Circular RNAs inducing the osteogenic differentiation of dental mesenchymal stem cells via microRNA sponging

Abstract

Circular RNAs (circRNAs) are a distinct type of nonlinear and noncoding RNAs endogenously expressed by pre-mRNA back-splicing and crucial in transcriptional and posttranscriptional regulation. CircRNAs can regulate cellular and molecular pathways through various mechanisms, such as microRNA sponging. Numerous studies have indicated the regulatory roles of circRNAs in the osteogenic differentiation of stem cells (SCs) isolated from different sources. Dental tissue-derived mesenchymal SCs (MSCs) have received considerable attention in artificial bone engineering, in which SCs are used to manufacture functional bone tissues to repair bone defects. Recently, studies have reported the regulatory roles of circRNAs in the osteogenic differentiation of dental-derived MSCs, such as apical papillae, dental pulp, and dental follicle SCs. This review aimed to discuss the findings of studies evaluating the contribution of circRNAs to the osteogenic differentiation of dental-derived MSCs.

Key Words: Circular RNA; Dentistry; MicroRNA; Osteogenesis; Dental stem cells

Core Tip: Several circular RNAs were found to induce osteogenesis in dental mesenchymal stem cells through acting as microRNA sponges leading to eliminating the inhibitory impacts of microRNAs on downstream target genes. These circular RNAs can help identify new molecular elements that provide diagnostic biomarkers and/or therapeutic targets for treating bone-associated dental disorders.

INTRODUCTION

Circular RNAs (circRNAs) are a distinct type of endogenous nonlinear noncoding RNAs formed by the back-splicing of pre-mRNAs, which play a crucial role in transcriptional and posttranscriptional regulation. They were first detected in 1991 and considered useless by-products with incorrect splicing[1-3]. CircRNAs were not found to be significant until reports of their widespread existence in mammals, particularly humans, in 2013[3-6]. They are expressed in various human cells[4] and show predominant cell/tissue-specific expression in the cytoplasm[5]. A high-throughput sequencing study showed that circRNAs are widely expressed by numerous human genes and exhibited higher expression than their corresponding homologous linear isoforms[7].

Compared with conventional linear noncoding RNAs, circRNAs have a stable head-to-tail closed circular structure created by the covalent binding of the 3’ and 5’ ends without a 5’ cap structure and 3’ polyadenylation tail[8]. Such a structure renders them resistant to exonuclease degradation and more stable than linear RNAs[9-13]. A circular loop is created by a particular alternative splicing event called back-splicing, by which an upstream splice acceptor is bound to a downstream splice donor[14] (Figure 1).

Figure 1 Diagrammatic of the biogenesis and sponging function of circular RNAs. Circular RNAs (circRNAs) are generated from pre-mRNAs. A loop structure is formed through back-splicing driven by the pairing of intronic complementary sequences. CircRNA is eventually formed by removing the intron sequences in the loop structure. CircRNAs sponge microRNAs to decrease their availability to bind and inhibit target mRNAs, leading to the expression of the target mRNAs. circRNA: Circular RNAs.

CircRNAs can influence various cellular and molecular functions through several mechanisms[15-18]. These including: (1) Acting as competitive endogenous RNAs, where circRNAs, among other transcripts, can competitively bind microRNAs (miRs) in the cytoplasm at conserved miR target sites, thereby counteracting the binding and inhibitory effects of mIR on downstream mRNA target genes (Figure 1); (2) Interacting with RNA-binding proteins; (3) Activating polymerase II machinery and U1 small nuclear ribonucleoproteins to regulate the expression of target genes; (4) Serving as scaffolding proteins that facilitate protein-protein interactions; and (5) Attaching to target proteins in the nucleus, which help stabilize these proteins so that they are not easily digested. Among the aforementioned mechanisms, circRNAs normally function as miRNA sponges[3,19,20] (Figure 1). For example, circRNA CDRlas contains several binding sites for miR-7, acting such a sponge for it and competing with its downstream target genes for binding, thereby negatively regulating the function of miR-7[21]. They have various biological functions and participate in cellular events, such as cell proliferation and differentiation[22-26]. Emerging studies have demonstrated the contribution of circRNAs to physiological and pathological events associated with cellular metabolism, inflammation, and apoptosis[27-29]. CircRNAs regulate gene expression in cancers and cardiovascular diseases. They are also employed as diagnostic biomarkers in cardiovascular and neuropsychiatric diseases[30,31].

Although most recent studies on circRNAs have focused on cancer, numerous studies have shown that they can regulate stem cell (SC) differentiation and tissue regeneration events[32-34]. Of note, circRNAs were overexpressed during the differentiation of induced pluripotent SCs in humans[32], and many studies have shown their participation in the proliferation and osteogenic differentiation of SCs by acting as miR sponges[35,36].

SCs are undifferentiated cells with high self-renewal and differentiation abilities for generating one or more types of specific cells. They serve as a primary resource in regenerative medicine, particularly in artificial bone engineering in which functional bone tissues are created as a scaffold or artificial environment for repairing bone defects. In recent years, SC therapy, particularly mesenchymal SCs (MSCs), has received considerable attention for the treatment of bone defects[37-41].

The teeth are the most natural, noninvasive source of SCs. Dental-derived MSCs (DMSCs) show great potential in regenerative medicine. These cells possess several important advantages, such as the absence of ethical concerns, easy accessibility, and ready availability, which make them of great interest for research. SCs self-renew to maintain a pool of cells that can be activated to replace terminally differentiated cells or enable wound healing, for example, the healing of periodontal tissues after surgery. DMSCs can differentiate into functional blood vessels and nerves. Several clinical trials have reported that transplanting DMSCs into disinfected necrotic teeth has allowed for the recovery of tooth vitality and vertical and horizontal root growth in immature teeth with incomplete root formation[42].

In recent years, numerous DMSCs have been isolated and characterized, such as MSCs from the gingiva, SCs from the apical papilla (SCAPs), dental follicle progenitor cells (DFPCs), SCs from exfoliated deciduous teeth, periodontal ligament SCs, and dental pulp SCs (DPSCs). Once exposed to special inducer conditions, similar to other SCs, DMSCs can differentiate into various tissue-like cells. Recently, several studies have reported the regulatory roles of circRNAs in the osteogenic differentiation of some DMSCs, including DPSCs, SCAPs, and DFPCs.

Osteogenesis is a biological process for the regeneration of lost bone volume. Tooth loss results in alveolar bone resorption. To suppress the stimulus generated by the periodontal ligament, the vestibular cortical bone undergoes resorption, resulting in the gradual disappearance of the marrow component of the alveolus. Consequently, the morphology of the alveolar ridge is altered, that is, when a few teeth are lost, the extent of the alveolar defect is affected; however, when more teeth are lost, a more noticeable atrophy occurs. In these conditions, the bone must be regenerated, taking advantage of the osteogenesis[43]. This review aimed to discuss the findings of various studies evaluating the effect of circRNAs on the osteogenic differentiation of DPSCs, SCAPs, and DFPCs (Table 1).

Table 1 The molecular targets and underlying mechanisms of circular RNAs inducing osteogenic differentiation of dental stem cells.

CircRNA	Target miRs	Underlying mechanisms	Experimental methods	Ref.
CircRNAs inducing osteogenic differentiation of human dental pulp stem cells
Hsa-circ-036872	MiR-143-3p	IGF2 signaling activation	In vitro culture of hDPSCs isolated from orthodontic patients. In vivo experiments in BALB/c nude mice for heterotopic implantation of hDPSCs for bone formation assay	[52]
CircFKBP5	MiR-708-5p	GIT2 signaling activation	In vitro culture of hDPSCs isolated from human dental pulp tissue	[34]
CircSIPA1L1	MiR-617	Smad3 signaling activation	In vitro culture of hDPSCs isolated from human dental pulp tissue. In vivo experiments in BALB/c nude mice for heterotopic implantation of hDPSCs for bone formation assay	[61]
CircLPAR1	Hsa-miR-31	Increased expression of SATB2	In vitro culture of hDPSCs isolated from human dental pulp tissue	[66]
Hsa-circ-0026827	MiR-188-3p	Enhancing Beclin-1-mediated autophagy	In vitro culture of hDPSCs isolated from human dental pulp tissue. In vivo experiments in BALB/c nude mice for heterotopic implantation of hDPSCs for bone formation assay	[68]
CircRNA124534	MiR-496	β-catenin signaling activation	In vitro culture of hDPSCs isolated from healthy pulp tissues derived from caries-free teeth of patients. In vivo experiments in BALB/c nude mice for heterotopic implantation of hDPSCs for bone formation assay	[67,73]
CircAKT3	MiR-206	Increased expression of CX43	In vitro culture of hDPSCs isolated from human dental pulp tissue. In vivo experiments in BALB/c nude mice for heterotopic implantation of hDPSCs for bone formation assay	[77]
CircRNAs inducing osteogenic differentiation of apical papilla stem cells
Hsa-circ-0008016	MiR-337-3p	FGF/FGFR signaling activation	In vitro culture of SCAPs derived from human apical papilla tissues isolated from orthodontic patients	[93]
Circ-ZNF236	MiR-218-5p	LGR4-mediated Wnt/β-catenin signaling activation	In vitro culture of SCAPs derived from human apical papilla tissues isolated from teeth with immature roots from healthy donors. In vivo experiments in a rat skull-impaired model for heterotopic implantation of SCAPs for bone formation assay	[94]
CircRNAs inducing osteogenic differentiation of dental follicle stem cells
CircFgfr2	MiR-133	Increased expression of DLX3, RUNX2, and BMP6 in TGF-β and MAPK signaling pathways	In vitro culture of dental follicles isolated from the tooth germs of the mandibular molars of rats	[120]

CircRNAs: Circular RNAs; IGF2: Insulin-like growth factor 2; hDPSC: Human dental pulp stem cell; GIT2: G-protein-coupled receptor-kinase interacting protein 2; Smad3: Mothers against decapentaplegic homolog 3; SATB2: Special AT-rich sequence-binding protein 2; CX43: Connexin 43; FGF: Fibroblast growth factor; FGFR: Fibroblast growth factor receptor; LGR4: Leucine-rich repeat-containing G-protein-coupled receptor 4; SCAP: Stem cell from the apical papilla; RUNX2: Runt-related transcription factor 2; BMP6: Bone morphogenetic protein 6; TGF-β: Transforming growth factor-β; MAPK: Mitogen-activated protein kinase; circLPAR1: Circular lysophosphatidic acid receptor 1.

LITERATURE REVIEW AND ELIGIBILITY CRITERIA

A systematic search was conducted in the electronic databases of Web of Science, Scopus, and PubMed to extract relevant articles from database inception to July 2024. The following keywords were used as part of the search strategy in each database: [“Circular RNAs” OR “circRNAs”] AND “dental stem cell” AND “osteogenesis”, and all articles were entered into the EndNote X9 reference manager software for screening. After removing duplicates, the remaining articles were assessed for selection. Articles published in English and in vitro, in vivo, and clinical studies evaluating circRNAs in the osteogenesis differentiation of DMSCs were included. Conversely, studies that did not provide sufficient data, review articles, and conference abstracts were excluded.

CIRCRNAS INDUCING THE OSTEOGENIC DIFFERENTIATION OF HUMAN DPSCS

Human DPSCs (hDPSCs) are a class of MSCs that can differentiate into specific cell types, such as odontoblasts, osteoblasts, adipocytes, chondrocytes, and neural-like cells[44-46]. hDPSCs are frequently utilized in regenerative medicine because of their high proliferation rate, plasticity in multilineage differentiation, and ease of acquisition. Compared with MSCs, hDPSCs also show higher osteogenic potential but lower immunogenicity[47]. During osteoblast differentiation, hDPSCs express osteoblast differentiation-associated markers such as osterix (OSX), alkaline phosphatase (ALP), osteocalcin (OCN), and runt-related transcription factors (RUNX)[48,49]. RUNX2 is a transcriptional regulator of early-stage osteogenesis that directly affects the expression of genes responsible for bone tissue enrichment and intracranial secretion[50]. The OSX gene, which participates in osteoblast differentiation and bone formation, is one of the downstream targets of RUNX2[51]. ALP, as an early marker of calcification, is associated with osteogenesis specificity and osteoblast activity[52]. Emerging evidence demonstrates that circRNAs are crucial in the osteogenic differentiation of hDPSCs by sponging miRs.

Osteogenic activity of hsa-circ-0036872 in hDPSCs

The detection of differentially expressed circRNAs by next-generation sequencing technology after osteogenic promotion indicated that hsa-circ-0036872, arising from an exon of the FURIN gene, was upregulated during osteogenic differentiation of hDPSCs[53]. Notably, reduced expression of hsa-circ-0036872 suppressed the osteogenic differentiation of hDPSCs[53]. A dual-luciferase reporter assay revealed miR-143-3p as a downstream target of hsa-circ-0036872[53]. MiR-143-3p was found to inhibit osteogenesis in periodontal ligament cells by regulating kruppel-like factor 5 and suppressing Wnt/β-catenin signaling[54]. During the osteogenic differentiation of hDPSCs, the expression of miR-143-3p was reduced, and hsa-circ-0036872 upregulation inhibited the miR-143-3p expression, whereas the upregulation of miR-143-3p suppressed the osteogenic differentiation of hDPSCs[53]. Further assays demonstrated that miR-143-3p could bind to the 3’-untranslated regions (3’UTRs) of insulin-like growth factor 2 (IGF2) and inhibit its expression[53]. IGF2, which regulates different cellular processes[55], is upregulated during the osteogenic differentiation of MSCs[56,57] and promotes osteogenesis by regulating the expression of Col1 and Runx2[58]. In hDPSCs, IGF2 overexpression could reverse the suppression of osteogenic differentiation resulting from hsa-circ-0036872 silencing or miR-143-3p upregulation[53]. Of note, hsa-circ-0036872 knockout inhibits heterotopic bone formation in vivo, further supporting its promotional effect on improving hDPSC osteogenesis[53]. These findings indicate that hsa-circ-0036872 can facilitate hDPSC osteogenesis by inhibiting miR-143-3p and subsequently activating IGF2 signaling.

Osteogenic activity of circFKBP5 in hDPSCs

CircFKBP5, another circRNA generated by the back-splicing of exons 3-6 of FKBP5, was found to promote the osteogenic differentiation of hDPSCs and suppress inflammation and apoptosis[35]. Pulpitis, clinically classified as irreversible and reversible, refers to the inflammatory responses of the pulp to opportunistic infections induced by commensal oral bacteria[59]. During the pulpitis event, proinflammatory cytokines induced via microbial infection could dysregulate hDPSC viability and differentiation[60]. As a model mimicking the inflammatory condition of pulpitis, lipopolysaccharide (LPS)-induced hDPSC inflammation caused a reduction in the circFKBP5 expression by decreasing its cytoplasmic stability, whereas circFKBP5 overexpression reduced LPS-promoted inflammation, apoptosis, and suppression of viability and osteogenic differentiation of hDPSCs[35]. As shown by the dual-luciferase reporter assay, circFKBP5 could exert these protective functions by directly targeting and inhibiting miR-708-5p, which was detected to target G-protein-coupled receptor (GPCR)-kinase interacting protein 2 (GIT2)[35]. In hDPSCs, miR-708-5p expression was elevated by LPS treatment, whereas miR-708-5p knockdown prevented LPS-induced hDPSC dysfunction[35]. Furthermore, GIT2 exerts a positive regulatory effect on the functions of osteoblasts and osteoclasts[59]. LPS-treated hDPSCs exhibited a reduced expression of GIT2, and GIT2 knockdown inhibited the protective effects of circFKBP5 on hDPSCs, suggesting that GIT2 could mediate the protective effects of circFKBP5 on LPS-treated hDPSCs[35]. CircFKBP5 could prevent LPS-promoted apoptosis, inflammation, and osteogenic differentiation inhibition of hDPSCs by the miR-708-5p/GIT2 axis[36]. To further support the osteogenic potential of circFKBP5, another study showed that circFKBP5 boosted the proliferation and osteogenic differentiation of bone marrow-derived MSCs by modulating the miR-205-5p/RUNX2 axis[61]. Altogether, circFKBP5 overexpression can efficiently inhibit apoptosis and inflammation and induce osteogenic differentiation of hDPSCs under pulpitis by the miR-708-5p/GIT2 axis.

Osteogenic activity of circSIPA1 L1 in hDPSCs

CircSIPA1 L1, expressed by a transcript encoding circSIPA1 L1, is upregulated under mineralization-promoting conditions and has been identified as a bone regeneration target. As confirmed by the altered expression of osteogenic-associated genes, such as ALP, OSX, and RUNX2, circSIPA1 L1 was found to promote the osteogenesis of DPSCs[62]. Notably, circSIPA1 L1 overexpression increased the expression of osteogenic markers, whereas circSIPA1 L1 knockdown caused a reduction in their expression[62]. The dual-luciferase reporter assay indicated that circSIPA1 L1 could exert an osteogenic effect on DPSCs by sponging miR-617 as a negative regulator of osteogenesis and further activating mothers against decapentaplegic homolog 3 (Smad3)[62]. CircSIPA1 L1 and miR-617 were colocalized in the DPSC cytoplasm, where circSIPA1 L1 could sponge miR-617[62]. Notably, circSIPA1 L1 and miR-617 were found upregulated and downregulated, respectively, during the osteogenic differentiation of DPSCs, whereas the circSIPA1 L1 knockdown or miR-617 overexpression significantly suppressed DPSC osteogenesis[62]. Another study found that miR-223-3p overexpression induced odontoblast differentiation of DPSCs by inhibiting Smad3 expression[63]. Smad3, as a miR-617 target, is a key regulator of transforming growth factor-β (TGF-β)[64], which mediates the production and degradation of the extracellular matrix[65]. After the induction of RUNX2 and TGF-β, Smad3 can be activated and accumulated to promote bone formation[66]. The Smad pathway also provoked the differentiation of mesenchymal progenitors by converging with RUNX2[66]. In summary, these findings indicate that circSIPA1 L1 stimulates DPSC osteogenesis by inhibiting miR-617 and subsequently activating Smad3.

Osteogenic activity of circular lysophosphatidic acid receptor 1 in hDPSCs

Evaluating the expression profiles of circRNAs in exosomes from osteogenic-induced DPSCs indicated increased expression of circular lysophosphatidic acid receptor 1 (circLPAR1) along with the osteogenic differentiation of DPSCs[67]. Furthermore, exosomes containing high levels of circLPAR1 exert an osteogenic effect on the recipient DPSCs[67]. Mechanistically, the dual-luciferase reporter assay revealed that circLPAR1 sponged hsa-miR-31 and eliminated its inhibitory effect on SATB2 and osteogenic differentiation of DPSCs, inducing osteogenesis of the recipient DPSCs[67]. Notably, the expression of SATB2 was upregulated in response to circLPAR1-mediated hsa-miR-31 inhibition, which caused the upregulation of its downstream genes associated with osteogenic differentiation, such as RUNX2, leading to the onset and progression of osteogenic differentiation of DPSCs[67].

Osteogenic activity of hsa-circ-0026827 in hDPSCs

Primary studies have indicated that hsa-circ-0026827 expression stimulated osteogenic differentiation[68]. Further studies have revealed that hsa-circ-0026827 expression was markedly increased during the osteoblast differentiation of DPSCs, whereas hsa-circ-0026827 knockdown inhibited the osteoblast differentiation of DPSCs[69]. This was further supported by experiments using heterotopic bone models, which demonstrated that hsa-circ-0026827 overexpression could induce heterotopic bone formation[69]. In mechanistic studies, dual-luciferase reporter assay demonstrated that hsa-circ-0026827 could induce the osteogenic differentiation of DPSCs by upregulating RUNX1 and Beclin-1-mediated autophagy by targeting and inhibiting miR-188-3p[69]. Interestingly, miR-188-3p overexpression could inhibit DPSC osteogenesis by directly inhibiting the expression of RUNX1 and Beclin-1, whereas miR-188-3p downregulation restored the osteogenic differentiation of DPSC[69]. RUNX1 is a crucial transcription factor that regulates osteogenic differentiation by affecting Smad1/5/8 and mitogen-activated protein kinase (MAPK) signaling[70]. Moreover, autophagy is a natural self-cannibalization process that not only has variety of biological effects but may also induce osteogenesis and suppress bone loss[71-73], further supporting hsa-circ-0026827-induced osteogenic differentiation of DPSCs through the enhancement of Beclin-1-mediated autophagy via the sponging of miR-188-3p.

Osteogenic activity of circRNA124534 in hDPSCs

CircRNA124534 has demonstrated a pivotal role in DPSC osteogenesis, showing significant upregulation in the process[68]. In vitro and in vivo experimental studies have uncovered that circRNA124534 overexpression could enhance the osteogenic differentiation of DPSCs, as evidenced by increased levels of the osteogenic-related genes OCN and RUNX2[74]. Mechanistically, circRNA124534, acting as an miR sponge, directly targets miR-496 and consequently upregulates β-catenin expression, eventually inducing DPSC osteogenesis[74]. The Wnt/β-catenin signaling is known to induce the osteogenic differentiation of DPSCs[75]. The Wnt/β-catenin signaling activation leads to the transcriptional activation of several downstream genes, such as cyclinDl and Runx2, which can regulate cell proliferation and osteogenic differentiation[76,77]. MiR-496 directly inhibited β-catenin expression, whereas miR-496 knockdown restored the β-catenin level and stimulated the osteogenic differentiation of DPSCs[74]. Notably, circRNA124534 and the β-catenin 3’UTR shared the same miR-496 response elements and competitively interacted with miR-496[74]. Therefore, overexpressed circRNA124534 could act as a competitive RNA molecule absorbing miR-496 and consequently reverse β-catenin inhibition, thereby inducing the osteogenic differentiation of DPSCs.

Osteogenic activity of circAKT3 in hDPSCs

Microarray analysis of the expression profiles of circRNAs during the osteogenic differentiation of hDPSCs indicated significant upregulation of circAKT3[78]. An in vitro investigation showed that circAKT3 knockdown led to the downregulation of osteogenic marker genes and inhibited osteogenic differentiation of hDPSCs[78]. Prediction and validation studies using a dual-luciferase reporter assay have indicated that circAKT3 could directly bind and inhibit miR-206 expression, and the latter was found to target connexin 43 (CX43) mRNA and inhibit hDPSC osteogenesis[78]. MiR-206 knockdown could reverse the suppressive effect of circAKT3 knockdown on osteogenic differentiation[78]. CX43, a positive regulator of SC osteogenesis, was predicted as a miR-206 target, and miR-206 upregulation and silencing could reduce and increase the CX43 expression, respectively[78]. Bone formation in vivo indicated that circAKT3 knockdown inhibited the expression of the osteogenic proteins OCN and COL1 and the formation of mineralized nodules[78]. Similarly, several studies have indicated that miR-206 upregulation suppressed osteoblast differentiation of osteoblasts and MSCs by targeting the 3’UTR sequence of CX43 mRNA and subsequently reducing the protein expression of CX43[79-82]. In conclusion, during hDPSC osteogenesis, circAKT3 can act as a positive regulator by directly sponging miR-206 and abolishing the inhibitory effects of miR-206 on CX43 expression.

CIRCRNAS INDUCING THE OSTEOGENIC DIFFERENTIATION OF SCAPS

SCAPs, a class of ectomesenchyme-derived cells in the apical tissue of immature teeth, have the abilities of multilineage differentiation and self-renewal[83]. They contribute to root development and pulp regeneration. SCAPs can induce the regeneration of tooth root-like tissues and bone/dentin-like structures[84-87] and differentiate into osteoblasts and odontoblasts[88]. They are more proliferative than DPSCs[89]. Given their advantages such as easy culture, differentiation capacity, and high proliferative rate and stable biological properties, SCAPs are suitable seed cells for pulp regeneration[90,91].

Several studies have shown the regulatory roles of circRNAs on SCAP osteogenesis. High-throughput sequencing detecting changes in circRNA expression profiles revealed 301 differentially expressed circRNAs during the osteogenic differentiation of SCAPs[92]. Validating the high-throughput sequencing data revealed that among the differentially expressed circRNAs, the expression levels of four circRNAs, including hsa-circ-0002538, has-circ-0003549, hsa-circ-0006618, and hsa-circ-0040809, were significantly increased, whereas the expression levels of three circRNAs, including hsa-circ-0002381, has-circ-0009031, and hsa-circ-0068958, were significantly downregulated[92]. The function of these circRNAs was predicted for miRs that circRNAs may bind and the target genes that miRs may regulate[92]. Subsequently, two networks of circRNA-miRNA-mRNA were created utilizing upregulated and downregulated circRNAs, showing complex interactions between circRNAs and miRNAs[92]. Another study evaluating the circRNA expression profiles also indicated that among the differentially expressed circRNAs, the expression of circRNAs, including ISPD, PRDM5, LATS2, and UTP4, was significantly high during the osteogenic differentiation of SCAPs, whereas the expression profiles of circRNAs, including METTL9, SRPK1, ARID1A, and EIF4E2, were significantly downregulated[93]. Similarly, a cross-talk circRNA-miR-mRNA network was constructed using such differentially expressed circRNAs, further supporting that competitive circRNA-miR-mRNA regulatory networks provide important mechanisms for SCAP osteogenesis[93]. Furthermore, in the aforementioned studies, predicting the potential functions of target genes using Gene Ontology and Kyoto Encyclopedia of Genes and Genomes analysis showed that the target genes of these differentially expressed circRNAs are enriched in different signaling pathways associated with osteogenic differentiation, such as the Wnt and MAPK signaling pathways[92,93].

A study reported that hsa-circ-0008016 could actively induce osteogenic differentiation while diminishing odontogenic differentiation of SCAPs[94]. Further analysis using the dual-luciferase reporter assay revealed that hsa-circ-0008016 could sponge the miR-337-3p molecule. Fibroblast growth factor receptor 1 (FGFR1), which has a critical role in osteoblast differentiation by maintaining the balance between bone formation and remodeling, was the target gene of miRNA-337-3p and hsa-circ-0008016[94]. Therefore, hsa-circ-0008016 can abolish the suppressive effect of miR-337-3p on FGF/FGFR signaling, thereby enhancing the osteogenic differentiation of SCAPs.

A subsequent study unveiled that hsa-circ-0000857 (circ-ZNF236) was the most significantly upregulated circRNA during SCAP osteogenesis[95]. Gain/loss-of-function studies on circ-ZNF236 indicated significant upregulation of the protein expressions of various osteogenesis-related genes, such as OSX, RUNX2, ALP, and DSPP, in SCAPs overexpressing circ-ZNF236 and significant downregulation in circ-ZNF236 knockdown ones, suggesting circ-ZNF236 as a positive regulator in SCAP osteogenesis[95]. These results were further supported by in vivo studies revealing that SCAPs overexpressing circ-ZNF236 induced bone formation in a rat skull-impaired model[95]. Mechanistic studies using a dual-luciferase reporter assay showed that the high circ-ZNF236 expression could enhance SCAP osteogenesis via miR-218-5p/leucine-rich repeat-containing GPCR4 (LGR4) signaling. Specifically, circ-ZNF236 was identified to sponge miR-218-5p to promote the osteogenic differentiation of SCAPs. MiR-218-5p negatively affected SCAP osteogenesis by directly targeting the mRNA of LGR4 at its 3’-UTR[95]. LGR4, also called GPR48, is a member of the largest cell surface molecule family of GPCRs contributing to the transmission of extracellular signals to the cytoplasm[96]. LGR4 knockout could reduce the osteogenic differentiation of SCAPs[95,97], whereas miR-218-5p inhibition blocks the effect of LGR4 knockout on SCAP osteogenesis[95]. Of note, circ-ZNF236 could indirectly increase LGR4 expression by absorbing miR-218-5p, thus inducing the osteogenic differentiation of SCAPs[95]. Other studies have shown that LGR4 contributes to the committed differentiation of DMSCs and osteogenic differentiation of SCAPs by inducing Wnt/β-catenin-mediated autophagy[97-103]. Autophagy induces the osteogenic differentiation of DPSCs[104-106] and MSCs[107]. Notably, the excessive expression of circ-ZNF236 could induce autophagosome formation, and circ-ZNF236-mediated autophagy activation could strengthen the committed differentiation ability of SCAPs[95]. Therefore, these findings indicate that circ-ZNF236 could promote autophagy by abolishing the suppressive effects of miR-218-5p on the LGR4-mediated Wnt/β-catenin pathway and stimulate osteogenic differentiation of SCAPs.

CIRCRNAS INDUCING THE OSTEOGENIC DIFFERENTIATION OF DFPCS

Periodontitis, as a chronic progressive inflammatory status, is identified through alveolar bone resorption and periodontal attachment loss, causing clinical traits such as tooth loosening[108]. Challenges limit the regeneration of lost alveolar bone[109]. In current clinical practice, the guided bone tissue regeneration strategy is applied to reconstruct the alveolar bone. However, it shows inconsistent clinical efficacy and limited applicability[110]. Investigations have recently focused on alveolar bone regeneration using SCs from dental follicle cells (DFCs). Dental follicles can form the periodontal tissues and thus exert a critical role during tooth growth[111,112]. It contains numerous SCs that can differentiate into osteoblasts, fibroblasts, and cementoblasts and is essential for the formation of alveolar bone, periodontal ligaments, and cementum[112]. DFCs are potential cells capable of differentiation and osteogenesis under inflammatory conditions and contribute significantly to tooth regeneration[113]. Experimental studies have shown the cementogenic, neurogenic, adipogenic, and osteogenic differentiation of DFCs[114-119], indicating the usage of DFCs in tooth regeneration. Many studies have demonstrated the vital contribution of circRNAs in DFC osteogenesis. Once osteogenic differentiation of rat DFCs begins, circRNAs have been detected to be differentially expressed; 128 downregulated and 138 upregulated circRNAs were detected, potentially involving MAPK and TGF-β signaling pathways[120]. Among them, the expression levels of circFgfr2 and its predicted downstream targets, bone morphogenetic protein-6 (BMP6) and miR-133, were found to be significantly altered in vitro and in vivo[120]. The expression level of circFgfr2 was markedly increased during DFC osteogenesis and postnatal development of dental follicles, suggesting a critical contribution of circFgfr2 to osteogenic differentiation of DFCs[120]. In addition, circFgfr2 overexpression in DFCs caused miR-133 downregulation and BMP6 upregulation, which are essential for DFC osteogenesis. Further validations using the dual-luciferase reporter assay indicated that circFgfr2 absorbs the miR-133a-3p molecule[121]. MiR-133a-3p was also found to inhibit DFC osteogenesis by targeting the osteogenic genes DLX3 and RUNX2[121]. DLX3 and RUNX2 are known as direct and indirect miR-133a-3p targets, respectively, and contribute to osteogenesis[122,123] as well as bone formation and mineralization in craniofacial bones[124-126]. During osteogenic differentiation, DLX3 acts as a transcription regulator of RUNX2[127,128], which is a BMP response gene required for early bone formation[129]. Interestingly, miR-133a-3p overexpression reduced the circFgfr2 level and the mRNA expression of DLX3 and RUNX2, whereas circFgfr2 overexpression could markedly decrease miR-133a-5p level and increased the expression levels of DLX3 and RUNX2[121]. These findings reveal that circFgfr2 overexpression accelerates the osteogenic differentiation of DFCs by reversing the suppressive effects of miR-133a-3p on the expression of DLX3, RUNX2, and BMP6. The osteogenic activities of circFgfr2 were further confirmed by in vivo implantation, which showed that circFgfr2 overexpression significantly enhanced the formation of new bone-like tissues with increased bone mineral regions[121]. Notably, histological analyses indicated that circFgfr2 overexpression increased the volume of the new bone in the scaffold tissue, in which osteogenic markers DLX3, COL1, OPN, and RUNX2 were significantly upregulated[121]. Therefore, circFgfr2 acts as an osteogenetic facilitator that can induce DFC osteogenesis via the miR-133a-3p/DLX3-RUNX2-BMP6 axis in TGF-β and MAPK signaling pathways. In summary, these findings indicate that circFgfr2 as a positive regulator during osteogenic differentiation can provide a novel therapeutic target to stimulate tooth-defect regeneration using DFCs.

PROMISING THERAPEUTIC PERSPECTIVE OF OSTEOGENIC CIRCRNAS IN ENDODONTIC DISEASES

The potential of the aforementioned circRNAs in inducing osteogenic differentiation of DMSCs can be an efficient tool for cell-free regenerative endodontic treatment of dental diseases where bone restoration is challenging and tooth loss occurs, such as periodontal disorders, osteogenesis imperfecta, caries, and trauma. Periodontitis, also known as gum disease, is an inflammatory condition arising from a serious gum infection that damages the alveolar bone, periodontal ligament, and cementum, destroying the bone that supports teeth[130-133]. Currently, no periodontal treatment facilitates the regeneration of the modified region and the lost periodontal tissue into a normal and functional structure. Notably, with the development of SC-based regenerative medicine, tooth regeneration has become an ideal and promising method in the treatment of endodontic diseases[134-136]. Compared with SC therapy, cell-free regenerative endodontic treatment is mainly based on inducing the redevelopment of endogenous SCs at the lesion site, avoiding both technical and ethical problems that SC transplantation needs to solve[137]. Therefore, the induction of the osteogenic differentiation of dental-derived SCs using the here-reported circRNAs may be a promising and efficient approach for new tooth formation and treatment of dental diseases related to limited bone restoration, such as periodontitis.

THERAPEUTIC POTENTIAL, CLINICAL PROSPECTIVE, AND TECHNICAL CHALLENGES OF CIRCRNAS

CircRNAs take on critical roles in various cancers, and research into their functions and mechanisms suggests promising clinical applications. Based on the current research, circRNAs may serve as potential diagnostic markers (such as has-circ-0006401 in colorectal cancer and circAXIN1 in gastric cancer), prognostic markers (such as circFNDC3B, circPLCE1, and circMAPK14 in colon cancer; circDIDO1 and circMAPK1 in gastric cancer; circ-FBXW7 in glioblastoma; and circASK1 in lung adenocarcinoma), or therapeutic targets[138]. More studies using patient-derived xenograft mouse models and other tumor-bearing mouse models have demonstrated that several circRNAs might be potential therapeutic targets for cancer. These include circPLCE1, which can be a potential therapeutic target for colorectal cancer, circGprc5a in bladder cancer, circAXIN1 in gastric cancer, and circAXIN1 in the lung metastasis of gastric cancer[138]. Despite their considerable potential as biomarkers or therapeutic targets, their potential for use in clinical practice has not been established. In the development of circRNA therapeutics, multiple challenges must be addressed, mainly the design and optimization of circRNAs, circularization efficiency of circRNAs, and chemical manufacturing and control process development of circRNAs[139].

Despite recent advances in understanding circRNA biogenesis and functions, technological obstacles remain in determining the physicochemical properties and mechanisms of action of circRNAs occurring at multiple levels. The main concern is that the circRNA-miRNA-mRNA stoichiometry required for efficiency competition in miR binding is rarely observed under physiological conditions. Furthermore, because circRNA sequences fully overlap with their cognate linear RNA isoforms processed from the same pre-mRNAs, dissecting the functional significance of circRNAs has been challenging. Resolving the contribution of a circRNA from its residing gene into an observable effect remains difficult. Improvements in methods to study these RNA circles without affecting their residing genes are key to understanding what they do in cells. Further challenges exist in assays to identify circRNA-binding proteins[140,141].

CONCLUSION

This review summarized the activities and underlying mechanisms of circRNAs in regulating the osteogenic differentiation of DMSCs, such as DPSCs, SCAPs, and DFPCs. Generally, all reported circRNAs induce osteogenesis in DMSCs by acting as miR sponges, eliminating the inhibitory effects of miRs on downstream target genes. In DPSCs, these circRNAs and corresponding molecular targets include hsa-circ-0036872 and the miR-143-3p/IGF2 axis, circFKBP5 and the miR-708-5p/GIT2 axis, circSIPA1 L1 and the miR-617/Smad3 axis, circLPAR1 and the has-miR-31/RUNX2, hsa-circ-0026827 and the miR-188-3p/Beclin-1 axis, circRNA124534 and the miR-496/β-catenin axis, and circAKT3 and the miR-206/CX43 axis. In SCAPs, they are hsa-circ-0008016 and the miR-337-3p/FGFR axis and circ-ZNF236 and the miR-218-5p/LGR4 axis. Moreover, DFPC osteogenesis was found to be induced by circFgfr2 and the miR-133a-3p/DLX3-RUNX2-BMP6 axis.

These findings can help identify new molecular elements that provide diagnostic biomarkers and/or therapeutic targets for the treatment of bone-associated dental disorders. However, more evidence is needed to interpret this phenomenon and explore the application of the results of laboratory studies to the clinic. Currently, research has focused on investigating the sponge function of circRNAs in dental bone regeneration, whereas the other completely different but equally important functions have rarely been mentioned. Well-designed in vivo experiments are warranted to further explore the effects of circRNAs on the living body and realize their early clinical application in disease diagnosis and treatment. A comprehensive understanding of circRNA functions and regulation in dental bone regeneration processes could reveal novel molecular pathways crucial for maintaining dental integrity and homeostasis. This knowledge could lead to the development of targeted therapeutic interventions for various dental disorders.