Introduction
Preeclampsia (PE), a specific complication of pregnancy, mainly causes injuries to the blood vessels and kidneys as well as long-term injuries, and remains the most important cause of maternal and neonatal death [1]. This complicated and serious pregnancy-related diseases is characterized by hypertension and proteinuria. Some cases of PE can uaually co-occur with gestational diabetes mellitus (GDM), for they may share common pathogrnrsis [2,3] and cause damage to multiple organs in the human body even in cases when prior to pregnancy, both blood pressure or blood glucose and renal function were normal in these patients.
As PE proceeds during pregnancy, blood pressure may continue to increase to more than 160/110 mmHg, and the level of proteinuria may reach greater than 5000 mg/24 h, developing into severe preeclampsia (sPE). Additional symptoms may develop as well, including HELLP syndrome (hemolysis, elevated liver enzymes and low platelets syndrome), persistent headache, and even a choked optic disc, resulting in vision loss [4]. Including poor placental vascular invasion, decreased 2-methoxyestradiol level, abnormal proteomics expression, many factors have been identified to participate in the pathogenesis of PE [5,6,7,8], but no more detailed or exact mechanisms were found. Due to the unknown pathogenesis of PE, prompt delivery of the fetus and placenta may be the only effective treatment in the later stages of the disease. Unfortunately, preterm delivery can result in further maternal and infant health problems. As a result, PE, and especially sPE, exists as one of the most common causes of mortality during pregnancy, with an increasingly higher incidence rate. Additional research is urgently needed to foster a better understanding of the pathogenic mechanisms involved in PE.
From the previous papers, non-coding RNAs, such as miRNAs, play a vital role in PE. Circular RNAs (circRNAs) are a special type of non-coding RNA in mammalian cells that interact closely with miRNAs; thus, circRNAs have attracted increasing attention from researchers due to their distinctive ring frame, which is in contrast the more common linear structures. As circRNAs have no 5' to 3' polarity or polyadenylated tails [9,10], they are immune to RNase and are expressed stably. circRNA expression is concentrated in some specific tissues or organs, particularly in the brain [11]. Most of the known circRNAs are produced from the back-splicing [12,13] of exons through three main mechanisms, including lariat-driven circularization, intron-pairing-driven circularization and the self-circularization of an intron to form circRNAs (ciRNAs) [14,15,16]. These features contribute to the performance of crucial physiological functions by circRNAs. Some specific circRNAs may have regulatory effects on gene expression [17] and human diseases, and some are correlated with the RNA binding protein Quaking [18]. The most well-known circRNA functions as a microRNA sponge by interacting with miRNA-7 [11,19]. As research has progressed, circRNAs have been demonstrated to be associated with atherosclerosis, neurological disorders, diabetes and cancer [20,21,22,23]. However, few studies in the field of gynecology and obstetrics have identified their potential significance in the onset of PE.
In this study, we quantified the expression level of circRNAs and identified the role that circRNAs played in placental tissue during the development of PE to provide a new avenue of research regarding the pathological mechanisms of PE.
Materials and Methods
Patients and sample collection
All human placental tissue samples were obtained from the Nanjing Medical University Affiliated Nanjing Maternal and Child Health Hospital from February 2014 to January 2015 (Table 1). Overall, the study included 75 patients. To form the experimental group, 40 samples were collected from patients with severe preeclampsia at gestational weeks ranging from 30 to 34 and with levels of proteinuria ranging from 5000 mg/24 h to 11,700 mg/24 h. The control group consisted of samples acquired from 35 healthy but premature births at 32.3 to 34 weeks gestation. Immediately following acquisition, the fresh placental tissue samples were placed in sterile, RNase-free 2.0 ml centrifuge tubes. Then, the samples were minced and allowed to soak in TRIzol. After these steps, all 75 samples were stored at -80 °C.
Six placental tissue samples, including three PE samples and three control samples, were sent to KANGCHEN (Shanghai, China) for the Arraystar circRNA Microarray analysis. On the basis of the chip results, we picked out several circRNAs as the candidate validation genes due to their high levels in both groups and their significant fold changes. The final choice of the validation genes was determined by the practicability of the designed divergent primers, which are described in greater detail below.
Total RNA extraction and reverse transcription
According to the instructions, we extracted total RNA from the samples using TRIzol reagent (Invitrogen, Karlsruhe, Germany) and the RNAprep pure tissue kit (TIANGEN) (DP431) in a step-by-step manner. The integrity of the extracted RNA was tested via 1% agarose gel electrophoresis using the following criteria: there could be up to three bands; the ratio of 28 S rRNA/18 S rRNA should be 2; and the 5 S rRNA should not be too bright. The purity of the extracted RNA was measured by a UV spectrophotometer using the following criteria: the 260/280 nm absorbance ratio of the qualified sample should be between 1.8 and 2.1, with 2.0 being considered best. According to the concentration of each sample, we added 1000 ng to the 20 µl reverse transcription system and then examined each sample using reverse transcription with random primers following the recommendations of the Thermo Scientific RevertAid First Strand cDNA Synthesis Kit.
Regular-PCR (R-PCR) and annealing temperature determination
Considering the unique ring structure of circRNAs, we designed corresponding divergent primers using Primer 3.0 according to the sequences provided by the microarray analysis results. These primers were synthesized by the Realgene Company (Nanjing, China). The integral internal reference gene we used was glyceraldehyde 3-phosphate dehydrogenase (GAPDH). The primers were designed to amplify a target sequence with a length of approximately 200 bp. Based on the instructions for Taq DNA Polymerase, we utilized 3 different temperature gradients (56, 59 and 62 °C) in a 25µl reaction system.
The thermal cycling conditions were as follows: start at 94 °C for 5 mins; 30 cycles of 94 °C for 30 s, a pre-selected annealing temperature for 30 s, and 72 °C for 30 s; and maintenance at 72 °C for 10 mins for full extension. The products of R-PCR were examined using 1.5% agarose gel electrophoresis under the following criteria: there should be only one band in the lane of the specific annealing temperature, and if there was one band in more than one lane, the brightest band was chosen; no existence of primer dimers or by-products; and compared to the DNA marker, the molecular weight must be equal to the size of the target fragment (approximately 200 bp). Finally, combining all the conditions above, three ideal up-regulated genes, hsa_circRNA_100782 (FC = 3.71), hsa_circRNA_102682 (FC = 3.59) and hsa_circRNA_104820 (FC = 5.96) were selected as the validation genes. Table 2 contains the sequences of the paired primers we designed. The best annealing temperatures were 59 °C for hsa_circRNA_100782 and hsa_circRNA_102682 and 56 °C for GAPDH and hsa_circRNA_104820.
qRT-PCR detection of target genes
Once the annealing temperature of each gene was identified, qRT-PCR was performed with SYBR. To eliminate the experimental random error, samples were loaded in triplicate and each well was treated identically. The data were analyzed using the 2-∆∆t method; to guarantee the accuracy of the results, all data are represented as the means ± SD of three independent experiments.
Statistical analyses
To analyze the complex and tedious data, both Statistical Program for Social Sciences Version 22 (SPSS) and GraphPad Prism 5.0 were utilized simultaneously. In addition, a t-test was also used to estimate the data, and p ≤ 0.05 was used to denote statistical significance. To determine the significance of these data for preeclampsia, we also established a receiver operating characteristic (ROC) curve for each circRNA. The area under the curve (AUC) was calculated for each respective circRNA.
Results
Characteristics of the study population
In total, 40 placenta samples of patients with PE and 35 placenta samples from corresponding premature births were collected in our study. The patients' characteristics are summarized in Table 1. No differences were observed between the two groups regarding age, height, weight, gestational week, ALT/AST, PLT, mode of delivery or the neonatal Apgar score (p > 0.05). However, the 24 h proteinuria, blood pressure and neonatal weight were significantly different between the two groups (p < 0.01).
Results of the microarray analysis
We accounted for the fold change (FC ≥ 2.0) and p-values (≤ 0.05) in this analysis. The expression of specific circRNAs was significantly different between groups. The general information pertaining to the detected circRNAs is shown in Figure 1. Altogether, 301 differentially expressed circRNAs were identified, of which 143 were up-regulated and 158 were down-regulated. Certain circRNAs with highly differential levels of expression are shown in Table 3 (FC ≥3.0).
Fig. 1
Detection results of all circRNAs. a. Scatter Plots The values for the X and Y axes are normalized signal values (log2 scaled). The green lines represent fold change lines. circRNAs above the top green line and below the bottom green line indicated more than a 2.0-fold change of circRNAs between the two groups. b. Volcano Plots The red point in the plot represents the differentially expressed circRNAs that were statistically significant. c. Histogram Considering the fold change and p-value, a total of 301 circRNAs were detected, of which 143 were up-regulated and 158 were down-regulated.
Evaluation of primers and products
To increase the rigor of the study and to enhance the efficacy of the primers that we designed, we measured both the specificity and sensitivity of the amplification products. After qRT-PCR, 1.5% agarose gel electrophoresis was used to test the uniqueness of the products and the relative amounts between the experimental and control groups (Fig. 2). According to the electrophoresis bands, only the targeted products and no primer dimers or non-specific amplification products were present, indicating that the divergent primers utilized for the circRNAs were effective and appropriate. The data in Figure 3 and the band brightness in Figure 2 show that the circRNA levels were significantly higher in PE group.
Fig. 2
Image showing testing of qRT-PCR products on a 1.5% agarose gel. The molecular weight of the marker is 10,000. According to the markers, the molecular weights of the target genes were approximately 200 - 250 bp. Only one band was present in each lane. In comparison, the bands in the control samples were not as bright as their corresponding PE samples, which shows the differential circRNA expression levels between PE samples and controls.
Fig. 3
The expression levels of circular RNAs in patients with PE and patients who delivered prematurely. The expression levels of hsa_circRNA_100782, hsa_circRNA_102682 and hsa_circRNA_104820 in each patient were compared. Higher ΔCt values indicate lower levels of expression. The expression levels of each gene were significantly higher versus their controls; all p-values < 0.05.
Values of circRNA in PE
These validation results agreed with the ROC curves (Fig. 4). The potential role of circRNAs in the pathogenesis of PE is of great importance. Furthermore, the area under the ROC curve for hsa_circRNA_100782, hsa_circRNA_102682 and hsa_circRNA_l04820 were 0.653, 0.774 and 0.995, respectively.
Discussion
circRNAs differ substantially from conventional linear RNAs and have recently become an important research topic due to their stable structures and high degrees of tissue specificity. circRNAs were first accidently observed in RNA viruses and were regarded as splicing errors or by-products during the onset of splicing [9,12] despite having been observed for decades in eukaryotic cells. While previously limited by existing technology, circRNA research has emerged as an important research topic only in recent years. circRNAs have highly conserved sequences and stable expression in different individuals [15]; these futures hint at their potential to mediate the occurrence of specific diseases in the absence of an external environment influence.
Exonic circRNAs might have extraordinary effects in cellular physiology, including miRNA binding, translational regulation, protein interactions, and even protein translation (found only in viruses) [23]. Additional studies have noted that circRNAs can function as miRNA sponges, which is to say circRNAs have many miRNA binding sites that competitively bind to miRNAs. Thus, circRNAs may alleviate the inhibitory effects of miRNAs on target molecules, thereby regulating gene expression levels. The most widely studied molecular sponge is antisense to the cerebellar degeneration-related protein1 transcript (CDR1as), which is located in the brains of humans and mice. CDR1as has approximately 74 miR-7 binding sites, and CDRlas over-expression can down-regulate miR-7 expression level [11,14,16]. This finding revealed a new therapeutic strategy for Alzheimer's disease.
The circRNAs isolated from the placental tissues of PE patients in our study also have several miRNA binding sites [24], and some even were associated with two different diseases like PE and gestational diabetes mellitus [25]. As shown in Table 4, many circRNAs have miRNA-17 binding sites, suggesting that these circRNAs can regulate the expression level of miRNA-17 in human placental tissues. miRNA-17 has been identified as one of the angiogenesis-associated miRNAs in the human placenta and was found to be highly expressed in PE placentas [30]. In the work of Chen and Wang [24] and Wang et al. [30], up-regulated miRNA-17 in the placenta could advance the process of PE by targeting the ephrin-B2/Eph receptor B4 (EPHB4) system, a classical pathway involved in trophoblast invasion; notably, the disorganization of this system exacerbates the process during PE. The differential expression of circRNAs could possibly up-regulate the expression levels of miRNA-17 through miRNA sponges, thereby contributing to the pathogenesis of PE. This possibility warrants further investigation.
Table 4
List of PE incidence-related miRNAs in placenta. miRNAs in bold were shown to participate in the pathogenesis of PE via different pathways in the placenta. Their references are given in brackets
The fact that the miRNA binding sites ofmiRNA-17 are strongly related to the onsetof PE is not a peculiar phenomenon, and from Table 4, we can tell that many other MREs of other circRNAs also play an important role in PE. These results suggested that the sponge functions of circRNAs may be highly significant and thus deserve further investigation. Despite the miRNA sponge functions, the interaction of proteins in blood corpuscles should also be mentioned. Recently, the work of Zhang et al. [31] showed that one plasma protein factor, endoglin (ENG), can combine with the up-regulated circRNA_101222 in blood corpuscles of preeclampsia patients; this combination of circRNA and ENG may be a potential biomarker for the early prediction and diagnosis of PE. While this finding reveals new information for researchers, further study is still warranted in the future.
It is believed that when PE occurs, the deficiency of placental trophoblast invasion may cause reduced placental flow and, eventually, placental ischemia, which is the initiating agent for subsequent disorders. Ischemia accompanied by hypoxia raises a series of complications regarding placental function, and some virulence factors are released into the blood. Therefore, the placenta is both the start and center of all the mechanisms responsible for PE. As the expression of circRNAs may differ between the placenta and the peripheral blood [32], we utilized placental tissues rather than serum in our study for a more circumstantial research.
Our study has some limitations. First, the sample selection was limited by a small number of PE samples, which was not large enough to establish definitive conclusions. Moreover, the choice of samples may not be generalizable to the general population. In future works, more samples should be collected to perform a detailed study. Equally important is that the samples were all collected from one hospital in a single year, which may have resulted in regional differences. Second, the study of circRNAs in PE has just started, and the functional analysis is imperfect; more work should be done to improve this shortcoming in the future. Through further study regarding the functions of circRNAs, our understanding of circRNA-related mechanisms of diseases could be improved, and the diagnostic accuracy and development of alternative prevention methods could be enhanced. Third, the expression of circRNAs in peripheral blood is essential for finding a suitable biomarker for earlier diagnoses of PE. In future studies, we also plan to analyze blood samples.
PE is an agnogenic disease occurring in pregnant women that primarily causes renal damage. PE cannot be identified until clinical manifestation becomes apparent, which is usually too late for clinical intervention measures. Thus, there is an urgent need to discover an ideal biomarker for PE before the condition progresses beyond the point of treatment. We hope that our research will open up new research directions regarding the pathogenesis of PE and will foster the development of effective breakthroughs.
To conclude, to identify the role that circRNAs play in PE placental tissues, our study analyzed the content of circRNAs in both patients with PE and patients who delivered prematurely. The results indicated that circRNAs expression differed significantly between the two groups and that circRNAs may have an important function as miRNA sponges. Our research represents a new breakthrough in the pathogenesis of PE. Of course, more work is needed to further uncover the molecular mechanisms of circRNAs and to reveal their deeper involvement in disease pathogenesis.
Acknowledgments
This study was supported by grants from the National Natural Science Foundation of China (No. 81571444, 81501341) and the Nanjing Medical Science and Technology Development Foundation (No. YKK15164).
Disclosure Statement
We declare that we have no financial and personal relationships with other individuals or organizations that can inappropriately influence our work. There are no professional or other personal interests of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, the manuscript entitled "Potential Significance of Circular RNA in Human Placental Tissues for Patients with Preeclampsia".


Get Permission
