This article is mentioned in:

Introduction

Cancer is the second leading cause of global mortality and a major modern public health issue. In 2022, ~20 million new cases and 9.7 million cancer-related deaths were reported worldwide (1). Drug resistance remains a major obstacle to extending the overall survival of patients with cancer, contributing to >90% of cancer mortality (2,3). The mechanisms underlying drug resistance include the inhibition of programmed cell death, aberrant metabolism, immune evasion and cancer stem cell (CSC) maintenance. These processes are driven by multiple factors, such as hypoxia and epigenetic regulation (4-8). While previous research has primarily focused on elucidating epigenetic genes and signaling pathway dysregulation in cancer cells (9,10), evidence has highlighted the critical role of the tumor microenvironment (TME) in drug resistance (11,12). The TME is highly complex, consisting of cancer, immune, stromal and endothelial cells, pericytes, and the extracellular matrix (ECM) (13). Cancer-associated fibroblasts (CAFs) are essential stromal cells within the TME, and regulate the TME through paracrine signaling, including the secretion of cytokines, chemokines, growth factors and exosomes, leading to the remodeling of the TME (14,15). This process promotes connective tissue proliferation, angiogenesis and metabolic processes, and intricately interacts with the signaling pathways of cancer cells, resulting in drug resistance (16,17). Furthermore, CAFs induce the formation of immune barriers by influencing the activity of immune cells directly and upregulating immune checkpoint molecules on the cancer cell surface (18,19).

Exosomes, which are membrane vesicles formed by the shedding of the cytoplasmic membrane, serve as crucial mediators for intercellular communication between CAFs and other cells within the TME (20). Non-coding RNAs (ncRNAs), such as long non-coding RNAs (lncRNAs), circular RNAs (circRNAs/circs) and microRNAs (miRNAs/miRs), are non-protein-coding transcripts and are evolutionarily conserved functional components of exosomes (21,22). miRNAs (17-25 nucleotides) bind to the 3′ untranslated region (UTR) of the target mRNA via the seed sequence, leading to translational repression or mRNA degradation (23). lncRNAs (>200 nucleotides) and circRNAs (covalently closed loop) regulate gene expression by interacting with transcription factors (TFs), RNA binding proteins (RBPs) or miRNAs (24,25). Ample evidence has demonstrated that these ncRNAs critically regulate cellular physiological processes and the pathogenesis of diseases, including cancer (21-25).

Recent studies have shown that exosomal ncRNAs act as key regulators, mediating crosstalk between tumor cells and CAFs, tumor-associated macrophages (TAMs), T cells, neutrophils and other cells to regulate tumor growth, metastasis, angiogenesis, immune responses and drug resistance (26-30). For example, M2 polarization of macrophages promotes tumor progression, and exosomal ncRNAs recruit macrophages in the TME and induce M2 polarization by targeting signaling pathways such as the NF-κB or STAT3 signaling pathways (31-33). Exosomal ncRNAs also serve a crucial role in the communication between CAFs and the TME, whereby ncRNAs can alter the phenotype of CAFs. In non-small cell lung cancer (NSCLC), circNOX4 promotes the production of IL-6 via the miR-329-5p/fibroblast-associated protein (FAP) axis, inducing the transition of normal fibroblasts (NFs) to the inflammatory CAF (iCAF) phenotype (34). Furthermore, several publications have demonstrated that CAF-derived exosomes promote the growth and metastasis of colorectal cancer (CRC), head and neck cancer, breast cancer (BC) and melanoma, and are involved in the development of drug resistance (35-38). Recent studies have shown that miRNAs derived from CAFs decrease lipid reactive oxygen species (ROS) levels and inhibit ferroptosis in cancer cells by targeting ChaC glutathione specific γ-glutamyl cyclotransferase 1 and long-chain acyl-CoA synthase 4 (ACSL4), leading to chemotherapy resistance (39,40). Additionally, ncRNAs secreted by CAFs induce stemness in bladder cancer cells by activating the STAT3 and mTOR signaling pathways, thereby creating supportive regions or conditions required for CAFs to exert their biological functions for chemoresistance in bladder cancer (41). Although these studies have revealed the potential role of exosomal ncRNAs in CAF-mediated tumor resistance, there is a lack of systematic elucidation of their core molecular mechanisms and clinical application potential across different tumor types, making it difficult for researchers to achieve a comprehensive understanding. The present review highlights the pivotal role of CAFs in cancer drug resistance. Based on current evidence, to the best of our knowledge, the present review represents the first systematic elaboration of the potential mechanisms through which exosomal ncRNAs mediate resistance to cancer therapies via bidirectional regulation between CAFs and tumor cells. Furthermore, the present review discusses promising research avenues and future directions. Notably, the present review discusses the clinical applications and challenges of exosomal ncRNAs as mediators of CAF-driven cancer resistance, and that targeting CAF-associated exosomal ncRNAs to modify the drug resistance-associated TME holds notable promise for overcoming resistance to chemotherapy, targeted therapy and immunotherapy. This approach thereby offers innovative strategies for the treatment of patients with therapy-resistant cancer in the future.

Origin, heterogeneity and function of CAFs

CAFs are typically elongated cells that lack the mutations commonly found in cancer cells and do not express markers indicating epithelial, endothelial or leukocyte lineages (42). CAFs interact with other cells in the TME through the secretion of metabolites, cytokines, chemokines and exosomes (43), serving a role in tumor cell proliferation and metastasis (44). CAFs are abundantly present in the TME of almost all solid tumors, actively participating in tumor growth and metastasis (42). This section focuses on the heterogeneity and plasticity of CAFs, and discusses their role in cancer (Fig. 1).

Figure 1 Role of CAFs in cancer. CAFs originate from various cell types, including tissue-resident fibroblasts, mesenchymal stem cells, endothelial cells, epithelial cells, adipocytes, smooth muscle cells and pericytes. Differentiated CAFs facilitate tumor matrix remodeling, promote growth, enhance metastasis, influence metabolism and contribute to drug resistance through the secretion of cytokines, chemokines, exosomes and metabolites. The figure was created by Adobe Illustrator CC 2019 (https://adobe.com/products/illustrator.html). CAF, cancer-associated fibroblast; CCL2, C-C motif chemokine ligand 2; CXCL12, C-X-C motif chemokine ligand 12; ECM, extracellular matrix; EMT, epithelial-mesenchymal transition; GAS6, growth arrest specific 6; HGF, hepatocyte growth factor.

Origin, heterogeneity and plasticity of CAFs

The current consensus is that CAFs originate from locally resident fibroblasts activated by various signaling molecules in the TME (42,45). These signaling molecules include TGF-β, hepatocyte growth factor (HGF), platelet-derived growth factor (PDGF), fibroblast growth factor 2, stromal derived factor-1 (SDF-1) and ROS (18). Wu et al (16) summarized the signaling pathways that promote the transformation of fibroblasts into CAFs, including the TGF-β, PI3K/AKT/mTOR and MAPK pathways. Additionally, substantial evidence suggests that CAFs can also differentiate from other cell types. Mesenchymal stem cells (MSCs) can differentiate into CAFs. In BC, bone marrow-derived MSCs are recruited to the TME and differentiate into PDGF receptor (PDGFR) α− CAFs (46,47). Furthermore, endothelial cells (48), epithelial cells (49), adipocytes (50), smooth muscle cells (51) and pericytes (52) are also potential sources of CAFs.

CAFs from different origins exhibit distinct functional roles in tumors. For instance, CAFs differentiated from bone marrow MSCs promote the growth and doxorubicin resistance of acute lymphoblastic leukemia by upregulating TGF-β (53). Endothelial-like and adipocyte-differentiated CAFs enhance cancer invasiveness and metastasis (54,55), whereas mesothelial cell-derived antigen-presenting CAFs induce the differentiation of CD4+ T cells into regulatory T cells (56). These diverse functions highlight the heterogeneity of CAFs. The heterogeneity of CAFs is a hot research topic, and various experts in this field have provided overviews of CAF heterogeneity, aiming to distinguish subtypes of CAFs with different functions (57-59). Based on their functions, CAFs express various markers such as PDGFRα/β, α smooth muscle actin (α-SMA), FAP and fibroblast-specific protein-1 (60). However, it is difficult to establish a unified standard for CAF subtypes. On one hand, CAFs lack specific surface markers (61). α-SMA is widely expressed in smooth muscle cells (62), while FAP can also be detected in macrophages (63). On the other hand, the markers of CAFs lack consistent expression patterns, leading to a vast number of possible marker combinations (64). The consensus is that there are two CAF subtypes: Myofibroblastic CAFs (myCAFs) and iCAFs (42). myCAFs are characterized by the expression of collagen genes and ECM proteins, participating in matrix remodeling and cellular invasion. iCAFs exhibit high expression of cytokine and chemokine genes, promoting a local inflammatory environment in tumors (65,66). In pancreatic cancer, myCAFs with high expression levels of α-SMA and MMP11 are located closer to cancer cells and exhibit a contractile phenotype that produces the ECM, while iCAFs, which are located away from cancer cells, express IL-6 and angiotensin II type 1 receptor, and serve a key role in shaping an immunosuppressive microenvironment (67,68). The heterogeneity of iCAFs and myCAFs is illustrated in Fig. 2 (42,65-68). In recent years, with the application of single-cell and spatial transcriptomics, the heterogeneity of CAFs has been further untangled (69,70). CAFs can be further classified into multiple distinct subtypes based on their differential surface markers. Although the categorization of these subtypes has not yet achieved widespread acceptance, it has deepened the understanding of CAF heterogeneity. For instance, Cords et al (71) classified CAFs into eight categories through pan-cancer single-cell sequencing: myCAFs (high expression of MMP11 and collagen type I α2 chain), iCAFs (high expression of IL-6 and CD34), vascular CAFs (high expression of NOTCH3, collagen type XVIII α1 chain and melanoma cell adhesion molecule), tumor-like CAFs (expressing VEGFA, CD10 and transmembrane protein 158), antigen-presenting CAFs [expressing human leucocyte antigen (HLA)-DRA, HLA-DRB1 and CD74], and the relatively rare interferon-responsive CAFs, reticular-like CAFs and dividing CAFs.

Figure 2 Heterogeneity of CAFs. CAFs can be classified into two classic subtypes: iCAFs and myCAFs. These two distinct subtypes of CAFs express different markers. iCAFs promote local inflammation in the tumor microenvironment, while myCAFs facilitate ECM deposition and modulation. The figure was created by Adobe Illustrator CC 2019. CAF, cancer-associated fibroblast; ECM, extracellular matrix; iCAF, inflammatory CAF; myCAF, myofibroblastic CAF.

The goal of antitumor therapy based on targeting of CAFs is to reduce the CAFs that promote tumorigenic phenotypes. Therefore, researchers have attempted to determine whether CAF-related molecules can be targeted to transform the phenotype of CAFs (57). Hypoxia-induced hypoxia-inducible factor 1-α (HIF-1α) promotes an increase in the number of α-SMA−/α-SMA-low iCAFs in pancreatic ductal adenocarcinoma (PDAC), while the differentiation of α-SMA+ myCAFs relies on HIF-2α (72). TGF-β inhibits the activation of the janus kinase (JAK)/STAT signaling pathway by suppressing the expression of IL-1 receptor type 1, thereby inhibiting the differentiation of iCAFs (73). These results suggest the plasticity of CAFs.

Role of CAFs in tumor development and progression

CAFs exert a marked effect on various aspects of the TME through the release of exosomes, cytokines, inflammatory ligands and ECM proteins. These actions ultimately manifest as proliferation, invasion and drug resistance in tumor cells (14,15).

The ECM is an important component of the TME and is considered to be a key regulatory factor in promoting tumor malignancy and invasion (74). The ECM is a non-cellular component present in all tissues, composed of collagen, fibronectin, laminin, glycosaminoglycans and proteoglycans (75). The ECM also interacts with factors such as EGF, TGF-β, HGF and VEGF, participating in tumor growth and angiogenesis (76). One of the most prominent abilities of CAFs is the remodeling of the ECM, which is attributed to their high expression of MMP (34), transglutaminase, lysyl oxidase and integrin α11β1 (77,78). ECM remodeling refers to the alteration of the synthesis, degradation and components of the ECM. CAFs alter the density and composition of the ECM, leading to its stiffening and degradation (79). In the TME, the binding of TGF-β to CAF surface receptors activates Smad2/3 and collagen type I α1 chain, thereby enhancing CAF synthesis of collagen, leading to collagen deposition (80). Additionally, the crosslinking enzymes produced by CAFs, in conjunction with force-mediated ECM remodeling, further increase the stiffness of the ECM (42). Higher ECM stiffness is associated with greater tumor invasiveness (81) as it promotes the release of the TF twist family bHLH transcription factor 1 (TWIST1) from the cytoplasmic binding protein G3BP stress granule assembly factor 2 (G3BP2) into the nucleus. The absence of G3BP2 can result in sustained nuclear localization of TWIST1, which, together with increased matrix stiffness, induces epithelial-mesenchymal transition (EMT) (82). Additionally, the stiffening of the ECM can affect the ability of drugs to penetrate cancer cells (83). In addition to increasing ECM stiffness, the MMPs secreted by CAFs lead to ECM degradation, creating pathways within the ECM that promote cancer cell proliferation and invasion, while also providing space for abundant angiogenesis (84-86). Furthermore, the ECM remodeling induced by CAFs maintains an inflammatory state in the TME and prolongs the release of proinflammatory factors such as IL-6, IL-8 and leukemia inhibiting factor (LIF) from cancer cells (87,88). ECM remodeling also affects the migration of infiltrating leukocytes, contributing to immune evasion by cancer cells (89).

CAFs secrete various cytokines, such as TGF-β, growth differentiation factor and IL-11, and activate the P38/MAPK, Smad and STAT3 signaling pathways on the surface of tumor cells to promote tumor cell proliferation and invasion (16). Tumor-promoting CAFs highly express integrin subunit β2, which activates the PI3K/AKT/mTOR signaling pathway through the induction of NADH oxidation, thereby facilitating cancer cell proliferation (90). CAFs also secrete exosomes carrying ncRNAs, which serve a profound role in the communication between CAFs and cancer cells by targeting multiple physiologically active mRNAs in cancer cells (91). In addition, CAFs are the primary source of the chemokine C-X-C motif chemokine ligand 12 (CXCL12) in the TME. The binding of CXCL12 to C-X-C motif chemokine receptor 4 (CXCR4) induces an increase in the number of regulatory T cells and a decrease in the number of CD8+ T cells, thereby suppressing the immune microenvironment (92,93). Simultaneously, CXCL12 impedes the recognition of cancer cells by anti-cytotoxic T lymphocyte-associated protein 4 and α-programmed cell death 1 ligand 1, thereby inducing tumor immune evasion (94). Additionally, periostin-positive myCAFs induce macrophage secreted phosphoprotein 1 expression through the IL-6/STAT3 axis and hinder effective T-cell infiltration and the immune response (95,96). Furthermore, crosstalk between CAFs and cancer cells is one of the main mechanisms of cancer metabolic reprogramming. CAF-derived exosomes and metabolic factors such as HIF-1α regulate glycolysis, and amino acid and lipid metabolism in cancer cells by modulating intracellular metabolic pathways (97). This process contributes to cancer growth, metastasis and drug resistance (97). The metabolic secretions of CAFs themselves also mediate tumor development. For instance, lactate produced by CAFs can enhance the expression of discs large homolog 5 in oral squamous cell carcinoma (OSCC) and activate the Hippo signaling pathway, promoting the stem cell properties of OSCC (98). CAF-derived amino acids provide essential nutrients and energy for cancer growth and invasion (99).

In addition, ECM synthesis and remodeling generate solid stress, leading to the compression of blood and lymphatic vessels. This exacerbates hypoxia in the TME, which promotes angiogenesis and transformation through the formation of new blood vessels (100,101). CAFs also secrete various angiogenic molecules, such as VEGF, PDGF, CXCL12 or HGF, to induce angiogenesis in the TME (102,103). The angiogenesis induced by CAFs not only increases the invasive capacity of cancer cells but also represents one of the contributing factors underlying the resistance of cancer cells to anti-EGFR drugs (104).

Exosomal ncRNAs from tumor cells regulate the activation and metabolism of CAFs

Under the stimulation of various factors in the TME, NFs undergo transformation into CAFs, acquiring enhanced proliferative, invasive and metabolic capabilities (105). Notably, this transition equips CAFs with functions that allow them to participate in tumor progression (45). Factors involved in the activation of CAFs include TGF-β, fibroblast growth factor, PDGF, Sonic hedgehog, IL-6, IL-1R and ncRNAs (16,91). These molecular factors activate various signaling pathways within the CAFs, such as the JAK/STAT and EGFR/erb-b2 receptor tyrosine kinase 2 signaling pathways, thereby shaping the heterogeneity of the influence of CAFs on tumor progression (73,106,107). Recent evidence suggests that dysregulation of ncRNAs within CAFs may represent an important mechanism mediating changes in their functional phenotype (108). Biologically functional ncRNAs are abundantly present in exosomes. They are transferred between cells via exosomes and serve a crucial role in cancer growth and drug resistance (109). Notably, CAF-mediated drug resistance often occurs following ncRNA dysregulation. For example, miR-9 secreted by triple-negative BC cells induces the differentiation of tumor-promoting CAFs by targeting EGF containing fibulin extracellular matrix protein 1 (110). miRNAs originating from cancer cells can also stimulate the expression of SDF-1, MMP3, MMP9, PDGF and chemokine CC motif ligand 7 by CAFs (111,112). These findings indicate that the activation of CAFs by exosomal ncRNAs induces drug resistance. Therefore, when discussing the role of ncRNAs in CAF-mediated drug resistance, it is essential to consider the activating effects of exosomal ncRNAs on CAFs (Fig. 3).

Figure 3 Exosomal ncRNAs secreted by tumor cells activate CAFs. These ncRNAs are taken up by NFs, mediating the activation of multiple signaling pathways that induce the conversion of NFs into CAFs. Furthermore, exosomal ncRNAs enhance the ability of CAFs to secrete cytokines and chemokines, thereby contributing to tumor progression. ↓, promotion; ⊥, inhibition. The figure was created by Adobe Illustrator CC 2019. B4GALT3, β-1,4-galactosyltransferase 3; CAF, cancer-associated fibroblast; CXCL12, C-X-C motif chemokine ligand 12; IGF2R, insulin-like growth factor 2 receptor; JAK2, janus kinase 2; miR, microRNA; ncRNA, non-coding RNA; NF, normal fibroblast; SOCS1, suppressor of cytokine signaling 1; TP53INP1, tumor protein P53 induced nuclear protein 1; TXNIP, thioredoxin interacting protein; ZBTB2, zinc finger and BTB domain containing 2.

Exosomal ncRNAs activate CAFs

Numerous studies have demonstrated that ncRNAs derived from cancer cells serve as critical regulators mediating the differentiation of NFs into CAFs. Treatment of NFs with gastric cancer (GC)-derived exosomes enhances their proliferative capacity and the expression of α-SMA and Vimentin (113). The expression of long intergenic non-protein coding RNA 691 (LINC00691) is increased in NFs, augmenting the ability of NFs to promote cancer cell proliferation and invasion. Subsequent experiments have demonstrated that LINC00691 promotes the differentiation of NFs into CAFs by activating the JAK2/STAT3 signaling pathway (113). Pang et al (114) isolated primary pancreatic fibroblasts from C57 mice and co-cultured them with the BxPC-3 and SW1990 pancreatic cancer cell lines. The results indicated that miR-155 secreted by pancreatic cancer cells in exosomes promoted the transformation of NFs into α-SMA+ and FAP+ CAFs by targeting tumor protein P53 induced nuclear protein 1 (TP53INP1). TP53INP1 is an important functional component of p53, and its inhibition leads to the inactivation of p53, resulting in the activation of CAFs (115,116). Based on clinical data, there is a positive association between exosomal miR-1247-3p and the malignancy and metastasis of hepatocellular carcinoma (HCC). This is due to the fact that exosomal miR-1247-3p secreted by HCC cells targets β-1,4-galactosyltransferase 3, facilitating the differentiation of CAFs and activating their β1-integrin-NF-κB signaling pathway (117). Activated CAFs secrete IL-6 and IL-8, which contribute to the progression of HCC (117). Notably, another study found that miR-21 secreted by HCC cells activated the pyruvate dehydrogenase kinase 1/AKT signaling pathway by targeting PTEN. This process induced the transformation of hepatic stellate cells into CAFs (118). Exosomal miR-21 has also been demonstrated to serve a role in head and neck squamous cell carcinoma (HNSCC); it induces the differentiation of NFs into CAFs by targeting YOD1, thereby promoting lymph node metastasis in HNSCC (119). The lncRNA Gm26809 and miR-211 are upregulated in exosomes from melanoma cells, inducing the formation of CAFs and increasing the expression of CAF markers (α-SMA and FAP) (120). Furthermore, miR-211 targets insulin-like growth factor 2 receptor and leads to the activation of the MAPK signaling pathway in CAFs, which results in increased proliferation and migration, and upregulation of pro-inflammatory genes in CAFs (121). BC cell-secreted exosomes contain various miRNAs, including miR-125b, miR-146a and miR-9, which are transferred to NFs and promote their differentiation into the CAF phenotype (122-124). For example, miR-125b, as an inhibitor of TP53INP1, elevates the expression of α-SMA, MMP and cytokines in NFs (122). Furthermore, NFs differentiate into CAFs following miR-9 uptake, enhancing their proliferation and migration. These CAFs also secrete miR-9, which targets BC cells, forming a positive feedback loop. The miR-9 secreted by CAFs promotes BC metastasis by reducing E-cadherin expression (124). The specific mechanisms of exosomal ncRNAs regulating CAF activation are described in Table I. Activated by these cancer cell-derived ncRNAs, CAFs secrete abundant angiogenic factors, IL-6, TGF-β, TNF-α and CXCL12, which act on cancer cells to promote their proliferation, invasion and drug resistance (118,125,126).

Table I Role of tumor cell-derived exosomal ncRNAs in CAF activation and metabolism.

Table I

Role of tumor cell-derived exosomal ncRNAs in CAF activation and metabolism.

First author/s, year	Origin	ncRNA	Molecular targets	Function	(Refs.)
Xia et al, 2023	GC	LINC00691	JAK2/STAT3	Induces the differentiation of NFs into CAFs	(113)
Pang et al, 2015	Pancreatic cancer	miR-155	TP53INP1	Induces the differentiation of NFs into CAFs	(114)
Fang et al, 2018	HCC	miR-1247-3p	B4GALT3	Induces the differentiation of NFs into CAFs	(117)
Zhou et al, 2018	HCC	miR-21	PTEN	Induces the differentiation of NFs into CAFs	(118)
Ye et al, 2023	Head and neck squamous cell carcinoma	miR-21	YOD1	Induces the differentiation of NFs into CAFs	(119)
Hu and Hu, 2019	Melanoma	Gm26809	Not clear	Induces the differentiation of NFs into CAFs	(120)
Dror et al, 2016	Melanoma	miR-211	IGF2R	Induces the differentiation of NFs into CAFs	(121)
Vu et al, 2019	BC	miR-125b	TP53INP1	Induces the differentiation of NFs into CAFs	(122)
Yang et al, 2020	BC	miR-146a	TXNIP	Induces the differentiation of NFs into CAFs	(123)
Baroni et al, 2016	BC	miR-9	Not clear	Induces the differentiation of NFs into CAFs	(124)
Tong et al, 2020	ESCC	POU3F3	Not clear	Induces the differentiation of NFs into CAFs	(125)
Wang et al, 2022	CRC	miR-155-5p/miR-146a-5p	SOCS1 (miR-155-5p) and ZBTB (miR-146a-5p)	Induces the differentiation of NFs into CAFs	(126)
Yan et al, 2018	BC	miR-105	Myc	Promotes metabolic reprogramming	(131)

[i] GC, gastric cancer; LINC00691, long intergenic non-protein coding RNA 691; JAK2, janus kinase 2; NF, normal fibroblast; TP53INP1, tumor protein P53 induced nuclear protein 1; HCC, hepatocellular carcinoma; B4GALT3, β-1,4-galactosyltransferase 3; IGF2R, insulin-like growth factor 2 receptor; TXNIP, thioredoxin interacting protein; POU3F3, POU class 3 homeobox 3; SOCS1, suppressor of cytokine signaling 1; ZBTB2, zinc finger and BTB domain containing 2; BC, breast cancer; CAF, cancer-associated fibroblast; CRC, colorectal cancer; ESCC, esophageal squamous cell carcinoma; miR, microRNA; ncRNA, non-coding RNA.

Exosomal ncRNAs induce the glycolysis of CAFs

CAFs maintain their metabolic characteristics of producing lactate and pyruvate through the epigenetic regulation of HIF-1α and glycolytic enzymes (127). This phenomenon is termed the 'reverse Warburg effect' in CAFs, where myCAFs shift to aerobic glycolysis, generating copious amounts of lactate and pyruvate (128). The lactate produced by CAFs is taken up and metabolized by cancer cells to generate NADH, ultimately producing ATP to support cancer cell proliferation and invasion (90). On the other hand, the lactate derived from glycolysis in cancer cells within the TME forms a positive feedback loop with CAFs. The lactate released by cancer cells stimulates the upregulation of IL-6 and collagen triple helix repeat-containing 1 expression in CAFs, which further enhances glycolysis in cancer cells and suppresses the immune microenvironment (129,130). A previous study has reported the role of tumor cell-secreted exosomal ncRNAs in the metabolic reprogramming of CAFs (Table I). In BC, secreted miR-105 induces the metabolic reprogramming of CAFs by targeting the Myc signaling pathway. When nutrients are sufficient, miR-105-reprogrammed CAFs enhance glucose and glutamine metabolism to fuel cancer cells. Under conditions of nutrient deprivation and accumulation of metabolic byproducts, CAFs convert metabolic waste such as lactate and ammonia into high-energy metabolites for detoxification (131). Recently, He et al (132) identified that circABCC4 induces glycolysis in CAFs by enhancing the interaction between pyruvate kinase M1/2 and karyopherin subunit α2, thereby promoting increased IL-8 secretion by CAFs and leading to oxaliplatin resistance in pancreatic cancer. However, to the best of our knowledge, existing evidence has not confirmed whether circABCC4 is delivered to CAFs via exosomes derived from pancreatic cancer cells (132).

CAF-derived exosomal ncRNAs promote drug resistance in cancer

The development of drug resistance in cancer is not only determined by intrinsic factors but also influenced by the complex regulatory network among cells in the TME. CAFs are the most crucial cells involved in the interaction between tumor cells and the stroma, and targeted therapy against CAFs holds great promise for improving cancer drug resistance (133). The use of chemotherapy drugs induces the transformation of NFs into CAFs, which secrete large quantities of cytokines and exosomes (134). It has been suggested that CAF-derived exosomes are ingested by cancer cells, mediating drug resistance in cancer cells by activating drug resistance-related pathways, regulating programmed cell death, EMT, drug transporters and immune evasion (135). These mechanisms are described in Table II and Fig. 4.

Figure 4 Exosomes derived from CAFs serve a crucial role in regulating drug resistance-related signaling pathways, apoptosis, ferroptosis, EMT, drug transporters and immune evasion in tumor cells. ncRNAs in CAFs are abnormally expressed and are delivered to tumor cells via exosomes. Once inside the tumor cells, these ncRNAs exert biological functions by targeting critical components involved in the regulation of signaling pathways, cell death, EMT, drug transporters and immune evasion, ultimately contributing to tumor drug resistance. ↓, promotion; ⊥, inhibition. The figure was created by Adobe Illustrator CC 2019. AFAP1-AS1, actin filament associated protein 1 antisense RNA 1; ANRIL, antisense non-coding RNA in the INK4 locus; CAF, cancer-associated fibroblast; CCAL, colorectal cancer-associated lncRNA; circ, circular RNA; DLEU1, deleted in lymphocytic leukemia 1; EMT, epithelial-mesenchymal transition; HLA, human leucocyte antigen; LINC00355, long intergenic non-protein coding RNA 355; miR, microRNA; ncRNA, non-coding RNA; PD-1, programmed cell death protein 1; PD-L1, programmed death-ligand 1; ROS, reactive oxygen species; TME, tumor microenvironment; UPK1A-AS1, uroplakin 1A antisense RNA 1.

Table II Cancer-associated fibroblast-derived exosomal ncRNAs promote tumor therapy resistance.

Table II

Cancer-associated fibroblast-derived exosomal ncRNAs promote tumor therapy resistance.

First author/s, year	ncRNA	Molecular targets	Function	Cancer type	Drugs	(Refs.)
Ren et al, 2018	H19	miR-141/β-catenin	Activates Wnt/β-catenin, and promotes stemness and drug resistance	CAC	Oxaliplatin	(150)
Deng et al, 2020	CCAL	HuR	Activates Wnt/β-catenin and promotes drug resistance	CRC	Oxaliplatin	(151)
Shan et al, 2021	miR-148b-3p	PTEN	Activates Wnt/β-catenin, and promotes EMT and drug resistance	Bladder cancer	Docetaxel and paclitaxel	(152)
Fang et al, 2022	miR-1323	PABPN1	Activates Wnt/β-catenin, and promotes cell proliferation, migration and invasion	CCa	Radiation	(154)
Zhuang et al, 2023	miR-146-5p	ARID1A/SOCS1	Activates STAT3 and promotes stem cell characteristics	Bladder cancer	Gemcitabine and cisplatin	(41)
Sun et al, 2024	miR-296-3p	SOCS6	Activates STAT3, and promotes cell proliferation, migration and invasion	Ovarian cancer	Paclitaxel	(162)
Zhao et al, 2021	miR-21	/	Activates STAT3	ESCC	Cisplatin	(163)
Deng et al, 2024	miR-21	/	Activates RAS/AKT/ERK	Pancreatic cancer	Gemcitabine	(174)
Yeung et al, 2016	miR-21	APAF1	Inhibits apoptosis	Ovarian cancer	Paclitaxel	(193)
Shi et al, 2022	miR-20a	PTEN	Activates PI3K/AKT and promotes drug resistance	NSCLC	Cisplatin	(169)
Sun et al, 2024	miR-20	RB1	Reduces cell cycle arrest	BC	CDK4/6i	(181)
Qu et al, 2022	circN4BP2L2	EIF4A3	Activates PI3K/AKT, and promotes stem cell characteristics and drug resistance	CRC	Oxaliplatin	(170)
Zhang et al, 2023	miR-1228-3p	PLAC8	Activates PI3K/AKT and promotes drug resistance	HCC	Sorafenib	(173)
Shan et al, 2020	miR-423-5p	GREM2	Activates TGF-β and promotes drug resistance	Prostate cancer	Paclitaxel	(177)
Richards et al, 2017	miR-146a	/	Promotes proliferation and drug resistance	Pancreatic cancer	Gemcitabine	(180)
Fang et al, 2019	miR-106b	TP53INP1	Promotes proliferation and drug resistance	Pancreatic cancer	Gemcitabine	(182)
Pan et al, 2022	miR-181d-5p	NCALD	Promotes drug resistance	CRC	5-FU	(183)
Yang et al, 2024	circ0067557	Lin28	Promotes drug resistance	CRC	5-FU	(184)
Gong et al, 2023	miR-200b-3p	ZEB1 and E2F3	Inhibits cell invasion, stemness maintenance and drug resistance	CRC	5-FU	(187)
Kang et al, 2024	miR-876-3p	GATA1	Inhibits apoptosis and downregulates the IGFBP-3/p53 axis	Oral squamous cell carcinoma	Cisplatin	(200)
Wang et al, 2021	miR-103a-3p	Bak1	Inhibits apoptosis	NSCLC	Cisplatin	(203)
Zhang et al, 2021	miR-24-3p	CDX2/HEPH	Inhibits apoptosis	CRC	Methotrexate	(204)
Qin et al, 2019	miR-196a	CDKN1B and ING5	Affects the cell cycle and inhibits apoptosis	Head and neck cancer	Cisplatin	(205)
Zhou et al, 2024	AFAP1-AS1	p-ATM and p-CHK2	Reduces DNA damage	Esophageal squamous cell carcinoma	Radiation	(207)
Zhang et al, 2022	UPK1A-AS1	X-ray repair cross-complementing protein 5/6	Mediates DNA repair through non-homologous end-joining	PDAC	Cisplatin	(208)
Zhao et al, 2023	DLEU1	ZFP36	Confers ferroptosis resistance and drug resistance	Glioma	Erastin	(217)
Qu et al, 2023	DACT3-AS1	miR-181a-5p/SIRT1	Increases sensitivity of cells to ferroptosis	GC	Oxaliplatin	(219)
Zhao et al, 2024	miR-432-5p	ChaC1	Inhibits GSH depletion and ferroptosis	Prostate cancer	Docetaxel and erastin	(39)
Zhang et al, 2020	miR-522	ALOX15	Induces acquired drug resistance	GC	Cisplatin and paclitaxel	(224)
Qi et al, 2023	miR-3173-5p	ACSL4	Induces acquired drug resistance	Pancreatic cancer	Gemcitabine	(40)
Hu et al, 2019	miR-92a-3p	Wnt/β-catenin	Induces tumor cell EMT	CRC	5-FU	(228)
Hu et al, 2024	LINC00355	miR-34b-5p /CrkL	Induces tumor cell EMT	CRC	5-FU, oxaliplatin and cisplatin	(231)
Luo et al, 2021	LINC00355	miR-34b-5p/ABCB1	Enhances the efflux of chemotherapy drugs	BC	Cisplatin	(236)
Zhang et al, 2022	miR-625-3p	CELF2	Induces tumor cell EMT and chemoresistance	CRC	Oxaliplatin	(232)
Zhang et al, 2017	ANRIL	MRP1 and ABCC2	Enhances the efflux of chemotherapy drugs	Oral squamous cell	Cisplatin	(237)
Kunou et al, 2021	miR-4717-5p	ENT2	Reduces gemcitabine uptake	Lymphoma	Gemcitabine	(238)
Yao et al, 2024	RP11-161H23.5	CNOT4/HLA-A	Promotes immune evasion	PDAC	/ (241)
Jiang et al, 2021	OIP5-AS1	miR-142-5p/PD-L1	Promotes immune evasion	Lung cancer	/ (242)

[i] CAC, colitis-associated cancer; HuR, human antigen R; CRC, colorectal cancer; EMT, epithelial-mesenchymal transition; CCa, cervical cancer; PABPN1, poly(A) binding protein nuclear 1; ARID1A, AT-rich interactive domain 1A; SOCS, suppressor of cytokine signaling; ESCC, esophageal squamous cell carcinoma; APAF1, apoptotic peptidase activating factor 1; RB1, retinoblastoma susceptibility; BC, breast cancer; CDK4/6i, cyclin-dependent kinase 4/6 inhibitors; EIF4A3, eukaryotic translation initiation factor 4A3; PLAC8, placenta associated 8; GREM2, gremlin 2; TP53INP1, tumor protein P53 induced nuclear protein 1; NCALD, neuronal calcium sensor-1; ZEB1, zinc finger E-box binding homeobox 1; E2F3, E2F transcription factor 3; GATA1, GATA binding protein 1; IGFBP-3, insulin-like growth factor binding protein 3; Bak1, Bcl-2 homologous antagonist/killer 1; CDX2, caudal-type homeobox transcription factor 2; HEPH, hephaestin; CDKN1B, cyclin-dependent kinase inhibitor 1B; ING5, inhibitor of growth family member 5; AFAP1-AS1, actin filament associated protein 1 antisense RNA 1; p-ATM, phosphorylated ataxia telangiectasia mutated; p-CHK2, phosphorylated checkpoint kinase 2; UPK1A-AS1, uroplakin 1A antisense RNA 1; PDAC, pancreatic ductal adenocarcinoma; DLEU1, deleted in lymphocytic leukemia 1; ZFP36, zinc finger protein 36; SIRT1, sirtuin 1; ChaC1, γ-glutamylcyclotransferase 1; GSH, glutathione; ALOX15, arachidonic acid 15-lipoxygenase-1; ACSL4, long-chain acyl-CoA synthase 4; CrkL, Crk-like adapter protein; ABCB1, ATP-binding cassette subfamily B member 1; CELF2, CUGBP, Elav-like family member 2; ANRIL, antisense non-coding RNA in the INK4 locus; MRP1, multidrug resistance-associated protein 1; ABCC2, ATP-binding cassette subfamily C member 2; ENT2, equilibrative nucleoside transporter 2; CNOT4, CCR4-NOT transcription complex subunit 4; HLA-A, human leukocyte antigen A; PD-L1, programmed death-ligand 1; 5-FU, fluorouracil; CCAL, colorectal cancer-associated lncRNA; circ, circular RNA; GC, gastric cancer; HCC, hepatocellular carcinoma; LINC00355, long intergenic non-protein coding RNA 355; miR, microRNA; ncRNA, non-coding RNA; NSCLC, non-small cell lung cancer.

Drug resistance-related signaling pathways

In cancer cells, the dysregulation and mutation of key nodal molecules in signaling pathways lead to the aberrant activation of multiple pathways, which form intricate cross-regulatory networks collectively contributing to the development of tumor drug resistance (136). For instance, TGF-β not only induces EMT in cancer cells by activating the core effector Smad, but also promotes PI3K/AKT activation by increasing transforming growth factor β receptor type I kinase activity or suppressing PTEN (137), while the PI3K/AKT pathway further reinforces Wnt signaling by phosphorylating GSK-3β to inhibit β-catenin degradation (138,139). AKT phosphorylates salt-inducible kinase 1, relieving its suppression of STAT3 (140), thereby enhancing JAK/STAT-mediated transcriptional activation of anti-apoptotic and proliferative genes (141). There exists a feedback activation mechanism in several signaling pathways, such as the AKT, ERK and STAT3 signaling pathways. When a single signaling pathway is inhibited, the crosstalk of other signaling pathways activates feedback, further driving the adaptive resistance of cancer cells (142). Functionally, these signaling pathways serve a synergistic role in the formation of tumor resistance by regulating the EMT, the maintenance of CSC stemness and metabolic reprogramming (143,144). There is an interplay of signaling pathways between CAFs and cancer cells (Fig. 5) (144). As crucial players in the regulation of cellular signaling pathways (145), ncRNAs exert critical roles in this regulatory process, which is described in detail subsequently.

Figure 5 Exosomal ncRNAs derived from cancer-associated fibroblasts mediate the activation of drug resistance-related signaling pathways in tumor cells. Specifically, exosomal ncRNAs activate the Wnt/β-catenin, STAT3 and PI3K/AKT signaling pathways, among others, thereby promoting tumor progression and drug resistance. ↓, promotion; ⊥, inhibition. The figure was created by Adobe Illustrator CC 2019. ARID1A, AT-rich interactive domain 1A; DVL, dishevelled; E2F3, E2F transcription factor 3; EIF4A3, eukaryotic translation initiation factor 4A3; HIF, hypoxia-inducible factor; IGF2BP1, insulin like growth factor 2 mRNA binding protein 1; JAK, janus kinase; NCALD, neurocalcin δ; ncRNA, non-coding RNA; P, phosphorylated; PABPN1, poly(A) binding protein nuclear 1; PLAC8, placenta associated 8; RB1, retinoblastoma susceptibility; RTK, receptor tyrosine kinase; SOCS1, suppressor of cytokine signaling 1; TCF/LEF, T-cell factor/lymphoid enhancer factor 1; TP53INP1, tumor protein P53 induced nuclear protein 1; ZEB1, zinc finger E-box binding homeobox 1.

Wnt/β-catenin

The Wnt/β-catenin pathway is a classic signaling cascade that drives cancer growth, metastasis and drug resistance in various malignancies, such as GC, HCC and BC (146). In the absence of Wnt ligands, GSK3β, Axin and adenomatous polyposis coli (APC) form a complex, promoting the phosphorylation of β-catenin. When Wnt protein ligands are present, they recruit cytoplasmic Dishevelled proteins to disrupt the Axin/GSK3β/APC complex, thereby rescuing β-catenin from inhibition. Subsequently, accumulated β-catenin enters the nucleus and binds to the TFs T-cell factor (TCF) and lymphoid enhancer factor 1, initiating the expression of a series of oncogenes (147). Activation of the Wnt/β-catenin pathway promotes the expression of glutathione (GSH) peroxidase 4 (GPX4) by forming a β-catenin/TCF4 transcriptional complex, thereby reducing lipid ROS production and drug-induced ferroptosis in GC cells (148). In the TME, HGF secreted by myCAFs enhances the cloning ability and tumorigenicity of colon CSCs by activating the Wnt/β-catenin signaling pathway (149).

Studies have shown that CAF-derived exosomal ncRNAs are also important activators of the Wnt/β-catenin pathway. Ren et al (150) revealed upregulated expression of lncRNA H19 in colitis-associated cancer (CAC). The authors further confirmed that CAF (specifically expressing FAP, ferroptosis suppressor protein 1, ACTA2 and CD90)-derived exosomes were the source of H19 in CAC cells. Overexpression of H19 promoted the proliferation, migration and invasion of CAC cells both in vivo and in vitro. Additionally, H19 increased the expression of CSC markers (Nanog, Oct4 and Sox2) and oxaliplatin resistance of CAC cells. Mechanistically, H19 activated the β-catenin signaling pathway by sponging miR-141, an inhibitor of β-catenin (150). Similarly, lncRNA colorectal cancer-associated lncRNA (CCAL) is transferred between CAFs and CRC cells via exosomes. Upon entering cancer cells, CCAL activates β-catenin by directly interacting with the mRNA-stabilizing protein human antigen R, inducing drug resistance in CRC (151). CAF-derived exosomal miR-148b-3p is a crucial factor driving bladder cancer development. In a xenograft mouse model, injection of CAF-exosomes into mice enhanced their resistance to docetaxel combined with paclitaxel, while the use of a miR-148b-3p inhibitor restored chemotherapy sensitivity (152). PTEN, a direct target of miR-148b-3p, reduces nuclear β-catenin levels by inhibiting AKT phosphorylation (153). Mechanistically, miR-148b-3p promoted the proliferation and EMT of bladder cancer cells through PTEN upregulation and suppression of the β-catenin pathway, thus endowing bladder cancer with drug resistance (152). In cervical cancer (CCa), miR-1323 is secreted by CAFs and is upregulated in CCa cells. Through functional assays, miR-1323 has been demonstrated to promote cell proliferation, migration, invasion and radiation resistance. Further investigation revealed that the effect of miR-1323 was achieved by targeting poly(A) binding protein nuclear 1 (PABPN1). Additionally, PABPN1 can recruit insulin-like growth factor 2 mRNA-binding protein 1 to further regulate GSK-3β, thereby influencing the activity of the WNT/β-catenin signaling pathway (154).

STAT3

The STAT family is one of the most important TF families in tumors, comprising STAT1-6, which share similar structures and exhibit a high degree of conservation (155). Among the STAT family, STAT3 is the most extensively studied member, and serves a pivotal role in tumor cell proliferation, invasion, stem cell characteristics and drug resistance (156). Previous research has indicated that the activation of STAT3 is a crucial pathway for CAFs to exert their effects in cancer progression. CAFs secrete IL-6 to activate the JAK/STAT3 signaling pathway in gallbladder cancer cells, promoting the metastasis of gallbladder cancer. Additionally, the increased number of CAFs is associated with poor prognosis in patients with gallbladder cancer (157). IL-10 derived from CD146+ CAFs facilitates angiogenesis and tumor growth in endometrial cancer via the JAK1/STAT3 axis (158). Cisplatin induces CAFs to produce more IL-11, which activates STAT3 through IL-11 receptor α, thereby enabling NSCLC cells to acquire cisplatin resistance (159). The binding of LIF to the LIF receptor also constitutes an important pathway for activating STAT3 and inducing cancer cell stemness (160). A previous study has revealed that circFARP1 upregulates LIF expression by sponging miR-660-3p and interacts with caveolin 1 to increase LIF secretion. The LIF secreted by CAFs activates STAT3 in pancreatic cancer cells, thereby reducing their sensitivity to gemcitabine (161).

Similarly, exosomal ncRNAs derived from CAFs promote cancer drug resistance by upregulating STAT3. miR-146a-5p expression is elevated in CAFs and miR-146a-5p enters bladder cancer cells via exosomes, and can target and inhibit the expression of AT-rich interactive domain 1A (ARID1A) (41). By targeting ARID1A to relieve the inhibitory effect of suppressor of cytokine signaling 1 on STAT3, miR-146a-5p indirectly activates STAT3, thereby promoting stem cell characteristics and chemotherapy resistance in bladder cancer. Furthermore, high levels of miR-146a-5p are positively associated with bladder cancer staging and recurrence risk (41). Sun et al (162) demonstrated that CAF-derived miR-296-3p in ovarian cancer promotes the proliferation, invasion and migration of ovarian cancer cells, and induces paclitaxel resistance. By analyzing the expression of circulating miR-296-3p in the blood, the chemotherapy sensitivity of patients with ovarian cancer can be predicted. Mechanistically, miR-296-3p directly targets suppressor of cytokine signaling 6 (SOCS6), an inhibitor of STAT3, thereby activating the STAT3 signaling pathway in ovarian cancer cells. Overexpression of SOCS6 can inhibit the activation of STAT3, thereby reversing the drug resistance in cancer cells induced by the elevation of miR-296-3p (162). Zhao et al (163) found that monocytic myeloid-derived suppressor cells (M-MDSCs) contribute to cisplatin resistance in esophageal squamous cell carcinoma (ESCC). CAF-secreted IL-6 and miR-21 synergistically induce the high immunosuppressive function of M-MDSCs. miR-21 promotes tumor α-SMA and CD11b expression, facilitates M-MDSC accumulation and is delivered via exosomes to induce M-MDSC generation (163). Substantial evidence has conclusively demonstrated that PTEN serves as a direct target of miR-21 (118,164,165). Similarly, the study by Zhao et al (163) demonstrated that miR-21 overexpression in M-MDSCs indirectly activated STAT3 by suppressing PTEN, thereby inducing MDSC generation and IL-6 autocrine signaling. Therefore, inhibiting STAT3 could restore cisplatin resistance induced by IL-6 and exosomal miR-21.

PI3K/AKT

The PI3K/AKT signaling pathway serves a crucial role in promoting cancer by regulating cancer cell survival and metabolism. PI3K phosphorylates phosphatidylinositol-4,5-bisphosphate to generate phosphatidylinositol-3,4,5-trisphosphate (PIP3). Subsequently, PIP3 recruits AKT to the plasma membrane. Once AKT is activated, it phosphorylates various substrates (166). mTOR, a common downstream effector of AKT, drives cancer progression by phosphorylating S6 kinase 1 and eIF-4E binding protein 1, and activates HIF-1α and LIPIN1 signaling to mediate cellular glycolysis and lipid metabolism (167).

Certain ncRNAs indirectly activate PI3K/AKT by targeting its upstream regulators. PTEN, a phosphoinositide phosphatase, dephosphorylates phosphatidylinositols (PtdIns) (3-5) P3 at the D3 position of the inositol ring to generate PtdIns (4,5) P2, serving as a direct antagonist of PI3K (168). CAF-derived miR-20a is transferred to cancer cells and upregulates the PI3K/AKT signaling pathway by inhibiting PTEN, thereby enhancing cisplatin resistance in NSCLC. Overexpression of PTEN reverses the upregulation of miR-20a in the PI3K/AKT pathway, inhibiting cell proliferation and cisplatin resistance, while promoting apoptosis (169). Qu et al (170) found that CAF-exosome circN4BP2L2 directly binds to eukaryotic initiation factor 4A (EIF4A3). EIF4A3 is an RNA-binding protein that is involved in the binding and subcellular localization of circRNAs. Studies have reported that EIF4A3 activates the PI3K/AKT signaling pathway through a competing endogenous RNA mechanism after binding with circRNA (171,172). However, the specific mechanisms of circN4BP2L2 in the study by Qu et al (170) have not been thoroughly elucidated. By activating PI3K/AKT, circN4BP2L2 promotes the proliferation and CSC characteristics of CRC cells, inhibits apoptosis and induces resistance to oxaliplatin in CRC cells. Furthermore, an in vivo experiment demonstrated that the inhibition of circN4BP2L2 can suppress the growth and drug resistance of CRC (170). HCC cells uptake CAF-secreted miR-1228-3p, which targets placenta associated 8, an inhibitor of PI3K/AKT, thus promoting proliferation, migration and invasion, and reducing the response of HCC cells to sorafenib (173). In addition, a study by Deng et al (174) demonstrated that HIF-1α was upregulated in CAFs under hypoxic conditions, enriching miR-21 in CAF exosomes by promoting its transcription. Subsequently, miR-21 acted on pancreatic cancer cells and maintained their stemness and gemcitabine resistance through the RAS/AKT/ERK axis.

Other signaling pathways

The TGF-β signaling pathway represents a crucial signaling pathway in the interaction between CAFs and cancer cells (80,175). SMAD is a core effector protein of TGF-β. Phosphorylated SMAD2, SMAD3 and SMAD4 form SMAD multimers, which translocate to the nucleus to regulate the transcription of genes associated with cancer growth and metastasis (175). However, research on the mechanisms by which CAF-derived exosomal ncRNAs regulate TGF-β in cancer cells remains relatively scarce. A study demonstrated that exosomal miR-423-5p could inhibit gremlin2, a SMAD-regulatory protein (176), promoting paclitaxel resistance in prostate cancer. Inhibition of TGF-β partially restored the sensitivity of cancer cells to paclitaxel. Additionally, treatment with exosomal miR-423-5p increased the expression of TGF-β. One study revealed miR-423-5p as a potential target in the TGF-β signaling pathway activated by CAFs (177). Snail is also a TF driven by TGF-β signaling, promoting the transcription of tumor EMT-related genes (178,179). Richards et al (180) demonstrated that exposure to gemcitabine increased the expression of the TF Snail in pancreatic CAFs, leading to increased release of exosomal miR-146a. This resulted in the proliferation and chemoresistance of pancreatic cancer epithelial cells. However, due to the increased expression of Snail in exosomes, the specific role of miR-146a was not determined in the study (180). CDK4/6 inhibitors represent a promising novel therapy for patients with estrogen receptor-positive BC (181). Through single-cell RNA sequencing analysis, miR-20 has been found to be enriched in CD63+ CAFs and to be transported to BC cells via exosomes. miR-20 could directly bind to the 3′ UTR of retinoblastoma susceptibility mRNA, inhibiting the therapeutic effect of CDK4/6 inhibitors on BC. The use of designed cyclic RGD-miR-20 sponge nanoparticles to inhibit miR-20 expression restored the sensitivity of BC to CDK4/6 inhibitors (181). In another study, exosomal miRNA-106b reduced the efficacy of gemcitabine in pancreatic cancer through the TP53INP1/P53 axis. Treatment with an miRNA-106b inhibitor reduced the levels of miRNA-106b in exosomes secreted by CAFs and decreased the resistance of pancreatic cancer to gemcitabine (182). Additionally, in CRC, CAF-derived exosomal miR-181d-5p and circ0067557 promote cell proliferation, migration, invasion and CRC resistance to fluorouracil (5-FU) (183,184). Mechanistically, miR-181d-5p is elevated in CAF-derived exosomes through RNA N6-methyladenosine methylation mediated by methyltransferase-like 3, and subsequently targets neurocalcin δ (NCALD) expression, thereby attenuating NCALD-mediated inactivation of the NF-κB signaling pathway (183,185). circ0067557 physically interacts with the CSC regulator Lin28, and knockdown of circ0067557 reduces Lin28 levels in CRC cells, and suppresses both malignant progression and drug resistance in CRC (184,186). Compared with normoxic CAFs, hypoxic CAFs exhibit deletion of miR-200b-3p, leading to upregulation of its targets zinc finger E-box binding homeobox 1 and E2F transcription factor 3 in CRC cells, which facilitates cell invasion, stemness maintenance and 5-FU resistance (187).

Resistance to cell death

Apoptosis, a form of programmed cell death, serves to maintain tissue homeostasis and eliminate harmful cells (188). Apoptosis is primarily initiated through two distinct pathways: The extrinsic and intrinsic pathways. The extrinsic pathway is mediated by cell surface death receptors, whereas the intrinsic pathway is activated by intracellular stress signals (such as DNA damage and oxidative stress) independent of cell surface receptors, with mitochondria serving as the central regulatory hub (189). Both pathways activate caspases, the key regulators of apoptosis, which are categorized into initiator caspases (such as caspase-8 and -9) and effector caspases (such as caspase-3, -6 and -7) (190). Upon activation, these caspases selectively cleave hundreds of intracellular substrate proteins, ultimately leading to cellular fragmentation into apoptotic bodies and programmed cell death (190,191). Abnormal regulation of the apoptotic system allows cancer cells to escape the apoptotic program, leading to unlimited proliferation and drug resistance (192). Research has highlighted the role of CAF-derived exosomal ncRNAs in cancer cell apoptosis. miR-21 is elevated in CAF exosomes and directly targets ovarian cancer cells, reducing their chemotherapy sensitivity and apoptosis. It has also been demonstrated that miR-21 directly binds to apoptotic peptidase activating factor 1 (APAF1) (193), an apoptotic protein that promotes the formation of apoptotic bodies through interactions with cytochrome c and deoxyadenosine triphosphate. The subsequent activation of caspase-9 and -3 by apoptotic bodies marks the initiation of the apoptotic program (194). By inhibiting APAF1, miR-21 reduces the expression of caspase-9 and -3 in cells, rendering ovarian cancer cells resistant to paclitaxel (193). Furthermore, miR-21 inhibits apoptosis by targeting programmed cell death 4 (PDCD4). In various gastrointestinal tumors, miR-21 has been demonstrated to be upregulated and to suppress PDCD4, thereby promoting tumor cell proliferation, invasion, migration and 5-FU resistance. Overexpression of PDCD4 can reverse the effects of miR-21, consequently reducing cell survival rates and drug resistance (195-197). The Bcl-2 family represents a critical group of proteins governing the intrinsic apoptotic pathway, consisting of anti-apoptotic members (such as Bcl-2, Bcl-xL and Bcl-w) and pro-apoptotic members [Bax/Bcl-2 homologous antagonist/killer (Bak)]. Specifically, Bax and Bak induce the formation of pores in the mitochondrial outer membrane, facilitating the release of cytochrome c and subsequent caspase activation, whereas Bcl-2 exerts its anti-apoptotic function by inhibiting Bax/Bak activation (198,199). Several exosomal ncRNAs are involved in the regulation of the Bcl-2 family. According to the expression levels of collagen type VI α6 chain, collagen type XXV α1 chain and collagen type XXVI α1 chain in OSCC-CAFs, Kang et al (64) classified CAFs into two distinct phenotypes: CAF-Ps (promote cancer development) and CAF-Ds (delay cancer development). CAF-Ps promote cancer progression while exhibiting low expression of the aforementioned genes (64). Subsequently, it was found that the exosomes secreted by CAF-Ps contained elevated levels of miR-876-3p. CAF-P-derived miR-876-3p inhibits the insulin-like growth factor-binding protein-3 (IGFBP-3)/p53 axis by targeting GATA-binding protein 1 in OSCC (200). Inhibition of IGFBP-3 reduces chemotherapy-induced apoptosis by upregulating the expression of Bcl-2, and downregulating Bax and p53. This process reduces the sensitivity of OSCC cells to cisplatin by inhibiting apoptosis (201,202). In NSCLC, CAF-secreted miR-103a-3p directly targets the pro-apoptotic protein Bak1, thereby inhibiting apoptosis in NSCLC cells and inducing cisplatin resistance (203). Additionally, miR-24-3p is transported via exosomes derived from CAFs and is highly expressed in CRC cells, which is associated with malignancy and drug resistance in CRC. Treatment of CAFs with a miR-24-3p inhibitor resulted in decreased expression of miR-24-3p in CRC cells. This led to reduced cell proliferation and enhanced apoptosis. Mechanistically, miR-24-3p inhibits apoptosis by downregulating caudal-type homeobox transcription factor 2/hephaestin, increasing the resistance of CRC cells to methotrexate (204). The packaging of ncRNAs into exosomes may be associated with RBPs such as Pumilio2 (199) and heterogeneous nuclear ribonucleoprotein A1 (hnRNPA1) (205). In head and neck cancer CAFs, hnRNPA1 binds to miR-196a, enriching it in CAFs and accelerating its packaging into exosomes through translocation between cells and exosomes. Subsequently, miR-196a targets cyclin-dependent kinase inhibitor 1B and inhibitor of growth family member 5 in head and neck cancer cells, affecting the cell cycle and apoptosis, respectively, and conferring cisplatin resistance to head and neck cancer (205).

DNA damage, particularly DNA double-strand breaks (DSBs), is a crucial factor in activating apoptotic signaling in cancer cells mediated by antitumor drugs (5). Numerous antitumor drugs, such as cisplatin, 5-FU and epirubicin, can induce DNA damage in cancer cells, thereby triggering apoptosis (206). Previous studies have shown that exosomal ncRNAs not only regulate apoptotic signaling pathways but also facilitate the repair of DNA double strands, thereby reducing the occurrence of apoptosis. In radiation-resistant ESCC, FAP+ CAFs transport lncRNA actin filament associated protein 1 antisense RNA 1 (AFAP1-AS1) to cancer cells via exosomes. AFAP1-AS1 promotes the expression of the DNA repair proteins phosphorylated (p-)ataxia telangiectasia mutated and p-checkpoint kinase 2, reducing the DNA damage marker protein γ-H2AX and inducing resistance to radiotherapy in cancer cells (207). Although not directly transmitted via exosomes, the overexpression of lncRNA uroplakin 1A antisense RNA 1 (UPK1A-AS1) in PDAC cells is induced by CAF-secreted IL-8. IL-8 promotes the expression of p65, which directly activates the transcription of UPK1A-AS1. Subsequently, UPK1A-AS1 serves as a molecular scaffold for the DSB repair proteins X-ray repair cross-complementing protein 5 and X-ray repair cross-complementing protein 6, mediating DNA repair through non-homologous end-joining, thus conferring cisplatin resistance to PDAC (208).

Ferroptosis, a regulated form of cell death distinct from apoptosis, was first proposed in 2012 (209,210). Ferroptosis is characterized by an increase in the small pool of Fe2+ and lipid peroxidation (8). The accumulation of Fe2+ within cells generates free radicals through the Fenton reaction and contributes to the production of phospholipid hydroperoxides (211). Simultaneously, Fe2+ catalyzes the generation of reactive ROS, thereby promoting lipid peroxidation. This overwhelming lipid peroxidation ultimately leads to cell death (212). The canonical regulatory pathway of ferroptosis relies on the GSH peroxidase GPX4, whose normal biological function depends on the support of GSH (213). Inhibition of GPX4 and GSH leads to the accumulation of lipid ROS, resulting in rupture of the cell membrane (214). Sorafenib can inhibit solute carrier family 3 member 2 and solute carrier family 7 member 11 (SLC7A11), promoting lipid peroxidation in HCC. This hinders the production of GSH and inhibits GPX4 expression, thereby upregulating ferroptosis in HCC cells (215). However, the regulatory network within the TME may enhance cancer cell resistance to ferroptosis by strengthening the antioxidant system, thus inducing drug resistance (215).

The activation of the cystine-glutamate antiporter SLC7A11 increases GSH production by transporting cysteine (216). In glioma, CAF-derived exosomes promote deleted in lymphocytic leukemia 1 (DLEU1) transcription through heat shock transcription factor 1 activation, thereby upregulating DLEU1 expression (217). Subsequently, DLEU1 interacts directly with ZFP36 and inhibits activating transcription factor 3 (ATF3), which binds to the SLC7A11 promoter and inhibits SLC7A11 in a p53-independent manner (218). This process confers ferroptosis resistance to glioma cells (217). Conversely, lncRNA DACT3-AS1 is transferred from CAF exosomes to GC cells. DACT3-AS1 inhibits cell proliferation, migration and invasion both in vivo and in vitro by sponging miR-181a-5p, which targets sirtuin 1 (SIRT1) (219). In this context, SIRT1 activates ATF3 by depleting NAD+, thereby inhibiting SLC7A11 and GPX. By suppressing miR-181a-5p, DACT3-AS1 promotes SIRT1-mediated ferroptosis, consequently increasing cellular sensitivity to 5-FU (220). γ-glutamylcyclotransferase 1 (ChaC1) is a key enzyme in the GSH degradation process (221). CAF-derived exosomes can hinder ferroptosis induced by erastin (a GSH inhibitor) via miR-432-5p, which directly targets to ChaC1. By inhibiting ChaC1 expression, miR-432-5p attenuates GSH depletion and ferroptosis in prostate cancer cells, thereby inducing docetaxel resistance (39). Additionally, ACSL4 and arachidonic acid 15-lipoxygenase-1 (ALOX15) are key enzymes in phospholipid metabolism and induction of lipid peroxidation, mediating the specific oxidation of polyunsaturated fatty acids-phosphates (222,223). In digestive system cancers, CAFs secrete miR-522 and miR-3173-5p, which target ALOX15 and ACSL4, respectively, leading to the accumulation of lipid ROS in cancer cells and inducing acquired drug resistance (40,224).

EMT in cancer cells

EMT is a process driven by EMT-related TFs that leads to the transformation of cells from an epithelial-like morphology to a mesenchymal morphology. Following EMT, cancer cells exhibit an elongated morphology and upregulated expression of mesenchymal markers such as N-cadherin and vimentin, which respectively reduce intercellular adhesion and enhance cell motility, and thus, the cells become more invasive (225). EMT has a strong association with cancer drug resistance. Multiple signaling pathways, such as Wnt/β-catenin and TGF-β, that promote EMT also activate multidrug resistance (MDR) genes (226). The adaptive changes in autophagy and metabolism during EMT in cancer cells also promote the occurrence of MDR (226,227). CAFs can promote EMT in cancer cells, with exosomal ncRNAs serving a role in this process (19). In CRC, CAFs directly transfer miR-92a-3p to CRC cells via exosomes. Upregulation of miR-92a-3p activates the Wnt/β-catenin signaling pathway (228). Furthermore, miR-92a-3p has been demonstrated to directly target and suppress F-box and WD repeat domain containing 7, an E3-ubiquitin ligase responsible for degrading the EMT TF zinc finger E-box binding homeobox 2 (229). Experimental results revealed that miR-92a-3p promoted cell proliferation, invasion and migration, while inducing CSC generation, EMT and resistance to 5-FU in CRC (228). Crk-like adapter protein (CrkL) has been identified as a key regulator of EMT, mediating C-C motif chemokine ligand 20/C-C motif chemokine receptor 6-induced EMT through activation of the AKT pathway (230). Long intergenic non-protein coding RNA 355 (LINC00355), present in CAF-derived exosomes, acts as a sponge for miR-34b-5p, which can target and inhibit CrkL. LINC00355 indirectly upregulates the expression of CrkL, promoting EMT and MDR in CRC cells (231). Additionally, exosomal miR-625-3p inhibits the tumor suppressor gene WW domain containing oxidoreductase (WWOX) by targeting CUGBP, Elav-like family member 2 (CELF2), which regulates SMAD3, a component of the TGF-β signaling pathway. Through the CELF2/WWOX axis, miR-625-3p promotes cell EMT and chemoresistance in CRC (232,233).

Drug transporters

The ability of cancer cells to acquire drug resistance manifests as reduced drug uptake and enhanced drug efflux, which is achieved through the abnormal upregulation of drug transporter proteins such as ABC transporters (234). Numerous ncRNAs, such as lncRNA cancer susceptibility 9, miR-27a and miR-451, serve a critical role in regulating drug transporters, and targeting these proteins may be an effective way to alleviate drug resistance in cancer cells (235). Exosomal LINC00355 derived from CAFs in BC promotes the expression of ATP-binding cassette subfamily B member 1 (ABCB1), a drug efflux transporter and a target of miR-34b-5p, by sponging miR-34b-5p. This increased expression enhances the efflux of chemotherapy drugs from cancer cells. Overexpression of miR-34b-5p or knockdown of ABCB1 can eliminate the effect of LINC00355 in CAF exosomes, thereby restoring the sensitivity of BC cells to cisplatin (236). Similarly, midkine in exosomes increases the expression levels of lncRNA antisense non-coding RNA in the INK4 locus (ANRIL) in OSCC. Elevated ANRIL expression is associated with advanced TNM staging and lymph node metastasis in OSCC. Quantitative PCR analysis has demonstrated that ANRIL upregulated the drug transporters multidrug resistance-associated protein 1 (MRP1) and ATP-binding cassette subfamily C member 2 (ABCC2), while ANRIL knockdown reduced their expression, indicating positive regulation of MRP1/ABCC2 by ANRIL. Additionally, ANRIL suppresses caspase-3 activation to inhibit apoptotic pathways, thereby promoting tumor cell proliferation and reducing cisplatin-induced apoptosis (237). Additionally, anti-pyrimidine drugs such as gemcitabine enter cancer cells through balanced uptake by nucleoside transporter 2 (ENT2). miR-4717-5p, one of the most abundant miRNAs in CAF-derived exosomes in lymphoma, enters lymphoma cells and inhibits ENT2 expression to reduce gemcitabine uptake, thereby decreasing the sensitivity of lymphoma to gemcitabine (238).

Immune evasion

Cytokines secreted by CAFs serve a crucial role in immune cell regulation within the TME (239). However, to the best of our knowledge, the function of ncRNAs derived from CAF-sourced exosomes, which serve as important mediators of intercellular communication in the TME, remains unexplored. HLA is a type of protein found on cell surfaces, and is essential for the immune recognition mechanism of CD8+ T cells (240). In PDAC, lncRNA RP11-161H23.5 derived from CAFs promotes immune evasion via downregulation of HLA-A expression through binding to CCR4-NOT transcription complex subunit 4, which enhances the degradation of HLA-A mRNA by shortening its poly(A) tail. Engineered extracellular vesicles loaded with RP11-161H23.5 small interfering RNA (siRNA) have been constructed to enhance immunotherapy effectiveness, emerging as a potential strategy for the treatment of pancreatic cancer (241). In lung cancer, CAFs interact with programmed cell death protein 1 (PD-1) on the surface of peripheral blood mononuclear cells (PBMCs) by upregulating the expression of programmed death-ligand 1 (PD-L1) in lung cancer cells (242). This process inhibits signaling pathways in PBMCs, such as the PI3K/AKT and Ras/MEK/ERK pathways, thus blocking the proliferation, differentiation and metabolic reprogramming of PBMCs (243). A study (242) has shown that the lncRNA OIP5-AS1 serves an important role by functioning as a sponge to adsorb miR-142-5p, which directly targets and inhibits PD-L1 expression. OIP5-AS1 enters lung cancer cells through exosomes secreted by CAFs, upregulating the expression levels of PD-L1 in lung cancer cells by inhibiting miR-142-5p. This process subsequently suppresses PBMC-induced cell death via the PD-L1/PD-1 axis (242,243).

Association between exosomal ncRNAs and soluble factors

The regulation of tumor drug resistance by CAFs is not mediated by a single pathway but rather involves the synergistic effects of multiple mechanisms. Evidence indicates that cytokines and chemoattractants secreted by CAFs also contribute to tumor progression and drug resistance (42). Activated CAFs secrete IL-6, which stimulates the activation of the JAK/STAT3 pathway in cancer cells and enhances TGF-β-induced p-Smad3 and Snail expression, thereby promoting EMT (244). Additionally, IL-6-activated JAK/STAT3 suppresses p53, attenuating the response of prostate cancer cells to doxorubicin (245). HGF derived from α-SMA+ CAFs activates c-Met and further promotes the PI3K/AKT and FRA1/hes related family bHLH transcription factor with YRPW motif 1 signaling cascades, inducing drug resistance in HCC, ovarian cancer and BC (246-248). Furthermore, SDF-1 secreted by CAFs upregulates SATB homeobox 1 expression in pancreatic cancer cells, enhancing their resistance to gemcitabine (249). The chemoattractant CXCL12, also secreted by CAFs, binds to the CXCR4 receptor on cancer cells, activating the Wnt/β-catenin pathway, and inducing EMT and cisplatin resistance in ovarian cancer (250). These findings suggest that various soluble factors secreted by CAFs and exosomal ncRNAs share common mechanisms of action, such as activating the TGF-β and Wnt/β-catenin pathways, thereby exerting synergistic effects. Compared with soluble factors, exosomal ncRNAs may serve as key regulatory nodes in signaling pathways, modulating multiple cascades and avoiding feedback activation upon inhibition of a single pathway (21,22). Furthermore, ncRNAs can induce cancer drug resistance through diverse mechanisms. For example, miR-21 mediates the activation of multiple drug resistance-related signaling pathways and promotes apoptosis resistance (163,174,193). Given the tissue- and cell-specific expression of ncRNAs, targeting of ncRNAs may represent a more comprehensive and personalized therapeutic strategy compared with targeting of soluble factors.

Unresolved issues and research visions

The present review summarizes the role of exosomal ncRNAs in CAF-mediated tumor drug resistance, noting the diversity of their mechanisms of action. However, current research does not encompass all the mechanisms associated with CAF-derived exosomal ncRNAs, and a number of these mechanisms remain to be validated by further studies. CAFs serve a crucial role in the metabolic regulation of cancer cells (97). Cytokines secreted by CAFs stimulate glycolysis in cancer cells by activating the HIF-1 and PI3K/Akt signaling pathways (251,252). Glycolysis provides tumors with sufficient energy in nutrient-deprived environments, supporting their survival and supplying ATP for drug efflux pumps (253). Concurrently, CAFs secrete substantial amounts of pyruvate and lactate, synergizing with cancer cell glycolysis to suppress the immune microenvironment (97). Only a limited number of studies have elucidated the association between CAFs and cancer lipid metabolism. Ippolito et al (254) demonstrated that lactate secreted by CAFs upregulated the expression levels of lipid metabolism-related genes in cancer cells. Aberrant lipid metabolism supplies energy and metabolic substrates to cancer cells. Additionally, its catabolic products act as signaling molecules to modulate drug resistance-related pathways (255). Furthermore, CAFs exhibit high expression of glutamine synthetase, leading to increased secretion of glutamine. The glutamine molecules are catabolized into glutamate in cancer cells to support their proliferation (256). CAF-derived exosomal ncRNA is also a crucial factor regulating metabolism. Targeting ncRNA can overcome cancer growth, metastasis and drug resistance mediated by the metabolic reprogramming of cancer cells (257). In oral cancer, lncRNA H19 secreted by CAFs upregulates the expression of 6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3, a key enzyme regulating the cellular glycolytic pathway, by acting as a sponge for miR-675-5p. This interaction allows H19 to induce increased glycolysis in oral cancer, thereby promoting tumor growth (258). In BC, exosomes derived from CAFs have been found to indirectly upregulate negative elongation factor complex member E (NELFE) by acting as a sponge for miR-4510. Long intergenic non-protein coding RNA 1711 activates thioredoxin through the miR-4510/NELFE axis, thereby promoting glycolysis, proliferation and invasion of BC cells (259). In addition to glycolysis, exosomes derived from CRC cells that contain HSPC111 could promote the formation of a pre-metastatic niche and facilitate CRC liver metastasis in mouse models. This occurs through the upregulation of acetyl-CoA levels in CAFs, which is mediated by lipid metabolism promoted by ATP citrate lyase phosphorylation (260). The accumulation of acetyl-CoA enhances H3K27 acetylation in CAFs, thus facilitating the expression and secretion of C-X-C motif chemokine ligand 5 (CXCL5). Furthermore, the CXCL5-C-X-C motif chemokine receptor 2 axis increases the excretion of CRC cell exosomes containing HSPC111, thereby establishing a positive feedback loop that promotes CRC cell liver metastasis (260). Although the role of ncRNAs derived from CAFs in drug resistance through metabolic reprogramming has been hypothesized, to the best of our knowledge, the association between the two has not been specifically confirmed by experimental data.

Various stresses such as antitumor drugs promote tumor autophagy (261). Autophagy facilitates tumor growth by recycling metabolic substances and regulating mitochondrial function (261). Autophagy involves >20 proteins encoded by autophagy-related genes. These proteins encapsulate cytoplasmic materials into the double-membrane vesicles of autophagosomes. After engulfing the materials, autophagosomes fuse with lysosomes, utilizing pH-sensitive hydrolases to degrade the engulfed substances (262). Although Zhu et al (263) demonstrated that CAF-derived lncRNA FAL1 promoted oxaliplatin resistance by inhibiting autophagy, it was only under conditions of excessive autophagy activation that it could exert its antitumor effect by inducing cell death (264). CAF-mediated autophagy in cancer cells contributes to drug resistance (265,266). The role and mechanism of exosomal ncRNAs in this process remain to be further investigated.

Exosomal ncRNAs secreted by CAFs may also influence other cells within the TME, including TAMs and endothelial cells (227). The role of crosstalk mediated by exosomal ncRNAs between CAFs and these cells in the context of drug resistance has not been extensively studied. Given that much of the research on the TME has concentrated on individual cell types, further exploration in this area could enhance the understanding of the complex regulatory networks within the TME and their impact on tumors. Additionally, the subtypes of CAFs when examining the functions of CAF-derived ncRNA have not been thoroughly investigated. This limitation impedes a more comprehensive understanding of the roles of CAFs expressing different surface receptors and their secreted ncRNAs in cancer, leading to issues of irreproducibility in research.

Clinical application and challenges

Given the various roles of CAFs in tumors, targeting CAFs (promoting their transformation into NFs or anticancer phenotypic CAFs) may benefit patients with tumors (267). In practice, achieving clinical benefits may not necessarily require the elimination or reprogramming of CAFs, but can also be achieved by blocking signals from CAFs. For example, targeting CXCL12 signaling can be considered as targeting CAFs, as they are the primary source of chemokines in numerous tumors (94). Indeed, targeting the communication process between CAFs and cancer cells is also an effective method to alter the function of CAFs (267), making ncRNAs important candidate targets for the treatment of CAF-mediated cancer drug resistance. Therapies targeting ncRNAs are based on RNA modification technologies that regulate their expression: Upregulating cancer-suppressing ncRNAs or downregulating cancer-promoting ncRNAs (268). Multiple methods have been used to target ncRNA expression to regulate tumor progression (268). For instance, the CRISPR-CRISPR associated protein 9 (Cas9) system (269), siRNA (270) and antisense DNA oligonucleotides containing locked nucleic acid (271) have been used to inhibit cancer-promoting ncRNAs, while mimics can be employed to overexpress cancer-suppressing ncRNAs (272). Furthermore, small activating RNA (saRNA), which induces gene expression by targeting promoter regions, has also been employed to upregulate target ncRNAs (273,274).

Given the inherent instability and high molecular weight of these RNA-based therapeutics in vivo, the delivery systems for these 'RNA cargos' are increasingly being optimized to ensure their precise and efficient transport to specific cells (275). Delivery systems are generally divided into viral delivery systems and non-viral delivery systems. Based on the type of carrier, the non-viral delivery systems can be categorized into nanoparticles (including lipid, polymer and inorganic nanoparticles), liposomes, polymer micelles and exosomes (275). Nanoparticles can enhance their efficacy through crosslinking agents or stimulus-responsive systems. For instance, some studies have utilized GSH and pH-sensitive nanoparticles. The pH-sensitive properties allow nanoparticles to accumulate in the acidic environment of tumor tissues, where they can further reduce the abundant GSH in cells to thiols, leading to micelle degradation and promoting the release of their contents (276,277). Simultaneously, these nanoparticles can also co-deliver chemotherapeutic drugs with ncRNA-targeted drugs to achieve a synergistic antitumor effect (278). Additionally, using viral vectors such as adenoviruses, adeno-associated viruses and lentiviruses to deliver nucleic acids that regulate ncRNAs is an effective strategy (268). Gene therapies targeting ncRNAs have entered clinical trials, and have demonstrated good tolerability and efficacy in some studies (279,280).

The timing of cancer diagnosis and the initiation of treatment are critical determinants of patient prognosis (281). Exosomes secreted by CAFs enter the circulatory system and are abundantly present in bodily fluids such as blood, urine and saliva, and remain stable in suitable environments (282,283). This allows the ncRNAs specifically expressed in exosomes to serve as important biomarkers to monitor the progression, prognosis and drug response of patients with cancer (284). Xu et al (285) developed a CAF-related miRNA model and found that miR-106b-5p was associated with the growth and recurrence of prostate cancer. CAF-derived exosomal miR-146a-5p has been detected in the bodily fluids of patients with bladder cancer, and its elevated levels are positively associated with the recurrence rate of bladder cancer (41). The findings of Zhou et al (207) indicated that AFAP1-AS1 was secreted by CAFs, and high levels of AFAP1-AS1 in plasma were associated with low sensitivity to definitive chemoradiotherapy in patients with ESCC.

Targeting the exosomal ncRNA derived from CAFs provides a promising strategy to overcome drug resistance in cancer. However, several urgent issues hinder its clinical application. In ncRNA-targeted therapies, off-target effects and systemic toxicity present formidable challenges (286). Despite the considerable potential of ncRNAs in cancer treatment, their off-target effects continue to be a barrier that must be addressed (286). Off-target effects can arise with any delivery system employed to target ncRNAs (286). Furthermore, systemic toxicity leading to patient intolerance remains a primary reason for the discontinuation of clinical trials (287). A phase I clinical trial was conducted to evaluate the efficacy and safety of a miR-34a liposomal mimetic in patients with advanced solid tumors. Among the 85 enrolled patients, a partial response was observed in three cases, while stable disease was achieved in 16 cases. However, all participants experienced adverse reactions, including fever, nausea and pain. The trial was ultimately terminated due to four fatalities attributed to severe adverse events (288). On the other hand, although delivery systems have advanced, different types of delivery systems still exhibit certain disadvantages. For instance, while liposomes and polymer nanoparticles can safely and effectively transfer large genes, their transfection efficiency and transgene expression are relatively low (289). Conversely, although viral vectors address the issue of transfection efficiency and allow for sustained gene expression, they are limited in their development due to high immunogenicity and low payload capacity (290). Despite these challenges, it is hypothesized that with the emergence of novel, innovative research and interdisciplinary collaboration, therapies targeting ncRNAs will achieve breakthroughs.

Conclusion

The TME serves a crucial role in cancer growth, metastasis and drug resistance. CAFs are a critical component of the TME, constructing a drug-resistant TME through a complex intercellular communication network. Accumulating evidence suggests that ncRNAs exert extensive biological functions in regulating cancer progression mediated by CAFs. The present review focuses on the bidirectional regulatory mechanisms of exosomal ncRNAs in CAF-mediated drug resistance. Exosomal ncRNAs derived from tumor cells activate multiple signaling pathways, inducing the differentiation of NFs into CAFs. This establishes a suitable niche for CAFs to regulate cancer drug resistance. Conversely, CAF-derived exosomal ncRNAs can be internalized by cancer cells, activating drug resistance-related signaling pathways such as Wnt/β-catenin, STAT3, TGF-β and PI3K/AKT. Furthermore, exosomal ncRNAs regulate cell death, EMT and drug transporter expression levels, and promote immune evasion, all of which contribute to cancer drug resistance. Additionally, CAF-derived exosomal ncRNAs may mediate drug resistance by regulating autophagy, metabolic reprogramming and the immune microenvironment, which are promising areas for future research

It is noteworthy that the specifically expressed ncRNAs in CAF-derived exosomes hold potential for the development of novel clinical diagnostic and therapeutic approaches. Exosomal ncRNAs in bodily fluids are readily accessible and quantifiable, serving as emerging biomarkers that reflect tumor TNM staging and therapeutic drug sensitivity in patients with cancer. Furthermore, RNA modification technologies, including siRNA, saRNA and CRISPR-Cas9, can disrupt intercellular communication and signaling cascades by regulating exosomal ncRNAs, thereby effectively reversing tumor drug resistance. Various delivery systems have been designed by researchers to achieve precise and efficient delivery of ncRNA-targeting therapeutics. Currently, ncRNA-targeting strategies are undergoing continuous advancement and entering clinical trials. With further research progress, intervention strategies targeting CAF-derived exosomal ncRNAs are expected to overcome the limitations of conventional therapies and provide innovative treatment approaches to combat tumor drug resistance.

Availability of data and materials

Not applicable.

Authors' contributions

JL wrote the majority of the manuscript and created the figures and tables. YH contributed to the manuscript writing. LF and MS assisted in figure preparation and manuscript revision. GH, ZW and FD revised the manuscript. YZ and YX revised the manuscript and prepared tables. YL revised the manuscript and supervised the process. Data authentication is not applicable. All authors have read and approved the final version of the manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Acknowledgements

Not applicable.

Funding

No funding was received.

References