Osteoporosis is increasing in prevalence around the world due to the aging population. Although osteoporosis can be attributed to an imbalance between osteoblastic bone formation and osteoclastic bone resorption [1
] as part of the natural aging process, the finer details of pathogenesis remain to be elucidated. Current treatment methods, including bisphosphonates, hormone replacement therapy, and immunotherapy, all carry the risk of side effects related to their mechanisms of action. Therefore, novel therapies are being developed. Further understanding of the underlying pathophysiology is paramount to developing these new therapies.
Micro RNA (miRNA) and long-non-coding RNA (lncRNA) are two such targets that have recently come into the spotlight due to their ability to control gene expression at the post-transcriptional level, providing epigenetic modification [2
]. miRNAs are a class of non-coding RNAs that are approximately 18–25 nucleotides long [4
]. It is thought that up to 60% of human protein-coding genes may be regulated by miRNAs [2
]. They bind to the 3-untranslated regions (3-UTR) of target genes, leading to messenger RNA (mRNA) degradation and transcription inhibition [2
]. The process of miRNA regulation is complex, as each miRNA binds to a number of targets, and several miRNAs target the same mRNA [2
]. They have been found to regulate most biological processes, including cell development, differentiation, proliferation, metabolism, and cell cycle regulation [2
]. They have also been found to regulate gene expression that controls osteoblast-dependent bone formation and osteoclast-related bone remodeling [3
lncRNAs are highly structured RNA transcripts longer than 200 nucleotides that do not translate into proteins [8
]. In fact, lncRNAs have very complex secondary and tertiary structures and the same degradation processes as mRNAs. The fact that they have a rapid turnover is due to their sponge function in binding the miRNAs that lead to a degradation of the lncRNA itself [6
]. They can act as signaling, decoy, and framework molecules, or as primers [9
]. Current evidence suggests that lncRNAs can act as chromatin and transcriptional as well as post-transcriptional regulators [8
]. With regards to osteoporosis, lncRNA is thought to be involved in the proliferation, apoptosis, and inflammatory response of the bone [10
The interaction between miRNAs and lncRNAs is also of current interest. Studies have attempted to link lncRNAs, miRNAs, and mRNAs together in a complex network, such as Hao et al.’s systematic analysis using the mandibles from ovariectomized mice [11
]. Fei and colleagues performed a small study in five Chinese women to identify the key lncRNAs in postmenopausal osteoporosis (PMOP) through RNA sequencing [12
]. After identifying various differentially expressed mRNAs (DEmRNAs) and differentially expressed lncRNAs (DElncRNAs), they constructed a DElncRNA-DEmRNA co-expression network [12
]. In a larger study, Zhou et al. identified lncRNAs in 73 Caucasian women with PMOP and established an mRNA/lncRNA co-expression network [13
]. The shared goal of the above studies was to provide a foundation for future investigations of lncRNAs in PMOP and help to develop biomarkers and drugs. In this review, we aim to summarize the current evidence on the actions of various miRNA and lncRNA. IsomiRNAs are grouped together.
The field of miRNAs and lncRNAs with regards to osteoporosis is still relatively new, and this is reflected in the quantity of research currently available. Often, there are conflicting results in the literature, such as miR-223 affording to both promote and inhibit osteoclastogenesis [3
]. Wijnen et al. suggest that miRNAs provide both positive and negative cross-talk between different regulatory pathways [3
], thereby leading to this phenomenon. Another possible explanation is that miRNA is present in different clinical specimens. For example, Mandourah et al. found that, while both miR-122-5p and miR-4516 were suitable biomarkers for osteoporosis, miR-122-5p was detectable in the serum, while miR-4516 was found in the plasma [33
]. A large cohort study of 682 women found that there was a lack of association between bone parameters and circulating levels of miRNAs, stating results were canceled out after age adjustment [69
]. The authors suggest this could be due to the fact that age was also strongly correlated with the serum levels of the 32 miRNAs they selected [69
]. On the other hand, another pilot study suggests that the combination of many miRNAs can help predict fragility fracture risk [70
]. Perhaps it makes sense that one cannot look at each miRNA in isolation, as the cellular processes of OP work in a synergistic fashion. Finally, many studies on miRNA deregulation lack control groups [7
]. These factors are worth bearing in mind when conducting future clinical trials.
The current review of literature was retrieved from studies of different sources (human and animals) and tissues (bone cells, serum, and tissue fluid), which may explain the many conflicting results. Thus, conflicting conclusions have been reported for most miRNAs and lncRNAs depending on cellular models used, animal studies or cohorts of humans, and even analytical methods. We have specified these sources and differences where possible. One of the advantages of animal studies is the convenience in conducting these studies. Some animals (such as mice) have a shorter growth cycle than that of a human being and are subject to investigations in a limited research period. Furthermore, gene knockout can be implemented in an animal model to investigate the depletion effect of genes, DNAs, as well as RNAs on the target organs, which cannot be performed in a human being. However, the major disadvantage of animal studies lies in the differences of genes and subsequent RNAs between animals and human beings, and thus generalization of the conclusions of animal studies to human beings is limited. In contrast, the results of human studies are observational but direct evidence. The major disadvantage of human studies is that intervention (such as gene knockout) cannot be implemented due to the consideration of moral hazards. In addition, the expression of genes, miRNAs, or lncRNAs in variant tissues (bone cells, serum, and tissue fluid) may be different, thus resulting in conflicting conclusions.
Due to the lack of clear-cut associations between PMOP and the expression of miRNAs and lncRNAs at present, there is still a long way to go before they can be used as potential noninvasive biomarkers. In addition, there are also many barriers still to overcome to transfer miRNA and lncRNA knowledge to the synthesis of clinical therapeutic drugs. Synthetic oligonucleotides mimicking miRNAs have a limited half-life due to degradation by nucleases in the bloodstream, and they also have a poor capacity to penetrate host cell membranes to reach their target cells [71
]. lncRNA degrades even more easily than miRNA due to their low structural stability [9
]. At present, antagomiRs, viruses, scaffold-based miRNA delivery, and extracellular vesicles have all been used as vectors in miRNA studies [71
]. AntagomiRs are direct mRNA inhibitors, but, unfortunately, are required at a high dose to work [71
]. Adeno-associated viruses are small vectors non-pathogenic to humans but are expensive [71
]. Scaffold-based miRNA delivery provides not only structural support but also a convenient environment for bone tissue growth [71
]. Finally, extracellular vesicles are natural bioabsorbable gene carriers that can recognize target cells, with the great advantages of oral administration and ease of long-term storage [71
]. There are also currently no reports of lncRNAs in current osteoporosis treatment up to date.