Analyses of bovine luteal fractions obtained by FACS reveals enrichment of miR-183-96-182 cluster miRNAs in endothelial cells

The corpus luteum (CL) is a transitory structure that forms during each estrous/menstrual cycle and plays a critical role in the establishment and maintenance of pregnancy as well as in regulation of cyclic reproductive activity in the non-pregnant female. Development of the CL from follicular remnants after ovulation involves complex morphological and functional changes at the cell and tissue levels, including follicular cell differentiation, tissue remodelling and rapid development of vascular elements, with blood vessels making up to 30% of the total volume of the bovine CL [1]. The CL is a highly heterogeneous organ containing steroidogenic cells (large and small luteal, derived from follicular granulosa and theca cells, respectively) and non-steroidogenic cells (endothelial, pericytes, immune cells, fibroblasts); of these, endothelial cells make up about 85% of total luteal cells during the early luteal phase, a proportion that decreases to 50% as the corpus luteum matures [2‐4].

Limited information is available on the molecular regulation of luteal development. miRNAs have been shown to be involved in follicular and luteal functions [5, 6]. Two miRNA clusters that are involved are miR-183-96-182 and miR-212-132 [7‐9]. In a recent study, we showed that these two clusters are dramatically upregulated during the follicle-luteal transition in cattle, and we further showed that one miRNA from each cluster, namely miR-96 and, to a lesser extent, miR-132, play a key role in regulating survival and steroid production of luteal cells of both cattle and humans [9]. Further, in situ hybridisation (ISH) analyses revealed that miR-132 is expressed in different cell compartments within the bovine CL. However, we did not succeed in detecting miR-96 using this approach, possibly due to its relatively lower expression levels in the CL compared to miR-132. Establishing the relative expression of the miR-183-96-182 cluster in different luteal cell fractions, particularly steroidogenic and endothelial, would further our understanding of the role of these miRNAs in luteal development, warranting the use of approaches alternative to ISH for this purpose.

Previous studies documented the isolation of luteal endothelial cells using a variety of approaches such as density gradient centrifugation, fluorescence activated cell sorting (FACS), and lectins or antibodies ligated to magnetic beads, with variable success [10‐13]. Moreover, following pioneering studies by Hansel’s group [14], separation of luteal steroidogenic cell populations based on cell size/density has been extensively reported using methods including unit gravity sedimentation, elutriation, flow cytometry or a combination of these [3, 15]. More recently, Quirk et al, (2013) used the lipophilic dye Nile Red [16] to distinguish steroidogenic from non-steroidogenic luteal cells by flow cytometry [17]. Following from these findings, in this study we set out to optimise a procedure for simultaneous isolation of endothelial and steroidogenic cells from bovine CL using FACS. Using this method, we showed that the miR-183-96-182 cluster is enriched in endothelial relative to steroidogenic cells of the bovine CL.

Early corpora lutea (days 1 to 4 of the oestrous cycle [18]) collected from three heifers at an abattoir were processed as described before [9] to obtain cell digests.

By using immunocytochemistry we confirmed that CD144 antibody and Nile Red selectively stained endothelial and steroidogenic cells, respectively, in luteal extracts (Fig. 1 a, b). The CD144 antibody used had been previously validated in bovine luteal endothelial cells by Pate’s group [13]. Accordingly, CD144 stained distinct capillary-like structures (Fig. 1a, left and middle panels), whereas Nile Red stained primarily large, round cells, indicative of steroidogenic cells, in the form of numerous bright dots indicating lipid globules (Fig. 1b left panel), in agreement with previous results [17].

Cells stained with CD144 antibody and Nile Red were analyzed using flow cytometry and single events were selected to exclude debris and small cell clumps from the samples (Fig. 2A a,b). Live (DAPI-negative) CD144+ cells were then collected as endothelial fraction (Fig. 2A c1). Live CD144- cells (Fig. 2A c2) were further analyzed to select a cell fraction with high Nile Red fluorescent signal (Nile Red^Hi; Fig. 2A d). As expected, control cells stained with both primary and secondary antibodies showed positive staining for CD144 (Fig. 2B a) but no shift in the signal for Nile Red (Fig. 2B b). Conversely, in cells stained with Nile Red, a positive signal was observed for Nile Red (Fig. 2C b) but not for CD144 (Fig. 2C a).

Expression analyses of the endothelial and steroidogenic markers, CD144 and HSD3B1, were performed using qPCR to confirm the identity of the sorted cell fractions. Each transcript was expressed predominantly in CD144+ and Nile Red^Hi fractions, respectively, as expected (Fig. 3a). Moreover, miR-126, a miRNA known to be highly expressed in endothelia [19], was expressed at much higher levels in CD144+ than in Nile Red^Hi cells, further indicating the efficacy of our sorting procedure. Of note, no attempt was made to determine the relative proportion of large and small luteal cells in the sorted Nile Red^Hi fraction; this was not possible due to the limited number of cells obtained by FACS and available for downstream analyses.

Lastly, using this procedure we set out to determine the relative expression of the miR-212-132 and miR-183-96-182 clusters in the sorted steroidogenic and endothelial cell fractions. Our results showed that miR-132 and miR-212 were expressed at similar levels in CD144+ and Nile Red^Hi fractions (Fig. 3b). Consistent with this, our previous in situ hybridization analyses showed non-specific expression of miR-132 in the bovine corpus luteum [9]. Moreover, the miR-212-132 cluster is known to be widely expressed across body tissues and cell types including ovarian steroidogenic cells [7, 9], and vascular cells [20]. Although, miR-212-132 may be expressed by both endothelial and steroidogenic cells, their relative expression in large and small luteal cells types and in different vascular cell components (e.g., endothelia and pericytes) was not determined in this study and should be investigated in the future. We also quantified the expression of two of the miRNAs in the second cluster, miR-96 and miR-183, but not miR-182 as in our experience this miRNA is barely detectable in the bovine ovary [9]. We found that miR-96 and miR-183 were expressed in both cell fractions. Moreover, the two miRNAs were expressed at > 4-fold higher levels in CD144+ than in Nile Red^Hi cells (Fig. 3b). This result was unexpected in light of our previous evidence indicating that miR-96 plays important roles in survival and progesterone production by luteal steroidogenic cells in both cattle and humans [9]. Intriguingly, evidence of an involvement of the miR-183-96-182 cluster in regulation of endothelial cells is very scarce, however in light of our findings this should be investigated further, particularly in relation to the CL. Moreover, the observation that miR-96 was expressed at higher levels in endothelial than steroidogenic cells raises the interesting possibility that some of the reported effects on regulation of steroidogenic cells in the CL may actually be mediated by miR-96 produced by endothelial cells, a possibility that should be explored in the future.

In summary, we show that endothelial and steroidogenic cell fractions can be effectively isolated simultaneously from CL. In addition, our results further expand on previous knowledge on the role of miRNAs in luteal development, specifically by providing evidence that miR-212-132 and, particularly, miR-183-96-182 may be important in functionally regulating not only steroidogenic cells but also endothelial cells in the corpus luteum (CL).