivery [20]. However, the mechanisms of preeclampsia are unknown. Preeclampsia (PE) has been separated into early-onset and late-onset sub-phenotypes. There is evidence that the severity of the maternal and foetal compromise is mainly related to the gestational age at the onset of preeclampsia [5]. Therefore, how to prevent and cure early onset preeclampsia has always been a difficult clinical problem that urgently needs to be solved in obstetrics. Early-onset PE is associated with a significantly higher frequency of placental lesions and abnormal placental morphology than late-onset PE [8,21]. Studies initially revealed the molecular basis of abnormal trophoblastic invasion in early onset preeclampsia [22,23]. However, there is still a lack of effective therapeutic targets to prevent and predict earlyonset preeclampsia. The existing mechanistic research is far from explaining the pathogenesis of early onset preeclampsia. The rapid development of genetic information technologies has given researchers a new understanding of the complex gene regulatory network in early onset preeclampsia. Research on the disease has gradually expanded from the proteomics to epigenetics, such as methylation and miRNAs [24,25]. Long non-coding RNAs compose the largest proportion of
Fig. 4. The differentially expressed lncRNAs were validated using quantitative real-time PCR and normalized to GAPDH mRNA levels. Control n = 20, EOPE n = 20. EOPE: the early-onset preeclampsia group; control: normal placental tissue before 36 weeks of gestation. **P b 0.01.
198
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199
Fig. 5. The lncRNA RP11-465L10.10 was downregulated in early-onset preeclampsia placental tissues and was shown to regulate the MMP9 gene. Control n = 20, EOPE n = 20. **P b 0.01.
non-coding RNAs throughout the human genome and have been identified to be one of the most important factors controlling gene expression [26,27]. Recently, He et al. reported the expression profiles of human lncRNAs in the placenta of term patients with preeclampsia [15]. Here, we first selected early-onset PE placental tissues (gestational weeks 30–36) for lncRNA microarray analysis and identified several dysregulated lncRNAs. Our results are not entirely in agreement with a previous study that identified 738 differentially expressed lncRNAs in the human placentas from patients with early-onset preeclampsia compared to in preterm controls. This may be because the placental samples used in the previous study were all from term placenta (gestational weeks 37–40). Because early-onset PE is the most severe sub-group of PE and is associated with a greater prevalence of placental lesions [7,8, 21], it is usually difficult to maintain pregnancy to term. Therefore, the distinct expression profiles of lncRNAs between early-onset PE and gestational age-matched preterm controls may be more meaningful. Further, we verified the microarray profiles using real-time PCR. The real-time PCR results and microarray data were consistent. A total of 15,646 upregulated and 13,178 downregulated lncRNAs were identified in the placental tissues of EOPE patients compared with the placental tissues from gestational age-matched controls by setting a filter of log fold-change N2. GO and pathway analysis were applied to deeply understand the function of the target genes. In our data, the lncRNAs were mainly associated with biological processes involved in placental development and the initiation of EOPE, such as cell motility [28], cell migration [29], cell adhesion [30] and vascular smooth muscle contractions [3]. We next focused on the lncRNAs confirmed to be differentially expressed in the placenta, specifically the lncRNAs potentially involved in the initiation of early-onset preeclampsia. Several of these lncRNAs are predicted to target the expression of many genes, including matrix metalloproteinase (MMP) [31], hypoxia-inducible factor 1-alpha inhibitor (HIF1AN) [32], transforming growth factor beta-3 (TGFβ3) [33] and complement component 1 Q subcomponent-binding protein (C1QBP) [34]. Functionally, many of these genes serve as mediators of trophoblastic invasion, inflammation, angiogenesis, and oxidative stress, which are events considered to participate in both normal placental development and the pathology of pregnancy disorders. By analysing the associated coding genes of the candidate lncRNAs that we randomly chose to verify our microarray data, we identified one of the downregulated lncRNA RP11-465L10.10 that was of interest to us. RP11-465L10.10 was located near the MMP9 gene, which was previously reported to facilitate trophoblast invasion [19]. During the past ten years, the role of MMP9 in trophoblast invasion and placental development has been well documented [35]. A threshold level of MMP9 at the maternal–foetal interface is required for normal placentation. Severe PE and its early onset phenotype are strongly associated with an MMP9 variant, which causes shallow decidual EVT invasion
that leads to incomplete vascular transformation and reduced uteroplacental blood flow [19]. The co-expression and correlation between RP11-465L10.10 and MMP9 implies their possible regulatory function in sophisticated placental cellular activities. To our knowledge, the current data are the first to indicate that lncRNAs are closely associated with trophoblast invasion and may influence the severity and onset of preeclampsia. In this study, we reported differentially expressed lncRNA profile between EOPE placental tissues and their gestational age-matched preterm control placental tissues, and showed that the differentially expressed lncRNAs were associated with trophoblast invasion. Although direct evidence for the biological relevance of lncRNAs and their alterations in expression in placental cells during normal and complicated pregnancies is difficult to achieve, our data and those from the other studies cited above suggest that validating the true target genes of the differentially expressed lncRNAs may help to establish the regulatory functions of lncRNAs in placental cellular activities. Taken together, our results provide several directions for future research and provide new insights into the pathogenesis of early-onset severe preeclampsia. Disclosure of interests The authors declare that they have no conflicts of interest. Contribution to authorship In this paper, Wei Long, Can Rui and Xuejing Song performed the experiments. Xiaonan Dai and Xuan Xue organized the statistical analysis of the data. Jun Li, Yuanqing Lu and Rong Shen contributed the reagents and materials. Jingyun Li and Hongjuan Ding analysed the data. Details of the ethics approval This study was approved by the Ethics Committee of the Obstetrics and Gynecology Hospital Affiliated to Nanjing Medical University (project no. NJFY-201253). Acknowledgements This work was financially supported by grants from the National Natural Science Foundation of China (81501257, 81501672) and the Natural Science Foundation of Jiangsu Province (BK20140082, BK20140083). References [1] J.O. Lo, J.F. Mission, A.B. Caughey, Hypertensive disease of pregnancy and maternal mortality, Curr. Opin. Obstet. Gynecol. 25 (2) (2013) 124–132.
W. Long et al. / Clinica Chimica Acta 463 (2016) 193–199 [2] B.M. Sibai, J.R. Barton, Expectant management of severe preeclampsia remote from term: patient selection, treatment, and delivery indications, Am. J. Obstet. Gynecol. 196 (6) (2007), 514.e1-9. [3] S.J. Sheppard, R.A. Khalil, Risk factors and mediators of the vascular dysfunction associated with hypertension in pregnancy, Cardiovasc. Hematol. Disord. Drug Targets 10 (1) (2010) 33–52. [4] I.P. Gaugler-Senden, A.G. Huijssoon, W. Visser, E.A. Steegers, C.J. de Groot, Maternal and perinatal outcome of preeclampsia with an onset before 24 weeks' gestation. Audit in a tertiary referral center, Eur. J. Obstet. Gynecol. Reprod. Biol. 128 (1–2) (2006) 216–221. [5] Z.S. Khodzhaeva, Y.A. Kogan, R.G. Shmakov, N.I. Klimenchenko, A.S. Akatyeva, O.V. Vavina, A.M. Kholin, K.T. Muminova, G.T. Sukhikh, Clinical and pathogenetic features of early- and late-onset pre-eclampsia, J. Matern. Fetal Neonatal Med.: the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet 1-7 (2015). [6] J.M. Roberts, M.J. Bell, If we know so much about preeclampsia, why haven't we cured the disease? J. Reprod. Immunol. 99 (1–2) (2013) 1–9. [7] M. Egbor, T. Ansari, N. Morris, C.J. Green, P.D. Sibbons, Morphometric placental villous and vascular abnormalities in early- and late-onset pre-eclampsia with and without fetal growth restriction, BJOG: an international journal of obstetrics and gynaecology 113 (5) (2006) 580–589. [8] G. Ogge, T. Chaiworapongsa, R. Romero, Y. Hussein, J.P. Kusanovic, L. Yeo, C.J. Kim, S.S. Hassan, Placental lesions associated with maternal underperfusion are more frequent in early-onset than in late-onset preeclampsia, J. Perinat. Med. 39 (6) (2011) 641–652. [9] L.A. Magee, A. Pels, M. Helewa, E. Rey, P. von Dadelszen, S.H.G. Committee, Diagnosis, evaluation, and management of the hypertensive disorders of pregnancy: executive summary, J. Obstet. Gynaecol. Can.: JOGC = Journal d'obstetrique et gynecologie du Canada: JOGC 36 (7) (2014) 575–576. [10] A. Troy, N.E. Sharpless, Genetic "lnc"-age of noncoding RNAs to human disease, J. Clin. Invest. 122 (11) (2012) 3837–3840. [11] P.J. Batista, H.Y. Chang, Long noncoding RNAs: cellular address codes in development and disease, Cell 152 (6) (2013) 1298–1307. [12] J.T. Kung, D. Colognori, J.T. Lee, Long noncoding RNAs: past, present, and future, Genetics 193 (3) (2013) 651–669. [13] U.A. Orom, R. Shiekhattar, Long noncoding RNAs usher in a new era in the biology of enhancers, Cell 154 (6) (2013) 1190–1193. [14] P.J. McKiernan, K. Molloy, S.A. Cryan, N.G. McElvaney, C.M. Greene, Long noncoding RNA are aberrantly expressed in vivo in the cystic fibrosis bronchial epithelium, Int. J. Biochem. Cell Biol. 52 (2014) 184–191. [15] X. He, Y. He, B. Xi, J. Zheng, X. Zeng, Q. Cai, Y. Ouyang, C. Wang, X. Zhou, H. Huang, W. Deng, S. Xin, Q. Huang, H. Liu, LncRNAs expression in preeclampsia placenta reveals the potential role of LncRNAs contributing to preeclampsia pathogenesis, PLoS One 8 (11) (2013), e81437. [16] Y. Zou, Z. Jiang, X. Yu, M. Sun, Y. Zhang, Q. Zuo, J. Zhou, N. Yang, P. Han, Z. Ge, W. De, L. Sun, Upregulation of long noncoding RNA SPRY4-IT1 modulates proliferation, migration, apoptosis, and network formation in trophoblast cells HTR-8SV/neo, PLoS One 8 (11) (2013), e79598. [17] S. Buckberry, T. Bianco-Miotto, C.T. Roberts, Imprinted and X-linked non-coding RNAs as potential regulators of human placental function, Epigenetics 9 (1) (2014) 81–89. [18] Y. Zhang, Y. Zou, W. Wang, Q. Zuo, Z. Jiang, M. Sun, W. De, L. Sun, Down-regulated long non-coding RNA MEG3 and its effect on promoting apoptosis and suppressing migration of trophoblast cells, J. Cell. Biochem. 116 (4) (2015) 542–550. [19] V. Plaks, J. Rinkenberger, J. Dai, M. Flannery, M. Sund, K. Kanasaki, W. Ni, R. Kalluri, Z. Werb, Matrix metalloproteinase-9 deficiency phenocopies features of preeclampsia
199
and intrauterine growth restriction, Proc. Natl. Acad. Sci. U. S. A. 110 (27) (2013) 11109–11114. [20] J.M. Roberts, C.A. Hubel, The two stage model of preeclampsia: variations on the theme, Placenta 30 (Suppl A) (2009) S32–S37. [21] D. Raymond, E. Peterson, A critical review of early-onset and late-onset preeclampsia, Obstet. Gynecol. Surv. 66 (8) (2011) 497–506. [22] C. Qiao, C. Wang, J. Zhao, C. Liu, T. Shang, Elevated expression of KiSS-1 in placenta of Chinese women with early-onset preeclampsia, PLoS One 7 (11) (2012), e48937. [23] Y. Yang, J. Zhang, Y. Gong, X. Liu, Y. Bai, W. Xu, R. Zhou, Increased expression of prostasin contributes to early-onset severe preeclampsia through inhibiting trophoblast invasion, J. Perinatol.: official journal of the California Perinatal Association 35 (1) (2015) 16–22. [24] A. Kulkarni, P. Chavan-Gautam, S. Mehendale, H. Yadav, S. Joshi, Global DNA methylation patterns in placenta and its association with maternal hypertension in preeclampsia, DNA Cell Biol. 30 (2) (2011) 79–84. [25] M. van Dijk, C. Oudejans, (Epi)genetic control of human trophoblast invasion, Front. Genet. 5 (2014) 38. [26] M. Guttman, I. Amit, M. Garber, C. French, M.F. Lin, D. Feldser, M. Huarte, O. Zuk, B.W. Carey, J.P. Cassady, M.N. Cabili, R. Jaenisch, T.S. Mikkelsen, T. Jacks, N. Hacohen, B.E. Bernstein, M. Kellis, A. Regev, J.L. Rinn, E.S. Lander, Chromatin signature reveals over a thousand highly conserved large non-coding RNAs in mammals, Nature 458 (7235) (2009) 223–227. [27] M. Guttman, J. Donaghey, B.W. Carey, M. Garber, J.K. Grenier, G. Munson, G. Young, A.B. Lucas, R. Ach, L. Bruhn, X. Yang, I. Amit, A. Meissner, A. Regev, J.L. Rinn, D.E. Root, E.S. Lander, lincRNAs act in the circuitry controlling pluripotency and differentiation, Nature 477 (7364) (2011) 295–300. [28] S.K. Gupta, S.S. Malhotra, A. Malik, S. Verma, P. Chaudhary, Cell signaling pathways involved during invasion and syncytialization of trophoblast cells, Am. J. Reprod. Immunol. (2015). [29] K.R. Palmer, T.J. Kaitu'u-Lino, R. Hastie, N.J. Hannan, L. Ye, N. Binder, P. Cannon, L. Tuohey, T.G. Johns, A. Shub, S. Tong, Placental-specific sFLT-1 e15a protein is increased in preeclampsia, antagonizes vascular endothelial growth factor signaling, and has antiangiogenic activity, Hypertension 66 (6) (2015) 1251–1259. [30] L. Anton, A.G. Brown, M.S. Bartolomei, M.A. Elovitz, Differential methylation of genes associated with cell adhesion in preeclamptic placentas, PLoS One 9 (6) (2014), e100148. [31] W. Li, K.M. Mata, M.Q. Mazzuca, R.A. Khalil, Altered matrix metalloproteinase-2 and -9 expression/activity links placental ischemia and anti-angiogenic sFlt-1 to uteroplacental and vascular remodeling and collagen deposition in hypertensive pregnancy, Biochem. Pharmacol. 89 (3) (2014) 370–385. [32] K. Fukushima, M. Murata, M. Hachisuga, K. Tsukimori, H. Seki, S. Takeda, K. Asanoma, N. Wake, Hypoxia inducible factor 1 alpha regulates matrigel-induced endovascular differentiation under normoxia in a human extravillous trophoblast cell line, Placenta 29 (4) (2008) 324–331. [33] R. Tal, The role of hypoxia and hypoxia-inducible factor-1alpha in preeclampsia pathogenesis, Biol. Reprod. 87 (6) (2012) 134. [34] C. Agostinis, R. Bulla, C. Tripodo, A. Gismondi, H. Stabile, F. Bossi, C. Guarnotta, C. Garlanda, F. De Seta, P. Spessotto, A. Santoni, B. Ghebrehiwet, G. Girardi, F. Tedesco, An alternative role of C1q in cell migration and tissue remodeling: contribution to trophoblast invasion and placental development, J. Immunol. 185 (7) (2010) 4420–4429. [35] C.J. Lockwood, M. Basar, U.A. Kayisli, O. Guzeloglu-Kayisli, W. Murk, J. Wang, N. De Paz, J.P. Shapiro, R.J. Masch, N. Semerci, S.J. Huang, F. Schatz, Interferon-gamma protects first-trimester decidual cells against aberrant matrix metalloproteinases 1, 3, and 9 expression in preeclampsia, Am. J. Pathol. 184 (9) (2014) 2549–2559.