Sage Journals: Discover world-class research

Abstract

BACKGROUND:

MicroRNAs (miRNAs), with noticeable stability and unique expression pattern in plasma of patients with various diseases, are powerful non-invasive biomarkers for cancer detection including endometrial cancer (EC).

OBJECTIVE:

The objective of this study was to identify promising miRNA biomarkers in plasma to assist the clinical screening of EC.

METHODS:

A total of 93 EC and 79 normal control (NC) plasma samples were analyzed using Quantitative Real-time Polymerase Chain Reaction (qRT-PCR) in this four-stage experiment. The receiver operating characteristic curve (ROC) analysis was conducted to evaluate the diagnostic value. Additionally, the expression features of the identified miRNAs were further explored in tissues and plasma exosomes samples.

RESULTS:

The expression of miR-142-3p, miR-146a-5p, and miR-151a-5p was significantly overexpressed in the plasma of EC patients compared with NCs. Areas under the ROC curve of the 3-miRNA signature were 0.729, 0.751, and 0.789 for the training, testing, and external validation phases, respectively. The diagnostic performance of the identified signature proved to be stable in the three public datasets and superior to the other miRNA biomarkers in EC diagnosis. Moreover, the expression of miR-151a-5p was significantly elevated in EC plasma exosomes.

CONCLUSIONS:

A signature consisting of 3 plasma miRNAs was identified and showed potential for the non-invasive diagnosis of EC.

Keywords

Plasma microRNA endometrial carcinoma diagnostic biomarker qRT-PCR

1. Introduction

Endometrial cancer (EC) is the third most common cancer among women worldwide [1]. However, there is no early detection method for asymptomatic women at the average risk of endometrial cancer [2]. Most guidelines recommend either transvaginal ultrasonography or endometrial biopsy as the initial detection but none of them are completely satisfactory. Diagnostic curettage is a widely used diagnostic method in EC patients. However, diagnostic curettage has several inherent defects, such as invasive and blindness in scraping tissue specimens together with missed diagnosis, and leads to different degrees of endometrial bleeding in some patients. Patients with endometrial cancer may make cancer cells to invade the muscle layer, which may result in tumor metastasis, causing serious consequences [3]. In addition, diagnostic curettage needs to be operated by experienced gynaecologists and possesses risk of missed diagnosis. Thus, identification of accurate and validated biomarkers for EC is needed to improve diagnosis and lead better survival outcome of EC.

MicroRNA (miRNA), a type of non-coding small RNA (18–25 nucleotides), functions as a post-transcrip- tional regulator via the repression of translation or direct degradation of target messenger RNA (mRNA) [4]. Many studies have unraveled its roles in the biological process of various cancers [5, 6, 7]. Meanwhile, circulating miRNAs show great potential of being cancer biomarkers for the stable existence in peripheral serum or plasma. Since blood samples are easy to obtain and the testing procedure is more convenient and cheaper, circulating biomarker can be used alone or combined with other traditional screening methods for preliminary screening before further invasive pathological and imaging examinations. It can help to identify the existence of cancer disease as early as possible, which can be employed for the screening of both undiagnosed suspected cases and healthy population. Several studies have shown that certain miRNAs could serve as highly accurate biomarkers for EC detection [8, 9, 10, 11, 12]. However, these findings lacked consistency between each other because of the varied research methods and study populations. In this study, we performed a four-stage study to determine the diagnostic value of plasma miRNAs using quantitative reverse transcription polymerase chain reaction (qRT-PCR). In addition, we also analyzed miRNA expression in EC tissues and plasma exosomes to promote the understanding of the possible existence form and biological function of the identified miRNAs.

2. Materials and methods

2.1 Information of participants

The study was conducted in accordance with the guidelines of the Hospital Ethics Committee and approved by the Institutional Review Boards of the First Affiliated Hospital of Nanjing Medical University (ID: 2016-SRFA-148). Totally, participants in this study included 93 patients and 79 normal controls (NCs) from the First Affiliated Hospital of Nanjing Medical University and Women’s Hospital of Nanjing Medical University between 2016 and 2017. The cases we collected may come from all over the country, but most of them originated in East China and their ethnic group is basically Han. Patients involved were diagnosed as EC by histopathological examination and didn’t accept any therapeutic intervention such as surgery, radiotherapy or drug treatment before sampling. These healthy volunteers were randomly recruited from the health examine center of the hospital. They were ensured to be EC-free after image examination. In addition, they were familiar with the project and volunteer to participate in this study and had signed the informed consent form. The criteria of sample inclusion have been summarized in Table S1.

2.2 Study design and sample preparation

The present study was divided into four stages (Fig. 1), and in the initial screening phase, 2 peripheral plasma pool samples from 20 EC patients and 1 pooled sample from 10 NCs were selected. Exiqon miRCURY-Ready-to-Use PCR-Human-panel-I+II-V1.M (Exiqon miRNA qPCR panel, Vedbaek, Denmark) was applied to identify dysregulated miRNAs preliminarily. Based on the candidate list obtained in the screening stage, qRT-PCR was used to verify the expression of each candidate miRNA. The external validation phase (27 EC vs. 23 NCs) was designed to finally verify the characteristics of the identified miRNA. In addition, tissue samples (24 EC vs. 24 NCs) and exosomes samples (32 EC vs. 32 NCs) were analyzed to explore the expression of the identified miRNAs.

Figure 1.

The flow chart of the experiment design. EC: endometrial cancer; NC: normal control; qRT-PCR: Quantitative real-time polymerase chain reaction.

Blood samples from EC patients and healthy controls were collected with ethylenediaminetetraacetic acid (EDTA)-containing tubes (BD, New Jersey, USA). Cell-free plasma was separated from blood samples within 6 hours. The plasma samples were separated using a two-step centrifugation protocol (350 reactive centrifugal force (RCF) for 10 min at 4 ${}^{\circ}$ C, 20,000 RCF for 10 min), then samples were kept at $-$ 80 ${}^{\circ}$ C. Formalin-fixed paraffin-embedded (FFPE) EC tissue specimens and adjacent normal tissue samples were obtained from 24 EC patients without preoperative chemoradiotherapy.

2.3 Isolation of exosomes

Exosomes were isolated from plasma using Exo-Quick Exosome Precipitation Solution (System Biosciences, Mountain View, CA, USA) as described by the manufacturer. Briefly, the exosomes pellets were precipitated from 200 $\mu$ L plasma and 100 $\mu$ L ExoQuick exosome precipitation solutions, incubated at 4 ${}^{\circ}$ C for at least 30 min, and then dissolved in 200 $\mu$ L of RNase-free water for further RNA extraction.

2.4 Extraction of RNA

RNA was isolated from 200 $\mu$ L plasma using mirVana Paris Kit (Ambion, Austin, TX, USA) according to the manufacturer’s protocol. Each sample was mixed with denaturing solution (Ambion, Austin, TX, USA), and 5 $\mu$ L synthetic C. elegans miRNA cel-miR-39 (5 mmol/L, RiboBio, Guangzhou, China) was added into the samples as the spike-in controls. Total miRNA was extracted from FFPE samples using the High Pure FFPE RNA Micro Kit (Ambion, Austin, TX, USA). Total RNA was eluted with 100 $\mu$ L RNase-free water and stored at $-$ 80 ${}^{\circ}$ C until further analysis. The concentration and purity of plasma RNA in this study were determined using Nanodrop 2000 spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA).

Table 1
Clinical characteristics of 93 EC patients and 79 normal controls

Variables	Training stage ( $n=$ 44)		Testing stage ( $n=$ 78)		External validation stage ( $n=$ 50)
	Cases (%)	Controls (%)	Cases (%)	Controls (%)	Cases (%)	Controls (%)
Number	22	22	44	34	27	23
Age
$\leqslant$ 50	8 (31.8)	13 (64.5)	14 (31.8)	21 (61.8)	9 (33.3)	13 (56.5)
$>$ 50	14 (63.6)	9 (35.5)	30 (68.2)	13 (38.2)	18 (66.7)	10 (43.5)
FIGO stage
I	19 (86.4)		38 (86.4)		26 (96.3)
II, III, IV	3 (13.6)		6 (13.6)		1 (3.7)
Menopausal status
Pre-	6 (27.3)	10 (45.4)	10 (22.7)	19 (55.9)	8 (29.6)	10 (43.5)
Post-	11 (50.0)	8 (36.4)	30 (68.2)	11 (32.3)	18 (66.7)	8 (34.8)
Unknow	5 (22.7)	4 (18.2)	4 (9.1)	4 (11.8)	1 (3.7)	5 (21.7)
Histological grade
G1	5 (13.6)		11 (25.0)		10 (37.0)
G2	12 (40.9)		26 (59.1)		13 (48.2)
G3	5 (13.6)		7 (15.9)		4 (14.8)

2.5 Quantitative real-time polymerase chain reaction (qRT-PCR)

The specific primers of reverse transcription (RT) and polymerase chain reaction (PCR) from Bulge-Loop ${}^{\text{TM}}$ miRNA qRT-PCR Primer Set (RiboBio, China) were used to amplify miRNAs. The process of RT and PCR were performed based on the manufacturer’s instructions. Then, in order to quantify the relative expression of mature miRNAs, SYBR Green-based (TaKaRa, Dalian, China) qRT-PCR method was detected in a LightCycler 480 (Roche 480, Germany) real-time thermal cycler. The expression of miRNAs in plasma or exosomes was calculated using the 2 ${}^{-\Delta\Delta\text{Ct}}$ method (cel-miR-39 as exogenous reference miRNA for plasma samples; RNU6B (U6) [U6] for tissue sample normalization; $\Delta$ Ct $=$ Ct ${}_{\text{miRNA}}$ $-$ Ct ${}_{\text{normalizer}}$ ; Ct: the threshold cycle) [13].

2.6 Statistical analysis

Statistical analyses and graph plotting were performed using SPSS 24.0 software (SPSS Inc., Chicago, IL, USA) and GraphPad Prism 7.0 (GraphPad Software, USA). To assess the comparison of miRNA expression between EC patients and healthy controls, nonparametric Mann-Whitney test were conducted. The relationship of miRNAs and clinical characteristics was evaluated by one-way ANOVA or $\chi^{2}$ test. Logistic regression analysis was used to establish a combined miRNA panel to predict the probability of EC. Receiver-operating characteristic (ROC) curves and the area under the ROC curve (AUC) were utilized to evaluate the sensitivity and specificity of identified miRNAs or the miRNA panels. When $P<$ 0.05, the result was considered to be statistically significant. Decision curve analysis (DCA) was performed using the “dca.R” (decisioncurveanalysis.org).

2.7 Bioinformatics analysis

The enrichment of overlapping target gene signaling pathways of these miRNAs was performed using the online tool of DIANA-miRPath v3.0 [14]. The potential biological processes and pathways involved by miRNAs were evaluated using the assembly of Gene Ontology (GO) biological process and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway databases.

3. Results

3.1 Characteristics of participants

A total of 172 plasma samples (93 EC vs. 79 NCs) were analyzed in this four-stage experiment. The schematic diagram of the experimental procedure was shown in Fig. 1 and the clinical characteristics of the study subjects were listed in Table 1. There was no significant difference in the distribution of age between EC patients and NCs during the four stages ( $P>$ 0.05).

Table 2
Endometrial cancer diagnostic capability of the signature and previously identified biomarkers (presented as the results of ROC curve analysis using the data from GEO dataset)

GEO	Researcher	Biomarker	AUC	95% CI	Sensitivity (%)	Specificity (%)
GSE35794	The present study	3-miRNA panel	0.903	0.767–1	100	77.8
	Andrea Ritter et al. [9]	miR-484 and miR-23a	0.847	0.652–1	77.8	75.0
	Wenhui Jia et al. [10]	miR-222, miR-223, miR-186, and miR-204	0.806	0.597–1	77.8	75.0
	Anna Torres et al. [8]	miR-449a and miR-1228	0.889	0.736–1	77.8	100
	Martina Montagnana et al. [11]	miR-222, miR-223, miR-186, and miR-204	0.806	0.597–1	77.8	75.0
GSE25405	The present study	3-miRNA panel	0.857	0.707–1	85.7	84.6
	Andrea Ritter et al. [9]	miR-484 and miR-23a	0.612	0.420–0.803	56.4	71.4
	Anna Torres et al. [8]	miR-449a and miR-1228	0.762	0.617–0.906	66.7	85.7
	Lin Wang et al. [12]	miR-15b, miR-27a, and miR-223	0.835	0.669–1	84.6	85.7

Figure 2.

Expression levels of the three identified miRNAs in the plasma of 93 EC patients and 79 female NCs. Horizontal line: mean with 95% confidence interval (CI). A: miR-142-3p; B: miR-146a-5p; C: miR-151a-5p. N: normal controls; T: tumor.

3.2 MiRNA quantification in screening phase

In the screening phase, 174 miRNAs were measured by the Exiqon miRCURY-Ready-to-Use PCR-Human-panel-I+II-V1.M based on the qRT-PCR platform in 2 peripheral plasma pools from 20 EC patients; 1 pooled sample came from 10 female NCs. The criteria of candidate miRNAs for further analysis were listed: (a) a Ct-value $<$ 37; (b) a Ct-value 5 lower than negative control (No Template Control); (c) at least a 1.5-fold altered expression in EC patients compared with healthy controls. Taken together, 30 differentially expressed miRNAs were selected as candidate miRNAs and were chosen to be further assessed via qRT-PCR (Table S2).

3.3 Evaluation of serum miRNAs by qRT-PCR

Firstly, in the training stage, 12 differentially expressed miRNAs (has-miR-142-3p, has-miR-24-3p, has-miR-151-5p, has-miR-155-5p, has-miR-423-3p, has-miR-146a-5p, has-miR-154-5p, has-miR-197-3p, has-miR-181a-5p, has-miR-29c-3p, has-miR-125b-5p, has-miR-143-3p) were identified. In the following testing stage, the plasma samples of 44 EC patients and 34 female NCs were used for validation, and 3 miRNAs were consistently upregulated. To verify the diagnostic capacity of the three miRNAs for EC patients, the expression levels of miRNAs were further assessed among an additional independent cohort of 27 EC patients and 23 NCs. Combined results showed that miR-142-3p, miR-146a-5p, and miR-151a-5p were significantly up-regulated in the plasma of EC patients (Fig. 2).

3.4 Evaluation of diagnostic value

We used ROC analysis to evaluate the performance of the 3 miRNAs in identifying EC patients from NCs. Firstly, we separately analyzed the three miRNAs in a combined cohort of 93 EC patients and 79 NCs. The AUCs were 0.689 (95% confidence interval (CI): 0.611–0.767) for miR-142-3p, 0.694 (95% CI: 0.616–0.772) for miR-146a-5p, and 0.680 (95% CI: 0.601–0.759) for miR-151a-5p, respectively (Fig. S1). Though single miRNA could distinguish EC from healthy controls, a combination of several miRNAs might provide stronger differentiation power than individual miRNA. The three miRNAs were combined as a panel using logistic regression analysis and the equation to calculate the probability of EC was: Logit (P) $=$ 0.630 $-$ 0.028 $\times$ miR-142-3p - 0.179 $\times$ miR-146a-5p - 0.075 $\times$ miR-151a-5p. As is demonstrated in Fig. 3A, the AUC of the 3-miRNA panel was 0.716 (95% CI: 0.640–0.793, Sensitivity $=$ 62%, Specificity $=$ 64.5%, cut off $=$ 0.528). As shown in the DCA curve, when the threshold probability is greater than 35%, the 3-miRNA signature owes better diagnostic performance (Fig. S2). Meanwhile, the diagnostic performance of the 3-miRNA panel was also assessed in the training, testing, and external validation stages separately, and the AUCs were 0.729 (95% CI: 0.580–0.879), 0.751 (95% CI: 0.645–0.858), and 0.789 (95% CI: 0.664–0.914), respectively (Fig. 3B–D). The three-miRNA signature might become a suitable model for the detection of EC.

Figure 3.

Receiver-operating characteristic (ROC) curves for the three-miRNA signature to discriminate EC patients from NCs. (A) the combined data of training, testing and external validation stages (93 EC vs. 79 NC): AUC $=$ 0.716, 95% CI: 0.640–0.793; (B) the training stage (22 EC vs. 22 NC): AUC $=$ 0.729, 95% CI: 0.580–0.879; (C) the testing stage (44 EC vs. 34 NC): AUC $=$ 0.751, 95% CI: 0.645–0.858; (D) the external validation stage (27 EC vs. 23 NC): AUC $=$ 0.789, 95% CI: 0.664–0.914. AUC: Areas under the curve; CI: Confidence interval; EC: endometrial cancer; NC: normal control.

3.5 Association of identified miRNAs with clinicopathological parameters

In addition, this study analyzed the correlation between the identified miRNAs and clinicopathological parameters of 93 EC patients. Based on the analysis of FIGO stages, there was no statistically significant difference in the expression of the three miRNAs in the early stage (I) and advanced stage (II+III+IV) (Fig. S3). Subgroup analyses based on histological grade were also conducted and miR-142-3p was significantly overexpressed in well-differentiated EC patients (Fig. S4). The expression of miR-142-3p and miR-146a-5p in EC patients in pre-menopausal status was significantly higher than that of post-menopausal status (Fig. S5). The ROC curves of early stage (I) and advanced stage (II+III+IV) compared with NC group are 0.720 (95% CI: 0.642–0.797) and 0.713 (95% CI: 0.547–0.878) respectively, which proves that 3-miRNA marker can distinguish EC patients at various stages from NCs (Fig. 4). The ROC analysis was further conducted for histological grade and menopausal status. As is shown in the Fig. 4, the 3-miRNA signature performs well in discriminating EC cases of different histological grade and menopausal status from healthy controls, with AUCs of 0.768 (95% CI: 0.666–0.870) for grade I, 0.733 (95% CI: 0.649–0.816) for grade II, 0.672 (95% CI: 0.547–0.797) for grade III, 0.792 (95% CI: 0. 673–0.911) for pre-menopause status and 0.678 (95% CI: 0. 587–0.768) for post-menopause status.

Figure 4.

ROC curves of the 3-miRNA panel for discriminating EC patients at different clinicopathological parameters from NCs. (A) Stage I: AUC $=$ 0.720, 95% CI: 0.642–0.797; (B) Stage II-IV: AUC $=$ 0.713, 95% CI: 0.547–0.878; (C) G1: AUC $=$ 0.768, 95% CI: 0.666–0.870; (D) G2: AUC $=$ 0.733, 95% CI: 0.649–0.816; (E) G3: AUC $=$ 0.672, 95% CI: 0.547–0.797; (F) Pre-menopausal: AUC $=$ 0.792, 95% CI: 0.673–0.911; (G) Post-menopausal: AUC $=$ 0.678, 95% CI: 0.587–0.768. ROC: Receiver-operating characteristic; AUC: Areas under the curve; CI: Confidence interval; EC: endometrial cancer; NC: normal control.

Figure 5.

External validation of the 3-miRNA signature in TCGA-UCEC (A), GSE35794 (B), and GSE25405 (C) datasets. A: TCGA-UCEC: AUC $=$ 0.680, 95% CI: 0.673–0.911, $P=$ 0.004; B: GSE35794: AUC $=$ 0.903, 95% CI: 0.767–1.000, $P=$ 0.014; C: GSE25405: AUC $=$ 0.857, 95% CI: 0.701–1.000, $P=$ 0.003. AUC: Areas under the curve; CI: Confidence interval.

3.6 External validation in public databases

The diagnostic value of the identified 3-miRNA signatures was further verified in three public datasets: The Cancer Genome Atlas Uterine Corpus Endometrial Carcinoma (TCGA-UCEC) dataset, GSE35794 dataset (18 tumor samples and 4 normal controls), and GSE25405 dataset (41 tumor samples and 7 normal controls) (Fig. 5). In the TCGA-UCEC dataset, the identified signatures still performed well in discriminating EC tumor tissues ( $n=$ 531) and normal tissues ( $n=$ 22). The corresponding AUC of the 3-miRNA signature for EC detection was 0.680 (95% CI: 0.578–0.782). In GSE35794, the AUCs of the 3-miRNA panels to distinguish EC patients from healthy people were 0.903 (95% CI: 0.767–1.000). In GSE25405, the AUCs of the 3-miRNA panels to distinguish EC patients from healthy people were 0.857 (95% CI: 0.701–1.000). In summary, the diagnostic value of the identified signature was successfully verified in the three public datasets.

3.7 Comparison with previously identified signatures

Comparing the diagnostic capabilities of the signature with the five previously reported signatures could similarly distinguish EC patients from healthy people. As is shown in Supplementary Table 2, the 3-miRNA signature in this study is superior to the other miRNA biomarkers in EC diagnosis.

3.8 Expression of identified miRNAs in tissue and plasma exosomes samples

In present study, the expression patterns of the three miRNAs were examined in 24 pairs of tissue samples of EC patients. However, none of them were significantly up-regulated in tumor tissues compared to their matched controls (Fig. 6). In addition, the expression of the three miRNAs was also examined in plasma exosomes from 32 EC patients and 32 female NCs. As is shown in Fig. 6, the significant dysregulation of exosomal miR-151a-5p was consistent with the dysregulation in plasma. No significant difference was found in miR-142-3p and miR-146a-5p expression.

Figure 6.

Expression of the three miRNAs in the tumor tissues and plasma exosomes. Horizontal line: mean with 95% CI. A: The expression of miR-142-3p in the tumor tissues; B: The expression of miR-146a-5p in the tumor tissues; C: The expression of miR-151a-5p in the tumor tissues; D: The expression of miR-142-3p in the plasma exosomes; E: The expression of miR-146a-5p in the plasma exosomes; F: The expression of miR-151a-5p in the plasma exosomes; N: normal controls; T: tumor.

3.9 Functional exploration of the identified miRNAs

We utilized DIANA-mirPath v3.0 (http://www.microrna.gr/miRPathv3), a miRNA pathway analysis web server, to predict the possible functions of the miR-142-3p, miR-146a-5p, and miR-151a-5p. These programs were used for gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) category analysis to investigate the significantly pathways. We identified 31 biological processes associated with the three identified miRNAs, such as MyD88-independent toll-like receptor signaling pathway, TRIF-dependent toll-like receptor signaling pathway and Fc-epsilon receptor signaling pathway. In human cancers, MyD88 protein acts as a bridge between the inflammatory signaling from the TLR/IL-1R and Ras oncogenic signaling pathway [15]. Through activation, TLR4 recruits adaptor proteins, MyD88 or TRIF, to trigger canonical and non-canonical signaling pathways that lead to different immune responses [16]. Targeted pathways’ heatmaps are shown in Fig. S6.

4. Discussion

EC is one of the most common malignant tumors in females. Although quantities of cases have been diagnosed at early stages, there are still a proportion of patients who show high-grade malignancy or recurrence and have shorter overall survival.

Circulating miRNAs, including miRNAs in serum, plasma, or circulating microcarriers such as peripheral exosomes, have particular advantages as diagnostic markers for cancers. Firstly, circulating miRNAs are quite easy to obtain without damage which permits the dynamic monitoring of patients and the routine examination of healthy people. Secondly, PCR, the critical step of miRNA detection, can be more sensitive since it maximums the difference signal between cases and controls. Compared with other conventional screening methods, miRNA detecting method is also cheaper, quicker, and more objective. Last but not least, the aberrant expression of certain miRNAs in blood circulation not only reflects the existence of disease status but also indicates the potential mechanisms of tumor development, progression, invasion, and metastasis. Circulating miRNA can be both a biomarker and a therapeutic target [17, 18]. Although the sensitivity and specificity of different miRNA biomarkers may fluctuate greatly, stable circulating miRNA signatures, combined with other traditional methods, can play a potent role in assisting clinical diagnosis.

We reviewed the current methodologies of detecting circulating miRNAs and analyzed several key factors affecting miRNA quantification. On the one hand, considering that not only race and population differences but also variety of samples will affect the outcome of the research, multicenter data is needed for further verification. On the other hand, there are multiple standardized methods for the quantitative analysis of circulating miRNA, including adding endogenous or exogenous reference. At present, there is no generally accepted endogenous reference for circulating miRNA analysis [19, 20]. It was reported that correcting sample processing differences by adding exogenous miRNAs may be a better method for quantitative analysis of circulating miRNAs [19, 21, 22]. In addition, inconsistent platforms usually show different results in the initial screening stage. Exiqon miRNA qPCR panel was proved to have better linearity and sensitivity in the quantification of miRNAs at relatively lower abundance than the TaqMan platform [23, 24, 25].

In recent years, the diagnostic potential of circulating miRNAs has been gradually discovered in EC. Potential biomarkers such as circulating miR-222, miR-223, miR-186, miR-204, miR-484, and miR-23a were continuously discovered, but these results lacked consistency within these studies. Through four-stage screening, we investigated the application of plasma miRNAs for EC diagnosis among 93 EC patients and 79 NCs in this study, and we found that a biomarker combining over-expressed miR-142-3p, miR-146a-5p, and miR-151a-5p in plasma could be utilized to diagnose EC.

During the screening stage, we used the Exiqon miRNA qPCR panel to initially screen differentially expressed plasma miRNAs in 2 EC and 1 NC pooled samples. In the testing and the training stage, we further used qRT-PCR to validate the candidate miRNAs from the screening stage. In this study, three plasma miRNAs (miR-142-3p, miR-146a-5p and miR-151a-5p) with significantly higher expression levels that could accurately discriminate EC patients from healthy women. A panel combing these three miRNAs was then built using logistic regression analysis for EC detection. The AUCs of the panel turned out to be 0.729, 0.751, and 0.789 for the training, testing, and external validation stage, respectively. Moreover, the expression levels of the three miRNAs were also examined in tissues and plasma-derived exosomes samples. However, none of them were significantly up-regulated in tumor tissues compared to their matched controls. MiR-151a-5p were up-regulated in the plasma exosomes of EC patients.

Deng et al. demonstrated that miR-142-3p was detected to be down-regulated in cervical cancer cells, and over expression of miR-142-3p inhibited proliferation and invasion of cervical cancer cells by targeting FZD7 [26]. In addition, Wang et al. validated miR-142-3p and miR-146a-5p as the part of miRNA signature that could serve as an independent predictor for overall survival (OS) of EC patients [27]. We found that the expression of miR-142-3p and miR-146a-5p in premenopausal EC patients was significantly higher than that in postmenopausal patients. It was reported that menopause could be an independent prognostic factor for EC and premenopausal EC patients would have a better prognosis. This further indicates that the mechanism and relationship between menopause and these two miRNAs are worthy of further study.

Zhou et al. once reported overexpression of miR-146a attenuates the effect of NIFK-AS1 on M2 polarization inhibition of macrophages and estrogen-induced endometrial cancer cell proliferation, migration and invasion [28]. It was reported that miR-146a-5p is involved in the pathogenesis of CC, and miR-146a has been shown to promote CC cell viability via targeting IRAK1 and TRAF6 [29]. In addition, miR-151a was identified as an oncomiRNA in non-small cell lung cancer. It has been shown to target E-cadherin mRNA and induce migration of proliferation, and partial EMTEMT [30]. However, there is little evidence of its role in EC. More research is needed to decipher the exact role of these identified miRNAs in EC.

The menstrual phase of the endometrium includes inflammation as a physiologic component [31]. menstruation involves the synthesis and release of inflammatory factors, including cytokines, growth factors, COX-2, and prostaglandins, as well as activation of NF- $\kappa$ B. Thus, in the absence of a pregnancy, the endometrium can be viewed as being cyclically exposed to a chronic inflammatory-like process [32, 33]. In addition to the circumstantial evidence linking inflammation and endometrial cancer just described, emerging laboratory data suggest a link as well [34, 35, 36]. Through KEGG and GO analysis, we know that the 3-miRNA marker is involved in various pathways related to immunity [16, 37]. The functional annotation results indicate that the 3-miRNA panel may play a key role in the occurrence and development of EC, and indirectly proves the reliability of the differences in miRNA expression found between EC patients and NCs in this study. These results indicate the potential for further large-scale verification and investigate the relevant mechanisms in future research.

In order to further verify the results of the study, we also explored plasma miRNAs for EC detection in tissues. However, in our study, the expression of the three miRNAs in tissue samples is not consistent with the miRNA plasma profile. We speculated that discrepancy between EC plasma and tissue samples might be due to the release and transportation of miRNAs in tumor and its microenvironment. Other biological factors might also affect the result, such as different biological states of tumors, influence of other non-tumor cells or some immunological factors. Circulating miRNA signals may reflect the changes in body fluids of patients with diseases, and the tissue miRNA profile may mainly show local organ diseases. Anyway, a larger number of tissue samples were in demand for further validation.

Exosomes are small cell-derived vesicles (30–120 nm) existing in almost all kinds of body fluids and can carry a variety of molecules including miRNAs. Exosomal miRNAs are associated with various cancers which shows their great potential of being diagnostic tools [38, 39, 40, 41]. Therefore, to explore the potential form and function of the three miRNAs, we further analyzed miRNA expression in plasma-derived exosomes. Exosomal miR-151a-5p was increased in EC compared to NCs. The findings might give some hints about cell-to-cell communication and cell-to-microenvironment communication mediated by the two exosomal miRNAs in the development and progression of EC.

Limitations of our study still remained. First of all, since the samples in this study were all from Chinese patients, the clinical outcomes of different races and biological behaviors were not considered. Secondly, the sample size in our study was relatively small and the diagnostic value in patients with advanced stage could not be well evaluated. Last but not least, the precise mechanisms by which these three miRNAs facilitate EC progression remained largely obscure. Our study was still in preclinical stage, and there would be a long way to go for future clinical application.

5. Conclusion

In summary, our study has identified and validated a three-miRNA signature (miR-142-3p, miR-146a-5p and miR-151a-5p) in plasma as a potential non-invasive biomarker for EC diagnosis.

Experimental subject statement

Approval for the present study was obtained from the Institutional Ethical Committee and participants have given written informed consent (ID: 2016-SRFA-148). The experiments were undertaken with the understanding and written consent of each subject, and that the study conforms with The Code of Ethics of the World Medical Association (Declaration of Helsinki), printed in the British Medical Journal (18 July 1964).

Authors’ contributions

Conception: Z.W and Z.H; Interpretation or analysis of data: Z.S.Y and C.W.F; Preparation of the manuscript: F.X.C; revision for important intellectual content: C.M.M, F.X.C, and L.C; Supervision: Z.C and L.C.Y.

Supplementary data

The supplementary files are available to download from http://dx.doi.org/10.3233/CBM-200972.

sj-docx-1-cbm-10.3233_CBM-200972.docx - Supplemental material

Supplemental material, sj-docx-1-cbm-10.3233_CBM-200972.docx

Footnotes

Acknowledgments

This work was supported by the Natural Science Foundation of China [Grant number: 81672400].

References

Neacşu

Marcu

M.L.

Stănică

C.D.

Brăila

A.D.

Pacu

Ioan

R.G.

Grigorescu

C.C.

and Ionescu

C.A.

, Clinical and morphological correlations in early diagnosis of endometrial cancer, Romanian Journal of Morphology and Embryology = Revue Roumaine de Morphologie et Embryologie 59 (2018), 527–531.

Tarney

C.M.

Wang

Bateman

N.W.

Conrads

K.A.

Zhou

Hood

B.L.

Loffredo

Tian

Darcy

K.M.

Hamilton

C.A.

Casablanca

Lokshin

Conrads

T.P.

and Maxwell

G.L.

, Biomarker panel for early detection of endometrial cancer in the Prostate, Lung, Colorectal, and Ovarian cancer screening trial, American Journal of Obstetrics and Gynecology 221 (2019).

Gottwald

Kubiak

Pasz-Walczak

Sek

Piekarski

Szwalski

Spych

Chałubińska-Fendler

Suzin

Tyliński

and Jeziorski

, The value of progesterone and estrogen receptors expression in tissue microarray method in prognosis of patients with endometrioid endometrial cancer, Ginekologia polska 84 (2013).

Zhou

Zhu

Cheng

Zhu

Shu

and Liu

, MiR-4728-3p could act as a marker of HER2 status, Cancer Biomark 15 (2015), 807–814.

Lee

Y.S.

and Dutta

, MicroRNAs in cancer, Annu Rev Pathol 4 (2009), 199–227.

S.R.

and Shi

Y.Q.

, Recent advances of miRNAs in the development and clinical application of gastric cancer, Chin Med J (Engl) 133 (2020), 1856–1867.

C.Y.

Zhang

W.W.

Xiang

J.L.

Wang

X.H.

and Wang

J.L.

, Identification of microRNAs as novel biomarkers for esophageal squamous cell carcinoma: a study based on The Cancer Genome Atlas (TCGA) and bioinformatics, Chin Med J (Engl) 132 (2019), 2213–2222.

Torres

Pesci

Ceccaroni

Paszkowski

Cassandrini

Zamboni

and Maciejewski

, Diagnostic and prognostic significance of miRNA signatures in tissues and plasma of endometrioid endometrial carcinoma patients, Int J Cancer 132 (2013), 1633–1645.

Ritter

Hirschfeld

Berner

Jaeger

Grundner-Culemann

Schlosser

Asberger

Weiss

Noethling

Mayer

and Erbes

, Discovery of potential serum and urine-based microRNA as minimally-invasive biomarkers for breast and gynecological cancer, Cancer Biomark 27 (2020), 225–242.

10.

Jia

Zhang

Gao

Zhang

and Xiang

, Identification of four serum microRNAs from a genome-wide serum microRNA expression profile as potential non-invasive biomarkers for endometrioid endometrial cancer, Oncol Lett 6 (2013), 261–267.

11.

Montagnana

Benati

Danese

Giudici

Perfranceschi

Ruzzenenete

Salvagno

G.L.

Bassi

Gelati

Paviati

Guidi

G.C.

Franchi

and Lippi

, Aberrant MicroRNA expression in patients with endometrial cancer, Int J Gynecol Cancer 27 (2017), 459–466.

12.

Wang

Chen

Y.J.

Shen

X.Z.

and Tu

R.Q.

, Circulating microRNAs as a fingerprint for endometrial endometrioid adenocarcinoma, PLoS One 9 (2014), e110767.

13.

Livak

K.J.

and Schmittgen

T.D.

, Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method, Methods 25 (2001), 402–408.

14.

Vlachos

I.S.

Zagganas

Paraskevopoulou

M.D.

Georgakilas

Karagkouni

Vergoulis

Dalamagas

and Hatzigeorgiou

A.G.

, DIANA-miRPath v3.0: deciphering microRNA function with experimental support, Nucleic Acids Res 43 (2015), W460–466.

15.

Wang

Zhang

and Yu

, Dual functional roles of the MyD88 signaling in colorectal cancer development, Biomed Pharmacother 107 (2018), 177–184.

16.

Lupi

L.A.

Cucielo

M.S.

Silveira

H.S.

Gaiotte

L.B.

Cesario

R.C.

Seiva

F.R.F.

and de Almeida Chuffa

L.G.

, The role of Toll-like receptor 4 signaling pathway in ovarian, cervical, and endometrial cancers, Life Sci 247 (2020), 117435.

17.

Wang

Peng

Wang

Qin

and Xue

, Circulating microRNAs as potential cancer biomarkers: the advantage and disadvantage, Clin Epigenetics 10 (2018), 59.

18.

Cui

Wang

Yao

Zhang

Xie

Cui

and Zhang

, Circulating MicroRNAs in cancer: potential and challenge, Front Genet 10 (2019), 626.

19.

Kroh

E.M.

Parkin

R.K.

Mitchell

P.S.

and Tewari

, Analysis of circulating microRNA biomarkers in plasma and serum using quantitative reverse transcription-PCR (qRT-PCR), Methods 50 (2010), 298–301.

20.

McDonald

J.S.

Milosevic

Reddi

H.V.

Grebe

S.K.

and Algeciras-Schimnich

, Analysis of circulating microRNA: preanalytical and analytical challenges, Clin Chem 57 (2011), 833–840.

21.

Sellars

M.J.

Vuocolo

Leeton

L.A.

Coman

G.J.

Degnan

B.M.

and Preston

N.P.

, Real-time RT-PCR quantification of Kuruma shrimp transcripts: a comparison of relative and absolute quantification procedures, J Biotechnol 129 (2007), 391–399.

22.

Zhou

Wang

Huang

Wang

Zhu

Shu

and Liu

, Prognostic value of miR-21 in various cancers: an updating meta-analysis, PLoS One 9 (2014), e102413.

23.

Zhang

Wang

Guo

and Cui

, Serum miRNAs panel (miR-16-2*, miR-195, miR-2861, miR-497) as novel non-invasive biomarkers for detection of cervical cancer, Sci Rep 5 (2015), 17942.

24.

Zhang

Zhan

Yan

and Wang

, Circulating MicroRNA-21 is involved in lymph node metastasis in cervical cancer by targeting RASA1, Int J Gynecol Cancer 26 (2016), 810–816.

25.

Jensen

S.G.

Lamy

Rasmussen

M.H.

Ostenfeld

M.S.

Dyrskjot

Orntoft

T.F.

and Andersen

C.L.

, Evaluation of two commercial global miRNA expression profiling platforms for detection of less abundant miRNAs, BMC Genomics 12 (2011), 435.

26.

Deng

Zhang

Wen

Miao

and Guo

, MicroRNA-142-3p inhibits cell proliferation and invasion of cervical cancer cells by targeting FZD7, Tumour Biol 36 (2015), 8065–8073.

27.

Wang

and Yang

, A six-microRNA signature predicts survival of patients with uterine corpus endometrial carcinoma, Curr Probl Cancer 43 (2019), 167–176.

28.

Zhou

Y.X.

Zhao

Mao

L.W.

Wang

Y.L.

Xia

L.Q.

Cao

Shen

and Chen

, Long non-coding RNA NIFK-AS1 inhibits M2 polarization of macrophages in endometrial cancer through targeting miR-146a, Int J Biochem Cell Biol 104 (2018), 25–33.

29.

Song

Ding

Cui

Liang

and Han

, miR-146a promotes cervical cancer cell viability via targeting IRAK1 and TRAF6, Oncol Rep 39 (2018), 3015–3024.

30.

Daugaard

Sanders

K.J.

Idica

Vittayarukskul

Hamdorf

Krog

J.D.

Chow

Jury

Hansen

L.L.

Hager

Lamy

Choi

C.L.

Agalliu

Zisoulis

D.G.

and Pedersen

I.M.

, miR-151a induces partial EMT by regulating E-cadherin in NSCLC cells, Oncogenesis 6 (2017), e366.

31.

Baird

D.T.

Cameron

S.T.

Critchley

H.O.

Drudy

T.A.

Howe

Jones

R.L.

Lea

R.G.

and Kelly

R.W.

, Prostaglandins and menstruation, European journal of obstetrics, gynecology, and reproductive biology 70 (1996), 15–17.

32.

King

A.E.

Critchley

H.O.

and Kelly

R.W.

, The NF-kappaB pathway in human endometrium and first trimester decidua, Molecular human reproduction 7 (2001), 175–183.

33.

Kelly

R.W.

, Pregnancy maintenance and parturition: the role of prostaglandin in manipulating the immune and inflammatory response, Endocrine reviews 15 (1994), 684–706.

34.

Vaskivuo

T.E.

Stenbäck

and Tapanainen

J.S.

, Apoptosis and apoptosis-related factors Bcl-2, Bax, tumor necrosis factor-alpha, and NF-kappaB in human endometrial hyperplasia and carcinoma, Cancer 95 (2002), 1463–1471.

35.

Uotila

P.J.

Erkkola

R.U.

and Klemi

P.J.

, The expression of cyclooxygenase-1 and -2 in proliferative endometrium and endometrial adenocarcinoma, Annals of Medicine 34 (2002), 428–433.

36.

St-Germain

M.-E.

Gagnon

Parent

and Asselin

, Regulation of COX-2 protein expression by Akt in endometrial cancer cells is mediated through NF-kappaB/IkappaB pathway, Molecular cancer 3 (2004), 7.

37.

Wang

Zhang

and Yu

, Dual functional roles of the MyD88 signaling in colorectal cancer development, Biomedicine & Pharmacotherapy = Biomedecine & Pharmacotherapie 107 (2018), 177–184.

38.

Lin

Huang

Liu

Chen

X.M.

Y.M.

Huang

L.F.

and Wang

X.Z.

, Exosomes: novel biomarkers for clinical diagnosis, ScientificWorldJournal 2015 (2015), 657086.

39.

Zhou

Bai

Xie

and Lu

, miRNA in plasma exosome is stable under different storage conditions, Molecules 19 (2014), 1568–1575.

40.

Warnecke-Eberz

Chon

S.H.

Holscher

A.H.

Drebber

and Bollschweiler

, Exosomal onco-miRs from serum of patients with adenocarcinoma of the esophagus: comparison of miRNA profiles of exosomes and matching tumor, Tumour Biol 36 (2015), 4643–4653.

41.

Taylor

D.D.

and Gercel-Taylor

, MicroRNA signatures of tumor-derived exosomes as diagnostic biomarkers of ovarian cancer, Gynecol Oncol 110 (2008), 13–21.

Three plasma-based microRNAs as potent diagnostic biomarkers for endometrial cancer

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Materials and methods

2.1 Information of participants

2.2 Study design and sample preparation

2.4 Extraction of RNA

Table 1 Clinical characteristics of 93 EC patients and 79 normal controls

2.6 Statistical analysis

2.7 Bioinformatics analysis

3. Results

3.1 Characteristics of participants

Table 2 Endometrial cancer diagnostic capability of the signature and previously identified biomarkers (presented as the results of ROC curve analysis using the data from GEO dataset)

3.3 Evaluation of serum miRNAs by qRT-PCR

3.4 Evaluation of diagnostic value

3.7 Comparison with previously identified signatures

3.8 Expression of identified miRNAs in tissue and plasma exosomes samples

4. Discussion

5. Conclusion

Experimental subject statement

Authors’ contributions

Supplementary data

sj-docx-1-cbm-10.3233_CBM-200972.docx - Supplemental material

Footnotes

Acknowledgments

References

Table 1
Clinical characteristics of 93 EC patients and 79 normal controls

Table 2
Endometrial cancer diagnostic capability of the signature and previously identified biomarkers (presented as the results of ROC curve analysis using the data from GEO dataset)