Expression Profile of the Chromosome 14 MicroRNA Cluster (C14MC) Ortholog in Equine Maternal Circulation throughout Pregnancy and Its Potential Implications

Equine chromosome 24 microRNA cluster (C24MC), the ortholog of human C14MC, is a pregnancy-related miRNA cluster. This cluster is believed to be implicated in embryonic, fetal, and placental development. The current study aimed to characterize the expression profile of this cluster in maternal circulation throughout equine gestation. The expression profile of miRNAs belonging to this cluster was analyzed in the serum of non-pregnant (diestrus), pregnant (25 d, 45 d, 4 mo, 6 mo, 10 mo), and postpartum mares. Among the miRNAs examined, 11 miRNAs were differentially expressed across the analyzed time-points. Four of these miRNAs (eca-miR-1247-3p, eca-miR-134-5p, eca-miR-382-5p, and eca-miR-433-3p) were found to be enriched in the serum of pregnant mares at Day 25 relative to non-pregnant mares. To further assess the accuracy of these miRNAs in differentiating pregnant (25 d) from non-pregnant mares, receiver operating characteristic (ROC) analysis was performed for each of these miRNAs, revealing that eca-miR-1247-3p and eca-miR-134-5p had the highest accuracy (AUCROC = 0.92 and 0.91, respectively; p < 0.05). Moreover, eca-miR-1247-3p, eca-miR-134-5p, eca-miR-409-3p, and eca-miR-379-5p were enriched in the serum of Day 45 pregnant mares. Among those miRNAs, eca-miR-1247-3p and eca-miR-409-3p retained the highest accuracy as shown by ROC analysis. GO analysis revealed that these miRNAs are mainly implicated in nervous system development as well as organ development. Using in situ hybridization, we localized eca-miR-409-3p in the developing embryo (25 d) and extra-embryonic membranes (25 and 45 d). In conclusion, the present study is the first to elucidate the circulating maternal profile of C24MC-associated miRNAs throughout pregnancy and to suggest that serum eca-miR-1247-3p, eca-miR-134-5p, and eca-miR-409-3p could be used as pregnancy-specific markers during early gestation (25 and 45 d). Overall, the high abundance of these embryo-derived miRNAs in the maternal circulation suggests an embryo-maternal communication during the equine early pregnancy.


Introduction
MicroRNAs (miRNAs) are a class of small non-coding RNA that post-transcriptionally regulate protein-coding mRNAs [1]. MiRNAs are expressed in a variety of cell types as biological regulators. They play a fundamental role in the regulation of a variety of developmental and physiological processes, such as cell proliferation, growth, metabolism, communication, apoptosis, and death [2][3][4]. Aberrant expression of miRNAs has been associated with many pathological disorders and diseases, including cancer and abnormalities during pregnancy [5][6][7][8]. Occasionally throughout the genome, a group of two or more miRNAs are transcribed from physically adjacent miRNA genes (within 10 kb), which form a cluster [9,10]. These miRNA clusters have been found to be exclusively or preferentially expressed in a tissue-specific manner [11]. This tissue expression is at least partly reflected in the circulation (i.e., circulating miRNAs) and could be detected in plasma and serum with much greater sensitivity than protein markers [12,13]. The measurability, sensitivity, and stability (approximately five day half-life) of circulating miRNAs [13,14] empowered them to emerge as biomarkers for several physiological and pathological conditions, including diabetes [15], cancer [16,17], cardiovascular diseases [18,19], and skeletal diseases [20].
We hypothesized that equine C24MC-associated miRNAs, the ortholog of human C14MC, will have a differential expression pattern in maternal circulation throughout gestation. Moreover, identifying the normal profile of these miRNAs in maternal circulation in comparison to non-pregnant mares and postpartum mares could positively impact the development of novel and non-invasive biomarkers for pregnancy diagnosis and prediction of pregnancy outcomes and/or complications. Therefore, this study was designed to evaluate the expression profile of equine C24MC-associated miRNAs in the serum of non-pregnant (diestrus), pregnant (25 days, 45 days, 4 months, 6 months, 10 months) and postpartum mares.

Equine C24MC Expression in Serum
To determine the expression pattern of C24MC-associated miRNAs in maternal serum, the expression of candidate miRNAs was evaluated in samples from pregnant (25 d, 45 d, 4 mo, 6 mo, and 10 mo), postpartum and non-pregnant (diestrous) mares. Out of the 26 tested miRNAs, 13 of them showed a persistent expression at all time-points (expressed in all the samples from at least one time points). The remaining 13 miRNAs were not detected in all the samples therefore were excluded from the study. Regarding the abundance of tested miRNAs in equine circulation, we found eca-miR-433-3p, eca-miR-1247-3p, eca-miR-134-5p, and eca-miR-411-5p with the greatest abundance in serum (p < 0.05). On the other hand, eca-miR-412-5p and eca-miR-379-5p showed the lowest abundance (p < 0.05) (Figure 1). Out of the tested miRNAs, 11 miRNAs were differentially expressed across the analyzed time-points (p < 0.05), whereas the expression profile of eca-miR-412-5p and eca-miR-432-5p did not show any significant fluctuation. The differential expression profile of tested miRNAs is depicted in Figures 2 and 3. The correlation between tested miRNAs expression profiles is illustrated in Table 1.

Gene Ontology Analysis for the Differnetially Expressed miRNAs
Computational target prediction for miRNAs was performed to identify putative mRNA targets. A total of 428 target mRNAs were predicted using mirDB.org and IPA for the differentially expressed miRNAs (eca-miR-134-5p, eca-miR-1247-3p, eca-miR-382-5p, eca-miR-409-3p, eca-miR-433-3p, and eca-miR-379-5p). GO enrichment (biological process) analysis and pathway analysis were carried out. GO analysis revealed that these miRNAs are mainly implicated in nervous system development ( Figure 5A). Pathway analysis of the same set of miRNAs revealed that they are associated with the Wnt signaling pathway and oxytocin receptor mediated signaling pathway, among others ( Figure 5B).

Localization of a Member of Equine C24MC in Equine Extra-Embryonic Membranes and 25 d Embryos
In order to study the cellular and subcellular localization of eca-miR-409-3p (a member of the C24MC), we performed miRNA in situ hybridization (ISH) on 25 d embryos along with the extra-embryonic membranes of the 25 and 45 d GA. Since eca-miR-409-3p had a differential expression pattern during pregnancy and presented a 100% homology to hsa-miR-409-3p, this miRNA was chosen as the best candidate for ISH analysis. At 25 d, eca-miR-409-3p was intensely expressed in extra-embryonic membranes ( Figure 6). In 25 d embryos, the expression of this miRNA was extensive in several developing tissues from the three germ layers ( Figure 6). Moreover, extra-embryonic membranes collected at 25 and 45 d shows strong cytoplasmic signal for eca-miR-409-3p specifically in trophoblastic cells (Figure 7).

Discussion
To the best of our knowledge, this is the first study elucidating the expression profile of equine C24MC-associated miRNAs in peripheral circulation of non-pregnant (diestrus), pregnant (25 d, 45 d, 4 mo, 6m, and 10 mo), and postpartum mares. In the present study, four equine miRNAs (eca-miR-1247-3p, eca-miR-134-5p, eca-miR-382-5p, and eca-miR-433-3p) were found to be enriched in the serum of Day 25 pregnant mares in comparison to non-pregnant mares. In support of our findings, it has been reported that serum miR-433-3p expression is upregulated in Day 19 and 24 pregnant cows [34] and during the first trimester of pregnancy in women [54]. Moreover, it has been reported that serum hsa-miR-1247-3p expression is reduced in women with ectopic pregnancy [37] and preeclampsia [39] in comparison to normal pregnant women. To further assess the accuracy of eca-miR-1247-3p, eca-miR-134-5p, eca-miR-382-5p, and eca-miR-433-3p in differentiating pregnant (25 d) from non-pregnant mares, ROC analysis was performed and indicated that eca-miR-1247-3p followed by eca-miR-134-5p retain the highest accuracy. It is worth noting that miR-134-5p is highly expressed during embryonic development and becomes restricted to the brain postnatally [55,56]. Moreover, it has been reported that miR-134-5p is uniquely correlated in maternal and fetal circulation in women [57]. These facts might suggest that the serum enrichment with eca-miR-134-5p during Days 25 and 45 of equine pregnancy is of embryonic origin. Altogether, our findings suggest that eca-miR-1247-3p and eca-miR-134-5p could be potential biomarkers for pregnancy establishment and/or early embryonic loss at 25 days of equine pregnancy. In equine, it has been suggested that changes in circulating miRNAs could be detected as early as Days 9-13 of pregnancy [58]. However, eca-miR-1247-3p and eca-miR-134-5p profiles were not investigated in the previously mentioned study. Therefore, further studies are

Discussion
To the best of our knowledge, this is the first study elucidating the expression profile of equine C24MC-associated miRNAs in peripheral circulation of non-pregnant (diestrus), pregnant (25 d, 45 d, 4 mo, 6m, and 10 mo), and postpartum mares. In the present study, four equine miRNAs (eca-miR-1247-3p, eca-miR-134-5p, eca-miR-382-5p, and eca-miR-433-3p) were found to be enriched in the serum of Day 25 pregnant mares in comparison to non-pregnant mares. In support of our findings, it has been reported that serum miR-433-3p expression is upregulated in Day 19 and 24 pregnant cows [34] and during the first trimester of pregnancy in women [43]. Moreover, it has been reported that serum hsa-miR-1247-3p expression is reduced in women with ectopic pregnancy [37] and preeclampsia [39] in comparison to normal pregnant women. To further assess the accuracy of eca-miR-1247-3p, eca-miR-134-5p, eca-miR-382-5p, and eca-miR-433-3p in differentiating pregnant (25 d) from non-pregnant mares, ROC analysis was performed and indicated that eca-miR-1247-3p followed by eca-miR-134-5p retain the highest accuracy. It is worth noting that miR-134-5p is highly expressed during embryonic development and becomes restricted to the brain postnatally [44,45]. Moreover, it has been reported that miR-134-5p is uniquely correlated in maternal and fetal circulation in women [46]. These facts might suggest that the serum enrichment with eca-miR-134-5p during Days 25 and 45 of equine pregnancy is of embryonic origin. Altogether, our findings suggest that eca-miR-1247-3p and eca-miR-134-5p could be potential biomarkers for pregnancy establishment and/or early embryonic loss at 25 days of equine pregnancy. In equine, it has been suggested that changes in circulating miRNAs could be detected as early as Days 9-13 of pregnancy [47]. However, eca-miR-1247-3p and eca-miR-134-5p profiles were not investigated in the previously mentioned study. Therefore, further studies are needed to elucidate the expression profile of eca-miR-1247-3p and eca-miR-134-5p during earlier stages of equine pregnancy.
The serum enrichment with eca-miR-1247-3p, eca-miR-134-5p, eca-miR-382-5p, and eca-miR-433-3p at 25 d pregnancy as well as eca-miR-1247-3p, eca-miR-134-5p, eca-miR-409-3p, and eca-miR-379-5p at 45 d pregnancy suggest that these miRNAs are involved in early pregnancy events. In the same line, we found a similar pattern in the expression profile of this cluster of miRNAs in equine chorioallantoic membranes [29]. Moreover, a high expression of C24MC at 45 d was followed by a gradual decline toward the end of the gestation. To gain further insight into the biology of these miRNAs' targets, GO enrichment (biological process) analysis and pathway analysis were carried out. GO analysis revealed that these miRNAs are mainly implicated in nervous system development as well as organ development, cell-cell signaling, and Wnt signaling pathway, among others. In support of our outcomes, it has been reported that the expression of both miR-1247-3p and miR-409-3p is significantly downregulated in embryonic neuroblastoma [48,49]. In line with this, ISH analysis of eca-miR-409-3p revealed extensive expression in various tissues of the developing embryo including the central nervous system, among others. Moreover, pathway analysis of the same set of miRNAs revealed that they are associated with Wnt signaling pathway and oxytocin receptor mediated signaling pathway, among others. Interestingly, it has been reported that oxytocin responsiveness is altered during early equine pregnancy, and reduced expression of the oxytocin receptor (OXTR) seems to be regulated at the posttranscriptional level rather than the transcriptional level [50]. Our prediction suggests that these miRNAs are potentially involved in oxytocin receptor mediated signaling pathway during early pregnancy in the mare.
In conclusion, the present study is the first to elucidate the maternal circulation profile of C24MC-associated miRNAs throughout equine pregnancy in comparison to non-pregnant (diestrus) and postpartum mares. Moreover, this study introduced serum eca-miR-1247-3p, eca-miR-134-5p, and eca-miR409-3p as possible non-invasive biomarkers for early pregnancy (25 and 45 d) in mare. We further localized one of these miRNAs (eca-miR-409-3p) in the equine 25 d embryo, demonstrating the expression of it in the central nervous system and developing organs. Overall, the high abundance of these embryo-derived miRNAs in the maternal circulation suggests an embryo-maternal communication during the equine early pregnancy.

Animal Use
All animal procedures were prospectively approved by and completed in accordance with the Institutional Animal Care and Use Committee of the University of Kentucky (Approval No. #2014-1341, Date: 15 January 2015). All horses used in this study were mixed-breed, ranging between 250 and 550 kg. Mares were housed on pasture with ad libitum grass hay. Gestational age (GA) was determined based on the day of ovulation (Day 0).

MiRNA Extraction, cDNA Synthesis, and RT-qPCR
MiRNAs were extracted from serum samples (400 µL) using the miRNeasy Micro Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions with the following modification: TRIzol™ LS (Life technologies, Carlsbad, CA, USA) was used as the lysis reagent; 200 µL of chloroform was used instead of 140 µL; and the final elution step was performed using 20 µL of RNase-free water instead of 14 µL. Complementary DNA synthesis was carried out using the miScript II RT Kit (Qiagen). Three µL of cDNA product from each sample were combined to make the pooled serum samples. Pooled samples were only used as RT-qPCR inter-plate controls. The expression of mature miRNAs was determined by RT-qPCR using the miScript SYBR Green PCR kits (Qiagen, Hilden, Germany) with the miScript Universal Primer along with miRNA-specific primers according to the manufacturer's guidelines. Primers were designed using miRprimer software (version 2.0) for the candidate miRNAs [51].
The expression of the C24MC-associated miRNAs (n = 26), which were previously tested in the equine CAM throughout gestation [29], was evaluated in the collected serum samples. The list of miRNA candidates and primer sequences is provided in Table 2. The primers' efficiency was verified on the pooled samples. Primers with CT values <35 that did not produce primer-dimers were used for further experimentation. Otherwise, primers were re-designed and re-tested. The primers that did not yield amplified PCR products in all the samples from at least one of the time-points were excluded from further analysis. Eca-miR-10a (gcagtaccctgtagatccga), eca-miR-21 (gcagtagcttatcagactgatg), and eca-Let-7a (gcagtgaggtagtaggttg) were used as reference genes for serum [52]. A DNA melting curve was generated to discriminate between specific and non-specific amplification products. Real-Time qPCR was performed in triplicate for all samples [53,54]. PCR efficiencies were calculated using LinRegPCR (version 2012.0; http://www.hartfaalcentrum.nl) to ensure that all primers resulted in PCR efficiencies between 1.8 and 2.1.

Target Prediction for the miRNA Which Were Highly Expressed during Early Pregnancy (25 and 45 d)
Predicted targets for the six highly expressed miRNAs at 25 and 45 d were selected from IPA (only the Experimentally Observed or High Predicted) and miRDB.org (Target Score >80). Next, to predict functions of C24MC-associated miRNAs, the biological functions and physiological pathways of the target mRNA were analyzed by DAVID and Protein ANalysis THrough Evolutionary Relationships Classification System (PANTHER; Release 13.1) ontology classification system, respectively [55,56].

Data Analysis
Delta CT (∆CT) values for serum samples were calculated where ∆CT = (the CT values of the miRNA of interest-the CT values of all three reference miRNAs [geometric mean]) [57,58]. Results are presented as −∆CT (negative ∆CT is more intuitive than ∆CT). The expressions of C24MC-associated miRNAs were compared across GA, postpartum and with samples from non-pregnant mares using one-way analysis of variance (one-way ANOVA) followed by a pairwise comparison of means using T-test. Significance was set at p < 0.05. Pearson's correlation was performed to analyze the relationship between the expression profile of all the tested miRNAs. Receiver operating characteristic (ROC) curve analysis was carried out to assess the accuracy of eca-miR-1247-3p, eca-miR-134-5p, eca-miR-382-5p, and eca-miR-433-3p (differentially expressed miRNAs at 25 d) in differentiating pregnant (25 d) from non-pregnant (diestrus) mares. Moreover, ROC curve analysis was carried out to assess the accuracy of eca-miR-1247-3p, eca-miR-134-5p, eca-miR-409-3p, and eca-miR-379-5p (differentially expressed miRNAs at 45 d) in distinguishing pregnant (45 d) from non-pregnant (diestrus) mares. The area under the ROC curve (AUC ROC ) was estimated and used to evaluate the accuracy of the tested miRNAs [59,60]. In brief, the AUC ROC was inferred as noninformative (AUC ROC ≤ 0.5), poorly accurate (0.5 < AUC ROC ≤ 0.7), moderately accurate (0.7 < AUC ROC ≤ 0.9), highly accurate (0.9 < AUC ROC < 1), or perfect (AUC ROC = 1) [61]. The AUC ROC for each parameter was compared with the expected value (AUC ROC = 0.5) under the null hypothesis of a noninformative test.