MIR-361-5P Mediates TGF-β Signalling to Promote Granulosa Cell Apoptosis Through Vegfa During Porcine Follicle Atresia

Background: Follicular atresia is an inevitable degenerative process of ovarian follicles in mammals. The molecular events involved in atresia, particularly in granulosa cell apoptosis, have always attracted researchers’ attention. It is known that VEGFA isdownregulated during follicular atresia in pig ovaries and serves as an apoptosis inhibitor of granulosa cells. Besides, the TGF-β signalling has been considered as a central trigger in granulosa cell apoptosis. However, the link between TGF-β signalling and VEGFA is missing. Results:We proved that miR-361-5p signicantly up-regulated during the atresia process andenhances GC apoptosis by direct targeting to VEGFA 3’UTR. In addition,we revealed that miR-361-5p coding geneMIR361was signicant downregulation bySMAD4, the central intracellular mediator of TGF-β signalling,through promoter binding. Conclusions: Our ndingsenriched the knowledge of VEGFA post-transcriptional regulation and completed the TGF-β/miR-361-5p/VEGFAregulatory networkin ovarian granulosa cell apoptosis. It offereduseful references for follicular atresia and ovarian physiological function studies.


Background
Follicular atresia is a common physiological phenomenon, which may happen at all stages of follicular development in mammals. In porcine primordial follicle reserve at puberty, there are approximately 5 million primordial follicles, most of which go through the atresia process when they passed the size of 1 mm of diameter during the antral stage. The atresia rate remarkably increases in antral follicles around 3-5 mm in diameter (Marchal et al., 2002). In the end, the majority of follicles disappear before maturation, and less than 14% can reach ovulation (Manabe et al., 2004).
Adequate blood ow is necessary for oxygen and nutrient supply in the ovary and possibly a rate-limiting step in the selection and maturation of dominant follicles destined for ovulation (Stouffer et al., 2001). It is known that the vascular endothelial growth factor (VEGF) family, which is composed of at least six members (VEGF A-F), is involved in the formation of blood vessels. VEGFA is the rst identi ed and mostly studied molecule, which is primarily known as a promoting factor of angiogenesis and vasculogenesis (López et al., 2017). It is considered that the function of VEGFA is achieved through its major tyrosine kinase receptor VEGFR2 (also known as kinase insert domain receptor, KDR) (Greenaway et al., 2004). The expression and function of VEGFA have been studied in mammal ovaries during follicular development. In human, VEGF mRNA and protein presented in theca layer and granulosa cells (GCs) during antral follicle stages, when vascular network developed actively (Yamamoto et al., 1997); In bovine, VEGFA mRNA also expressed in both theca interna and GCs, while the VEGF protein level increased with developmental stages of follicle growth (Berisha et al., 2000); In pig, expression levels of two VEGFA isoforms (VEGF120 and VEGF164) in GCs and levels of two receptors (VEGFR1 and VEGFR2) in theca cells appeared to be higher in medium and large follicles than in small ones (Shimizu et al., 2002). According to our earlier study in pig, VEGFA mRNA level signi cantly decreased in early atretic follicles compare to the healthy ones (Zhang et al., 2018a), which implied a potential role of VEGFA during follicular atresia.
Direct binding microRNAs such as miR-26a (Chai et  It has been reported that the canonical transforming growth factor (TGF)-β signalling pathway is involved in the proliferation and apoptosis of porcine ovarian granulosa cells. In our previous studies, SMAD family member 4 (SMAD4), the terminal regulatory molecule in TGF-β signalling, was identi ed as an anti-apoptosis factor in granulosa cells (Zhang et al., 2010). Interestingly, this function of SMAD4 was related to its direct binding to the promoter of a few miRNA coding genes and negative regulation of miRNA expression (Xing et al., 2016), which implied a possible regulatory axis of SMAD4-miRNAsfunctional genes in GCs. Thus, we performed the study to identify a potential regulatory axis of SMAD4/miR-361-5p/VEGFA in porcine GCs. The study not only lled the gap of miRNA regulated VEGFA during follicular atresia but also provided evidence for the anti-apoptosis role of SMAD4 through transcriptional regulation of miRNAs in GCs.

Follicle Collection
Ovaries were obtained from mature sows at a local slaughterhouse. Individual antral follicles, approximately 3 to 5 mm in diameter, were dissected from the ovaries after quickly washing twice with 75% ethanol and physiologic saline using small scissors and ne forceps and then classi ed into healthy follicles (HFs) and atretic follicles (AFs) according to follicle shape, GC density, and hormone levels (Zhang et al., 2018a).

Cell Culture and Transfection
Primary GCs were obtained from HFs by syringing with a 20-gauge needle and cultured as previously described . HEK293 cells were cultured as previously described . VEGFA siRNA, Smad4 siRNA, miR-361-5p mimic, and miR-361-5p inhibitor were synthesized by GenePharma (Shanghai, China) (Supplementary Table S1). After 12 h of culture, the porcine GCs were transfected with the appropriate plasmids or oligos using Lipofectamine 2000 and Opti-MEM (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's protocol.

RNA extraction and qRT-PCR
Total RNA was extracted from the follicles and GCs using Trizol reagent (Invitrogen, Carlsbad, CA, USA). Then the extracted total RNA was reverse transcribed to cDNA by Super M-MLV RTase Synthesis Kit and qRT-PCR was performed using SYBR Premix Ex Taq (Takara, Dalian, China) on the ABI StepOne system (Applied Biosystems, Carlsbad, CA, USA) according to the manufacturer's instructions. GAPDH was used as an internal control. For miRNA detection,frst-strand cDNA synthesis and qRT-PCR were performed using miRNA two-step qRT-PCR SuperMix (TransGen, Beijing, China). U6 was used as internal control. For each gene, controls for each primer set containing no cDNA were included on each plate, and the reaction was repeated three times for every sample. The primers for qRT-PCR are listed in Supplementary Table   S2.

uorescent in situ hybridization (FISH)
The FAM-labeled probe (5'-GTACCCCTGGAGATTCTGATAA-3 ') was speci cally synthesized for miR-361-5p, and DAPI was used to stain the cell nuclei. The GCs were cultured on coverslips, xed in 4% paraformaldehyde (containing DEPC) for 20 min, washed with shaking in PBS (PH7.4) for three times, and proteinase K (20ug/ml) was nally added for 5 min for digestion. Then all the procedures were conducted according to the manufactory's instruction (Sevicebio, Wuhan, China). Finally, the images were acquired on a Nikon upright uorescence microscope (Nikon DS-U3, Japan). Each experiment was performed three times.

Plasmid construction
The 3′-UTR fragments of VEGFA containing putative target sites of miR-361-5p and the promoter fragments of the miR-361-5p coding gene (MIR361) containing putative SMAD4 binding sites were ampli ed from porcine genomic DNA and veri ed by sequencing. The VEGFA 3′-UTR fragment was digestion with Nhel and Xbal (Thermo, Waltham, MA, USA), and then cloned into pmirGLO Dualluciferase miRNA Target Expression Vector (Promega Corporation, Madison, WI, USA). The MIR361 promoter fragment was digestion with NheI and SacI and then cloned into pGL-3 reporter vector (Promega, USA). The mutant plasmids of the miR-361-5p putative binding site were generated by ClonExpress Entry One Step Cloning Kit (Vazyme, Nanjing, China) according to the manufacturer's protocol. The successful mutations were identi ed by sequencing technology. The overexpression plasmid pcDNA3.1-SMAD4 plasmid was generated previously by our group (Du et al., 2018). Primers used here are detailed in Supplementary Table S3.

Luciferase Reporter Assays
After a transfection period of 24 h, the cells and lysates were collected. A Dual-Luciferase Reporter Assay System (Promega Corporation, Madison, WI, USA) was used to quantify luciferase activities following the manufacturer's instructions. Fire y luciferase activity was normalized to Renilla luciferase activity. Each experiment was performed six times.

Statistical analysis
All data are presented as means ± S.E.M. Prism 5 software (GraphPad Software) was used to perform statistical analysis. Two-tailed Student's t-test was used to evaluate the signi cance when two groups were compared. When three or more groups were compared, a one-way analysis of variance test was performed, and Turkey's test to determine signi cance between groups. P-value of < 0.05 and 0.01 was considered signi cant and extremely signi cant differences, respectively.

VEGFA is downregulated in atretic follicles
To investigate the VEGFA level during follicular atresia, we rst determined the location of VEGFA in antral follicles by immunohistochemistry. The results showed that a positive reaction was observed in both theca and granulosa cells, and the brown staining was stronger in HFs (GCs closely arranged) than in AFs (GCs loosely arranged and partially dropped into the follicular cavity) (Fig. 1.A-D). In addition, mRNA levels of VEGFA detected in the whole follicle, GC, and theca cell (TC) by GeneChip Porcine GenomeArray (detailed data was contained in reference (Zhang et al., 2018a)) or qRT-PCR also suggested a signi cantly higher expression in HFs compare to atretic ones ( Fig. 1.E-G). The ELISA detection of VEGFA showed a slight but signi cant decrease of VEGFA content in follicular uid during the atresia process ( Fig. 1.H). These results suggested that VEGFA decreases during porcine follicular atresia.

miR-361-5p is up-regulated in atretic follicles
To investigate the possible function of miR-361-5p during follicle atresia, we detected its location and expression levels in healthy and atretic follicles. FISH results showed that miR-361-5p distributed in both TCs and GCs, and the signal was stronger in atretic follicles compare to the healthy ones ( Fig. 2

.A-F).
Further qualitative measurement in the whole follicle, GC, TC, and follicular uid by µPara o™ micro uidic chip (detailed data was contained in reference (Lin et al., 2012)) or qRT-PCR, respectively, also suggested a signi cantly higher expression of mir-361-5p in atretic follicles compare to healthy ones in each component of the follicle (Fig. 2.G-J). These results implied that miR-361-5p was involved in the atresia process and may involve in post-transcriptional regulation of functional genes during follicular atresia.

miR-361-5p regulates VEGFA by direct 3'UTR binding
To further investigate the possible function of miR-361-5p in the regulation of VEGFA expression, the direct targeting of VEGFA by miR-361-5p was rst predicted by bioinformatic methods and con rmed by luciferase reporter assay (Fig. 3.A). Next, we cultured porcine GCs, transfected with miR-361-5p mimics or inhibitor, and then detected the mRNA/protein levels of VEGFA in GCs and VEGFA secretion in culture media. The results showed that both mRNA (Fig. 3.B, C) and protein (Fig. 3.D, E) levels of VEGFA were signi cantly decreased after transfection of miR-361-5p mimics, while increased after transfection of miR-361-5p inhibitor (Fig. 3.B-E). Besides, we noticed that VEGFA secretion in culture media was slightly increased after transfection of miR-361-5p inhibitor (Fig. 3.F, G). These results suggested that miR-361-5p negatively affected VEGFA expression by direct binding to its 3'UTR in porcine GCs.

SMAD4 involved in miR-361-5p mediated VEGFA expression
To examine whether the expression of miR-361-5p was under the control of TGF-β signalling pathway, we predicted the promoter region of the miR-361-5p coding gene, MIR361. We identi ed 4 SMAD-binding elements (SBEs) within the region (Fig. 5.A). Next, we con rmed the promoter activity of the MIR361 promoter region with SBEs using a dual-luciferase reporter assay (Fig. 5.B). To further investigate the effect of SMAD4 on MIR361 promoter, the increase and knockdown of SMAD4 were achieved by SMAD4 overexpression plasmid and siRNA, respectively. The luciferase reporter assays suggested that SMAD4 had a negative effect on the MIR361 promoter ( Fig. 5.C, D). In addition, in cultured porcine GCs, overexpression of SMAD4 did not show an apparent effect on miR-361-5p, but knockdown of SMAD4 signi cantly enhanced miR-361-5p expression (Fig. 5. E, F). Finally, both qRT-PCR and WB suggested that knockdown of SMAD4 resulted in a signi cant decrease of VEGFA expression (Fig. 5. G, H). These results indicated that SMAD4 regulated MIR361 transcription negatively as a trans-acting element and thereby adjusted VEGFA expression. In this study, we con rmed a direct interaction between miR-361-5p and VEGFA, which was mainly produced in GCs and play essential roles in angiogenesis, GCs function and oocyte development in antral follicles (Gao et al., 2020). In addition, our study explored the transcriptional regulation of the miR-361-5p coding gene, MIR361, by SMAD4 and thus identi ed a straightforward regulatory network of SMAD4/miR-361-5p/VEGFA. Brie y, miR-361-5p reduces VEGFA mRNA expression by direct binding, thus promotes GC apoptosis, while SMAD4 increases VEGFA level through negative regulation of MIR361 expression through promoter binding at the transcription level (Fig. 6).

Discussion
It is known that SMAD4 is the universal mediator that plays a role in canonical TGF-β signal transduction into the nucleus, where the SMAD complexes regulate gene transcription positively or negatively with different coactivator or corepressor factors (Papoutsoglou et al., 2019). TGF-β signalling has been noticed in porcine ovarian granulosa cells after a SMAD4 knockdown study . Since then, the apoptosis-inducing capacity of TGF-β signalling was gradually proved. Some studies suggested that SMAD4 affects the FSH response because knockdown of SMAD4 signi cantly inhibited FSH-induced GC proliferation and estradiol production (Wang et al.). Also, SMAD4 was proved to downregulate miR-143 expression by promoter binding, thus resist GC apoptosis caused by miR-143 targeting of FSHR (Xing et al., 2016). Our ndings, however, implied another possible function of TGF-β signalling during follicular atresia through the VEGFA-mediated angiogenesis and vasculogenesis processes. Although it is well known that expression of VEGFA can be regulated at the transcriptional level by several cis-acting mechanisms, its regulation by TGF-beta seems to be at the post-transcriptional level. For example, reduced stability of the VEGFA protein through TGF-β signalling mediated ubiquitination and degradation was observed in colon carcinoma (Geng et al.). Our results provided further evidence that TGF-β signaling could reduce VEGFA level through miRNA mediated mechanisms, which added knowledge to the functional study of TGF-β signaling.

Conclusions
Our data provide direct evidence that miR-361-5p is up-regulated during follicular atresia and enhanced GC apoptosis by direct targeting to the 3'UTR of VEGFA mRNA and downregulating its expression. In addition, the TGF-β signalling might play a part in VEGFA mediated GC apoptosis by transcriptional regulation of miR-361-5p expression via its common mediator SMAD4. Overall, our ndings enriched the knowledge of VEGFA post-transcriptional regulation in ovarian GC apoptosis, provide novel insights into the mechanism underlying follicular atresia and ovarian physiological function studies in ovaries of mammals.      Transcription factor SMAD4 was involved in miR-361-5p mediated VEGFA expression. A: Schematic diagram showing the genome location of the miR-361-5p coding gene and potential SMAD4 binding sites. B: The promoter activity ofMIR361upstream region con rmed by dual-luciferase reporter assay; C,D:

Abbreviations
Overexpression and knockdown of SMAD4 weaken and enhanced MIR361 promoter activity respectively; E: The expression of miR-361-5p was not affected after SMAD4 overexpression; F: The expression of miR-361-5p was up-regulated after SMAD4 knockdown; G,H: The expression of VEGFA mRNA and protein levels was down-regulated after SMAD4 knockdown. n = 3 cell culture wells per group. Data are expressed as the mean ± SEM. *p < 0.05, **p < 0.01.