SnoRNAs and SNHGs in bone development, homeostasis, and disease

Summary

Small nucleolar RNAs (snoRNAs) guide ribosomal RNA modification and regulate broader RNA-based programs, whereas snoRNA host gene-derived long non-coding RNAs, including SNHG-family transcripts, participate in post-transcriptional and epigenetic regulation. This review synthesizes evidence from skeletal stem/progenitor cells, osteoblast- and osteoclast-lineage models, chondrocytes, bone tumor systems, patient samples, and extracellular vesicle or circulating RNA studies to define how snoRNAs and SNHGs shape bone development, remodeling, and disease. Across osteoporosis, osteoarthritis, bone tumors, and fracture healing, these RNAs influence osteogenic differentiation, osteoclastogenesis, cartilage homeostasis, tumor progression, repair responses, and intercellular communication through ribosome-associated control, ceRNA networks, transcript stability, and chromatin- or protein-associated mechanisms. We also discuss their potential as biomarkers and RNA-targeted therapeutic candidates, together with challenges in annotation, functional validation, delivery, and clinical translation. This framework highlights snoRNAs and SNHGs as regulatory components of skeletal biology and disease.

Graphical abstract

cell biology; molecular biology

Introduction

Small nucleolar RNAs (snoRNAs) are an abundant class of small non-coding RNAs that localize predominantly to the nucleolus and function primarily in the processing and chemical modification of ribosomal RNAs and other nuclear RNAs. By assembling into small nucleolar ribonucleoprotein complexes (snoRNPs), snoRNAs contribute to ribosome biogenesis, RNA maturation, and broader gene-regulatory programs. Based on conserved sequence motifs, associated core proteins, and subnuclear localization, snoRNA-related guide RNAs are commonly grouped into box C/D snoRNAs (SNORDs), box H/ACA snoRNAs (SNORAs), and small Cajal body-specific RNAs (scaRNAs).^1,2 Box C/D snoRNAs form characteristic K-turn structures via the C-box (RUGAUGA) and D-box (CUGA), recruiting fibrillarin methyltransferase to direct 2′-O-methylation of rRNAs and snRNAs. Box H/ACA snoRNAs adopt a hairpin-hinge-hairpin-tail conformation and guide site-specific pseudouridylation of target RNAs through dyskerin-containing H/ACA snoRNP complexes.^3,4 Distinct from canonical nucleolar snoRNAs, scaRNAs localize to Cajal bodies through CAB box-dependent mechanisms and mainly guide the modification of spliceosomal snRNAs.^5,6 Moreover, snoRNAs encompass a few exceptions such as mitochondrial RNA processing (MRP) RNA, hybrid snoRNAs, and snoRNAs potentially involved in non-canonical functions like pre-mRNA splicing and chromatin regulation. Nearly half of the snoRNAs in the human snoRNAome have unknown functions and are therefore termed “orphan” snoRNAs.⁷ This classification system is highly conserved across eukaryotes and archaea, revealing the core architecture of snoRNA-mediated RNA modifications as an ancient and fundamental mechanism for regulating life processes. These molecular properties provide the basis for understanding why snoRNAs are increasingly being implicated in skeletal development, bone homeostasis, and bone-related disorders.

In vertebrates, the majority of snoRNAs are embedded within introns of protein-coding and non-coding genes, collectively termed snoRNA host genes.⁴ Following transcription, host-gene-derived primary transcripts undergo splicing, through which intronic sequences are excised and further processed into mature snoRNAs. Meanwhile, some mature long non-coding host transcripts independently exert regulatory functions, modulating cellular processes through multiple mechanisms including competitive endogenous RNA (ceRNA) effects, epigenetic regulation, transcription factor interactions, and protein stability control.^8,9,10 Although many snoRNAs are generated from host-gene transcripts, including lncRNAs of the small nucleolar RNA host gene (SNHG) family and non-SNHG host genes such as GAS5, UHG, and U50HG,^11,12,13 the mature snoRNA and the processed host transcript should not be assumed to act as a single functional unit. In this review, we therefore discuss snoRNAs broadly and consider SNHG-family lncRNAs as related but potentially independent regulatory transcripts in skeletal biology.

Results

SnoRNAs and SNHGs regulate bone development and homeostasis

SnoRNAs and SNHGs regulate osteogenic differentiation and bone formation

Bone marrow mesenchymal stem cells (BMSCs) are key functional cells within the bone marrow microenvironment, playing an indispensable role in regulating bone metabolic homeostasis. Through their potent osteogenic differentiation potential, BMSCs determine the quantity and activity of pre-osteoblasts, thereby regulating the rate and quality of bone formation.¹⁴ Concurrently, BMSCs coordinate the dynamic equilibrium between bone resorption and formation via paracrine and cell-cell interactions,^15,16 maintaining bone homeostasis. When BMSC function is impaired or depleted, homeostasis of the bone marrow microenvironment is disrupted. This leads to heightened immune-inflammatory responses and bone metabolic imbalance, forming the core pathological basis for degenerative bone diseases such as osteoporosis.¹⁷ In osteoporosis and related skeletal disorders, impaired osteogenic differentiation and senescence of BMSCs are increasingly recognized as important pathogenic events. Emerging studies indicate that non-coding RNAs, including snoRNAs and SNHGs, participate in the regulation of BMSC fate, osteogenic differentiation, and age-related functional decline. Small RNA profiling has revealed significant changes in snoRNA expression during human BMSC (hBMSC) senescence, suggesting that snoRNA dysregulation may contribute to stem-cell aging and reduced osteogenic competence.¹⁸ Accordingly, clarifying how snoRNAs and SNHGs regulate BMSC biology may improve mechanistic understanding of bone metabolic disorders and help identify candidate targets for future precision interventions.

The application of BMSCs represents a promising therapeutic option for patients with immune system disorders and degenerative diseases. In addition to soluble mediators, BMSCs secrete extracellular vesicles (EVs) carrying functional small RNAs that may reflect key biological properties of the parental cells.¹⁹ Notably, the small RNA cargo of BMSC-derived EVs is enriched in snoRNAs and miRNAs, which together account for approximately 64–71% of the total small RNA content,²⁰ highlighting snoRNAs as a potentially important but underexplored component of BMSC-mediated signaling. Evidence from stem cell-derived exosome studies further suggests that these vesicular RNAs are biologically active in recipient cells. For example, cortical bone marrow-derived stem cell (CBSC)-derived exosomes reduced scar size and improved cardiac function after ischemic injury, at least in part through mechanisms involving altered snoRNA signaling in cardiac fibroblasts.²¹ Although this finding arises from a non-skeletal context, it supports the broader concept that exosomal snoRNAs can mediate functionally relevant intercellular communication.

Among snoRNAs with potential relevance to skeletal phenotypes, SNORD116 has also attracted attention because of its involvement in Prader-Willi syndrome (PWS), a neurogenetic imprinting disorder caused by loss of paternally expressed genes on chromosome 15q11–q13. PWS is frequently associated with skeletal abnormalities, including reduced bone mass and low bone mineral density, and Snord116-deficient mice recapitulate several of these phenotypes, such as low birth weight and decreased bone mass.²² These observations suggest that Snord116 contributes to skeletal development and bone homeostasis, although its direct mechanistic role in BMSC osteogenesis remains to be further clarified. Moreover, hypothalamic restoration of Snord116 expression in young germline knockout mice increased energy expenditure,²³ indicating that both central metabolic regulation and peripheral skeletal effects may be relevant to the pathogenesis and therapeutic targeting of PWS.

The skeletal functions of SNHGs should be considered within the broader context of snoRNA host-gene biology. “SnoRNA host gene” is a genomic category that extends beyond the SNHG-named lncRNA family, as classical non-SNHG host genes such as GAS5, UHG, and U50HG^11,12,13 also encode snoRNAs. Although many snoRNAs are embedded within host-gene transcripts, mature snoRNAs and processed host transcripts do not necessarily act as a single regulatory unit. Thus, this Review discusses snoRNAs broadly but focuses the host-gene-related analysis primarily on SNHG-family lncRNAs, for which most skeletal mechanistic evidence is currently available. Within this defined scope, emerging evidence indicates that SNHG-family lncRNAs are important modulators of bone metabolism through diverse mechanisms, including transcriptional regulation, epigenetic control, and ceRNA activity. Several SNHGs are upregulated during osteogenic differentiation and appear to directly promote osteogenesis in mesenchymal stem cells. For example, SNHG5 expression increases during the osteogenic differentiation of hBMSCs, and its silencing suppresses osteogenesis while inducing apoptosis. Mechanistically, SNHG5 functions as a ceRNA by interacting with miR-582-5p to regulate RUNX3 expression.²⁴ In addition, Yin Yang 1 (YY1) directly binds to the SNHG5 promoter to regulate its transcription, while SNHG5 also targets GDF5 via miR-212-3p and subsequently activates Smad1/5/8 signaling, further promoting osteogenesis.²⁵ SNHG5 has likewise been reported to enhance the osteogenic differentiation of periodontal ligament stem cells through the miR-23b-3p/Runx2 axis.²⁶ Similarly, SNHG14 is positively associated with BMSC osteogenic differentiation. Its downregulation reduces alkaline phosphatase activity and decreases osteopontin and osteocalcin expression. Mechanistically, SNHG14 promotes NEDD4L mRNA stability, enhances FOXA2 ubiquitination, and relieves FOXA2-mediated repression of PCP4 transcription, thereby facilitating osteogenesis through the NEDD4L/FOXA2/PCP4 axis.²⁷ SNHG16 also promotes osteogenic differentiation in hBMSCs by regulating the miR-485-5p/BMP7 axis.²⁸ In addition, SNHG3 has been reported to enhance osteogenesis through activation of BMP2 signaling,²⁹ although the specific cellular context should be further clarified.

Beyond these direct pro-osteogenic effects, several SNHGs appear to modulate osteogenesis under specific microenvironmental conditions. SNHG1 promotes cartilage regeneration and angiogenesis by enhancing STAT3 phosphorylation, reducing reactive oxygen species (ROS) accumulation, and regulating mitochondrial energy metabolism,³⁰ suggesting that certain SNHGs may support skeletal repair by improving cellular stress adaptation and the regenerative microenvironment. Under inflammatory conditions, TNF-α downregulates SNHG7 during the osteogenic differentiation of hDPSCs, whereas SNHG7 restores osteogenic capacity by suppressing miR-6512-3p, highlighting its potential relevance in inflammation-associated bone loss and periodontitis.³¹ In contrast, SNHG8 acts as a mechanosensitive lncRNA in PDLSCs. Its expression is reduced in response to mechanical stimulation, and SNHG8 knockdown enhances osteogenesis-related gene expression, indicating that it participates in the epigenetic regulation of mechanically induced osteogenic remodeling.³²

Collectively, SNHGs function as important regulators of bone formation by promoting osteogenic differentiation and modulating the cellular response to oxidative stress, inflammation, and mechanical cues through multiple transcriptional, post-transcriptional, and signaling mechanisms. These findings identify SNHGs as candidate molecular targets for the diagnosis and treatment of skeletal disorders such as osteoporosis and periodontitis.

SnoRNAs and SNHGs regulate osteoclastogenic differentiation and bone resorption

Bone marrow macrophages (BMMs) are a pivotal cell population within the bone marrow microenvironment, possessing dual functions in both innate immune responses and bone metabolism regulation. They play a central role in maintaining bone homeostasis and mediating pathological bone destruction.³³ BMMs serve not only as the direct precursors of osteoclasts but also differentiate into functionally active multinucleated osteoclasts under the synergistic drive of receptor activator of nuclear factor κB ligand (RANKL) and macrophage colony stimulating factor (M-CSF), thereby mediating bone resorption.³⁴ Concurrently, BMMs establish intricate paracrine regulatory networks with BMSCs and osteoblasts through the secretion of inflammatory cytokines (such as TNF-α and IL-6) and growth factors, thereby bidirectionally modulating the dynamic equilibrium between bone formation and resorption.³⁵ Given that BMMs originate from the myeloid differentiation pathway of hematopoietic stem cells (HSCs), the functional state and fate determination of HSCs directly govern the pool homeostasis and activation potential of BMMs and their osteoclast precursor cells.³⁶ Recent studies indicate that snoRNAs, as a significant class of non-coding RNA molecules, may contribute to bone homeostasis by indirectly regulating BMM generation and functional activity through effects on HSC self-renewal, myeloid differentiation, and cellular phenotype.³⁷ Accordingly, clarifying how snoRNAs regulate the HSC-BMM axis may improve mechanistic understanding of abnormal bone resorption and related bone metabolic disorders.

Research has revealed that the SNORD113–114 cluster participates in regulating the renewal of HSCs. Maternal knockout (Mat-KO) of the SNORD113–114 cluster disrupts translational mechanisms, inducing nucleosomal stress. This allows p53 to evade Mdm2-mediated proteasomal degradation, thereby preventing HSC apoptosis.³⁸ U3 snoRNA activates purified DNA-PK and triggers phosphorylation of DNA-PKc at residue T2609. Blocking phosphorylation of DNA-PKcs at the T2609 site, but not at the S2056 site, results in Ku protein complex-dependent defects in 18S rRNA processing, impairs global protein synthesis in hematopoietic cells, and induces bone marrow failure (BMF) in mice.³⁹ rRNA 2′-O-methylation (2′-O-Me) contributes to translational control and is guided by box C/D snoRNAs within fibrillarin-containing snoRNP complexes. NPM1 is an important regulator of this process because it binds C/D box snoRNAs and supports proper rRNA 2′-O-methylation.⁴⁰ Inactivation of Npm1 in adult HSCs leads to BMF. Germline mutations in NPM1 have been identified in dyskeratosis congenita patients with BMF, demonstrating a loss of snoRNA-binding capacity.⁴¹ Overexpression of HMGB1 resulted in reduced methylation modifications at the 28S rRNA sites Am2388, Cm2409, and Gm2411, thereby elevating ribosomal RNA expression levels and enhancing protein synthesis. Bioprocess analysis revealed that genes upregulated by HMGB1 expression were highly enriched in positive regulation of transcription at RNA polymerase II promoters, negative regulation of osteoclast differentiation, positive regulation of apoptosis, and DNA-dependent transcription. HMGB1 selectively binds to the H/ACA box snoRNA SNORA63, thereby promoting transcription of genes involved in osteoclast differentiation and apoptosis.⁴²

Research into SNHGs in osteoclast differentiation and bone resorption remains in its infancy. Mechanical stimulation downregulated SNHG5 and increased the RANKL/OPG ratio in human periodontal ligament cells. Overexpression of SNHG5 suppressed RANKL expression and regulated osteoclast differentiation via C/EBPβ.⁴³ Research has demonstrated that SNHG15 is significantly upregulated in RANKL/M-CSF-induced human monocytic THP-1 cells, where it directly promotes osteoclast differentiation and migratory capacity by regulating key markers including cathepsin K (CTSK), matrix metalloproteinase 9 (MMP-9), activated T cell nuclear factor 2 (NFAT2), and tartrate-resistant acid phosphatase (TRAP).⁴⁴ These findings suggest that SNHGs can regulate bone metabolism through the osteoblast-osteoclast coupling. SNHGs function both as osteoblast-promoting factors and, under specific microenvironments, drive bone resorption, thereby maintaining the dynamic equilibrium of bone remodeling (Figure 1).

Figure 1.

Regulatory roles of snoRNAs and SNHGs in bone metabolism

(A) SnoRNAs and SNHGs in osteogenic differentiation. BMSC-derived pre-osteoblasts undergo osteogenic differentiation to form bone. snoRNAs modulate this process through RNA modification and related regulatory mechanisms. SNHG5 promotes osteogenesis via the YY1/SNHG5 and miR-212-3p/GDF5/Smad1/5/8 axes, and as a ceRNA for miR-582-5p to regulate RUNX3. SNHG14 enhances NEDD4L mRNA stability, promoting FOXA2 ubiquitination and subsequent PCP4 upregulation. SNHG16 facilitates osteogenic differentiation through the miR-485-5p/BMP7 axis.

(B) SnoRNAs and SNHGs in osteoclast differentiation and bone resorption. HSCs differentiate into BMMs, which further differentiate into multinucleated osteoclasts under RANKL and M-CSF stimulation, leading to bone resorption. SNORD113-114 regulates HSC self-renewal through translational control and p53-mediated apoptosis. U3 snoRNA activates DNA-PK and maintains rRNA processing. Box C/D snoRNAs guide, and NPM1 supports, rRNA 2′-O-methylation associated with translational control. SNORA63, in association with HMGB1, may influence ribosomal function and osteoclast-related gene transcription. SNHG15 promotes osteoclast differentiation by regulating CTSK, MMP9, NFAT2, and TRAP, and counteracts miR-381-3p.

SnoRNAs and SNHGs contribute to bone-related diseases

SnoRNAs and SNHGs exhibit altered expression profiles across multiple skeletal disorders and have been implicated in regulating pathological processes. These molecules function through diverse signaling pathways by integrating mechanisms spanning transcription, post-transcriptional regulation, and intercellular communication. Together, these observations broaden our understanding of the role of snoRNAs and SNHGs in bone metabolism and provide a conceptual framework for exploring targeted interventions in bone remodeling (Figure 2).

Figure 2.

SnoRNAs and SNHGs in bone diseases and healing

(A) Osteoporosis. SNORD118 deletion activates p53 and reduces BMSC proliferation. Altered snoRNA-guided rRNA modification may impair translational efficiency and reduce osteogenic protein synthesis. SNHG14 and SNHG16 promote osteogenesis via miR-185-5p/WISP2 and miR-485-5p/BMP7 axes, respectively. SNHG18 knockout inhibits YAP nuclear translocation, accelerating bone loss.

(B) OA. RMRP mutations and SNORD26/SNORD96A dysregulation disrupt cartilage homeostasis. sdRNA-D43 targets NRF1/WIPI2 to inhibit mitophagy. SNHG1 reduces ROS and activates STAT3 to promote chondrogenesis. SNHG5 and SNHG7 protect chondrocytes via the miR-26a/SOX2 and miR-485-5p/FSP1 or miR-324-3p/DUSP1 axis, respectively.

(C) Bone tumors. In MM, ACA11/SCARNA22 drives ribosome biogenesis and proliferation in a ROS-dependent manner; SNORD78 promotes oncogenic rRNA methylation. In osteosarcoma, altered snoRNA expression contributes to Dox resistance, partly through GADD45A/TOP2A-related regulation. Mutant p53 upregulates snoRNAs via Ets2 to facilitate metastasis. SNHG1, SNHG6, and SNHG12 promote tumor progression through miR-577/WNT2B, EZH2/KLF6, and miR-320a/MCL1 axes, respectively.

(D) Fracture healing. Serum SNHG1 is elevated in delayed healing and inhibits osteogenic activity via miR-181a-5p. Hypoxia-induced SNHG2 and SNHG7 promote BMSC proliferation and migration. SNHG5 enhances bone formation through GDF5.

SnoRNAs and SNHGs in osteoporosis

Osteoporosis is a systemic bone disorder characterized by bone mass loss and microstructural deterioration, with its core pathological mechanism residing in the dynamic imbalance between osteoblast-mediated bone formation and osteoclast-driven bone resorption. As precursors to osteoblasts, the proliferation and differentiation capacity of BMSCs directly determine the rate of bone formation, while the functional decline of BMSCs is recognized as a key driver of senile osteoporosis.⁴⁵ SnoRNAs and SNHGs are increasingly recognized as post-transcriptional regulators of bone metabolism, finely modulating the fate of BMSCs and osteoblast function by controlling ribosomal biosynthesis, translational efficiency, and bone-related signaling networks. Depletion of Snord118 leads to p53 activation, increased cell death, reduced proliferation, and premature osteogenic differentiation of MSCs, leading to suture growth and premature cranial suture opening. Snord118 deficiency causes dysregulation of ribosomal protein translation and downregulation of complement pathway genes. Complement C3a receptor 1 (C3ar1) knockout exacerbates MSC and cranial suture defects in mutant mice. Activation of the complement pathway can rescue MSC cell and suture growth defects.⁴⁶ Moreover, abnormal snoRNA-mediated rRNA modifications—such as reduced 2′-O-methylation guided by SNORD123—may impair ribosomal translation efficiency.⁴⁷ This may contribute to insufficient synthesis of key osteogenic proteins, including COL1A1 and OCN, thereby providing a potential mechanistic basis for osteoporosis-related bone mass reduction.

In osteoporosis, accumulating evidence suggests that SNHGs commonly function as positive regulators of osteogenesis, and their downregulation in patient-derived hBMSCs contributes to impaired bone formation. A recurrent mechanism is that SNHGs act as ceRNAs, sequestering specific miRNAs and thereby relieving repression of osteogenesis-related target genes. Representative examples include SNHG14 and SNHG16, both of which are significantly downregulated in hBMSCs from patients with osteoporosis. SNHG14 promotes osteogenic differentiation by regulating the miR-185-5p/WISP2 axis,⁴⁸ whereas SNHG16 exerts a similar pro-osteogenic effect through the miR-485-5p/BMP7 pathway.²⁸ These findings indicate that distinct SNHGs may converge on a common regulatory logic in osteoporosis, namely, the maintenance of osteogenic competence through SNHG/miRNA/mRNA ceRNA networks. Not all SNHGs, however, act through the same mechanism. In contrast to the ceRNA-dominated modes of SNHG14 and SNHG16, SNHG18 appears to regulate bone homeostasis through a stem/progenitor cell-related signaling pathway. SNHG18 has been identified as a key osteogenic regulator in LepR⁺ cells, where its deletion accelerates bone loss and suppresses Yap nuclear translocation, while its upregulation restores bone mass and Yap signaling in osteoporosis models.⁴⁹ In addition, SNHG-related pathways may also have translational relevance, as zoledronic acid has been reported to alleviate colorectal cancer-associated osteoporosis partly through modulation of the SNHG16/miR-146a/TRAF6 axis.⁵⁰ Collectively, these findings suggest that SNHGs contribute to osteoporosis pathogenesis through both shared ceRNA-mediated mechanisms and distinct signaling-associated pathways, underscoring their potential as candidate biomarkers and therapeutic targets for bone loss disorders.

Beyond osteoporosis, SNHGs have also been implicated in pathological bone formation. SNHG12, for instance, is upregulated in ectopic ossification tissues and promotes aberrant osteogenesis through the miR-199a-5p/Fzd4/Wnt/β-catenin pathway.⁵¹ Although this process is distinct from osteoporosis, it further supports the broader concept that SNHGs participate in pathological bone remodeling across different disease contexts by modulating osteogenic lineage programs.

Notably, biologically relevant snoRNA host genes in skeletal disease are not restricted to the SNHG-named family. GAS5 is a classical non-coding multi-snoRNA host gene and has been implicated in several bone-related conditions.⁵² In experimental osteoporosis settings, GAS5 has been reported to promote osteoblast differentiation through the UPF1/SMAD7 axis,⁵³ whereas circulating GAS5 is elevated in some patient cohorts with osteoporosis and bone fragility,⁵⁴ suggesting potentially context- and specimen-dependent behavior. GAS5 has also been linked to osteoarthritis (OA) and osteosarcoma,^55,56 underscoring that the host-gene landscape relevant to bone disease extends beyond SNHG-family lncRNAs.

Overall, snoRNAs and SNHGs have emerged as important regulators of osteoporosis by coordinating translational control, osteogenic signaling, and BMSC fate determination. These findings not only deepen our understanding of the non-coding RNA network underlying bone loss, but also support the development of candidate snoRNA- and SNHG-based biomarkers and targeted interventions for osteoporosis.

SnoRNAs and SNHGs in osteoarthritis

OA is characterized by progressive articular cartilage degeneration driven by chondrocyte senescence, inflammatory activation, extracellular matrix (ECM) breakdown, and impaired reparative capacity. Increasing evidence indicates that snoRNAs and SNHGs participate in these processes by regulating chondrocyte differentiation, mRNA translation, mitochondrial homeostasis, and stress-responsive signaling pathways.

Although chondro-hair hypoplasia (CHH) is a developmental skeletal disorder rather than OA, it provides important conceptual evidence that dysregulation of snoRNA-related non-coding RNAs can profoundly affect cartilage biology. CHH is a rare form of epiphyseal chondrodysplasia characterized by disproportionate short stature and abnormal growth plate development, and it is caused by mutations in RNA component of mitochondrial RNA processing endoribonuclease (RMRP), the non-coding RNA component of RNase MRP. During chondrogenic transdifferentiation, CHH cells display marked upregulation of BMP, fibroblast growth factor (FGF), and IGF-1 signaling, indicating that RMRP dysfunction can alter multiple aspects of chondrocyte fate and signaling regulation.⁵⁷ Although CHH is not an OA phenotype per se, these findings suggest that snoRNA-related ncRNAs are essential for cartilage development and homeostasis, thereby providing a basis for understanding their involvement in degenerative cartilage diseases such as OA.

In OA, several snoRNAs appear to regulate cartilage homeostasis more directly. SNORD26 and SNORD96A have been implicated in chondrogenesis, rRNA-associated regulation, and OA-related gene expression, and altered levels of these snoRNAs affect both chondrogenic differentiation and the expression of cartilage-associated genes.⁵⁸ U3 snoRNA is also critical for maintaining the translational capacity of chondrocytes in OA. Perturbation of U3 expression alters chondrocyte phenotype, while proteomic analyses indicate that reduced U3 levels significantly disrupt protein synthesis and inflammatory pathways.⁵⁹ In addition, the snoRNA-derived fragment sdRNA-D43 is upregulated in damaged regions of OA cartilage and promotes chondrocyte senescence by targeting NRF1 and WIPI2 to suppress PINK1/Parkin-mediated mitophagy.⁶⁰ These findings suggest that snoRNA dysregulation in OA extends from defects in cartilage differentiation and ribosome-associated protein synthesis to mitochondrial quality control and senescence-related degeneration.

Compared with snoRNAs, SNHGs in OA are more commonly involved in the post-transcriptional and signaling-level regulation of inflammatory injury, apoptosis, and ECM metabolism. Several SNHGs exert broadly protective effects on chondrocytes under inflammatory conditions. SNHG1 promotes chondrogenic differentiation and angiogenesis of BMSCs in vivo and in vitro, partly by reducing ROS, increasing mitochondrial membrane potential and ATP production, and sustaining STAT3 phosphorylation.³⁰ In chondrocytes, SNHG1 also alleviates IL-1β-induced metabolic dysfunction by reducing the expression of matrix-degrading enzymes, including MMPs and ADAMTS family members, and by modulating miR-16-5p-dependent p38 MAPK and NF-κB signaling.⁶¹ Moreover, SNHG1 overexpression inhibits chondrocyte apoptosis in OA mouse models through activation of the PI3K/Akt pathway and suppression of autophagy.⁶²

A similarly protective pattern has been reported for SNHG5, SNHG7, and SNHG15. SNHG5 is downregulated in OA cartilage and suppresses IL-1β-induced chondrocyte apoptosis through the miR-10a-5p/H3F3B axis, while also enhancing chondrocyte proliferation via SNHG5/miR-26a/SOX2 signaling and attenuating inflammatory responses and ECM degradation through the miR-181a-5p/TGFBR3 pathway.^63,64,65 SNHG7 is reduced in IL-1β-treated chondrocytes and OA cartilage, and its overexpression alleviates inflammatory injury, oxidative stress, and cell death. Notably, SNHG7 can be delivered by BMSC-derived exosomes and protects chondrocytes through both the miR-485-5p/FSP1⁶⁶ and miR-324-3p/DUSP1 axes,⁶⁷ highlighting an intercellular communication mechanism in OA regulation. SNHG15 is likewise decreased in OA cartilage and IL-1β-stimulated chondrocytes; its overexpression suppresses ECM degradation and promotes chondrocyte survival and differentiation by regulating the miR-7/KLF4 axis and downstream β-catenin-related signaling.⁶⁸ Not all SNHGs appear to act in a protective manner, however. SNHG9 is downregulated in OA and inhibits chondrocyte apoptosis by suppressing miR-34a through a methylation-associated mechanism,⁶⁹ whereas SNHG14 may play a distinct and potentially deleterious role. In lipopolysaccharide-induced chondrocytes, inhibition of SNHG14 attenuates inflammatory responses, apoptosis, and ECM degradation through miR-137,⁷⁰ suggesting that the functions of SNHGs in OA are context-dependent rather than uniformly protective.

Collectively, these findings indicate that snoRNAs and SNHGs contribute to OA pathogenesis through complementary but distinct mechanisms. snoRNAs are more closely linked to cartilage developmental programming, ribosome-associated translation, and mitochondrial quality control, whereas SNHGs mainly regulate inflammatory signaling, apoptosis, ECM remodeling, and intercellular communication through ceRNA networks and related pathways. Together, they form an important non-coding RNA regulatory layer underlying cartilage degeneration and may represent candidate biomarkers and therapeutic targets for OA.

SnoRNAs and SNHGs in bone tumors

The regulatory networks of snoRNAs and SNHGs in bone malignancies have attracted increasing attention in bone tumor studies. Evidence indicates that these non-coding RNAs may influence core pathological events in bone malignancies—including osteosarcoma and multiple myeloma (MM)—by driving abnormal ribosomal biosynthesis, remodeling the tumor microenvironment, and regulating post-transcriptional gene silencing. These mechanisms have been associated with chemotherapy resistance, metastatic dissemination, and poor prognosis. Representative mechanisms are summarized below.

ACA11/SCARNA22, an orphan H/ACA-type scaRNA encoded within an intron of NSD2,⁷¹ is a key factor in the response to standard chemotherapy regimens for MM. ACA11 increases nuclear area and cell number in MM cells, alongside argyrophilic nuclear organizing regions. ACA11 upregulates ribosome biogenesis, 47S pre-rRNA synthesis, and protein synthesis in a ROS-dependent manner.⁷² Poor prognosis in MM is associated with chromosomal aberrations such as gain(1q), del(17p) and translocation t(4;14).⁷³ Analysis of 71 newly diagnosed MM patients and four relapsed/refractory MM patients revealed that SNORD78, SNORD75, and SNORD47 were all overexpressed and located on 1q25. SNORD78 exhibited elevated expression in patients with gain(1q) and was associated with poorer prognosis. SNORD78-mediated rRNA methylation may exert an oncogenic role in MM.⁷⁴ ACA11 overexpression promotes cell proliferation in MM cell lines and primary splenic B cells from mice, and this highly proliferative phenotype is dependent on increased ROS levels.⁷⁵

In osteosarcoma, altered snoRNA expression has also been linked to chemotherapy resistance. Research reveals a global upregulation of snoRNAs in human doxorubicin (Dox)-resistant osteosarcoma cells. Among these, overexpression of SNORD3A, SNORA13, and SNORA28 in Dox-sensitive cells reduced Dox cytotoxicity. Within these cells, GADD45A and MYC proto-oncogene (MYC) were upregulated, while TOP2A was downregulated. The same expression pattern was detected in Dox-resistant cells. Silencing GADD45A/MYC and overexpressing TOP2A counteracted the doxorubicin resistance induced by snoRNAs.⁷⁶ A mouse osteosarcoma model has been established to investigate tumor metastasis. This model employs lineage-specific Cre to express the gain-of-function mutant p53R172H in osteoblasts. Tumor cells were additionally genetically labeled with Cre-inducible GFP, enabling sensitive tracking of metastatic nodules in secondary organs. Results revealed a significantly increased incidence of pulmonary metastasis in p53-mutant mice.⁷⁷ Comprehensive transcriptomic RNA sequencing analysis of tumor samples revealed a set of snoRNAs highly upregulated in p53-mutant tumors. Homozygous deletion of Ets2 in p53-mutant mice resulted in marked downregulation of these snoRNAs, reversing the tumor metastasis phenotype of mutant p53 while having no effect on osteosarcoma development.⁷⁸ These findings suggest that Ets2 is the primary effector of mutant p53 in the metastasis of osteosarcoma. The development of anti-metastatic therapeutics, such as inhibitors blocking the interaction between Ets2 and mutant p53, offers a potential therapeutic avenue for treating osteosarcoma. SNORA7A expression correlates with the activation of oncogenic signaling pathways and poor survival rates in cancer patients. SNORA7A has been implicated in H19-associated regulatory networks involving osteogenesis, DNA damage response, and osteosarcoma suppression, suggesting that dysregulated H19–SNORA7A signaling may contribute to Li-Fraumeni syndrome-associated osteosarcoma.⁷⁹

In osteosarcoma, multiple SNHG family members are aberrantly upregulated and commonly associated with aggressive clinicopathological features, including larger tumor size, advanced tumor-node-metastasis (TNM) stage, lymph node metastasis, recurrence, and poor survival. Rather than acting as isolated biomarkers, these SNHGs appear to form an oncogenic regulatory network that promotes tumor progression primarily through ceRNA-dependent mechanisms. A recurrent pattern is that SNHGs function as molecular sponges for tumor-suppressive miRNAs, thereby derepressing downstream oncogenic targets involved in proliferation, migration, invasion, and metastasis. SNHG1 is a representative example, as it promotes osteosarcoma progression through both the miR-577/WNT2B/Wnt/β-catenin axis⁸⁰ and the miR-326/NOB1⁸¹ pathway. Similarly, SNHG5 enhances osteosarcoma cell proliferation, invasion, and migration by acting as a ceRNA for miR-26a and activating ROCK1 signaling.⁸² SNHG15 also exerts oncogenic effects through miRNA-mediated mechanisms, as it directly interacts with miR-141⁸³ and miR-346⁸⁴ to regulate downstream targets including TRAF4, thereby promoting proliferation, invasion, migration, and survival of osteosarcoma cells. SNHG16 likewise facilitates osteosarcoma malignancy by modulating the miR-488/ITGA6 axis⁸⁵ and by relieving miR-205-mediated repression of ZEB1,⁸⁶ thus contributing to cell migration, invasion, and epithelial-mesenchymal transition. In addition, SNHG8 has been linked to enhanced osteosarcoma proliferation and invasion,⁸⁷ at least in part through suppression of miR-542-3p.⁸⁸ Beyond these shared ceRNA-based oncogenic effects, some SNHGs are particularly relevant to clinical stratification or therapeutic response. SNHG4 is significantly upregulated in osteosarcoma tissues and cell lines, and its high expression correlates with larger tumor size and poor prognosis, while its knockdown inhibits cell survival and invasive potential.⁸⁹ SNHG12, by contrast, is notable for its role in chemotherapy resistance. Elevated SNHG12 expression is associated with Dox resistance and unfavorable overall survival in osteosarcoma patients, and mechanistic studies indicate that SNHG12 promotes chemoresistance through the miR-320a/MCL1 axis.⁹⁰ Collectively, these findings suggest that SNHGs contribute to osteosarcoma progression through convergent ceRNA-mediated oncogenic programs as well as therapy resistance-associated pathways. This mechanistic framework not only explains how distinct SNHGs drive malignant behavior in osteosarcoma but also highlights their potential value as candidate prognostic biomarkers and therapeutic targets.

SnoRNAs and SNHGs in fracture healing

Fracture healing is a highly coordinated regenerative process that depends on the proliferation, migration, survival, and osteogenic differentiation of skeletal precursor cells, as well as adaptation to the local injury microenvironment. Emerging evidence suggests that SNHGs participate in multiple aspects of this process and may influence whether repair proceeds normally or progresses toward delayed healing.

Some SNHGs appear to be associated with impaired fracture repair. In patients with delayed fracture healing, serum SNHG1 levels are significantly elevated. Functionally, SNHG1 suppresses the proliferation of MC3T3-E1 mouse pre-osteoblastic cells, reduces osteogenic marker expression, and promotes apoptosis, whereas upregulation of miR-181a-5p partially reverses these inhibitory effects on osteogenic differentiation and cell survival.⁹¹ These findings suggest that SNHG1 may serve not only as a candidate biomarker of delayed fracture healing but also as a potential negative regulator of osteogenesis during repair. By contrast, other SNHGs seem to support regenerative responses during fracture healing. SNHG7 is downregulated in femoral neck fracture tissue, and its knockdown inhibits BMSC proliferation and migration, suppresses TGF-β signaling, induces apoptosis, and reduces osteoblastic activity in vitro,⁹² indicating that SNHG7 may facilitate bone repair by sustaining BMSC function and osteogenic signaling. Consistent with a pro-regenerative role, cranial defect models have shown that knockdown of SNHG5, together with GDF5 suppression, inhibits new bone formation in vivo,²⁵ further supporting the importance of SNHG-mediated regulation in skeletal regeneration. In addition, SNHG2 has been reported to respond to hypoxic stimuli and to participate in fracture healing,⁹³ suggesting that certain SNHGs may also contribute to microenvironmental adaptation during the repair process. Collectively, these findings indicate that SNHGs regulate fracture healing in a context-dependent manner, with some contributing to delayed repair and others promoting bone regeneration. This highlights their potential translational value as candidate biomarkers for healing prognosis and as candidate therapeutic targets for enhancing fracture repair.

SnoRNAs and SNHGs have biomedical potential in skeletal disease

SnoRNAs and SNHGs as candidate diagnostic tools for bone-related diseases

Following bone injury or under pathological conditions, osteoblasts alter EV production and modify snoRNA expression, highlighting the potential of snoRNAs and SNHGs as candidate diagnostic molecules for bone-related diseases. Large quantities of snoRNAs exist in the circulation, where they are protected from RNase degradation by binding to high-density lipoproteins or by being encapsulated within EVs.⁹⁴ Osteocytic populations utilize EVs for intercellular communication, particularly during bone remodeling. SNHG7, which is abundant in BMSC-derived exosomes, can be endocytosed by chondrocytes, and overexpression of SNHG7 in BMSC-derived exosomes suppresses chondrocyte inflammation and ferroptosis in OA.⁶⁶ Moreover, altered expression profiles of specific snoRNAs, such as SNORD3A, SNORA73B, SNORD46, and SNORA26, have been observed in osteosarcoma patients and may serve as prognostic markers.⁹⁵ In patients with bone metastatic cancer, the expression of snoRNAs such as SNORD78, SNORD75, and SNORD47 in plasma and tumor tissue correlates with the risk of bone metastasis.⁷⁴ Therefore, circulating snoRNAs may have value as candidate biomarkers for bone diseases. SnoRNAs and SNHGs can be detected in plasma, serum, and extracellular vesicles,⁹⁶ and their potential utility in liquid-biopsy-based stratification is being actively explored. Nevertheless, technical and biological challenges persist in detecting and quantifying circulating snoRNAs, with biases potentially arising at multiple stages, including RNA isolation, library preparation, sequencing, and bioinformatics analysis.⁹⁷ Accordingly, current results should be interpreted with caution, and rigorous cross-cohort validation, standardized preanalytical procedures, and reproducible analytical pipelines will be essential before clinical utility can be established. The development of small RNA-sensitive RNA sequencing and bioinformatics tools will be important for providing the technical foundation for validating snoRNA biomarkers in bone diseases.

SnoRNAs and SNHGs as therapeutic targets

Compared with protein-targeted drug discovery, direct pharmacological targeting of snoRNAs is likely to be more difficult in many settings. Nucleic acid therapeutics (NATs), small molecule drugs and gene editing technologies have undergone rapid development, offering possible strategies for snoRNA intervention. Although these approaches were initially developed to target proteins and mRNAs,⁹⁸ their technical frameworks are conceptually adaptable to applications targeting snoRNAs and SNHGs. In fact, targeted strategies against snoRNAs have shown preliminary promise in bone malignancies such as osteosarcoma. The most commonly employed approach for snoRNA targeting is NATs, particularly antisense oligonucleotides (ASOs), gapmers, and locked nucleic acid oligonucleotides. Consequently, through delivery optimization approaches such as ligand-conjugated modifications and encapsulation within lipid/polymeric nanoparticles, the pharmacokinetic properties, target affinity, and biological stability of NATs have been systematically enhanced.⁹⁹ In osteosarcoma, abnormally expressed snoRNAs influence tumor proliferation and metastasis by regulating signaling pathways such as MAPK, Wnt, and TGF-β. Consequently, targeting these snoRNAs with ASOs may inhibit tumor progression. Furthermore, SNHG6 acts as an EZH2 scaffold in chondrosarcoma to promote the silencing of the tumor suppressor gene KLF6, suggesting that targeting the SNHG6-EZH2 interaction may represent a potential epigenetic therapeutic strategy.¹⁰⁰

In addition, the development of RNA-targeting small molecules more broadly remains constrained by challenges in identifying selective ligandable pockets, accounting for RNA conformational dynamics, and achieving sufficient intracellular specificity.¹⁰¹ Accordingly, although RNA-binding small molecules remain conceptually attractive, snoRNA-directed therapeutic development may in many cases be more tractable through antisense-based approaches, modulation of snoRNP-associated proteins, or interference with snoRNA biogenesis and function rather than through direct small-molecule binding alone.¹⁰²

For SNHGs, therapeutic intervention may be somewhat different. Because many SNHGs function as longer regulatory transcripts that act through ceRNA networks,¹⁰³ chromatin-associated mechanisms,¹⁰⁴ or RNA-protein scaffolding,¹⁰⁵ ASOs and gapmers may be more suitable than indiscriminate small-molecule screening in many contexts.¹⁰⁶ In such cases, therapeutic success will depend not only on knockdown efficiency but also on defining the disease-relevant transcript isoform, minimizing off-target hybridization, and identifying the dominant functional mechanism in a given skeletal context. More broadly, RNA-targeting small molecules remain an expanding area of drug discovery, particularly for structured RNAs,¹⁰⁷ although this strategy is still at an early stage for skeletal disease applications.

Targeted drug delivery to snoRNAs and SNHGs

The utilization of targeted ligands—such as peptides, antibodies, aptamers, inhibitors, or small molecules—may improve the efficiency of drug delivery and tissue targeting. The structural specificity of bone tissue presents both opportunities and barriers for targeted delivery in snoRNA-based therapies. Given the characteristics of the bone tissue pathological microenvironment and advances in targeted ligand technology, the precise delivery of snoRNAs and SNHGs for bone-related diseases warrants further investigation. Future research may integrate dual-targeting strategies responding to the acidic pH of the bone microenvironment and cell surface markers. Injectable, pH-responsive, or protease-sensitive hydrogels releasing oligonucleotides targeting snoRNAs may have utility in bone defect and fracture healing models.^108,109 It is noteworthy that while the hypovascular nature of bone tissue and the characteristics of its dense matrix limit the efficiency of conventional delivery methods, they create conditions conducive to sustained action for local sustained-release systems, such as those utilizing mineralized collagen scaffolds.^110,111 Clinical translation necessitates addressing issues concerning the retention of activity following chemical modification of oligonucleotide cargoes, off-target effects, and immunogenicity, alongside establishing pharmacokinetic evaluation criteria for bone tissue-specific delivery. RNA-based gene therapy requires therapeutic RNAs to exert their effects within target cells without triggering undesirable immune responses. Viral vectors and non-viral delivery systems have been developed to protect RNAs from degradation and enhance their delivery to target cells.¹¹² With the advancement of snoRNA functional studies in bone diseases and the integration of smart responsive nanomaterials, targeted ligand-mediated snoRNA/SNHG therapies may support future precision approaches for bone disorders. RNA interventions for bone-related disorders such as osteoporosis and bone tumors necessitate in-depth analysis of pathological mechanisms within the bone microenvironment, precise identification of skeletal-cell-specific molecular targets, and development of delivery systems tailored to the physiological characteristics of bone tissue. Given the anatomical characteristics of bone tissue—its low vascularization and dense matrix—developing delivery vehicles for non-coding RNAs (snoRNAs and SNHGs) that possess both bone-targeting capability and transmembrane functionality¹¹³ represents a core technical challenge. For snoRNA-directed strategies, effective delivery may also require access to the appropriate subcellular compartment, including the nucleus and, in some cases, the nucleolus, which further raises the translational threshold. Consequently, comprehensively elucidating the molecular mechanisms of bone disorders, identifying therapeutically actionable targets, and optimizing delivery system efficacy will be important determinants of the clinical translation of RNA therapeutics in the field of skeletal diseases (Figure 3).

Figure 3.

Clinical translation of snoRNAs and SNHGs in bone diseases

(A) Circulating snoRNAs are protected from RNase degradation by high-density lipoproteins (HDL) binding or EV encapsulation, serving as candidate biomarkers for bone remodeling and pathology. BMSC-derived exosomal SNHG7 may modulate chondrocyte inflammation. SNORD3A/46/73B/26 have been associated with osteosarcoma prognosis, whereas SNORD78/75/47 have been linked to bone metastasis. Technical challenges include preanalytical and analytical variability, such as RNA isolation, library prep, and bioinformatics biases.

(B) Therapeutic targeting. ASOs, small molecules, and gene editing represent potential strategies for snoRNA intervention. Antisense-based approaches may help suppress osteosarcoma-associated pathways, whereas targeting the SNHG6-EZH2-KLF6 axis may represent an epigenetic strategy for chondrosarcoma. Direct pharmacological targeting of snoRNAs remains challenging because of RNA structure, intracellular accessibility, and target selectivity.

(C) Bone-targeted delivery. Ligand-conjugated carriers, including peptides, antibodies, and aptamers, may improve tissue targeting. Platforms: injectable hydrogels, pH/protease-responsive systems, scaffolds, and viral/non-viral vectors. Barriers: low vascularization, dense ECM, off-target effects, immunogenicity, and activity loss. For snoRNA-directed strategies, effective delivery may also require access to the nucleus or nucleolus. Bone-specific pharmacokinetic evaluation criteria remain to be established.

Discussion

Summary of skeletal functions of snoRNAs and SNHGs

Taken together, the available literature suggests that the skeletal functions of snoRNAs and SNHGs can be provisionally organized into several recurrent mechanistic axes. First, many snoRNA-associated effects converge on ribosome biology, RNA modification, and translational control, thereby influencing osteogenic differentiation, osteoclast precursor fitness, cartilage homeostasis, and tumor cell adaptation. Second, a substantial fraction of the reported skeletal phenotypes can be interpreted as changes in cell-state regulation, including lineage allocation of BMSCs, activation of osteoclast-lineage cells, and stress adaptation of chondrocytes. Third, the SNHGs discussed in skeletal disease are predominantly linked to post-transcriptional and epigenetic regulatory programs, including ceRNA activity, transcript stability, and chromatin- or protein-associated regulation.

Overall, snoRNAs and SNHGs have emerged as important regulators of bone development and bone-related diseases. Previously, snoRNAs were thought to function mainly in 2′-O-methylation and pseudouridylation. However, additional roles have now been proposed, including functions related to N4-acetylcytidine (ac4C) modification, antisense regulation, and microRNA-like activity.¹¹⁴ In this review, we examine representative and mechanistically supported cases of snoRNA involvement in bone development, function, and disease, considering their diagnostic and therapeutic implications.

Although several upstream regulatory principles of snoRNAs and SNHGs appear to be shared across diseases, their biological outputs are strongly shaped by tissue context.¹¹⁵ In hematologic diseases such as leukemia, these molecules are more often linked to hematopoietic lineage specification, stem-cell maintenance, ribosome biogenesis, translational control, malignant proliferation, and treatment resistance.¹¹⁶ By contrast, in skeletal biology, the currently available evidence places greater emphasis on lineage allocation and functional coupling among BMSCs, osteoblasts, osteoclast-lineage cells, and chondrocytes. Accordingly, in bone physiology and bone disease, snoRNAs and SNHGs are discussed primarily in relation to osteogenic differentiation, bone resorption, cartilage homeostasis, ECM remodeling, inflammatory microenvironments, and mechanical adaptation. Thus, the major distinction is not necessarily the existence of entirely different molecular rules, but rather the cell-type-specific pathways and pathological outputs through which these non-coding RNAs are manifested.

Methodological challenges and conceptual gaps

This review should be interpreted as a focused synthesis of snoRNAs and SNHGs in skeletal biology, rather than as a comprehensive catalog of all snoRNA host genes. Given the current state of the field, a fully exhaustive catalog of reported snoRNA- and SNHG-associated changes in skeletal disease risks overemphasizing studies that remain descriptive or technically uneven. We therefore interpret the literature with an emphasis on studies that provide functional perturbation, mechanistic support, or in vivo relevance.

Several limitations should be considered when interpreting the current evidence. We treat circulating snoRNAs as emerging but still insufficiently validated biomarker candidates, rather than established clinical tools. Although snoRNAs and SNHG-family transcripts are genomically linked in many loci, they should not be regarded as unified functional entities by default, because host transcripts can exert biological effects that are independent of the embedded snoRNAs.¹³ In addition, many skeletal studies are still based primarily on differential expression analyses followed by limited functional validation in single disease models.

In parallel, snoRNA annotation remains incomplete, a substantial proportion of snoRNAs still lack validated targets or clearly defined functions, and transcriptome-wide identification of snoRNA interaction partners has only recently become technically feasible.¹¹⁷ For SNHGs, an additional challenge is that the relationship between the mature host transcript and the embedded snoRNA is often inferred rather than directly dissected. Accordingly, current evidence should be interpreted cautiously, and future progress will require integrated studies combining accurate annotation, perturbation-rescue experiments, target-mapping technologies, and cell-type-specific skeletal models.

Future directions

Several directions now merit particular attention. First, a deeper mechanistic understanding is required of how combinations of snoRNA-mediated rRNA modifications determine the preferential translation of specific mRNAs during osteogenic differentiation, osteoclast activation, cartilage degeneration, and tumor adaptation. Second, the potential role of snoRNAs and SNHGs in the bone microenvironment should be defined more precisely. Pathological signals from the bone microenvironment, including tumor-derived factors, mechanical stress, inflammatory cues and metabolic changes, may dynamically influence the expression of snoRNAs and SNHGs. At present, however, this reciprocal framework should be regarded as an emerging research direction rather than a fully established principle.

An additional emerging dimension of snoRNA biology is the unexpected detection of snoRNAs at the outer cell surface. Recent studies profiling surface RNAs in mammalian cells have shown that, in addition to glycoRNAs, multiple classes of membrane-associated non-coding RNAs can be detected extracellularly, including snoRNAs and scaRNAs.¹¹⁸ Functional evidence from immune systems further suggests that cell-surface RNAs can participate in intercellular signaling, for example by regulating neutrophil recruitment.^119,120 Although direct evidence in skeletal cells is still lacking, this concept may be relevant to bone science because osteoimmunology, marrow-derived progenitors, and inflammatory cell trafficking are integral to bone homeostasis and pathological bone loss. The biological significance of surface snoRNAs in osteoblasts, osteoclast-lineage cells, BMSCs, or chondrocytes therefore warrants further investigation.

Finally, translating these discoveries into clinical practice remains a major challenge. Although preclinical models using ASOs targeting snoRNAs have shown promise,¹²¹ efficient and cell-selective delivery of RNA therapeutics to skeletal tissues remains difficult because of the low vascularity, dense ECM, and complex cellular composition of bone. Future translational studies should prioritize bone-adapted delivery systems, cell-type-specific targeting strategies, and rigorous validation of biomarker performance across independent cohorts. Taken together, snoRNAs and SNHGs represent a valuable but still incompletely characterized layer of skeletal regulation, and further mechanistic and translational studies will be required before their full diagnostic and therapeutic potential can be defined.

Limitations of the study

This review has several limitations. First, the current literature on snoRNAs and SNHGs in skeletal biology remains uneven across disease contexts. Compared with osteoporosis, OA, and osteosarcoma, evidence related to fracture healing, osteoclast-lineage regulation, and bone-specific snoRNA mechanisms is still limited. Second, many available studies rely on differential expression analyses and in vitro perturbation experiments, whereas in vivo validation, cell-type-specific models, rescue experiments, and direct target-mapping approaches remain insufficient. Third, SNHG-family transcripts and their embedded snoRNAs are often discussed together because of their genomic relationship, but their mature transcript functions are not always experimentally separated. Therefore, functional attribution to the host lncRNA, the embedded snoRNA, or both should be interpreted cautiously. Fourth, circulating snoRNAs and SNHGs are promising biomarker candidates, but their clinical utility remains constrained by cohort size, sample heterogeneity, RNA isolation procedures, sequencing depth, normalization strategies, and cross-platform reproducibility. Finally, although RNA-targeted therapeutic strategies are conceptually attractive, efficient and cell-selective delivery to skeletal tissues remains a major translational barrier because of the dense ECM, low vascularity, and cellular complexity of bone.

Acknowledgments

This research was supported by Natural Science Foundation of Liaoning Province (2025-MSLH-389).

Author contributions

Conceptualization, M.W. and L.L.; writing – original draft, W.Z. and Y.L.; writing – review and editing, W.Z., Y.L., and Q.W.; supervision, T.C. and Y.G.; funding acquisition, L.L. All authors reviewed and approved the final manuscript.

Declaration of interests

The authors declare no competing interests.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work, the authors used ChatGPT (OpenAI) only for English grammar polishing and language refinement. The tool was not used to generate scientific content, analyze data, select references, interpret results, or draw conclusions. After using this tool, the authors reviewed and edited the text as needed and take full responsibility for the content of the publication.