Small Non-Coding-RNA in Gynecological Malignancies

Simple Summary Gynecologic malignancies are among the leading cause of female mortality worldwide, and their management is complicated by late diagnosis and acquired therapy resistance. Although altered DNA code, leading to aberrant protein expression, is indispensable for cancer initiation and progression, from the current literature it is clear that, not only proteins, but also noncoding RNA, which does not translate into proteins, can also be instrumental. Based on their size, noncoding RNA, are further classified into long and small noncoding RNA. Here, we have comprehensively reviewed the literature about the role of small noncoding RNAs in gynecological malignancies, and discussed how these small noncoding RNA can be vital for diagnosis and therapy. Abstract Gynecologic malignancies, which include cancers of the cervix, ovary, uterus, vulva, vagina, and fallopian tube, are among the leading causes of female mortality worldwide, with the most prevalent being endometrial, ovarian, and cervical cancer. Gynecologic malignancies are complex, heterogeneous diseases, and despite extensive research efforts, the molecular mechanisms underlying their development and pathology remain largely unclear. Currently, mechanistic and therapeutic research in cancer is largely focused on protein targets that are encoded by about 1% of the human genome. Our current understanding of 99% of the genome, which includes noncoding RNA, is limited. The discovery of tens of thousands of noncoding RNAs (ncRNAs), possessing either structural or regulatory functions, has fundamentally altered our understanding of genetics, physiology, pathophysiology, and disease treatment as they relate to gynecologic malignancies. In recent years, it has become clear that ncRNAs are relatively stable, and can serve as biomarkers for cancer diagnosis and prognosis, as well as guide therapy choices. Here we discuss the role of small non-coding RNAs, i.e., microRNAs (miRs), P-Element induced wimpy testis interacting (PIWI) RNAs (piRNAs), and tRNA-derived small RNAs in gynecological malignancies, specifically focusing on ovarian, endometrial, and cervical cancer.


Gynecologic Cancers
In the United States, in 2020, gynecologic cancers constituted~6% of all cancer patients, accounting for~5.3% of all cancer related deaths. Typically, the primary, and most effective, treatment for these cancers is optimal surgery; however, surgery is effective only if the disease is diagnosed at an early stage. Due to recent clinical advances, including the availability of effective screening tools, both endometrial and cervical cancer are increasingly likely to be diagnosed at an early stage. Disappointingly, ovarian cancer is designated the "silent killer", because it is often diagnosed at an advanced stage where curative treatment is difficult; treatment of ovarian cancer is often further challenged by recurrence and acquired drug resistance. These malignancies have been reviewed in detail elsewhere [1][2][3][4], and a brief summary is presented in the following sections: In the primary pathway, piRNA precursors are transcribed from piRNA clusters by RNA polymerase 2 (RNA Pol2). Antisense primary piRNAs are cleaved, trimmed into short fragments, and their 3 ends are 2 -O-methylated and then loaded onto PIWI family proteins. In the secondary amplification pathways, also known as the ping-pong cycle, PIWI proteins associate with antisense piRNA and cleave piRNA precursors in the sense strand, or PIWI proteins associate with sense piRNA and cleave antisense piRNA precursors in the sense strand. The incorporated RNA is therefore processed into a mature secondary piRNA by trimming and modification, likely by the same mechanisms that generate a primary piRNA. (B) The biological functions of piRNAs: In the nucleus, PIWI-piRNA complexes can repress the transposon expression by methylation of the transposon region or chromatin modification around the transposon region. In the cytoplasm, piRNAs can facilitate mRNA maturation, cause cleavage of mRNAs through a miRNA-like mechanism, and degrade retrotransposon-associated mRNAs.
Increasing evidence suggests that PIWI proteins are linked to the hallmarks of cancer, such as cell proliferation, anti-apoptosis, genomic instability, invasion, and metastasis. Due to their restricted expression PIWI are classified as cancer/testis antigen (CTA) and are considered as excellent targets for diagnostic and prognostic biomarkers, and immunotherapy [62]. The expression of piRNA or PIWI protein in non-germline cancers is in line with the well-established phenomenon of cancer/germline genes, which describes the aberrant expression of germline-specific genes in non-germline cancers [61,[63][64][65][66]. This provides new possibilities for anticancer therapies through the targeting of PIWI proteins, and which may have fewer side effects due to their restricted expression.

piRNA in Gynecological Cancers
Increasing functional evidence supports the involvement of piRNAs in the regulation of epigenetic changes in tumorigenesis [67][68][69], along with posttranscriptional mRNA and protein stability regulation. It has been suggested that the PIWI-piRNA complex contributes to cancer development and progression by promoting a stem-like state of cancer cells, or cancer stem cells (CSCs). It has been reported that CSCs represent the cells that have undergone epithelial-mesenchymal transition (EMT) and acquired metastatic capacities. Such epigenetic alterations allow cancer cells to adapt to changes in their microenvironment. Epigenetic global changes in cancer include DNA hypo-methylation, histone hypo-acetylation, and gene-specific DNA hyper-methylation, leading to oncogene activation (R-ras, cyclin D2) [70], and tumor suppressor silencing (RB1, p16) [71]. In cancer tissues, aberrantly expressed piRNAs implicate global hypo-methylation and local hyper-methylation as potential cancer-specific features [69,72].
Abnormal expression of piRNAs is emerging as a crucial regulator in cancer cell proliferation, apoptosis, invasion, and migration. In gynecologic malignancies the study of piRNA expression and their pathophysiological significance remains exploratory. Singh et al. [73] used RNA sequencing to identify piRNAs in normal ovary (pi RNA #219), endometrioid (pi RNA #256), and serous ovarian cancer (pi RNA #234). Although sample numbers used in the study were small, the authors reported 159 and 143 differentially expressed piRNAs in endometrioid and serous ovarian cancer, respectively. The differentially expressed piRNAs in endometrioid ovarian cancer were comprised of 74 upregulated and 77 down-regulated piRNAs, while those in serous ovarian cancer included 56 upregulated and 81 downregulated. Specific findings showed that piR-52207 was upregulated in endometrioid ovarian cancer, and piR-52207 and piR-33733 were increased in serous ovarian cancer [73]. Upregulated piR-52207 targets NUDT4, MTR, EIF2S3, and MPHOSPH8, which promote endometrioid ovarian cancer cell proliferation, migration, and tumorigenesis. In serous ovarian cancer, piR-33733 targets LIAS3 -UTRs, whereas piR-52207 binds ACTR10 and PLEKHA5 3 -UTRs and 5 -UTRs, leading to increased anti-apoptotic and decreased pro-apoptotic proteins. Thus, piR-52207 and piR-33733 promote ovarian cancer oncogenes via involvement in multiple cell-signaling pathways at the post-transcriptional level, supporting them as possible therapeutic targets for ovarian cancer [73].
In endometrial cancer, studies utilizing small-RNA sequencing and microarrays, have shown a significant difference in the expression pattern of piRNAs between normal, hyperplastic, and the neoplastic endometrium [74]. Utilizing the stringent thermodynamic parameters for RNA-RNA binding, the authors showed each piRNA to be complementary to a number of mRNAs, ranging from 28 to 308. In total, 1526 mRNA targets were predicted; differential expression analysis on paired sample groups revealed that 170 of the predicted 1526 were differentially expressed (|FC| ≥ 1.5 and p-value 0.001) in hyperplastic and/or tumor tissues [74]. In cervical cancer cell lines piR-651 has been shown to be upregulated [75] however, apart from this one example the details of piRNA status in cervical cancer remains largely unexplored.

Future Perspectives
Cancer and germ cells share several essential characteristics, including high proliferation rates and self-renewal abilities; in addition, cancer cells may re-activate cancer testis antigen (CTA) genes, whose expression is usually restricted to the germline, and silenced in adult somatic tissues. The expression of germline genes in cancer reflects the aberrant activation of a silenced developmental program that leads to escape from cell death, immune evasion, and invasiveness, thus contributing to the molecular mechanisms of carcinogenesis [61]. Among cancer testis antigens, PIWI-like (PIWIL) genes, belonging to the Ago family, are frequently deregulated in several malignancies, including cervical [60, 76,77], endometrial [78], and ovarian cancer [58,59,76], and these proteins, along with piRNA, are involved in various aspects of malignancy, and associated with advanced tumor stage and poor prognosis. Along with piRNA, the PIWI proteins are prominently expressed in cancer cells, making them both useful biomarkers for cancer diagnosis and possible druggable targets [79].

tRNA-Derived Small RNAs
Transfer RNA (tRNA), one of the most abundant cellular ncRNAs, is important for protein translation. Recent research has shown that tRNAs are not always the terminally differentiated molecules; fragments derived from tRNAs are a source of small regulatory RNAs, known as tRNA-derived small RNAs (tsRNAs) [80]. Based on the cleavage site, tsRNAs can be divided into two main types: (1) transfer RNA-derived RNA fragments (tRFs) approximately 14 to 30 nucleotides in length, and derived from mature or precursor tRNAs; and (2) tRNA halves or tiRNAs, 29 to 50 nucleotides in length, induced by stress, and produced by specific cleavage at the anticodon loop of mature tRNA. tRFs are further subdivided into tRF-1s, tRF-3s, tRF-5s, and internal tRFs (i-tRFs or tRF-2s), while tiRNAs are divided into 5 tiRNA and 3 tiRNA ( Figure 2) [81].
The tRF-1, also known as 3 U-tRF, originates from the 3 untranslated regions (UTR) of pre-tRNA through RNase Z digestion, with the characteristic of a poly-U sequence [82]. The tRF-5s are generated from cleavage in the D-loop or the arm region between the D-loop and anticodon loop of mature tRNA, and include the intact sequence of the 5 end of mature tRNA [83]. The tRF-3s originate from cleavage in the T-loop, and end with trinucleotides "CCA" [84]. The i-tRFs derive from an internal region of mature tRNA, and include the anticodon loop and part of the D-and T-loops [85] (Figure 2). Most human i-tRFs are 20 or 36 nucleotides long [86]. The tRF-5s occur predominantly in the nucleus; large numbers of tRF-5s are present in HeLa cell nucleoli [83]. tRF-3s and tRF-1 are more abundant in the whole cell fraction than the nuclear fraction, suggesting that both species occur exclusively in the cytoplasm [87]. Each class of tRFs is generated by specific ribonucleases and regulated by specific pathways.
One of the earliest discovered classes of tsRNAs were the stress (and starvation) induced tRNA fragments called tiRNA (tiR), or tRNA halves [88]. In mammals, tiRNAs are generated through cleavage by ribonuclease (RNAse (A or T)) angiogenin (ANG) within the anticodon loops of mature tRNAs [89]. This ANG cleavage produces two types of tiRNA: a tiRNA-5 and a tiRNA-3, which are the 5 and 3 half of mature tRNA, respectively [89] ( Figure 2). The production of tiRNAs is induced by stress, such as starvation, oxidative stress, heat shock, UV irradiation, or viral infection [88,90]. The upregulation of ANG under certain conditions positively correlates with increased tiRNA levels [91][92][93]. RNH1, an ANG inhibitor interacting with ANG in the cytoplasm, is a negative regulator of tiRNA generation [90]. Several studies have shown that methylation of mature tRNA by DNA methyltransferase DNMT2 or cytosine-5 methyltransferase NSun2 enhances tRNA resistance to ANG cleavage [94,95].
Both tRFs and tiRNAs play important roles in tumorigenesis [85][86][87]96,97], and are promising diagnostic biomarkers and therapeutic targets for cancer. tRFs can modulate protein translation and interact with ribosomes and aminoacyl tRNA synthetases [98,99]. In addition, tRFs can associate with Ago and PIWI proteins in a cell-type specific manner, potentially affecting gene expression ( Figure 3). Moreover, the interaction between RNAbinding proteins and tRFs has been linked with cancer development and metastasis [85,100]. Biogenesis and function of tRNA-derived small RNAs: Different types of tRNA-derived RNA fragments produced from either pre-tRNAs or mature tRNAs. The tRF-1 series is produced by RNase Z (or ELAC2) cleavage of the pre-tRNA during tRNA processing. Mature tRNA can be cleaved in the anticodon loop by angiogenin (ANG) to produce the 5 -tiRNA and 3 -tiRNA series under stress conditions. tRF-2 is a tRNA fragment containing an anti-codon loop generated by an unknown cleavage method. Cleavage in the T-loop results in the production of the 3 -tRF series. The 5 -tRF series is derived from the 5 -end of mature tRNAs by endonucleolytic cleavage and exonuclease digestion in the D-loop.
In cancer models, tsRNAs promote cell proliferation and cell cycle progression by regulating the expression of oncogenes and proto-oncogenes. Functional studies show that tRFs and tiRNAs may bind to RNA binding proteins such as Y-box binding protein 1 (YBX1) and prevent transcription, inactivate initiation factor eIF4G/A, promote translation of ribosomal proteins, activate aurora kinase A (the regulator of mitosis) by binding to cytochrome C, or promote the assembly of stress granules that helps cells survive under adverse conditions [105][106][107][108]. A recent study indicated that two specific tRFs derived from tRNA Lys-CTT and tRNA Phe-GAA are good indicators of progression free survival (PFS), and thus are candidate prognostic markers in prostate cancer [106]. The combination of tRF-3 with argonaute 3 (Ago3) and Ago4 binding to mRNA allows mRNA-degrading enzymes to degrade the target mRNA [101]; tRFs function like miRNAs to inhibit cancer-associated gene expression. Binding with the 3 untranslated region (3 UTR) of target mRNA, Argonaute (Ago) protein and other proteins form an RNA-induced silencing complex (RISC). (B,C) Inhibition of translation: By binding to small ribosomal subunits, tRF inhibits peptidyl transferase activity that results in reduced protein abundance of the target gene [99,102]. 5 tiRNA inhibits translation by forming a RNA G-quadruplex (RG4s) that replaces the translation initiation complex eIF4G/eIF4E on the mRNA cap [103,104].

tRNA Derived Small RNAs in Gynecologic Cancer
A significant impact of the deregulated tsRNAs has been demonstrated in various malignancies, including gynecological cancers; these biological functions of tRFs are Agodependent. In ovarian cancer tRF5 Glu regulates breast cancer anti-estrogen resistance 3 (BCAR3) mRNA levels by direct binding to the 3 untranslated region (UTR) [109]. In colon, breast, and ovarian cancer patients, as well as corresponding cell lines, the expression level of ts-101 and ts-46 (tRF-1s) correlates with chromatin structure, cell survival, cell proliferation, clonal growth, and apoptosis. The expression of tRFs also correlates with oncogene activation and ovarian cancer progression [110]. Reanalysis of existing RNA-sequencing data, from 180 serum samples, including 15 healthy controls, 46 benign tumors, 22 borderline tumors, and 97 ovarian cancer patients, revealed that tsRNAs cover a high proportion of total small RNA, and are non-random degradation products in serum (ranging from 2.5-29.4%), and which are enriched in several specific types of related tRNA (e.g., Gly-tRNA), and can predict abnormal cell proliferation with high accuracy [111]. Another group using serum samples from ovarian cancer patients and healthy donors, along with ovarian cancer cell lines, have shown differential expression of tRF; they showed that tRF-03357 promoted SK-OV-3 cell proliferation, migration, and invasion, as well as downregulating HMBOX1 [112]. In cervical cancer, preliminary studies using biopsy samples demonstrated that the expression of 5S rRNA, tRNA Arg , and tRNA Sec was significantly elevated in the HPV16-containing samples when compared to the HPV-negative biopsies [113]. In Hela cells, 5 tRFs derived from tRNA Gln are produced in abundance [83]. Currently, tRNA/tiRNA status in endometrial cancer remains unexplored.

Future Perspectives
tsRNAs are unique sequences derived from tRNA precursors, and generated in the nucleus. tsRNAs are frequently dysregulated in various cancers, including the gynecologic malignancies. Since tsRNAs can accompany both Ago proteins (like miR) and PIWI proteins (like piRNAs), they can regulate gene expression both pre-transcriptionally (like piRNA) and post-transcriptionally (like miR). Like piRNAs, tsRNAs are produced as singlestranded molecules, and can interact with DNA and the histone methylation machinery, suggesting a role in the pre-transcriptional regulation of gene expression. Like miRNAs, ts-53 (previously known as miR-3676) interacts with the 3 UTR of TCL1, supporting a role for tsRNAs in the posttranscriptional regulation of gene expression [81,114].
Rapidly proliferating tumor cells often overcome a deficient blood supply, resulting in a microenvironment with limited oxygen and nutrients. Tumor cells adapt to this stress with varying strategies, thus ensuring survival and proliferation [115]. Generation of tsRNA from tRNAs under stress is an important pathway, and the biological function of tsRNA mainly supports cell survival under stress. Additionally, tsRNAs can be detected in the urine and serum from cancer patients [111,[116][117][118], suggesting their potential as molecular diagnostic markers. For example, high-throughput RNA sequencing in breast cancer patients showed that tsRNA blood levels closely relate to the pathological characteristics [119]; similarly, in breast cancer and prostate cancer, hormone-dependent tsRNA are expressed in abundance and enhance the proliferation of cancer cells [91]. Thus, a tsRNA database of different tumors may represent a new diagnostic tool for cancer management [114,120]. In summary, tsRNAs are implicated in tumor onset, progression, and drug response, and thus represent potential therapeutic targets and/or diagnostic markers.

Micro-Ribonucleic Acid (miR)
Micro-ribonucleic acids (miRs/micro-RNAs) are short (18-25 nucleotides), evolutionarily conserved, and endogenously expressed regulatory RNA molecules, which belong to the family of ncRNAs. The miRs were first detected in the early 1990s in Caenorhabditis elegans [121], and later studies confirmed their presence in almost all species [122,123]. Although most miRs modestly alter expression of their target genes, the intricate network of miR target genes and downstream effectors plays a profound role in the regulation of biological pathways [124][125][126].
Among the non-coding RNAs, miR are the most widely studied, and apart from classical 3 mRNA targeting and cytosolic expression, miR are also known for their nuclear expression [127]; the biogenesis and mRNA regulatory functions of miR have been extensively reviewed by Jacob O'Brien and others [30,31,128,129]. In brief, miRs are synthesized in the nucleus by DNA polymerase II [130] as a long double-stranded precursor called pri-miR. This pri-miR is cleaved at specific sites by the RNAse drosha inside the nucleus, producing a precursor miR (pre-miR) [131]. The pre-miR is exported to the cytoplasm by the exportin 5 protein, where it is processed by dicer into mature miR. Mature miRs are then activated through binding to the RNA-induced silencing complex (RISC); via the RISC, miRs can regulate their target mRNAs, leading to translational repression or degradation. The sequence at the 5 end of the mature miR is called the "seed region", or "seed sequence". The seed sequence of the miR binds the complementary sequence within the 3 untranslated region (3 UTR) of target mRNAs [131,132]. Some miRs also interact with the 5 UTR [133], coding sequence [134], and promoter regions of their targets [135]. Perfect or near-perfect complementarity between the miR and its mRNA target results in mRNA degradation, while imperfect complementarity leads to translational inhibition [136]. Several parameters, including the subcellular location of the miR, the quantity of both miR and its target mRNA, and the affinity, modulate the miR-mRNA interaction [128].
miRs are the critical modulators that regulate gene expression at the post-transcriptional level; they govern the stability and translation of protein coding mRNAs and are thus involved in almost every biological process, including regulation of cell division, dif-ferentiation, growth, and apoptosis. Dysregulation of miRs is instrumental in various pathophysiologies including neurodegenerative, inflammatory, metabolic, and cardiovascular diseases and, significantly for this review, cancer [137,138]. Altered expression of hundreds of miRs is closely associated with tumor development, invasion, metastasis, and drug resistance in cancer [139,140]. The miR expression profile differs significantly between normal and cancerous tissues, localized and aggressive cancer, and across type and stages of cancer.
Differential miR expression in normal and tumor cells has been reported in numerous studies, and suggests a robust association of altered miR expression with cancer pathogenesis and progression [141,142]. More than half of all annotated human miR genes are located in cancer-associated genomic regions that are amplified, deleted, or translocated in cancer [143]. Mechanistically, miRs regulate tumor suppressor genes and oncogenes; based on the regulated gene, miRs can be either oncogenic or tumor suppressive. The altered expression of miRs that control tumor suppressor genes and oncogenes results in cancer. Dysregulated miR expression has several causes, including genetic alteration, epigenetic changes, and SNPs in miR coding genes, as well as defects in factors regulating miR biogenesis.

miR in Gynecologic Cancer
miRs are estimated to control over 50% of the activities of all protein coding genes [144], and are involved in regulation of almost all cellular processes [145]. Multiple studies have shown that dysregulation of miRs leads to a variety of human diseases, including cancer [146]. The relationship between miR expression and gynecologic cancer is well documented [147][148][149][150], but the significance of the cumulative effect of miR expression may not have been fully realized as each miR targets multiple genes associated with various cellular processes. Herein we summarize the key research findings on the significance of miR in gynecological cancer progression, prognosis, and diagnosis. We also present a compressive analysis of the miR expression in ovarian, endometrial, and cervical cancer, grouping them based on their expression. These miR are further analyzed using a systems biology approach for target prediction, and these target proteins are then analyzed for associated pathways to find the significance of altered miR in respective gynecological cancers. Based on the number of predicted target genes and associated pathways, miR can be best prioritized for possible disease interventions. A text-mining program called IRIDESCENT [151] was used to document details of the different miRs associated with ovarian, endometrial, and cervical cancer. IRIDESCENT identified the co-occurrences of miRNAs and gynecological cancer within MEDLINE titles and abstracts by analyzing public databases (e.g., Entrez, OMIM, and Disease Ontology) for a thesaurus of names and synonyms of miRNAs and diseases. The strength of association between gynecologic cancer and miR was calculated by the frequency of co-mention, (+0.5 for every abstract, +0.8 for every sentence). Associations and directionality, i.e., up-or downregulation, was confirmed manually for the respective types of cancer. Deregulated miRs were further grouped based on their expression in respective cancers, and potential significance was analyzed using miRNET [152].

Role of miRs in Ovarian Cancer Pathogenesis
Ovarian cancer is characterized by wide-scale deregulation of miRs, and aberrant expression of miRs in ovarian cancer is known to correlate with histotype, lymphovascular, and organ invasion [153]. We have reported that miR-15a and miR-16 are under-expressed in ovarian cancer tumor samples compared to normal tissue, and that ectopic expression of these miRs significantly inhibits ovarian cancer progression in preclinical tumor models [154,155]. Additionally, we have demonstrated that miR-195 is under-expressed in ovarian cancer and regulates ovarian cancer progression by regulating the expression of Mitochondrial Calcium Uptake 1 (MICU1) [156]. miR have been shown to play a key role in all the stages of tumorigenesis, and are also significant for diagnosis, prognosis, and therapeutics. Differential miR expression has been shown to be associated with ovarian cancer progression, omental metastasis, and drug resistance. Compared to normal ovarian tissue, miR-200a, miR-141, miR-200c, and miR-200b were significantly upregulated, whereas miR-199a, miR-140, miR-145, and miR-125b1 were among the most downregulated miRNAs in ovarian tumor tissues [153]. In another study miR-146a and miR-150 were reported to be significantly associated with omental metastasis and cisplatin resistance, and were shown to induce spheroid formation [157]. miR-200 and miR-429 are associated with recurrence and survival rates of ovarian cancer; their increased expression inhibits cancer metastasis [158].
Presently, miRs are being evaluated for their potential utility as therapy candidates. In our previous reports, we have shown that the nanoliposomal delivery of miR-15a and miR-16 in combination, in a pre-clinical chemo-resistant orthotopic mouse model of ovarian cancer, demonstrated a striking reduction in tumor burden compared to cisplatin alone [154], while the importance of miR in cancer therapy has been reviewed previously [169,170].
Here, we have listed micro-RNAs deregulated in ovarian cancer using the text-mining program IRIDESCENT. The deregulated miRs were manually grouped by their expression levels compared to the control group (Table 1). Fifty-three miRs were reported to be upregulated, and sixty-eight miRs were downregulated in ovarian cancer. The miRNet web tool (https://www.mirnet.ca/ accessed between December 2020 and January 2021) was used for microRNA-gene target prediction (using miRTarBase v8.0 [171], and the parameters of degree and betweenness were used to derive the network, setting the cutoff for the degree filter at five. The network was further used to elucidate the biological processes and pathways of these upregulated, and downregulated targets, and networks are presented in Figures 4 and 5. In the ovarian cancer upregulated miR group, a total of 7605 gene targets were found; of special note miR-20a-5p regulates 14.1% and miR 106a-5p regulates 9.4% of target genes. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis enrichment was performed for these miR targets genes (degree cutoff 5); which revealed 24 significantly enriched pathways (cutoff p > 0.001), these altered pathways are known for their significant role in cancer progression, ( Table 2). Gene ontology enrichment for biological process (GO-BP) and molecular function (GO-MF) were also analyzed; using p > 0.001 as the cutoff, GO-BP and GO-MF respectively identified 67 and 24 pathways as enriched, and the top 10 pathways are shown in Table 2.
Similar analyses for the downregulated miRs in ovarian cancer ( Figure 2) identified 9287 gene targets. miR-26b-5p, miR-519d, miR-15a, and miR-15b regulated 20.8%, 11.3%, 8.6%, and 8.9% of the target genes, respectively. KEGG, GO-BP, and GO-MF enrichment analysis of the targets respectively identified 41, 95, and 38 enriched pathways. The top 10 pathways for each analysis are shown in Table 2.     While mir-21 and mir-155 are the most studied miRs in ovarian cancer, our analysis suggests that the genes targeted by miR-20a-5p and miR-26b-5p may play a major role in the progression of ovarian cancer (Figure 2). The most prominent targets of these miRs are phosphatase and tensin homolog (PTEN), cyclin dependent kinase inhibitor 1A (CDKN1A), MDM2 proto-oncogene (MDM2), superoxide dismutase 2 (SOD2), high mobility group box 1 (HMGB1), insulin like growth factor 1 receptor (IGF1R), WEE1 G2 checkpoint kinase (WEE1), forkhead box K1 (FOXK1), and thioredoxin interacting protein (TXNIP), which play a crucial role in ovarian cancer progression. From the pathway enrichment analysis, these deregulated miRs were shown to be involved in essential cellular processes and key signaling pathways, i.e., cell proliferation, regulation of cell cycle and apoptosis, adherens junction, focal adhesion, regulation of actin cytoskeleton, regulation of transcription, p53 signaling pathway, TGF-beta signaling pathway, Erbb signaling pathway, neurotrophin signaling pathway, Wnt signaling pathway, Jak-STAT signaling pathway, and insulin signaling pathways ( Table 2). Our systemic analysis of these deregulated miRs, their target genes, and the target pathways, which are instrumental in cancer progression, reinforced the importance of these miRs in ovarian cancer, and may help prioritize miRs as targets for therapeutic evaluation.
Here, we have compiled a list of micro-RNAs that are deregulated in endometrial cancer using the text-mining program IRIDESCENT (Table 3). We report nineteen upregulated and twenty-seven downregulated miRs in endometrial cancer. The miRNet web tool (https://www.mirnet.ca/) was used for microRNA-gene target prediction, and the parameters of degree and betweenness were used to draw the network, and were further used to elucidate the predicted biological processes and pathways of these upregulated, and downregulated targets; the networks are presented in Figures 6 and 7. In endometrial cancer, upregulated miR were predicted to target a total of 4224 genes. Among these target genes, 14.5% and~26% are targeted by miR-21-5p and miR-106a-5p, respectively. Pathway enrichment analysis was performed on the target genes using KEGG, GO-BP, and GO-MF, as described above, and these methods respectively identified 32, 96, and 34 enriched pathways. A similar analysis was performed using the downregulated miRs; 5749 gene targets were identified, a network was generated with degree filter 5, and miR-195-5p and miR-424-3p respectively regulated 11.1% and 8.2% of the target genes. KEGG, GO-BP, and GO-MF enrichment analysis respectively identified 57, 100, and 97 pathways. The top 10 KEGG pathways, biological processes, and the molecular functions altered by the miRs are shown in Table 4. A literature search showed miR-205 to be the most studied miR in endometrial cancer, however, our target analysis showed miR-21-5p and miR-106-5p to be the most interconnected miRs that can regulate multiple targets responsible for endometrial cancer progression. The most prominent targets of these miRs are PTEN, IGF1R, Notch Receptor 2 (NOTCH2), cyclin D1 (CCND1), TXNIP, AGO2, and cyclin dependent kinase 6 (CDK6). The pathway enrichment analysis showed that these deregulated miRs are involved in critical cellular processes and key signaling pathways, i.e., cell cycle, proliferation, programmed cell death, cellular response to stress, p53 signaling pathway, mTOR signaling pathway, insulin signaling pathway, notch signaling pathway, NOD-like receptor signaling pathway, inositol phosphate metabolism, and adipocytokine signaling pathway (Table 4), all of which support the importance of these miRs in the progression of endometrial cancer.   Role of miRs in Cervical Cancer Pathogenesis miR deregulation plays a key role in the malignant transformation of cervical cancer, and, along with their target genes, miRs have been exploited for both prognostic and therapeutic strategies in cervical cancer. HPV causes about 91% of cervical cancers [354], and the high-risk HPV strains HPV-16 and HPV-18 in keratinocyte cell lines have been shown to regulate expression of different miRs [237]. The study reports that both HPV strains induced the expression of miR-16, miR-25, miR-92a, and miR-378, while they inhibited the expression of miR-22, miR-27a, miR-29a, and miR-100. Moreover, data from cervical-tissue specimens correlated with in vitro results; miR-25, miR-92a, and miR-378 were increased with cervical cancer progression, which supports the significance of HPV mediated miR expression in cervical cancer pathogenesis [237]. Mounting evidence supports the significance of miR expression in cervical cancer, e.g., miR-21, which is reportedly overexpressed in various malignancies [355], promotes cell proliferation in HeLa cervical carcinoma cells by targeting programmed cell death 4 (PDCD4) expression. In addition, miR-20a [356], as well as miR-106a [357], suppressed the migration and invasion of cervical cancer cells by targeting TIMP2.  Pathway enrichment analysis for deregulated miR in endometrial cancer: Gene targets for miR were predicted using the miRNet web tool (https://www.mirnet.ca/) and miRTarBase v8.0. (A,D) KEGG pathway analysis enrichment, gene ontology enrichment for (B,E) biological process (GO-BP) and (C,F) molecular function (GO-MF) were analyzed for upregulated (A-C) and downregulated (D-F) miR target genes, the 10 most significant pathways based on the network generated using degree cutoff 5 are shown.
The potential of miR for early detection and as a prognostic biomarker in cervical cancer has also been explored. Expression profiles of circulating miRs in cervical cancer patient samples compared to healthy volunteers demonstrated that the expression of miR-20a and miR-203 was upregulated in the cervical cancer patients' sera, and circulating levels of miR-20a could be useful for the detection of lymph-node metastasis [358]. Various reports have shown the correlation of miR expression with cervical cancer stage; miR-494 is upregulated, while the expression of miR-195 and miR-144 was downregulated in patients with stage IB cervical carcinoma [359][360][361]. In Stages I and II, the levels of miR-375, miR-145, and miR-124 were downregulated, while miR-99a/b, miR-92a, miR-150, and miR-21 levels were upregulated [359,[362][363][364][365][366][367][368]. In cervical cancer patients, levels of miR-218 were similarly decreased in plasma and tumor samples, and levels correlated with tumor stage [369]. Differentially expressed miRs in the serum of cervical cancer patients, compared with controls, measured using Solexa sequencing, identified 12 miRs of interest. Five were upregulated, i.e., miR-21, miR-29a, miR-200a, miR-25, and miR-485-5p, and likely are credible alternatives to current tumor markers, such as squamous cell carcinoma (SCC) antigen and carbohydrate antigen 125 (CA125) [370]. Elevated serum levels of miR-205 correlate with both cervical cancer tumor stage and decreased survival rates. Some of the aforementioned miRs may be viable diagnostic and prognostic biomarkers for cervical cancer progression. Circulating levels of miR-646, miR-141, and miR-542-3p are significantly different when assayed before and after surgical procedures [371], and may be useful in monitoring the health status of patients after treatment.
Here, we have compiled a list of micro-RNAs deregulated in cervical cancer (Table 5); thirty-six miRs were reported to be upregulated and seventy downregulated in cervical cancer. The miRNet web tool, as discussed above, was used for target gene prediction, and parameters of degree and betweenness were used to draw the network, and were further used to elucidate the predicted biological processes and pathways of the upregulated, and downregulated targets. The miR and the target protein networks are shown in Figures 8 and 9. In cervical cancer, target prediction using upregulated miR resulted in 6957 gene targets; mir-106b-and miR-106a respectively regulate 15.7% and 10.3% of the target genes. Networks for these targets were generated and pathway enrichment analysis performed using KEGG, GO-BP, and GO-MF parameters, as before; they identified 19, 94, and 10 enriched pathways, respectively (Table 6). A similar analysis with the downregulated miR, generated 9715 gene targets, of which miR15a-5p targeted 7.4%, and miR195-5p regulated 6.6%. KEGG, GO-BP, and GO-MF enrichment analyses were applied and respectively identified 51, 96, and 55 pathways ( Table 6). From our collective analysis of miR and their target genes in cervical cancer based on the number of target genes and associated networks, the most important upregulated miRs were miR-106, miR-20a, and miR-519d, while the most important downregulated were miR-15a and miR-195. The most prominent targets of these miRs are PTEN, CRK proto-oncogene, adaptor protein (CRK), SOD2, TXNIP IGF1R, BCL2 apoptosis regulator (BCL2), and FOXK1, which play a crucial role in cervical cancer progression. Based on the network analysis these deregulated miRs are involved in important cellular pathways and key signaling processes, i.e., cell cycle, adherens junction, pyrimidine metabolism, focal adhesion, cell proliferation, nuclear transport; and the p53 signaling pathway, Wnt signaling pathway, PPAR signaling pathway, G1 phase of mitotic cell cycle, ErbB signaling pathway, insulin signaling pathway, and MAPK signaling pathway ( Table 6). All these cellular processes and signaling pathways are instrumental in the progression of cervical cancer, and further substantiate the significance of these miRs in cervical cancer.

Future Perspectives
Ongoing research reveals the diagnostic potential and therapeutic promise of miRs for multiple cancers, including those of the female reproductive system. However, their diagnostic and/or therapeutic translation in terms of helping patients in need has not been fully apprehended, and before this becomes a reality, our understanding of the diverse roles of miRs must become complete. There is a certain level of redundancy in gene regulation by miRs, as a single miR can target multiple genes, and a single gene can be the target for multiple miRs. Thus, any miR selected as a therapeutic target must be comprehensively investigated, all target genes must be identified, and the outcome(s) of their combined inhibition should be carefully evaluated. Conflicting reports on the functionality of a single miR within a specific gynecologic cancer, as well as across different gynecologic cancers, suggest that we do not fully understand the complex nature of miR-mediated regulation, which may be tissue and cell-context specific. As an example, miR-26a was first shown to promote ovarian cancer proliferation through its suppression of ER-α [481], while subsequently, it was shown to inhibit proliferation of ovarian cancer cells through its regulation of CDC6 [482]. In a cervical cancer model, miR-26a inhibited cell proliferation, migration, and invasion in vitro, and also inhibited tumor growth in vivo in a xenograft model [483]. Similarly, miR-31 [484][485][486], miR-214 [181,487], miR-494 [361,488,489], and miR-222 [490,491] have different reported roles in different gynecologic cancers. This conflict arises from the targeting of multiple genes by the same miR, and the effects must be evaluated at a systems level using systems biology approaches in order to translate miRs into successful cancer therapies. Enrichment analysis based on the validated/predicted targets, as we have shown herein, will undoubtedly help in selecting the best miRs for a specific diagnostic, therapeutic, or predictive purpose. Concurrent with efforts to overcome the inherent barriers to miR use, strategies to manipulate miR levels in cancer patients should be developed further. This will permit the rapid translation of constructive findings to the clinic for the benefit of millions of gynecologic cancer patients worldwide.

Conclusions
We remain in the early stages of decoding the essential role of the small noncoding RNAs which were previously believed to be transcriptional debris. It is now well established that these small non-coding RNAs play a key role in gene regulation, and have a profound role in the development of multiple cancers, including gynecologic malignancies. Our understanding of the roles of piRNA-and tRNA-derived small RNAs is still very limited; interestingly, much valuable information is available for miR, and a few studies are at the preliminary stages of development for cancer therapy, with some clinical trials in progress for ovarian cancer (ClinicalTrials.gov Identifier: NCT03738319, NCT03776630, NCT01970696, NCT02253251, NCT01391351, NCT03877796), endometrial cancer (ClinicalTrials.gov Identifier: NCT02983279, NCT01119573, NCT03824613), and cervical cancer (ClinicalTrials.gov Identifier: NCT04087785). Based on this accumulating evidence, it is possible that in the near future miRs in plasma may be used for the diagnosis and prognosis of various cancers. One limitation of these miR expression data may be the lack of evaluation among diverse ethnic groups; however, as more information about the diverse types of small non-coding RNA and their target networks become available, it will assist us in refining their possible diagnostic and therapeutic applications. The primary causes of greater cancer related mortality, such as in ovarian cancer, are late diagnosis and acquired resistance; advancing the use of small noncoding RNA in cancer diagnosis and therapy will enhance early detection and therapeutic intervention, and will certainly be a significant step forward in the management of gynecologic malignancies.