Edited by: Junbai Wang, Oslo University Hospital, Norway
Reviewed by: Toni Ibrahim, Romagnolo Scientific Institute for the Study and Treatment of Tumors (IRCCS), Italy; Luisella Bocchio Chiavetto, Centro San Giovanni di Dio Fatebenefratelli (IRCCS), Italy
†These authors have contributed equally to this work
This article was submitted to Genomic Medicine, a section of the journal Frontiers in Genetics
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
Osteoporosis (OP) is a systemic bone disease with a series of clinical symptoms. The use of screening biomarkers in OP management is therefore of clinical significance, especially in the era of precision medicine and intelligent healthcare. MicroRNAs (miRNAs) are small, non-coding RNAs with the potential to regulate gene expression at the post-transcriptional level. Accumulating evidence indicates that miRNAs may serve as biomarkers for OP prediction and prevention. However, few studies have emphasized the role of miRNAs in systems-level pathogenesis during OP development. In this article, literature-reported OP miRNAs were manually collected and analyzed based on a systems biology paradigm. Functional enrichment studies were performed to decode the underlying mechanisms of miRNAs in OP etiology and therapeutics in three-dimensional space, i.e., integrated miRNA–gene–pathway analysis. In particular, interactions between miRNAs and three well-known OP pathways, i.e., estrogen–endocrine, WNT/β-catenin signaling, and RANKL/RANK/OPG, were systematically investigated, and the effects of non-genetic factors on personalized OP prevention and therapy were discussed. This article is a comprehensive review of OP miRNAs, and bridges the gap between an understanding of OP pathogenesis and clinical translation.
According to the World Health Organization definition, OP is a systemic bone disease with the characteristics of decreased bone mass, bone destruction, decreased bone mineral density (BMD), increased bone fragility, and a high risk of fracture that may be caused by mild trauma or even in the absence of trauma (
miRNAs widely exist in eukaryotes. They are a class of small non-coding RNAs with about 21 to 24 nucleotides, and hold the power to regulate gene expression at the post-transcriptional level. Accumulating evidence has convinced that miRNAs are functional players in multiple cellular abilities and biological processes, such as the proliferation, cycle, and apoptosis of cells; the development of complex diseases, including cancers, cardiovascular diseases, and neurodevelopmental diseases, is often closely related to the deregulation of miRNAs. In recent years, extensive efforts have been made to show that miRNAs play important roles in OP occurrence and development, especially in the osteoporotic fracture. For instance, miRNAs are involved in bone metabolism in OP through regulating target genes associated with human mesenchymal stem cells (hMSCs), bone cells osteoblasts and osteoclasts, which are important components during bone remodeling (
Biomarkers or markers are measurable and evaluable substances that can indicate change within biological systems. They are traceable and can predict disease occurrence and progression with high sensitivity and specificity, therefore exhibiting great potential for the diagnosis, prognosis, and therapy of various human diseases, including OP. With advances in biomedical research, plenty of studies have reported miRNAs as biomarkers for OP management, and reviews on OP miRNA biomarkers have been conducted to further the understanding of OP pathogenesis. For example,
This study provides a comprehensive review of current frontiers in the use of miRNA alterations for precision OP medicine and management. First, the definition of OP and its clinical symptoms are briefly introduced and its underlying pathogenic mechanisms presented. In particular, three crucial pathways, i.e., the estrogen–endocrine pathway, the WNT/β-catenin signaling pathway, and the receptor activator of nuclear factor-κB ligand (RANKL)/receptor activator of nuclear factor-κB (RANK)/osteoprotegerin (OPG) pathway, are emphasized as clues for systems-level “miRNA–gene–pathway–OP” deciphering. Second, reported miRNA alterations are manually collected and summarized by integrating OP sample source information (e.g., tissue, blood, and cell line) and subtype classifications [e.g., post-menopausal OP (PMO), senile OP (SOP)]. The targets of the reported OP miRNAs are computationally identified and investigation into OP pathogenesis is performed at gene, pathway, and cross level (i.e., integrating genetic and non-genetic factors for OP understanding). Here the relationship between the reported miRNA dysregulation and the three OP-associated pathways is explored, based on a combination of functional enrichment analysis and literature survey. Finally, challenges and perspectives for the integration and fusion of OP data and knowledge for future clinical translation and personalized OP healthcare are discussed.
Osteoporosis is a systemic bone disease characterized by decreased BMD or bone mineral content, resulting in increased bone fragility and an increased tendency to fracture. The quality and density of bone are constantly improved by the process of bone remodeling. Bone remodeling is bone formation by osteoblasts and the resorption of osteoclasts. It is a process of repeated renewal that eventually leads to a stable skeletal state. According to a recent epidemiological survey, about 200 million people worldwide have OP, and about 9 million of them have osteoporotic fractures. The risk of fracture in OP patients is up to 40%, and spine, hip, and wrist are especially prone to osteoporotic fractures. The mortality of osteoporotic patients is high due to fractures of the spine and hip. OP is a chronic disease with many causative factors. According to the clinical classification, OP can be divided into two types: primary and secondary. Primary OP can be classified into consists of three subtypes – that type I PMO, type II SOP, and type III idiopathic OP, with fracture the most common complication. Secondary OP is OP arising as a complication of a primary condition. Prior to fracture, there is usually no clinical manifestation. The disease is more common in women than in men, and is most common in post-menopausal women and the elderly. Interactions between genetic and environmental factors play an important role in its occurrence and development (see
Interactions between genetic and non-genetic factors in OP development. The reported factors may change the balance of bone remodeling. Bone density and structural integrity are dependent on bone-remodeling mechanisms associated with the function of osteocytes, osteoblasts, and osteoclasts. BMC, bone mineral content; BMD, bone mineral density; CRP, C-reactive protein; IL, interleukin; IGF, insulin-like growth factor; OPG, osteoprotegerin; PTH, parathyroid hormone; RAAS, renin angiotensin aldosterone system; RANK, receptor activator for nuclear factor-κB; RANKL, receptor activator for nuclear factor-κB ligand; TGF, transforming growth factor; TNF, tumor necrosis factor; (+): increase or promote; (–): reduce or inhibit.
Recent studies on OP have focused on the estrogen, RANKL/RANK/OPG (OPG: osteoprotegerin), and WNT signaling pathways. These three signaling pathways have their own signal transduction targets and are closely interrelated, forming a complex system to regulate bone metabolism in OP. Earlier observations showed that estrogen has a therapeutic effect on PMO, with the bone mineral content of 63 ovariectomized patients in one study increasing significantly following estrogen treatment (
The search terms used for the collection of information on previously reported OP miRNAs were “osteoporosis[tiab] AND (miRNA∗[tiab] OR microRNA∗[tiab]) AND (biomarker∗[tiab] OR marker∗[tiab] OR indicator∗[tiab] OR predict∗[tiab] OR therapeutic target∗[tiab]).” As shown in
Literature-reported osteoporosis miRNAs.
Report ID | Official symbol | Osteoporosis type | Expression | Sample | Experimental method | AUC | PMID |
miR-181c-5p | miR-181c-5p | PMO | Down | Blood | RT-PCR | NA | 31872255 |
miR-497-3p | miR-497-3p | PMO | Down | Blood | RT-PCR | NA | 31872255 |
miR-133a-3p | miR-133a-3p | PMO | Up | Blood | NA | NA | 31023966 |
let-7c | let-7c | PMO | Up | Blood | NA | NA | 30379578 |
miR-23b-3p | miR-23b-3p | PMO | NA | Blood | qRT-PCR | NA | 30171938 |
miR-140-3p | miR-140-3p | PMO | NA | Blood | qRT-PCR | NA | 30171938 |
miR-485-5p | miR-485-5p | NA | Up | Blood | qRT-PCR | NA | 30070309 |
miR-122-5p | miR-122-5p | NA | Down | Blood | RT-qPCR | 0.666 | 29849050 |
hsa-miR-4516 | miR-4516 | NA | Down | Blood | RT-qPCR | 0.727 | 29849050 |
miR-148a-3p | miR-148a-3p | NA | Up | Blood | qPCR | NA | 27900532 |
miR-30b-5p | miR-30b-5p | PMO | Down | Blood | qPCR | 0.793 | 27821865 |
miR-103-3p | miR-103-3p | NA | Down | Blood | qPCR | 0.8 | 27821865 |
miR-142-3p | miR-142-3p | NA | Down | Blood | qPCR | 0.789 | 27821865 |
miR-328-3p | miR-328-3p | NA | Down | Blood | qPCR | 0.874 | 27821865 |
miR-122-5p | miR-122-5p | NA | Up | Blood | RT-PCR | NA | 26163235 |
miR-125b-5p | miR-125b-5p | NA | Up | Blood | RT-PCR | NA | 26163235 |
miR-21-5p | miR-21-5p | NA | Up | Blood | RT-PCR | NA | 26163235 |
miR-194-5p | miR-194-5p | PMO | Up | Blood | qRT-PCR | NA | 26038726 |
miR-21 | miR-21-5p | PMO | Down | Blood | NA | NA | 25231354 |
miR-133a | miR-133a | PMO | Up | Blood | NA | NA | 25231354 |
miR-21 | miR-21-5p | NA | Up | Blood | qPCR | 0.63 | 24431276 |
miR-23a | miR-23a-3p | NA | Up | Blood | qPCR | 0.63 | 24431276 |
miR-24 | miR-24-3p | NA | Up | Blood | qPCR | 0.63 | 24431276 |
miR-93 | miR-93-5p | NA | Up | Blood | qPCR | 0.68 | 24431276 |
miR-100 | miR-100-5p | NA | Up | Blood | qPCR | 0.69 | 24431276 |
miR-122a | miR-122-5p | NA | Up | Blood | qPCR | 0.77 | 24431276 |
miR-124a | miR-124-3p | NA | Up | Blood | qPCR | 0.69 | 24431276 |
miR-125b | miR-125b-5p | NA | Up | Blood | qPCR | 0.76 | 24431276 |
miR- 148a | miR-148a-3p | NA | Up | Blood | qPCR | 0.61 | 24431276 |
miR-503 | miR-503-5p | PMO | Down | Blood | qRT-PCR | NA | 23821519 |
miR-137 | miR-137 | NA | Up | Bone | RT-PCR | NA | 29786747 |
miR-331 | miR-331-3p | NA | Down | Bone | NA | NA | 26329309 |
miR-21 | miR-21-5p | NA | Up | Bone | qPCR | 0.63 | 24431276 |
miR-23a | miR-23a-3p | NA | Up | Bone | qPCR | 0.63 | 24431276 |
miR-24 | miR-24-3p | NA | Up | Bone | qPCR | 0.63 | 24431276 |
miR-25 | miR-25-3p | NA | Up | Bone | qPCR | NA | 24431276 |
miR-100 | miR-100-5p | NA | Up | Blood | RT-PCR | 0.89 | 31532098 |
miR-100 | miR-100-5p | NA | Up | Bone | qPCR | 0.69 | 24431276 |
miR-125b | miR-125b-5p | NA | Up | Bone | qPCR | 0.76 | 24431276 |
miR-422a | miR-422a | PMO | Up | Monocyte | qRT-PCR | NA | 24820117 |
miR-133a | miR-133a | PMO | Up | Monocyte | qRT-PCR | NA | 22506038 |
Classification of the collected OP miRNAs based on sample source. Red: up-regulated; green: down-regulated; blue: up-regulated and down-regulated in different studies; black: expression level not available.
The filtering criteria for miRNA collection is: (1) Sample sources must originate from human; (2) Circulating miRNA expression profiles of OP patients should be used; (3) Samples should be separated into case and control groups; (4) The expression of miRNA should mainly be validated through low-throughput (e.g., qRT-PCR) methods.
Through literature searching, we identified several miRNAs for OP diagnosis. Among them, ROC curves were performed for some prominently regulated miRNAs. The area under the curve (AUC) is used to determine the diagnostic value of each miRNA. The AUC values were higher than 0.6, which indicated their diagnostic value for OP. For example,
There are some miRNAs that are significant for the potential treatment of OP. For example, miR-503 (also known as: miR-503-5p) in circulatory mononuclear cells played an important role in the regulation of bone resorption in PMO. The overexpression of miR-503 inhibits osteoclast formation induced by RANKL, making it a potential target for OP treatment (
We found that miR-137 could serve as a prognostic biomarker for OP. This miRNAs were derived from cells in bone tissue samples of OP patients. Liu and Xu found that the expression of miR-137 in osteoporotic fracture patients was significantly higher than that without fracture (
Some miRNAs have also been proved to be potential biomarkers of OP. For example,
To decode the pathogenic role of the collected miRNAs in OP, we performed functional analysis from an integrated “miRNA-gene-pathway” angle using computational tools and methods. Among them, Gene ontology (GO) involves gene and gene product vocabulary which can be divided into three categories: biological process (BP) and cell components (CC), molecular function (MF). The Kyoto Encyclopedia of Genes and Genomes (KEGG) is a manually created database, where the biochemical processes of cells, such as membrane transport, cell cycle, metabolism and signal transmission, are illustrated. Ingenuity Pathway Analysis (
To investigate the regulatory role of collected miRNAs in OP pathogenesis, the target genes of the OP miRNAs were identified by integrating the miRNA–gene relationships from different databases (
Gene ontology analysis was performed on the identified targets of the collected OP miRNAs using the online tool DAVID (
The 10 gene ontology terms most significantly enriched by targets of the reported miRNAs.
Category | GO terms | Number of enriched genes | Adj. |
BP | Positive regulation of transcription from RNA polymerase II promoter | 281 | 1.37E-23 |
Positive regulation of transcription, DNA-templated | 246 | 2.14E-22 | |
Negative regulation of transcription from RNA polymerase II promoter | 247 | 4.10E-22 | |
Protein phosphorylation | 223 | 1.72E-21 | |
Transcription from RNA polymerase II promoter | 232 | 5.30E-21 | |
Negative regulation of transcription, DNA-templated | 234 | 8.09E-21 | |
Positive regulation of apoptotic process | 185 | 1.38E-20 | |
Ephrin receptor signaling pathway | 223 | 1.69E-20 | |
Viral process | 214 | 6.61E-20 | |
Cell–cell adhesion | 260 | 1.18E-19 | |
CC | Nucleoplasm | 406 | 4.14E-29 |
Cytosol | 463 | 8.54E-26 | |
Cytoplasm | 477 | 2.18E-25 | |
Nucleus | 468 | 7.60E-25 | |
Membrane | 265 | 2.38E-22 | |
Extracellular exosome | 346 | 5.28E-18 | |
Cell–cell adherens junction | 219 | 5.42E-14 | |
Focal adhesion | 562 | 1.03E-11 | |
Intracellular membrane-bounded organelle | 562 | 1.03E-11 | |
Protein complex | 163 | 1.87E-11 | |
MF | Protein binding | 180 | 3.16E-18 |
Transcription factor binding | 400 | 5.37E-16 | |
Chromatin binding | 188 | 1.67E-12 | |
Protein kinase binding | 141 | 8.16E-11 | |
Protein serine/threonine kinase activity | 163 | 1.06E-10 | |
Protein kinase activity | 112 | 1.39E-10 | |
Transcription factor activity, sequence-specific DNA binding | 133 | 1.12E-09 | |
Cadherin binding involved in cell–cell adhesion | 70 | 6.65E-08 | |
Ubiquitin protein ligase binding | 62 | 6.74E-08 | |
Enzyme binding | 115 | 2.44E-07 |
The identified miRNA targets were further mapped and compared with those of collected OP-associated genes. A final total of 96 genes were found to be overlapped and regulated by 19 OP miRNAs (see
KEGG (
The 10 most significantly enriched KEGG and IPA signaling pathways were selected for further analysis. In KEGG, as shown in
The 10 pathways most significantly enriched by targets of the reported miRNAs.
Category | Pathway terms | Number of enriched genes | Adj. |
KEGG | Pathways in cancer | 140 | 1.29E-17 |
Proteoglycans in cancer | 50 | 3.32E-11 | |
Prostate cancer | 51 | 1.40E-10 | |
Renal cell carcinoma | 42 | 3.74E-09 | |
Small cell lung cancer | 42 | 1.33E-08 | |
FoxO signaling pathway | 42 | 7.83E-08 | |
Pancreatic cancer | 33 | 6.32E-07 | |
Hepatitis B | 55 | 5.36E-06 | |
Chronic myeloid leukemia | 27 | 7.45E-06 | |
Non-small cell lung cancer | 36 | 1.67E-05 | |
IPA | Molecular mechanisms of cancer | 151 | 7.26E-28 |
Senescence pathway | 61 | 1.15E-19 | |
HGF signaling | 63 | 6.08E-19 | |
p53 signaling | 61 | 1.44E-17 | |
Hepatic fibrosis signaling pathway | 71 | 1.45E-17 | |
Role of tissue factor in cancer | 118 | 3.04E-17 | |
NGF signaling | 83 | 6.55E-17 | |
ERK/MAPK signaling | 62 | 7.21E-17 | |
Glucocorticoid receptor signaling | 60 | 3.21E-16 | |
Pancreatic adenocarcinoma signaling | 74 | 1.45E-15 |
Considering the complexity of OP, an integrated “miRNA–gene–pathway” analysis was conducted for the systems-level decoding of OP pathogenesis. As a biomarker for PMO diagnosis, miR-133a-3p regulates important target genes related to OP, such as VEGFA, SP1, COL1A1, EGFR, etc. These genes are concentrated in key sites of pathways closely related to OP, such as “Osteoarthritis Pathway,” “MAPK signaling pathway,” “TGF-beta signaling pathway,” “Glucocorticoid Receptor Signaling,” “ERK/MAPK Signaling,” etc. In particular, target gene SP1 plays an important role in signal transduction in these pathways. MiR-142-3p, a potential biomarker of PMO, and its target genes (CCND1, STAT1, ACVR1, TNFRSF11B, IRS2, CDC25B) also enrich and participate in signal transduction in these important pathways. In addition, CCND1, ACVR1 are enriched in Wnt/beta-catenin Signaling and IRS2 in RANK Signaling in osteoclasts. These signaling pathways have been previously reported to be closely related to bone metabolism in OP, while other target genes of miR-194-5p (target OP-associated genes: ACVR2B, PRKAR1A, RB1, SERPINE2, etc.), miR-21-5p (target OP-associated genes: TGFB1, JAG1, THBS1, SP1TGFB1, TNSF11B, EGFR, etc.), miR-23b-3p (target OP-associated genes: CCND1, FGF2, PDGFA, LRP5, etc.), and miR-30b-5p (target OP-associated genes: CCNE2, ACVR1, SMAD1, etc.), which are potential biomarkers of PMO and also mostly concentrated in the key sites of these pathways. Here miR-21-5p was found to be down-regulated in the serum of patients with OP and so could be used as a diagnostic biomarker for PMO. Interestingly, its target gene, TGFB1, is involved in extracellular signal transduction through the MAPK signaling pathways; thus miR-21-5p may affect the metabolism of bone cells by regulating TGFB1 in this pathway, influencing the occurrence and transformation of OP. In addition, the target genes (CCND1, CCNE2, CDKN1A, VEGFA, FGF2, CHEK1, etc.) of miR-503-5p, which are potential therapeutic target for the treatment of PMO, are also enriched in some important pathways, such as “Cell cycle,” “Pathways in cancer,” “p53 signaling pathway,” “AMPK signaling,” “PI3K/AKT signaling,” “Wnt/beta-catenin Signaling,” et al. Another miRNA,
Since the development of OP is closely related to the dysfunction of three well-studied pathways, i.e., the estrogen–endocrine, WNT/β-catenin signaling, and RANKL/RANK/OPG pathways, exploring the underlying relationships between the reported miRNAs and these three pathways based on a systems biology viewpoint would enhance our pathogenic understanding of OP. Estrogen deficiency is the main cause of PMO, and estrogen replacement therapy is clinically effective for PMO. The relationship between estrogen signaling and OP miRNAs is shown in
The regulatory role of miRNAs in OP-associated pathways.
According to previous reports, WNT signaling and the RANKL/RANK/OPG pathway may influence bone metabolism and bone mass. As described in
The development of OP is affected by a combination of genetic and non-genetic factors, thus cross-level analysis of OP pathogenesis, i.e., the integration of molecular events (e.g., miRNAs, genes, and pathways) with signatures at individual (e.g., clinical phenotype) and population level (e.g., lifestyle, living environment) is essential. The abnormal expression of miRNAs in OP patients regulates the function of their target genes and participates in some key biological signaling pathways associated with bone metabolism in OP. For example, miR-124, is also down-regulated in the serum of OP patients, and it may weaken the inhibition of NFATC1 and promote bone resorption of osteoclasts. In addition, the over-expression of let-7c in Wnt signaling pathway reduces osteoblasts proliferation by inhibiting target genes SCD1 under oxidative stress (
In addition to molecular factors, surgical intervention, lifestyle, and living environment are important in OP initiation and progression. For example, PMO often develops following physiological menopause; however, ovariectomy may also lead to estrogen deficiency, another reason for the high incidence of OP in post-menopausal women. Previous studies have shown that appropriate physical exercise can increase BMD, and a healthy diet and adequate nutritional intake are also positive lifestyle changes that influence OP prevention and treatment.
Osteoporosis prevention and treatment remains a challenge and the development of integrative methods for precision medicine in the management of OP is urgently required. On one hand, miRNAs hold the power to enable the decoding of the hidden pathogenesis of OP and are potential drug targets for OP therapy. On the other hand, non-drug therapies, i.e., lifestyle and environmental interventions, are also effective in OP management and would be advantageous in personalized OP treatment and healthcare.
Osteoporosis is a systemic bone disease with increased incidence with age. Due to their pathological role in the development of OP, miRNAs have gradually been validated as biomarkers for OP monitoring, such as early diagnosis, prognosis tracking, and personalized therapy.
In this article, a total of 28 previously reported OP miRNAs were collated and analyzed, based on systems biology approaches. Most of these miRNAs were of value in OP diagnosis, therapeutics and prognosis. To decode the regulatory role of the collected miRNAs, the targets were then identified from public databases. OP-associated genes were mined and integrated for a systems-level analysis of OP pathogenesis. Based on functional enrichment analysis, significantly enriched KEGG and IPA terms were discovered, such as “ERK/MAPK signaling pathway,” “Glucocorticoid Receptor Signaling,” “TGF-β signaling pathway,” “p53 signaling,” “Estrogen Receptor Signaling,” “RANK Signaling in Osteoclasts,” and “Wnt/β-catenin Signaling.” All of these pathways have been reported to be associated with bone metabolism in OP. Moreover, most of the target genes of biomarker miRNAs were located at the key sites of these pathways, which convinced us of the functional importance of miRNAs in OP development. For example, miR-21 may affect the metabolism of osteoporotic osteocytes by regulating TGF-β1 in the FoxO and MAPK signaling pathways, and miR-503 regulates RANK, participating in osteoclast proliferation, differentiation, and activity. According to previous reports, the development of OP is closely related to dysfunction in three well-known pathways, i.e., the estrogen–endocrine pathway, the WNT/β-catenin signaling pathway, and the RANKL/RANK/OPG pathway. To better understand the heterogenicity of OP, the relationships among miRNAs, pathways, and different OP subtypes were closely analyzed. For example, miR-221 is a potential treatment of PMO. Due to estrogen deficiency, miR-221 inhibits RUNX2 to regulate MSC differentiation into osteoblasts. MiR-27a and miR-29a participate in WNT/β-catenin signaling by regulating the associated target genes. But these miRNAs as biomarkers of OP need to be further validated. miR-21 and miR-124a regulate gene PDCD4 and NFATC1, respectively, and have opposite effects in the RANKL/RANK signaling pathway. In addition to genetic factors, some non-genetic elements, e.g., lifestyle and living environment, were found to be essential in OP epidemiology; thus cross-level analysis is of great significance for OP precision medicine and healthcare.
This study provides a comprehensive review of reported OP miRNAs and integrates “miRNA–gene–pathway” knowledge to explain the pathogenic role of miRNAs in OP genesis; however, some limitations exist. First, the alteration of the collected miRNAs is only validated in the experimental stage, e.g., using cell lines, model animals, and human samples; translational application of biomarkers in OP clinical management needs to be promoted. Second, the effects of non-genetic factors on miRNA–OP interactions are insufficiently investigated. With such complexity and heterogenicity in OP development, the importance of lifestyle and environmental factors on OP epigenetics should be reasonably considered. Third, the initial intention was to consider the interaction mechanism between different types of OP and miRNA alterations, but due to the limitations of existing research data, it has not been better expanded. Finally, but importantly, only 28 miRNAs were recorded in this study. Considering the papers for miRNA selection, limitations still need to be concerned. For example, many of the studies identified the change of miRNA expression based on small sample validation, further clinical tests using multi-center sample sources should be performed to evaluate the sensitivity and specificity of the miRNA candidates. On the other hand, it is difficult to verify whether a miRNA could serve as a true biomarker due to the design of research scheme, the selection of reference standard, and the reliability of result evidence. In this review, we mainly selected miRNAs validated by low-throughput experiments such as PCR to ensure the biomarker potential. For clinical translation, integrated computational prediction and experimental validation should be encouraged for decoding miRNAs in OP pathogenesis at the systems biology level. With advances in experimental and computational biology, more miRNA biomarkers will be identified and verified for the personalized prevention and treatment of OP, and methodology used in OP data analysis can reasonably be expected to improve, creating a systematic framework for use in intelligent medicine and healthcare.
The data supporting this review are from previously reported studies and datasets, which have been cited. The processed data are available at tables/figures in the
HH, XH, and YL collected the data. HH, XH, YZ, RW, and YL reviewed the data and performed the bioinformatics analyses. All authors drafted and revised the manuscript. HH and XH contributed equally to this study. BS and YL conceived and supervised this study jointly.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The authors gratefully thank Prof. Luonan Chen at Key Laboratory of Systems Biology, Shanghai Institute of Biological Sciences, Chinese Academy of Sciences for providing the IPA analysis.
The Supplementary Material for this article can be found online at: