{"paper_id":"3bca93e4-5304-4f29-b5e8-dca6ed1af925","body_text":"Postpartum maternal reproductive function restoration usually involves two\naspects, including the recovery of ovarian function and completed uterine\ninvolution. Among the uterus and ovaries, the uterus is the reproductive\norgan that experiences the greatest change in morphology and function as it\nshowed endometrial regeneration and reconstruction of uterine function. The\nrecovery of maternal uterus after delivery to a non-pregnant state is known\nas uterine involution. This plays a vital role in ensuring normal\nreproductive function and cyclical estrus in postpartum female animals\n(Sheldon and Dobson, 2004; Mahdi and Khallili, 2008).\nThe uterus is a dynamic and complex organ that interacts with many deep\ntissues, ovaries, cerebrum, and other organs. In many of the aforementioned\nconnections, it is evident that the uterus and ovaries are dependent and\ninteract with each other in morphology and regulation of endocrine\nsecretion. Structurally, the ovaries are located at the posterolateral side\nat the end of the uterus and are connected to the uterus through Fallopian\ntubes. There are many nerve plexi and blood vessels between the uterus and\novaries. At the uterus, the internal iliac artery supplies the uterus and is\nanastomosed with the ovarian arteries (branch of the aorta) and is connected to\nthe inferior and superior vaginal blood vessels. At the same time, the lymph\nnodes of the uterine body are extended to the pelvic cavity on one side and\nthe para-aortic lymph nodes on the other side (Ercoli et al., 2010; Lanciego\net al., 2012; Abe et al., 2014; Hamadeh et al., 2018). At the endocrine\nsecretion level, the ovaries synthesize and secrete steroid hormones such as\nE 2  and P 4 . These two hormones are transported to uterine tissues\nvia the utero-ovarian plexus (UOP) and play important roles in the\nproliferation, differentiation, and functional changes of the endometrium\nand myometrium. A study showed that E 2  not only stimulates myometrial\nactivity, promotes uterine muscle excitation, and accelerates uterine muscle\ncontraction and action potential frequencies of individual fibers (Batra,\n1980), but also promotes intrauterine PGF 2 α  secretion, OTR\nsynthesis and collagenase activity, and cervix maturation. Similarly,\nP 4  secreted by the corpus luteum not only inhibits cervix maturation\nand myometrial excitation but also simultaneously decreases myometrial\nspontaneous electrical activity and OXT sensitivity, thereby affecting the\nphysiological functions of the uterus (Hafez and Hafez, 2000; Rodriguez Blanquet,\n2003). Conversely, PGF 2 α  and/or PGE 2  synthesized and\nreleased by uterine tissues are secreted into uterine veins and are locally\ntransported into ovarian tissues through the unique vascular structure of\nthe UOP to regulate corpus luteum maintenance and degradation. Therefore,\nchanges in the structures and functions of these tissues created\npreconditions for ensuring the cyclicity and continuity in breeding of\nfemale animals in the husbandry industry.\nIn recent years, a large volume of studies found that smooth muscle cells\nare rich in many types of stable miRNAs, and these miRNAs play important\nroles in the proliferation, hypertrophy, and differentiation of smooth\nmuscle cells (Carletti and Christenson, 2009; Li, 2014; Nothnick, 2015).\nThis suggests that a large number of miRNAs may participate in postpartum\nmaternal uterine involution, particularly the recovery of uterine smooth\nmuscle contraction function, and these miRNAs may play specific regulatory\nroles during this process. In view of this, we employed miRNA HiSeq deep\nsequencing to construct miRNA libraries of the ovary and uterus of\npostpartum ewes from treatment group (intramuscular injection of\nergometrine) and control group (intramuscular injection of physiological\nsaline) to screen miRNAs in the ovary–uterus axis that participate in\npostpartum maternal uterine involution. GO annotation and KEGG signaling\npathway analysis were used to predict target genes and the networks\nregulated by these genes. The results provide an important reference for\ncomprehensive understanding and examining of the molecular network\nregulatory mechanisms in postpartum maternal uterine involution.\n\nForty healthy adult Kazakh ewes that had similar ages (3–4 years), weighing\n45–50 kg, free from uterine diseases, and with normal reproductive function\nwere selected. The ewes were housed in the Shihezi University experimental\nstation under the same housing environment. Methylergometrine and\nphysiological saline were injected intramuscularly, and postpartum uterine\ninvolution model construction was completed to ensure significant phenotypic\nchanges in uterine involution in postpartum ewe populations. There were 15\newes in the treatment group, and 0.2 mg ergometrine was injected into the\nmedial thigh in the rear limb at 1 d after delivery. There were 15 ewes in\nthe control group, and an equal volume of physiological saline for medical\nuse was injected in the same site 1 d after delivery. Following that,\ndynamic changes in the uterus were monitored every day. In addition, early\nweaning was carried out 7 d after delivery in which the ewes and lambs in\nthe aforementioned groups were separated (lambs were transferred to another\npen for milk replacer feeding by dedicated staff).\nThe Tringa Vet portable B-mode ultrasound device was used to monitor uterine\ninvolution in postpartum ewes. Uterine involution was considered to have\nended when the uterine cavity was completely sealed, no residual liquid was\npresent in the uterine cavity, and uterus diameter was less than 2 cm.\nFollowing that, three postpartum ewes that met the aforementioned criteria were\nselected from each group and recorded as the ergometrine-treated group with\nfast uterine involution (UF) and physiological saline-treated group with\nslow uterine involution (US). After ewes were sacrificed, the ovary and\nuterus were collected, numbered, and placed in liquid nitrogen for storage\nbefore sampling.\nAfter the aforementioned samples were collected, chloroform, isopropanol,\nand 70 % ethanol were used to extract total RNA from the samples.\nFollowing that, total RNA from the UF and US group samples was sent to\nHonortech Co. Ltd. on dry ice, and the Agilent 2100 Bioanalyzer, Kaiao K5500\nmicrospectrophotometer, and Agilent RNA 6000 Nano Kit were used to measure\nRNA integrity and purity. Qualified samples were used to construct ovary and\nuterus miRNA libraries.\nThe aforementioned postpartum ovine ovary and uterus miRNA libraries in the\nUF and US groups with constructed uterine involution were marked as fast\nuterine involution–ovary (UFO) vs. slow uterine involution–ovary (USO), fast\nuterine involution–uterus (UFU) vs. slow uterine involution–uterus (USU).\nFollowing that, the Agilent 2100 Bioanalyzer and ABI StepOnePlus Real-Time\nPCR System were used for quality and yield measurements of the constructed\nlibraries. Lastly, the Illumina HiSeq 2500 platform was used for\nhigh-throughput sequencing of qualified sequencing libraries using the SE50\nsequencing strategy.\nThe different sequences obtained after sequencing were first filtered to\nobtain reliable target sequences, and the quality and lengths of these\nsequences and inter-sample common sequences were tallied. Following that,\ntarget sequences were classified and annotated to obtain the various\ncomponent and expression level information in each sample. Annotated small\nRNA fragments were used for the prediction of novel miRNAs. At the same\ntime, log2 ratio and scatterplot were used to analyze and screen\ndifferentially expressed miRNAs between groups. Lastly, the selected\ndifferentially expressed known miRNAs and novel miRNAs were used for cluster\nanalysis, target gene prediction, and GO functional annotation and KEGG\npathway annotation of target genes.\nTwelve miRNAs, each with six miRNAs in the ovary and uterus libraries, were\nrandomly selected to verify the reliability of the sequencing results.\nmiRNAs and mRNAs were reverse-transcribed to cDNA with the miRcute miRNA\nFirst-strand cDNA Kit (TIANGEN, Beijing, China). qRT-PCR was then performed\nusing the LightCycler96 qRT-PCR System (Stratagene, USA) and miRcute miRNA\nPremix SYBR (TIANGEN), following the manufacturer's instructions. All\nreactions were performed in triplicate. U6 RNA was chosen as an endogenous\ninternal control, and the relative expression levels were calculated based\non the 2 - Δ Δ Ct  method. The miRNA-specific primers are listed\nin Table S1A and B in the Supplement.\nBased on the joint analysis of all miRNA libraries, the proposed candidate\nmiRNAs and potential target genes were finally selected, and total RNA was\nextracted from different tissues of the same as the UF vs. US group and batch\nfor real-time fluorescence detection of miRNAs and mRNA. These potential\ntarget gene primers are listed in Table S1C. All experiments were repeated\nthree times with three replicates for each sample.\n\nB-mode ultrasound is an effective and reliable method to monitor postpartum\nuterine involution status in sheep. Two days after ewe delivery, the\nuterine horns were large in volume and contain more lochia. A linear array\nprobe B-mode ultrasound was used for transrectal examination of postpartum\nuterine involution changes but cannot completely observe the status of the\nentire uterus. From day 3 after delivery onwards, the uterine horns of\npostpartum ewes showed significant contraction. However, after the cervix\nand uterine body have completely closed, the uterine cavity in the uterine\nhorn was still partially opened. This shows that completion of uterine horn\nrecovery can be used as a marker for uterine recovery. Based on the\naforementioned uterine involution determination criteria, we used B-mode\nultrasound to track and monitor changes in the largest cross section during\nuterine horn recovery in the postpartum ewes (Fig. 1). Results showed that\nthe uterine involution duration in the UF group (17 d) was significantly\nshorter than the US group (27 d). This difference in uterine involution\nduration between postpartum ewes from the two groups was significantly\ndifferent, proving that construction of animal models for different duration\nof uterine involution was successful and can be used for subsequent\nexperiments.\nTrends for maximum cross-sectional diameter of the postpartum\nuterine horn between UF and US group.\nTo ensure accuracy in experimental data and improve the quality of data\nobtained, data preprocessing was carried out based on annotation information\nto finally obtain final clean reads for the UF and US group samples. Results\nshowed that 41 262 294, 42 539 226, 42 355 026, and 40 960 450 raw reads\nwere obtained from the UFO, USO, UFU, and USU libraries, respectively.\nFollowing that, adapters, contaminated sequences, and low-quality reads were\nremoved to obtain clean data, and Rfam and miRbase database annotation were\nused for data preprocessing to obtain 40 321 542 and 40 702 382 clean reads\nfrom the UFO and USO libraries respectively, which accounted for 97.72 %\nand 95.68 % of raw reads, respectively (Table 1). Similarly, 40 978 134\nand 39 652 778 clean reads were obtained from the UFU and USU libraries,\nrespectively, which accounted for 96.75 % and 96.81 % of total reads,\nrespectively (Table 1). This shows that the quality of miRNA libraries\nconstructed was good and can be used for subsequent analysis.\nPercentage and distribution of sequencing results in four libraries.\nAfter removing contamination, all clean reads that were 18 nt and above in\nlength were used for genome localization, classification, and annotation.\nFirst, small RNA sequences were aligned with non-coding small RNA and RNA\nrepeat sequences, introns, and exons in Rfam database 11.0 and NCBI GenBank\ndatabase. Results are shown in Fig. 2a, b, c, and d. Alignment results\nshowed that most of the small RNAs in the UFO (Fig. 2a), USO (Fig. 2b), UFU (Fig. 2c), and\nUSU (Fig. 2d) databases were miRNAs. Following that, known and unknown miRNAs in\nthe two databases were used as data sources for subsequent analysis.\nStatistics of the distribution characteristics of non-coding small\nRNAs between UF and US group.  (a)  UFO,  (b)  USO,  (c)  UFU, and  (d)  USU.\nCompared with the US group, 16 significant differentially expressed miRNAs\nwere found in the UFO libraries, of which 4 and 12 were known and unknown\nmiRNAs, respectively. Among these miRNAs, the four known miRNAs were\nupregulated (Table 2). At the same time, 54 significant differentially\nexpressed miRNAs were found in the UFU libraries, of which 5 and 49 were\nknown and unknown miRNAs, respectively. Among these five known miRNAs, two were\nupregulated, and three were downregulated. Table 2 shows some known\ndifferentially expressed miRNAs. It should be worth noting that we found one\ncommon differentially expressed miRNA in the ovary and uterus libraries,\nwhich was oar-miR-200a. In addition, Table 3 shows the position and mature\nsequences of the 12 most significant differentially expressed unknown miRNAs\nfrom the UF and US libraries in the genome.\nThe significantly differential expression of known miRNA.\nRelatively higher abundance 12 novel miRNA position on the genome\nand its sequence in UF and US libraries.\nTo further validate the reliability of the four miRNA libraries, we randomly\nselected six differentially expressed miRNAs from the ovary and uterus\nlibraries, including four known miRNAs and two unknown miRNAs for RT-PCR\nvalidation. The 2 - Δ Δ Ct  method was used to calculate the\nrelative expression level of the aforementioned miRNAs (Fig. 3a and b).\nResults showed that the expression level differences and library sequencing\nhigh-throughput differences of the six miRNAs were consistent, indicating that\nthe ovary and uterus library data obtained are trustworthy and reliable.\nValidation of the Solexa sequencing results in the ovarian and\nuterine libraries. Data are mean  ±  SEM. Panels  (a)  and  (b)  show ovary and\nuterus libraries, respectively.\nCurrently, there are a few software programs used for the prediction of miRNA\ntargets. In this study, four software programs (such as TargetScan, miRDB, miRWalk, and\nmiRTarBase) were used to predict miRNA target genes, and the intersecting\ntarget genes were finally selected as target genes for candidate miRNAs.\nResults showed that there are 39 target genes of differentially expressed\noar-miR-200a that are common to the ovary and uterus libraries (Fig. 4).\nPreliminary screening and analysis were used for GO annotation and KEGG\npathway enrichment analysis of the aforementioned target genes.\nVarious types of software for predicting the target genes of\nmiRNAs. The numbers in the intersection of the four circles refer to the\ncandidate target genes of miRNAs.\nThe aforementioned 39 target gene sets were used for molecular function,\nbiological process, and cellular component annotation and enrichment\nanalysis. Results showed that GO enrichment of target genes in this study\nare involved in the regulation of many biological processes, such as\nreproduction, multi-organism process, biological adhesion, and biological\nregulation (Fig. 5). At the same time, KEGG signaling pathway enrichment\nanalysis of the above target genes found that a large number of candidate\ngenes are enriched in microRNAs in cancer, glycosylphosphatidylinositol\n(GPI)-anchor biosynthesis, Hippo signaling pathway, and Hippo signaling\npathway-multiple species (Fig. 6).\nAll the target genes were analyzed using taxonomic annotation and\nenrichment analysis in terms of molecular function, biological process, and\ncell composition.\nAll the target genes were conducted using KEGG signaling pathway\nenrichment analysis.  P  value  <  0.05 was considered statistically\nsignificant.\nGO and KEGG enrichment analyses based on miRNA target prediction and target\ngenes found that one miRNA-target gene pair was involved in microRNAs in\ncancer, which was miR-200a- ZEB1 . At the same time, another miRNA-target gene\npair is involved in the Hippo signaling pathway, i.e., miR-200a- YAP1 . Based\non this, we used the same treatment group tissue samples for qRT-PCR for\nfurther validation of the reliability of the aforementioned screening\nresults (Fig. 7). Results showed that significant differential expression of\nmiRNA-200a and its potential target genes ( ZEB1  and  YAP1 ) are present in\novary and uterus tissues, and the expression trends of the two are\nnegatively correlated. This shows that typical miRNA-target gene negative\nregulatory relationships may be present between them. This provided a\ndirection and basis for subsequent mining of functional regulatory\nmechanisms.\nExpression levels of the candidate miRNAs and their target genes in\nthe ovarian and uterine samples. All the samples are consistent with the\nlibrary samples.\n\nThe uterus is an important site for fetal development and plays an important\nrole in life formation and development in mammals (Browne et al., 2015). At\nthe same time, the uterus is also a functionally complex endocrine organ\nthat can secrete many bioactive substances to regulate local and systemic\nphysiological and pathological processes and ensure the dynamic equilibrium\nin interactions and endocrine secretion with other organs and tissues.\nDuring pregnancy, the uterus of the mother will undergo a series of changes\nas the fetus develops, including the increase in uterine volume and weight,\nthe increase in blood flow in various parts of the uterus, and changes in\nnerve distribution, tissue hormone level, and their receptors. When the\nfetus is full term, the uterus will undergo a series of changes as delivery\noccurs to ensure that all changes that occurred during pregnancy will\nrecover or be restored to pre-pregnancy status to fully prepare for the next\npregnancy (Kaelin Agten et al., 2018).\nThe uterus and ovaries are dependent on and interact with each other in\nmorphology and regulation of endocrine secretion during recovery of\nreproductive function in female animals. For example, granulosa cells in the\novaries secrete E 2 , while the corpus luteum secretes P 4 . These two\nhormones are transported by the UOP arteriovenous anastomosis to the uterus\nto regulate cyclical changes in the endometrium and myometrial contraction.\nConversely, the endometrium and myometrium will synthesize and secrete\nprostaglandin (PG) family hormones, which are transported from the UOP blood\nvessels to ovary–corpus luteum tissues and participate in ovary–corpus\nluteum degeneration. Currently, many studies proved that a large number of\nmiRNAs are present in ovary and uterus tissues. These miRNAs play important\nroles in granulosa cell proliferation, sex hormone synthesis and secretion,\ncorpus luteum apoptosis, and endometrial regeneration and remodeling. It is\nreported that miR-23a, miR-23b, miR-542-3p, miR-211, and miR-17-5p can\ntarget several key genes in the PGs-P 4 -E 2  axis in ovary–uterus\ntissues, such as the expression of  COX2 ,  StAR , and  CYP19A1  to participate in\nregulating granulosa cell proliferation, differentiation, and steroid\nbiosynthesis and secretion (Donadeu et al., 2012). At the same time, Akhtar\nand Haqqi (2012) found that miR-199a can target and regulate COX2 expression to\nparticipate in the synthesis of intrauterine PGF 2 α  so that the\nformer plays a key role in uterine contraction. In\ncontrast to previous studies, we carried out an alignment analysis of\ndifferentially expressed miRNAs in the ovary and uterus libraries and found\none common miRNA, which was miR-200a. In addition, four target-prediction\nsoftware programs, including TargetScan, miRDB, miRWalk, and miRTarBase, were employed,\nand we found 39 target genes for miR-200a. These were combined with KEGG\nsignaling pathway enrichment analysis, which showed that a miRNA-target gene\npair, miR-200a- ZEB1 , was involved in microRNAs in cancer. At the same time,\nanother miRNA-target gene pair, miR-200a- YAP1 , is involved in the Hippo\nsignaling pathway. MiR-200a is a member of the miR-200 family, and a large\nvolume of studies have proved that it participates in many physiological\nregulation processes. For example, Williams (2014) showed that miR-200a can\ndirectly inhibit  STAT5b  increase and can decrease  20 \n α \n -HSD  expression to\nregulate myometrial P 4  level to participate in regulating uterine\ncontraction. In fact, it is worth mentioning that miR-200\nfamily members can regulate the expression of ZEB family members ( ZEB1  and\n ZEB2 ) to inhibit genes associated with smooth muscle contraction ( CX43  and\n OTR ) to maintain myometrial quiescence. Conversely, when both  ZEB1  and  ZEB2 \nsimultaneously decrease,  OTR  and  CX43  genes will be significantly\nupregulated and induce myometrial excitation. At the same time, a study\nfound that elevation in the concentration of P 4  secreted by ovaries can\ninduce  ZEB1  expression, thereby weakening the inhibitory effects of miR-200\nfamily members on ZEB family members (Wu and DeMayo, 2017). Similarly, Wu and DeMayo (2017)\nconducted a study on in vitro transfected uterine trophoblast cells and\nfound that miR-200a inhibition can increase the proliferation, migration,\nand invasion of uterine trophoblasts, while transfection with  ZEB1  siRNA can\ninhibit miR-200a expression. Therefore, the\naforementioned studies showed that miR-200a may target ZEB family members\n(such as  ZEB1 ) in the postpartum mother to participate in postpartum uterine\nremodeling, including uterine involution and repair of the intrauterine\nenvironment. In addition, we also noticed that another target gene of\nmiR-200a is  YAP1 , which is a core gene in the Hippo pathway. This pathway\nplays a role in regulating in vivo balance, oncogenesis, and regeneration.\nChen et al. (2011) reported that the YAP1/Hippo pathway plays a crucial role in\ncontrolling the differentiation of myofibroblasts and fibrosis development. At the same time, Song et al. (2016) found that compared with\nnormal endometrium,  YAP  overexpression in the YAP/Hippo signaling pathway\nwill not only lead to endometriosis but also upregulate  CTGF  and  BCL-2 \nexpression to increase proliferation and decrease apoptosis. In summary, many studies showed that some regulatory circuits may be\npresent in ovaries and uterus in the postpartum mothers. That is, the\n ZEB1 -miR-200a- YAP1  pathway, which connects the uterus–ovary axis and\nPG-E 2 /P 4  endocrine regulatory networks to participate in the\npostpartum maternal uterine remodeling, plays important regulatory roles\nduring this process.\nIn conclusion, we constructed postpartum ovine ovary and uterus libraries\nand successfully screened out eight known differentially expressed miRNAs, of\nwhich one differentially expressed intersection miRNA-200a. In addition,\nbioinformatics analysis found two pairs of miRNA-target genes –\nmiRNA-200a- ZEB1  and  YAP1  – may participate in regulating postpartum maternal\nuterus morphology and function recovery through the ovary–uterus exits\nconstructed by  ZEB1 -miR-200a- YAP1  pathway. This provides a direction and a\nsolid foundation for an in-depth examination of the molecular network\nregulatory mechanisms in postpartum uterine involution in female animals.","source_license":"CC-BY-4.0","license_restricted":false}