The role of circular RNAs as miRNA sponges in the mechanisms and therapeutic potential of triple negative breast cancer

Triple-negative breast cancer (TNBC) is an aggressive subtype of breast carcinoma which lacks estrogen receptors, progesterone receptors and HER2 along with limited therapeutic options mainly based on chemotherapy. In this review, we outline the emerging function of circRNAs as key regulators in TNBC pathogenesis. CircRNAs are endogenous non-coding RNAs with a closed-loop structure, in contrast to the linear form. In TNBC cells, the underlying molecular mechanism mainly relies on their functions as a competitive sponge of miRNAs, which can absorb or bind to microRNAs (miRNAs) and hence regulate the expression of target genes. Such sponging can result in the activation of oncogenes or repression of tumor suppressor genes, which eventually affect cellular proliferation, apoptosis, and drug sensitivity. Crucial mechanisms include certain circRNAs, such as circEPSTI1, circRAD18, and hsacirc0000199 that enhance tumorigenesis and resistance to chemotherapy by targeting tumor-suppressor miRNAs and activation of oncogenic pathways (e.g., PI3K/Akt/mTOR pathway or Wnt/β-catenin). The potential clinical implications and dysregulation of circRNA-miRNA axes are highlighted, indicating that these may constitute promising diagnostic or prognostic markers by their stability in biofluids. Additionally, this review outlines the innovative treatment approaches regarding these interactions which have recently been addressed and described, novel methods include ASOs therapy, CRISPR/Cas system and nanoplatforms that may help to get over current therapeutic drawbacks in treating TNBC patients.

Highlights

• CircRNAs act as potent miRNA sponges in TNBC, driving tumorigenesis, metastasis, and chemoresistance through oncogenic pathway activation.

• Aberrant circRNA–miRNA axes show strong clinical relevance, offering stable, non‑invasive biomarkers for TNBC diagnosis, prognosis, and patient stratification.

• Emerging therapeutic strategies targeting circRNAs—including ASOs, CRISPR/Cas systems, and nanoplatforms—hold promise to overcome current treatment limitations in TNBC.

Introduction

​Triple-negative breast cancer (TNBC) is precisely characterized by the molecular absence of estrogen receptors (ER), progesterone receptors (PR), and human epidermal growth factor receptor 2 (HER2) expression [1]. This intrinsic lack of key molecular targets means that therapeutic strategies commonly effective in other breast cancer (BC) subtypes, which rely on targeting these specific receptors, are rendered ineffective in TNBC [2]. The non-expression of ER, PR, and HER2 on the cellular surface forces TNBC treatment away from targeted endocrine and HER2-directed therapies, leaving conventional chemotherapy as the primary systemic intervention [3]. This molecular signature contributes significantly to TNBC’s aggressive phenotype, which is clinically manifested by a higher proliferative index and greater metastatic potential compared to receptor-positive BC [4].

There are numerous types of non-coding RNAs (ncRNAs) that do not form peptides, play non-ignorable regulatory roles in gene expression [5]. In particular, microRNAs (miRNAs) are small ncRNAs of approximately 22 nucleotides, which have a central role as post-transcriptional regulators [6]. They do so by base-pairing to matching sequences in messenger RNA (mRNA), leading to mRNA degradation or prevention of protein translation [7]. The regulatory function of miRNAs is extensive, participating in various life activities including cell proliferation, differentiation, apoptosis and metabolism [8]. Dysregulated miRNAs are commonly found in cancer and can serve as oncogenes or tumor suppressors to foster cancer initiation, progression and metastasis [9]. Their power to modulate complex gene networks highlights their relevance in deciphering the regulatory complexity of cellular pathways, being especially important in diseases like TNBC [10].

Circular RNAs (circRNAs), a special class of endogenous ncRNAs, have a closed loop structure that allows them to form the covalently [11]. This circular structure is formed by a back-splicing event, in which the 5’ splice site downstream is linked to the 3’ splice site upstream on a pre-mRNA transcript [12]. It is widely accepted that one of their best-characterized roles implicates them as “miRNA sponges” or “miRNA decoys”, by which the lncRNAs are designed with multiple binding sites complementary to various miRNAs, and succorously bind with them to prevent further interaction between target mRNAs [13]. This sponging effect results in the release of miRNA target genes and thus modulates protein expression [14]. Besides, circRNAs might also sequester RNA-binding proteins, modify transcriptional activities or even translate into proteins themselves, which enable them a greater than expected functional flexibility [15]. (Fig. 1)

Fig. 1.

Schematic overview of non-coding RNA–mediated regulation in triple-negative breast cancer (TNBC). Cancerous cells within TNBC tumors express diverse non-coding RNAs, including circular RNAs (circRNAs) and long non-coding RNAs (lncRNAs), which compete with microRNAs (miRNAs) for binding to target mRNAs. This competitive endogenous RNA (ceRNA) network modulates mRNA availability and gene expression, contributing to cancer-related cellular phenotypes

The focus of this review is to review a complex crosstalk involving circRNAs and miRNAs, and specifically, circRNA-miRNA axes in TNBC. In particular, it seeks to explain the role of such regulatory loops in TNBC aggressiveness and resistance to the current therapies. This review presents a summary of the important circRNAs and miRNAs that are implicated in the development and progression of TNBC and their contribution to the development of the tumor in terms of proliferation, apoptosis, metastasis, and drug resistance. This work summarizes the existing evidence to highlight the clinical importance of circRNA-miRNA interactions as a basis to develop translational studies and therapies in the future.

Biogenesis and properties of circRNAs

Formation and classification

​​ CircRNAs belong to a novel endogenous family of ncRNAs that are hardly expressed and are presented in a more stable fashion than linear RNA molecules because of the lack of free 5’ and 3’ ends [16, 17]​. This special structure makes circRNAs different from regular linear RNAs and is the inherent foundation of their biological functions [18]. The existence of circRNAs was originally seen in plant viroids and later also noticed in eukaryotic cells by electron microscopy over 40 years ago [19]. At first, they were typically thought to be non-functional splicing intermediates or intronic lariats. But since the development of high-throughput sequencing of RNA and computer big-data processing, thousands circRNAs have been discovered in various cell types and organisms [20]. (Fig. 2)

Fig. 2.

Precursor mRNA (pre-mRNA) transcribed by RNA Polymerase II undergoes canonical splicing and back-splicing events in the nucleus. Back-splicing, facilitated by spliceosomal components such as U1 snRNP, leads to the formation of circular RNAs (circRNAs), which lack free 5′ and 3′ ends and are thus resistant to exonuclease-mediated degradation. Following nuclear export, circRNAs accumulate in the cytoplasm where they exert diverse regulatory functions. These include acting as microRNA (miRNA) sponges to modulate miRNA availability and activity, interacting with RNA-binding proteins to influence their localization and function, and in some cases, serving as templates for cap-independent translation

CircRNAs are mainly generated by a non-canonical alternative splicing mechanism called backsplicing [21]. This mechanism is a covalent binding of an introns downstream splice donor site to an upstream splice acceptor site within a pre-mRNA transcript. Back-splicing can be catalyzed by canonical spliceosomal machinery [22]. Multiple mechanisms have been proposed for circRNA biogenesis, such as lariat-driven circularization through exon skipping or direct pairing of a 5’ splice site with an intronic 3’ splice site. Complementary sequences, as in the case of inverted repeat Alu elements in flanking introns, can promote backsplicing by proximal stimulation of splice sites [23]. Certain protein factors, such as RNA-binding proteins (RBPs), also have an important effect in the control of backsplicing efficiency and circRNA generation [24]. For instance, the oncogenic circDOCK1 (2–27) biogenesis is regulated by an epithelial splicing regulator ESRP1, which binds to a GGU-containing repeat of intron 1 and promotes one of its alternative splicing types [25]. Another is NOVA2 which functions as a global circRNA biogenesis factor in the developing brain via binding to YCAY clusters in flanking introns [26].

According to the composition, circRNAs can be generally divided into three groups: exonic circRNA (ecircRNA) which is the most abundant and composed only of exonic sequences usually enriched by exonuclease degradation of linear RNA able to reach very high levels [27]; intronic circRNA (ciRNA) which are originated from introns could escape debranching process after splicing and thus contain 2’-5’ junction remnants from the intron lariat [28]; or circles containing both exonic and intronic sequences such as exon-intron circRNA (EIciRNAs) [29]. Apart from the canonical form, sub-exonic circRNAs and tricRNAs have been recently discovered as relatively less-investigated types of circRNAs [30]. It should be mentioned that from one gene, a number of circRNAs could be derived, with indications that their expression may not necessarily correlate with that of the linear host gene transcript [31].

Stability, localization, and biological roles

Notably, circRNAs are more stable due to their circular conformation compared with the linear RNA sequences. Unlike linear RNAs, whose free 5’ and 3’ ends are easily degradable by exonucleases, circRNAs are not susceptible to exonuclease-mediated degradation [32, 33]. ​ This natural resistance to degradation is what leads the circRNAs to have longer half-lives than their linear siblings, and are very stable species in cells. For instance, circRNAs were more stable compared to their accompanying linear mRNA in vivo. This perennially is a good feature for their application in diagnostics and therapeutics [34]. CircRNAs can be removed from cells through extracellular vesicle co-precipitation, among other methods [35].

The subcellular localization of circRNAs is important in the identification of their functional significance and regulatory characteristics as their sites are responsible for their interactions with molecules [36]. CircRNAs are highly expressed among mammalian cells and tissues, frequently showing a specific cell type-, tissue-, or developmental stage-related expression profile [37]. Numerous circRNAs are enriched in the cytoplasm of eukaryotic cells [38]. Most cytoplasmic circRNAs serve as miRNA sponge or associate with protein [39]. In contrast, ciRNAs and EIciRNAs are mainly in the nucleus [40]. With the help of deep neural network, studies are being carred to make prediction about cell line-specific circRNA subcellular localization.

CircRNAs are involved in a variety of biological processes and may be critical regulators in health states as well as diseases. One of the best characterized roles is their function as sequestering miRNA molecules, which can in turn reduce their capacity to regulate a set of mRNAs through competing endogenous RNA (ceRNA) [41]. For example, circEPSTI1 binds to miR-4753 and miR-6809 as a miRNA sponge to regulate BCL11A expression and affect TNBC proliferation and apoptosis [42]. Moreover, circRNAs are able to bind to RNA-binding proteins (RBPs) and act as scaffolds for RBPs interacting with each other or as decoys blocking the interactions of RBPs with their endogenous targets, which implicates protein locati0n and activity [43]. For instance, circATP5C1 accelerates the progression of TNBC partly via binding with IGF2BP2 to increase the secretion of CSF-1 [44]. Some nuclear circRNAs, such as EIciRNAs also involve in transcription regulation by interacting with RNA polymerase II to increase the expression of their host genes and are believed to function in chromatin modification or DNA methylation [45]. For example, circ-ITCH suppresses the Wnt/β-catenin pathway by absorbing miR-7, miR-17 and miR-214 [46]. Traditionally, circRNAs were regarded as ncRNA molecules; however, recently several circRNAs have also been discovered to encode active peptides and proteins based on cap-independent translation including the IRES (Internal Ribosome Entry Sites), including muscle-differentiation protein from circ-ZNF609 [47]. An example is circ-HER2, which is translated into a 103-amino acid peptide (HER2-103) in TNBC cells, and HER2-103 facilitates the formation of EGFR/EGFR homodimers and EGFR activity [48]. Their diverse functions illustrate their significance in gene regulation and cellular behavior, and the impressive stability of circRNAs along with their high abundance in biofluids makes them promising reagents for disease diagnosis and potential therapeutic targets.

Mechanistic roles of circRNA–miRNA axes in TNBC

​ Dysregulation of these circRNA-miRNA axes is a feature of TNBC, affecting important cellular behaviors, such as proliferation, apoptosis, migration and drug sensitivity [49]. ​ Elucidating these complex interactions is critical for the identification of new therapeutic targets in the progression of TNBC.

Tumorigenesis and cell proliferation

​​ CircRNAs play a vital role in tumorigenesis, and cell proliferation associated with TNBC due to the presence of molecular sponges being formed [50]. (Table 1) ​ This process results in the induction of oncogenes and activation of multiple downstream oncogenic pathways. For instance, circEPSTI1 (hsacircRNA000479) is significantly highly expressed in TNBC tissues and cells, the bound miR-4753 and miR-6809 respectively to mediate BCL11A expression as well as affect pro-proliferative and anti-apoptotic ability [42]. This axis functions through a ceRNA mechanism, affecting the survival of TNBC patients [42]. ​ This axis acts through the ceRNA network, influencing TNBC patient survival [49]. Also, circRAD18 (hsacirc0002453) was greatly upregulated in TNBC and positively related to tumor size, clinical stage, and pathological grade. It sequesters miR-208a and miR-3164 resulting in the upregulation of insulin-like growth factor 1 (IGF1) and fibroblast growth factor 2 (FGF2) expression driving TNBC progression [51].

Table 1.

CircRNAs acting as miRNA sponges in TNBC tumorigenesis and proliferation

CircRNA	Targeted miRNA(s)	Downstream oncogenic pathway/target genes	Functional outcomes in TNBC	Refs.
circEPSTI1	miR-4753, miR-6809	BCL11A expression	Promotes proliferation, inhibits apoptosis	[42]
circRAD18	miR-208a, miR-3164	IGF1 and FGF2 expression	Promotes proliferation, migration, and tumor growth	[51]
hsacirc0000199	miR-206, miR-613	PI3K/Akt/mTOR signaling	Enhances proliferation, migration, invasion, and chemo-tolerance	[52]
circ-ERBB2	miR-136-5p	Pyruvate Dehydrogenase Kinase 4 (PDK4), Warburg effect	Facilitates metabolic reprogramming, proliferation, invasion	[54]
circPLK1	miR-296-5p	PLK1 pathway	Promotes tumor growth and invasion	[109]
circIFI30	miR-520b-3p	CD44	Promotes proliferation and metastasis	[57]
circSEPT9	miR-637	LIF/Stat3 signaling pathway	Enhances proliferation, migration and invasion of TNBC cells	[110]
hsa_circ_0131242	hsa-miR-2682	Not specified	Promotes TNBC progression	[111]

Another functional circRNA, hsa_circ_0000199, is significantly overexpressed in TNBC than non-TNBC tissues. The circRNA targets miR-206 and miR-613, which are neutralized by hsacirc0000199. Silencing hsacirc0000199 increases chemo-sensitivity and suppresses the proliferation, migration, and invasion of TNBC cells via deactivating the PI3K/Akt/mTOR signaling pathway [52]. The Wnt/β-catenin pathway is also facilitated by some circRNAs in TNBC; for example, CircRNA_069718 promotes the proliferation and invasion of TNBC cells through activation of the Wnt/β-catenin pathway to up-regulate β-catenin, c-myc and cyclin D1 [53]. ​ In addition, circ-ERBB2 (hsacirc0007766) is also highly expressed in TNBC and promotes proliferation and invasion through miR-136-5p sponge effect to enhance pyruvate dehydrogenase kinase 4 (PDK4) expression, leading to the development of Warburg effect [54]. ​ In addition, a ceRNA network analysis identified the dysregulated pathways in BC while its genes were enriched for TGF-beta signaling and PI3K-Akt signaling pathways, implicating their significance of pathogenesis [55].

Metastasis and invasion

​ Metastasis of TNBC is a multistep process that frequently includes epithelial-mesenchymal transition (EMT), enhanced cellular migration, and extracellular matrix (ECM) remodeling [50]. (Table 2) ​ In TNBC, circRNA-miRNA interactions are a key part of these processes resulting in an aggressive metastatic behavior. For instance, circRNA_0044556 leads to reduced TNBC cell sensitivity towards adriamycin and enhanced migration by sponging miR-145 and targeting the NRAS proto-oncogene [56]. ​ For instance, the migratory function of circRNA_0044556 on TNBC cells was reversed by overexpression of miR-145 [56]. ​ Another study reports a significantly up-regulated expression of circIFI30 in TNBC tissues and cells, and its high level is positively correlated with the clinical TNM stage and pathological grade [57]. ​ Downregulation of circIFI30 significantly suppresses the migration and invasion ability of TNBC cells, indicating its role in promoting metastasis through circIFI30/miR-520b-3p/CD44 axis [57].

Table 2.

CircRNA–miRNA axes regulating metastasis and invasion in TNBC

CircRNA	Targeted miRNA(s)	Affected EMT or extracellular matrix-related genes/pathways	Functional role in TNBC Metastasis and invasion	Refs.
circRNA_0044556	miR-145	NRAS proto-oncogene	Promotes cell migration, viability, and adriamycin resistance	[56]
circIFI30	miR-520b-3p	CD44	Promotes cell migration, invasion, and metastasis	[57]
circ-ERBB2	miR-136-5p	Pyruvate Dehydrogenase Kinase 4 (PDK4)	Promotes cell migration and invasion by facilitating the Warburg effect	[54]
hsacirc0131242	hsa-miR-2682	Not explicitly specified, but affects cell proliferation and migration	Promotes cell proliferation and migration	[58]
circPLK1	miR-296-5p	PLK1	Facilitates tumor growth and invasion	[109]
circSCYL2	Not specified	EMT-related pathways	Downregulated in TNBC EMT models; overexpression inhibits migration and invasion	[59]
circPSMA1	miR-637	Akt1-β-catenin axis	Facilitates the tumorigenesis, metastasis, and immunosuppression of TNBC	[112]

Circ-ERBB2 is also involved in the aggressive metastasis ability of TNBC; its downregulation inhibits proliferation, migration and invasion of TNBC cells in vitro [54]. ​ In addition, hsacirc0131242 exhibits a significantly high expression in TNBC tissues and cell lines and relates to advanced TN status, poor clinical features [58]. ​Inhibition of hsacirc0131242 suppresses the progression of BC cells, and it acts as a sponge for hsa-miR-2682, thereby promoting cell proliferation and migration [58]. ​On the other hand, certain circRNAs could have a repressive effect on metastasis; for instance, circSCYL2 is down-regulated in BC tissues and cells and its upregulation can suppress cell migration and invasion that might act to hinder EMT advancement [59]. ​ Together, these cases illustrate the complex regulatory loops mediated by circRNAs and miRNAs that determine migratory and invasive properties of TNBC cells, which may be targeted for anti-metastasis therapy.

Chemoresistance

​ Chemotherapy is still a mainstay of treatment in TNBC, but its efficacy is seriously hampered by chemo-resistance acquisition which accounts for high rates of treatment failure and disease recurrence [60]. ​(Table 3) ) circRNAs control resistance to drugs in TNBC by sponging miRNAs that regulate key functions, such as drug efflux, apoptosis and survival signaling [61]. ​ For example, hsacirc0000199 promotes chemo-tolerance in TNBC through regulating the miR-206/miR-613-meditated PI3K/Akt/mTOR pathway [52]. ​Overexpression of hsa_circ_0000199 is associated with the larger size of tumor, advanced TNM class and poor 3-year overall survival, suggesting its effects in chemoresistance induction. Knocking down this circRNA increases the chemo-sensitivity of TNBC patients by upregulating expression of miR-206/miR-613 and downregulating the PI3K/Akt/mTOR pathway [52].

Table 3.

CircRNAs contributing to chemoresistance in TNBC and their mechanisms

CircRNA	miRNA Target(s)	Drug resistance mechanism	Associated chemotherapeutic agents	Clinical outcome correlations	Refs.
hsacirc0000199	miR-206, miR-613	Activates PI3K/Akt/mTOR signaling, promoting proliferation, migration, autophagy, chemo-tolerance	Adriamycin, taxanes	High expression predicts larger tumor size, advanced TNM grade, poor 3-year survival	[52]
circINTS4	miR-129-5p	Upregulates POM121, promoting proliferation, migration, invasion, autophagy, and chemoresistance	Adriamycin (ADR)	Overexpression correlated with ADR resistance in TNBC cells	[62]
circRNA_0044556	miR-145, miR-665	Sponges miRNAs, upregulates NRAS, promotes cell viability, migration, and inhibits apoptosis	Adriamycin (ADM), Paclitaxel (PTX)	Associated with ADM and PTX resistance, malignant progression	[56]
circEPSTI1	miR-4753, miR-6809	Regulates BCL11A expression, affecting proliferation and apoptosis, influencing chemo-resistance	General TNBC chemotherapy	High levels correlate with reduced survival in TNBC patients	[42]

​ Another circRNA, circINTS4, was markedly upregulated in adriamycin (ADR)-resistant TNBC cells. It serves as a ceRNA of miR-129-5p to target gene POM121; thus, it promotes proliferation, migration, invasion and autophagy and induces chemoresistance [62]. ​ Knocking down circINTS4 abrogates ADR-resistant tumor growth in vivo, thus making it a potential target for reversing chemoresistance [62]. ​ Likewise, circRNA_0044556 is upregulated in TNBC tissues and cells, including adriamycin resistance cells. It can also inhibit the TNBC cells’ susceptibility to adriamycin and paclitaxel through sponging miR-145 and miR-665 respectively, and regulating their downstream target genes such as NRAS [56]. ​This circRNA exerts a cell-proliferative, drug resistance- and migration-promoting and apoptosis-restraining activity. CircEPSTI1 could also influence poor chemotherapy response and, in TNBC patients high expression of this circRNA was shown to be associated with decreased survival rates, possibly through effects on apoptosis pathways under BCL11A regulation [42]. ​ These results imply that interfering circRNA-miRNA networks could provide novel avenues to conquer chemoresistance and enhance antitumor treatment in TNBC.

Clinical relevance and therapeutic potential

​ The discovery of aberrant circRNA-miRNA axes in TNBC has provided new opportunities for clinical practice, from diagnostic and prognostic actors to therapeutic targets. Considering the aggressive nature and the paucity of targeted therapies for TNBC, such ncRNAs appear as promising candidates to better manage patients. The specific properties of circRNAs, such as their stability and tissue type-specific expression, make them particularly suitable for clinical applications [63].

Diagnostic and prognostic biomarkers

​ CircRNAs show significant stability and high detectability in multiple biological fluids such as plasma and serum exosomes, indicating that they could be promising non-invasive liquid biopsy markers [64]. ​ They are of covalently closed-loop conformation and resistant to exonuclease-mediated degradation, which may explain their high stability in circulation [65]. ​ This stability enables quantitative detection by techniques such as RT-qPCR, which makes them applicable as biomarkers [66]. ​ Several circRNAs have been found to be associated with TNBC development, metastasis, and prognosis.​.

For example, the overexpression circEPSTI1 in TNBC specimens is an independent prognostic indicator associated with poor survival of patients with TNBC [42]. ​ Likewise, circIFI30 is highly expressed in TNBC and correlates with advanced clinical TNM stage, histological grade, and poor prognosis [57]. ​ In liquid biopsies, circTRPS1 was identified to be higher in cancer plasma when compared with normal controls (showing a tendency towards significance) and it occurred significantly more often in ER-negative BC [66].

​ Other plasma circRNAs such as circVAV3, circDNAH14, and circCDYL were also higher in cancer tissues compared to normal ones, with circERBB2 and circCDYL significantly correlating with metastatic disease [66]. ​A diagnostic model of these three plasma circRNAs (hsacirc0000091, hsacirc0067772 and hsacirc0000512) exhibited high sensitivity and specificity for BC diagnosis, in which the expression level of plasma hsacirc0000091 was significantly associated with axillary lymph node metastasis, TNM stage as well as prognosis [67]. ​ These findings highlight the great potential of circRNAs as reliable diagnostic, prognostic, and predictive biomarkers for TNBC personalized therapy. (Table 4)

Table 4.

CircRNAs as diagnostic and prognostic biomarkers in TNBC

CircRNA	Sample type	Detectability method	Correlation with TNBC progression and patient outcomes	Refs.
circEPSTI1 (hsacircRNA000479)	TNBC tissues	Microarray, ISH, RT-qPCR	Upregulated; high levels correlate with reduced survival and poor prognosis; independent prognostic marker	[42]
circRAD18	TNBC tissues, cell lines	Microarray, RT-qPCR	Upregulated; positively correlated with T stage, clinical stage, pathological grade; independent risk factor	[51]
circIFI30	TNBC tissues, cells	RNA sequencing, RT-qPCR, ISH	Upregulated; positively correlated with clinical TNM stage, pathological grade, and poor prognosis	[57]
circPLK1	TNBC cell lines/tissues	Microarray, RT-qPCR	Upregulated; associated with poor survivals	[109]
hsacirc0000199	TNBC tissues, cell lines	RT-qPCR	Higher in TNBC tissues; predictive of large tumor size, advanced TNM grade, high Ki-67, and poor 3-year survival	[52]
has-circRNA-LPXN	Serum exosomes, TNBC tissues	RNA-sequencing, RT-qPCR	Downregulated; valuable for breast cancer diagnosis and prognosis	[113]
has-circRNA-HSDL2	Serum exosomes, TNBC tissues	RNA-sequencing, RT-qPCR	Upregulated; valuable for breast cancer diagnosis and prognosis, plays vital roles in TNBC development	[113]
circTRPS1, circVAV3, circDNAH14, circCDYL, circERBB2, circWDR78	Plasma	Quantitative PCR	CircTRPS1 higher in cancer plasma, more common in ER-negative BC; circVAV3, circDNAH14, circCDYL elevated in cancer; circCDYL & circERBB2 correlated with metastatic disease	[66]
hsacirc0000091	Plasma	Microarray, RT-qPCR	Correlates with axillary lymph node metastasis, TNM stage, and prognosis; potential prognostic biomarker	[114]
circCFL1	TNBC cells and tissues	Not specified	High expression in TNBC; prognostic potential	[115]
hsa_circ_0044234	TNBC cells and tissues	RT-PCR	Downregulation in TNBC; diagnostic potential	[116]

Therapeutic targeting strategies

​ The complex regulatory functions of circRNA-miRNA pathways in TNBC render them promising candidates for developing novel therapeutic approaches. Recent work has focused on several sophisticated strategies to modulate these interactions to impair tumor growth, metastasis, and chemotherapy resistance. Such strategies involve antisense oligonucleotides (ASOs), CRISPR/Cas systems and next-generation delivery nanoplatforms. (Fig. 3)

Fig. 3.

Schematic overview of emerging therapeutic strategies targeting circRNA–miRNA regulatory axes in Triple-Negative Breast Cancer (TNBC). Approaches such as antisense oligonucleotides (ASOs), CRISPR/Cas systems, and nanoparticle-based delivery platforms are being developed to modulate oncogenic RNA interactions, aiming to suppress tumor growth, reduce metastasis, and overcome chemoresistance. These interventions represent promising avenues for translating molecular insights into effective clinical treatments

Antisense oligonucleotides (ASOs)

​ ASOs are artificial nucleic acid analogs that are developed to bind with a high degree of specificity to target RNA such as cancerous circRNAs or miRNAs and functionally inhibit them. The antagonism of a circRNA by ASOs can relieve the repression on a tumor suppressor miRNA, and recover their suppression function against oncogenes [68]. ​ Alternatively, ASOs may be tailored to block oncogenic miRNAs and consequently abrogate their function of silencing tumor suppressor genes [69]. ​ These successes have led to preclinical findings that ASOs can robustly inhibit miRNA function in mammalian cells and promote inhibition of both proliferation and metastasis [69].

CRISPR/Cas systems

CRISPR/Cas strategies provide robust genome editing platforms that can be modified to direct cleavage and knockout of circRNA or miRNA genes at the DNA level [70]. ​ This permits specific knockout of oncogenic RNA molecules or their transcriptional control to be modulated [71]. ​ For instance, CRISPR/Cas9 systems have been applied to knockout miRNA targets in cell lines with success but there are differences in efficiencies [72]. ​ Progress of these systems is encouraging for in vivo gene editing in TNBC models, particularly when combined with novel delivery mechanisms.

​.

Nanoparticle carriers

One of the major hurdles for ASOs and CRISPR/Cas systems are to specifically and efficiently deliver them to TNBC cells in vivo, especially because TNBC is heterogeneous and resistant to conventional treatment [73]. ​ Hybrid nanoparticles, including lipid-based and hyaluronic acid-based nanomedicines, have also been prepared to improve the delivery process and cellular uptake [74]. ​ For such nanoplatforms, they can be functionalized with targeting ligands in order to precisely target TNBC and BC stem cell-like cells (BrCSCs) for enhancing the accumulation of tumor tissues and decreasing side effects to normal organs. Endo/lysosomal escape and nuclear localization are, moreover, important design requirements for successful delivery [74].

​ Off-target specificity and the risk of potential off-target effects remain a significant concern for RNA-based therapies. ASO design will have to take into account the minimization of nonspecific hybridization on non-target RNAs or unwanted splicing changes [75]. ​ Chemically modified nucleotides and truncated ASOs can also lower infidelity for off target binding or non-specific immunostimulation [75]. ​ In case of CRISPR/Cas systems, careful design of the sgRNA is important to achieve specificity for the target gene, with few off-target cuts in the genome [76]. ​ Bioinformatics programs together with experimental validation, even taking into account the tolerance to mismatch, are necessary for rational design of effective therapeutic agents [77]. ​Although preclinical results are promising, improvements in methods of delivery and improved molecular design will be needed to translate these approaches into effective and safe clinical therapies for TNBC [73].

circRNA–miRNA axes in other breast cancer subtypes (comparative insights)

circRNA–miRNA interactions in HER2 + and ER+ breast cancers

​ In addition to TNBC, circRNA-miRNA regulatory axes have also been reported in ER + and HER2-positive BC and are often woven into receptor-mediated transcriptional and signaling pathways. Dysregulated circRNAs in HER2 + tumors regulate downstream effectors of HER2 amplification which include MAPK and PI3K -like cascades in direct and indirect manners. As an example, circGFRA1 enhances HER2 + tumor progression by absorbing miR-1228 and increasing the expression of AIFM2, which stimulates proliferation and migration. In a similar fashion, circDOCK1 promotes the metastatic spread through the circDOCK1/miR-1385p/GRB7 signaling, one of the pathways tightly interconnected with HER2-dependent signal transduction [78]. Circ-ERBB2 is yet another example of subtype specificity, which enhances the progression of the tumor by regulating transcription factors, including TFAP2C, via miRNA, which supports HER2-linked oncogenic signaling [79]. ER + BC circRNA-miRNA interaction during estrogen-dependent transcriptional regulation is commonly regulated by ER signaling as the primary determinant of the regulatory RNA landscape [80]. CircPGR which is an estrogen induced circRNA is considered a ceRNA that contains regulating cell-cycle advancement and growth through miRNA sequestration [80]. The remarkable plasticity of circPGR expression in ER + tumours, and its sensitivity to endocrine signaling, highlights the incorporation of circRNA miRNA networks into hormone-regulated hierarchies of control [80]. Contrary to this, circRNAs that have tumor suppressive properties like hsacirc0087378 demonstrate the two-way regulation of circRNA [81]. Taken together, these findings imply that circRNA-miRNA axes in ER + and HER2 + BC are major fine-tuning mechanisms of receptor-related pathways, and not direct oncogenic drivers.

Comparative analysis: TNBC vs. other subtypes

Comparative studies show that the unique molecular structures of breast-cancer subtypes determine circRNA-miRNA networks. The tumours of ER + and HER2 + are driven by the predominant receptor-mediated signaling pathways, which coordinate transcriptional programmers and therapeutic sensitivity [82]. By contrast, TNBC do not have estrogen, progesterone or HER2 receptors [83] and have to rely on alternative oncogenic drivers like MYC amplification [84], PI3K/AKT/mTOR, metabolic reprogramming, immune evasion and high cellular plasticity [85]. Such an inherent deviation is potentially going to affect the structure, as well as the functional effect of the circRNA-miRNA regulatory networks. CircRNA-miRNA convergences in TNBC are more often centered on proliferation, epithelial-mesenchymal transition, chemoresistance, and survival signaling pathways, which indicates the lack of receptor-based control mechanisms [86]. CircRNAs herein have a potential role as central modulators that stabilize oncogenic signaling strength or counteract cellular stress response contributing to the aggressive and therapy-resistant phenotype of TNBC [87, 88]. On the other hand, interactions between circRNA and miRNA in ER + and HER2 + subtypes are generally executed via pre-existing receptor-based systems, which tune pathway strength, but not alter signaling orientation [89].

Notably, the current evidence suggests that the circRNA functionality is much more context-specific, with some circRNAs having divergent or even opposite roles across subtypes based on varying miRNA presence, transcriptional regulation and downstream signaling contexts [90]. ER-responsive element-regulated CircRNAs in ER + tumours might have lost significance in TNBC, and CircRNAs facilitating MYC or PI3K-driven programmers in TNBC might have minor effects on luminal disease [91]. However, there is a lack of direct comparative functional research across subtypes, and this is where the niche of knowledge is. On the whole, the relative scarcity of the prevailing receptor signaling in TNBC seems to increase the role of non-receptor regulatory strata, such as circRNA-miRNA interactions. This difference implies that circRNA-miRNA networks can be subtype-specific dependencies in regulation and therapeutic vulnerability especially in TNBC. Further studies using systematic, cross-subtype studies with quantitative and functional validation would be essential in establishing the role of circRNAs as ubiquitous regulators or context-dependent drivers of BC development.

Challenges and future perspectives

Notwithstanding the growth of research of circRNAs along with miRNA in TNBC, there are limited technical and biological limitations which currently restrict their translation potential [92]. Overcoming these challenges will be essential for the effective utilization of circRNA-miRNA networks in the diagnosis and treatment of this aggressive cancer type [92].

Technical and biological limitations

Despite the fact that the circRNA as miRNA sponge paradigm has gained wide scholarly exposure, its biological implications are constrained by quantitative and contextual imperatives that are often not addressed [93]. The high copy numbers of circRNAs required to be effective in miRNA sponging have been demonstrated to be extremely large; they are unlikely to be attained under physiological conditions, and thus ceRNA interactions are not always associated with biologically significant regulation [94]. The sponge activity of many circRNAs, therefore, seems to be context-dependent, with only effects of moderate magnitude, or dependent on the evidence provided by luciferase assays that are insufficient to measure the stoichiometry in vivo and cell complexity [95, 96]. Consequently, research in the future should also go beyond simplified configurations of sponges to absolute quantification techniques and undertaking functional validation of the circRNAs in biologically competent systems in order to define the regulatory effects of circRNAs with more specificity [97].

A major bottleneck is the precise and sensitive identification and quantification of circRNAs, which are frequently less abundant than their cognate linear RNAs, in addition to introducing complications, such as contamination with linear RNAs during RNASeq or reverse-transcription-based quantitative PCR [98]. The low abundance of EV-associated RNAs requires extremely sensitive systems for their detection, whose sensitivity to noise may compromise the reliability of results and may result in an underestimation or even a false negative readout of true expression levels [98]. Rigorous validation is needed to distinguish real circRNAs from artifacts due to technical issues (e.g., trans-splicing or incomplete reverse transcription) [99]. In addition, methods such as RNase R digestion used to enrich circRNAs are not 100% efficient and residual linear RNA contamination or degradation of some circRNAs may affect the profile of detected circRNA [98]. Current sequencing methods are often biased toward polyadenylated, linear RNAs and suboptimal for the detection of non-polyadenylated circRNAs; therefore, specialized bioinformatic analysis is required to identify novel back-spliced junctions, which may lead to false-positive results [98]. RT-qPCR quantification requires the design of divergent primers, which can be difficult due to sequence similarity with linear isoforms or low expression abundance, as well as lack of standardized external controls and normalization methods [98].

The functional validation of specific circRNA-miRNA axes remains complex; precisely demonstrating that a circRNA acts as a sponge for a particular miRNA and that this interaction directly mediates a downstream biological effect necessitates sophisticated experimental designs [5]. These designs include gain- and loss-of-function studies using molecular tools like miRNA mimics, inhibitors, and CRISPR/Cas-mediated circRNA modifications [5]. For instance, ASOs can inhibit oncogenic circRNAs or miRNAs, while overexpression studies can reveal pro-tumorigenic effects, and CRISPR/Cas systems allow for precise knockout or modulation of circRNA/miRNA expression [5]. However, the inherent redundancy within ceRNA networks, where multiple circRNAs or other ncRNAs can sponge the same miRNA, further complicates the attribution of specific functions to individual circRNA-miRNA interactions [6]. This redundancy makes it difficult to isolate the precise functional contribution of any single circRNA-miRNA pair, thereby complicating experimental validation and the development of targeted therapies [6].

Another significant hurdle involves the development of appropriate in vitro and in vivo models that accurately replicate the intricate complexity of TNBC biology and the dynamic interplay of circRNA-miRNA axes [7]. Traditional 2D cell cultures often fail to capture the tumor microenvironment, including interactions with stromal cells, immune cells, and the extracellular matrix, all of which can influence circRNA expression and function [8]. While 3D organoid models and patient-derived xenografts (PDXs) offer more physiologically relevant platforms, they are costly, time-consuming, and still may not fully recapitulate the heterogeneity and metastatic cascade observed in human patients [9]. Moreover, the tissue- and cell-specificity of circRNA expression and function means that findings from one model system may not be universally applicable, necessitating a careful consideration of model selection and extensive validation across diverse contexts [100]. Challenges also exist in the precise identification of target miRNAs for circRNAs, as computational prediction tools, while useful for initial identification, often yield numerous false positives due to accuracy limitations, requiring laborious experimental validation through techniques like RNA immunoprecipitation and luciferase reporter assays [101].

Translational and clinical integration

Translating the promising findings on circRNA-miRNA axes in TNBC from bench to bedside faces substantial regulatory, ethical, and logistical challenges [102]. For circRNAs to be adopted as diagnostic or prognostic biomarkers, standardized protocols for sample collection, processing, and analysis are essential to ensure reproducibility and comparability across different clinical laboratories [103]. Rigorous validation in large, multi-center clinical cohorts is imperative to establish their sensitivity, specificity, and predictive power, as well as to define clinically relevant cut-off values for their expression [104]. Regulatory agencies will require extensive data demonstrating the clinical utility and cost-effectiveness of these biomarkers before they can be routinely implemented in patient management [105]. Ethical concerns are raised about the utility of genetic material in diagnostic use, or in how it pertains to patient privacy and informed consent, especially if germline information is unintentionally disclosed [15]. The management of incidental findings must be approached by appropriate and ethical planning, including clear consent procedures as well as receiving established guidance on disclosure and genetic counseling [106].

With regard to therapeutic approaches, the creation of RNA-targeted therapeutics like ASOs and CRISPR/Cas systems gives rise to a number of intricate technical and safety issues [68]. The effective and safe delivery of these molecules to TNBC cells in vivo is still a challenge: even if nanoparticle carriers seem promising, their stability, targeting specificity, cellular uptake and endosomal escape must be optimized to succeed in delivering the treatment while avoiding side-effects as well as systemic toxicity [73]. These carriers are engineering to include targeting agents in order to optimize tumor accumulation and minimize systemic toxicity although, robust strategies have yet be developed to overcome the specificity issues as well as off-targets effects of ASO but also the surface receptor compatibility when CRISPR/Cas technology is used [73]. The long-term safety studies are also required to assess the possible immunogenicity or cumulative toxicities for repeated administrations. Registration approval of such novel therapies will require extensive preclinical information on pharmacokinetics, pharmacodynamics and toxicology, as well as a stepwise clinical process to evaluate in humans its security, dosage and efficacy. There will also be a need to overcome IP barriers and ensure affordable access to these cutting-edge therapies for their effective clinical translation.

Cooperative initiatives between basic scientists, clinicians, pharmaceutical industry and regulatory authorities will be needed to deal with the many complexities involved in bridging the “bench-to-bedside” gap and therefore for the benefit of TNBC patient’s outcome [107]. Solid collaboration between basic scientists, bioinformaticians, oncologists, pathologists and pharmaceutical companies is essential to provide standardized protocols and more physiologic models [107]. Dedicated funding for translational research is essential to address financial barriers, including the conduct of large-scale clinical validation studies and long-term safety evaluations, and the regulatory process [108]. Active engagement with regulators from a very early stage is critical to preventing any roadblocks in terms of approval processes, whilst partnerships between companies will then allow the sharing of knowledge and resources between those involved in drug development to ensure that basic science findings are constantly tested in their usefulness to patients [108]. ​ Taken together, unravelling the regulatory mechanisms of circRNA-miRNA axes in TNBC will provide a valuable roadmap for advancing translational research and therapeutic success.

Conclusion

​ The systematic study of circRNA-miRNA axes in TNBC emphasizes the vital regulatory functions and great implications for interpreting its aggressive biology. Not only are ncRNAs regulators, they are active players in the major hallmarks of cancer by which cells transform into tumor entities and acquire developing features such as uncontrolled proliferation, escaping apoptosis, gain in metastatic potential and resistance to chemotherapeutic treatment. The remarkable stability and cell/tissue-specific expression profiles of circRNAs also mean that they are among the most promising candidates for new diagnostic and prognostic biomarkers, with potential for early detection and individualized treatment stratification.

The revelation of the basis, where circRNAs sponge distinctive miRNAs to activate cancerous signaling cascades or inactivate tumor-inhibiting ones, has created potential therapeutic approaches. There is intense interest in developing strategies for precise targeting and modulation of these complex interactions using ASOs, gene-editing tools like CRISPR/Cas systems and sophisticated delivery nanoplatforms. These targeted therapies have the capacity to limit tumor growth, metastasis as well as importantly, to overcome chemoresistance in TNBC where a tremendous unmet need exists.

Although from basic science to the clinic there are barriers such as improving the delivery system, verifying its target specificity and pursuing rigorous clinical validation, we have no doubt that circRNA-miRNA axes possess therapeutic promise in TNBC. More investigations and interdisciplinary interactions between basic scientists and clinicians are necessary to convert these promising discoveries into clinical applications that prove effective, safe and have a direct impact on survival benefit and life quality in patients afflicted with TNBC.

Author contributions

The study was conceptualized by SM-D. Data compilation and organization were undertaken by SHK, AK-D and PA. AA and SH contributed to the design of the figures. MMP provided critical revisions to the draft, while AG and SH was responsible for its authorship. Throughout the process, SM-D offered consistent supervision. All authors have reviewed and approved the final version of the manuscript. The authors declare that all data were generated in-house, that no paper mill was used and that no AI tool has been used for the generation of text or figures.

Funding

This study is financially supported by the Tabriz University of Medical Sciences. (Grant Number: 73169)

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Competing interests

The authors declare no competing interests.