Serum microRNAs as non-invasive diagnostic biomarkers for intrahepatic cholestasis of pregnancy

Abstract

Objectives: Intrahepatic cholestasis of pregnancy (IHCP) causes itching, preterm birth, and stillbirth. However, there is no accurate diagnostic method for IHCP. Currently, circulating microRNAs (miRNAs) have become candidate biomarkers for the diagnosis of multiple diseases. Here, we investigated the diagnostic value of miRNAs in IHCP and aimed to predict the molecular mechanism of IHCP pathogenesis. Methods: We analyzed differentially expressed miRNAs in both women with IHCP and normal pregnant women. The selected candidate miRNAs were validated in 46 IHCP cases and 46 normal pregnant subjects, and we constructed receiver operator characteristic curves of miRNAs. Pearson correlations between levels of total bile acid (TBA) and differentially expressed miRNAs were also calculated. In addition, we clustered functionally significant biological pathways using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses. Results: The expression levels of 13 miRNAs were remarkably upregulated while the other 35 miRNAs were significantly downregulated, in women with IHCP (P≤0.05) when compared with healthy pregnant women. The areas under the curves of miRNA-7706, miRNA-877-3p, and miRNA-128-3p were higher than 0.90, indicating more reliable diagnosis of IHCP. The Pearson analysis showed that the levels of these miRNAs were positively correlated to TBA level. Additionally, the results of bioinformatics analysis revealed that the differentially expressed miRNAs mainly influenced fatty acid biosynthesis, the endoplasmic reticulum ubiquitin ligase complex, and the p53, and mammalian target of rapamycin (mTOR) and AMP-activated protein kinase (AMPK) signaling pathways. Conclusion: The panel of three-miRNAs (miRNA-7706, miRNA-877-3p, and miRNA-128-3p) may be a useful noninvasive diagnostic biomarker of IHCP.

Introduction

Intrahepatic cholestasis of pregnancy (IHCP), a pregnancy-specific liver abnormality that occurs most often during the third trimester, is characterized by pruritus and high level of liver total bile acid (TBA) [1]. The incidence of IHCP ranges from 1%-15% globally [2]. The maternal prognosis is usually benign, but there is a clear association between IHCP and the elevated risk of adverse perinatal outcomes such as preterm birth, fetal distress, and stillbirth [3,4]. Although genetic, endocrinological, nutritional, and environmental factors have been implicated in IHCP, the etiology and diagnostic criteria of this condition are yet to be fully understood [5].

An accurate diagnosis plays an essential role in the treatment of IHCP. Currently, the diagnosis of IHCP is based on elevated serum TBA levels and symptoms of pruritus and jaundice [6]. Therefore, TBA is often used as a diagnostic marker for IHCP. However, its levels can fluctuate with fasting state or gestational age [7,8], new prognostic and diagnostic biomarkers for IHCP are still urgently needed.

MicroRNAs (miRNAs) are a large family of small, noncoding RNAs of about 21-22 nucleotides and are regulators of essential biological processes [9], including cell apoptosis, proliferation, inflammation, and metabolism [10-12]. miRNAs are known to control approximately 30% of human genes that encode proteins [13] and inhibit target gene expression mainly through translational inhibition or degradation of messenger RNAs (mRNAs) [14]. Previous studies have shown that changes in the expression of circulating miRNAs are associated with numerous diseases, such as hepatic carcinoma [15], stomach cancer [16], lung cancer [17], cholestasis [18], and alcoholic liver disease [19]. Currently, miRNAs in peripheral blood have been utilized as biomarkers for the diagnoses of acute and chronic liver disorders [20-22]. miRNAs have also been shown to reflect the physiological state of pregnancy and can be used as prognostic and diagnostic biomarkers to indicate complications in pregnancy such as preeclampsia and gestational diabetes mellitus [23,24]. Thus, miRNAs may be further developed as important diagnostic indicators of IHCP.

In this study, we measured the expression levels of various miRNAs in patients with IHCP and healthy pregnant women and found miRNAs that were either remarkably upregulated or significantly downregulated. We have identified a panel of miRNAs that may be a useful noninvasive diagnostic biomarker for IHCP. The differentially expressed miRNAs identified by bioinformatics analysis have been found to be associated with fatty acid biosynthesis, the endoplasmic reticulum ubiquitin ligase complex, p53, mammalian target of rapamycin (mTOR), and the AMP-activated protein kinase (AMPK) signaling pathways, which will further our understanding and thus, increase the diagnostic and therapeutic value of such miRNAs for IHCP.

Materials and methods

Sample collection: patients and healthy controls

We conducted a case-control study to measure maternal peripheral blood miRNA levels indicative of an IHCP diagnosis. Peripheral blood specimens were drawn from women with IHCP (n=46) and control subjects (n=46) by researchers in the Department of Obstetrics and Gynecology, Tongji Hospital, Huazhong University of Science and Technology. All participants were Chinese women with singleton pregnancies. This research was approved by the Human Ethics Committee of the same institute (No. TJ-IRB20210520). Informed consent was obtained from all participants.

The inclusion criteria for IHCP were [25,26]: (1) pruritus and jaundice; (2) elevated levels of aminotransferases, bilirubin, and TBA; and (3) normalization of cholestasis after delivery. The exclusion criteria were [27]: (1) presence of skin diseases or chronic liver and gall bladder diseases; (2) abnormal liver and kidney function, gestational diabetes, pregnancy-induced hypertension syndrome, or hematological diseases; and (3) pregnancy with preeclampsia, uterine fibroids, abnormal uterine development, placental adhesion or placenta accreta, placenta previa, or placental abruption. Only healthy pregnant women with normal coagulation and cognitive function were enrolled in the control group [26]. Peripheral blood specimens were collected in PAXgene tubes (PreAnalytiX GmbH, Feldbachstrasse, Hombrechtikon, Switzerland) for testing. Blood specimens were kept overnight at 4°C and stored until further analysis at -80°C.

RNA extraction

Peripheral blood specimens were thawed at 25°C for 2 h. Then, total RNA and miRNAs were isolated and purified using the PAXgene Blood miRNA kit (PreAnalytiX GmbH), following the kit’s protocols. Quantification of purified miRNAs was conducted using the miScript^® PCR system (Takara, Dalian, China). The yields and integrity of total RNA were examined using the Agilent High Sensitivity Reagents kit (Agilent Technologies, Santa Clara, CA, USA), employing the Agilent 2100 Bioanalyzer.

MiRNA expression profiling

The Qubit™ dsDNA HS Assay kit (Invitrogen, Carlsbad, California, USA, Q32854) was used for cDNA synthesis. MiRNA expression profiling was conducted using the Agilent miRNA array, designed to detect 2,705 miRNA sequences. Data normalization and quality control were performed using data collection software (Illumina). Data preprocessing was based on the default 90^th percentile-normalization method. Benjamini-Hochberg-adjusted P values of 0.05 and threshold values of less than or equal to 2 and greater than 2-fold were used to identify differentially expressed miRNAs.

Quantitative, real-time PCR (qRT-PCR) assays

Nine candidate miRNAs (miRNA-7706, miRNA-877-3p, miRNA-128-3p, miRNA-1306-5p, miRNA-30c-5p, miRNA-3613-5p, miRNA-379-5p, miRNA-4772-5p, and miRNA-204-5p) were selected for miRNA expression profiling, and U6 was used as an internal control. The sequences of these miRNAs were chosen according to the miRNA sequences retrieved from the miRbase database (Table 1).

Table 1.

Sequences of miRNAs based on the miRNA sequences obtained from the miRBase database

miRNA	Species	miRBase ID	Sequence (5’ to 3’)
miR-7706	Human	MI0025242	UGAAGCGCCUGUGCUCUGCCGAGA
miR-877-3p	Human	MI0005561	UCCUCUUCUCCCUCCUCCCAG
miR-1306-5p	Human	MI0006443	CCACCUCCCCUGCAAACGUCCA
miR-128-3p	Human	MI0000447	UCACAGUGAACCGGUCUCUUU
miR-30c-5p	Human	MI0000736	UGUAAACAUCCUACACUCUCAGC
miR-3613-5p	Human	MI0016003	UGUUGUACUUUUUUUUUUGUUC
miR-379-5p	Human	MI0000787	UGGUAGACUAUGGAACGUAGG
miR-4772-5p	Human	MI0017414	UGAUCAGGCAAAAUUGCAGACU
miR-204-5p	Human	MI0000284	UUCCCUUUGUCAUCCUAUGCCU

To confirm the candidate miRNAs detected by the microarrays, SYBR Green-based qRT-PCR was employed to analyze the relative miRNA expression. The RNA template was reverse transcribed, and the cDNA product was used for qRT-PCR. Using an miRNA-X-miRNA-qRT-PCR-TB-Green kit (Takara, Dalian, China), miRNA was polyadenylated using poly(A) polymerase and subsequently reverse transcribed into cDNA. Next, qRT-PCR analysis of the target genes was conducted using SYBR Green (Takara), following the kit protocol. PCR was initiated using a denaturing step at 37°C for 1 h and 87°C for 5 min, followed by 40 cycles of annealing at 95°C for 5 s and at 60°C for 20 s.

Functional and pathway analyses

The Gene Ontology (GO) database (http://www.geneontology.org/) was used to analyze the molecular functions of the differentially expressed miRNA genes (P≤0.05). Meanwhile, the relevant signaling pathway of each target gene was analyzed using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (http://www.kegg.jp/). The Fisher’s exact test was applied to determine the P value for each pathway. A P value of less than or equal to 0.05 was deemed statistically significant.

Statistical analysis

Statistical tests were conducted with a one-way analysis of variance followed by Tukey’s test using GraphPad Prism 8. An independent sample t-test was employed to compare the characteristics of healthy women and those with IHCP in this study. We constructed receiver operator characteristic (ROC) curves of miRNAs from peripheral blood (with 95% confidence intervals) to assess the diagnostic value of biomarkers of IHCP. Pearson correlations between serum TBA levels and differentially expressed miRNA levels were also calculated. All values are shown as the mean ± standard deviation. A P value of less than or equal to 0.05 was deemed statistically significant.

Results

Clinical features of women with IHCP and normal pregnant women

Table 2 summarizes the clinical features of women with IHCP and normal pregnant women. Maternal age and gestational week did not remarkably differ between the groups (P>0.05). Serum levels of TBA, direct bilirubin (DBIL), total bilirubin (TBIL), gamma-glutamyltransferase (GGT), alkaline phosphatase (ALP), alanine transaminase (ALT), and aspartate aminotransferase (AST) were significantly elevated in women with IHCP compared to normal controls (P<0.05).

Table 2.

Main clinical characteristics of patients with ICP and healthy pregnant women

Variable	Control (N=46)	ICP (N=46)	P value
Maternal age, years	31.91±3.21	29.83±3.41	0.254
Gestational week	38.09±2.17	37.98±2.87	0.213
TBA, μmol/L	3.02±2.19	38.34±34.20*	0.014
TBIL, μmol/L	4.68±2.03	10.72±8.58*	0.061
DBIL, μmol/L	2.17±0.83	8.22±8.16*	0.031
AST, IU/L	16.73±2.69	151.08±166.20*	0.024
ALT, IU/L	9.34±1.86	225.04±196.28*	0.006
ALP, IU/L	93.45±6.55	203.58±85.83*	0.003
GGT, IU/L	11.82±3.43	49.00±29.99*	0.003

TBA: total bile acid; TBIL: total bilirubin; DBIL: direct bilirubin; AST: aspartate aminotransferase; ALT: alanine transaminase; ALP: alkaline phosphatase; GGT: glutamyltransferase.

^*P<0.05 vs. control.

Differentially expressed miRNAs between women with IHCP and normal pregnant women

To determine differential miRNA expression in peripheral blood samples of women with IHCP and normal pregnant women, we employed miRNA array technology (Agilent) and determined the expression profiles of 2,705 mature human miRNAs. Of these, 13 miRNAs were remarkably upregulated (P≤0.05), including miRNA-7706, miRNA-877-3p, miRNA-128-3p, miRNA-1284, miRNA-1306-5p, miRNA-4479, miRNA-483-3p, miRNA-99a-5p, miRNA-561-5p, and miRNA-30c-5p, and 35 were downregulated (P≤0.05), including miRNA-3613-5p, miRNA-379-5p, miRNA-4772-5p, and miRNA-204-5p (Figure 1).

Figure 1.

Heatmap of miRNA expression in patients with IHCP compared to controls. The heatmap shows the expression of 48 miRNAs that demonstrated significant differences between patients with IHCP and the control group. Compared to the control group, patients with IHCP were clustered based on the upregulation and downregulation of 13 and 35 miRNAs, respectively (P≤0.05). Each column represents an individual sample; columns A1-A3 represent the controls, and columns B1-B3 represent patients with IHCP. Each row corresponds to different miRNAs. The color scale illustrates the miRNA levels: blue, low expression; red, high expression.

Verification of differential miRNA expression profiles for an IHCP diagnosis

To determine their potential as biomarkers in patients with IHCP, nine candidate miRNAs (five upregulated: miRNA-7706, miRNA-877-3p, miRNA-128-3p, miRNA-1306-5p, and miRNA-30c-5p; four downregulated: miRNA-3613-5p, miRNA-379-5p, miRNA-4772-5p, and miRNA-204-5p) were selected for further validation by qRT-PCR. U6 served as the normalization control. We found that all selected miRNAs could be detected in the collected samples, and the differentially expressed miRNAs between the IHCP and control groups were significantly different (P≤0.05). The nine differentially expressed miRNAs between the IHCP and control groups were shown in Figure 2. The results showed that miRNA-7706, miRNA-877-3p, miRNA-128-3p, miRNA-1306-5p, and miRNA-30c-5p were markedly upregulated (P≤0.05), whereas miRNA-3613-5p, miRNA-379-5p, miRNA-4772-5p, and miRNA-204-5p were significantly downregulated (P≤0.05) in women with IHCP compared to normal pregnant women.

Figure 2.

Differential expression of selected miRNAs in women with IHCP and healthy pregnant women. Expression of each miRNA was verified using qRT-PCR in 92 (46 control and 46 IHCP) peripheral blood specimens. *P<0.05 vs. control, **P<0.01 vs. control.

MiRNA-7706, miRNA-877-3p, and miRNA-128-3p could be used as biomarkers to predict the risk of IHCP

To assess the diagnostic values of differentially expressed miRNAs in patients with IHCP, we constructed ROC curves and analyzed the areas under the curves (AUCs) of miRNA-7706, miRNA-877-3p, miRNA-128-3p, miRNA-1306-5p, and miRNA-30c-5p. As shown in Figure 3, the AUCs were 1.00, 0.99, 0.97, 0.86, and 0.85 for miRNA-7706 (Figure 3A), miRNA-877-3p (Figure 3B), miRNA-128-3p (Figure 3C), miRNA-1306-5p (Figure 3D), and miRNA-30c-5p (Figure 3E), respectively. The AUCs of miRNA-7706, miRNA-877-3p, and miRNA-128-3p were higher than 0.90, indicating that these three miRNAs had sufficiently diagnostic value in predicting the risk of IHCP. The diagnostic values of miRNA-7706, miRNA-877-3p, and miRNA-128-3p at cut-off points of 1.00, 0.98, and 0.85, respectively, were assessed. The specificities of miRNA-7706, miRNA-877-3p, and miRNA-128-3p were 0.74, 0.77, and 0.76, respectively, while the sensitivities of miRNA-7706, miRNA-877-3p, and miRNA-128-3p were 0.74, 0.74, and 0.72, respectively. These results indicate that these miRNAs had sufficiently diagnostic value in predicting the risk of IHCP. Therefore, we explored the correlation of miRNA-7706, miRNA-877-3p, and miRNA-128-3p with clinical parameters and further investigated the potential value of these miRNAs in the pathogenesis of IHCP.

Figure 3.

ROC curves to explore the diagnostic values of selected individual miRNAs in patients with IHCP. A: miRNA-7706; B: miRNA-877-3p; C: miRNA-128-3p; D: miRNA-1306-5p; E: miRNA-30c-5p.

Correlation between miRNA-7706, miRNA-877-3p, and miRNA-128-3p and clinical parameters in women with IHCP

To examine the correlation between miRNAs and clinical parameters, we performed a correlation analysis between miRNA-7706, miRNA-877-3p, and miRNA-128-3p expression and TBA levels. The expression of these miRNAs positively correlated with TBA levels (r=0.20, P=0.17; r=0.83, P<0.001; and r=0.43, P=0.003 for miRNA-7706, miRNA-877-3p, and miRNA-128-3p, respectively) (Figure 4). Our findings suggest that these miRNAs may be potential biomarkers of an IHCP diagnosis.

Figure 4.

Correlations between peripheral blood miRNA levels and TBA levels in patients with IHCP. A: miRNA-7706; B: miRNA-877-3p; C: miRNA-128-3p.

Functional and pathway analyses

To further evaluate the functions of the identified miRNAs, we clustered functionally significant biological pathways using GO and KEGG pathway analyses. The GO enrichment terms associated with miRNAs in women with IHCP were determined. As shown in Figure 5A and 5B, the target genes were enriched in various biological processes such as “biological regulation”, “cellular component organization or biogenesis”, “cellular process”, “immune system process”, “localization”, “metabolic process”, and “positive regulation of gene silencing by miRNA”. The target genes also appeared responsible for the formation of cellular components, such as “ER ubiquitin ligase complex”, “cell junction”, “membrane”, “membrane-enclosed lumen”, and “X chromosome”. Several molecular functions such as “antioxidant activity”, “transporter activity”, “protein tag”, “structural molecule activity”, “extracellular ATP-gated cation channel activity”, “beta-galactoside alpha-2,3-sialyltransferase activity”, “GPI-linked ephrin receptor activity”, and “profilin binding” were also identified as possibly regulated by these miRNAs.

Figure 5.

Functional and pathway analyses of differentially expressed miRNAs in women with IHCP. A: GO enrichment classification of differentially expressed miRNAs in women with IHCP, including cellular components (green), molecular functions (blue), and biological processes (red); B: KEGG pathways of differentially expressed miRNAs in patients with IHCP; C and D: Top 30 significantly enriched GO terms and KEGG pathways, respectively.

The signaling pathways modulated by these miRNAs were also evaluated using the KEGG database. The differentially expressed miRNAs participated in fatty acid biosynthesis, endoplasmic reticulum ubiquitin ligase complex, p53 signaling pathway, and mTOR and AMPK signaling pathways (Figure 5C and 5D).

Discussion

In our study, we demonstrated that the expression levels of miRNA-7706, miRNA-877-3p, miRNA-128-3p, miRNA-1306-5p, and miRNA-30c-5p were increased in women with IHCP. Among them, strong correlations were found between levels of miRNA-7706, miRNA-877-3p, miRNA-128-3p, and TBA in women with IHCP, suggesting that these three miRNAs may be of certain clinical relevance as noninvasive biomarkers of IHCP.

IHCP is a pregnancy-specific disease with an incidence of 0.2%-5.6%, affecting 0.1%-2% of pregnant women [28,29]. IHCP may cause adverse perinatal outcomes, including fetal asphyxia and preterm labor [30,31]. Thus, an early and accurate diagnosis of IHCP is essential. However, the pathogenesis of IHCP and its relationship with the concentrations of biochemical markers remain unknown [32]. Currently, diagnosis of IHCP is based on symptoms of pruritus and elevated levels of TBA, ALT, ASL, and ALP levels. TBA is the most frequently used diagnostic marker for IHCP. However, TBA levels can fluctuate, depending on fasting state or gestational age [33]. Thus, it is important to identify more specific and sensitive diagnostic markers to enhance the diagnosis for women with IHCP.

MiRNAs play a vital role in post-transcriptional target gene regulation by targeting the 3’ untranslated region of mRNAs to reduce protein translation and/or increase mRNA degradation s [34,35]. Many studies have reported miRNAs as novel molecules that could be used as diagnostic, prognostic, and therapeutic biomarkers [36]. Currently, miRNA profiling studies for liver disease are gaining considerable attention [37-39]. A study of patients with primary biliary cholangitis (PBC) showed that miRNA-505-3p and miRNA-197-3p were differentially expressed compared to healthy controls [40]. Sakamoto et al. reported that the elevation of let-7 miRNA and miRNA-520a-5p can be used as biomarkers in patients with refractory PBC [38]. Fernandez-Ramos et al. showed that miRNA-873-5p expression significantly increases in patients with primary sclerosing cholangitis [39]. Zhao and co-workers found that miRNA-132, miRNA-222, and miRNA-29a expression was significantly decreased in women with gestational diabetes mellitus [41]. Therefore, after using miRNA arrays to obtain miRNA profiles of healthy women and those with IHCP, validation of selected miRNAs using qRT-PCR to identify miRNA biomarkers could potentially be used to diagnose IHCP.

In this study, we observed that 48 miRNAs were differentially expressed in women with IHCP compared with the control group (Figure 1). Among candidate miRNAs, miRNA-7706, miRNA-877-3p, miRNA-128-3p, miRNA-1306-5p, and miRNA-30c-5p were significantly increased in women with IHCP. Moreover, the AUCs of miRNA-7706, miRNA-877-3p, and miRNA-128-3p were 1.00, 0.99, and 0.97, respectively (Figure 3), and Pearson correlations revealed that these three miRNAs were strongly correlated with TBA level (Figure 4), indicating that the expression levels of these miRNAs might provide reliable diagnostic value for IHCP. However, the molecular mechanisms by which these miRNAs are differentially regulated during IHCP pathogenesis have not been fully elucidated. The possible contributing factors may include environmental, nutritional, endocrinological, and genetic factors. Most studies have found that estrogen levels in women with IHCP are significantly increased, which may be the main reason for its pathogenesis. Besides, estrogen plays a critical role in regulating (either suppressing or stimulating) miRNA expression [42]. There are reports that the estrogen receptor regulates miRNAs via binding its promoter in an estrogen-dependent manner, which include upregulation and downregulation of miRNAs, such as miRNA-17-92, miRNA-106a-363, and miRNA-125b [42,43]. Therefore, it is necessary to conduct further studies on the regulatory mechanism of differentially expressed miRNAs.

Results of a bioinformatics analysis revealed that differentially expressed miRNAs may be involved in localization, metabolism, the endoplasmic reticulum ubiquitin ligase complex, and the p53, mTOR, and AMPK signaling pathways in women with IHCP (Figure 5). Studies have shown that the p53, mTOR, and AMPK signaling pathways are involved in the pathogenesis of cholestasis [44-46], including bile acid homeostasis and the regulation of hepatic polarity, inflammation, oxidative stress, and fibrosis. Previous studies have reported that the activation of AMPK signaling disrupts bile acid homeostasis and promotes cholestasis [47]. Further, AMPK is responsible for the modulation of liver polarity, inflammation, and fibrosis in cholestasis and can affect the pathogenesis of cholestatic liver injury. Chen et al. revealed the promising role of p53 in modulating bile acid disposition-related genes via the enzymes and transporters responsible for bile acid metabolism [48]. Other studies reported that p53 can promote apoptosis during cholestatic liver injury [49,50]. These differentially expressed miRNAs may regulate the aforementioned signaling pathways and participate in the pathogenetic process of IHCP. Thus, our study provides novel insights into the pathophysiology of IHCP and potential targets for its treatment.

To summarize, we identified significantly altered and specifically regulated miRNAs in women with IHCP compared with healthy controls. Our results suggest that miRNA-7706, miRNA-877-3p, and miRNA-128-3p may serve as a panel of potential biomarkers for the diagnosis of IHCP. The functional roles of these three miRNAs remain unclear. Due to the limitation of in the present study (suah as small sample size), further studies are needed to corroborate our findings.

Appendix A. supplementary data

The original datas of miRNA expression profiling on the Supplementary File.

Disclosure of conflict of interest

None.