Toms et al. Journal of Ovarian Research (2021) 14:54 https://doi.org/10.1186/s13048-021-00802-3 RESEARCH Open Access Small RNA sequencing reveals distinct nuclear microRNAs in pig granulosa cells during ovarian follicle growth Derek Toms1* , Bo Pan2, Yinshan Bai2,3 and Julang Li2 Abstract Nuclear small RNAs have emerged as an important subset of non-coding RNA species that are capable of regulating gene expression. A type of small RNA, microRNA (miRNA) have been shown to regulate development of the ovarian follicle via canonical targeting and translational repression. Little has been done to study these molecules at a subcellular level. Using cell fractionation and high throughput sequencing, we surveyed cytoplasmic and nuclear small RNA found in the granulosa cells of the pig ovarian antral preovulatory follicle. Bioinformatics analysis revealed a diverse network of small RNA that differ in their subcellular distribution and implied function. We identified predicted genomic DNA binding sites for nucleus-enriched miRNAs that may potentially be involved in transcriptional regulation. The small nucleolar RNA (snoRNA) SNORA73, known to be involved in steroid synthesis, was also found to be highly enriched in the cytoplasm, suggesting a role for snoRNA species in ovarian function. Taken together, these data provide an important resource to study the small RNAome in ovarian follicles and how they may impact fertility. Keywords: Microrna, smallRNA, snoRNA, Granulosa cells, Pig, Ovarian, Ovary, Follicle, Reproductive biology, Next generation sequencing, RNA-seq, Cellular fractionation, Nucleus, Cytoplasm Introduction species are transcribed from chromosomal DNA, and Increasing ovarian follicle size has long been an import- most are processed in both the nucleus and cytoplasm. ant indicator of oocyte quality, with oocytes obtained Accordingly, a complex system of transport exists to from large follicles showing consistently higher in vitro dynamically localize small RNA within the cell. For ex- and in vivo developmental potential than their counter- ample, after processing into mature miRNA by Dicer in parts from small follicles [3, 51]. In addition to increased the cytoplasm, strands loaded onto Ago2 may be subse- expression of key growth factors like GDF9 and BMP4, quently imported back into the nucleus via Importin 8 we and others have shown differences in microRNA [39]. The availability of complex secondary structures (miRNA) expression in the somatic cells of these follicles between RNA and DNA species permit these RNA to [44–46, 54, 55, 57]. Recent investigations have looked at function in a myriad of ways. In the nucleus, miRNA the distinct populations of small RNA that perform key can form triple helices with chromosomal DNA, to regu- biological functions in the cytoplasmic and nuclear com- late gene expression [48]. Even degraded fragments of partments of cells [19, 41, 47, 60]. All of these RNA miscoded messenger RNA have been shown to partici- pate in the assembly of such scaffolds [9]. Although * Correspondence: [email protected] small nucleolar RNAs (snoRNAs) are principally in- 1Department of Comparative Biology and Experimental Medicine, Faculty of volved with RNA modification in the nucleus, they are Veterinary Medicine, University of Calgary, Calgary, AB, Canada shuttled to the cytoplasm during periods of stress [20, Full list of author information is available at the end of the article © The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Toms et al. Journal of Ovarian Research (2021) 14:54 Page 2 of 12 31]. Despite evidence that such RNA species have func- lower in the nuclear fraction compared to cytoplasmic tional roles in both the nuclear and cytoplasmic com- fractions. partments of the cell, little systematic analysis has been done to study them at a population level. Here, we MicroRNAs have distinct subcellular localization patterns present a survey of small RNA species present in gilt After mapping reads to annotated swine miRNA, we granulosa cells obtained from small and large preovula- looked at the relationship between samples using princi- tory follicles, herein referred to as SGCs and LGCs. pal component analysis and unsupervised clustering. While we found a few significant differences between The first principle component comprised 54% of the SGCs and LGCs, we did reveal a diverse network of variance and separated nuclear samples from cytoplas- small RNA that showed distinct subcellular localization. mic samples. The second principal component that com- This small RNAome of ovarian granulosa cells will pro- prised 11% of the variance in the data was related to vide an important resource for studying their subcellular batch effect, although interestingly this primarily affected function during follicle growth. the variance of nuclear samples. This batch effect was corrected for in subsequent statistical analysis. Expres- Results sion of miRNA from distinct subcellular components Validation of subcellular fractionation clustered together without any obvious differences be- Following cell fractionation, intact nuclei were visualized tween miRNA from SGC and LGC (Fig. 3a). We next by phase-contrast and fluorescence microscopy, after examined expression of miRNA using a generalized lin- chromosome staining with Hoescht 33342 (Fig. 1a). Be- ear model (see Methods) comparing all four groups fore proceeding with deep sequencing of small RNAs, (SGC nuclei, LGC nuclei, SGC cytoplasm, and LGC we confirmed the quality and purity of the nuclear and cytoplasm). While most miRNA that showed differences cytoplasmic fractions. The RNA integrity number (RIN) in expression between the nucleus and cytoplasm were for each sample was determined by capillary electro- common between SGC and LGC, several miRNA dif- phoresis: LGC cytosol, 8.62 ± 0.40; LGC nucleus, 6.92 ± fered in their subcellular localization between granulosa 0.17; SGC cytosol, 9.02 ± 0.22; SGC nucleus, 6.78 ± 0.46. cells from the two stages of follicular development. Fur- Western blotting revealed the purity of each fraction by ther examination of nuclear miRNA revealed seven dif- the exclusion of glyceraldehyde 3-phosphate dehydro- ferentially expressed between SGC and LGC (Fig. 3b). genase from the nucleus and lamin B from the cytosol We then ranked miRNA based on their fold change (Fig. 1b). RT-qPCR analysis of the nuclear spliceosomal between the nucleus and cytoplasm (Supplemental Ta- RNA U6 revealed 1000-fold higher expression in the nu- bles 1 and 2). SGC showed more significantly different cleus, as expected (Fig. 1c), confirming clear separation miRNAs between these subcellular compartments than of the nucleus and cytosol fractions. LGC (101 versus 83, respectively). To confirm our sequencing results, we analyzed ex- Overview of small RNA sequencing data pression of 12 miRNA by digital droplet RT-PCR in our Four nuclear and four cytoplasmic RNA pools were se- fractionated samples. All analyzed miRNA showed iden- quenced from both LGCs and SGCs. After trimming tical enrichment, either cytoplasmic or nuclear, and a adapter sequences, reads with a sequence length less significant correlation of expression ratios (R = 0.50, P = than 18 nt were discarded. Remaining sequencing reads 0.013) between the two technologies (Fig. 4). had lengths between 18 and 50 nt (Fig. 2a). The majority of reads (97%) were between 22 and 25 nt in length. Some nuclear miRNAs are predicted to target promoter Twenty percent of these reads could not be mapped to regions the pig genome and 7% mapped to coding regions. With specific miRNAs observed to be enriched in the We mapped the remaining reads to S. scrofa small nucleus of granulosa cells and previous work demon- RNA databases and looked broadly at RNA species strating that small RNA in the nucleus can regulate tran- breakdown in nuclear and cytoplasmic fractions (Fig. 2b). scription, we examined the potential targeting of these MicroRNA, both mature sequences and hairpin precur- nuclear miRNAs to genomic promoter regions. Because sors, made up the majority of the mapped reads in both of a dearth of annotation for pig promoter regions, we cell fractions (55% of the nuclear reads; 61% of cytoplas- used the fact that mature miRNAs are highly homolo- mic reads) followed by 30.4% of nuclear reads mapping gous [13] and looked first at binding sites in human to snoRNA, while 16.3% of cytoplasmic reads remained promoter regions that were then mapped to homologous uncharacterized as “miscellaneous RNA”. As expected, regions in the pig. Our analysis revealed 417 potential considerably fewer cytoplasmic reads mapped to binding sites at the promoter regions across the human snoRNA, while reads mapping to ribozyme, transfer genome, of which 55 had identical matches to homolo- RNA and mitochondrial RNA were five- to ten-fold gous regions in the pig genome (Supplemental Table 3 Toms et al. Journal of Ovarian Research (2021) 14:54 Page 3 of 12 Fig. 1 Isolation of nuclear and cytoplasmic fractions. a Isolated nuclei were visualized at 400X magnification to confirm purity, left-to-right: phase contrast, Hoescht 33342, and trypan blue.