Alternative Splicing: A Critical Regulator in Human Bone Biology and Tumor Progression.

Introduction

Alternative splicing (AS) is an important post-transcriptional regulatory mechanism that enables a single gene to produce multiple distinct RNA and protein isoforms, thereby expanding the diversity of gene expression in eukaryotes. AS was discovered in 1977, when studies on the mRNA of adenovirus first revealed that a single gene can generate multiple mRNA transcripts through different splicing patterns [ 1 , 2 ].

Generally, RNA splicing can be categorized into 4 categories: group I self-splicing, group II self-splicing, nuclear mRNA spliceosomal splicing, and nuclear transfer RNA (tRNA) enzymatic splicing [ 3 – 5 ]. Group I splicing mainly occurs in genes of eukaryotic organelles, such as mitochondria and chloroplasts, whereas group II splicing occurs in genes of fungal mitochondria and plant chloroplasts [ 3 , 4 , 6 ]. Nuclear tRNA enzymatic splicing is mediated by specific endonucleases and has been extensively studied in yeast, a classical model organism for this mechanism [ 7 ]. Nuclear mRNA spliceosomal splicing is the process by which the spliceosome removes introns from nuclear mRNA, encompassing both constitutive splicing and AS [ 8 ]. Researchers commonly use the term “constitutive splicing” to refer to the basic form of splicing in which all exons are joined together without selection. However, as cells undergo developmental transitions and encounter increasing demands for protein function, a single gene can produce different RNA variants through AS, thereby encoding diverse protein isoforms [ 5 ]. Approximately 95% of multi-exon genes in the human genome undergo AS [ 9 ].

Precursor mRNA (pre-mRNA) splicing in humans is precisely mediated by the spliceosome and involves accurate recognition of splice sites and branch points, followed by intron removal and exon ligation [ 10 ]. AS is regulated by cis-acting elements and trans-acting factors and plays a critical role in bone biology [ 11 ]. The principal cell types in bone tissue, including osteoblasts, osteoclasts, and chondrocytes, undergo differentiation and perform functions precisely regulated by AS [ 12 ]. In addition, key components of the bone matrix, such as collagen and fibronectin, are subject to AS regulation, which influences the structure and function of bone tissue. However, most existing studies on AS in bone biology remain fragmented, with limited integration across osteogenesis, bone resorption, and bone matrix metabolism. This gap may obscure the overall connections of AS in regulating bone formation, resorption, and matrix homeostasis. In this review, we systematically summarize recent advances in AS within bone biology, emphasizing novel mechanistic insights and potential translational value.

Under pathological conditions, aberrant AS patterns are closely associated with the development and progression of bone tumors [ 13 ]. AS plays a critical role in the invasion and metastasis of osteosarcoma (OS), Ewing sarcoma (EwS), chondrosarcoma (CS), and various metastatic bone cancers, such as those originating from the prostate, bladder, and breast [ 11 ]. AS not only regulates tumor cell proliferation, migration, and apoptosis, but also contributes to the remodeling of both the tumor microenvironment and bone metastatic niche. Thus, AS variants serve as potential biomarkers for the diagnosis, prognosis, and therapeutic targeting of bone tumors [ 14 ]. Elucidating the specific mechanisms of AS and its roles in bone biology and tumor progression will provide novel insights and therapeutic strategies for the diagnosis and treatment of bone-related diseases (Fig. 1 ). Our review systematically summarizes the latest advances in the field, covering the fundamental processes and regulatory factors of human pre-mRNA splicing; the roles of key AS regulators in bone biology and bone tumors; the functions of representative AS events in bone remodeling and tumor progression; the molecular mechanisms by which RNA-binding proteins (RBPs) regulate AS in bone biology and bone tumors; and the potential translational value of AS targets in clinical applications.

Splicing Process and Regulation of Pre-mRNA in Humans

General splicing process in humans

Despite the inherent complexity of AS in humans, this section provides an overview of the most recently characterized proteins and molecular steps involved in the splicing process. Primary transcripts are single-stranded RNA products synthesized by DNA transcription. After processing, various mature RNA products, such as messenger RNA (mRNA), tRNA, and ribosomal RNA (rRNA), are produced. The pre-mRNA is a primary transcript, which is transcribed and synthesized by the DNA template in the nucleus, and mRNA is formed through the processes of 5′-cap addition, splicing, editing, and polyadenylation. Untreated pre-mRNA in the nucleus, collectively known as heterogeneous nuclear RNA (hnRNA), forms a heterogeneous nuclear ribonucleoprotein (hnRNP) after binding to a protein. Complexes of small RNA, nucleoproteins, and their attached introns are known as spliceosomes [ 10 ]. Noncoding small nuclear RNAs (snRNAs) bind to proteins to form small nuclear RNA protein particles (snRNPs), which play a precise splicing function. snRNAs (U1, U2, U4, U5, and U6 snRNAs) provide specificity for different types of splicing and identify various types of introns by base pairing (Table 1 ).

Intron 5′ splice site (5′ SS), branching point (BP) adenosine, and 3′ splice site (3′ SS) were identified during splicing [ 15 ]. In humans, nucleotide sequences flanking the 5′ SS and BP adenosine were not properly conserved, and the trinucleotide YAG sequence at the 3′ SS was preceded by a region of consecutive pyrimidine bases [ 16 ]. From a chemical perspective, the 2′ OH group of adenines located at the intron BP site initiates an attack on the phosphodiester bond, known as branching, at the intron 5′ SS. The 2′ oxygen of the first intron (G) at the 5′ SS was linked to BP adenosine, resulting in the formation of a lariat intron and the connection of the exons′ 3 OH groups [ 17 ]. Subsequently, the exon 3′ OH initiated an attack on the 5′-phosphate group of the first guanine base in the following exon, thereby substituting the 3′-terminal (known as exon ligation) of the lariat intron, ultimately leading to the fusion of the 2 exons. This process is known as constitutive splicing.

The spliceosome was not pre-assembled but rather constituted a sophisticated molecular machine (Fig. 2 A). Currently, 10 different spliceosome states have been defined: E, A, pre-B, B, B act , B * , C, C * , P, and intron lariat spliceosome (ILS) complexes, and 4 distinct splicing phases: assembly, activation, catalysis, and disassembly. In addition to the 5 snRNPs, many splicing factors including RNA helicases and the nineteen complex (NTC) contribute to spliceosome function, comprising approximately 100 proteins. In the assembly phase, the U1 snRNP of the spliceosome recognizes the 5′-GU splicing sequence of the intron, and the zinc finger domain of U1-C touches the RNA duplex directly to stabilize the 5′ SS/U1 snRNA interaction. Human splicing factors, such as TIA1 (T-cell intracellular antigen 1), LUC7L (Luc7-like protein), PRPF40 (pre-mRNA processing factor 40), and RBM25 (RNA binding motif protein 25), are involved in promoting the binding of U1 snRNP to weak 5′ SS [ 18 , 19 ]. Afterwards, splicing factor 1 (SF1) is attached to the BP site of pre-mRNA, while the cofactor U2AF65-U2AF35 heterodimer acts together with the downstream polypyrimidine tract and 3′ SS to create the E complex. The DEAD (Asp-Glu-Ala-Asp)-box helicase replaces SF1 and U2AF (U2 small nuclear RNA auxiliary factor), recruiting U2 snRNP to the BP sequences to combine with U1 snRNP to form the lariat intron structure known as the A complex. The U4/U6.U5 tri-snRNP, the largest pre-assembled complex of spliceosome-containing active sites, was subsequently recruited to build the pre-B complex. U6 snRNA is paired with 5′ SS and folded to make a splice site. U4 snRNA maintained U6 in its pre-catalytic shape. U5 snRNA tethers 5′ SS exons together or aligns 5′ SS and 3′ SS during exon connection.

In the activation phase, PRP28 (pre-mRNA processing factor 28) helicase helps U1 snRNP break away from 5′ SS [ 20 ]. Subsequently, the U6 ACAGAGA loop is bound to 5′ SS to form the B complex. Structural alterations in the B complex facilitate the assembly of the U5-200kD helicase associated with U4 snRNP to disentangle the U4/U6 pair, resulting in the formation of splice active sites. After the U4 snRNP is released, U2 and U6 snRNAs pair to form the B act complex [ 10 ]. The 4 U6 phosphates in the U2/U6 helix form a catalytic center. Under the stability of catalytic metal ions, NTC, and U2 core proteins, 5′ SS are incorporated into the catalytic center. However, SF3B in the B act complex isolates U2 snRNP and removes the 2′ OH nucleophile of BP from the active site, whereas SF3A and RNF113A (ring finger protein 113A) are sequestered at 5′ SS. Upon entering the catalysis phase, the DHX16 (DEAH-box helicase 16) helicase remodels the B act complex, enabling docking of the branch helix at the BP and 5′ SS, after which the active site catalyzes the branching reaction to form the B* complex. The C complex is formed when the phosphate group of the first intron is broken at 5′ SS (G+1 position) and connected to the 2′ OH of adenosine at the BP. The B* and C complexes are functionally similar, representing the spliceosome at the beginning (B*) and completion (C) of the branching step. Subsequently, DHX38 helicase restructures the C complex, leading to the transformation of the spliceosome into the C* complex before exon ligation [ 21 ]. 3′ SS is introduced into the active site of the C* complex, and with PRPF18 and synthetic lethal with U5 snRNA 7 homolog (SLU7), 5′ SS and BP enter the process of docking recognition. In the last stage of disassembly, the spliceosome forms a P complex under the action of exon ligation factors, such as CDC40 (cell division cycle 40), SLU7, CACTIN (coactivator of transcription 2), SDE2, FAM32A, and NKAP127 (nuclear karyopherin 127) [ 22 – 24 ]. At its active site, the 3′ SS, BP, and 5′ SS are aligned to allow the 3′ SS to initiate a nucleophilic attack on the exon 3′ OH of 5′ SS. After exon ligation, mRNA is released, and the P complex is transformed into the ILS. Under the influence of DHX15, TFP11, and C2ORF3, snRNP and NTC in the ILS are recycled and re-enter the splicing cycle.

Cryoelectron microscopy, cryoelectron tomography, and other methodologies have contributed to the rapid advancement of structural biology by gradually elucidating pre-mRNA splicing mechanisms [ 25 ]. Spliceosomes are composed of various proteins with an active core that functions as an RNA-based metalloenzyme [ 26 ]. This enzyme catalyzes 2 trans-esterification events: branching and exon ligation. The catalytic core remains constant during the transition from B to the B act complex until mRNA release occurs. Furthermore, spliceosomes preserve both structural and mechanistic characteristics of yeast and humans. RNA helicases are crucial for facilitating core formation, regulating splice sites, and restructuring protein–RNA or RNA–RNA networks. The splicing process is further regulated by cis-acting elements and trans-acting factors, with both tissue specificity and disease specificity. This regulation results in 7 modes of AS of pre-mRNA: exon skipping (ES), intron retention (IR), alternative 5′ SS (A5SS), mutually exclusive exons (MXE), alternative 3′ SS (A3SS), alternative promoters (AP), and alternative polyadenylation (Fig. 2 B). AS enhances gene expression complexity to adapt to the dynamic microenvironment, regulates cell growth and differentiation to determine cell fate, generates protein diversity without altering genomic integrity, improves coding efficiency by conserving genomic resources, and contributes to evolutionary innovation [ 27 ].

AS process regulated by cis-acting elements

Theoretically, the cis-regulatory elements of pre-mRNA can be located anywhere within the transcript. However, most cis-acting elements are concentrated within 100 to 300 nucleotides of splice sites and exhibit distinct sequence characteristics [ 28 ]. Splicing enhancers are unique RNA motifs that influence AS by binding to splicing-related proteins. They can be classified into exonic splicing enhancers (ESEs) and intronic splicing enhancers (ISEs). The ESE motif is a purine-rich sequence of 3 to 8 nucleotides that is usually located near the end of the exon of the coding sequence. The ISE motif is typically 6 to 10 nucleotides rich in G and T, which can be categorized into 6 groups according to the features of the motif sequence; ISEs are often far away from exons (>500 nt) [ 29 , 30 ]. Previous studies have proposed that the SR family and SR-related splicing factors gathered on the ESE allow for both constitutive and AS, or the ability to select splice sites, by building complex networks of contacts between snRNPs [ 31 , 32 ]. For instance, an ESE mutation occurring in exon 18 of BRCA1 (breast cancer 1 gene) inhibits SF2 by identifying the splice site, thereby preventing the translation of mRNA in Clinical Genomics [ 33 ]. Mutation of ESE in adenomatous colonic polyposis leads to the loss of the exon splicing boundary of adenomatous polyposis coli, a suppressor gene resulting in the removal of exon 14 to produce a truncated protein that accelerates the development of familial adenomatous polyposis [ 34 ]. Unlike ESE, ISE functions as a binding site for RBPs, including hnRNPH1 (heterogeneous nuclear ribonucleoprotein H1), hnRNPF (heterogeneous nuclear ribonucleoprotein F), Fox-1 (feminizing locus on X homolog protein 1), and Fox-2, rather than for SR proteins [ 30 , 35 ]. In certain individuals, a 40-bp ISE was identified within the Alu repeat element of intron 20 of the ataxia-telangiectasia mutated (ATM) kinase gene, where it promotes the preservation of ATM exons through hnRNPA1 (heterogeneous nuclear ribonucleoprotein A1)/DAZAP1 (DAZ associated protein 1) binding [ 36 , 37 ]. A 19-bp ISE deficiency in intron 3 of proteolipid protein 1 (PLP1) in mice led to axonal damage and broad astrocyte proliferation, which is potentially associated with the hnRNPH/hnRNPF-mediated skipping of PLP1 exon 3.

Similarly, splicing silencers can be divided into exonic splicing silencer (ESS) and intronic splicing silencer (ISS). The ESS motif is rich in G and U and low in C, with a length between 4 and 10 nucleotides. The ISS motif exhibits a high abundance of A and U, and can be divided into 4 distinct groups varying in length from 3 to 5 nucleotides [ 30 ]. The general view is that the ESS and ISS limit the formation of snRNP assemblies around the spliced site by interacting with the hnRNP family [ 38 ]. An illustrative example is the ability of ESS to facilitate exon skipping in CD45 exons 4, 5, and 6. After the activation of T cells, hnRNPL (heterogeneous nuclear ribonucleoprotein L) interacts with the PSF (PTB-associated splicing factor) protein on the ESS, leading to the skipping of exon 4, 5, or 6, which creates a smaller CD45 variant that homodimerizes easily, accompanied by weak phosphatase activity [ 39 , 40 ]. Naturally occurring alterations in the ESS of CD45 contribute to increased susceptibility to autoimmune disorders and viral infections [ 41 ]. Furthermore, the presence of only 1 or 2 repeats of the ISS motif, especially GCACC or UGCACC, could effectively prevent exon splicing through YB1 binding in intron analysis for several genes [ 29 ]. Exon 7 skipping of survival motor neuron 2 (SMN2) pre-mRNA, a key gene in patients with spinal muscular atrophy (SMA), led to abnormal growth of motor neurons, whereas targeting 2 ISSs (UAGGGU and UAGGUC) in the proximal intron region of exon 7 flanks promoted the inclusion of exon 7 of SMN2 by inhibiting hnRNPA1/A2-mediated splicing, highlighting the therapeutic potential of ISS in diseases.

However, cis-acting elements in AS are worth reviewing. The mechanism of binding of the splicing factor to the RNA motif was characterized by adaptability, and the regulation of the RNA-protein network exhibited considerable plasticity in various clinical situations. In addition, a study of the RNA recognition motif of the SR and hnRNP protein families indicated that they exhibited antagonistic or cooperative influences on the cis-regulatory elements in AS and were susceptible to the influence of other splicing factors [ 5 , 31 , 42 ]. Although the characterization of cis elements has advanced, the methods for identifying cis-acting elements need to be improved. The predominant technique for identifying short splicing sequences is the cell-based fluorescence-activated screening approach; however, no studies have used histological identification methods [ 29 , 30 , 43 ].

AS process regulated by trans-acting factors

Splicing is accompanied by profound conformational rearrangements and is regulated by cis-acting elements, trans-acting factors, and splicing factors. This intricate regulation hints at the existence of a “splicing code”. RBPs participate directly or indirectly in AS regulation, and their splicing-regulatory activity is primarily determined by the composition and specificity of the sequences they recognize. Among these, serine/arginine-rich (SR) proteins and hnRNPs are the best characterized.

Humans possess at least 12 distinct genes encoding SR proteins, distributed across various chromosomes [ 44 , 45 ]. SR proteins are characterized by an arginine/serine-rich (RS) domain and at least one RNA recognition motif (RRM) that ensures RNA interactions. Upon phosphorylation, SR proteins bind to ESE through their RS domains and promote exon inclusion by recruiting U1 snRNP and U2 snRNP. The greater the density of serine–arginine repeats within the ESE, the higher the relative activity of the SR proteins [ 46 ]. The proximity to introns further enhances their ability to promote exon inclusion. A subset of SR proteins can also bind to introns within pre-mRNAs to interfere with spliceosome assembly [ 47 ]. For instance, following adenoviral infection, serine/arginine-rich splicing factor 1 (SFRS1) binds intronic silencer elements in adenoviral pre-mRNA, preventing U2 snRNP binding to the 3′ SS [ 48 ].

hnRNPs form another classical protein family involved in AS regulation [ 49 ]. At least 19 genes encoding hnRNPs have been identified, each of which is capable of producing multiple variants [ 50 ]. Similar to SR proteins, hnRNP family members contain at least one RRM domain alongside auxiliary regions, such as RGG or KH motifs, to facilitate interactions with pre-mRNA and proteins. hnRNPs are generally considered negative regulators of AS. For example, hnRNP L suppresses AS by binding CA-rich sequences and interfering with U1 snRNP binding to the 5′ SS of the SLC2A2 pre-mRNA [ 51 ]. Similarly, hnRNP A1 inhibits the inclusion of variable exon N1 in c-Src and competes with SFRS1 for binding [ 52 ]. hnRNP A1 binds to the UAG sequence through its RRM1 and RRM2 domains, facilitating the skipping of exon 13 in lysyl oxidase like 2 (LOXL2), ultimately promoting the expression of a shorter LOXL2 variant [ 53 ].

Other RBPs involved in AS regulation were predominantly enriched in the RRM, Zinc-finger, and KH domains. Notably, 86% of RRMs bind to single-stranded mRNA [ 54 ]. Among 230 human RBPs containing RRMs, 29% were functionally annotated to be involved in AS [ 54 ]. Zinc-finger domains not only bind RNA to participate in AS but also interact with DNA and other proteins [ 55 , 56 ]. Furthermore, 57% of the zinc finger proteins exhibit AS dysregulation when knocked out in embryonic stem and neuroblastoma cells [ 57 ]. KH domains, first identified in hnRNPK (heterogeneous nuclear ribonucleoprotein K) homology, are composed of approximately 70 amino acids and exhibit high specificity and affinity to C-rich polypyrimidine motifs in introns near 5′ SS, facilitating exon splicing [ 58 ]. In addition to the hnRNP and SR families, other proteins with RRM domains that are involved in AS include RBFOX, PTBP, MBNL, and CELF. Proteins containing KH domains include QKI, KHDRBS1, Nova, FUS, SF1, and PCBP. Those with zinc finger domains include CELF, ZFP36, RBM, PRDM, ZNF277, ZNF768, ZFP36L1, ZFP36L2, WT1, ZMAT2, and TRIM28 (Table 2 ).

Role of AS in Bone Biology

Osteoblasts

We now turn to the roles of AS in the regulation of bone biology. Osteogenesis and chondrogenesis are critical physiological processes in bone biology in which AS plays a substantial role (Fig. 3 ). During the pubertal phase of skeletal development, osteoblasts become highly active, depositing new bone matrix along the periosteal surface and secreting osteocalcin and lipocalin to modulate systemic metabolism [ 59 ]. The prevailing view is that osteoblasts, key participants in osteogenesis, originate from bone marrow mesenchymal stem cells (BMSCs) [ 60 ]. The process by which stem cells differentiate into osteoblasts is called osteogenic differentiation, and numerous AS events contribute to this process. For instance, one study has identified 3 AS variants of the heat shock protein TID (encoded by DNAJA3): TID-L, TID-I, and TID-S. Among these, TID-L and TID-I variants were increased during the differentiation of BMSCs into osteoblasts, whereas TID-S was minimally expressed, suggesting its involvement in osteogenic differentiation [ 61 ]. Filamin B, a newly discovered RBP, regulates the AS of transcription factors Tpx2 and Evc, which are associated with early osteoblast development and thereby influence osteoblast differentiation [ 62 ]. In addition, long AS variants of PNO40 (partner of NOL1 homolog), a core protein of the 60S ribosome, potentially promote osteogenic differentiation and inhibit senescence by down-regulating runt-related transcription factor 2 (Runx2) and up-regulating reactive oxygen species (ROS) [ 63 ]. Another study has demonstrated that variant I of transcription factor IID subunit TATA-box binding protein associated factor 14 (TAF14) is abundantly expressed in human mesenchymal stem cells (hMSCs), whereas variant II is relatively rare. When the transcription activation factor homology (TAFH) domain encoded by TAF14 exons 5 and 7 is absent, hMSCs are driven toward chondrogenic differentiation, thereby blocking other differentiation pathways [ 64 ]. Compared with demethylases, the N6-methyladenosine (m6A) methyltransferase Mettl3 is up-regulated in BMSCs, and it regulates the expression of Vegfa and its AS variants, thereby promoting osteogenic differentiation [ 65 ].

Noncoding RNAs also modulate AS programs to regulate osteogenic differentiation. Specifically, a long non-coding RNA called lnc-protein phosphatase 2 regulatory subunit A beta (PPP2R1B) interacts with heterogeneous nuclear ribonucleoprotein L-like (HNRNPLL), leading to the down-regulation of variant 201 and the up-regulation of variant 203 of PPP2R1B. This results in the dephosphorylation and stabilization of β-catenin, facilitating its nuclear translocation to activate the Wnt (Wingless/Integrated)/β-catenin pathway and promote the expression of Runx2 and OSX (Osterix), ultimately enhancing osteogenesis [ 66 ]. Similarly, the co-transcription factor, CBFB (core-binding factor subunit beta), which interacts with Runx2, undergoes extensive AS in vivo, resulting in the predominance of the CBFB2 variant over the CBFB1 variant. This restricts osteogenic differentiation to appropriate levels, thereby benefiting skeletal development [ 67 ]. Variants of WNK lysine-deficient protein kinase 1 (Wnk1), collagen type IX alpha 1 chain (Col9a1), and cyclin-dependent kinase inhibitor 2a (CdkN2a) dominate different stages of osteoblast development, suggesting their potential role as key players in osteogenic differentiation [ 68 ]. Moreover, AS programs produce novel miRNAs that regulate osteogenic differentiation. For example, exon 4 of amelogenin (encoded by the AMELX gene) is spliced to form miR-exon4, which is proposed to promote osteoblast differentiation by regulating Runx2 expression [ 69 ].

Bone and cartilage formation cannot be separated from external mechanical stimulation, and current research indicates that AS is involved in the mechanical stimulation-induced formation of bone and cartilage. A previous study has demonstrated that mechanical stimulation activates the AS factor serine/arginine-rich splicing factor 1 (SRSF1) in osteoblasts, leading to the up-regulation of an insulin-like growth factor 1 (IGF1) splice variant named mechano growth factor (MGF). The MGF variant arises from the specific retention of exon 5 in IGF1 pre-mRNA, which alters the reading frame and causes premature termination of translation, resulting in a truncated protein with mechanosensitive capacity [ 70 , 71 ]. The MGF variant ultimately inhibits osteogenic differentiation via the extracellular signal-regulated kinase (ERK) signaling pathway while promoting osteoblast proliferation and migration [ 72 ]. Additionally, the AS factor SRSF1 responds to mechanical stimulation by co-regulating the AS of cyclin D1 with the SWI/SNF complex subunit SMARCE1 (SWI/SNF-related matrix-associated actin-dependent regulator of chromatin subfamily E member 1). This process reduces cyclin D1a levels and increases cyclin D1b levels [ 73 ].

During the osteogenic differentiation of BMSCs, the process can be divided into 4 stages: lineage commitment of osteoprogenitor cells, proliferation of pre-osteoblasts, matrix maturation, and matrix mineralization [ 74 , 75 ]. In the lineage commitment stage, filamin B acts as a key upstream RBP, directly influencing the normal development of osteoprogenitor cells by regulating the AS of osteogenesis-related genes Tpx2 and Evc [ 62 ]. At the same time, high expression of the TAF4v1 variant together with lnc-PPP2R1B/HNRNPLL synergistically enhances RUNX2 signaling, thereby promoting osteogenic differentiation [ 64 , 66 ]. The 2 variants of Cbfb (Cbfb1 and Cbfb2) serve as co-transcription factors of RUNX2 and exert complementary roles in regulating the transcription of downstream osteogenesis-related genes. Among them, Cbfb2 is indispensable for skeletal development, whereas Cbfb1 is more efficient in enhancing RUNX2 activity [ 67 ]. In the proliferation stage, SRSF1 mediates the AS of IGF1 to generate the MGF variant, which activates the ERK signaling pathway to promote pre-osteoblast proliferation [ 72 ]. Meanwhile, SRSF1 also regulates the AS pattern of Cyclin D1, thereby maintaining the dynamic balance of cells under a highly proliferative state [ 73 ]. During the matrix maturation stage, the up-regulation of the TID-L and TID-I variants enhances the osteoblastic phenotype, characterized by increased alkaline phosphatase activity and calcium deposition [ 61 ]. At the same time, the pNO40 variant, which lacks exon 4, promotes the transition of osteoblasts from a “differentiation-proliferative state” to a “matrix maturation state” by down-regulating RUNX2 signaling [ 63 ]. In vivo, novel AS patterns of genes such as Col9a1, Wnk1, and CdkN2a can also be observed in mice, suggesting potential involvement in osteogenic maturation, although their precise functions remain to be elucidated [ 68 ]. In the matrix mineralization stage, splicing of the amelogenin (AMELX) gene produces miR-exon4, which maintains RUNX2 at an appropriate level to support the sustained progression of matrix mineralization [ 69 ]. The 3 AS transcript variants of the VWC2L gene (Vwc2l-1, Vwc2l-2, and Vwc2l-3) promote matrix mineralization in MC3T3-E1 osteoblasts, accompanied by increased OSX expression, with Vwc2l-1 showing the strongest effect [ 76 , 77 ]. Moreover, they exhibit differential regulation of markers such as Col1a2 and Bsp, suggesting that they may jointly promote osteogenic differentiation and mineralization through activation of transforming growth factor beta (TGF-β) superfamily signaling [ 77 ]. In summary, the 4 stages of osteogenic differentiation are characterized by distinct AS features, with various AS factors and events exerting stage-specific regulatory functions during lineage commitment, cell proliferation, matrix maturation, and matrix mineralization (Fig. 4 ).

Overall, we observed the broad involvement of AS during osteogenic differentiation, spanning multiple dimensions including RUNX2 transcription factor, noncoding RNAs, and mechanical stimulation. Although existing studies have identified and validated many AS variants related to osteogenic differentiation, such as TID, TAF4, and CBFB2, most remain at the single-gene level, limiting the possibility of constructing a complete AS regulatory network. In the future, under the premise of sufficient sequencing depth and robust AS analysis, exploring the changes in AS regulatory networks during the differentiation of BMSCs into osteoblasts will be an important direction. Moreover, if antisense oligonucleotides (ASOs) or CRISPR tools are applied to in vivo models to target specific AS variants, this will provide the most direct functional validation and causal evidence to clarify their specific roles in osteogenic differentiation.

Osteoclasts

Osteoclasts are multinucleated cells that are primarily responsible for bone resorption and play a crucial role in maintaining bone homeostasis (Fig. 3 ). Derived from the monocyte–macrophage lineage, osteoclasts undergo differentiation and maturation into functional cells, a process that is tightly regulated by multiple signaling pathways. Accumulating evidence has demonstrated that AS plays a critical role in osteoclast biology. To date, numerous AS events have been identified to impact osteoclast differentiation. One study has revealed that cytoplasmic polyadenylation element binding protein 4 (CPEB4), an RBP, colocalizes with SRSF5 and SRSF6 through liquid–liquid phase separation, stabilizing the normal AS variant inhibitor of DNA binding 2 (ID2)-201 [ 78 ]. In the absence of CPEB4, abnormal variants ID2-202 and ID2-203 are up-regulated, impairing the inhibitory effect of ID2 on osteoclast differentiation and subsequently affecting osteoclastogenesis [ 78 ]. Furthermore, a study on multiple myeloma has reported that the exosome-derived AS factor SFRS8 regulates the AS of calcyclin binding protein (CACYBP) in the bone marrow microenvironment. This leads to an increase in CACYBP variant 2 (lacking exon 2) and a concomitant decrease in variant 1 (lacking exon 1), ultimately inducing osteoclast differentiation via the Wnt/β-catenin pathway [ 79 ]. The RANKL (receptor activator of NF-kappaB ligand)–RANK (receptor activator of NF-kappaB) signaling pathway is essential for osteoclast survival. Interestingly, a truncated AS variant of RANK, termed vRANK, antagonizes the anti-apoptotic effects of RANKL, thereby inhibiting osteoclastogenesis [ 80 ]. Moreover, studies have confirmed that zBTB (zinc finger and BTB domain-containing protein) family proteins, including leukemia/lymphoma-related factor (LRF) and osteoclast zinc finger protein (OCZF), jointly regulate the AS of Bcl, which influences osteoclast survival. In the presence of LRF/OCZF, the degradation of RBP Sam68 (Src-associated in mitosis of 68 kDa) is promoted, resulting in an up-regulation of the Bcl extra large (Bcl-xl; anti-apoptotic) AS variant and a loss of the Bcl extra short (Bcl-xs; pro-apoptotic) variant, thereby maintaining osteoclast survival [ 81 ].

Additionally, osteoclast activity is modulated by various hormones and matrix molecules, such as calcitonin, parathyroid hormone, galectins, and protein tyrosine phosphatase-osteoclastic (PTP-oc) [ 82 ]. Several studies have demonstrated the involvement of AS in these regulatory processes. For instance, 2 newly identified human calcitonin receptor (CTR) variants, H-CTR5 and H-CTR6, are expressed in giant cell tumors, but their functional role remains unclear [ 83 ]. Another research group identified a mouse osteoclast-specific AS variant of the CTR, P3-CTR, which is thought to play an essential role in modulating osteoclast activity [ 84 ]. Additionally, 3 alternatively spliced products of parathyroid hormone-related protein (PTHrP) were identified: PTHrP-1-139, PTHrP-1-141, and PTHrP-1-173. Among them, PTHrP-1-139 exhibited the most pronounced osteoclastogenic effects. In a mouse model of breast cancer bone metastasis, mice carrying PTHrP-1-139 exhibit extensive osteolytic lesions and high serum calcium levels [ 85 ]. In addition to generating distinct variants via AS, the intronic promoters within glomerular epithelial protein 1 in osteoclasts can drive tissue-specific transcription programs, producing a truncated variant, PTP-oc. This variant regulates tyrosine kinase signaling through dephosphorylation, inhibits RANKL signaling, and contributes to osteoclast survival [ 86 ]. The galectin family members (LGALS) are closely related to angiogenesis and consist of 11 members, all of which are expressed at the protein level in humans. LGALS1 inhibits osteoclast activity [ 87 , 88 ]. However, recent studies have identified 2 AS-derived galectin-8 variants in osteoclasts: a long variant (LGALS8-L) and a short variant (LGALS8-S). LGALS8-L, owing to its slower binding kinetics to bone matrix proteoglycans and an additional flexible loop, is proposed to obstruct certain protein interactions, thereby suppressing bone resorption. By contrast, LGALS8-S interacts with more proteins, particularly those associated with secretory vesicles (e.g., lysosome-associated membrane protein 1/2 and chloride channel 7), which are critical for the integrity of the sealing zone and bone resorption activity [ 89 ]. AS functions as a switch that determines the ratio of LGALS8-L to LGALS8-S, and directly influences the dynamic balance between osteoclast resorption and migration. Furthermore, a medium-length variant of LGALS8, LGALS8-M, has been identified in endothelial cells. These 3 galectin-8 variants mediate vascular and lymphatic vessel formation by interacting with membrane proteins, such as cluster of differentiation 166 (CD166), integrins, and CD44, suggesting that the AS of LGALS8 in osteoclasts has diverse biological functions [ 90 ].

Overall, we found that AS is involved in osteoclast differentiation, survival, and functional regulation, ranging from signaling pathway activity such as RANKL and Wnt/β-catenin, to specific RBP regulation such as CPEB4, SFRS8, and CTR, as well as the shaping of the osteoclast microenvironment such as LGALS8. Mechanistically, AS variants including vRANK, Bcl-xl, and LGALS8-L have been confirmed to exert functional counter-regulatory effects, inhibiting osteoclast activity and acting as “functional switches” for maintaining bone homeostasis. In vitro, through high-throughput sequencing, we propose to further investigate the AS network dynamics during osteoclast maturation, especially under different external stimuli such as mechanical loading and drug intervention. In vivo, by leveraging single-cell sequencing and spatial transcriptomics, monitoring the spatiotemporal RBP changes underlying the formation and maintenance of the osteoclast sealing zone will help to reveal the core functions of AS in bone resorption.

Chondrocytes

Chondrocytes are the only cell type present in the cartilage tissue and are primarily responsible for maintaining the structural integrity and functional stability of the cartilage [ 91 ]. MSC-derived chondrocytes undergo proliferation, differentiation, and maturation during cartilage development, ultimately reaching a terminally differentiated state embedded within the extracellular matrix (ECM) (Fig. 3 ). This process is strictly regulated by various molecular pathways and characterized by extensive AS events [ 92 ]. For instance, one study has demonstrated that TAF4 generates multiple AS variants, including TAF4-v2, TAF4-v4, and TAF4-v5, by altering exons 6 and 7 of the hTAF4-TAFH domain [ 64 ]. These variants collectively promote mesenchymal stem cell proliferation and preferentially direct differentiation toward the chondrogenic lineage [ 64 ]. Similarly, the AS variant of Wnt-induced secreted protein 1 (WISP1), WISP1v (lacking exon 3), is up-regulated during the terminal differentiation of chondrocytes, suggesting its critical role in the transition to endochondral ossification [ 93 ]. Another splice variant, WISP1vx, lacking exons 3 and 4, is thought to be involved in the oncogenic transformation of chondrocytes [ 93 ]. Recent studies have revealed that activation of A2AR regulates AS of p53 in chondrocytes, leading to the degradation of full-length p53 and increasing the expression of the anti-aging variant Δ133p53α [ 94 ]. Δ133p53α is a p53 AS variant lacking the N-terminal transactivation domain, which resists cellular senescence by antagonizing full-length p53 [ 94 – 96 ]. Δ133p53α mitigates p53-related senescence signaling by down-regulating p21 and p16 or altering their subcellular localization, thereby improving chondrocyte function, delaying cartilage degeneration, and maintaining chondrocyte homeostasis [ 94 ]. Similarly, insulin-like growth factor 1 (IGF1) precursors undergo AS to produce the proIGF-1A, proIGF-1B, and proIGF-1C variants [ 97 ]. The EA peptide of proIGF-1A (encoded by exons 4 and 5) enhances heparin binding via N-glycosylation, prolonging its extracellular activity and promoting the secretion of IGF-1, which stimulates the synthesis of matrix proteins, such as proteoglycans and type II collagen, facilitating cartilage growth and repair [ 97 ]. By contrast, proIGF-1B inhibits IGF-1 secretion without directly enhancing cartilage repair, whereas proIGF-1C, also known as MGF, aids cartilage regeneration by promoting stem cell migration, proliferation, and differentiation through IGF-1 receptor binding [ 98 ]. Interestingly, transforming growth factor beta 1 (TGF-β1) signaling mediates spatially distinct AS of superficial zone protein (SZP) in chondrocytes from load-bearing regions of the medial femoral condyle. SZP produces 3 variants: SZP-A (including exons 4 and 5), SZP-B (including exon 4 but lacking exon 5), and SZP-C (lacking exons 4 and 5). The SZP-A and SZP-B variants enhance ECM binding, whereas SZP-C exhibits weaker matrix binding and exists freely in the synovial fluid [ 99 ]. AS also plays a critical role in the pathophysiology of the cartilage. A recent study has shown that prednisolone induces AS in glycosyltransferase-like domain-containing protein 1 (Gtdc1) exon 4, producing circ-Gtdc1. Circ-Gtdc1 ubiquitinates and degrades the AS factor SRSF1, thereby inhibiting chondrocyte proliferation and ECM synthesis, ultimately leading to cartilage dysplasia [ 100 ]. Supplementation with exogenous circ-Gtdc1 improves joint cartilage quality, providing a potential therapeutic target for early intervention in fetal-origin osteoarthritis (OA) [ 100 ].

Chondrocytes and ECM maintain a “symbiotic relationship”, jointly preserving cartilage structure and function. ECM factors markedly influence chondrocytes. Collagen type II alpha 1 chain (COL2A1), a crucial ECM component of cartilage, has 4 AS variants: IIA (COL2A1IIA), IIB (COL2A1IIB), IIC (COL2A1IIC), and IID (COL2A1IID) [ 101 , 102 ]. The abundance of these splice variants changed during chondrocyte differentiation, suggesting a regulatory role at different stages of differentiation [ 103 ]. A study demonstrated that disruption of the 5′ SS in COL2A1 exon 2 suppresses IIC expression, increases IIA and IID mRNA levels, and decreases IIB mRNA levels, resulting in enhanced expression of embryonic type II collagen in the ECM. This indicates a functional regulatory role for COL2A1 exon 2 [ 104 ]. Another essential ECM component, fibronectin 1 (FN1), has garnered attention due to its role in chondrocyte maturation via AS regulation. Specifically, TGF-β1 modulates the AS of SRp40 (SRSF5), generating 2 variants, SRp40 and SRp40LF. During the proliferative phase, SRp40 promotes FN1 variants containing extra domain A (FN1-EDA), thereby accelerating chondrocyte maturation [ 105 ]. Conversely, SRp40LF inhibits the FN1-EDA variant, suggesting that the balance between these variants is a critical AS regulatory node for chondrocyte maturation [ 105 ]. Recent studies have shown that the RNA helicase DDX5 directly participates in the AS process by regulating FN1 AS to suppress the FN1-EDA variant, thereby reducing cartilage fibrosis [ 106 ]. DDX5 also regulates procollagen-lysine, 2-oxoglutarate 5-dioxygenase 2 (PLOD2) AS by inhibiting the production of the full-length PLOD2 variant (PLOD2-WT) while maintaining PLOD2-SE14 expression, reducing fibrosis-associated collagen crosslinking and preventing cartilage fibrosis [ 106 ]. Furthermore, another FN1 AS variant, FN1-208, was differentially expressed between osteoarthritic and normal cartilage samples. Further studies have revealed that the up-regulation of FN1-208 negatively affects ECM deposition in chondrocytes, warranting further investigation into its functional role [ 107 ].

Overall, we observed that specific AS events in cartilage components, such as WISP1, IGF1, COL2A1, and FN1, are associated with chondrocyte development and functional phenotypes. The SRSF5-derived variants SRp40 and SRp40LF display temporal expression and functional antagonism, respectively dominating the proliferative and maturation phases of chondrocytes, and serve as a key AS checkpoint regulating cartilage development. However, most studies remain at the single-gene level and lack a systematic positioning of AS events within regulatory networks. The interactions of AS with epigenetic modifications, post-translational modifications, and noncoding RNAs in chondrocytes have not yet been systematically explored. Moreover, under pathological conditions such as OA or rheumatoid arthritis, systematically dissecting AS patterns in chondrocytes and validating functional targets will represent a promising research direction.

Bone matrix

The bone matrix comprises various components, including matrix proteins, collagen fibers, noncollagenous proteins, and proteoglycans [ 108 , 109 ]. The physical properties of matrix proteins are altered by AS (Fig. 3 ), proteolytic processing, and the formation of homo- and hetero-oligomers [ 110 ]. Covalently crosslinked type II collagen is the major collagen component of the cartilage ECM, providing tensile strength to the cartilage. The protein encoded by COL2A1 forms homotrimeric procollagen, which undergoes proteolytic processing to assemble crosslinked fibers in the ECM of cartilage tissue [ 111 ]. Exon 2 of COL2A1 encodes a conserved domain of 69 amino acids in the NH 2 -terminal region of type IIA procollagen [ 112 ]. Cartilage progenitor cells predominantly produce variants containing exon 2 (types IIA and IID), whereas differentiated chondrocytes mainly generate variants lacking exon 2 (type IIB) [ 104 , 112 ]. Types IIA and IID, which are embryonic variants, are coexpressed during chondrogenesis and dominate the early stages of embryonic development before being down-regulated [ 113 ]. Type IIB is the only variant expressed and synthesized by fully differentiated chondrocytes and is more abundant in the epiphyseal cartilage of late embryonic and postnatal mice [ 113 ]. Homozygous mice engineered to retain exon 2 of type IIA showed persistent expression of the type IIA protein in the trabecular bone in the subepiphyseal and metaphyseal regions, without apparent effects on embryonic development [ 114 ]. Proliferative and hypertrophic chondrocytes, which typically synthesize type IIB procollagen, develop normally even when producing only type IIA procollagen [ 114 ].

Exon 2 of COL2A1 also contains an additional A5SS, resulting in a truncated mRNA variant (type IIC). In vivo suppression of type IIC AS using site-directed mutagenesis revealed no skeletal or gross morphological differences but led to elevated levels of variant type IIA and IID procollagen and reduced levels of type IIB variants in differentiated chondrocytes, indicating that type IIC AS is indispensable for ECM maturation in cartilage [ 104 ]. Antibodies targeting peptide sequences spanning exons 1 to 3 of mouse COL2A1, such as IIBN, which specifically recognizes type IIB proteins, have been developed. Using IIBN, high type IIB procollagen levels were detected in the hypertrophic chondrocytes of wild-type neonatal mice, whereas no signals were observed in the resting or proliferative zones, suggesting that chondrocytes efficiently transported type IIB procollagen from the endoplasmic reticulum to the bone matrix [ 115 ]. Another antibody specific to type IIB, pNIIB52, is effective for detecting type IIB in human, canine, and bovine cartilage, but fails to detect type IIB in mouse or rat cartilage owing to differences in the exon 1/3 sequences targeted by the antibody [ 113 ]. Owing to limitations in conventional AS detection methods, where the amplification efficiencies of polymerase chain reaction (PCR) products of different sizes tend to vary, researchers have developed alternative transcript-qPCR (AT-qPCR) based on a standard plasmid containing multiple amplicons. Using AT-qPCR, we observed that insulin-induced chondrogenic differentiation in ATDC5 cells primarily expressed types IIA and IID [ 116 ]. Compared to conventional PCR, AT-qPCR yields a lower IIB/IIA product ratio, which is closer to the true results [ 116 ]. Improving AS detection methods represents a key direction for future research.

The matrilin family, a critical component of the cartilage ECM, mediates interactions between collagen fibers and other matrix components [ 117 ]. An AS event between the second vWFA domain and the coiled-coil domain of matrilin-2 (encoded by matrilin-2 [MATN2]) gives rise to 2 variants: a short variant (MATN2S) and a long variant (MATN2L) [ 118 ]. The long variant of matrilin-2 can form multimers, whereas the short variant can only form oligomers, thereby affecting its post-translational modifications [ 118 ]. Alterations in the matrilin family are involved in abnormalities in the molecular network of the bone matrix.

Overall, the different AS variants of type II collagen represented by COL2A1 (IIA, IIB, IIC, and IID) are not expressed randomly, but rather undergo highly programmed spatiotemporal transitions, which constitute one of the reasons ensuring proper development and maturation of the bone matrix. The study of AS function should not be limited to its regulation of gene expression levels, but should also focus on a new dimension—how AS reshapes protein domain composition to regulate protein conformation, oligomeric state, and functional activity. MATN2, which generates variants with distinct polymerization capacities through AS, exemplifies this regulatory layer. By using cryo-electron microscopy to observe structural differences among AS variants at the protein conformation level, a critical frontier direction will emerge. Traditional PCR methods are limited by amplification efficiency bias, making it difficult to accurately quantify the proportion of AS variants, while the emerging AT-qPCR technique, though improved, still cannot achieve in situ detection of AS variants at single-cell resolution. As a result, we are not yet able to precisely define the distribution and functions of AS variants across the distinct zones of cartilage differentiation—resting, proliferative, hypertrophic, calcified, and ossification—or directly connect AS events with specific cellular states. The development of new AS detection technologies with single-cell and spatial resolution will be essential to advance mechanistic insights in this field.

Bone remodeling

Although we have discussed in detail the AS features of osteogenic cells and the bone matrix, the role of AS in dynamic bone remodeling has not yet received sufficient attention. Bone remodeling in humans is considered a cell optimization process essential for survival and is influenced by multiple factors, including osteocytes, mechanical stress, the RANKL/osteoprotegerin (OPG) ratio, PTH, estrogen, TGF-β1, and vitamin D [ 119 , 120 ]. RANKL, encoded by TNFSF11 , is primarily secreted by osteoblasts, bone marrow stromal cells, T cells, and synovial cells. It binds to RANK on the surface of osteoclast precursors and mature osteoclasts, activating downstream signals such as NF-κB and thereby promoting osteoclast differentiation, maturation, and bone resorption [ 121 ]. As a regulatory mechanism, osteoblasts under physiological conditions—including bone homeostasis, mechanical stimulation, or estrogen signaling—express OPG (encoded by TNFRSF11B ), a high-affinity soluble decoy receptor that binds RANKL, preventing its interaction with RANK and inhibiting osteoclast activation to maintain remodeling balance [ 122 ]. In addition to osteoblasts, OPG is also secreted by various other tissues, including the heart, liver, and B cells. Its expression is regulated by hormones, inflammatory cytokines, and signaling pathways such as Wnt/β-catenin and Notch. OPG is often insufficiently up-regulated in response to increased RANKL expression, resulting in an elevated RANKL/OPG ratio and enhanced osteoclastogenic activity [ 123 ]. Taken together, the RANKL–RANK–OPG axis serves as the core regulatory system of bone remodeling, integrating endocrine, immune, and mechanical signals through precise modulation of osteoclast activity to maintain bone homeostasis and structural renewal.

RANKL exists in 2 forms: membrane-bound and soluble (sRANKL). Systematic analyses have shown that sRANKL is mainly generated by extracellular proteolytic cleavage of membrane-bound RANKL via metalloproteinases such as MMP-7 or a disintegrin and metalloprotease 17 (ADAM17), representing a post-translational regulatory mechanism [ 124 ]. In bone tissue, the Tnfsf11 gene undergoes AS to produce 3 major transcript variants: RANKL1 (full-length), RANKL2 (partial deletion of exon 2, resulting in a truncated intracellular domain), and RANKL3 (lacking exons 2 to 4, encoding a soluble RANKL and constituting one of the sources of sRANKL) [ 125 , 126 ]. Classical evidence comes from multiple myeloma cell lines, which secrete active sRANKL through AS and post-translational modifications, thereby inducing TRAP + multinucleated cell formation from human or mouse bone marrow mononuclear cells and promoting osteoclastogenesis [ 127 ]. Unlike interleukins, PTH under sustained stimulation uniquely induces Tnfsf11 expression while simultaneously facilitating proteolytic processing of full-length RANKL to release sRANKL into the extracellular space [ 124 ]. The AS variant ΔCAPRI, which lacks one exon within the RasGAP domain, is highly expressed in primary mouse osteoblasts but exhibits low expression in osteoclasts [ 128 ]. ΔCAPRI activates the Ras/matrix metallopeptidase 14 (MMP14) axis and enhances RANKL shedding, thereby increasing sRANKL production and ultimately influencing bone remodeling [ 128 ]. Collectively, RANKL AS variants cooperatively regulate sRANKL production, bridging transcriptional and post-translational mechanisms as a key regulatory module in bone remodeling.

Three AS variants of RANK have been identified: RANK-a (lacking exon 9), RANK-b (lacking exons 8 and 9), and RANK-c (lacking exons 7 to 9) [ 129 ]. RANK-c lacks NF-κB activation capability and, when co-expressed with full-length RANK, suppresses downstream signal transduction, suggesting that it serves as an intrinsic negative regulator of the RANK pathway. However, the specific AS factors involved in generating these variants remain unknown [ 129 ]. In murine RAW264.7 cells, a truncated RANK splice variant named vRANK has been identified. This variant contains a novel exon 2a inserted between exons 1 and 2, which introduces a premature stop codon, resulting in a protein lacking the transmembrane and intracellular domains and retaining only part of the signal peptide [ 129 ]. Compared to full-length RANK, vRANK fails to activate the NFATc1 signaling pathway, strongly inhibits osteoclast formation, and promotes apoptosis, possibly by competing with RANK for RANKL binding and functioning as an endogenous antagonist of RANKL signaling [ 129 ]. In human brain tissue, a 42-bp splice variant named RANK-e5a lacking exon 5 has been identified. This variant alters the conformation of the TNFR cysteine-rich domain, reducing its RANKL binding affinity [ 130 ]. Although RANK-e5a retains the intracellular domain, its shortened extracellular domain weakens the ability to activate downstream NF-κB signaling upon RANKL binding [ 130 ].

In addition, Periostin-4, an AS variant of Periostin lacking exons 17 and 21, has been identified as the predominantly expressed isoform in primary human osteoblasts. Functionally, Periostin-4 acts as an extracellular signal integrator that substantially promotes osteoblast differentiation and mineralization in vivo [ 131 ]. To date, only a single transcript has been identified for OPG, and no AS variants have been reported.

Overall, we observed that the classical RANKL–RANK–OPG signaling axis of bone remodeling is also regulated by AS, highlighting that AS not only functions at the level of post-transcriptional modification but also plays an important role in intercellular communication (Fig. 5 ). Mechanistically, the upstream RBP-mediated AS regulation of the RANKL–RANK–OPG axis and its temporal dynamics remain insufficiently explored and hold important research value. In the future, approaches such as cross-linking and immunoprecipitation sequencing (CLIP)-seq and CRISPR screening will help to construct a more comprehensive RBP network of bone remodeling. For target validation, an important direction will be to generate gene-edited mouse models of AS variants, for example, by specifically deleting vRANK, to observe the net effects on bone remodeling homeostasis and bone loss under pathological conditions. Such animal models will not only reveal the independent functions of AS variants but also provide new experimental tools for studying bone metabolic diseases such as osteoporosis, thereby laying the foundation for therapies targeting AS variants or small molecules.

Distinct AS in bone and the musculoskeletal system

Compared with other tissues, the characteristics of tissue-specific AS in the musculoskeletal system remain an intriguing area of study. Brinegar et al. [ 132 ] conducted RNA sequencing (RNA-seq) across different developmental stages of mouse skeletal muscle—including E18.5 (embryonic day 18.5), PN2 (postnatal day 2), PN14, PN28, and adulthood (22 weeks)—to investigate dynamic changes in AS. Their results showed that AS events were most dynamic between PN2 and PN14. Among the 768 AS events identified between E18.5 and adulthood, 32% occurred within the first 2 postnatal weeks, with ES accounting for 77% of them. These events were significantly enriched in calcium-handling genes [ 132 ]. Notably, 3 AS variants of calcineurin A genes exhibited a transition from fetal-type to adult-type splicing patterns after birth. This transition involved increased inclusion of exons 10a and 13, which altered the conformation of the autoinhibitory domain and calmodulin-binding region, thereby modulating phosphatase activity and calcium sensitivity [ 132 , 133 ]. Collectively, these findings suggest that AS of calcium signaling-related genes constitutes a key component of tissue-specific AS regulatory networks, although more direct experimental evidence is needed.

Bone cells are responsible for regulating more than 99% of the body’s calcium content. Calcium functions as a hormonal and cytokine-like regulator that participates in intercellular communication and cell–matrix interactions [ 134 ]. At least 6 classes of calcium channels have been identified in osteoblasts and osteoclasts, including voltage-gated calcium channels (VGCCs), store-operated calcium entry channels, transient receptor potential channels, ATP-gated P2X channels, inositol 1,4,5-triphosphate receptors, and mechanically activated Piezo channels. These channels involve at least 18 genes that cooperatively regulate intracellular calcium homeostasis [ 135 ]. Among them, CACNA1C , a member of the VGCC family, is highly expressed in both chondrocytes and osteoblasts. Conditional knockout of CACNA1C in mice leads to abnormal endochondral ossification, characterized by shortened limbs and digital defects [ 136 , 137 ]. Exons 8 and 8a of CACNA1C are mutually exclusive, and the G406R mutation in exon 8a enhances OPG secretion, thereby suppressing osteoclast activity and altering skeletal phenotypes [ 136 , 138 ]. These findings indicate that AS of CACNA1C plays a critical role in the fine-tuning of VGCC function in bone-specific contexts. Similarly, P2RX5 , a member of the P2X receptor family, is highly expressed in osteoclasts and plays a role in bone immune regulation [ 139 ]. The AS site at exon 10 of the human P2RX5 gene is influenced by a single-nucleotide polymorphism, where the G allele (as opposed to the T allele) promotes ES, resulting in an AS variant lacking key structural domains [ 139 , 140 ]. The resulting truncated P2RX5 variant encodes a protein that fails to localize to the plasma membrane and cannot form a functional ATP-gated calcium channel, thus losing its capacity to mediate calcium influx [ 141 , 142 ]

Piezo2, a mechanosensitive ion channel, mediates mechanical sensing in bone tissue and plays a role in skeletal development [ 143 ]. To date, at least 17 AS variants of Piezo2 have been identified, primarily involving exons 10, 18, 19, 33, 35, and 40 [ 144 ]. In neurons, Piezo2 variants containing exon 35 (E35) exhibit faster inactivation kinetics and greater mechanical responsiveness [ 144 ]. Notably, E35 is highly expressed in bone-associated sensory neurons (e.g., Ntrk3 + cells), but nearly absent in nociceptors (e.g., Mrgprd + cells), suggesting that bone afferents are proposed to regulate Piezo2 responsiveness and mechanosensitivity through AS [ 144 ]. In addition, a novel AS variant of Piezo1, termed Piezo1.1, has been identified in mouse myogenic precursor C2C12 cells. This variant results from skipping of exon 30 and exhibits enhanced mechanical sensitivity and ion conductivity [ 145 ]. The tissue-specific splicing of calcium channel genes in bone is primarily governed by RBPs with bone cell-restricted expression, while mechanical stimulation likely acts as a critical modulatory factor. Such adaptive AS may also occur, to some extent, in other cells responding to mechanical stimulation or calcium fluxes, such as cardiomyocytes and vascular smooth muscle cells. However, the occurrence and functional relevance of Piezo1/2 AS events in osteoblasts and osteoclasts remain poorly studied, and direct evidence is still lacking.

Taken together, the AS characteristics of calcium channel genes serve as potential molecular signatures of tissue-specific AS regulation. Nevertheless, the detection of AS events largely relies on high-throughput RNA-seq, which demands sufficient sequencing depth to ensure sensitivity and accuracy [ 146 ]. Under physiological conditions, bone tissue is densely mineralized, rich in hydroxyapatite and other inorganic components, and characterized by low cellularity. These properties limit the efficiency of high-quality RNA extraction in vivo, thereby posing challenges for AS profiling and downstream transcriptomic analyses [ 147 ]. With the growing accumulation of bone tissue-related single-cell datasets, systematically resolving the AS patterns of distinct bone cell subpopulations will facilitate deeper secondary analyses of these data and yield novel mechanistic insights [ 148 – 154 ]. The characteristic AS patterns of bone cell subpopulations will represent an important direction for future investigation.

AS Impact on Bone Tumors

Osteosarcoma

Beyond essential roles in maintaining bone homeostasis under physiological conditions, AS also contributes notably to the pathogenesis of bone tumors. OS is the most common primary malignant bone tumor and is highly prone to distant metastasis [ 155 ]. AS and RBPs are involved in various OS progression processes (Fig. 6 ). RNA-seq results from 18 OS and adjacent nontumor samples revealed differences in leptin receptor overlapping transcript (LEPROT) gene variants between tumor and adjacent nontumor tissues [ 156 ]. Five LEPROT variants are overexpressed in OS tissues, whereas 22 variants are overexpressed in adjacent tissues, indicating that LEPROT is relatively down-regulated in OS tissues [ 156 ]. This highlights the importance of comprehensively considering all the variants generated by AS and their overall trends when evaluating whether a gene is up-regulated or down-regulated during OS. Overexpression of U1 small nuclear ribonucleoprotein 70K (SNRNP70), a component of the spliceosome, promotes OS migration, whereas its knockdown inhibits metastasis in vivo [ 157 ]. SNRNP70 facilitates exon 10 skipping in CD55, producing a shorter variant that enhances OS cell proliferation, migration, and metastasis [ 157 ]. In OS cell lines with low metastatic potential, AS-related proteins such as apoptosis chromatin condensation inducer 1, poly(A)-binding protein, heterogeneous nuclear ribonucleoprotein A2/B1, block of proliferation 1, and gem-associated protein 5 (GEMIN5) are up-regulated [ 158 ]. GEMIN5, a critical component of the spliceosome snRNP, affects OS cell migration when either overexpressed or knocked down [ 158 ].

SRSF1, a trans-acting factor that regulates AS, also influences tumor behavior in OS. Markedly up-regulation of SRSF1 expression was observed in OS samples and cell lines. SRSF1 overexpression enhances migration and invasion, while reducing apoptosis, whereas SRSF1 knockdown produces the opposite effects [ 159 ]. SRSF1 primarily triggers SE and MXE events, and is closely associated with the Wnt, NOTCH, and PI3K (phosphoinositide 3-kinase)–protein kinase B (AKT) pathways. SRSF1 knockdown altered the variant ratios of dermokine and SCAT1 (small cysteine-rich protein 1) [ 159 , 160 ]. Interestingly, the AS factors SRSF3 and SRSF1 were coexpressed in OS with mutual regulatory interactions. They synergistically regulate the AS of numerous genes associated with cell proliferation and cell cycle [ 161 ]. SRSF3 prevents cryptic intron missplicing in SRSF1 exon 4, ensuring the expression of normal SRSF1 variants, while SRSF1 promotes SRSF3 transcription without altering its AS pattern [ 161 ]. This mutual regulation between AS factors plays an important role in OS progression.

Vascular endothelial growth factor A (VEGFA)-AS events provide another representative example of OS. AS of VEGFA exons 5 to 8 generates variants with different amino acid lengths, such as VEGF 121, VEGF 165, and VEGF 189 [ 162 ]. AS at the 3′ SS of the terminal exon in VEGF 165 produces the VEGF 165b variant, which binds to VEGFR-2 but does not activate signaling pathways, thereby indirectly inhibiting the pro-angiogenic function of VEGF 165 [ 163 , 164 ]. In OS, the splicing factor Y-box binding protein 1 (YBX1) promotes the production of VEGF 165 variants while suppressing VEGF 165b, ultimately driving OS progression [ 165 ]. YBX1 also regulates the AS of DDX3X and contributes to the progression of OS [ 166 ].

Methyltransferase-like 3 (METTL3), METTL14, and Wilms tumor 1-associated protein (WTAP), key m6A modifiers, also regulate AS in OS [ 167 ]. In U2OS cells, knockdown of METTL3 induces widespread AS alterations involving 1,803 genes, which are enriched in cell cycle-related pathways, suggesting that METTL3 may play a critical role in cell cycle control of bone tumors through AS regulation [ 168 ]. METTL3 also promotes m6A modification of the 3′ UTR of the splicing factor proline- and glutamine-rich (SFPQ) mRNA, enhancing its stability [ 168 ]. In addition, previous studies have systematically elucidated the functional interplay between m6A modification and AS in cancer [ 169 ]. Methylated RNA immunoprecipitation sequencing data revealed that over 70% of AS sites within 200 bp were associated with m6A peaks mediated by METTL14, which enhances OS tumorigenicity by promoting m6A modifications in the coding region of meningioma 1 (MN1) [ 167 , 170 ]. Interestingly, AS in OS also produces noncoding RNAs that influence tumor progression. An intronic region of SRSF1 generates a 280-bp Lnc-RNA (termed Lnc-KASRT) via AS, which promotes SRSF1 binding to Krüppel-like factor 6 (KLF6) pre-mRNA, resulting in more KLF6 splice variant 1 (KLF6-SV1) than the wild-type KLF6 variant [ 171 ]. KLF6-SV1 enhances OS cell migration, invasion, and tumorigenic capacity by down-regulating p21 and up-regulating CCND1, MMP1, and MMP9 expression [ 171 ]. KLF6-SV1 also affects the PI3K/AKT pathway [ 172 ].

IGF1, widely studied in OS, also undergoes extensive AS with all its variants containing exons 3 and 4, whereas AS occurs primarily in exons 5 and 6 [ 173 ]. IGF1 consists of 6 exons and 5 introns, encoding IGF1Ea (including only exon 6), IGF1Eb (including only exon 5), and IGF1Ec variants (including both exons 5 and 6) [ 173 , 174 ]. Different OS cell lines exhibit varying dominant IGF1 variants. For instance, MG63 expresses only IGF1Ea and IGF1Ec, with IGF1Ec expression being higher than that in MNNG-HOS and HOS. IGF1Ec promotes cell cycle progression in MG63 cells, enhancing their proliferation and migration. U2OS cells predominantly express IGF1Eb. Both IGF1Eb and IGF1Ec enhance OS malignancy, whereas IGF1Ea did not. Overall, the proportion and functions of IGF1 AS variants vary among OS cell lines and warrant further investigation. Interestingly, dihydrotestosterone enhances IGF1Ea and IGF1Ec expression and induces IGF1Eb expression in MG63 cells, thereby partly explaining the more severe OS in male adolescents [ 175 ]. In OS, IGF1 variants differ fundamentally from those in osteoblasts, such as MGF. MGF can be regarded as a specialized form of IGF1Ec, whose transient induction exerts a functional role in skeletal repair. By contrast, the sustained overexpression of IGF1Ec and IGF1Eb in OS drives oncogenic processes. Moreover, mechanical stimulation in osteoblasts is intermittent, enabling a self-limiting reparative response, whereas pathological stimulation in OS is persistent, leading to irreversible activation of tumor-promoting pathways. Accordingly, the AS pattern of IGF1 in osteoblasts reflects an adaptive and transient response, while in OS, it is locked into a constitutive pro-tumorigenic state.

Studies of TP53 mutations in OS revealed that approximately 20% of the cases exhibit somatic rearrangements with breakpoints in intron 1, primarily upstream of the Hp53int1 variant [ 176 , 177 ]. In 50% of OS cases, intron 1 frameshift alterations were detected, whereas the coding sequence remained unchanged, suggesting that p53 mutations lead to abnormal AS, producing dominant-negative p53 variants that impair the normal function of p53 tetramers [ 178 , 179 ]. AS also plays a critical role in the response to genotoxic stress. Previous studies have indicated that HPV16 E6 overexpression in U2OS cells enhances AS activity. Although it did not alter sensitivity to mitomycin C treatment, it up-regulated the AS factors SFRS6 and SFRS7 [ 180 ], suggesting that AS is implicated in regulating the response to external damage or stress in the absence of p53. Human double minute X (Hdmx), an inhibitor of p53, also exhibits abnormal AS patterns in OS. The exclusion of exon 6 generated a truncated protein, Hdmx-S, with a higher p53 binding affinity, accompanied by a reduction in full-length Hdmx (Hdmx-FL) [ 181 , 182 ]. High Hdmx-S/Hdmx-FL ratios are associated with rapid metastatic progression and poor overall survival in patients with high-grade OS, making them more effective prognostic markers than p53 mutations [ 183 ]. This is attributable to elevated Hdmx-S levels driven by abnormal AS in p53-inactivated OS, which promotes tumor growth [ 183 , 184 ].

Interleukin-24 (IL-24) has garnered attention for its ability to reduce OS proliferation and stemness via Notch and Wnt/β-Catenin signaling, without affecting normal cells [ 185 , 186 ]. In U2OS, IL-24 undergoes AS to produce 7 variants that primarily affect exons 2, 3, and 5, which are regulated by SFRS6. SFRS6 overexpression increases the proportion of IL-24 variants lacking exons 2, 3, and 5 [ 187 ]. These variants activate caspases 3 and 7, inducing apoptosis in U2OS cells, but not in nontumor cell lines, such as NOK [ 187 ]. Targeting specific IL-24 AS variants represents a promising therapeutic strategy for the treatment of OS.

Ewing sarcoma

EwS is a malignant bone tumor that primarily affects adolescents between 10 and 19 years of age [ 188 ]. Its malignancy is predominantly driven by FET-ETS (fused in Ewing tumor family–erythroblast transformation-specific) fusion genes, resulting from the fusion of FET family genes with ETS family genes within tumor cells. This fusion leads to oncogenic transformation and malignant progression by disrupting the cellular physiological signaling network [ 188 ]. According to current knowledge, AS occurs across multiple processes of malignant progression of EwS, including the oncogenic fusion gene itself (Fig. 7 ). For instance, one study identified 9 AS variants of EWS-FLI1 (Ewing sarcoma friend leukemia integration 1) and one AS variant of EWS-ERG (Ewing sarcoma erythroblast transformation-specific gene) fusion genes in 23 EwS tissue samples, involving EWSR1 (Ewing sarcoma RNA binding protein 1) exon 8 and multiple exons of FLI1, such as exons 5 and 8 [ 189 ]. Approximately 21.7% of EwS tissues harbor multiple fusion variants, often with at least one in-frame variant capable of encoding functional oncogenic proteins [ 189 ]. The diverse AS variants of fusion genes in EwS affect growth factor levels and cytokine signaling pathways, thus playing a crucial role in EwS pathogenesis.

Although various oncogenic fusion proteins have been identified in EwS, more than 80% of cases involve the EWS-FLI1 fusion protein, which primarily functions as a transcription factor that dysregulates target gene expression. AS plays an important role in EWS-FLI1-induced oncogenic transformation. Proteomic studies targeting AS in EWS-FLI1 have shown that EWS-FLI1 integrates into the spliceosome network and directly interacts with several AS factors, including DDX5, hnRNPK, and PRPF6, to regulate the AS of target genes, such as CLK1 (CDC-like kinase 1), CASP3 (caspase-3), PPFIBP1 (protein tyrosine phosphatase fibronectin type III domain-containing 1), and TERT (telomerase reverse transcriptase). Compared to wild-type TERT (TERT-WT), AS of TERT induced by EWS-FLI1 down-regulation generates the γ-TERT variant (lacking exon 11), which exhibits higher telomerase activity, potentially reflecting a stress compensation mechanism of EWS-FLI1 [ 190 ]. Similarly, EWS-FLI1 directly binds the ARID1A-L variant (including exon 8) rather than the ARID1A-S variant (lacking exon 8) to form a stable chromatin remodeling BRG1/BRM-associated factor complex, supporting EwS cell proliferation and survival [ 191 ]. Reduced ARID1A-S expression decreases EWS-FLI1 stability, disrupts chromatin remodeling, and suppresses tumor growth [ 191 , 192 ]. The YK-4-279 inhibitor, which disrupts the interaction between EWS-FLI1 and ARID1A-L, induces ARID1A-S expression, leading to apoptosis, highlighting the critical regulatory role of ARID1A variants in EwS development [ 191 ]. The EWS domain of the EWS-FLI1 fusion protein is closely associated with AS. EWS recruits U1snRNP (U1 small nuclear ribonucleoprotein) and U2AF65 to the splice site of exon 6 of FAS (Fas cell surface death receptor), promoting their inclusion and enhancing FAS-mediated apoptosis [ 193 ]. By contrast, EWS-FLI1 shows reduced activity at FAS splice sites, suppressing the production of pro-apoptotic variants and aiding EwS cells to evade apoptosis [ 193 ]. Ultraviolet (UV) radiation induces the lysine acetylation of EWSR1. This modification inhibits its binding to U1 small nuclear ribonucleoprotein C (U1C), a spliceosomal component, and consequently disrupts the inclusion of exon 6 in checkpoint kinase 2 (CHK2). This leads to aberrant CHK2 splicing and impaired DNA damage repair [ 194 ]. The loss of exon 6 reduces CHK2 activity, impairs DNA damage response signaling, and promotes carcinogenesis [ 195 ]. Interestingly, UV exposure weakens the binding of EWS to target RNAs and induces transient accumulation in the nucleoli. This process triggers AS changes in ABL proto-oncogene 1 (ABL1), CHK2, and mitogen-activated protein kinase kinase kinase kinase 2 (MAP4K2), which regulate DNA damage responses [ 196 ]. Specifically, the absence of EWS alters 5′ SS selection of ABL1 exon 3, increasing the ABL1-Exon4A variant while decreasing the ABL1-ΔExon4A variant. ABL1-Exon4A reduces functional Abelson tyrosine kinase (c-ABL) protein levels, impairing DNA repair capacity, whereas ABL1-ΔExon4A maintains normal c-ABL function, aiding cell survival after DNA damage. Similarly, EWS maintains the normal inclusion of MAP4K2 exon 9, preserving MAP4K2 kinase activity and its role in stress response signaling. Overall, EWS plays a pivotal role in UV-induced genotoxic stress by specifically regulating the AS of ABL1, CHK2, and MAP4K2, thus affecting DNA damage responses and cell fate decisions. The formation of EWS-FLI1 in EwS attenuates EWS-mediated protection of DNA repair via AS regulation [ 197 ].

More recently, a study revealed that the FLI1 domain of EWS-FLI1 also mediates AS of target gene pre-mRNAs. It interacts with the AS factor RBFOX2 through the conserved C-terminal domain of FLI1 to regulate the AS of adducin 3, increasing the expression of its long variant and inhibiting the mesenchymal phenotype of ES cells, thereby reducing the metastatic potential [ 198 ]. The anti-EWS-FLI1 drug YK-4-279 disrupts the interaction between EWS-FLI1 and protein chaperones p68 (encoded by DDX5 ) and RHA, impairing EWS-FLI1-induced AS and downstream oncogenic signaling [ 190 ]. Similarly, YK-4-279 promotes apoptosis by increasing the expression of the pro-apoptotic variants of myeloid cell leukemia sequence 1 (MCL1) and B-cell lymphoma 2 (BCL2) [ 199 ].

In addition to oncogenic fusion proteins, widespread AS events occur in EwS with ribonucleoproteins that mediate oncogenic transformation via AS. A study has demonstrated that heterogeneous nuclear ribonucleoprotein M (hnRNPM) and the AS factor SRSF3 synergistically bind the intron 5 and exon 6A regions of DEAH-box helicase 9 (DHX9) to inhibit exon 6A inclusion, producing the DHX9-ΔExon6A and DHX9-Exon6A variants [ 200 ]. hnRNPM promotes the production of the non-toxic DHX9-ΔExon6A variant, stabilizing DHX9 expression and enhancing EwS cell proliferation [ 200 ]. Conversely, without hnRNPM inhibition, the DHX9-Exon6A variant is produced and rapidly degraded via nonsense-mediated decay (NMD), reducing DHX9 protein levels [ 200 ]. Thus, heightened spliceosomal activity directly contributes to drug resistance in EwS. Similarly, widespread AS programs mediated by hnRNPM have been observed following PI3K pathway inhibition in ES cells, indicating the role of AS in drug resistance [ 201 ]. In EWS-FLI1 variants, hnRNPH1 promoted the inclusion of EWS exon 8 and enhanced the expression of EWS-FLI1 variants [ 202 ]. By contrast, hnRNPH1 plays a weaker role in wild-type EWS, where exon 8 is typically excluded, forming the EWS-ΔExon8 variant [ 202 ]. This distinction highlights hnRNPH1’s specific role in regulating exon 8 splicing within the EWS-FLI1 fusion pre-mRNA, supporting EwS oncogenesis. Other scattered studies have reported AS events in EwS, such as CAPER-α scattered studies in which exon 8 is produced and rapidly degraded via nonsense [ 203 , 204 ]. In addition, 2 C-X-C motif chemokine receptor 4 (CXCR4) AS variants (CXCR4-3 and CXCR4-4) and a truncated death receptor 4 (DR4) variant were identified. Although CXCR4-3 shows strong ligand responsiveness, its function remains unclear while the truncated DR4 variant retains sensitivity to TRAIL (TNF-related apoptosis-inducing ligand) [ 205 , 206 ].

Chondrosarcoma

CS is a slow-growing primary malignant bone tumor and the second most common malignant bone tumor after OS. AS contributes to CS progression via multiple mechanisms (Fig. 8 ). A study has reported that all-trans retinoic acid (ATRA) increased the release of sulfated glycosaminoglycans from the human CS cell line HCS-2/8, suggesting that ECM remodeling in CS is associated with AS [ 207 , 208 ]. ATRA induces the production of the long AS variant ADAM28L, rather than the short variant ADAM28S. The former exhibits greater proteoglycan-degrading activity, which is closely linked to ECM remodeling [ 208 , 209 ].

The CCN family influences the ECM through distinct ligands [ 210 ]. Among these, cellular communication network factor 4 (CCN4), encoded by WISP1, is a tumor growth and metastasis suppressor protein that was previously considered essential for endochondral ossification in CS [ 210 ]. However, an AS study of CCN4 challenged this view. While the full-length CCN4 protein does not affect chondrocyte differentiation, its AS variant WISP1v, which lacks exon 3, promotes differentiation and increases alkaline phosphatase levels [ 93 ]. Interestingly, HCS-2/8 cells harbor a unique AS variant of CCN4, referred to as WISP1vx, which lacks exons 3 and 4, and encodes only a truncated IGFBP (insulin-like growth factor binding protein) module protein that is incapable of influencing CS differentiation and mineralization [ 93 ]. Further studies are required to elucidate the functions of WISP1vx. The ARHGAP family encodes proteins with RhoGAP domains that regulate GTPase activity. In CS, rho GTPase-activating protein 10 (ARHGAP10) has 2 AS variants: ARHGAP10 variant A (comprising exons 1 to 23, forming a full-length protein) and ARHGAP10 variant B (comprising exons 1 to 5 and intron 5, forming a truncated protein) [ 211 ]. ARHGAP10 variant A inhibits epithelial–mesenchymal transition (EMT) by suppressing the PI3K/AKT/GSK3β (glycogen synthase kinase 3 beta) pathway; however, the role of variant B in CS remains unclear [ 212 ].

FN1 encodes various ECM-related proteins through AS, with at least 20 variants. The skipping of exon EIIIA during chondrogenesis leads to the production of the FN1EIIIA− variant [ 213 , 214 ]. The EIIIA and EIIIB domains of FN1 are typically retained during embryogenesis, wound repair, and in tumors but are skipped in healthy adult tissues or more differentiated cartilage [ 215 , 216 ]. In rat CS cell lines, FN1 shows a variable region + connecting segment (V + C)-specific variant, low EIIIB expression, and effective inclusion of EIIIA (FN1EIIIA+), while retaining the undifferentiated chondrocytic AS pattern [ 214 ]. Intriguingly, EIIIA retention depends on approximately 200 nucleotides of the upstream intron (part of intron 32), and PTBP1 (polypyrimidine tract binding protein 1) overexpression promotes EIIIA retention [ 214 ]. Analyses of 14 osteochondrogenic tumor samples, including CS, also confirmed the inclusion of multiple variable exons, such as EIIIA and EIIIB, and the variable region, suggesting that AS facilitates CS in a dynamic remodeling state [ 217 ]. Matrix metalloproteinase 13, also known as collagenase 3, is highly expressed in tissue repair and degenerative joint diseases and is capable of degrading the cartilage matrix. It generates 4 AS variants: collagenase 3 (COL3)-APS, COL3-DEL, COL3-9B-2, and COL3-ATS [ 218 , 219 ]. In human CS cell lines such as SW1353, COL3-9B-2 arises from AS at intron 9, inserting the small exon 9B (related to Alu elements) that encodes a secreted protein [ 219 ]. However, whether COL3-9B-2 differs in its degradation activity from that of classical MMP-13 remains unknown. Aggrecan is a large aggregating proteoglycan in the ECM that confers elasticity to the cartilage. Its core contains G1, G2, and G3 globular domains, with the G3 domain exhibiting extensive AS [ 220 ]. Compared with normal chondrocytes, 2 additional AS variants of Aggrecan G3 are present in CS, involving partial or complete loss of epidermal growth factor-like 1 (EGF1), EGF2, and short complement repeat (SCR) domains, although their functions remain unclear [ 221 ]. Tenascin-C (TNC), a hexameric ECM protein, is highly expressed during tissue repair and in the CS. The proportion of AS variants is correlated with CS malignancy [ 222 , 223 ]. Compared to small TNC variants (lacking parts of the fibronectin type III domain [FN-III] repeat domain), large TNC variants (containing complete FN-III repeat domains) were more abundant in the CS cell line JJ012. As the small TNC/large TNC ratio decreases, CS progression accelerates, metastatic potential increases, and prognosis worsens [ 223 ]. The small TNC variants inhibited JJ012 adhesion in conjunction with fibronectin, whereas the large TNC variants did not [ 224 ]. Thus, TNC-AS variants serve as critical prognostic markers of CS.

Type XI collagen consists of 3 chains: α1(XI), α2(XI), and α3(XI), with multiple exons subject to selective inclusion or skipping, generating various AS variants [ 225 ]. In normal cartilage, α1(XI), α2(XI), and α3(XI) predominantly form “cartilage-type” AS variants with minimal exon inclusion [ 226 , 227 ]. However, in CS, α1(XI) predominantly exists as long variants (including exons 6a and 8), while α2(XI) tends to retain short “cartilage-type” variants, indicating a differentiation imbalance or cellular heterogeneity in CS [ 227 ]. Notably, the increase in long α3(XI) variants often correlates with enhanced CS malignancy and invasiveness, reflecting a reversion to an embryonic-like state [ 227 ].

Compared to normal tissues, CS expresses tumor/testis antigens, such as chondrosarcoma antigen gene (with at least 3 AS variants) and TRAG-3 (tumor-recognized antigen gene 3; with at least 4 AS variants), which enable normal expression of the melanoma antigen gene (MAGE) [ 228 ]. MAGE has become a critical target for autologous T-cell therapy, suggesting that CS could potentially serve as a candidate for T-cell therapy [ 229 , 230 ]. An extraskeletal myxoid CS subtype is characterized by an EWS/NOR1 fusion, producing the EWS/NOR1 fusion protein. This protein binds U1C (a U1 snRNP particle component), influencing distal 5′ SS usage in pre-mRNA, ultimately compensating for snRNP function [ 231 , 232 ]. This suggests that CS involves fusion proteins that directly participate in AS regulation.

Metastatic bone cancers

Bone metastases are common and can be classified as osteoblastic and osteolytic lesions based on their distinct characteristics (Fig. 9 ). These metastases often lead to complications, such as pathological fractures, spinal cord compression, and bone pain [ 233 ]. In lung cancer, truncated AS variants of ADAM8 exhibit structural alterations that expose integrin-binding domains, enhancing interaction with integrins [ 234 ]. This increases osteoclast activity, drives osteolytic destruction, and promotes the development and expansion of bone metastases [ 234 ]. Prostate cancer (PCa) frequently metastasizes to bone, and PCa cells often employ androgen receptor (AR) AS to evade androgen deprivation therapy or AR signaling inhibitors [ 235 ]. The AR variant AR-V7, which lacks the ligand-binding domain (corresponding to exons 4 to 8) but retains the constitutively active N-terminal domain, cannot be targeted by AR inhibitors. This variant transcriptionally activates AR signaling and is considered a key driver of resistance [ 236 – 238 ]. In castration-resistant PCa (CRPC), bone metastases exhibit increased aggressiveness and osteoblastic lesions in CRPC. AR-V7 enhances PCa bone metastasis by increasing the chromatin accessibility of EMT-related genes and up-regulating the transcription of the metastasis driver gene SOX9 (sex-determining region Y-box 9), highlighting the critical role of AS in bone metastasis [ 239 ]. AR-V7 has been implicated in shaping the immunosuppressive phenotype of PCa. Prophylactic vaccination targeting AR-V7 induces robust antitumor immune responses in preclinical models, but its efficacy is reduced in immunosuppressive settings where local CD8 + T-cell exclusion may counteract vaccine effects [ 240 ]. Transcriptomic analyses show that AR-V7 overexpression down-regulates pathways such as “leukocyte migration” and “chemokine-mediated signaling pathway”, and similar suppression of immune-related gene programs has been reported in breast cancer [ 240 , 241 ]. These findings indicate that AR-V7 may promote immune evasion through direct or indirect mechanisms. In addition, other AR variants, such as AR-V1 and AR-v567es, are expressed in PCa bone metastases. AR-V1 is composed of exons 1 to 3 and the cryptic exon CE1, whereas AR-v567es retains exons 1 to 4 but lacks exons 5 to 7, resulting in a frameshift and premature stop codon after the 10th amino acid of exon 8, forming a truncated protein [ 242 ]. These variants also lack the ligand-binding domain but exhibit constitutive activity, directly activating downstream AR transcription programs, particularly cell cycle-related genes such as cyclin-dependent kinase 1 (CDK1), thereby promoting the progression of bone metastases [ 242 , 243 ]. The CRPC-specific Erb-B2 receptor tyrosine kinase 3 (ERBB3) AS variant p45-sErbB3 lacks the transmembrane domain and the C-terminal sequences, transforming ERBB3 from a transmembrane protein into a paracrine factor [ 244 ]. p45-sErbB3 is closely associated with CRPC progression, stimulating osteoblast proliferation and differentiation and promoting the secretion of bone sialoprotein and other factors in the bone microenvironment, contributing to PCa bone metastases [ 245 ]. Cathepsin Z (CTSZ) mediates the degradation of the AS factor TRA2A via the proteasomal pathway, thereby promoting the expression and secretion of specific IL-32 variants and lifting the molecular suppression of macrophage infiltration [ 246 ]. This process activates the ITGA5/PI3K signaling axis in macrophages, markedly enhancing the recruitment of M2 macrophages and ultimately accelerating the metastatic progression of PCa [ 246 ].

Certain AS factors influence PCa bone metastasis. For instance, overexpression of RBM3 (RNA binding motif protein 3) and other factors in the bone microenvironment contributes to the regulation of cell cycle–related genes such as CDK1, thereby promoting the metastatic potential [ 247 ]. CD44, composed of 10 constant exons and 10 variable exons (v1 to v10) inserted between exons 5 and 16, generates distinct CD44 variants [ 248 ]. RBM3 suppresses the CD44v8 to v10 variant and increases the CD44s (standard CD44, lacking variable exons) variant, reducing PCa invasiveness and metastatic ability and offering a new avenue for addressing PCa bone metastases [ 249 ]. Moreover, CD44 variants serve as biomarkers of tumor recurrence and progression in ampullary adenocarcinoma, particularly CD44v3-10 and CD44v6-10 [ 250 ]. In metastatic prostate adenocarcinoma, the AS factor HSPB1 (heat shock protein family B member 1) potentially regulates AS events in genes, such as SRC, EGFR (epidermal growth factor receptor), MAPT (microtubule-associated protein tau), APP (amyloid precursor protein), and PRKCA (protein kinase C alpha), thereby contributing to bone metastasis [ 251 ]. Interestingly, intronic BP site switching may play a role in PCa. In normal bone tissue versus metastatic PCa, BP site switching was observed in the first intron of HBB (hemoglobin subunit beta), potentially disrupting U2 snRNA binding [ 252 ]. This suggests that tumor-specific AS environments guide the selection of intronic BP sites. The prostate-specific membrane antigen produces at least 3 AS variants: prostate-specific membrane (PSM), PSM-C, and PSM-D. However, only PSM-D is expressed at higher levels in lymph nodes and bone metastases than in primary tumors, indicating its specific role in PCa metastasis [ 253 ].

Bladder cancer (BLCA), a malignant lesion originating from bladder epithelial cells, is the second most common genitourinary malignancy that frequently metastasizes to the bones. In patients with BLCA and bone metastases, the AS factor JUP (junction plakoglobin) was identified as a regulator of the AP of SMOX (spermine oxidase) and the exon-skipping event of ITGB4 (integrin subunit beta 4), which influences BLCA metastasis [ 254 ]. Additionally, in mesothelioma, the AS factors HSPA1A (heat shock protein family A member 1A) and DDX3Y (DEAD-box helicase Y) regulate the AS of SNX5 (sorting nexin 5), affecting immune evasion and bone metastasis [ 255 ].

Osteopontin (OPN) facilitates bone metastasis in breast cancer and PCa and generates 3 AS variants: OPN-a (full-length), OPN-b (lacking exon 5), and OPN-c (lacking exon 4) [ 256 , 257 ]. Overexpression of OPN-a promotes the adhesion of non-small cell lung cancer (NSCLC) cells to bone tissue via αvβ3 integrin, and blocking αvβ3 integrin reduces adhesion [ 258 ]. Compared to adjacent tissues, OPN-b is highly expressed in soft tissue sarcomas and is associated with tumor staging, whereas OPN-c is correlated with tumor stage and lymph node metastasis in soft tissue sarcomas [ 257 ]. However, the roles of OPN-b and OPN-c in bone metastasis remain unclear.

In breast cancer bone metastases, parathyroid hormone-related protein (PTHrP) facilitates osteolytic destruction via 3 AS variants: PTHrP-1-139, PTHrP-1-141, and PTHrP-1-173 [ 259 ]. PTHrP-1-139, which includes exons 6 and 7, exhibits high secretion and osteolytic activity, leading to hypercalcemia. PTHrP-1-141 includes exons 6, 7, and 8, which have lower secretion and osteolytic activity [ 85 , 260 ]. PTHrP-1-173, the longest variant containing exons 6 to 9, exhibits minimal secretion and osteolytic activity [ 85 ]. These AS variants selectively promote osteoclast formation and bone resorption, providing a growth advantage for breast cancer bone metastases, and offering a novel evaluation metric for patients with breast cancer bone metastases.

Cysteine-rich secreted proteins (secreted protein acidic and rich in cysteine [SPARC]) enhance bone metastasis and EMT in glioblastoma, melanoma, and PCa by modulating cell–matrix interactions [ 261 ]. However, SPARC suppresses proliferation, migration, and bone metastasis in breast cancer, as validated in gene-knockout mice [ 261 – 263 ]. SPARC undergoes AS to produce at least 5 variants (SMOC1 [SPARC-related modular calcium-binding protein 1], SMOC2 [SPARC-related modular calcium-binding protein 2], SPOCK [SPARC-related osteonectin-like protein], follistatin-like 1, and SPARC-like-1), with N- or C-terminal domain deletions altering integrin and growth factor binding, exhibiting effects opposite to those of full-length SPARC in the adhesion, migration, and EMT processes [ 264 ]. For example, in breast cancer, SPOCK interacts with SIX1 to promote EMT and cell proliferation via the AKT/mTOR (mechanistic target of rapamycin) pathway [ 265 ]. Notably, SPARC variant generation and function are regulated by key signaling molecules such as TGF-β in bone metastases [ 264 , 266 ]. As TGF-β accumulates in the bone matrix and is activated during osteolytic destruction in bone metastases, it forms a positive feedback loop with signals such as PTHrP, VEGF, and interleukin-11, enhancing the invasiveness and EMT levels of bone metastases [ 267 ]. Furthermore, TGF-β directly regulates CD44, p120 catenin, and fibroblast growth factor receptor 2 (FGFR2) [ 267 ]. Due to mutually exclusive exon IIIb and exon IIIc usage, FGFR2 generates 2 AS variants, FGFR2b and FGFR2c, with FGFR2b primarily expressed in epithelial cells and FGFR2c in mesenchymal cells [ 268 ]. TGF-β stimulation increases FGFR2c levels, enhancing sensitivity to fibroblast growth factor 2 and promoting EMT, suggesting that inhibiting TGF-β signaling could mitigate tumor bone metastases through AS modulation [ 269 ]. The timeline highlights the key discoveries of aberrant AS in bone metastasis over the past 20 years (Fig. 10 ).

Physiological and pathological AS in bone

We summarized the characteristics of AS in bone tissue under physiological and pathological conditions, which can be compared across several dimensions (Table 3 ). In terms of shared features, both states rely on core splicing regulators, such as serine/arginine-rich splicing factors (e.g., SRSF1 and SRSF3) and hnRNPs, which coordinate exon selection and splice site recognition. In addition, the predominant types of splicing events—ES, IR, and MXE—are observed in both bone development and tumor progression. Third, as a post-transcriptional regulatory mechanism, AS dynamically modulates gene expression in both contexts, influencing osteogenic differentiation, matrix remodeling, and cellular responses to external stimuli.

In bone tissue, AS is far more than simply “cutting out the exon”. AS constructs distinct transcriptomic landscapes under physiological and pathological states. In homeostatic conditions, AS functions like a precise engineer, regulating the differentiation rhythm of bone cells, mediating responses to mechanical stimuli, and maintaining dynamic remodeling of the bone matrix. Upon transition into a pathological state, the role of AS shifts subtly but decisively, becoming a driver of tumor progression. In OS, ES, and bone metastases, AS ceases to be a tool for homeostatic regulation and instead participates in amplifying oncogenic signaling, promoting therapeutic resistance, and enabling immune evasion. By generating splice variants that promote angiogenesis, EMT, migration, or resistance to apoptosis, AS directly facilitates tumor cell adaptation to complex microenvironments and enables distant dissemination. This view is further supported by studies of AS in other malignancies [ 11 , 16 , 270 ]. Of note, certain AS events themselves serve as functional boundaries between physiological and pathological states. For example, AS of VEGFA produces VEGF165 and VEGF165b isoforms, with the former promoting angiogenesis and the latter acting as an endogenous inhibitor. Up-regulation of YBX1 disrupts this balance by favoring VEGF165 expression, thereby facilitating tumor progression. Similarly, AS of oncogenes such as IGF1 and TP53 can shift their functional roles from promoting osteogenesis to driving tumorigenesis, depending on the splicing pattern.

Although AS-related research in other pathological bone conditions (e.g., osteoporosis, OA, and hereditary bone diseases) is relatively limited, we have summarized the key findings in these fields. First, in osteoporosis-related AS research, conditional knockout of NIBAN2 in male mouse osteoblasts does not affect overall body weight or developmental processes, but leads to insufficient femoral mineralization and an osteoporotic phenotype [ 271 ]. Mechanistically, NIBAN2 stabilizes the binding of core components of the HNRNPU splicing complex (e.g., DDX39B and RBMX) to promote retention of RUNX2 exon 6 (containing the nuclear localization sequence), thereby generating a functional RUNX2 protein capable of nuclear translocation and transcriptional activation of osteogenic differentiation [ 271 ]. In addition, with aging, the expression of the AS factor YBX1 is reduced in mouse BMSCs, a change closely associated with decreased bone mass and enhanced adipogenic differentiation [ 272 ]. Conditional knockout of Ybx1 using Prx1-Cre in mice accelerates bone loss, whereas Ybx1 overexpression in BMSCs promotes osteogenesis [ 272 ]. Mechanistically, YBX1 deficiency induces aberrant AS of multiple genes associated with osteogenesis and aging (e.g., Fn1, Sp7, Spp1, Sirt2, and Nrp2), thereby accelerating BMSC senescence and driving adipogenic lineage commitment [ 272 ]. Moreover, glucocorticoids, through the HDAC4-FOXC1 epigenetic axis, down-regulate SRSF1 expression in osteoblasts, resulting in up-regulation of Bmf v2 and v3 AS variants, which enhances osteoblast apoptosis [ 273 ]. This finding highlights the critical role of SRSF1 in glucocorticoid-induced osteoporosis and osteonecrosis [ 273 ]. In human cohorts with fractures, genome-wide association studies combined with transcriptomic analyses have identified candidate AS genes associated with osteoporosis, including USP48, MAEA, and KLHL8 [ 274 ]. These statistical associations still require further functional studies to confirm causality and elucidate potential mechanisms.

Secondly, AS has also been implicated in OA. Differential AS events have been identified between OA and healthy meniscus tissues, involving apoptosis-related genes such as BCL2L13 , XAF1 , and NF2 , as well as OA progression-related genes including IL-16 and PRG4 [ 275 – 277 ]. In a study of 101 primary chondrocytes (FN-f stimulation vs. PBS control), a total of 7,188 splicing quantitative trait loci (sQTLs) were mapped [ 278 ]. Colocalization of sQTLs with OA GWAS signals further identified 6 potential OA-associated AS genes, namely, PBRM1 , WWP2 , COLGALT2 , GCAT , HMGN1 , and RNF144B [ 278 ], providing a foundation for future mechanistic studies and targeted interventions in OA. The AS factor DDX5 is down-regulated during OA progression, leading to increased proportions of full-length FN1 and PLOD2 variants. This shift drives cartilage fibrosis and degradation by up-regulating fibrosis-related genes (e.g., COL2A1 , COL1A1 , and COL1A2 ) and matrix-degrading genes (e.g., MMP13 and NOS2 ) [ 106 ]. Conditional knockout of Ddx5 in chondrocytes demonstrated that targeted regulation of Fn1 and Plod2 can partially rescue the OA phenotype, suggesting that restoration of DDX5-mediated AS balance may represent a novel therapeutic strategy for OA [ 106 ]. In a post-traumatic OA mouse model, the CALCA gene was shown to generate 2 distinct AS products: the procalcitonin/calcitonin (PCT/CT) variant, which attenuated cartilage degeneration and subchondral bone loss but exacerbated synovial inflammation, and the αCGRP variant, which promoted osteophyte formation [ 279 ]. These findings highlight that AS products derived from the same gene may exert divergent or even opposing functions during OA progression [ 279 ].

Finally, in the field of hereditary bone diseases, approximately 90% of osteogenesis imperfecta (OI) cases are caused by mutations in the type I collagen α-chain genes (collagen type I alpha 1 chain/collagen type I alpha 2 chain [COL1A1/COL1A2]), leading to either quantitative defects (reduced synthesis) or qualitative defects (structural abnormalities) of type I collagen [ 280 , 281 ]. One familial study identified a 562-bp deletion spanning the end of exon 34 to the middle of intron 36 in COL1A1, with the mutant allele undergoing aberrant splicing to generate 3 transcripts (IR, multi-exon skipping, and cryptic donor usage). As a result, some in-frame transcripts failed to produce stable α1(I) chains, accompanied by an abnormal increase in type III collagen [ 282 ]. This AS defect was consistent with the severe type III OI phenotype observed in patients [ 282 ]. Another study reported that a COL1A1 donor splice site mutation (c.1155+1G>C) caused exon 17 skipping, thereby disrupting α1(I) chain integrity and producing a typical autosomal dominant OI phenotype [ 283 ]. Similarly, a donor splice site mutation in COL1A2 intron 46 (IVS46+2T>C) activated an upstream cryptic splice site, leading to a 17-bp deletion, frameshift, and premature termination. This led to the loss of the α2(I) chain and the exclusive formation of unstable α1(I) homotrimers, consequently giving rise to a dominant OI phenotype [ 284 ]. In addition, a splice site mutation in intron 3 of TMEM38B (c.455-7T>G) was identified in Chinese Han OI patients, which impaired TRIC-B ion channel function and contributed to autosomal recessive OI [ 285 ]. Collectively, these findings highlight AS as an important molecular mechanism in OI pathogenesis and a potential therapeutic target. Beyond OI, mutations in COL2A1 affecting exon 12, exon 21, and the intron 20/exon 21 junction have been closely associated with chondrodysplasias [ 286 – 289 ]. Splice site mutations in PHEX are closely linked to X-linked hypophosphatemia [ 290 – 293 ], while splice alterations in TCIRG1 are associated with osteopetrosis [ 294 – 297 ].

In summary, AS under physiological and pathological conditions displays markedly different characteristics, providing new opportunities for future research (Table 4 ). For example, by integrating single-cell sequencing and spatial transcriptomics, it will be possible to systematically dissect the AS patterns of osteoblasts, chondrocytes, and osteoclasts under mechanical stimulation, thereby constructing an atlas of mechanically regulated AS. In this direction, several groups have already developed dedicated computational algorithms to decode the complexity of AS networks [ 298 – 300 ]. Furthermore, exploring the AS networks involved in the crosstalk between immunity and bone metabolism is another promising avenue. Osteoimmunology has already produced many interesting findings, such as T cells and B cells directly regulating osteoclastogenesis through the RANKL–RANK–OPG axis [ 301 , 302 ]. However, the role of AS in osteoimmunology remains unexplored, which represents a potential innovation point. Finally, whether AS variants or AS factors can serve as therapeutic targets could first be prospectively validated in patient cohorts using liquid biopsy [ 303 ]. After identifying patient subgroups dependent on specific AS events, targeted drug development could then be pursued, which holds promise to significantly improve the success rate of clinical trials [ 303 , 304 ].

Advances in AS-Targeted Therapeutics

Following the previous sections that highlighted the critical roles of AS in bone biology and tumor progression, we now turn our attention to its clinical relevance and translational potential. Targeted therapies developed based on AS mechanisms can be categorized into 3 approaches: targeting abnormal AS variants, targeting cis-regulatory elements of pre-mRNA, and targeting AS factors [ 305 – 307 ]. Targeting abnormal AS variants or cis-regulatory elements entails the precise delivery of exogenous agents to correct redundant or pathogenic AS variant functions in defined tissues or organs. Potential strategies include the use of ASO, gene editing, RNA aptamers, siRNA, and shRNA, which offer the advantage of high specificity and enable personalized treatment [ 306 ]. However, a major limitation is that nucleic acid drugs such as ASO typically require administration via injection and face difficulty crossing the blood–brain barrier. In addition, unmodified ASOs are unstable in vivo, prone to degradation by nucleases, and trigger immune responses, thus necessitating chemical modifications to enhance stability and reduce immunogenicity [ 308 , 309 ]. Targeting AS factors essentially involves activating or inhibiting specific protein activities through small-molecule drugs to reverse abnormal splicing patterns [ 310 , 311 ]. Small-molecule therapy can be administered orally or intravenously, providing better bioavailability and widespread systemic distribution, but the main drawback lies in its limited tissue specificity [ 312 ]. A single AS factor is often involved in the AS of multiple genes, which could result in non-specific effects and potential side effects, thereby increasing the complexity of predicting drug responses [ 313 , 314 ]. We have summarized AS-related targeted therapeutics that are either approved or currently undergoing clinical trials, and explored their potential applications in bone-related diseases (Table 5 ).

A classic example comes from the development of therapeutic strategies for SMA. SMA is primarily caused by deletions or mutations in the SMN1 gene, leading to a deficiency of SMN protein [ 315 ]. Fortunately, a homologous gene, SMN2, exists in the human genome, but due to exon 7 exclusion, SMN2 predominantly generates unstable or non-functional protein [ 315 ]. Nusinersen is an 18-nucleotide ASO that targets the ISS-N1 region downstream of intron 7 in SMN2 pre-mRNA, a cis-acting element that inhibits exon 7 inclusion [ 316 ]. By masking this region, Nusinersen promotes exon 7 retention, increases the proportion of full-length SMN2 mRNA, and elevates the expression of functional SMN protein, thereby partially compensating for the deficiency of SMN1 [ 317 – 319 ]. Among patients treated with Nusinersen, 51% of infants achieved motor developmental milestones (e.g., independent sitting, rolling, and standing) post-treatment, compared to 0% in the control group, and also exhibited prolonged survival without ventilatory support, leading to a substantial improvement in quality of life for patients with SMA [ 320 , 321 ]. Risdiplam, a diazine derivative, is an orally administered small-molecule splicing modulator that selectively targets 2 sites in SMN2 pre-mRNA, namely, the exon 7 ESE and the intron 7 5′ SS, in order to promote exon 7 inclusion and thereby increase functional SMN protein production to alleviate SMA symptoms [ 322 , 323 ]. Clinical studies of Risdiplam in patients with SMA types 1, 2, and 3, including infants, children, and young adults, demonstrated improvements in motor function, increased survival, and favorable tolerability, offering an effective oral therapeutic option for SMA patients across various types and age groups [ 324 , 325 ]. Another splicing modulator, Branaplam, had entered clinical studies for SMA and Huntington’s disease but was discontinued due to safety concerns [ 326 , 327 ]. Mechanistically, Branaplam modulates the 5′ SS through a sequential 3-step binding process: U1-C first reversibly binds U1 snRNP, then recognizes aberrant 5′ SS featuring a -1A bulge, ultimately forming a Branaplam-U1 snRNP-5′ SS RNA complex that stabilizes splice site recognition and promotes exon retention [ 328 ]. Taken together, insights gained from the therapeutic development of SMA have profound implications for splicing-targeted strategies. First, precise identification of pathogenic AS variants is foundational to the success of ASO-based therapy. Second, promoting exon retention represents a mechanistically clear and clinically validated pre-mRNA intervention strategy. Third, targeting AS factors offers broad regulatory potential but necessitates careful evaluation of off-target splicing alterations and associated safety risks.

E7107 is the first small-molecule spliceosome inhibitor to enter clinical trials. It targets the SF3B1 subunit and disrupts the recognition of the branch point in pre-mRNA by the U2 snRNP, thereby blocking spliceosome assembly and inducing NMD and reductions in mature mRNA levels [ 329 – 331 ]. In a Phase I trial for advanced solid tumors, although 8 patients exhibited stable disease following E7107 treatment, the trial was terminated due to 2 cases of E7107-associated irreversible vision loss [ 329 ]. Subsequent preclinical studies found that E7107 strongly induced apoptosis in uveal melanoma and pancreatic ductal adenocarcinoma cell lines, although clinical trial data remain unavailable [ 330 , 332 ]. Similarly, newer SF3B1-targeting inhibitors have emerged, including the Sudemycin family and H3B-8800, which demonstrate stronger splicing interference and cytotoxicity in tumor cells with comparatively milder effects on normal cells, indicating potential clinical applicability [ 333 , 334 ]. Among them, the Phase I clinical trial of H3B-8800 observed alleviation of red blood cell transfusion dependence in a subset of myelodysplastic syndrome (MDS) patients harboring SF3B1 missense mutations. However, due to the absence of objectively defined clinical responses, further clinical development was discontinued [ 335 ]. In summary, experience from the development of SF3B1 inhibitors offers several important insights. Although AS factors are often hyperactivated in tumors, their ubiquitous expression across normal tissues poses a substantial risk of toxicity, particularly to functionally sensitive organs such as the eyes. Moreover, while mutations in splicing factors confer increased drug sensitivity, such genetic alterations alone are insufficient to achieve complete tumor eradication. These findings underscore the necessity for splicing-targeted therapies to account for tissue-specific regulatory dynamics and the functional consequences of modulating distinct AS events.

The AS variant of the PKM gene represents a representative case of ASO application. In human PKM pre-mRNA, exons 9 and 10 are mutually exclusive, generating 2 AS variants: PKM1 (containing exon 9) and PKM2 (containing exon 10) [ 336 ]. Up-regulation of PKM2 promotes aerobic glycolysis, providing substrates for anabolic processes such as lipid and nucleotide synthesis, while also activating transcription factors including HIF-1α, β-catenin, and STAT3, thereby enhancing cellular proliferative capacity. This mechanism plays a critical role in the progression of inflammatory bone diseases and OS [ 337 – 339 ]. A PKM ASO modified with constrained ethyl (cEt) bases and optimized with a phosphorothioate backbone effectively blocks a cis-regulatory element that promotes exon 10 inclusion in PKM pre-mRNA, inducing a controlled switch from PKM2 to PKM1 splicing. This approach has been shown to suppress the progression of hepatocellular carcinoma and glioblastoma [ 336 , 340 ]. These findings suggest that in bone-related diseases characterized by aberrant PKM2 up-regulation, PKM ASO could represent a promising therapeutic strategy, though its efficacy remains to be further validated in clinical trials. Additional ASO targets, including BCL-X, MNK2, MDM4, and PD-L1, have been extensively discussed in previous reviews [ 305 ].

Protein arginine methyltransferase 5 (PRMT5) regulates pre-mRNA splicing by methylating AS factors, including members of the Sm and SR protein families [ 341 ]. At least 12 PRMT5-targeting compounds are undergoing clinical investigation [ 342 ]. Mechanistically, in diseases harboring AS factor mutations (such as MDS and chronic myelomonocytic leukemia), tumor cells become dependent on PRMT5 to maintain a “barely stable” splicing system; thus, PRMT5 inhibition induces synthetic lethality, selectively triggering tumor cell death [ 343 , 344 ]. Furthermore, homozygous deletion of methylthioadenosine phosphorylase (MTAP) leads to the accumulation of MTA, which renders cells dependent on the residual activity of PRMT5, thereby disrupting AS and exposing a novel therapeutic vulnerability [ 345 , 346 ]. PRMT5 inhibitors have demonstrated encouraging results in clinical trials for various solid tumors, including NSCLC, pancreatic cancer, biliary tract cancer, and esophageal carcinoma. Moreover, PRMT5 regulates the AS of key DNA repair factors TIP60 and SUV4-20H2, affecting their enzymatic activities and chromatin-modifying functions, thereby participating in the homologous recombination (HR) repair pathway [ 347 ]. Based on this, PRMT5 inhibition can induce an “HR deficiency” phenotype, and its combination with PARP inhibitors is expected to induce synthetic lethality [ 348 ]. Inhibition of PRMT5 also enhances osteogenic differentiation of mesenchymal stromal cells while reducing their proliferative capacity [ 347 , 349 , 350 ]. Additionally, PRMT5 inhibition suppresses osteoclast differentiation and impairs chondrocyte development, suggesting a critical role for PRMT5 in skeletal development [ 351 , 352 ]. In soft tissue sarcomas, including OS, PRMT5 inhibition reduces glycolytic rates and slows tumor cell proliferation [ 353 ]. Collectively, evidence highlights PRMT5 inhibitors as promising therapeutic targets for bone tumors, particularly in tumor subtypes exhibiting high dependency on PRMT5 and harboring specific splicing factor mutations. In addition, considering the essential function of PRMT5 in maintaining HR repair, its inhibition in combination with PARP inhibitors offer a novel approach for the precision treatment of bone malignancies.

Serine/arginine-rich protein-specific kinase (SRPK), as kinases, activate SR proteins through phosphorylation, alter their nuclear and cytoplasmic localization, promote spliceosome assembly, and participate in the pathogenesis of NSCLC, triple-negative breast cancer, leukemia, and gliomas [ 354 – 358 ]. In leukemia, SRPK enhances the splicing regulatory capacity of SRSF1 on cyclin A1 pre-mRNA by phosphorylating SRSF1 at S422, thereby accelerating cell cycle progression and promoting tumor cell proliferation [ 357 ]. The development of SRPK inhibitors has evolved through 3 stages: traditional ATP-site kinase inhibition (e.g., SRPINX31 and SRPIN340), specific inhibition of protein–protein interaction interfaces (e.g., DBS1), and covalent binding to target residues (e.g., C-DBS) [ 359 ]. Notably, SRPK inhibitors can shift VEGF splicing toward the VEGF165b variant (anti-angiogenic), thereby inhibiting tumor angiogenesis [ 360 ]. By modulating VEGF, SRPK inhibitors have entered clinical trials for diabetic macular edema, demonstrating therapeutic potential in reducing retinal vascular leakage [ 361 ]. SRPK1 inhibitors have shown potent anti-angiogenic activity in vitro by increasing VEGF165b levels, although their anti-tumor and anti-angiogenic effects in intrahepatic cholangiocarcinoma models have not yet been validated in vivo [ 362 ]. Overall, as the “activation switch” of SR proteins, SRPK represents a promising drug target for inhibiting tumor cell proliferation and neovascularization in bone-related malignancies.

In clinical and translational studies of bone-related diseases, we summarized validated AS-related cases. In PCa, detection of the AR-V7 variant in circulating tumor cells has been used to predict the efficacy of antiandrogen drugs [ 363 ]. This approach is particularly suitable for patients with bone metastatic CRPC, as bone is the most common metastatic site and often indicates therapeutic resistance [ 237 ]. In OA, chondrocytes re-express the COL2A1IIA variant in a “reverted” manner, and detection of COL2A1IIA-derived fragments in synovial fluid or blood, such as PIIANP, is commonly used to evaluate cartilage remodeling and disease activity [ 364 ]. PIIANP fragments have also been applied to assess cartilage damage in Kashin–Beck disease [ 365 ]. Soluble biomarkers based on COL2A1 AS variants can reflect OA progression earlier and with greater sensitivity than imaging examinations [ 101 , 366 ]. There are also representative cases in the field of AS-related therapies. Denosumab, used for the treatment of osteoporosis and giant cell tumor of bone, specifically recognizes and neutralizes certain sRANKL proteins generated by AS, thereby blocking the RANKL–RANK signaling pathway and efficiently suppressing osteoclast activity [ 367 ]. Finally, in bone defects or non-union fractures, miR-exon4 derived from exon 4 of AMELX promotes osteogenic differentiation by directly suppressing Nfia/Prkch and up-regulating RUNX2 in osteoblasts [ 69 , 368 ]. Leveraging advances in miRNA therapeutic delivery, the AS-derived miR-exon4 represents a compelling candidate for bone regeneration strategies, which requires further in vivo validation.

Small-molecule inhibitors targeting AS factors have encountered frequent clinical failure primarily due to off-target toxicities and limited tissue specificity, emphasizing the urgent need for strategies that enhance selective delivery in the context of bone tumor therapy. SF3B1 inhibitors such as E7107 were discontinued following cases of irreversible vision loss, while H3B-8800, despite alleviating transfusion dependence in a subset of MDS patients, did not achieve clearly defined clinical responses. PRMT5 inhibitors demonstrated efficacy in synthetic lethality models but raised safety concerns because of their broad roles in DNA repair and bone development. SRPK inhibitors can modulate VEGF AS and suppress angiogenesis, but their in vivo validation remains limited. To surmount these limitations, the development of next-generation strategies that achieve greater precision and tissue selectivity is imperative. Potential strategies include the design of nanocarriers responsive to the bone tumor microenvironment to achieve active targeting and minimize normal tissue exposure. The application of PROTAC technology to selectively degrade AS factors may provide a more precise therapeutic modality. Moreover, computer-aided screening coupled with AI-based predictive modeling offers opportunities to optimize the structural and pharmacological properties of small-molecule inhibitors. Finally, localized sustained-release delivery following bone tumor resection and patient stratification guided by AS variant biomarkers to enrich for populations most likely to benefit with minimal toxicity represent promising directions for the refinement of AS-targeted therapeutic strategies.

Overall, AS has emerged as a highly promising translational research direction in the diagnosis and treatment of bone-related diseases (Table 6 ). From diagnostic and prognostic biomarkers such as AR-V7 and COL2A1IIA/PIIANP, to therapeutic molecules such as Denosumab and miR-exon4, and further to nucleic acid drugs and small-molecule inhibitors under development such as ASO, AS modulators, and PRMT5/SRPK inhibitors, current evidence indicates that AS research is steadily moving from “mechanistic exploration” toward “clinical implementation”. The therapeutic success of AS-targeting drugs in SMA provides a strong rationale to accelerate their clinical translation in bone tumors, OA, osteoporosis, and bone regeneration. Future studies should focus on precisely defining pathogenic AS events under disease contexts, achieving tissue- and cell type-specific targeted intervention, and utilizing novel delivery systems such as bone-targeted nanocarriers to enhance drug accumulation at lesion sites. Therefore, elucidating the pathogenic mechanisms of AS in bone diseases and actively promoting interdisciplinary collaboration will be critical pathways to achieve precision diagnosis and targeted therapy in the future.

Conclusion and Future Perspectives

Over the past 2 decades, our understanding of AS processes in human tissues and their functions in bone biology has greatly expanded. Core spliceosome components and key factors regulating AS, initially identified in yeast splicing systems, have gradually been observed to have functional homologs in human cells, providing crucial insights into the conservation and specificity of eukaryotic AS. AS is extensively involved in bone tissue development, cellular differentiation, matrix remodeling, and activation of critical signaling pathways. The differentiation processes of various cell types in mature bone tissue, such as osteoblasts and osteoclasts, highly depend on AS regulation of key genes, which finely tune cell fate determination and functional differentiation pathways, thereby maintaining bone homeostasis and regenerative capacity. Notably, bone tissue exhibits distinct AS patterns of calcium channel genes. AS events in the RANK–RANKL signaling pathway contribute to bone remodeling, highlighting the critical role of AS in the dynamic development of the musculoskeletal system. These findings not only deepen our understanding of AS regulatory mechanisms in bone biology but also offer new directions for investigating the pathogenesis of bone remodeling disorders and developing AS-based therapeutic strategies. This review systematically compares the roles of AS in bone biology, bone tumor progression, and bone metastasis, revealing both shared mechanisms and context-specific regulation. Beyond intrinsic cellular AS regulation, it also highlights the pivotal role of AS within the bone microenvironment. Furthermore, from a clinical perspective, it comprehensively summarizes therapeutic strategies targeting AS and proposes their potential directions in translational research, providing new insights for future precision diagnosis and therapy.

In bone tumors, the ratio of different AS variants of the same gene is closely correlated with poor prognosis, suggesting that AS variants serve as critical biomarkers for tumor progression. Moreover, AS is widely implicated in biological processes, such as tumor proliferation, migration, apoptosis, angiogenesis, DNA damage repair, EMT, microenvironment remodeling of bone metastases, and TGF-β signaling pathways. Notably, given the subtle differences between AS variants, which typically involve only a single exon or short sequence, conventional antibody-based detection methods often lack specificity. Consequently, most studies have relied on PCR-based approaches for detecting AS. The clinical application of AS-derived biomarkers currently faces 2 major bottlenecks: insufficient detection accuracy and difficulties in quantification. RT-qPCR is limited by primer specificity and amplicon length, making it difficult to distinguish highly similar or longer AS variants, often resulting in reduced sensitivity. Therefore, more advanced methods such as droplet digital (dd) PCR, long-read sequencing, or targeted RNA capture sequencing are needed to improve detection accuracy. AT-qPCR has been proposed for variant quantification, but it lacks validation and standardization. In the future, establishing stable internal reference systems will be essential to ensure reproducibility. Developing highly specific antibodies that target proteins encoded by different AS variants is a promising direction for future studies. Abnormal AS patterns in tumors have emerged as a frontier in cancer research. However, the mechanisms and functions of aberrant AS in bone tumors remain insufficiently explored, and related studies are still in their infancy. In particular, it is necessary to conduct large-scale clinical cohort studies to validate prognostically relevant AS variants in OS, ES, and CS, in order to further establish their clinical utility in diagnosis, molecular subtyping, and prognostic evaluation. Moreover, scRNA-seq and long-read sequencing will be instrumental in constructing comprehensive splicing landscapes of skeletal cells during development and disease, and establishing bone disease-specific “splicing code” databases.

We also recognize the clinical translational potential of AS-related therapeutic strategies. Currently, 3 major approaches have been developed for AS-targeted therapy: directly targeting aberrant AS variants, modulating cis-regulatory elements of pre-mRNA, and regulating the activity of AS factor proteins. Among them, ASOs, siRNAs, and gene-editing technologies offer high specificity but face challenges such as poor delivery efficiency, limited biological stability, and immunogenicity. In contrast, small molecules targeting AS factors demonstrate more favorable in vivo distribution and pharmacokinetics, though they carry the risk of splicing off-target effects. Finally, the development of precise and highly specific molecular detection strategies for AS variants constitutes a critical avenue for future research.

Clinically, the successful application of AS-targeted therapies such as Nusinersen and Risdiplam in treating SMA provides strong proof of concept for the feasibility of precise AS modulation. These cases highlight that the accurate identification of AS variants, well-defined mechanisms of splice site regulation, and optimization of drug selectivity are key determinants in future drug development. In bone tumors, AS targets such as PKM and VEGF have shown therapeutic potential through ASO-mediated metabolic reprogramming and anti-angiogenesis. As an upstream regulator of AS factors, PRMT5 inhibition exerts synthetic lethality in tumor cells harboring splicing mutations and modulates the splicing of DNA repair regulators like TIP60 and SUV4-20H2, showing synergistic efficacy when combined with PARP inhibitors. Importantly, PRMT5 is also involved in regulating the differentiation of bone cell lineages; its inhibition enhances osteogenesis while suppressing osteoclast and chondrocyte differentiation, indicating its dual role in shaping the tumor microenvironment. Therefore, future research should focus on systematically identifying functional AS events in bone tumors, elucidating the crosstalk between splicing regulation and metabolic, immune, and DNA repair pathways, and developing ASO drugs with splicing specificity as well as tissue-selective AS factor inhibitors, to advance precision AS-targeted therapies in the clinical management of bone tumors.

Future research on AS in bone-related diseases could prioritize several critical directions. These include the comprehensive identification of functional AS events to clarify their specific roles in bone development, remodeling, and disease progression; the delineation of AS regulatory networks to uncover the hierarchical organization and pathway crosstalk mediated by key splicing factors; the development of splicing-specific and tissue-selective therapeutic strategies, including small-molecule inhibitors and ASOs; and the large-scale clinical validation of prognostically relevant AS variants to establish their utility in diagnosis, molecular subtyping, and risk stratification of bone tumors. Advancing these areas will not only deepen our mechanistic insights into AS-mediated regulation in bone biology and malignancies but also accelerate the clinical translation of AS-targeted therapies.

Acknowledgments

Funding: The study was financed by the National Key Research and Development Program of China (2022YFC2406100), the National Natural Science Foundation of China (Nos. 82503968, 82473221, 82272709, and 82072979), the Hubei Provincial Natural Science Foundation (No. 2025AFB240), the Natural Science Foundation of Hubei Province (No. ZRMS2022000165), and the Wuhan Yingcai and Science Foundation of Union Hospital (No. 2022xhyn023). L.C. was funded by the China Scholarship Council.

Author contributions: L.C., Y.H., M.A.R.-C., and K.J. contributed equally to the conception, drafting, and organization of the manuscript. E.C. and J.Y. performed grammar checking and language polishing of the manuscript. Z.S., B.A.V.T., and W.W. supervised the overall project, provided guidance on the manuscript structure, and finalized the content. All authors reviewed and approved the final version of the manuscript.

Competing interests: B.A.V.T. reports research grants from Polaris, Pfizer, Merck, Tracon Pharma, and GlaxoSmithKline; consulting fees from Bayer, Cytokinetics Inc., Deciphera Pharmaceuticals, Daiichi Sankyo, EcoR1, Advenchen, Putnam, Salarius Pharmaceuticals, Boxer Capital LLC, Acuta Capital Partners LLC, Aadi Biosciences, Race Oncology Limited, Hinge Bio, Inc., and Kronos Bio, Inc.; honoraria for educational talks from Iteration Therapeutics, Inc. and Total Health Conference; fees for participation in Data/Safety Advisory Boards for Apexigen, Daiichi Sankyo, Epizyme, Bayer, PTC Therapeutics, Aadi Biosciences, Boehringer Ingelheim, Agenus, Regeneron Pharmaceuticals, Advenchen, and Curtis; and speaking fees from Caris, Janssen, Lilly, Target Oncology, Bionest Partners, and Intellisphere LLC. All other authors declare that they have no competing interests.

References

1. Chow LT, Gelinas RE, Broker TR, Roberts RJ. An amazing sequence arrangement at the 5’ ends of adenovirus 2 messenger RNA. Cell. 1977;12(1):1–8.

2. Berget SM, Moore C, Sharp PA. Spliced segments at the 5’ terminus of adenovirus 2 late mRNA. Proc Natl Acad Sci USA. 1977;74(8):3171–3175.

3. Gomes R, Silva K, Theodoro RC. Group I introns: Structure, splicing and their applications in medical mycology. Genet Mol Biol. 2024;47(Suppl 1): Article e20230228.

4. Hale MA, Johnson NE, Berglund JA. Repeat-associated RNA structure and aberrant splicing. Biochim Biophys Acta Gene Regul Mech. 2019;1862(11–12): Article 194405.

5. Bradley RK, Anczuków O. RNA splicing dysregulation and the hallmarks of cancer. Nat Rev Cancer. 2023;23(3):135–155.

6. Fedorova O, Zingler N. Group II introns: Structure, folding and splicing mechanism. Biol Chem. 2007;388(7):665–678.

7. De Robertis EM, Black P, Nishikura K. Intranuclear location of the tRNA splicing enzymes. Cell. 1981;23(1):89–93.

8. Fica SM, Tuttle N, Novak T, Li NS, Lu J, Koodathingal P, Dai Q, Staley JP, Piccirilli JA. RNA catalyses nuclear pre-mRNA splicing. Nature. 2013;503(7475):229–234.

9. Jiang W, Chen L. Alternative splicing: Human disease and quantitative analysis from high-throughput sequencing. Comput Struct Biotechnol J. 2021;19:183–195.

10. Plaschka C, Newman AJ, Nagai K. Structural basis of nuclear pre-mRNA splicing: Lessons from yeast. Cold Spring Harb Perspect Biol. 2019;11(5): Article a032391.

11. Tao Y, Zhang Q, Wang H, Yang X, Mu H. Alternative splicing and related RNA binding proteins in human health and disease. Signal Transduct Target Ther. 2024;9(1):26.

12. Šromová V, Sobola D, Kaspar P. A brief review of bone cell function and importance. Cells. 2023;12(21):2576.

13. Dai Z, Sun Y, Maihemuti M, Jiang R. Genome-wide identification of alternative splicing and splicing regulated in immune infiltration in osteosarcoma patients. Front Genet. 2023;14:1051192.

14. Song X, Li X, Ge Y, Song J, Wei Q, He M, Wei M, Zhang Y, Chen T, Zhao L. Alternative splicing events and function in the tumor microenvironment: New opportunities and challenges. Int Immunopharmacol. 2023;123: Article 110718.

15. Marasco LE, Kornblihtt AR. The physiology of alternative splicing. Nat Rev Mol Cell Biol. 2023;24(4):242–254.

16. Zhang Y, Qian J, Gu C, Yang Y. Alternative splicing and cancer: A systematic review. Signal Transduct Target Ther. 2021;6(1):78.

17. Bessa C, Matos P, Jordan P, Gonçalves V. Alternative splicing: Expanding the landscape of cancer biomarkers and therapeutics. Int J Mol Sci. 2020;21(23):9032.

18. Li X, Liu S, Jiang J, Zhang L, Espinosa S, Hill RC, Hansen KC, Zhou ZH, Zhao R. CryoEM structure of Saccharomyces cerevisiae U1 snRNP offers insight into alternative splicing. Nat Commun. 2017;8(1):1035.

19. Bai R, Wan R, Yan C, Lei J, Shi Y. Structures of the fully assembled Saccharomyces cerevisiae spliceosome before activation. Science. 2018;360(6396):1423–1429.

20. Yeh FL, Chang SL, Ahmed GR, Liu HI, Tung L, Yeh CS, Lanier LS, Maeder C, Lin CM, Tsai SC, et al. Activation of Prp28 ATPase by phosphorylated Npl3 at a critical step of spliceosome remodeling. Nat Commun. 2021;12(1):3082.

21. Aísa-Marín I, García-Arroyo R, Mirra S, Marfany G. The Alter retina: Alternative splicing of retinal genes in health and disease. Int J Mol Sci. 2021;22(4):1855.

22. Dvinge H, Kim E, Abdel-Wahab O, Bradley RK. RNA splicing factors as oncoproteins and tumour suppressors. Nat Rev Cancer. 2016;16(7):413–430.

23. Chao Y, Jiang Y, Zhong M, Wei K, Hu C, Qin Y, Zuo Y, Yang L, Shen Z, Zou C. Regulatory roles and mechanisms of alternative RNA splicing in adipogenesis and human metabolic health. Cell Biosci. 2021;11(1):66.

24. Chen M, Manley JL. Mechanisms of alternative splicing regulation: Insights from molecular and genomics approaches. Nat Rev Mol Cell Biol. 2009;10(11):741–754.

25. Yan C, Wan R, Shi Y. Molecular mechanisms of pre-mRNA splicing through structural biology of the spliceosome. Cold Spring Harb Perspect Biol. 2019;11(1): Article a032409.

26. Zhan X, Lu Y, Shi Y. Molecular basis for the activation of human spliceosome. Nat Commun. 2024;15(1):6348.

27. Verta JP, Jacobs A. The role of alternative splicing in adaptation and evolution. Trends Ecol Evol. 2022;37(4):299–308.

28. Zhang XH, Leslie CS, Chasin LA. Dichotomous splicing signals in exon flanks. Genome Res. 2005;15(6):768–779.

29. Gao Y, Lin KT, Jiang T, Yang Y, Rahman MA, Gong S, Bai J, Wang L, Sun J, Sheng L, et al. Systematic characterization of short intronic splicing-regulatory elements in SMN2 pre-mRNA. Nucleic Acids Res. 2022;50(2):731–749.

30. Wang Y, Ma M, Xiao X, Wang Z. Intronic splicing enhancers, cognate splicing factors and context-dependent regulation rules. Nat Struct Mol Biol. 2012;19(10):1044–1052.

31. Rong S, Buerer L, Rhine CL, Wang J, Cygan KJ, Fairbrother WG. Mutational bias and the protein code shape the evolution of splicing enhancers. Nat Commun. 2020;11(1):2845.

32. Manley JL, Tacke R. SR proteins and splicing control. Genes Dev. 1996;10(13):1569–1579.

33. Cartegni L, Chew SL, Krainer AR. Listening to silence and understanding nonsense: Exonic mutations that affect splicing. Nat Rev Genet. 2002;3(4):285–298.

34. Plawski A, Banasiewicz T, Borun P, Kubaszewski L, Krokowicz P, Skrzypczak-Zielinska M, Lubinski J. Familial adenomatous polyposis of the colon. Hered Cancer Clin Pract. 2013;11(1):15.

35. Wang E, Dimova N, Cambi F. PLP/DM20 ratio is regulated by hnRNPH and F and a novel G-rich enhancer in oligodendrocytes. Nucleic Acids Res. 2007;35(12):4164–4178.

36. Pagani F, Buratti E, Stuani C, Bendix R, Dörk T, Baralle FE. A new type of mutation causes a splicing defect in ATM. Nat Genet. 2002;30(4):426–429.

37. Pastor T, Pagani F. Interaction of hnRNPA1/A2 and DAZAP1 with an Alu-derived intronic splicing enhancer regulates ATM aberrant splicing. PLOS ONE. 2011;6(8): Article e23349.

38. Yabas M, Elliott H, Hoyne GF. The role of alternative splicing in the control of immune homeostasis and cellular differentiation. Int J Mol Sci. 2015;17(1):3.

39. Tong A, Nguyen J, Lynch KW. Differential expression of CD45 isoforms is controlled by the combined activity of basal and inducible splicing-regulatory elements in each of the variable exons. J Biol Chem. 2005;280(46):38297–38304.

40. Rothrock CR, House AE, Lynch KW. HnRNP L represses exon splicing via a regulated exonic splicing silencer. EMBO J. 2005;24(15):2792–2802.

41. Jez J. Encyclopedia of biological chemistry: Third edition . Amsterdam (Netherlands): Elsevier; 2021; 1-6:1-4822.

42. Kornblihtt AR, Schor IE, Alló M, Dujardin G, Petrillo E, Muñoz MJ. Alternative splicing: A pivotal step between eukaryotic transcription and translation. Nat Rev Mol Cell Biol. 2013;14(3):153–165.

43. Wang Y, Wang Z. Systematical identification of splicing regulatory cis-elements and cognate trans-factors. Methods. 2014;65(3):350–358.

44. Shepard PJ, Hertel KJ. The SR protein family. Genome Biol. 2009;10(10):242.

45. Slišković I, Eich H, Müller-McNicoll M. Exploring the multifunctionality of SR proteins. Biochem Soc Trans. 2022;50(1):187–198.

46. Bei M, Xu J. SR proteins in cancer: Function, regulation, and small inhibitor. Cell Mol Biol Lett. 2024;29(1):78.

47. Ibrahim EC, Schaal TD, Hertel KJ, Reed R, Maniatis T. Serine/arginine-rich protein-dependent suppression of exon skipping by exonic splicing enhancers. Proc Natl Acad Sci USA. 2005;102(14):5002–5007.

48. Kanopka A, Mühlemann O, Akusjärvi G. Inhibition by SR proteins of splicing of a regulated adenovirus pre-mRNA. Nature. 1996;381(6582):535–538.

49. Fu XD, Ares M Jr. Context-dependent control of alternative splicing by RNA-binding proteins. Nat Rev Genet. 2014;15(10):689–701.

50. Carpenter B, MacKay C, Alnabulsi A, MacKay M, Telfer C, Melvin WT, Murray GI. The roles of heterogeneous nuclear ribonucleoproteins in tumour development and progression. Biochim Biophys Acta. 2006;1765(2):85–100.

51. Heiner M, Hui J, Schreiner S, Hung LH, Bindereif A. HnRNP L-mediated regulation of mammalian alternative splicing by interference with splice site recognition. RNA Biol. 2010;7(1):56–64.

52. Rooke N, Markovtsov V, Cagavi E, Black DL. Roles for SR proteins and hnRNP A1 in the regulation of c-src exon N1. Mol Cell Biol. 2003;23(6):1874–1884.

53. Pan D, Long L, Li C, Zhou Y, Liu Q, Zhao Z, Zhao H, Lin W, Zheng Z, Peng L, et al. Splicing factor hnRNPA1 regulates alternative splicing of LOXL2 to enhance the production of LOXL2Δ13. J Biol Chem. 2024;300(7): Article 107414.

54. Agarwal A, Bahadur RP. Modular architecture and functional annotation of human RNA-binding proteins containing RNA recognition motif. Biochimie. 2023;209:116–130.

55. Soni K, Jagtap PKA, Martínez-Lumbreras S, Bonnal S, Geerlof A, Stehle R, Simon B, Valcarcel J, Sattler M. Structural basis for specific RNA recognition by the alternative splicing factor RBM5. Nat Commun. 2023;14(1):4233.

56. Gosztyla ML, Zhan L, Olson S, Wei X, Naritomi J, Nguyen G, Street L, Goda GA, Cavazos FF, Schmok JC, et al. Integrated multi-omics analysis of zinc-finger proteins uncovers roles in RNA regulation. Mol Cell. 2024;84(19):3826–42.e8.

57. Han H, Braunschweig U, Gonatopoulos-Pournatzis T, Weatheritt RJ, Hirsch CL, Ha KCH, Radovani E, Nabeel-Shah S, Sterne-Weiler T, Wang J, et al. Multilayered control of alternative splicing regulatory networks by transcription factors. Mol Cell. 2017;65(3):539–53.e7.

58. Nazarov IB, Bakhmet EI, Tomilin AN. KH-domain poly(C)-binding proteins as versatile regulators of multiple biological processes. Biochemistry. 2019;84(3):205–219.

59. Dirckx N, Moorer MC, Clemens TL, Riddle RC. The role of osteoblasts in energy homeostasis. Nat Rev Endocrinol. 2019;15(11):651–665.

60. Bensreti H, Alhamad DW, Gonzalez AM, Pizarro-Mondesir M, Bollag WB, Isales CM, McGee-Lawrence ME. Update on the role of glucocorticoid signaling in osteoblasts and bone marrow adipocytes during aging. Curr Osteoporos Rep. 2023;21(1):32–44.

61. Krakowian D, Lesiak M, Auguściak-Duma A, Witecka J, Kusz D, Sieroń AL, Gawron K. Analysis of the TID-I and TID-L splice variants’ expression profile under in vitro differentiation of human mesenchymal bone marrow cells into osteoblasts. Cells. 2024;13(12):1021.

62. Wang X, Jia Q, Yu L, Huang J, Wang X, Zhou L, Mijiti W, Dong S, Xie Z, Ma H. Filamin B knockdown impairs differentiation and function in mouse pre-osteoblasts via aberrant transcription and alternative splicing. Heliyon. 2024;10(20): Article e39334.

63. Lin YM, Wu CC, Chang YC, Wu CH, Ho HL, Hu JW, Chang RC, Wang CT, Ouyang P. Target disruption of ribosomal protein pNO40 accelerates aging and impairs osteogenic differentiation of mesenchymal stem cells. Biochem Biophys Res Commun. 2016;469(4):903–910.

64. Kazantseva J, Kivil A, Tints K, Kazantseva A, Neuman T, Palm K. Alternative splicing targeting the hTAF4-TAFH domain of TAF4 represses proliferation and accelerates chondrogenic differentiation of human mesenchymal stem cells. PLOS ONE. 2013;8(10): Article e74799.

65. Tian C, Huang Y, Li Q, Feng Z, Xu Q. Mettl3 regulates osteogenic differentiation and alternative splicing of Vegfa in bone marrow mesenchymal stem cells. Int J Mol Sci. 2019;20(3):551.

66. Peng S, He T, Liu Y, Zheng L, Zhong Y, Niu Z, Zhang M, Yang S. Lnc-PPP2R1B mediates the alternative splicing of PPP2R1B by interacting and stabilizing HNRNPLL and promotes osteogenesis of MSCs. Stem Cell Rev Rep. 2023;19(6):1981–1993.

67. Jiang Q, Qin X, Kawane T, Komori H, Matsuo Y, Taniuchi I, Ito K, Izumi SI, Komori T. Cbfb2 isoform dominates more potent Cbfb1 and is required for skeletal development. J Bone Miner Res. 2016;31(7):1391–1404.

68. Khayal LA, Grünhagen J, Provazník I, Mundlos S, Kornak U, Robinson PN, Ott C-E. Transcriptional profiling of murine osteoblast differentiation based on RNA-seq expression analyses. Bone. 2018;113:29–40.

69. Le MH, Warotayanont R, Stahl J, Den Besten PK, Nakano Y. Amelogenin Exon4 forms a novel miRNA that directs ameloblast and osteoblast differentiation. J Dent Res. 2016;95(4):423–429.

70. Matheny RW Jr, Nindl BC, Adamo ML. Minireview: Mechano-growth factor: A putative product of IGF-I gene expression involved in tissue repair and regeneration. Endocrinology. 2010;151(3):865–875.

71. Hameed M, Orrell RW, Cobbold M, Goldspink G, Harridge SD. Expression of IGF-I splice variants in young and old human skeletal muscle after high resistance exercise. J Physiol. 2003;547(Pt 1):247–254.

72. Yi Q, Liu H, Feng J, Wu Y, Sun W, Ou M, Tang L. Splicing factor-modulated generation of mechano growth factor regulates physiological processes in osteoblasts under mechanical stimuli. Cell Adhes Migr. 2019;13(1):322–331.

73. Feng J, Xu X, Fan X, Yi Q, Tang L. BAF57/SMARCE1 interacting with splicing factor SRSF1 regulates mechanical stress-induced alternative splicing of cyclin D1. Genes. 2021;12(2):306.

74. Lian JB, Stein GS, Javed A, Wijnen AJ, Stein JL, Montecino M, Hassan MQ, Gaur T, Lengner CJ, Young DW. Networks and hubs for the transcriptional control of osteoblastogenesis. Rev Endocr Metab Disord. 2006;7(1-2):1–16.

75. Stein GS, Lian JB, Stein JL, Van Wijnen AJ, Montecino M. Transcriptional control of osteoblast growth and differentiation. Physiol Rev. 1996;76(2):593–629.

76. Almehmadi A, Ohyama Y, Kaku M, Alamoudi A, Husein D, Katafuchi M, Mishina Y, Mochida Y. VWC2 increases bone formation through inhibiting Activin signaling. Calcif Tissue Int. 2018;103(6):663–674.

77. Ohyama Y, Katafuchi M, Almehmadi A, Venkitapathi S, Jaha H, Ehrenman J, Morcos J, Alijamaan R, Mochida Y. Modulation of matrix mineralization by Vwc2-like protein and its novel splicing isoforms. Biochem Biophys Res Commun. 2012;418(1):12–16.

78. Arasaki Y, Hayata T. The RNA-binding protein Cpeb4 regulates splicing of the Id2 gene in osteoclast differentiation. J Cell Physiol. 2024;239(4): Article e31197.

79. Zhang Y, Yu X, Sun R, Min J, Tang X, Lin Z, Xie S, Li X, Lu S, Tian Z, et al. Splicing factor arginine/serine-rich 8 promotes multiple myeloma malignancy and bone lesion through alternative splicing of CACYBP and exosome-based cellular communication. Clin Transl Med. 2022;12(2): Article e684.

80. Mukai S, Kitazawa R, Ishii J, Kondo T, Hakozaki A, Horiuchi K, Haraguch R, Mori K, Kitazawa S. Identification and analysis of function of a novel splicing variant of mouse receptor activator of NF-κB. Mol Cell Biochem. 2011;350(1-2):29–38.

81. Xu X, Shobuike T, Shiraki M, Kamohara A, Hirata H, Murayama M, Mawatari D, Ueno M, Morimoto T, Kukita T, et al. Leukemia/lymphoma-related factor (LRF) or osteoclast zinc finger protein (OCZF) overexpression promotes osteoclast survival by increasing Bcl-xl mRNA: A novel regulatory mechanism mediated by the RNA binding protein SAM68. Lab Invest. 2022;102(9):1000–1010.

82. Takegahara N, Kim H, Choi Y. Unraveling the intricacies of osteoclast differentiation and maturation: Insight into novel therapeutic strategies for bone-destructive diseases. Exp Mol Med. 2024;56(2):264–272.

83. Beaudreuil J, Balasubramanian S, Chenais J, Taboulet J, Frenkian M, Orcel P, Jullienne A, Horne WC, Vernejoul MC, Cressent M. Molecular characterization of two novel isoforms of the human calcitonin receptor. Gene. 2004;343(1):143–151.

84. Anusaksathien O, Laplace C, Li X, Ren Y, Peng L, Goldring SR, Galson DL. Tissue-specific and ubiquitous promoters direct the expression of alternatively spliced transcripts from the calcitonin receptor gene. J Biol Chem. 2001;276(25):22663–22674.

85. Guise TA, Yin JJ, Thomas RJ, Dallas M, Cui Y, Gillespie MT. Parathyroid hormone-related protein (PTHrP)-(1-139) isoform is efficiently secreted in vitro and enhances breast cancer metastasis to bone in vivo. Bone. 2002;30(5):670–676.

86. Amoui M, Baylink DJ, Tillman JB, Lau KH. Expression of a structurally unique osteoclastic protein-tyrosine phosphatase is driven by an alternative intronic, cell type-specific promoter. J Biol Chem. 2003;278(45):44273–44280.

87. Thijssen VL. Galectins in endothelial cell biology and angiogenesis: The basics. Biomolecules. 2021;11(9):1386.

88. Takeuchi T, Oyama M, Tamura M, Arata Y, Hatanaka T. Reduced form of Galectin-1 suppresses osteoclastic differentiation of human peripheral blood mononuclear cells and murine RAW264 cells in vitro. Biomolecules. 2024;14(1):121.

89. Roy M, Mbous Nguimbus L, Badiane PY, Goguen-Couture V, Degrandmaison J, Parent JL, Brunet MA, Roux S. Galectin-8 modulates human osteoclast activity partly through isoform-specific interactions. Life Sci Alliance. 2024;7(5): Article e202302348.

90. Troncoso MF, Ferragut F, Bacigalupo ML, Cárdenas Delgado VM, Nugnes LG, Gentilini L, Laderach D, Wolfenstein-Todel C, Compagno D, Rabinovich GA, et al. Galectin-8: A matricellular lectin with key roles in angiogenesis. Glycobiology. 2014;24(10):907–914.

91. Pettenuzzo S, Arduino A, Belluzzi E, Pozzuoli A, Fontanella CG, Ruggieri P, Salomoni V, Majorana C, Berardo A. Biomechanics of chondrocytes and chondrons in healthy conditions and osteoarthritis: A review of the mechanical characterisations at the microscale. Biomedicine. 2023;11(7):1942.

92. Chen H, Tan XN, Hu S, Liu RQ, Peng LH, Li YM, Wu P. Molecular mechanisms of chondrocyte proliferation and differentiation. Front Cell Dev Biol. 2021;9: Article 664168.

93. Yanagita T, Kubota S, Kawaki H, Kawata K, Kondo S, Takano-Yamamoto T, Tanaka S, Takigawa M. Expression and physiological role of CCN4/Wnt-induced secreted protein 1 mRNA splicing variants in chondrocytes. FEBS J. 2007;274(7):1655–1665.

94. Friedman B, Larranaga-Vera A, Castro CM, Corciulo C, Rabbani P, Cronstein BN. Adenosine A2A receptor activation reduces chondrocyte senescence. FASEB J. 2023;37(4): Article e22838.

95. Mondal AM, Zhou H, Horikawa I, Suprynowicz FA, Li G, Dakic A, Rosenhal B, Ye L, Harris CC, Schlegel R, et al. Δ133p53α, a natural p53 isoform, contributes to conditional reprogramming and long-term proliferation of primary epithelial cells. Cell Death Dis. 2018;9(7):750.

96. Aoubala M, Murray-Zmijewski F, Khoury MP, Fernandes K, Perrier S, Bernard H, Prats AC, Lane DP, Bourdon JC. p53 directly transactivates Δ133p53α, regulating cell fate outcome in response to DNA damage. Cell Death Differ. 2011;18(2):248–258.

97. Shi S, Kelly BJ, Wang C, Klingler K, Chan A, Eckert GJ, Trippel SB. Human IGF-I propeptide a promotes articular chondrocyte biosynthesis and employs glycosylation-dependent heparin binding. Biochim Biophys Acta Gen Subj. 2018;1862(3):567–575.

98. Brisson BK, Barton ER. New modulators for IGF-I activity within IGF-I processing products. Front Endocrinol. 2013;4:42.

99. DuRaine GD, Chan SM, Reddi AH. Effects of TGF-β1 on alternative splicing of superficial zone protein in articular cartilage cultures. Osteoarthr Cartil. 2011;19(1):103–110.

100. Liu L, Hong Y, Ma C, Zhang F, Li Q, Li B, He H, Zhu J, Wang H, Chen L. Circular RNA Gtdc1 protects against offspring osteoarthritis induced by prenatal prednisone exposure by regulating SRSF1-Fn1 signaling. Adv Sci. 2024;11(20):e2307442.

101. McAlinden A. Alternative splicing of type II procollagen: IIB or not IIB? Connect Tissue Res. 2014;55(3):165–176.

102. Liu X, Dong H, Gong Y, Wang L, Zhang R, Zheng T, Zheng Y, Shen S, Zheng C, Tian M, et al. A novel missense mutation of COL2A1 gene in a large family with stickler syndrome type I. J Cell Mol Med. 2022;26(5):1530–1539.

103. McAlinden A, Johnstone B, Kollar J, Kazmi N, Hering TM. Expression of two novel alternatively spliced COL2A1 isoforms during chondrocyte differentiation. Matrix Biol. 2008;27(3):254–266.

104. Hering TM, Wirthlin L, Ravindran S, McAlinden A. Changes in type II procollagen isoform expression during chondrogenesis by disruption of an alternative 5’ splice site within Col2a1 exon 2. Matrix Biol. 2014;36:51–63.

105. Han F, Gilbert JR, Harrison G, Adams CS, Freeman T, Tao Z, Zaka R, Liang H, Williams C, Tuan RS, et al. Transforming growth factor-beta1 regulates fibronectin isoform expression and splicing factor SRp40 expression during ATDC5 chondrogenic maturation. Exp Cell Res. 2007;313(8):1518–1532.

106. Liu Q, Han M, Wu Z, Fu W, Ji J, Liang Q, Tan M, Zhai L, Gao J, Shi D, et al. DDX5 inhibits hyaline cartilage fibrosis and degradation in osteoarthritis via alternative splicing and G-quadruplex unwinding. Nat Aging. 2024;4(5):664–680.

107. Hoolwerff M, Tuerlings M, Wijnen IJL, Suchiman HED, Cats D, Mei H, Nelissen RGHH, Zwaag HMJ, Ramos YFM, Almeida RC, et al. Identification and functional characterization of imbalanced osteoarthritis-associated fibronectin splice variants. Rheumatology (Oxford). 2023;62(2):894–904.

108. Lin X, Patil S, Gao YG, Qian A. The bone extracellular matrix in bone formation and regeneration. Front Pharmacol. 2020;11:757.

109. Wang X, Chen S, Nan H, Liu R, Ding Y, Song K, Shuai J, Fan Q, Zheng Y, Ye F, et al. Abnormal aggregation of invasive cancer cells induced by collective polarization and ECM-mediated mechanical coupling in coculture systems. Research. 2021;2021:9893131.

110. Ma C, Du T, Niu X, Fan Y. Biomechanics and mechanobiology of the bone matrix. Bone Res. 2022;10(1):59.

111. Deng H, Huang X, Yuan L. Molecular genetics of the COL2A1-related disorders. Mutat Res Rev Mutat Res. 2016;768:1–13.

112. Donoso LA, Edwards AO, Frost AT, Ritter R III, Ahmad N, Vrabec T, Rogers J, Meyer D, Parma S. Clinical variability of stickler syndrome: Role of exon 2 of the collagen COL2A1 gene. Surv Ophthalmol. 2003;48(2):191–203.

113. Aubert-Foucher E, Mayer N, Pasdeloup M, Pagnon A, Hartmann D, Mallein-Gerin F. A unique tool to selectively detect the chondrogenic IIB form of human type II procollagen protein. Matrix Biol. 2014;34:80–88.

114. Lewis R, Ravindran S, Wirthlin L, Traeger G, Fernandes RJ, McAlinden A. Disruption of the developmentally-regulated Col2a1 pre-mRNA alternative splicing switch in a transgenic knock-in mouse model. Matrix Biol. 2012;31(3):214–226.

115. Patra D, DeLassus E, McAlinden A, Sandell LJ. Characterization of a murine type IIB procollagen-specific antibody. Matrix Biol. 2014;34:154–160.

116. McAlinden A, Shim KH, Wirthlin L, Ravindran S, Hering TM. Quantification of type II procollagen splice forms using alternative transcript-qPCR (AT-qPCR). Matrix Biol. 2012;31(7–8):412–420.

117. Klatt AR, Becker AK, Neacsu CD, Paulsson M, Wagener R. The matrilins: Modulators of extracellular matrix assembly. Int J Biochem Cell Biol. 2011;43(3):320–330.

118. Li L, Zhang L, Shao Y, Wang G, Gong R, Wang Z, Peng J, Wang S, Genochio D, Zhao B, et al. Distinct roles of two alternative splice variants of matrilin-2 in protein oligomerization and proteolysis. Mol Med Rep. 2012;6(5):1204–1210.

119. Bolamperti S, Villa I, Rubinacci A. Bone remodeling: An operational process ensuring survival and bone mechanical competence. Bone Res. 2022;10(1):48.

120. Xiao H, Li W, Qin Y, Lin Z, Qian C, Wu M, Xia Y, Bai J, Geng D. Crosstalk between lipid metabolism and bone homeostasis: Exploring intricate signaling relationships. Research. 2024;7:0447.

121. Infante M, Fabi A, Cognetti F, Gorini S, Caprio M, Fabbri A. RANKL/RANK/OPG system beyond bone remodeling: Involvement in breast cancer and clinical perspectives. J Exp Clin Cancer Res. 2019;38(1):12.

122. Tobeiha M, Moghadasian MH, Amin N, Jafarnejad S. RANKL/RANK/OPG pathway: A mechanism involved in exercise-induced bone remodeling. Biomed Res Int. 2020;2020:6910312.

123. Boyce BF, Xing L. Functions of RANKL/RANK/OPG in bone modeling and remodeling. Arch Biochem Biophys. 2008;473(2):139–146.

124. Heckt T, Keller J, Peters S, Streichert T, Chalaris A, Rose-John S, Mell B, Joe B, Aming M, Schinke T. Parathyroid hormone induces expression and proteolytic processing of Rankl in primary murine osteoblasts. Bone. 2016;92:85–93.

125. Takegahara N, Kim H, Choi Y. RANKL biology. Bone. 2022;159: Article 116353.

126. Wright HL, McCarthy HS, Middleton J, Marshall MJ. RANK, RANKL and osteoprotegerin in bone biology and disease. Curr Rev Musculoskelet Med. 2009;2(1):56–64.

127. Ikeda T, Kasai M, Utsuyama M, Hirokawa K. Determination of three isoforms of the receptor activator of nuclear factor-kappaB ligand and their differential expression in bone and thymus. Endocrinology. 2001;142(4):1419–1426.

128. Hikita A, Kadono Y, Chikuda H, Fukuda A, Wakeyama H, Yasuda H, Nakamura K, Oda H, Miyazaki T, Tanaka S. Identification of an alternatively spliced variant of Ca2 +- promoted Ras inactivator as a possible regulator of RANKL shedding. J Biol Chem. 2005;280(50):41700–41706.

129. Papanastasiou AD, Sirinian C, Kalofonos HP. Identification of novel human receptor activator of nuclear factor-kB isoforms generated through alternative splicing: Implications in breast cancer cell survival and migration. Breast Cancer Res. 2012;14(4):R112.

130. Sirinian C, Papanastasiou AD, Zarkadis IK, Kalofonos HP. Alternative splicing generates a truncated isoform of human TNFRSF11A (RANK) with an altered capacity to activate NF-κB. Gene. 2013;525(1):124–129.

131. Kuebart T, Oezel L, Gürsoy B, Maus U, Windolf J, Bittersohl B, Grotheer V. Periostin splice variant expression in human osteoblasts from osteoporotic patients and its effects on interleukin-6 and osteoprotegerin. Int J Mol Sci. 2025;26(3):932.

132. Brinegar AE, Xia Z, Loehr JA, Li W, Rodney GG, Cooper TA. Extensive alternative splicing transitions during postnatal skeletal muscle development are required for calcium handling functions. eLife. 2017;6: Article e27192.

133. Al-Shanti N, Stewart CE. Ca2+/calmodulin-dependent transcriptional pathways: Potential mediators of skeletal muscle growth and development. Biol Rev Camb Philos Soc. 2009;84(4):637–652.

134. Blair HC, Schlesinger PH, Huang CL, Zaidi M. Calcium signalling and calcium transport in bone disease. Subcell Biochem. 2007;45:539–562.

135. Hao Y, Yang N, Sun M, Yang S, Chen X. The role of calcium channels in osteoporosis and their therapeutic potential. Front Endocrinol. 2024;15:1450328.

136. Pitt GS, Matsui M, Cao C. Voltage-gated calcium channels in nonexcitable tissues. Annu Rev Physiol. 2021;83:183–203.

137. Striessnig J, Nakao A, Mori Y. Voltage-gated Ca2+ channels. Lessons from knockout and knock-in mice. In: Zamponi GW, Weiss N, editors. Voltage-gated calcium channels . Cham: Springer International Publishing; 2022. p. 255–357.

138. Splawski I, Timothy KW, Sharpe LM, Decher N, Kumar P, Bloise R, Napolitano C, Schwartz PJ, Joseph RM, Candouris K, et al. Ca V 1.2 calcium channel dysfunction causes a multisystem disorder including arrhythmia and autism. Cell. 2004;119(1):19–31.

139. Illes P, Müller CE, Jacobson KA, Grutter T, Nicke A, Fountain SJ, Kennedy C, Schmaizing G, Jarvis MJ, Stojikovic SS, et al. Update of P2X receptor properties and their pharmacology: IUPHAR review 30. Br J Pharmacol. 2021;178(3):489–514.

140. Kotnis S, Bingham B, Vasilyev DV, Miller SW, Bai Y, Yeola S, Chanda PK, Bowlby MR, Kaftan EJ, Samad TA, et al. Genetic and functional analysis of human P2X5 reveals a distinct pattern of exon 10 polymorphism with predominant expression of the nonfunctional receptor isoform. Mol Pharmacol. 2010;77(6):953–960.

141. Duckwitz W, Hausmann R, Aschrafi A, Schmalzing G. P2X5 subunit assembly requires scaffolding by the second transmembrane domain and a conserved aspartate. J Biol Chem. 2006;281(51):39561–39572.

142. Abramowski P, Ogrodowczyk C, Martin R, Pongs O. A truncation variant of the cation channel P2RX5 is upregulated during T cell activation. PLOS ONE. 2014;9(9): Article e104692.

143. Qin L, He T, Chen S, Yang D, Yi W, Cao H, Xiao G. Roles of mechanosensitive channel Piezo1/2 proteins in skeleton and other tissues. Bone Res. 2021;9(1):44.

144. Szczot M, Pogorzala LA, Solinski HJ, Young L, Yee P, Le Pichon CE, Chesler AT, Hoon MA. Cell-type-specific splicing of Piezo2 regulates mechanotransduction. Cell Rep. 2017;21(10):2760–2771.

145. Geng J, Liu W, Zhou H, Zhang T, Wang L, Zhang M, Li Y, Shen B, Li X, Xiao B. A plug-and-latch mechanism for gating the mechanosensitive piezo channel. Neuron. 2020;106(3):438–451.e6.

146. Liu Y, Ferguson JF, Xue C, Silverman IM, Gregory B, Reilly MP, Li M. Evaluating the impact of sequencing depth on transcriptome profiling in human adipose. PLOS ONE. 2013;8(6): Article e66883.

147. Faraldi M, Mangiavini L, Conte C, Banfi G, Napoli N, Lombardi G. A novel methodological approach to simultaneously extract high-quality total RNA and proteins from cortical and trabecular bone. Open Biol. 2022;12(5): Article 210387.

148. Liu F, Ding Y, Xu Z, Hao X, Pan T, Miles G, Wang S, Wu YH, Liu J, Bado IL, et al. Single-cell profiling of bone metastasis ecosystems from multiple cancer types reveals convergent and divergent mechanisms of bone colonization. Cell Genom. 2025;5(7): Article 100888.

149. Bandyopadhyay S, Duffy MP, Ahn KJ, Sussman JH, Pang M, Smith D, Duncan G, Zhang I, Huang I, Huang J, et al. Mapping the cellular biogeography of human bone marrow niches using single-cell transcriptomics and proteomic imaging. Cell. 2024;187(12):3120–3140.e29.

150. Lin W, Li Y, Qiu C, Zou B, Gong Y, Zhang X, Tian D, Sherman W, Sanchez F, Wu D, et al. Mapping the spatial atlas of the human bone tissue integrating spatial and single-cell transcriptomics. Nucleic Acids Res. 2025;53(2): Article gkae1298.

151. Yang W, Wang Y, Mo K, Chen W, Xie X. Single-cell RNA sequencing reveals multiple immune cell subpopulations promote the formation of abnormal bone microenvironment in osteoporosis. Sci Rep. 2024;14(1):29493.

152. Wang H, He X, Ma M, Dou T, Wei Y, Rux D, Qin L, Yang Y, Zhu Y, Yao L. Integrating spatial and single-cell transcriptomics to characterize mouse long bone fracture healing process. Commun Biol. 2025;8(1):887.

153. Tsukasaki M, Huynh NC, Okamoto K, Muro R, Terashima A, Kurikawa Y, Komatsu N, Pluemsakunthai W, Nitta T, Abe T, et al. Stepwise cell fate decision pathways during osteoclastogenesis at single-cell resolution. Nat Metab. 2020;2(12):1382–1390.

154. Lin Z, Chen Z. Single-cell RNA sequencing reveals key signaling pathways and biological functions in giant cell tumors of bone. Discov Oncol. 2025;16(1):573.

155. Beird HC, Bielack SS, Flanagan AM, Gill J, Heymann D, Janeway KA, Livingston JA, Roberts RD, Strauss SJ, Gorlick R. Osteosarcoma. Nat Rev Dis Primers. 2022;8(1):77.

156. Rothzerg E, Ho XD, Xu J, Wood D, Märtson A, Maasalu K, Koks S. Alternative splicing of leptin receptor overlapping transcript in osteosarcoma. Exp Biol Med. 2020;245(16):1437–1443.

157. Li W, Wang L, Tian W, Ji W, Bing D, Wang Y, Xu B, Feng J, Zhang P, Liang H, et al. SNRNP70 regulates the splicing of CD55 to promote osteosarcoma progression. JCI Insight. 2024;9(24): Article e185269.

158. Lee JH, Horak CE, Khanna C, Meng Z, Yu LR, Veenstra TD, Steeg PS. Alterations in Gemin5 expression contribute to alternative mRNA splicing patterns and tumor cell motility. Cancer Res. 2008;68(3):639–644.

159. Li S, Huang X, Zheng S, Zhang W, Liu F, Cao Q. High expression of SRSF1 facilitates osteosarcoma progression and unveils its potential mechanisms. BMC Cancer. 2024;24(1):580.

160. Fu Y, Huang B, Shi Z, Han J, Wang Y, Huangfu J, Wu W. SRSF1 and SRSF9 RNA binding proteins promote Wnt signalling-mediated tumorigenesis by enhancing β-catenin biosynthesis. EMBO Mol Med. 2013;5(5):737–750.

161. Ajiro M, Jia R, Yang Y, Zhu J, Zheng ZM. A genome landscape of SRSF3-regulated splicing events and gene expression in human osteosarcoma U2OS cells. Nucleic Acids Res. 2016;44(4):1854–1870.

162. Peach CJ, Mignone VW, Arruda MA, Alcobia DC, Hill SJ, Kilpatrick LE, Woolard J. Molecular pharmacology of VEGF-A isoforms: Binding and signalling at VEGFR2. Int J Mol Sci. 2018;19(4):1264.

163. Boudria A, Abou Faycal C, Jia T, Gout S, Keramidas M, Didier C, Lemaitre N, Manet S, Coll J-L, Toffart A-C, et al. VEGF 165 b, a splice variant of VEGF-A, promotes lung tumor progression and escape from anti-angiogenic therapies through a β1 integrin/VEGFR autocrine loop. Oncogene. 2019;38(7):1050–1066.

164. Bates DO, Cui TG, Doughty JM, Winkler M, Sugiono M, Shields JD, Peat D, Gillatt D, Harper SJ. VEGF 165 b, an inhibitory splice variant of vascular endothelial growth factor, is down-regulated in renal cell carcinoma. Cancer Res. 2002;62(14):4123–4131.

165. Quan B, Li Z, Yang H, Li S, Yan X, Wang Y. The splicing factor YBX1 promotes the progression of osteosarcoma by upregulating VEGF 165 and downregulating VEGF 165 b. Heliyon. 2023;9(8): Article e18706.

166. Cao L, Jia K, Van Tine BA, Yu Y, Peng Y, Chen X, Pan Q, Yang W, Zhang Z, Shao Z, et al. KPNA2 promotes osteosarcoma progression by regulating the alternative splicing of DDX3X mediated by YBX1. Oncogene. 2025;44(26):2186–2200.

167. Xu Y, Zhang Y, Luo Y, Qiu G, Lu J, He M, Wang Y. Novel insights into the METTL3-METTL14 complex in musculoskeletal diseases. Cell Death Discov. 2023;9(1):170.

168. Kim Y, Shin S, Kwon S, Moon K, Baek SV, Jo A, Kim HS, Hwang GH, Bae S, Kim YH, et al. METTL3 regulates alternative splicing of cell cycle-related genes via crosstalk between mRNA m 6 a modifications and splicing factors. Am J Cancer Res. 2023;13(4):1443–1456.

169. Zhu ZM, Huo FC, Zhang J, Shan HJ, Pei DS. Crosstalk between m6A modification and alternative splicing during cancer progression. Clin Transl Med. 2023;13(10): Article e1460.

170. Li HB, Huang G, Tu J, Lv DM, Jin QL, Chen JK, Zou YT, Lee DF, Shen JN, Xie XB. METTL14-mediated epitranscriptome modification of MN1 mRNA promote tumorigenicity and all-trans-retinoic acid resistance in osteosarcoma. EBioMedicine. 2022;82: Article 104142.

171. Chen K, Li C, Huang S, Chen Y, Zhu X. LncRNA KASRT serves as a potential treatment target by regulating SRSF1-related KLF6 alternative splicing and the P21/CCND1 pathway in osteosarcoma: An in vitro and in vivo study. Front Oncol. 2021;11: Article 700963.

172. Chen H, Chen L, Sun L, Zhen H, Li X, Zhang Q. A small interfering RNA targeting the KLF6 splice variant, KLF6-SV1, as gene therapy for gastric cancer. Gastric Cancer. 2011;14(4):339–352.

173. Tahimic CG, Wang Y, Bikle DD. Anabolic effects of IGF-1 signaling on the skeleton. Front Endocrinol. 2013;4:6.

174. Kasprzak A, Szaflarski W. Role of alternatively spliced messenger RNA (mRNA) isoforms of the insulin-like growth factor 1 (IGF1) in selected human tumors. Int J Mol Sci. 2020;21(19):6995.

175. Philippou A, Armakolas A, Panteleakou Z, Pissimissis N, Nezos A, Theos A, Kaparelou M, Armakolas N, Pneumaticos SG, Koutsilieris M. IGF1Ec expression in MG-63 human osteoblast-like osteosarcoma cells. Anticancer Res. 2011;31(12):4259–4265.

176. Hainaut P, Pfeifer GP. Somatic TP53 mutations in the era of genome sequencing. Cold Spring Harb Perspect Med. 2016;6(11): Article a026179.

177. Ribi S, Baumhoer D, Lee K, Edison TAS, Madan B, Zhang K, Kohlmann WK, Yao F, Lee WH, et al. TP53 intron 1 hotspot rearrangements are specific to sporadic osteosarcoma and can cause Li-Fraumeni syndrome. Oncotarget. 2015;6(10):7727–7740.

178. Chen X, Bahrami A, Pappo A, Easton J, Dalton J, Hedlund E, Ellison D, Shurtleff S, Wu G, Wei L, et al. Recurrent somatic structural variations contribute to tumorigenesis in pediatric osteosarcoma. Cell Rep. 2014;7(1):104–112.

179. Marcel V, Dichtel-Danjoy ML, Sagne C, Hafsi H, Ma D, Ortiz-Cuaran S, Olivier M, Hall J, Mollereau B, Hainaut P, et al. Biological functions of p53 isoforms through evolution: Lessons from animal and cellular models. Cell Death Differ. 2011;18(12):1815–1824.

180. Filippov V, Filippova M, Duerksen-Hughes PJ. The early response to DNA damage can lead to activation of alternative splicing activity resulting in CD44 splice pattern changes. Cancer Res. 2007;67(16):7621–7630.

181. Grawenda AM, Møller EK, Lam S, Repapi E, Teunisse AF, Alnæs GI, Borresen-Dale AL, Kristensen VN, Goding CR, Jochemsen AG, et al. Interaction between p53 mutation and a somatic HDMX biomarker better defines metastatic potential in breast cancer. Cancer Res. 2015;75(4):698–708.

182. Wade M, Wong ET, Tang M, Stommel JM, Wahl GM. Hdmx modulates the outcome of p53 activation in human tumor cells. J Biol Chem. 2006;281(44):33036–33044.

183. Lenos K, Grawenda AM, Lodder K, Kuijjer ML, Teunisse AF, Repapi E, Grochola LF, Bartel F, Hogendoorn PCW, Wueri P, et al. Alternate splicing of the p53 inhibitor HDMX offers a superior prognostic biomarker than p53 mutation in human cancer. Cancer Res. 2012;72(16):4074–4084.

184. Bartel F, Schulz J, Böhnke A, Blümke K, Kappler M, Bache M, Schmidt H, Wurl P, Taubert H, Hauptmann S. Significance of HDMX-S (or MDM4) mRNA splice variant overexpression and HDMX gene amplification on primary soft tissue sarcoma prognosis. Int J Cancer. 2005;117(3):469–475.

185. Zhuo B, Wang X, Shen Y, Li J, Li S, Li Y, Wang R. Interleukin-24 inhibits the phenotype and tumorigenicity of cancer stem cell in osteosarcoma via downregulation Notch and Wnt/β-catenin signaling. J Bone Oncol. 2021;31: Article 100403.

186. Hu Q, Zhang Y, Wang P, Zhou M, Hu Z, Liu C, Liu M, Wu L, Liu X, Liang D. IL-24 armored CAR19-T cells show enhanced antitumor activity and persistence. Signal Transduct Target Ther. 2021;6(1):14.

187. Whitaker EL, Filippov V, Filippova M, Guerrero-Juarez CF, Duerksen-Hughes PJ. Splice variants of mda-7/IL-24 differentially affect survival and induce apoptosis in U2OS cells. Cytokine. 2011;56(2):272–281.

188. Jia K, Cao L, Yu Y, Jing D, Wu W, Van Tine BA, Shao Z. Signaling pathways and targeted therapies in Ewing sarcoma. Pharmacol Ther. 2025;266: Article 108765.

189. Patócs B, Németh K, Garami M, Arató G, Kovalszky I, Szendrői M, Fekete G. Multiple splice variants of EWSR1-ETS fusion transcripts co-existing in the Ewing sarcoma family of tumors. Cellular Oncology (Dordrecht, Netherlands). 2013;36(3):191–200.

190. Selvanathan SP, Graham GT, Erkizan HV, Dirksen U, Natarajan TG, Dakic A, Yu S, Liu X, Paulsen MT, Ljungman ME, et al. Oncogenic fusion protein EWS-FLI1 is a network hub that regulates alternative splicing. Proc Natl Acad Sci USA. 2015;112(11):E1307–E1316.

191. Selvanathan SP, Graham GT, Grego AR, Baker TM, Hogg JR, Simpson M, Batish M, Crompton B, Stegmaier K, Tomazou EM, et al. EWS-FLI1 modulated alternative splicing of ARID1A reveals novel oncogenic function through the BAF complex. Nucleic Acids Res. 2019;47(18):9619–9636.

192. Alfert A, Moreno N, Kerl K. The BAF complex in development and disease. Epigenetics Chromatin. 2019;12(1):19.

193. Paronetto MP, Bernardis I, Volpe E, Bechara E, Sebestyén E, Eyras E, Valcárcel J. Regulation of FAS exon definition and apoptosis by the Ewing sarcoma protein. Cell Rep. 2014;7(4):1211–1226.

194. Zhang T, Wang Z, Liu M, Liu L, Yang X, Zhang Y, et al. Acetylation dependent translocation of EWSR1 regulates CHK2 alternative splicing in response to DNA damage. Oncogene. 2022;41(29):3694–3704.

195. Antoni L, Sodha N, Collins I, Garrett MD. CHK2 kinase: Cancer susceptibility and cancer therapy—Two sides of the same coin? Nat Rev Cancer. 2007;7(12):925–936.

196. Paronetto MP, Miñana B, Valcárcel J. The Ewing sarcoma protein regulates DNA damage-induced alternative splicing. Mol Cell. 2011;43(3):353–368.

197. Flores G, Grohar PJ. One oncogene, several vulnerabilities: EWS/FLI targeted therapies for Ewing sarcoma. J Bone Oncol. 2021;31:100404.

198. Saulnier O, Guedri-Idjouadiene K, Aynaud MM, Chakraborty A, Bruyr J, Pineau J, O’Grady T, Mirabeau O, Grossetête S, Galvan B, et al. ERG transcription factors have a splicing regulatory function involving RBFOX2 that is altered in the EWS-FLI1 oncogenic fusion. Nucleic Acids Res. 2021;49(9):5038–5056.

199. Zöllner SK, Selvanathan SP, Graham GT, Commins RMT, Hong SH, Moseley E, Parks S, Haladyna JN, Erkizan HV, Dirksen U, et al. Inhibition of the oncogenic fusion protein EWS-FLI1 causes G(2)-M cell cycle arrest and enhanced vincristine sensitivity in Ewing’s sarcoma. Sci Signal. 2017;10(499): Article eaam8429.

200. Palombo R, Verdile V, Paronetto MP. Poison-exon inclusion in DHX9 reduces its expression and sensitizes Ewing sarcoma cells to chemotherapeutic treatment. Cells. 2020;9(2):328.

201. Passacantilli I, Frisone P, De Paola E, Fidaleo M, Paronetto MP. hnRNPM guides an alternative splicing program in response to inhibition of the PI3K/AKT/mTOR pathway in Ewing sarcoma cells. Nucleic Acids Res. 2017;45(21):12270–12284.

202. Neckles C, Boer RE, Aboreden N, Cross AM, Walker RL, Kim BH, Kim S, Schneekloth JS Jr, Caplen NJ. HNRNPH1-dependent splicing of a fusion oncogene reveals a targetable RNA G-quadruplex interaction. RNA. 2019;25(12):1731–1750.

203. Huang G, Zhou Z, Wang H, Kleinerman ES. CAPER-α alternative splicing regulates the expression of vascular endothelial growth factor₁₆₅ in Ewing sarcoma cells. Cancer. 2012;118(8):2106–2116.

204. Rennel E, Waine E, Guan H, Schüler Y, Leenders W, Woolard J, Sugiono M, Gillatt D, Kleinerman E, Bates D, et al. The endogenous anti-angiogenic VEGF isoform, VEGF165b inhibits human tumour growth in mice. Br J Cancer. 2008;98(7):1250–1257.

205. Sand LG, Jochemsen AG, Beletkaia E, Schmidt T, Hogendoorn PC, Szuhai K. Novel splice variants of CXCR4 identified by transcriptome sequencing. Biochem Biophys Res Commun. 2015;466(1):89–94.

206. Picarda G, Surget S, Guiho R, Téletchéa S, Berreur M, Tirode F, et al. A functional, new short isoform of death receptor 4 in Ewing’s sarcoma cell lines may be involved in TRAIL sensitivity/resistance mechanisms. Mol Cancer Res. 2012;10(3):336–346.

207. Iacobescu GL, Corlatescu AD, Serban B, Spiridonica R, Costin HP, Cirstoiu C. Genetics and molecular pathogenesis of the chondrosarcoma: A review of the literature. Curr Issues Mol Biol. 2024;46(11):12658–12671.

208. Hikichi Y, Yoshimura K, Takigawa M. All-trans retinoic acid-induced ADAM28 degrades proteoglycans in human chondrocytes. Biochem Biophys Res Commun. 2009;386(2):294–299.

209. Hubeau C, Rocks N, Cataldo D. ADAM28: Another ambivalent protease in cancer. Cancer Lett. 2020;494:18–26.

210. Singh K, Oladipupo SS. An overview of CCN4 (WISP1) role in human diseases. J Transl Med. 2024;22(1):601.

211. Katoh M, Katoh M. Characterization of human ARHGAP10 gene in silico. Int J Oncol. 2004;25(4):1201–1206.

212. Lin LL, Yang F, Zhang DH, Hu C, Yang S, Chen XQ. ARHGAP10 inhibits the epithelial-mesenchymal transition of non-small cell lung cancer by inactivating PI3K/Akt/GSK3β signaling pathway. Cancer Cell Int. 2021;21(1):320.

213. Spada S, Tocci A, Di Modugno F, Nisticò P. Fibronectin as a multiregulatory molecule crucial in tumor matrisome: From structural and functional features to clinical practice in oncology. J Exp Clin Cancer Res. 2021;40(1):102.

214. Flanagan M, Liang H, Norton PA. Alternative splicing of fibronectin mRNAs in chondrosarcoma cells: Role of far upstream intron sequences. J Cell Biochem. 2003;90(4):709–718.

215. Murphy PA, Jailkhani N, Nicholas SA, Del Rosario AM, Balsbaugh JL, Begum S, Kimble A, Hynes RO. Alternative splicing of FN (fibronectin) regulates the composition of the arterial wall under low flow. Arterioscler Thromb Vasc Biol. 2021;41(1):e18–e32.

216. Hensel JA, Heineman BD, Kimble AL, Jellison ER, Reese B, Murphy PA. Identification of splice regulators of fibronectin-EIIIA and EIIIB by direct measurement of exon usage in a flow-cytometry based CRISPR screen. Sci Rep. 2021;11(1):19835.

217. Matsui Y, Nakata K, Araki N, Ozono K, Fujita Y, Tsumaki N, et al. Alternative splicing of fibronectin transcripts in osteochondrogenic tumors. Anticancer Res. 2001;21(2A):1103–1106.

218. Li S, Pritchard DM, Yu LG. Regulation and function of matrix metalloproteinase-13 in cancer progression and metastasis. Cancer. 2022;14(13):3263.

219. Tardif G, Dupuis M, Reboul P, Geng CS, Pelletier JP, Ranger P, et al. Identification and differential expression of human collagenase-3 mRNA species derived from internal deletion, alternative splicing, and different polyadenylation and transcription initiation sites. Osteoarthr Cartil. 2003;11(7):524–537.

220. Kiani C, Chen L, Wu YJ, Yee AJ, Yang BB. Structure and function of aggrecan. Cell Res. 2002;12(1):19–32.

221. Matsui Y, Araki N, Tsuboi H, Tsumaki N, Nakata K, Kawabata H, Eyre DR, Yoshikawa H. Differential expression of aggrecan mRNA isoforms by chondrosarcoma cells. Anticancer Res. 2002;22(6C):4169–4172.

222. Wang Y, Wang G, Liu H. Tenascin-C: A key regulator in angiogenesis during wound healing. Biomolecules. 2022;12(11):4169–4172.

223. Ghert MA, Jung ST, Qi W, Harrelson JM, Erickson HP, Block JA, Scully SP. The clinical significance of tenascin-C splice variant expression in chondrosarcoma. Oncology. 2001;61(4):306–314.

224. Ghert MA, Qi WN, Erickson HP, Block JA, Scully SP. Tenascin-C splice variant adhesive/anti-adhesive effects on chondrosarcoma cell attachment to fibronectin. Cell Struct Funct. 2001;26(3):179–187.

225. Nallanthighal S, Heiserman JP, Cheon DJ. Collagen type XI alpha 1 (COL11A1): A novel biomarker and a key player in cancer. Cancer. 2021;13(5):935.

226. Liu Z, Lai J, Jiang H, Ma C, Huang H. Collagen XI alpha 1 chain, a potential therapeutic target for cancer. FASEB J. 2021;35(6): Article e21603.

227. Matsui Y, Kimura T, Tsumaki N, Nakata K, Yasui N, Araki N, Hashimoto N, Uchida A, Ochi T. Splicing patterns of type XI collagen transcripts act as molecular markers for osteochondrogenic tumors. Cancer Lett. 1998;124(2):143–148.

228. Lin C, Mak S, Meitner PA, Wolf JM, Bluman EM, Block JA, Terek RM. Cancer/testis antigen CSAGE is concurrently expressed with MAGE in chondrosarcoma. Gene. 2002;285(1–2):269–278.

229. Weon JL, Potts PR. The MAGE protein family and cancer. Curr Opin Cell Biol. 2015;37:1–8.

230. Hong DS, Van Tine BA, Biswas S, McAlpine C, Johnson ML, Olszanski AJ, Clarke JM, Araujo D, Blumenschein GR, Kebria P, et al. Autologous T cell therapy for MAGE-A4 + solid cancers in HLA-A*02 + patients: A phase 1 trial. Nat Med. 2023;29(1):104–114.

231. Ohkura N, Yaguchi H, Tsukada T, Yamaguchi K. The EWS/NOR1 fusion gene product gains a novel activity affecting pre-mRNA splicing. J Biol Chem. 2002;277(1):535–543.

232. Filion C, Labelle Y. The oncogenic fusion protein EWS/NOR-1 induces transformation of CFK2 chondrogenic cells. Exp Cell Res. 2004;297(2):585–592.

233. Coleman RE, Croucher PI, Padhani AR, Clézardin P, Chow E, Fallon M, Guise T, Colangeli S, Capanna R, Costa L. Bone metastases. Nat Rev Dis Primers. 2020;6(1):83.

234. Hernández I, Moreno JL, Zandueta C, Montuenga L, Lecanda F. Novel alternatively spliced ADAM8 isoforms contribute to the aggressive bone metastatic phenotype of lung cancer. Oncogene. 2010;29(26):3758–3769.

235. Liadi Y, Campbell T, Dike P, Harlemon M, Elliott B, Odero-Marah V. Prostate cancer metastasis and health disparities: A systematic review. Prostate Cancer Prostatic Dis. 2024;27(2):183–191.

236. Antonarakis ES, Lu C, Wang H, Luber B, Nakazawa M, Roeser JC, Chen Y, Mohammad TA, Chen Y, Fedor HL, et al. AR-V7 and resistance to enzalutamide and abiraterone in prostate cancer. N Engl J Med. 2014;371(11):1028–1038.

237. Sobhani N, Neeli PK, D’Angelo A, Pittacolo M, Sirico M, Galli IC, Roviello G, Nesi G. AR-V7 in metastatic prostate cancer: A strategy beyond redemption. Int J Mol Sci. 2021;22(11):5515.

238. Ho Y, Dehm SM. Androgen receptor rearrangement and splicing variants in resistance to endocrine therapies in prostate cancer. Endocrinology. 2017;158(6):1533–1542.

239. Han D, Labaf M, Zhao Y, Owiredu J, Zhang S, Patel K, Venkataramani K, Steinfeld JS, Han W, Li M, et al. Androgen receptor splice variants drive castration-resistant prostate cancer metastasis by activating distinct transcriptional programs. J Clin Invest. 2024;134(11): Article e168649.

240. Marek RD, Halabi S, Wang ME, McBane J, Wei J, Wang T, Yang X, Liu C, Lei G, Lyerly HK, et al. Vaccination against androgen receptor splice variants to immunologically target prostate cancer. Vaccine. 2024;12(11):1273.

241. Katleba KD, Ghosh PM, Mudryj M. Beyond prostate cancer: An androgen receptor splice variant expression in multiple malignancies, non-cancer pathologies, and development. Biomedicine. 2023;11(8):2215.

242. Zhang X, Morrissey C, Sun S, Ketchandji M, Nelson PS, True LD, Vakar-Lopez F, Vessella RL, Plymate SR. Androgen receptor variants occur frequently in castration resistant prostate cancer metastases. PLOS ONE. 2011;6(11): Article e27970.

243. Hörnberg E, Ylitalo EB, Crnalic S, Antti H, Stattin P, Widmark A, Bergh A, Wikstrom P. Expression of androgen receptor splice variants in prostate cancer bone metastases is associated with castration-resistance and short survival. PLOS ONE. 2011;6(4): Article e19059.

244. Chen N, Ye XC, Chu K, Navone NM, Sage EH, Yu-Lee LY, Logothetis CJ, Lin SH. A secreted isoform of ErbB3 promotes osteonectin expression in bone and enhances the invasiveness of prostate cancer cells. Cancer Res. 2007;67(14):6544–6548.

245. Lin SH, Cheng CJ, Lee YC, Ye X, Tsai WW, Kim J, Pasqualini R, Arap W, Navone NM, Tu S-M, et al. A 45-kDa ErbB3 secreted by prostate cancer cells promotes bone formation. Oncogene. 2008;27(39):5195–5203.

246. Chen S, Deng B, Liu Y, Xie L, Xu J, Hu R, Zhao F, Song G, Ni R, Liu H, et al. The CTSZ-TRA2A-IL32 axis defines a targetable macrophage-dependent pathway in metastatic prostate cancer. J Transl Med. 2025;23(1):865.

247. Zhang S, Lv C, Niu Y, Li C, Li X, Shang Y, Zhang Y, Zhang Y, Zhang Y, Zeng Y, et al. RBM3 suppresses stemness remodeling of prostate cancer in bone microenvironment by modulating N6-methyladenosine on CTNNB1 mRNA. Cell Death Dis. 2023;14(2):91.

248. Choi ES, Kim H, Kim HP, Choi Y, Goh SH. CD44v8-10 as a potential theranostic biomarker for targeting disseminated cancer cells in advanced gastric cancer. Sci Rep. 2017;7(1):4930.

249. Zeng Y, Wodzenski D, Gao D, Shiraishi T, Terada N, Li Y, Vander Griend DJ, Luo J, Kong C, Getzenberg RH, et al. Stress-response protein RBM3 attenuates the stem-like properties of prostate cancer cells by interfering with CD44 variant splicing. Cancer Res. 2013;73(13):4123–4133.

250. Wu CL, Chao YJ, Yang TM, Chen YL, Chang KC, Hsu HP, Shan YS, Lai MD. Dual role of CD44 isoforms in ampullary adenocarcinoma: CD44s predicts poor prognosis in early cancer and CD44ν is an indicator for recurrence in advanced cancer. BMC Cancer. 2015;15:903.

251. Zhu J, Zhang J, Hu P, Fan M, Song D, Yin H, Yan P, Xian S, Li Z, Guo J, et al. Identification of bone metastatic and prognostic alternative splicing signatures in prostate adenocarcinoma. Biochem Genet. 2023;61(6):2242–2259.

252. Pineda JMB, Bradley RK. Most human introns are recognized via multiple and tissue-specific branchpoints. Genes Dev. 2018;32(7–8):577–591.

253. Schmittgen TD, Teske S, Vessella RL, True LD, Zakrajsek BA. Expression of prostate specific membrane antigen and three alternatively spliced variants of PSMA in prostate cancer patients. Int J Cancer. 2003;107(2):323–329.

254. Huang R, Zheng Z, Xian S, Zhang J, Jia J, Song D, Yan P, Yin H, Hu P, Zhu X, et al. Identification of prognostic and bone metastatic alternative splicing signatures in bladder cancer. Bioengineered. 2021;12(1):5289–5304.

255. Huang R, Zheng Z, Liu S, Yan P, Song D, Yin H, Hu P, Zhu X, Chang Z, Liu Y, et al. Identification of prognostic and bone metastasis-related alternative splicing signatures in mesothelioma. Cancer Med. 2021;10(13):4478–4492.

256. Pang X, Gong K, Zhang X, Wu S, Cui Y, Qian BZ. Osteopontin as a multifaceted driver of bone metastasis and drug resistance. Pharmacol Res. 2019;144:235–244.

257. Briones-Orta MA, Avendaño-Vázquez SE, Aparicio-Bautista DI, Coombes JD, Weber GF, Syn WK. Osteopontin splice variants and polymorphisms in cancer progression and prognosis. Biochim Biophys Acta Rev Cancer. 2017;1868(1):93–108.A.

258. Hao C, Cui Y, Hu MU, Zhi X, Zhang L, Li W, Wu W, Cheng S, Jiang WG. OPN-a splicing variant expression in non-small cell lung cancer and its effects on the bone metastatic abilities of lung cancer cells in vitro. Anticancer Res. 2017;37(5):2245–2254.

259. Kane JF, Johnson RW. Re-evaluating the role of PTHrP in breast cancer. Cancer. 2023;15(10):2670.

260. Devys A, Lortholary A, Audran M. PTHrP and breast cancer. Bull Cancer. 2001;88(11):1075–1080.

261. Tai IT, Tang MJ. SPARC in cancer biology: Its role in cancer progression and potential for therapy. Drug Resist Updat. 2008;11(6):231–246.

262. Ma J, Gao S, Xie X, Sun E, Zhang M, Zhou Q, Lu C. SPARC inhibits breast cancer bone metastasis and may be a clinical therapeutic target. Oncol Lett. 2017;14(5):5876–5882.

263. Sangaletti S, Stoppacciaro A, Guiducci C, Torrisi MR, Colombo MP. Leukocyte, rather than tumor-produced SPARC, determines stroma and collagen type IV deposition in mammary carcinoma. J Exp Med. 2003;198(10):1475–1485.

264. Feng J, Tang L. SPARC in tumor pathophysiology and as a potential therapeutic target. Curr Pharm Des. 2014;20(39):6182–6190.

265. Xu M, Zhang X, Zhang S, Piao J, Yang Y, Wang X, Lin Z. SPOCK1/SIX1axis promotes breast cancer progression by activating AKT/mTOR signaling. Aging. 2020;13(1):1032–1050.

266. Trivedi T, Pagnotti GM, Guise TA, Mohammad KS. The role of TGF-β in bone metastases. Biomolecules. 2021;11(11):1643.

267. Morrison CD, Parvani JG, Schiemann WP. The relevance of the TGF-β paradox to EMT-MET programs. Cancer Lett. 2013;341(1):30–40.

268. Yao D, Dai C, Peng S. Mechanism of the mesenchymal-epithelial transition and its relationship with metastatic tumor formation. Mol Cancer Res. 2011;9(12):1608–1620.

269. Shirakihara T, Horiguchi K, Miyazawa K, Ehata S, Shibata T, Morita I, Miyazono K, Saitoh M. TGF-β regulates isoform switching of FGF receptors and epithelial-mesenchymal transition. EMBO J. 2011;30(4):783–795.

270. Haaren MJH, Steller LB, Vastert SJ, Calis JJA, Loosdregt J. Get spliced: Uniting alternative splicing and arthritis. Int J Mol Sci. 2024;25(15):8123.

271. Zhang S, Yang Z, Xie Y, Zhang Y, Chen Z, Lv X, Deng Z, Huang Z, Cai L, Wei R. Identification of NIBAN2-regulated RUNX2 alternative splicing presents novel strategies for antagonizing osteoporosis. Adv Sci. 2025;12(17):e2416536.

272. Xiao Y, Cai GP, Feng X, Li YJ, Guo WH, Guo Q, Huang Y, Su T, Li C-J, Luo X-H, et al. Splicing factor YBX1 regulates bone marrow stromal cell fate during aging. EMBO J. 2023;42(9): Article e111762.

273. Luo H, Wang T, Xie Z, Li F, Yang C, Dong W, Wu J, Wang Q, Xu F, Liu J, et al. Glucocorticoids regulate the expression of Srsf1 through Hdac4/Foxc1 axis to induce apoptosis of osteoblasts. Commun Biol. 2025;8(1):566.

274. Liu Y, Shen H, Greenbaum J, Liu A, Su KJ, Zhang LS, Zhang L, Tian Q, Hu H-G, He J-S, et al. Gene expression and RNA splicing imputation identifies novel candidate genes associated with osteoporosis. J Clin Endocrinol Metab. 2020;105(12):e4742–e4757.

275. Zhang Z, Dong L, Tao H, Dong Y, Xiang W, Tao F, Zhao Y. RNA-binding proteins potentially regulate the alternative splicing of apoptotic genes during knee osteoarthritis progression. BMC Genomics. 2024;25(1):293.

276. Li C, Wei P, Wang L, Wang Q, Wang H, Zhang Y. Integrated analysis of transcriptome changes in osteoarthritis: Gene expression, pathways and alternative splicing. Cartilage. 2023;14(2):235–246.

277. Katsoula G, Steinberg J, Tuerlings M, Coutinho de Almeida R, Southam L, Swift D, Meulenbelt I, Wilkinson JM, Zeggini E. A molecular map of long non-coding RNA expression, isoform switching and alternative splicing in osteoarthritis. Hum Mol Genet. 2022;31(12):2090–2105.

278. Byun S, Shine J, Coryell P, Kramer NE, D’Costa S, Thulson E, Parkus SM, Chubinskaya S, Loeser RF, Diekman BO, et al. Response splicing quantitative trait loci in primary human chondrocytes identify putative osteoarthritis risk genes. Nat Commun. 2025;16(1):7932.

279. Jiang S, Xie W, Knapstein PR, Donat A, Albertsen LC, Sevecke J, Erdmann C, Appelt J, Fuchs M, Hildebrant A, et al. Transcript-dependent effects of the CALCA gene on the progression of post-traumatic osteoarthritis in mice. Commun Biol. 2024;7(1):223.

280. Forlino A, Marini JC. Osteogenesis imperfecta. Lancet. 2016;387(10028):1657–1671.

281. Marini JC, Forlino A, Bächinger HP, Bishop NJ, Byers PH, Paepe A, Fassier F, Fratzl-Zelman N, Kozloff KM, Krakow D, et al. Osteogenesis imperfecta. Nat Rev Dis Primers. 2017;3:17052.

282. Wang Q, Forlino A, Marini JC. Alternative splicing in COL1A1 mRNA leads to a partial null allele and two in-frame forms with structural defects in non-lethal osteogenesis imperfecta. J Biol Chem. 1996;271(45):28617–28623.

283. Gug C, Caba L, Mozos I, Stoian D, Atasie D, Gug M, Glorduza EV. Rare splicing mutation in COL1A1 gene identified by whole exomes sequencing in a patient with osteogenesis imperfecta type I followed by prenatal diagnosis: A case report and review of the literature. Gene. 2020;741: Article 144565.

284. Nicholls AC, Valler D, Wallis S, Pope FM. Homozygosity for a splice site mutation of the COL1A2 gene yields a non-functional proα 2(I) chain and an EDS/OI clinical phenotype. J Med Genet. 2001;38(2):132–136.

285. Lv F, Xu XJ, Wang JY, Liu Y, Asan, Wang JW, Song LJ, Song TW, Jiang Y, Wang O, et al. Two novel mutations in TMEM38B result in rare autosomal recessive osteogenesis imperfecta. J Hum Genet. 2016;61(6):539–545.

286. Chen L, Yang W, Cole WG. Alternative splicing of exon 12 of the COL2A1 gene interrupts the triple helix of type-II collagen in the Kniest form of spondyloepiphyseal dysplasia. J Orthop Res. 1996;14(5):712–721.

287. Yang W, Cole WG. Low basal transcripts of the COL2A1 collagen gene from lymphoblasts show alternative splicing of exon 12 in the Kniest form of spondyloepiphyseal dysplasia. Hum Mutat. 1998;Suppl 1:S1–S2.

288. Bruni V, Spoleti CB, La Barbera A, Dattilo V, Colao E, Votino C, Colao E, Votino C, Bellacchio E, Perrotti N, et al. A novel splicing variant of COL2A1 in a fetus with achondrogenesis type II: Interpretation of pathogenicity of in-frame deletions. Genes. 2021;12(9):1395.

289. Viakhireva I, Bychkov I, Markova T, Shatokhina O, Karandasheva K, Udalova V, Bekhtereva Y, Ryzhkova O, Skoblov M. The molecular complexity of COL2A1 splicing variants and their significance in phenotype severity. Bone. 2024;181: Article 117013.

290. BinEssa HA, Zou M, Al-Enezi AF, Alomrani B, Al-Faham MSA, Al-Rijjal RA, Meyer BF, Sji Y. Functional analysis of 22 splice-site mutations in the PHEX, the causative gene in X-linked dominant hypophosphatemic rickets. Bone. 2019;125:186–193.

291. Yamamoto A, Nakamura T, Ohata Y, Kubota T, Ozono K. Phenotypes of a family with XLH with a novel PHEX mutation. Hum Genome Var. 2020;7:8.

292. Li SS, Gu JM, Yu WJ, He JW, Fu WZ, Zhang ZL. Seven novel and six de novo PHEX gene mutations in patients with hypophosphatemic rickets. Int J Mol Med. 2016;38(6):1703–1714.

293. Fraga G, Herreros MA, Pybus M, Aza-Carmona M, Pilco-Teran M, Furlano M, Garcia-Borau MJ, Torra R, Ars E. A mild presentation of X-linked hypophosphatemia caused by a non-canonical splice site variant in the PHEX gene. Genes. 2024;15(6):679.

294. Sobacchi C, Pangrazio A, Lopez AG, Gomez DP, Caldana ME, Susani L, Vezzoni P, Villa A. As little as needed: The extraordinary case of a mild recessive osteopetrosis owing to a novel splicing hypomorphic mutation in the TCIRG1 gene. J Bone Miner Res. 2014;29(7):1646–1650.

295. Zirngibl RA, Wang A, Yao Y, Manolson MF, Krueger J, Dupuis L, Mendoza-Londono R, Voronov I. Novel c.G630A TCIRG1 mutation causes aberrant splicing resulting in an unusually mild form of autosomal recessive osteopetrosis. J Cell Biochem. 2019;120(10):17180–17193.

296. Bliznetz EA, Tverskaya SM, Zinchenko RA, Abrukova AV, Savaskina EN, Nikulin MV, Kirillov AG, Ginter EK, Polyakov AV. Genetic analysis of autosomal recessive osteopetrosis in Chuvashiya: The unique splice site mutation in TCIRG1 gene spread by the founder effect. Eur J Hum Genet. 2009;17(5):664–672.

297. Amirfiroozy A, Naghinejad M, Rezamand A, Farhangi H, Golchehre Z, Jalali H, Tahari M, Keramatipour M. A comprehensive report of the clinical and mutational profiles of 30 Iranian malignant infantile osteopetrosis patients. Mol Cell Probes. 2025;79: Article 102014.

298. Fu Y, Kim H, Roy S, Huang S, Adams JI, Grimes SM, Lau BT, Sathe A, Ji HP, Zhang NR. Single cell and spatial alternative splicing analysis with nanopore long read sequencing. Nat Commun. 2025;16(1):6654.

299. Kabza M, Ritter A, Byrne A, Sereti K, Le D, Stephenson W, Sterne-Weiler T. Accurate long-read transcript discovery and quantification at single-cell, pseudo-bulk and bulk resolution with isosceles. Nat Commun. 2024;15(1):7316.

300. Van Hecke M, Beerenwinkel N, Lootens T, Fostier J, Raedt R, Marchal K. ELLIPSIS: Robust quantification of splicing in scRNA-seq. Bioinformatics. 2025;41(2): Article btaf028.

301. Tsukasaki M, Takayanagi H. Osteoimmunology: Evolving concepts in bone-immune interactions in health and disease. Nat Rev Immunol. 2019;19(10):626–642.

302. Okamoto K, Takayanagi H. Osteoimmunology. Cold Spring Harb Perspect Med. 2019;9(1): Article a031245.

303. Heitzer E, Haque IS, Roberts CES, Speicher MR. Current and future perspectives of liquid biopsies in genomics-driven oncology. Nat Rev Genet. 2019;20(2):71–88.

304. Larson MH, Pan W, Kim HJ, Mauntz RE, Stuart SM, Pimentel M, Zhou Y, Knudsgaard P, Demas V, Aravanis AM, et al. A comprehensive characterization of the cell-free transcriptome reveals tissue- and subtype-specific biomarkers for cancer detection. Nat Commun. 2021;12(1):2357.

305. Lv X, Sun X, Gao Y, Song X, Hu X, Gong L, Han L, He M, Wei M. Targeting RNA splicing modulation: New perspectives for anticancer strategy? J Exp Clin Cancer Res. 2025;44(1):32.

306. Suñé-Pou M, Limeres MJ, Moreno-Castro C, Hernández-Munain C, Suñé-Negre JM, Cuestas ML, Sune C. Innovative therapeutic and delivery approaches using nanotechnology to correct splicing defects underlying disease. Front Genet. 2020;11:731.

307. Murphy AJ, Li AH, Li P, Sun H. Therapeutic targeting of alternative splicing: A new frontier in cancer treatment. Front Oncol. 2022;12: Article 868664.

308. Scharner J, Aznarez I. Clinical applications of single-stranded oligonucleotides: Current landscape of approved and in-development therapeutics. Mol Ther. 2021;29(2):540–554.

309. Dhuri K, Bechtold C, Quijano E, Pham H, Gupta A, Vikram A, Bahal R. Antisense oligonucleotides: An emerging area in drug discovery and development. J Clin Med. 2020;9(6):2004.

310. El Marabti E, Abdel-Wahab O. Therapeutic modulation of RNA splicing in malignant and non-malignant disease. Trends Mol Med. 2021;27(7):643–659.

311. Bashari A, Siegfried Z, Karni R. Targeting splicing factors for cancer therapy. RNA. 2023;29(4):506–515.

312. Wu K, Kwon SH, Zhou X, Fuller C, Wang X, Vadgama J, Wu Y. Overcoming challenges in small-molecule drug bioavailability: A review of key factors and approaches. Int J Mol Sci. 2024;25(23):13121.

313. Scotti MM, Swanson MS. RNA mis-splicing in disease. Nat Rev Genet. 2016;17(1):19–32.

314. Urbanski LM, Leclair N, Anczuków O. Alternative-splicing defects in cancer: Splicing regulators and their downstream targets, guiding the way to novel cancer therapeutics. Wiley Interdiscip Rev RNA. 2018;9(4): Article e1476.

315. Mercuri E, Sumner CJ, Muntoni F, Darras BT, Finkel RS. Spinal muscular atrophy. Nat Rev Dis Primers. 2022;8(1):52.

316. Hoy SM. Nusinersen: A review in 5q spinal muscular atrophy. CNS Drugs. 2021;35(12):1317–1328.

317. Corey DR. Nusinersen, an antisense oligonucleotide drug for spinal muscular atrophy. Nat Neurosci. 2017;20(4):497–499.

318. Aartsma-Rus A, Takeda S. A historical perspective on the development of antisense oligonucleotide treatments for Duchenne muscular dystrophy and spinal muscular atrophy. J Neuromuscul Dis. 2025; 10.1177/22143602251317422.

319. Hua Y, Vickers TA, Okunola HL, Bennett CF, Krainer AR. Antisense masking of an hnRNP A1/A2 intronic splicing silencer corrects SMN2 splicing in transgenic mice. Am J Hum Genet. 2008;82(4):834–848.

320. Finkel RS, Mercuri E, Darras BT, Connolly AM, Kuntz NL, Kirschner J, Chiriboga AC, Saito K, Servais L, Tizzano E, et al. Nusinersen versus sham control in infantile-onset spinal muscular atrophy. N Engl J Med. 2017;377(18):1723–1732.

321. Moultrie F, Chiverton L, Hatami I, Lilien C, Servais L. Pushing the boundaries: Future directions in the management of spinal muscular atrophy. Trends Mol Med. 2025;31(4):307–318.

322. Baranello G, Darras BT, Day JW, Deconinck N, Klein A, Masson R, Mercuri E, Rose K, El-Khairi M, Gerber M, et al. Risdiplam in type 1 spinal muscular atrophy. N Engl J Med. 2021;384(10):915–923.

323. Balińska N, Ostrowska K, Studzińska S. Short review on currently used sample preparation and determination methods of risdiplam. J Sep Sci. 2025;48(5): Article e70173.

324. Masson R, Mazurkiewicz-Bełdzińska M, Rose K, Servais L, Xiong H, Zanoteli E, Baranello G, Bruno C, Day JW, Deconinck N, et al. Safety and efficacy of risdiplam in patients with type 1 spinal muscular atrophy (FIREFISH part 2): Secondary analyses from an open-label trial. Lancet Neurol. 2022;21(12):1110–1119.

325. Oskoui M, Day JW, Deconinck N, Mazzone ES, Nascimento A, Saito K, Vuillerot C, Baranello G, Goemans N, Kirschner J, et al. Two-year efficacy and safety of risdiplam in patients with type 2 or non-ambulant type 3 spinal muscular atrophy (SMA). J Neurol. 2023;270(5):2531–2546.

326. Childs-Disney JL, Yang X, Gibaut QMR, Tong Y, Batey RT, Disney MD. Targeting RNA structures with small molecules. Nat Rev Drug Discov. 2022;21(10):736–762.

327. Keller CG, Shin Y, Monteys AM, Renaud N, Beibel M, Teider N, Peters T, Faller T, St-Cyr S, Knehr J, et al. An orally available, brain penetrant, small molecule lowers huntingtin levels by enhancing pseudoexon inclusion. Nat Commun. 2022;13(1):1150.

328. White DS, Dunyak BM, Vaillancourt FH, Hoskins AA. A sequential binding mechanism for 5’ splice site recognition and modulation for the human U1 snRNP. Nat Commun. 2024;15(1):8776.

329. Hong DS, Kurzrock R, Naing A, Wheler JJ, Falchook GS, Schiffman JS, Faulkner N, Pilat MJ, O’Brien J, LoRusso P. A phase I, open-label, single-arm, dose-escalation study of E7107, a precursor messenger ribonucleic acid (pre-mRNA) splicesome inhibitor administered intravenously on days 1 and 8 every 21 days to patients with solid tumors. Investig New Drugs. 2014;32(3):436–444.

330. Sciarrillo R, Terrana F, Comandatore A, Supadmanaba IGP, Wang B, Hassouni BE, Mantini G, Jansen G, Avan A, Carbone D, et al. Exploring splicing modulation as an innovative approach to combat pancreatic cancer: SF3B1 emerges as a prognostic indicator and therapeutic target. Int J Biol Sci. 2024;20(8):3173–3184.

331. Finci LI, Zhang X, Huang X, Zhou Q, Tsai J, Teng T, Agrawal A, Chan B, Irwin S, Karr C, et al. The cryo-EM structure of the SF3b spliceosome complex bound to a splicing modulator reveals a pre-mRNA substrate competitive mechanism of action. Genes Dev. 2018;32(3–4):309–320.

332. Nguyen JQN, Drabarek W, Leeflang A, Brands T, Bosch TPP, Verdijk RM, Werken HJG, Riet J, Paridaens D, Klein A, et al. The impact of spliceosome inhibition in SF3B1-mutated uveal melanoma. Invest Ophthalmol Vis Sci. 2024;65(12):11.

333. Wu G, Fan L, Edmonson MN, Shaw T, Boggs K, Easton J, Rusch MC, Webb TR, Zhang J, Potter PM. Inhibition of SF3B1 by molecules targeting the spliceosome results in massive aberrant exon skipping. RNA. 2018;24(8):1056–1066.

334. Seiler M, Yoshimi A, Darman R, Chan B, Keaney G, Thomas M, Agrawal AA, Caleb B, Csibi A, Sean E, et al. H3B-8800, an orally available small-molecule splicing modulator, induces lethality in spliceosome-mutant cancers. Nat Med. 2018;24(4):497–504.

335. Steensma DP, Wermke M, Klimek VM, Greenberg PL, Font P, Komrokji RS, Yang J, Brunner AM, Carraway HE, Ades L, et al. Phase I first-in-human dose escalation study of the oral SF3B1 modulator H3B-8800 in myeloid neoplasms. Leukemia. 2021;35(12):3542–3550.

336. Voss DM, Kral AJ, Sim G, Utama R, Lin KT, Cizmeciyan C, Schafer B, Cunniff PJ, Vakoc CR, Caruthers MH, et al. PKM splice-switching ASOs induce upregulation of dual-specificity phosphatases and dephosphorylation of ERK1/2 in hepatocellular carcinoma. J Biol Chem. 2025;301(4): Article 108345.

337. Deng Y, Hou M, Wu Y, Liu Y, Xia X, Yu C, Yu J, Yang H, Zhang Y, Zhu X. SIRT3-PINK1-PKM2 axis prevents osteoarthritis via mitochondrial renewal and metabolic switch. Bone Res. 2025;13(1):36.

338. Li M, Li F, Zhu C, Zhang C, Le Y, Li Z, Wang Q. The glycolytic enzyme PKM2 regulates inflammatory osteoclastogenesis by modulating STAT3 phosphorylation. J Biol Chem. 2025;301(4): Article 108389.

339. Pu F, Liu J, Jing D, Chen F, Huang X, Shi D, Wu W, Lin H, Zhao L, Zhang Z, et al. LncCCAT1 interaction protein PKM2 upregulates SREBP2 phosphorylation to promote osteosarcoma tumorigenesis by enhancing the Warburg effect and lipogenesis. Int J Oncol. 2022;60(4):44.

340. Ma WK, Voss DM, Scharner J, Costa ASH, Lin KT, Jeon HY, Wilkinson JE, Jackson M, Rigo F, Bennett CF, et al. ASO-based PKM splice-switching therapy inhibits hepatocellular carcinoma growth. Cancer Res. 2022;82(5):900–915.

341. Mulvaney KM, Blomquist C, Acharya N, Li R, Ranaghan MJ, O’Keefe M, Rodriguez DJ, Young MJ, Kesar D, Pal D, et al. Molecular basis for substrate recruitment to the PRMT5 methylosome. Mol Cell. 2021;81(17):3481–95.e7.

342. Jiao Z, Huang Y, Gong K, Liu Y, Sun J, Yu S, Zhao G. Medicinal chemistry insights into PRMT5 inhibitors. Bioorg Chem. 2024;153: Article 107859.

343. Watts JM, Bradley TJ, Thomassen A, Brunner AM, Minden MD, Papadantonakis N, Abedin S, Baines AJ, Barbash O, Gorman S, et al. A phase I/II study to investigate the safety and clinical activity of the protein arginine methyltransferase 5 inhibitor GSK3326595 in subjects with myelodysplastic syndrome and acute myeloid leukemia. Blood. 2019;134:2656.

344. Fong JY, Pignata L, Goy PA, Kawabata KC, Lee SC, Koh CM, Musiani D, Massignani E, Kotini AG, Penson A, et al. Therapeutic targeting of RNA splicing catalysis through inhibition of protein arginine methylation. Cancer Cell. 2019;36(2):194–209.e9.

345. Belmontes B, Slemmons KK, Su C, Liu S, Policheni AN, Moriguchi J, Tan H, Xie F, Aiello DA, Yang Y, et al. AMG 193, a clinical stage MTA-cooperative PRMT5 inhibitor, drives antitumor activity preclinically and in patients with MTAP-deleted cancers. Cancer Discov. 2025;15(1):139–161.

346. Slemmons KK, Su C, Liu S, Yu P-W, McEntee M, Yuan E, Lee J, Lu J, Horning A, Hindoyan A, et al. Abstract 4225: Targeting RNA splicing sensitizes MTAP-deleted cancers to MTA-cooperative PRMT5 inhibitors. Cancer Res. 2025;85(8_Supplement_1):4225.

347. Hamard PJ, Santiago GE, Liu F, Karl DL, Martinez C, Man N, Mookhtiar AK, Duffort S, Greenblatt S, Verdun RE, et al. PRMT5 regulates DNA repair by controlling the alternative splicing of histone-modifying enzymes. Cell Rep. 2018;24(10):2643–2657.

348. Carter J, Hulse M, Sivakumar M, Burtell J, Thodima V, Wang M, Agarwal A, Vykuntam K, Spruance J, Bhagwat N, et al. PRMT5 inhibitors regulate DNA damage repair pathways in cancer cells and improve response to PARP inhibition and chemotherapies. Cancer Res Commun. 2023;3(11):2233–2243.

349. O’Brien S, Butticello M, Thompson C, Wilson B, Wyce A, Mahajan V, Kruger R, Mohammad H, Fedoriw A. Inhibiting PRMT5 induces DNA damage and increases anti-proliferative activity of Niraparib, a PARP inhibitor, in models of breast and ovarian cancer. BMC Cancer. 2023;23(1):775.

350. Kota SK, Roening C, Patel N, Kota SB, Baron R. PRMT5 inhibition promotes osteogenic differentiation of mesenchymal stromal cells and represses basal interferon stimulated gene expression. Bone. 2018;117:37–46.

351. Dong Y, Song C, Wang Y, Lei Z, Xu F, Guan H, Chen A, Li F. Inhibition of PRMT5 suppresses osteoclast differentiation and partially protects against ovariectomy-induced bone loss through downregulation of CXCL10 and RSAD2. Cell Signal. 2017;34:55–65.

352. Ramachandran J, Liu Z, Gray RS, Vokes SA. PRMT5 is necessary to form distinct cartilage identities in the knee and long bone. Dev Biol. 2019;456(2):154–163.

353. Verbeke S, Bourdon A, Guegan JP, Leroy L, Chaire V, Richard E, Bessede A, Italiano A. Antitumor effects of PRMT5 inhibition in sarcomas. Cancer Res Commun. 2023;3(11):2211–2220.

354. Malvi P, Janostiak R, Chava S, Manrai P, Yoon E, Singh K, Harigopal M, Gupta R, Wajapeyee N. LIMK2 promotes the metastatic progression of triple-negative breast cancer by activating SRPK1. Oncogene. 2020;9(8):77.

355. Wang C, Zhou Z, Subhramanyam CS, Cao Q, Heng ZSL, Liu W, Fu X, Hu Q. SRPK1 acetylation modulates alternative splicing to regulate cisplatin resistance in breast cancer cells. Commun Biol. 2020;3(1):268.

356. Huang JQ, Duan LX, Liu QY, Li HF, Hu AP, Song JW, Lin C, Huang B, Yao D, Peng B, et al. Serine-arginine protein kinase 1 (SRPK1) promotes EGFR-TKI resistance by enhancing GSK3β Ser9 autophosphorylation independent of its kinase activity in non-small-cell lung cancer. Oncogene. 2023;42(15):1233–1246.

357. Jang SW, Yang SJ, Ehlén A, Dong S, Khoury H, Chen J, Persson JL, Ye K. Serine/arginine protein-specific kinase 2 promotes leukemia cell proliferation by phosphorylating acinus and regulating cyclin A1. Cancer Res. 2008;68(12):4559–4570.

358. Chang Y, Wu Q, Tian T, Li L, Guo X, Feng Z, Zhou J, Zhang L, Zhou S, Feng G, et al. The influence of SRPK1 on glioma apoptosis, metastasis, and angiogenesis through the PI3K/Akt signaling pathway under normoxia. Tumour Biol. 2015;36(8):6083–6093.

359. Cai G, Bao Y, Li Q, Hsu PH, Xia J, Ngo JCK. Design of a covalent protein-protein interaction inhibitor of SRPKs to suppress angiogenesis and invasion of cancer cells. Commun Chem. 2024;7(1):144.

360. Li Q, Zeng C, Liu H, Yung KWY, Chen C, Xie Q, Zhang Y, Wan SWC, Mak BSW, Xia J, et al. Protein-protein interaction inhibitor of SRPKs alters the splicing isoforms of VEGF and inhibits angiogenesis. iScience. 2021;24(5): Article 102423.

361. Malhi NK, Allen CL, Stewart E, Horton KL, Riu F, Batson J, Amoaku W, Morris JC, Arkilll KP, Bates DO. Serine-arginine-rich protein kinase-1 inhibition for the treatment of diabetic retinopathy. Am J Physiol Heart Circ Physiol. 2022;322(6):H1014–h27.

362. Supradit K, Boonsri B, Duangdara J, Thitiphatphuvanon T, Suriyonplengsaeng C, Kangsamaksin T, Janvilisri T, Tohtong R, Yacqub-Usman K, Grabowska AM, et al. Inhibition of serine/arginine-rich protein kinase-1 (SRPK1) prevents cholangiocarcinoma cells induced angiogenesis. Toxicol In Vitro. 2022;82: Article 105385.

363. Bastos DA, Antonarakis ES. CTC-derived AR-V7 detection as a prognostic and predictive biomarker in advanced prostate cancer. Expert Rev Mol Diagn. 2018;18(2):155–163.

364. Daghestani HN, Jordan JM, Renner JB, Doherty M, Wilson AG, Kraus VB. Serum N-propeptide of collagen IIA (PIIANP) as a marker of radiographic osteoarthritis burden. PLOS ONE. 2017;12(12): Article e0190251.

365. Lian W, Liu H, Sun LY, Liu YQ, Cui SL, Wang Y, Song QQ, Deng Q, Wang SP, Cao YH, et al. Serum levels of PIICP, PIIANP, and PIIBNP are decreased in patients with an endemic osteochondropathy, Kashin-Beck disease. J Orthop Surg Res. 2018;13(1):128.

366. Kraus VB, Karsdal MA. Clinical monitoring in osteoarthritis: Biomarkers. Osteoarthr Cartil. 2022;30(9):1159–1173.

367. Li B, Wang P, Jiao J, Wei H, Xu W, Zhou P. Roles of the RANKL-RANK axis in immunity—Implications for pathogenesis and treatment of bone metastasis. Front Immunol. 2022;13: Article 824117.

368. Shemirani R, Lin G, Abduweli Uyghurturk D, Le M, Nakano Y. An miRNA derived from amelogenin exon4 regulates expression of transcription factor Runx2 by directly targeting upstream activators Nfia and Prkch. J Biol Chem. 2022;298(5): Article 101807.

369. Kondo Y, Oubridge C, Roon A-MM, Nagai K. Crystal structure of human U1 snRNP, a small nuclear ribonucleoprotein particle, reveals the mechanism of 5′ splice site recognition. eLife. 2015;4: Article e04986.

370. Kattah NH, Kattah MG, Utz PJ. The U1-snRNP complex: Structural properties relating to autoimmune pathogenesis in rheumatic diseases. Immunol Rev. 2010;233(1):126–145.

371. Feltz C, Hoskins AA. Structural and functional modularity of the U2 snRNP in pre-mRNA splicing. Crit Rev Biochem Mol Biol. 2019;54(5):443–465.

372. Kao CY, Cao EC, Wai HL, Cheng SC. Evidence for complex dynamics during U2 snRNP selection of the intron branchpoint. Nucleic Acids Res. 2021;49(17):9965–9977.

373. Leung AK, Nagai K, Li J. Structure of the spliceosomal U4 snRNP core domain and its implication for snRNP biogenesis. Nature. 2011;473(7348):536–539.

374. Vidovic I, Nottrott S, Hartmuth K, Lührmann R, Ficner R. Crystal structure of the spliceosomal 15.5kD protein bound to a U4 snRNA fragment. Mol Cell. 2000;6(6):1331–1342.

375. Wood KA, Eadsforth MA, Newman WG, O’Keefe RT. The role of the U5 snRNP in genetic disorders and cancer. Front Genet. 2021;12: Article 636620.

376. Newman AJ. The role of U5 snRNP in pre-mRNA splicing. EMBO J. 1997;16(19):5797–5800.

377. Mroczek S, Dziembowski A. U6 RNA biogenesis and disease association. Wiley Interdiscip Rev RNA. 2013;4(5):581–592.

378. Montemayor EJ, Didychuk AL, Yake AD, Sidhu GK, Brow DA, Butcher SE. Architecture of the U6 snRNP reveals specific recognition of 3’-end processed U6 snRNA. Nat Commun. 2018;9(1):1749.

379. Dunn EA, Rader SD. Secondary structure of U6 small nuclear RNA: Implications for spliceosome assembly. Biochem Soc Trans. 2010;38(4):1099–1104.

380. Stanek D, Rader SD, Klingauf M, Neugebauer KM. Targeting of U4/U6 small nuclear RNP assembly factor SART3/p110 to Cajal bodies. J Cell Biol. 2003;160(4):505–516.

381. Miller MJ, Kroenke MA, Barger T, Manning MS, Sohn W, Graham K, Mytych DT, Gupta S. Inhibition of RANKL is critical for accurate assessment of anti-drug antibody incidence to denosumab in clinical studies. J Immunol Methods. 2025;540: Article 113864.

382. Deeks ED. Denosumab: A review in postmenopausal osteoporosis. Drugs Aging. 2018;35(2):163–173.

383. Cretu C, Gee P, Liu X, Agrawal A, Nguyen TV, Ghosh AK, Cook A, Jurica M, Larsen NA, Pena V. Structural basis of intron selection by U2 snRNP in the presence of covalent inhibitors. Nat Commun. 2021;12(1):4491.

384. Wheeler EC, Martin BJE, Doyle WC, Neaher S, Conway CA, Pitton CN, Gorelov RA, Donahue M, Jann JC, Abdel-Wahab O, et al. Splicing modulators impair DNA damage response and induce killing of cohesin-mutant MDS and AML. Sci Transl Med. 2024;16(728): Article eade2774.