CircRNAs are a class of non-coding RNA molecules characterized by covalently closed circular structures, which distinguish them from traditional linear RNAs (Fang et al., 2025). Unlike linear RNAs, circRNAs lack both a 5′ cap and a 3′ poly(A) tail, making them resistant to exonuclease-mediated degradation and thereby conferring high stability within the cellular environment (Ding et al., 2025; Greco et al., 2025). This stability allows circRNAs to persist over extended periods in complex intracellular contexts and to participate in the regulation of various biological processes (Sanati and Ghafouri-Fard, 2025). The discovery of circRNAs has significantly expanded our understanding of the RNA world and established a new and important subfield within non-coding RNA research.

CircRNAs exhibit structural diversity and are mainly classified into four types: exonic circRNAs (ecircRNAs), circular intronic RNAs (ciRNAs), exon-intron circRNAs (EIciRNAs), and tRNA-derived circRNAs (tricRNAs) (Yang Y. et al., 2025; Lv et al., 2025; Schmidt and Matera, 2020). Among them, ecircRNAs are the most abundant and are formed by back-splicing of one or more exons (Lin et al., 2021). They are predominantly located in the cytoplasm, where they act as competing endogenous RNAs (ceRNAs) by sponging miRNAs or interacting with proteins to regulate gene expression (Xi et al., 2025). ciRNAs, derived from retained intronic sequences, are typically localized in the nucleus and regulate host gene transcription by interacting with transcriptional machinery such as RNA polymerase II (Shafaghat et al., 2025). EIciRNAs contain both exonic and intronic regions and also tend to function in the nucleus, for example by recruiting U1 small nuclear ribonucleoproteins (snRNPs) to enhance the transcription of host genes (Li et al., 2021). TricRNAs represent a relatively new subclass of circRNAs, generated from tRNA precursors under stress conditions, and are thought to be involved in cellular stress responses, although their precise functions remain to be fully elucidated (Robic and Kühn, 2020).

The biogenesis of circRNAs mainly relies on a back-splicing mechanism, which differs from canonical linear RNA splicing (Hatzimanolis et al., 2025). In back-splicing, a downstream 5′ splice donor site is covalently linked to an upstream 3′ splice acceptor site, resulting in the formation of a closed circular RNA molecule (Sur et al., 2025). This process is regulated by multiple factors, including genomic structural features, RBPs, and cis-acting elements (Jiang et al., 2025). For instance, reverse complementary sequences within flanking introns—such as Alu repeats—can facilitate circularization via base pairing. Meanwhile, RBPs such as QKI, MBL, and FUS can bind specific RNA motifs and promote the circularization process (Srinivasan et al., 2025; Wang et al., 2025b; Zhu et al., 2025). Additionally, components of the spliceosome and their cofactors, including SF3B1 and U2AF65, also participate in circRNA formation, underscoring the highly regulated nature of circRNA biogenesis (Hollander et al., 2016; Paira and Borden, 2025).

In terms of intracellular localization, circRNAs exhibit clear subcellular distribution preferences. EcircRNAs are primarily localized in the cytoplasm, where they regulate post-transcriptional gene expression by interacting with miRNAs or proteins (Liu et al., 2025). A well-known example is CDR1as (ciRS-7), which acts as a sponge for miR-7 to modulate the expression of its target genes, thereby influencing cellular proliferation and differentiation (Lou et al., 2024). In contrast, ciRNAs and EIciRNAs are enriched in the nucleus, where they modulate transcriptional activity by interacting with transcription complexes or chromatin-modifying factors (Song W. et al., 2024). These distinct localization patterns reflect the functional diversity of circRNAs and suggest that they may exert context-dependent biological effects in different cellular compartments.

CircRNA expression is highly tissue-specific and developmentally regulated. Distinct circRNA expression profiles have been observed across various tissues and cell types, with some circRNAs being specifically upregulated or downregulated under particular pathological conditions, such as cancer, cardiovascular disease, or neurodegenerative disorders (Ali et al., 2025; Mohammadpour et al., 2025; Sun M. et al., 2025). This context-dependent expression pattern makes circRNAs attractive candidates for disease biomarkers or therapeutic targets. In the case of DCM, certain circRNAs have been found to regulate key pathological processes such as myocardial fibrosis, oxidative stress, and metabolic dysregulation, offering new directions for clinical diagnosis and treatment (Rai et al., 2020).

The degradation of circRNAs in vivo is predominantly characterized by “endonuclease-mediated” pathways. Due to the absence of 5′ and 3′ termini, exonucleases generally cannot act directly, making circRNA turnover dependent on multiple intracellular endonucleolytic mechanisms. Recent studies employing in vivo and ex vivo models have further demonstrated that circRNAs are not “nondegradable,” but instead exist in a dynamic balance between homeostasis and stress responses: in aged brain tissues, abundant circRNAs can accumulate over the long term, yet their half-lives remain finely regulated by specific ribonucleases and epitranscriptomic marks such as m6A (Xu F. et al., 2024). Among these, m6A-dependent degradation represents the most prominent selective pathway under homeostatic conditions. The m6A reader protein YTH N6-methyladenosine RNA binding protein 2 (YTHDF2) recognizes methylated circRNAs and recruits RNase P/MRP endoribonuclease complex via the adaptor protein heat-responsive protein 12 (HRSP12) to facilitate cleavage; conversely, demethylases such as alkB homolog 5 (ALKBH5) and fat mass and obesity-associated protein (FTO) are thought to attenuate this degradation process. Multiple systematic reviews and experimental studies published between 2024 and 2025 have consistently identified this axis as a critical regulatory hub in the interplay between m6A modification and circRNA stability (Liu F. et al., 2024).

miRNA/Argonaute 2 (AGO2)-mediated “slicing-type” degradation highlights the decisive role of sequence complementarity. When circRNAs are highly complementary to specific miRNAs (such as CDR1as and miR-671), AGO2 can catalyze endonucleolytic slicing, directly resulting in circRNA breakage and clearance. In recent years, this mechanism has been revalidated in systems including neural tissues and cancer models, providing a unified explanation for the dual “sponge–slicing” outcomes: once the interaction shifts from “loose adsorption” to “near-perfect complementarity,” degradation becomes the predominant fate (Feng et al., 2023).

Another mechanism, termed structure-mediated RNA decay (SRD), represents a “shape-sensing” pathway independent of sequence features. SRD relies on UPF1 RNA helicase and G3BP stress granule assembly factor 1 (G3BP1) to recognize highly ordered or folded circRNA structures and promote endonucleolytic cleavage, thereby selectively reducing the steady-state levels of structurally complex circRNAs. Initially described at the transcriptome-wide level, subsequent studies in diverse model systems have suggested that interfering with UPF1 or G3BP1 preferentially upregulates highly structured circRNAs, indicating that SRD contributes to balancing the “structural diversity” of the circular RNAome (Fischer et al., 2020; Ren et al., 2022).

A major advance reported in 2025 was the identification of the ribonuclease DIS3, independent of exosomes, as a broad-spectrum circRNA degradation factor. Inhibition or depletion of DIS3 across multiple cell lines and animal tissues led to the upregulation of more than half of detected circRNAs, with minimal impact on their linear counterparts. This suggests that DIS3 exerts relatively “preferential” endonucleolytic activity toward circular forms. Further subcellular fractionation experiments supported its cytoplasmic function independent of exosomal pathways. These findings provide strong evidence that DIS3 represents a leading candidate for a “universal circRNA degradation pathway” under homeostatic conditions (Latini et al., 2025; Tao et al., 2025).

It is important to emphasize that these pathways do not operate in isolation. m6A modifications can alter the spatial conformation and binding repertoire of circRNAs, thereby influencing their routing between the YTHDF2–HRSP12–RNase P/MRP pathway and the SRD/UPF1–G3BP1 pathway. Innate immune activation, through RNase L, can induce global circRNA clearance that temporarily suppresses the fine-tuned regulation of circRNA pools by other homeostatic pathways. Meanwhile, AGO2/miRNA slicing reflects greater tissue- or developmental-stage specificity (Liu F. et al., 2024). Depending on tissue type, age, and pathological stress, the relative weighting and hierarchy of these regulatory axes can shift, shaping the heterogeneous “landscape of circRNA degradation” in vivo (Liu F. et al., 2024; Karimi et al., 2025).

In the pathogenesis of DCM, insulin resistance and chronic hyperglycemia collectively lead to significant disturbances in myocardial glucose metabolism (Mann et al., 2025). Under physiological conditions, cardiomyocytes rely primarily on the insulin–PI3K/Akt signaling pathway to mediate the translocation of glucose transporter type 4 (GLUT4) to the plasma membrane, thereby maintaining normal glucose uptake and energy homeostasis (Ferreira et al., 2018). However, under pathological DCM conditions, impaired insulin signaling disrupts GLUT4 translocation, markedly reducing glucose uptake efficiency in cardiomyocytes (Bertrand et al., 2020). Prolonged hyperglycemia also facilitates the aberrant accumulation of advanced glycation end-products (AGEs), which bind to their receptor (RAGE) and further suppress insulin signaling, forming a vicious cycle of “hyperglycemia–insulin resistance” (Giraldo-Gonzalez et al., 2025).

Recent studies have identified various circRNAs involved in the fine-tuned regulation of glucose metabolism, contributing to the pathological progression of DCM. For instance, circRNA_0071336 modulates glucose metabolic homeostasis in cardiomyocytes by sponging miR-93-5p and regulating GLUT4 expression (Yan et al., 2022). circIGF1R significantly influences cardiac fibroblast proliferation by modulating key enzymes associated with carbohydrate metabolism (Schmidt et al., 2025). Circ_0000284 specifically inhibits GLUT4 translocation in hepatocytes and contributes to arsenic-induced insulin resistance in a type 2 diabetes mellitus (T2DM) model (Xu S. et al., 2024). In another study, circHIPK3 promoted hyperglycemia and insulin resistance by sponging miR-192-5p and upregulating the transcription factor FOXO1, providing new insights and therapeutic targets for glucose metabolic disorders (Cai et al., 2020). Moreover, circPIP5K1A ameliorates insulin resistance, lipid metabolic disturbances, and inflammatory responses by targeting miR-552-3p to regulate ENO1 expression (Song G. et al., 2024). These findings advance understanding of circRNA-mediated regulation in glucose metabolism and support their potential as early diagnostic biomarkers or therapeutic targets in DCM.

Cardiac hypertrophy is a common pathological change in DCM. Within this context, multiple studies report significant alterations in circRNA expression in cardiac tissue and circulation. For example, the conserved circRNA DICAR is downregulated in diabetic hearts; its deficiency leads to spontaneous cardiac dysfunction, cardiomyocyte hypertrophy, and fibrosis, whereas DICAR overexpression alleviates the DCM phenotype, suggesting a protective role in myocardial homeostasis (Yuan et al., 2023).

At the molecular level, circRNAs often function as competing endogenous RNAs (ceRNAs) or miRNA sponges, regulating pro- or anti-hypertrophic signaling. By adsorbing miRNAs that target pro-hypertrophic genes, circRNAs relieve suppression of these targets, promoting increased cell size and protein synthesis, ultimately driving cardiac hypertrophy (Cheng et al., 2024; Xu Z. et al., 2024).

CircRNAs also regulate programmed cell death and metabolic reprogramming under diabetic stress conditions (hyperglycemia, lipotoxicity, insulin resistance). Certain circRNAs influence mitochondrial function, oxidative stress, or inflammatory pathways, modulating cardiomyocyte susceptibility to apoptosis, pyroptosis, or PANoptosis. For instance, circ-OGDH can promote or regulate PANoptosis in cardiomyocytes, indirectly influencing hypertrophy and functional impairment (Guan et al., 2025).

Myocardial interstitial remodeling and fibroblast activation, closely associated with hypertrophy, are also regulated by circRNAs. By altering fibroblast metabolism or functioning within ceRNA networks, circRNAs modulate collagen synthesis and fibroblast proliferation. Targeting pro-fibrotic circRNAs attenuates fibroblast activation and ventricular wall thickening, highlighting circRNAs as key regulators of fibrotic signaling and hypertrophic progression (You et al., 2025). Some circRNAs additionally act via noncanonical mechanisms, interacting with myosin-associated proteins to regulate protein stability or translation. For example, circ-0001283 exacerbates cardiac hypertrophy by modulating myosin light chain, illustrating mechanistic diversity and context-dependence (Wang W. et al., 2025).

The impact of circRNAs on hypertrophic phenotypes is highly dependent on tissue type, disease stage, and pathological stressors. Biphasic or opposing effects may occur for the same circRNA in different models or disease phases. High-throughput sequencing and functional screens have identified numerous circRNAs related to metabolism, cell death, and fibrosis, but challenges such as specificity, delivery, safety, and cross-species conservation remain before clinical translation (Mei et al., 2024; Yao et al., 2024).

Oxidative stress and subsequent inflammation are hallmark pathological features of DCM, primarily initiated by excessive reactive oxygen species (ROS) under hyperglycemia (Chen and Guo, 2025). Mitochondrial electron transport chain (ETC.) dysfunction from imbalanced glucose and fatty acid oxidation leads to electron leakage, generating superoxide (O₂•^-) (Li T. et al., 2025). Hyperglycemia also activates NADPH oxidase (NOX), especially NOX2/NOX4, while AGEs binding to RAGE amplifies ROS production via a positive feedback loop (Li T. et al., 2025). Excess ROS oxidatively modifies IKKβ kinase, triggering IκBα degradation and NF-κB nuclear translocation, inducing transcription of pro-inflammatory cytokines such as TNF-α and IL-6. Activation of the NLRP3 inflammasome via caspase-1 promotes IL-1β and IL-18 maturation, directly damaging cardiomyocytes and activating fibroblasts, leading to ECM deposition and fibrosis (Peng et al., 2022). ROS-induced lipid peroxidation, protein carbonylation, and DNA damage ultimately impair cardiomyocyte structure and function, resulting in systolic and diastolic dysfunction (Lei et al., 2025).

CircRNAs participate in regulating oxidative stress in DCM through multiple mechanisms. CircHIPK3 promotes ROS generation by sponging miR-20b-5p and upregulating ATG7 (Qiu et al., 2021). CircSlc8a1 enhances oxidative stress in H9c2 cells via the miR-673-5p/TFRC axis (Wu and Du, 2024). Circ-AMOTL1 is upregulated in diabetic myocardium, and its silencing improves cardiac function, reduces fibrosis, and decreases MARCKS expression, indicating pathogenic involvement (Yang Y. et al., 2023). Conversely, circFOXP1 exerts cardioprotective effects by lowering ROS through miR-9-3p binding (Rong et al., 2025).

CircRNAs also modulate inflammatory responses in DCM. CircPIP5K1A upregulation in STZ-induced diabetic rats correlates with inflammation, and its downregulation alleviates insulin resistance and inflammation via miR-552-3p/ENO1 regulation (Song G. et al., 2024). Circ_0003928 regulates high-glucose-induced oxidative stress and inflammation via the miR-31-5p/MAPK6 axis in HK-2 cells (Bao et al., 2024). CircANKRD36 silencing in T2DM rats upregulates miR-145, targets XBP1, and mitigates inflammation (Lu et al., 2021). Studies in diabetic nephropathy also highlight circRNA involvement in inflammation and fibrosis, e.g., circ-ITCH via miR-33a-5p/SIRT6 and circTAOK1 via miR-142-3p/SOX6 axes (Liu et al., 2021; Liu SY. et al., 2024). These renal findings provide mechanistic insights for cardiac circRNA research and therapeutic development in DCM.

Myocardial fibrosis is a critical pathological feature of DCM, characterized by excessive ECM accumulation, particularly collagen. Chronic hyperglycemia and metabolic dysregulation activate cardiac fibroblasts, enhancing ECM synthesis and impairing degradation, resulting in structural and functional myocardial impairment (Sun J. et al., 2025). AGEs binding to RAGE activate pro-fibrotic pathways, including TGF-β/Smad signaling, while ROS further enhance fibrosis (Qin et al., 2025; Yue et al., 2017). Insulin resistance disrupts energy and lipid metabolism, promoting apoptosis and reparative fibrosis (Xu H. et al., 2024).

Fibrotic remodeling increases myocardial stiffness and reduces ventricular compliance, causing diastolic dysfunction (Smati et al., 2025). Progressive fibrosis replaces cardiomyocytes, reducing contractile units and resulting in systolic dysfunction and heart failure. Fibrosis also disrupts electrical conduction, increasing arrhythmia risk (Cunha et al., 2022).

CircHIPK3 is widely expressed in the heart, liver, and brain and is upregulated in DCM, promoting myocardial fibrosis via miR-152-3p/TGF-β2 and miR-29b-3p pathways (Wang et al., 2021; Liu et al., 2020). High-throughput sequencing identified circPHF20L1, circCLASP1, and circSLC8A1 as key regulators, with axes including circCLASP1/miR-182-5p/Wnt7a, circSLC8A1/miR-29b-1-5p/Col12a1, and circPHF20L1/miR-29a-3p/Col6a2 (Yuan et al., 2024). Circ-AMOTL1 and circRNA_010567 also contribute to fibrosis via MARCKS and TGF-β1 regulation, respectively (Yuan et al., 2024; Zhou and Yu, 2017).

Some circRNAs exert anti-fibrotic effects. DICAR overexpression alleviates fibrosis through a VCP-Med12–mediated degradation pathway (YUAN et al., 2023). CircRNA_012164 and circRNA_42623 regulate fibrosis via miR-9-5p–related pathways, and knockdown reverses fibrotic phenotypes (WANG H. et al., 2024; WANG H. et al., 2022).The mechanism by which circRNAs regulate DCM is shown in Figure 2.

Programmed cell death (PCD) of cardiomyocytes plays a central role in the onset and progression of DCM (Zhang et al., 2024). High-glucose-induced PCD contributes not only to cardiomyocyte loss but also aggravates inflammation and promotes myocardial fibrosis, collectively accelerating cardiac dysfunction (Xuan and Zhang, 2023). PCD encompasses various cell death modalities, including apoptosis, autophagy, pyroptosis, and necroptosis, which interact to form a complex regulatory network in DCM pathogenesis.

Emerging evidence suggests that high-glucose conditions alter the expression profiles of numerous circRNAs, which modulate specific PCD pathways affecting cardiomyocyte viability and function. For instance, CDR1as, a circRNA with prognostic potential, is upregulated in DCM, while its knockdown significantly attenuates cardiomyocyte apoptosis. Mechanistic studies reveal that CDR1as inhibits ubiquitination of mammalian sterile 20-like kinase 1 (MST1), thereby activating the Hippo signaling pathway to suppress apoptosis (Shao et al., 2022). Similarly, circMAP3K5 is upregulated in response to hyperglycemia and promotes apoptosis of H9c2 cells by sponging miR-22-3p and upregulating death-associated protein kinase 2 (DAPK2), identifying the circMAP3K5/miR-22-3p/DAPK2 axis as a potential therapeutic target (Shen et al., 2024). CircHIPK3 also plays an anti-apoptotic role by downregulating PTEN, a negative regulator of survival pathways, thereby protecting AC16 cells from high-glucose-induced apoptosis (Jiang et al., 2022). High-throughput RNA sequencing has further identified several apoptosis-related circRNAs—such as mmu_circ_0000652, mmu_circ_0000547, mmu_circ_0001058, mmu_circ_0000680, and novel_circ_0004285—that may modulate early-stage diabetic myocardial apoptosis through competitive miRNA binding (Dong et al., 2020).

In terms of pyroptosis, circ_0071269 is significantly upregulated in H9c2 cells under high-glucose conditions (Fu et al., 2022). It promotes pyroptosis and inflammation via the miR-145/GSDMA axis, while its knockdown mitigates cytotoxicity and enhances cell viability (Fu et al., 2022). DICAR, previously described as a protective circRNA in DCM, inhibits cardiomyocyte pyroptosis, potentially via a VCP-Med12 degradation mechanism (Yuan et al., 2023). Another pyroptosis-associated circRNA, CACR (hsa_circ_0076631), is elevated in both diabetic patient serum and high-glucose-treated cardiomyocytes (Yang et al., 2019). CACR sponges miR-214-3p, relieving suppression of caspase-1, thereby enhancing pyroptotic signaling. CACR knockdown attenuates caspase-1 activation, while miR-214-3p inhibition partially reverses this effect, supporting the therapeutic relevance of the CACR/miR-214-3p/caspase-1 axis (Yang et al., 2019). PYRCR, another pyroptosis-related circRNA, protects against ischemia/reperfusion (I/R)-induced cardiac injury by modulating Drp1 activity via DRG2, offering a novel strategy for pyroptosis inhibition (Chen X. Z. et al., 2025).

In other PCD types, circOGDH expression is elevated in diabetic mouse myocardium, accompanied by upregulation of PANoptosis-related proteins (Guan et al., 2025). CircOGDH specifically regulates RIPK3 via the HMGB1 pathway, triggering necroptosis and exacerbating cardiac injury, providing mechanistic insights into DCM (Guan et al., 2025). Regarding autophagy, circMKLN1 is upregulated in serum of STZ-induced diabetic mice and acts as a sponge for miR-26a-5p to regulate autophagy in high-glucose/methylglyoxal-treated human retinal microvascular endothelial cells (hRMECs) (Yang J. et al., 2023). Silencing circMKLN1 inhibits excessive autophagy and inflammation, offering novel clues for exploring aberrant autophagy in cardiomyocytes and its role in DCM progression (YANG J. et al., 2023). The mechanism of CircRNAsPCD is shown in Figure 3.

It is noteworthy that in the pathological progression of DCM, different circRNAs may exert diametrically opposed effects. Several studies have demonstrated that under comparable conditions of hyperglycemia, hypoxia, or oxidative stress, some circRNAs, such as circFOXP1, exhibit cardioprotective effects by suppressing oxidative stress, attenuating apoptosis, and mitigating inflammatory responses, whereas others, such as circHIPK3, may aggravate disease progression by inducing mitochondrial fission, increasing ROS production, and directly damaging cardiomyocytes. These functional discrepancies may result from the interplay of several molecular mechanisms.

First, subcellular localization and target site specificity are critical determinants of circRNA functional diversity. Certain circRNAs localize to mitochondria or interact with proteins regulating mitochondrial dynamics—for instance, circHIPK3 promotes mitochondrial fission and elevates ROS levels by acting on DRP1—thereby directly amplifying oxidative stress (Li X. et al., 2025). In contrast, other circRNAs reside in the cytoplasm or nucleus and function through sponging miRNAs or interacting with RNA-binding proteins (RBPs), suppressing pro-oxidative or pro-apoptotic pathways and thus conferring protective effects (Rong et al., 2025).

Second, the functional output of circRNAs is strongly influenced by the abundance and binding affinity of miRNAs within the competing endogenous RNA (ceRNA) network. Whether a circRNA can effectively “release” its downstream targets depends on the number of miRNA binding sites, the binding affinity, and the cellular expression level of the circRNA itself. In contexts where miRNAs are highly expressed or target gene networks are complex, the sponging capacity of certain circRNAs may be markedly diminished, leading to a weak or absent protective phenotype.

Third, RBP- and splicing factor–mediated functional switching is also crucial. CircRNAs may serve as scaffolds or molecular decoys for RBPs, or indirectly regulate their host linear transcripts or other genes by altering RBP accessibility. Under different pathological conditions, alterations in the expression or activity of splicing factors such as the serine/arginine-rich splicing factor (SRSF) family or polypyrimidine tract-binding protein 1 (PTBP1) may further influence circRNA biogenesis, isoform composition, and functional orientation. For example, SRSF4 has been reported to associate with cardiac circFOXP1 and modulate its downstream functions (Song et al., 2025). Thus, the same circRNA may display protective or pathogenic phenotypes depending on the splicing or binding context.

In addition, epitranscriptomic modifications such as N⁶-methyladenosine (m6A) and translational potential profoundly shape circRNA functions. m6A modification not only regulates circRNA stability and nucleo-cytoplasmic distribution but also facilitates cap-independent translation, resulting in functional peptides or altered RBP-binding profiles, thereby broadening circRNA-mediated effects. Under pathological conditions, altered expression of m^6A-associated enzymes, including methyltransferase-like proteins (METTL3/14), demethylases (FTO/ALKBH5), and YTH domain-containing family (YTHDF) reader proteins, may dynamically reprogram the functional properties of specific circRNAs, shifting them between roles as “noncoding RNAs” and “peptide-coding transcripts,” and ultimately producing divergent phenotypes (Qin et al., 2022).

Moreover, disease stage and expression kinetics cannot be overlooked. CircRNAs that are upregulated in the early stress response may serve compensatory protective roles, such as transiently suppressing apoptosis or promoting repair; however, sustained overexpression in a chronic hyperglycemic environment may convert them into pathogenic drivers that exacerbate inflammation or fibrosis. This time- and dose-dependent effect explains why cross-sectional studies often yield inconsistent results and underscores the necessity of longitudinal and functional studies.

Finally, methodological variability may amplify apparent contradictions across studies. Model selection (e.g., streptozotocin [STZ]-induced mice, high-fat diet [HFD] models, in vitro high-glucose or oxidative stress treatment), endpoints of observation (transcriptional levels, ROS detection, apoptosis rate, or cardiac function), and intervention strategies (knockdown, overexpression, antisense oligonucleotides [ASO], or CRISPR-Cas13 approaches) all influence the comparability of findings (Yao et al., 2024; Qin et al., 2022). The research progress of circular RNAs is shown in Table 1.

Although significant progress has been made in elucidating the role of circRNAs in DCM—particularly in regulating programmed cardiomyocyte death, inflammatory responses, mitochondrial dysfunction, and fibrosis—the field remains in its infancy and faces several unresolved challenges that demand further investigation. Current studies exploring the relationship between circRNAs and DCM predominantly rely on in vitro stress models and animal systems, yet these approaches have inherent methodological constraints.

First, precise quantification of circRNAs in cardiac tissue remains challenging. Conventional sequencing combined with RNase R enrichment is prone to bias, short-read sequencing cannot discriminate between isoforms, and single-cell or spatial approaches are limited by sequencing depth, all of which compromise the accuracy of expression profiling (Nielsen et al., 2022; Vromman et al., 2025). Second, validation of targets and mechanisms is insufficient. Most studies rely on predictive models of “miRNA sponging” or RBP interactions without robust supporting evidence from CLIP assays, reporter systems, or stoichiometric analyses, leaving their bona fide roles in the myocardium uncertain. Cell-type heterogeneity and dynamic changes in splicing factors further complicate interpretation (MontAñés-Agudo et al., 2023). Third, in vivo functional studies remain limited. While siRNA or ASOs are widely used, they often fail to distinguish circular from linear transcripts. Newer approaches such as circular-specific ASOs (cASOs) and CRISPR-Cas13 have improved specificity, yet challenges remain in delivery efficiency, off-target effects, and immunogenicity (Ai et al., 2022; Shi and Wu, 2024). Finally, the choice of models and endpoints significantly influences extrapolation. STZ, db/db, and high-fat diet (HFD) models differ in their metabolic and inflammatory features, while in vitro hyperglycemia or ROS stimulation poorly recapitulates the complex cardiac milieu. Moreover, many studies assess transcript levels or oxidative stress indices without direct correlation to cardiac structure or function. Collectively, these factors contribute to inconsistent findings across studies, underscoring the need for integration of long-read and single-cell sequencing, targeted interventions, and longitudinal functional evaluation to enhance both reliability and translational relevance (Bibi et al., 2025).

Given the current paucity of circRNA-focused studies in DCM, this review has, in part, referenced findings from other disease models such as diabetic kidney disease (DKD) and myocardial ischemic injury. These conditions share common pathological features with DCM, including chronic hyperglycemia and oxidative stress, which drive mitochondrial dysfunction and excessive ROS production, ultimately triggering apoptosis, autophagy, and inflammatory responses. Moreover, they all involve dysregulation of noncoding RNA networks, including circRNAs, contributing to fibrosis, cell death, and metabolic derangements. These shared mechanisms provide a rationale for cross-model comparison, offering useful insights into the potential roles of circRNAs in DCM.

Nevertheless, caution is warranted when extrapolating from external models to DCM. Pathological contexts differ significantly, shaping both circRNA expression and function. In DKD, the major drivers include glomerular hyperglycemic stress and basement membrane remodeling, whereas ischemia/reperfusion (I/R) injury reflects acute hypoxia-reoxygenation. In contrast, DCM is characterized by chronic hyperglycemia, insulin resistance, metabolic reprogramming, and AGE accumulation. These differences mean that the same circRNA may exhibit distinct phenotypes across tissues or models, depending on variations in cell composition, miRNA/RBP backgrounds, and metabolic state (Liu and Zhao, 2022; Pan M. et al., 2025). DCM also exhibits unique features that limit direct extrapolation, including cardiomyocytes’ high dependency on energy metabolism and calcium homeostasis, the central role of mitochondrial dysfunction, and a prolonged course of metabolic inflammation and fibrosis. These factors may reshape circRNA biology—affecting subcellular localization, miRNA sponging capacity, and RBP interactions—ultimately altering whether their effects in the myocardium are pathogenic or protective (Fang et al., 2025).

To improve extrapolative validity, validation in cardiac-specific contexts is essential. This includes reproducing circRNA expression profiles and functions in myocardial tissue or isolated cardiac cell types (cardiomyocytes, fibroblasts, endothelial, or immune cells); applying single-cell/spatial transcriptomics and long-read sequencing to resolve isoforms and subcellular localization; and employing back-splice junction (BSJ)-specific interventions (e.g., cASOs or optimized CRISPR-Cas13 platforms) delivered via AAV9 or other cardiotropic systems, with rigorous evaluation of dose–response, off-target activity, and immunogenicity (Guan et al., 2025; Wu et al., 2023).

In the future, promoting the clinical application of circRNAs in DCM will require overcoming critical challenges in both human sample research and therapeutic delivery. The procurement of human cardiac tissue is constrained by ethical considerations and limited availability, while serum circRNA levels are susceptible to individual variability and comorbidities, resulting in inconsistent detection reliability (Yan et al., 2024). Therefore, relying solely on differential expression profiles is insufficient to establish causality. Integrating single-cell omics, spatial transcriptomics, and CRISPR-based functional validation technologies will be essential to elucidate the precise role of circRNAs in disease mechanisms.

Simultaneously, the complex cardiac microenvironment imposes stringent demands on therapeutic delivery systems. Current delivery vehicles are generally limited by issues such as immunogenicity, low efficiency, and lack of cell-type specificity. Future efforts should prioritize developing nano-delivery systems with low immunogenicity, controllable release kinetics, and cardiomyocyte-specific targeting to enable efficient and clinically safe translation of circRNA-based therapies.

In summary, circRNAs retain significant promise as both potential biomarkers and therapeutic targets for DCM, underscoring their continued value in further research and clinical translation.