microRNAs and long non‐coding RNAs as biomarkers for polycystic ovary syndrome

Abstract Polycystic ovary syndrome (PCOS) is known as the most common metabolic/endocrine disorder among women of reproductive age. Its complicated causality assessment and diagnostic emphasized the role of non‐coding regulatory RNAs as molecular biomarkers in studying, diagnosing and even as therapeutics of PCOS. This review discusses a comparative summary of research into microRNAs (miRNAs) and long non‐coding RNAs (lncRNAs) that are molecularly or statistically related to PCOS. We categorize the literature in terms of centering on either miRNAs or lncRNAs and discuss the combinatory studies and promising ideas as well. Additionally, we compare the pros and cons of the prominent research methodologies used for each of the abovementioned research themes and discuss how errors can be stopped from propagation by selecting correct methodologies for future research. Finally, it can be concluded that research into miRNAs and lncRNAs has the potential for identifying functional networks of regulation with multiple mRNAs (and hence, functional proteins). This new understanding may eventually afford clinicians to control the molecular course of the pathogenesis better. With further research, RNA (with statistical significance and present in the blood) may be used as biomarkers for the disease, and more possibilities for RNA therapy agents can be identified.

| 655 TAMADDON eT Al. and there is a growing clinical need for biomarkers that are both sensitive and specific enough for PCOS diagnosis. 9 Biomarker research can also enable clinicians to achieve earlier diagnosis, molecularly subtype the disease in the clinic and cast light on the underlying molecular mechanisms of the disease and its derivatives, mainly cancer and diabetes. 10 In modern medicine, the multilateral relationship between the RNAs and different proteins determines how molecular disease conditions change functions. We know that RNAs have regulatory functions in the way of protein production. It may follow that the disorders in protein levels that lead to many diseases could have first happened at the RNA level. Earlier changes could be diagnosed in the RNA level as an early di- ing candidates as molecularly significant biomarkers of disease. 11,12 RNAs vary greatly in roles, size, conformation and sequence.
Non-coding RNAs (ncRNAs) (not translated into proteins) with complex regulatory functions have gene and protein regulatory functions. Two main groups are miRNAs with about 22 nucleotides and long non-coding RNAs (lncRNA). 13,14 Both have roles in the cell's physiology, especially in post-transcriptional regulations and could have a signature on every single disease or any biological malfunction. Some present in blood could be easily accessible through different body fluid samples. 15,16 Hence, their role has been discovered as a diagnostic biomarker for many diseases such as PCOS 17 and cancer. 18 Through advances in sequencing technology and systems biology over the past decade, their application in discovering the molecular mechanism of complex diseases such as PCOS has become higher, leading to the discovery of new precise and feasible biomarkers. 19,20 Figure 1 represents the biomarkers for PCOS analysis.
In this review, we first overview the prominent research groups working on miRNAs and PCOS, and then those working on lncRNAs.
We follow with a discussion of combinatory and creative studies into the molecular basis of PCOS. We discuss how each research approach can be helpful on its own or when combined or integrated with other available methodologies. We also discuss the innovative methods and approaches that allowed leading research groups to gather key findings and share our insights into how these findings may determine future research into the molecular mechanisms and biomarkers for PCOS. This article also includes descriptive finding data for the found RNA biomarkers.

| miRNAs and lncRNAs
The functional molecular role of RNAs has been observed in many key processes, such as their enzymatic role in protein translation at ribosomes. 21 Any other types of RNA, other than coding RNAs, that translate to proteins are called non-coding RNAs, which can either be a direct transcript of a gene or a messenger RNA intron. miRNAs are short non-coding single-stranded RNAs (22-23 nucleotides (nt) long), with an important role in gene regulatory processes. 22 They act like core elements in the transcriptome, mediating correlations between different genes representing RNAs or proteins. They can bind to various RNAs and biological sites in order to regulate cell functions. 23 Besides intra-cellular functions, inter-cellular relationships are also heavily dependent on miRNAs, and hence their presence is abundant in body fluids, and they have become a hotspot in biomarker research, particularly for PCOS. 11,12 Long non-coding RNAs, on the other hand, are normally more than 200nt long and can have either linear or circular conformation.
lncRNAs are core to many cellular functions such as gene regulation, transcription, chromatin modification and epigenetic regulation. 24 According to a recent discovery, there are many more lncRNAs in line to be discovered or functionally understood. With the presence in exosomes and being a means of cellular crosstalk overall, they are good candidates for being a biomarker in body fluids, just like miRNAs. 25 A wide range of RNA therapies are in the line for coming years, with applications in personalized medicine. 26 The technology to synthesize desired nucleotide sequences in vitro has achieved recent worldwide prominence in synthesizing an RNAbased COVID-19 vaccine.

| miRNAs as biomarkers for PCOS
One of the major symptoms of PCOS is the presence of numerous small cysts in the ovaries. One method used to diagnose PCOS in the clinic is internal sonography of the ovaries. 5,27 Sonography, however, normally might be uncomfortable for the patient and cannot distinguish between different disorders that come with cysts. Blood tests to determine testosterone levels are another common clinical approach to diagnose PCOS. However, high levels of hormones such as testosterone are only observed in the late stages of PCOS, where the patient is at an elevated risk for developing a diabetic response.
Genetically, early diagnosis with hormone levels is nearly impossible.
In general, there are two main approaches to validate a reliable biomarker for the disease. One is to use the potential of sequencing and blotting to differentially investigate a pool of possible RNAs and proteins in two populations of healthy and PCOS affected cells and patient blood samples or follicular fluid. From this pool, the most significant biomarkers are selected. The second is to investigate the molecular mechanism and biological significance of a known ncRNA biomarker for the syndrome.
As far as finding new biomarkers is concerned, there is diversity but promise in the results. Che et al 28 identified a pool of 55 differentially expressed miRNAs in PCOS conditions, from which the most significantly differentiated (miR-27a-5p) had a strong correlation with the incidence of cancer in patients with PCOS. In studies similar to this one, the researchers compile a set of differential data with various statistical significances. Key meta-analysis has been performed on these data sets by Deswal et al. 29 They used an initial group of 79 miRNAs from 21 studies, reported to be differentially expressed, and only three of which were reported in more than three studies. After the meta-analysis, they reported miR-29a-5p and miR-320 as significant biomarkers for PCOS. They moved further in their meta-analysis, and for the three most significant markers found, they performed a genetic and functional analysis, as shown in Figure 3.
These key meta-analyses determine the significance of results by evaluating their replicability in the literature, preventing the possible scientific error from propagation by eliminating the unrepeated biomarkers. This kind of noise cancellation is vital in research, especially when we will use the found biomarkers for further genetic and functional analysis.
Several statistically significant miRNAs have already been identified as candidates for a clinical PCOS biomarker. The most statistically significant miRNAs included 381-3p, 29a-5p, 93, 320, 3188, 612, 509-3p, 547-3p, −5p, 20-3p and −5pn (for a complete list, refer to Figure 1, Table 1 and Table 2). In order to further verify the scientific significance of each biomolecule, researchers must investigate its molecular role or path of efficacy. This is the F I G U R E 1 Biomarkers for PCOS, a schematic for how molecularly disease condition affects biomolecule levels, from cells to intercellular connections, such as gap junctions between cumulus and granulosa cells. Cross-talks between ovarian tissue cells and oocytes and exosomes mediating the intercellular networks, all exposed to follicular fluid and hence blood serum environment approach that most groups have taken when researching PCOS at the molecular level.
Deswal et al. have demonstrated that functional analysis can be effective, even without discovering a new biomarker for the disease.
They performed a bioinformatics analysis of recent findings in the field. After selecting three significant miRNAs from the existing literature, they investigated the pathways and functions related to them. All three miRNA biomarkers were strongly correlated with the insulin cycle (Figure 2A), adding further evidence for a meaningful correlation between insulin metabolism and PCOS conditions. This correlation cannot easily be called causation, and even if causation is to be determined, it could be in either direction between diabetic and PCOS conditions.
Research into PCOS gene ontology and pathway encyclopaedia analysis has also integrated functional analysis. 30 Figure 2B). These results were not individualistic and were corroborated with many other groups in sharing common functions such as cell proliferation and follicular development. 33,34 Recent publications report correlations between various RNAs and signaling networks, working towards identifying possible complex protein-RNA matrices. 30 Classically, we thought of signaling networks as solid protein expression rules. To recapitulate, we specifically investigated the proteins when studying their relationships and called them signaling networks. From that standpoint, numerous studies correlated miR-NAs and lncRNAs to key factors in some classical signaling networks.
The idea is only to correlate some types of RNAs to some key proteins, according to their significance in classical signaling networks.
For example, LncRNA-MALAT1 has been correlated to TGFβ signaling, 36 exosomal circLDLR to Jak-STAT 37  However, this has to be performed through more experimental investigations and big data build-up and analysis.
As drugs are normally agents that oppose the molecular mecha-    -No lncRNA correlated with anti-mullerian hormone (AMH) levels, insulin resistance (HOMA-IR) or the free androgen index (FAI). -LncRNAs differ between anovulatory PCOS and control women in the follicular phase of the menstrual cycle Tan et al. 83 LncRNA SRA1 -LncRNA SRA1 gene single-nucleotide polymorphism correlated to polycystic ovary syndrome Jiao et al. 84 -LncRNA and mRNA profiles in follicular fluid from mature and immature ovarian follicles of healthy women and women with PCOS, construction of the mRNA/ lncRNA network -Good example of systematic transcriptome-wide analysis Fawzy et al. 85 Circ-LncRNAs: H19, GAS5 -Also, associated with type 2 diabetes Huang et al. 44 LncRNA-PWRN2 -Construction of a lncRNA-PWRN2-ceRNA network suggests its potential roles in oocyte nuclear maturation in PCOS patients Wu et al. 89 Lnc-OC1 -Its downregulation associated with PCOS, in granulosa cells Lin et al. 90 LncRNA GAS5 -Downregulation of lncRNA-GAS5 may contribute to insulin resistance in PCOS patients -From serum Liu et al. 91 LncRNA-Xist -Xist downregulation may be involved in PCOS and is correlated with adverse pregnant outcomes in PCOS -From serum Wang et al. 92 lncRNA-H19 -high-throughput lncRNA sequencing of follicular fluid exosomes in non-PCOS infertility patients and PCOS infertility patients -In exosomes from follicular fluid -lncRNA-H19 represented the largest node and was predicted to have the potential to interact with 15 target miRNAs
Yang et al. 95 LncRNA-BANCR -Role in PCOS by promoting apoptosis in granulosa cells -From cells of IVF patients Wang et al. 54 LncRNA-GAS5 -LncRNA-GAS5 is upregulated in polycystic ovary syndrome and regulates cell apoptosis and the expression of IL−6 in granulosa cells -From blood plasma Sun et al. 53 lncRNA-H19 -lncRNA H19 acts as a ceRNA to regulate the expression of CTGF by targeting miR−19b in PCOS -On KGN cell line -H19 could promote cell proliferation and decrease cell apoptosis Zhang et al. 36 LncRNA-MALAT1 -lncRNA-MALAT1 is involved in the pathogenesis of PCOS through TGFβ signaling in granulosa cells -A nice biomolecule for possible future RNA therapy, repeated in literature Wang et al. 107 LncRNA-H19 -Metformin and sitagliptin combination therapy is effective for PCOS with insulin resistance through upregulation of lncRNA-H19. To summarize, co-treatment induced H19 expression via suppressing the PI3K/AKT-DNMT1 pathway.

| lncRNAs as biomarkers for PCOS
Efforts for better insight into the competing endogenous(ce)RNA networks affecting PCOS have also led to higher knowledge of the molecular networks and functions in the endocrine and female reproductive system overall. Overall, the literature indicates that lncR-NAs play a key role in making up endogenous networks of bilaterally effective relationships between proteins, nucleic acids and various other types of RNAs (mainly mRNAs and miRNAs). 24 As a biomarker such as miRNAs, lncRNAs will be significant in accordance with their central regulatory role in the transcriptome. 43 Various groups have    showed that oocyte lncRNAs could have roles in chromatin remodeling, cell pluripotency and also in the early development of the embryo. They did so by analyzing the differential lncRNA expression profiles in human oocytes and cumulus cells. A manner pretty widespread among groups, as a list of prominent studies regarding lncRNAs and PCOS with similar methods, is available in Table 2.

| CON CLUS ION
Polycystic ovary syndrome (PCOS) is a disorder in which women of reproductive age suffer, causing disorder in levels of hormones.
Having direct relation with infertility issues and causing long-term health damage by mechanism (like increasing the chance for diabetes), PCOS comes with an excessive need for early and precise diagnosis. The correlation between hormone levels and PCOS conditions indicates that endocrine cellular function in epithelial levels has been significantly affected. This opens up the possibility for developing diagnostic biomarkers and miRNAs and small non-coding RNAs for use in a clinical setting. In this study, we review recent research into how the syndrome exploits miRNAs and lncRNAs in its molecular path and whether this presence and effect makes RNA a biomarker for the diagnosis of PCOS. We also compared the type and methodology of research for prominent groups and their pros and cons. We make the following suggestions for future research: First, there is a need for more metaanalysis to summarize the statistical data of all the sequencing data from various groups. That is essential in eliminating the errors inherent in sequencing and any bias in selecting the biomolecules of interest in different groups. With rapid sequencing technology, it is not too optimistic to expect reliable RNA biomarkers in clinical use worldwide in the near future. Second, a great deal of research is needed to find more combinatory RNA networks formed of different types of known RNAs in combination with proteins and functional signaling networks. This is where a great insight for PCOS therapeutics is observed. With more unveiling of the molecular mechanism of PCOS, there will be significant core RNAs discovered as the best candidates for RNA immunotherapy development.

ACK N OWLED G M ENTS
Authors are thankful to all the supports provided by TUMS.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing not applicable-no new data generated for the review article.