Bone is a frequent site of metastasis. MicroRNAs (miRNAs) and small nucleolar RNAs (snoRNAs) are two classes of small non-coding RNAs (sncRNAs) that posttranscriptionally regulate gene expression in cells. While miRNAs have been largely investigated in the context of bone metastasis, snoRNAs have been poorly studied. However, there is evidence that snoRNAs can give rise to a specific class of miRNAs (called sno-miRNAs), thus sharing features with miRNAs. Another common ground between miRNAs and snoRNAs is the fact that both can circulate in biological fluids, such as blood and lymph. Compared to other RNAs, their small size as well as their interaction with core proteins protect them from a massive degradation both as free and embedded forms, making sncRNAs stable, secreted, circulating molecules. As an embedded form, they are usually within extracellular vesicles (EVs) that derive from cells. EV-embedded and/or circulating microRNAs, mainly, and snoRNAs have been pointed out as important players in bone metastasis by preparing the pre-metastatic niche, directly and indirectly affecting the activities of osteoclasts and osteoblasts, and acting as mediators within cells to support cancer cell growth in bone.
1. MiRNA and SnoRNA Roles in the Formation of a Pre-Metastatic Niche
The bone marrow stroma, which is enriched with cytokines and growth factors, is an advantageous environment for the homing and outgrowth of metastatic cells
[1]. Besides the intrinsic characteristics of bone, as an attractive and fertile metastatic site, there is evidence that bone niches can be further created or promoted by the remote action of primary tumour cells
[2]. Pre-clinical and clinical research have provided evidence that EVs, as systemic factors, could create ideal conditions at distant sites, allowing disseminated cancer cells to colonise bone marrow. Thus, primary cancer cells can remotely ‘educate’ distant sites in bone, by secreting EVs, to further receive and host disseminated cancer cells even before the metastatic process starts at the primary site. Indeed, it has been demonstrated that EVs can be internalised by resident cells, such as osteoclasts and osteoblasts, and modulate their maturation and activity
[3]. For example, the internalisation of EVs produced by multiple myeloma cells can promote the differentiation of osteoclasts
[4], while prostate cancer cell-derived EVs inhibit it
[5]. Moreover, the pre-treatment of animals with tumour cell-derived exosomes increases metastatic burden in murine models of prostate cancer, making the bone marrow the preferential target for these tumour cells
[6].
Some studies aiming to characterise the EV content also explored the role of ncRNAs as molecular players of changes induced by EVs. Up to now, EV-derived miRNAs are the best studied ncRNAs from the exosomal cargo, and several of them have been shown to play a direct role in cancer progression. First, the fundamental role of the EV-miRNA content has been proven by a study showing that, while wild-type EVs produced by prostate cancer cells contribute to creating a pre-metastatic niche in bone, EVs produced by Dicer-depleted tumour cells have less effects on bone cells
[7]. Then, a direct effect of specific EV-cargo miRNAs on the preparation of pre-metastatic niche was explored in various studies (
Figure 1). For example, breast cancer-secreted miR-105 can be delivered, embedded in EVs, to endothelial cells and can promote tumour metastasis
[8]. Although this work mainly focuses on lung and brain metastases, these findings might be extended also to bone. Slightly more controversial, there is evidence that the well-known oncogene suppressor miRNA miR-200, produced by primary breast cancer cells and then encapsulated in EVs, promotes metastasis
[9]. Another study on breast cancer showed that tumour-derived, exosome-delivered miR-20a-5p facilitates osteoclastogenesis by targeting SRCIN 1, previously known for its role in cancer progression
[10]. A similar study has been conducted in lung cancer, where the exosomal miR-214 can be released by both lung cancer cells and osteoclasts, mutually and positively contributing to osteoclast activation in bone, thereby favouring formation of osteolytic metastases
[11]. In prostate cancer, tumour-derived miR-141-3p embedded in EVs can be taken up by osteoblasts, promoting their activity, and indirectly compromising the function of osteoclasts to ultimately promote the formation of osteoblastic bone metastases
[12]. Moreover, EVs-embedded miRNAs can have synergic effects with secreted proteins. For instance, EV-embedded miR-19a together with the integrin-binding sialoprotein (IBSP) are secreted by breast cancer cells, and they synergistically influence the bone microenvironment
[13]. While IBSP creates an osteoclast-enriched niche, exosomal miR-19a induces osteoclastogenesis, two factors that contribute to creating a favourable site for breast cancer metastasis
[13]. Another study identified exosomal miR-940 as being highly expressed in prostate cancer cells, which usually induces an osteoblastic phenotype in the bone metastatic microenvironment
[14]. Interestingly, the artificial expression of miR-940 in breast cancer cells, which usually produce osteolytic bone metastasis lesions, induces extensive osteoblastic lesions in animals by promoting osteoblast maturation
[14]. This study is particularly interesting as it demonstrates how it is possible to reprogram the bone metastatic microenvironment through the secretion of a single miRNA, in this case miR-940, by modulating osteoblast activity. This clearly shows how tumour-derived miRNAs are powerful regulators of gene expression, and how important is to track down these modulations for a more comprehensive understanding of metastasis.
Figure 1. The role of microRNAs in the pre-metastatic niche in bone. MiRNAs have been largely investigated as EV cargo from cancer cells able to remotely affect the activity of cells in distant organs, such as osteoblasts and osteoclasts in bone. Here, the researches reported some examples of miRNAs from breast cancer (miR-105, miR-200, miR-20a-5p, miR-19a, etc.), lung cancer (miR-214, etc.) and prostate cancer (miR-141-3p, miR-940, etc.) that, as EV cargo, can circulate in blood or lymphatic vessels and reach distant sites, such as bone. In bone, EVs can be taken up by osteoblasts (e.g., miR-141-3p, miR-940) or osteoclasts (e.g., miR-20a-5p, miR-214, miR-19a), and modulate their activity and/or maturation.
The role of EVs-embedded snoRNAs in the formation of a pre-metastatic niche in bone has never been investigated. However, the development of a RNA-seq approach, including the thermostable group II intron reverse transcriptase sequencing (TGIRT-seq), dedicated to small structured ncRNA and based on the use of group II intron-encoded RTs instead of low-fidelity retroviral RTs, demonstrated that snoRNAs can be detected and identified in EVs
[15][16]. Such an approach led to surprising clinical observations, suggesting that snoRNAs may serve as novel peripheral blood plasma-EV-derived biomarkers for monitoring astronauts’ health
[17]. Indeed, this study revealed that several snoRNAs, including SNORA74A, were significantly dysregulated in peripheral blood plasma EVs from astronauts 3 days after the shuttle missions, compared to 10 days before the missions. Whether snoRNA-associated EV might be involved in metastatic progression clearly needs to be addressed. Overall, the lack of knowledge on the role of snoRNA in premetastatic niche formation suggests that a more comprehensive study aiming to explore the secretome from primary cancers—and the potential role of circulating small RNAs, besides miRNAs, in the remote control of distant organs—is a very attractive opportunity to better understand bone metastasis mechanisms.
2. MiRNA and SnoRNA Roles in the Vicious Cycle in Bone
Since both osteoclasts and osteoblasts are the main regulators of bone homeostasis, they play an important role in allowing the seeding and sustaining outgrowth of metastatic cells
[18][19]. Thus, EV-derived cargo that can affect any of the bone resident cells may be responsible for not only creating the pre-metastatic niche, but also promoting a series of events that establish a complex crosstalk between bone resident and metastatic cells. This concept is known as the vicious cycle. While miRNAs have clearly been shown to act as cell mediators in this vicious cycle, the role of snoRNAs remains largely unknown (
Figure 2).
Figure 2. The role of microRNAs and snoRNAs in the vicious cycle in bone. MiRNA and snoRNA release by metastatic cancer cells in bone is directly involved in sustaining a positive feedback loop (also called ‘vicious cycle’) between cancer, osteoclast, and osteoblast cells that further worsen the unbalance of bone homeostasis due to the presence of metastasis. Once metastatic cells disseminate to bone through the blood circulation, they can seed and proliferate in this new micro-environment. Here, cancer-derived miRNAs can directly affect activities of both osteoclasts (miR-214, miR-19a, miR-26a-5p, miR-27a-3p, miR-30e-5p, miR-92a-1-5p, etc.) and osteoblasts (miR-92a-1-5p, SNORD166, etc.) as well as the relationship between osteoclasts and osteoblasts, leading to bone lesion formation. Additionally, miRNAs that derive from osteoclasts (miR-214, etc.) and osteoblasts can directly sustain the growth of metastatic cells in the bone marrow.
Regarding miRNAs, several examples can be cited. MiR-214, which has been found to be highly expressed in lung adenocarcinoma, is also shown to mediate intercellular communication between osteoclasts and osteoblasts
[11]. In this study, exosomal tumour-derived miR-214 is proposed as further intercellular mediator between the primary tumour and osteoclasts. Specifically, tumour-exosomal miR-214 stimulates osteoclast differentiation, consequently increasing bone resorption, and the availability of cytokines and growth factors in the bone environment. Moreover, miR-214 can be secreted from osteoclasts. Thus, targeting miR-214 might be a good strategy to interrupt the vicious cycle at the bone metastatic site
[11]. In breast cancer bone metastasis, exosomal miR-19a and the integrin-binding sialoprotein (IBSP), derived from oestrogen-receptor positive breast cancer cells, induce osteoclastogenesis and create a bone microenvironment enriched with mature osteoclasts
[13], which is known to attract metastatic cancer cells. In prostate cancer, EVs that derive from prostate cancer cells increase osteoblastic activity and metabolism and impair bone resorption
[20]. In this study, miR-26a-5p, miR-27a-3p, and miR-30e-5p have been identified as the abundant cargo of these EVs. Moreover, these miRNAs are involved in the suppression of the BMP-2-induced osteogenesis, suggesting a role in the suppression of bone resorption
[20], and pointing out the importance of miRNAs as EV-cargo effectors. Another study identified miR-92a-1-5p as an abundant miRNA in exosomes from prostate cancer cells that directly target collagenase 1-A1 (Col1A1), promoting osteoclast differentiation and inhibiting osteoblastogenesis
[21], which is quite a surprising finding knowing that prostate cancer metastases in bone have usually an osteoblastic phenotype. However, this illustrates how bone remodelling can be remotely modulated by miRNAs to allow the future hosting of cancer cells.
Although not precisely studied in the context of the vicious cycle in bone, the relationships between bone homeostasis and snoRNAs have been identified in different pathologies. First, in vitro treatment of primary osteoclasts with an anti-HIV drug (Tenofovir), which promotes loss of bone mineral density, significantly reduced SNORD32A expression, although its role in osteoclast dysfunction remains to be determined
[22]. Second, SNORD116 loss in a mouse model of the Prader–Willi familial syndrome is sufficient to reduce both bone mineral content and density, by reducing osteoblast differentiation without alteration in osteoclastogenesis
[23][24]. Moreover, SNORD116 along with other snoRNAs have been found dysregulated in serums and tissues of mice affected by osteoarthritis and joint ageing, further suggesting the potential use of snoRNAs as biomarkers
[25]. Alteration in snoRNA pattern has also been observed in senescent BMSCs, which are capable of self-renewal into different cell types including osteoblasts, reinforcing the association between snoRNAs and physiological bone formation. In cancer, it has been shown that snoRNAs contribute to the metastatic potential of p53-induced osteosarcoma
[26]. Indeed, the deletion of the transcription factor ETS2 in conditional osteoblast mutant p53 mice reduces the expression of a panel of 24 snoRNAs and reverses the metastatic phenotype of mutant p53 without affecting osteosarcoma development. Overall, these data support the notion that snoRNAs contribute to osteoclast/osteoblast balance in a physiological context and/or its imbalance in different diseases, including cancer. However, the role of snoRNAs in the vicious cycle in bone has not been investigated yet.
Of particular interest is the recent discovery that the tropism of tumour-derived exosomes can be site-specific due to the expression of specific proteins at the exosome surface
[27]. In bone, it has been identified that L-plastin, an actin-binding protein, as a component of exosomes from breast cancer cells, is able to activate osteoclasts
[28]. The same study demonstrated that peroxiredoxin-4 (PRDX-4) is also implicated in this process, and that higher levels of both L-plastin and PRDX-4 are associated with a higher risk to develop bone metastasis in breast cancer patients
[28]. MiRNAs and snoRNAs, being cargo, cannot directly drive the tropism of exosomes, which happens thanks to protein–protein recognition. However, they can regulate the expression of proteins that can be expressed at the exosome surface
[29], suggesting a regulatory role for miRNAs and snoRNAs in the tropism of EVs.