Research progress on RNA-binding proteins in breast cancer.

Breast cancer is the most common malignancy among women, and the abnormal regulation of gene expression serves an important role in its occurrence and development. However, the molecular mechanisms underlying gene expression are highly complex and heterogeneous, and RNA-binding proteins (RBPs) are among the key regulatory factors. RBPs bind targets in an environment-dependent or environment-independent manner to influence mRNA stability and the translation of genes involved in the formation, progression, metastasis and treatment of breast cancer. Due to the growing interest in these regulators, the present review summarizes the most influential studies concerning RBPs associated with breast cancer to elucidate the role of RBPs in breast cancer and to assess how they interact with other key pathways to provide new molecular targets for the diagnosis and treatment of breast cancer. Copyright: © Zhang et al.

Introduction

Breast cancer is the most commonly reported type of cancer and the most common cause of cancer-associated mortality in women worldwide (1). Due to the high complexity and heterogeneity of breast tumors, the molecular mechanisms associated with their occurrence and development are still not fully understood (2). Breast cancer is categorized into the luminal A, luminal B, human epidermal growth factor receptor 2 (HER2)-positive and triple-negative breast cancer (TNBC) subtypes according to the expression status of the estrogen receptor (ER), progesterone receptor (PR) and HER2. Notably, ~70% of breast tumors are ER-positive and rely on estrogen for growth (3). Endocrine therapy and selective ER modulators or aromatase inhibitors are the gold standards of treatment for ER+ breast cancer; however, long-term endocrine therapy can cause the development of acquired resistance (4). HER2-positive breast cancer is mainly treated with targeted drugs, but there are inevitable side effects (5). TNBC is a subset of breast cancer characterized by the loss of ER, PR and HER2 expression. The TNBC subtype is defined by its high metastasis rate, fast tumor growth, early recurrence, low survival rate and lack of effective treatment (6,7). Therefore, there is an urgent need to identify more molecules to serve as diagnostic and therapeutic targets for breast cancer.

Abnormal regulation of gene expression is the main cause of breast cancer development. With in-depth studies of the genetic and epigenetic mechanisms underlying breast cancer, increasing evidence has shown that gene expression in these cells is regulated in a number of ways (8–10). RNA-binding proteins (RBPs) are involved in various aspects of RNA metabolism, including RNA splicing, polyadenylation, sequence editing, RNA transport, maintenance of RNA stability and degradation, intracellular localization and translation control (11,12). Any significant change or interference in the expression or activity of RBPs that regulate these basic cell functions can lead to different diseases, including cancer (13). RBPs are essential binding partners of intracellular RNA and engage in highly dynamic interactions with other proteins, coding RNAs and noncoding RNAs to form functional units called ribonucleoprotein complexes (RNPs); these complexes serve an important role in the regulation of post-transcriptional gene expression in a variety of ways. RBPs can rapidly and effectively alter gene expression, particularly in response to microenvironmental changes (14). It is clear that RBPs are abnormally regulated in breast cancer, thereby affecting the expression and function of oncoproteins and tumor suppressor proteins. Determining the intricate network of interactions between RBPs and breast cancer-related RNA targets will therefore contribute to a better understanding of tumor biology and potentially reveal new targets for cancer treatment. Based on these considerations, the present review aimed to critically discuss the role of RBPs in the occurrence, development and metastasis of breast cancer.

Eukaryotic translation initiation factor (eIF)4A3 and eIF4E

Translation initiation in eukaryotes is instigated by the binding of the eIF4F complex to the M7GTP cap structure of mRNA (15). The eIF4F complex contains the cap-binding protein eIF4E, the scaffold protein eIF4G and the ATP-dependent DEAD box RNA helicase eIF4A.

An isoform of eIF4A, eIF4A3, is the core component of the exon junction complex (16). eIF4A3 has been identified to serve an important role in post-transcriptional regulation, and is thought to be critical for mRNA splicing, transport, translation and monitoring (17,18). eIF4A3 has been reported to be overexpressed at the transcriptional level in common malignancies, and its overexpression in breast cancer is associated with shorter distant metastasis-free, and disease-free and overall survival (19). In addition, the knockdown of eIF4A3 has been shown to result in a defect in nonsense-mediated decay (20,21), a quality control system that can degrade mRNA containing premature stop codons to prevent the accumulation of dysfunctional RNAs and proteins (22).

The Coexpedia database predicted the coexpression of genes related to eIF4A3 function (23). eIF4A3 was coexpressed with 312 genes, and the corresponding network revealed that eIF4A3 may regulate apoptosis and the cell cycle through a variety of tumor-related signaling pathways to promote tumor cell migration, invasion and drug resistance. The coexpression partners of eIF4A3 were closely related to the regulation of apoptosis, Wnt signaling, EGFR signaling, MAPK signaling and TNF/NF-κB signaling (Table I) (24,25).

Table I.

Summary of the cellular functions of RBPs in BC.

RBP	Expression in BC	Functions	Pathways/targets	(Refs.)
eIF4A3	Upregulation	Promotes migration, invasion and chemoresistance; inhibits apoptosis; regulates the cell cycle	Wnt, EGFR, MAPK and TNF/NF-κB signaling pathways	(24,25)
eIF4E	Upregulation	Promotes proliferation, migration and angiogenesis; inhibits apoptosis	Cyclin D1	(26–33)
HuR	Upregulation	Promotes inflammation and apoptosis; inhibits adhesion and angiogenesis; regulates the cell cycle	GATA-3, FOXO1, HOx-A5, Yes, Wnt-5A, IGFIR, ERBB2, p21, p53, BRCA1, TNFSF12, CASP2, BAX, IL-8, COX-2, CSF-1-R, CD9, VEGF-A, THBS1, MMP-9 and ΔNp63	(34–71)
IMPs
IMP1	Downregulation	Inhibits proliferation, migration and invasion	E-cadherin, β-actin, α-actinin, Arp2/3, RGS4, GDF15, IGF2, PTG2 and β-catenin	(86–91)
IMP2	Upregulation	Upregulates the autoimmune response, proliferation and migration	PR	(92–94)
IMP3	Upregulation	Promotes migration, invasion, stemness and chemoresistance	CD164, MMP-9, SLUG, Wnt5B, BCRP and PR	(95–98,101,105)
LIN28	Upregulation	Promotes proliferation, migration, stemness, radioresistance and chemoresistance; regulates the cell cycle	MYC, HMGA2 and PI3K-mTOR pathways; cyclin D1/D2, CDC25A, CDK34, CDK6, RKIP, RAD51, RAD21, FANCD2, CDC25, P-gp, let-7, Rb, p21 and Bcl-XL	(113–127)
MSI	Upregulation	Promotes proliferation, chemoresistance and stemness; regulates the cell cycle; inhibits apoptosis, migration and invasion	p21Cip1, Wnt4, β-catenin and Notch signaling pathways; ERBB2 (MSI-1); ERα (MSI-2); CD44, GBX2, Vimentin, EGFR, DNA-PKCS and LIFR	(129,132,133, 136,137, 148–150)
RBM38	Downregulation	Inhibits proliferation, migration, invasion, the EMT and; chemoresistance induces cell arrest in the G1 phase	p21, C-Myc, p63, MDM2, p53, PR, Erα, PTEN, ZO-1, Mutp53 and STARD13-correlated ceRNA network	(158,161–163,168–171)
SAM68	Upregulation	Promotes proliferation, migration and invasion	Bcl-XL, CD44, SGCE, cardiophilin, cyclin D1, Src, BRK, P59fyn, PI3K, PRMT, FBP21, FBP309, PRMT1, p21 and p27	(172,173,176–181, 186,190,191,194)

BC, breast cancer; RBP, RNA-binding protein.

There is growing evidence suggesting that eIF4E phosphorylation may be involved in numerous key processes in tumor biology, including cell proliferation, transformation, apoptosis, tumor metastasis and angiogenesis. A previous study reported that eIF4E is overexpressed and hyperphosphorylated in various subtypes of breast cancer; overexpression of eIF4E can lead to increased protein levels of cyclin D1, which is required for cells to enter S phase and is associated with cell proliferation (26) (Fig. 1). The activity of eIF4E is regulated by phosphorylation of two MAPK-interacting kinases (MNKs, i.e., MNK1 and MNK2) at a single residue (Ser209) (27,28). Inhibition of MNKs can reduce cyclin D1 synthesis and inhibit the proliferation of breast cancer cells (29). Robichaud et al (30) reported that eIF4ESer209A mice, knock-in mice in which serine 209 was replaced by an alanine residue, were resistant to lung metastasis of breast cancer, and cells isolated from these mice had impaired invasion activities. The same group also revealed that in a mouse model of breast cancer, phosphorylation of eIF4E promoted neutrophil survival and accumulation, thereby promoting metastasis to the lung (31). Chrestensen et al (32) demonstrated that the levels of phosphorylated eIF4E were increased in HER2-overexpressing breast cancer cell lines, whereas inhibition of MNKs by the MNK inhibitor CGP 57380 reduced the proliferation of these cell lines. These studies point to the key role of MNKs in metastatic events. Notably, activation of the MNK signaling pathway downstream of EGFR/HER2 promotes XIAP expression and NF-κB activity in inflammatory breast cancer (33). Taken together, these data suggest that eIF4E inhibition is a promising target for treating breast cancer.

Figure 1.

eIF4E, its target genes and its modulators in cellular pathways. Arrows indicate activation.

Hu-antigen R (HuR)

HuR is a post-transcriptional regulator of RNA binding and a member of the Hu/ELAV family (34). Clinical studies have shown that elevated HuR expression levels and cytoplasmic expression patterns are associated with breast malignancies (35). HuR expression levels have been revealed to be elevated in atypical ductal hyperplasia, ductal carcinoma in situ and ductal infiltrating carcinoma compared with the levels in healthy tissue (36–38). Cytoplasmic HuR expression has also been associated with malignancy, and binding of HuR to different target mRNAs can result in differential regulation of tumor-related processes. Notably, HuR has been shown to promote the growth of nontumorigenic MCF-10A cells and ER+ MCF-7 cells (39–41), but to have little effect on the growth of the highly tumorigenic ER-negative MDA-MB-231 cell line (42).

HuR recognizes and binds the U-rich sequence in the 3′-UTR of the target mRNA (43), which regulates mRNA stability and the translation of various genes involved in the formation, development, metastasis and treatment of breast cancer (44–47). HuR has been reported to regulate the mRNA expression levels of coding proteins involved in transcription [e.g.,GATA-3 (48,49), FOXO1 (50) and HOx-A5 (51,52)], cell signal transduction [e.g., Yes (53), WnT-5A (54), IGFIR (55) and ERBB2 (56)], cell cycle progression [e.g., p21 (57–59), p53 (60) and BRCA1 (61)], apoptosis [e.g., TNFSF12, CASP2 and BAX (62)], inflammation [e.g., IL-8 (63), COX-2 (64) and CSF-1-R (65)], adhesion and angiogenesis (e.g., CD9 (66), VEGF-A (67), THBS1 (67) and MMP-9 (68)] (Fig. 2; Table I). These processes are associated with cancer and malignant transformation. In addition, the direct effect of HuR on the translation efficiency may be positive or negative. Overall, HuR has been demonstrated to bind 38 protein-coding mRNAs in breast cell lines to post-transcriptionally regulate their expression (69). In most cases, HuR stabilized the target mRNA, but the CD9 antigen mRNA levels were markedly downregulated in MDA-MB-231 HuR-overexpressing cells and upregulated in cells with small interfering RNA (siRNA)-mediated HuR silencing (66). HuR is also thought to influence translation as evidenced by its inhibition of Wnt-5A (70), ΔNp63 (71), IGF-1-R (55) and BRCA1 (61) protein production.

Figure 2.

HuR, its target genes and its modulators in cellular pathways. Arrows indicate activation and blunted lines indicate inhibition.

Exon-intron splicing and polyadenylation of mRNA occurring in the nucleus can also be modulated by HuR (41,71–73). Moreover, HuR can be transported along with bound mRNAs from the nucleus to the cytoplasm (74), and this change in subcellular localization appears to be related to the regulation of HuR function (75). In turn, HuR mRNA and protein expression levels are affected by some proteins and microRNAs (miRNAs/miRs), such as miR-519 (76); hormones, such as 17β-estradiol (77); cyclic GMP-elevating agents, such as nitric oxide (78); and drugs.

The aforementioned findings indicated that HuR may be a promising drug target for the treatment of breast cancer. Several chemicals, such as dehydromutactin (79), MS-444 (79), okicenone (79), quercetin (40), b-40 (40), b-41 (40), mitoxanthrone (80), CMLD-2 (81) and 15,16-dihydrotanshinone (82), have been reported to bind HuR and disrupt the interaction among HuR dimers, HuR and mRNA. However, more research is required to fully understand its potential as a drug target and determine which subtypes of breast cancer and at which stages of the disease HuR-targeting drugs would be most effective.

IMPs

IMPs, also known as insulin-like growth factor 2 (IGF2) mRNA-binding proteins, are highly conserved oncofetal RBPs. Three mammalian IMP paralogs (IMP1-3) are expressed in the majority of organs during embryogenesis, and are considered to serve an essential role in cell migration, metabolism and stem cell renewal. These three homologous genes regain their physiological function in malignant cells and are expressed in a wide range of cancer types, where their expression is often associated with poor prognosis (83–85).

IMP1 has been shown to be commonly expressed in normal adult breast epithelial cells, as well as in mouse and human breast tumor cells (86). Activation of IMP1 may help maintain cell polarity and directional movement by regulating the localization of E-cadherin, β-actin, α-actinin and Arp2/3 complex mRNA; this inhibits chemotaxis and metastasis of breast cancer cells (87–89) (Fig. 3; Table I). In addition, IMP1 expression may upregulate RGS4 mRNA, and downregulate GDF15, IGF2 and PTG2 mRNA, inhibiting tumor cell proliferation and invasion (90). IMP1 expression is regulated by the Wnt pathway, and β-catenin has been shown to bind the IMP1 promoter and activate the gene (91). IMP1 has also been observed to stabilize β-catenin mRNA, suggesting that a positive feedback loop exists between β-catenin and IMP1 in which they regulate the expression of each other (91). High levels of (promoter methylation and significant downregulation of IMP1 expression have been observed in all metastatic breast cancer cell lines, including MTLn3, MDA-MB-435, MDA-MB-231 and 4T1, but promoter methylation and downregulation of IMP1 expression were only slight in nonmetastatic cell lines, including MTC and T47D cells (87,88). It can be concluded from these in vitro and in vivo studies that IMP1 methylation events may become more frequent with increased breast cancer grade, which leads to more IMP1 silencing and downregulation events, resulting in dysregulated effects on IMP1-targeted mRNAs. Notably, IMP1 lacking the KH3/4 domain has been shown to have no RNA-binding activity, thus resulting in the loss of inhibitory function against tumor progression (90).

Figure 3.

IMPs, their target genes and their modulators in cellular pathways. Arrows indicate activation and blunted lines indicate inhibition.

Compared with the luminal or apocrine subtypes, basal breast cancer tissues show overexpression of IMP2 (92). Liu et al (93) further reported that IMP2 was overexpressed in breast cancer tissue and upregulated the autoimmune response. In mice with fusiform and squamous differentiated breast cancer, CCN6/WISP3 knockdown could upregulate IMP2 expression (94). Furthermore, CCN6 secreted from normal mammary epithelium can inhibit the protein expression of IMP2 in breast cancer tissue, thereby regulating tumor growth (94). Thus, it may be inferred that the overexpression of IMP2 is not only a potential biomarker of the occurrence of breast cancer but may also considered a new diagnostic factor.

IMP3 is preferentially expressed in TNBC (95) and is associated with the function of breast cancer stem cells (CSCs) (96). Upon induction by the EGFR signaling-activated mitogen-activated protein kinase pathway in these tumors, IMP3 expression is inhibited by 3βa-diol, the ligand of ERβ (97). Experiments performed in cell lines suggest that ERβ indirectly inhibits IMP3 based on the discovery that ERβ inhibits EGFR transcription (97). Moreover, IMP3 has been reported to bind several key mRNAs that may promote migration and invasion, including CD164 and MMP-9 (97) (Fig. 3; Table I). IMP3 has been shown to promote stem-like properties in TNBC via the stem cell factor SOX2, the transcriptional target of SLUG; in addition, IMP3 may mediate the post-transcriptional regulation of SLUG (98). SLUG, a member of the SNAIL family of transcriptional repressors, is preferentially expressed in TNBC and is involved in the occurrence of these tumors (99,100). Furthermore, IMP3 can indirectly stabilize WNT ligand (WNT5B) mRNA by targeting the miR145-5P binding site on WNT5B mRNA, resulting in TAZ activation by alternative WNT signaling. IMP3 and WNT5B also work together to promote SLUG transcription, which is required for the nuclear localization and activation of TAZ (101). TAZ is preferentially expressed in TNBC, which is of great significance for breast CSCs and poorly differentiated tumors (102). Breast cancer resistance protein (BCRP), also known as ABCG2, is a member of the ATP-binding cassette transporter family and is a major effector of drug resistance in breast cancer (103). It has been shown that doxorubicin and mitoxantrone, two chemotherapeutic drugs used in the treatment of breast cancer, are effluxed from cells by BCRP (104). IMP3 may promote chemotherapy resistance by binding BCRP mRNA and regulating its expression (105).

It has been demonstrated that miR-200a directly suppresses IMP2/3 expression by targeting the 3′-UTR. By contrast, IMP2 and IMP3 may suppress the transcription of miR-200a via recruitment of the CCR4-NOT transcription complex subunit 1 complex to destabilize PR mRNA. Therefore, the identified IMP2/3-miR-200a-PR axis double-negative feedback loop may serve as a novel potential therapeutic target for TNBC (106) (Fig. 3; Table I).

LIN28

LIN28 has been identified as an RBP involved in breast cancer proliferation, metastasis, drug resistance and stem cell renewal. LIN28 has two RNA-binding motifs: A cold shock domain and a Cys-Cys-His-Cys zinc finger domain (107). Mammals produce two similar LIN28 isoforms: LIN28A and LIN28B, which share the same primary structure and function but differ in more subtle ways (108). For example, LIN28B has nuclear localization signals (NLSs) and a nucleolar localization signal, and is mainly localized to the nucleus and nucleoli; however, LIN28A is mainly located in the cytoplasm. The key functions of LIN28 fall into two classifications: Let-7-dependent and let-7-independent. miRNAs of the let-7 family are key inhibitory targets of LIN28 and serve as effective tumor suppressors via post-transcriptional inhibition of a variety of oncogenic mRNAs (109). Previous studies have indicated that LIN28 is transcriptionally activated by upstream factors, such as C-Myc, NF-κB, Src and Wnt (110,111).

Breast cancer cells are characterized by a hypoxic microenvironment and extracellular acidosis. During carbonic anhydrase IX (CAIX)-mediated adaptation to hypoxia and acidosis in carcinogenesis, CAIX increases the LIN28 protein levels, followed by pyruvate dehydrogenase kinase 1 upregulation and enhanced glycolysis. CAIX is a hypoxia-induced pH regulator that modulates LIN28/Let-7 axis-mediated metabolic metastasis and stemness in breast cancer cells (112). Arguably, the most fundamental feature of LIN28 in breast cancer cells is its ability to promote proliferation. Numerous studies on tumor proliferation have shown that LIN28 acts as an oncogene by inhibiting let-7, resulting in the dysregulation of multiple genes directly or indirectly regulated by let-7, including components of MYC (such as C-Myc, N-Myc and L-Myc), HMGA2 and PI3K-mTOR pathways (113–116) (Fig. 4). In terms of the cell cycle, LIN28 may promote the proliferation of tumor cells by inhibiting let-7 and increasing the expression of cyclin-related factors, such as cyclin D1/D2, CDC25A, CDK34 and CDK6 (117) (Table I). In addition, the LIN28/let-7 axis has been reported to be involved in preventing angiogenesis in breast cancer. Isanejad et al (118) reported that the expression of aggressive breast cancer cell markers (such as Ki67 and ERα) or tumor vascular markers (such as HIF-1α, CD31 and VEGF) could be downregulated through the let-7a pathway when combined with hormone therapy.

Figure 4.

LIN28, its target genes and its modulators in cellular pathways. Arrows indicate activation and blunted lines indicate inhibition.

LIN28 may also contribute to bone metastasis of breast tumors expressing Raf kinase inhibitory protein (119). The overexpression of LIN28 in breast cancer cells has been reported to significantly reduce the expression of E-cadherin and increase the expression of vimentin, thus promoting metastasis (120). Furthermore, let-7a can inhibit cell migration in breast cancer cell lines by significantly blocking the direct binding target of LIN28 (121). Low expression of LIN28 and overexpression of let-7 can increase the radiosensitivity of cancer cells by reducing the expression of RAS oncogenes and DNA-related genes, such as RAD51, RAD21, FANCD2 and CDC25 (122–124). In terms of drug resistance, the overexpression of LIN28 has been demonstrated to reduce the sensitivity of cells to chemotherapy by inhibiting miR-107 expression, as well as the RNA and protein expression levels of C-Myc and P-gp (125). Other investigators have reported that LIN28-induced chemotherapy resistance is associated with let-7, Rb, p21 and Bcl-XL, thus clarifying the complex relationship between LIN28 and tumor drug resistance (126,127). These findings provide evidence for the potential therapeutic effect of a strategy targeting LIN28 against breast cancer metastasis.

Musashi (MSI)

MSI is an evolutionarily conserved RBP, which modifies translation by binding to (G/A)U1-3(AGU) motifs in the 3′-UTR of its target mRNAs (128). Two different proteins have been identified: MSI-1 and MSI-2. MSI-1 is associated with a variety of tumor types, and its function is essential for tumor growth in breast cancer, colon cancer, medulloblastoma and glioblastoma. MSI-1 regulates apoptosis, differentiation, proliferation and cell cycle progression by mediating different post-transcriptional processes (129–131). In addition, MSI-1 may be a target gene of the Wnt pathway; the target protein p21 (Cip1) negatively regulates Wnt4 and β-catenin, through which MSI-1 expression can be automatically regulated (132,133). This finding is consistent with MSI-1 activating Notch and Wnt activity, and the nuclear localization of β-catenin in mammary epithelial cells (134). In addition, the overexpression of C-Myc has been reported to result in a significant increase in MSI-1 protein and mRNA expression (135). MSI-1 was shown to be strongly correlated with the Notch pathway, and to serve an important role in stem cell self-renewal and cell fate determination (136,137). Notch is activated by the sequential proteolytic cleavage of its membrane-associated form to a constitutively active intracellular form [Notch intracellular domain (NIC)], which serves as a transcription coactivator (138). The maintenance of NIC is affected by the negative regulator Numb, which ubiquitinates and targets NIC for proteasomal destruction (139,140), and its expression is inhibited by MSI-1 (139) (Fig. 5). The NIC and Notch ligands Jagged and Delta have been shown to be highly expressed in breast cancer (141,142), and it has been reported that 50% of high-grade human breast cancer cases exhibit a loss of Numb expression, which is negatively correlated with the tumor grade (143). Furthermore, MSI-1 has been shown to be highly expressed in ~40% of primary breast tumors and 100% of lymph node-positive tumors and corresponds to a poor prognosis for survival (129). In breast cancer cell lines, MSI-1 expression is also associated with HER2 activity (129), and HER2-induced cell proliferation and cyclin D1 expression depend on Notch activation and Numb inhibition (136). Therefore, these findings suggested that MSI-1 may have an important role in mediating the progression of HER2-positive breast cancer.

Figure 5.

MSI, their target genes and their modulators in cellular pathways. Arrows indicate activation and blunted lines indicate inhibition.

MSI-2 has been identified as a novel ubiquitination target protein of deleted in breast cancer 2 (DBC2) (137). DBC2, also known as Rho-related BTB domain-containing protein 2, is classified as a tumor suppressor gene (144–146) that functions as a substrate-specific adaptor protein for a novel class of Cullin-3-based E3 ubiquitin ligases (147). MSI-2 interacts directly with DBC2, and this interaction promotes MSI-2 polyubiquitination and proteasomal degradation in breast cancer cells (137). This provides evidence that DBC2 may inhibit the occurrence of breast cancer via MSI-2 ubiquitination. In ER+ breast cancer, MSI-2 expression is highly enriched, and MSI-2 expression is significantly correlated with estrogen receptor 1 (ESR1) expression (Fig. 5; Table I). Furthermore, MSI-2 can increase the stability of the ESR1 protein by binding to the 3′-UTR of ESR1 mRNA (148). In summary, MSI-2 may act as the upstream regulator of ESR1, thus having clinical significance in ER+ breast cancer.

Previous studies have shown that both MSI-1 and MSI-2 are highly expressed in TNBC (128,149). In breast cancer, the expression levels of CD44, GBX2 and the mesenchymal protein vimentin have been shown to be downregulated upon MSI knockdown (150); these three proteins are key stem cell markers of breast malignant tumors (151–153). This strongly points to the effect of MSI on proliferation and potential apoptosis. Furthermore, MSI protein knockdown can reduce radiotherapy resistance in breast cancer by downregulating EGFR and DNA-PKCS expression (150). However, MSI protein knockdown can lead to higher cell invasiveness and increased migration in vivo, possibly due to downregulation of the LIF receptor (LIFR) (150). In breast cancer, LIFR is referred to as a metastasis inhibitor and is upstream of the Hippo-YAP pathway: high LIFR expression has been reported to inhibit metastasis by inactivating the transcription coactivator YAP through a cascade of polyphosphorylation processes (154). In summary, these data suggested that targeting MSI has potential therapeutic value; however, since metastasis is a key determinant of overall survival, increased cell migration and invasion are clear and troubling consequences of MSI silencing. Further offsetting the invasive and migratory nature of MSI silencing is critical before determining its therapeutic value in breast cancer.

RNA-binding motif protein 38 (RBM38)

RBM38, also known as RNPC1, belongs to the RRM family of RBPs, and may affect the proliferation, cell cycle arrest and epithelial-mesenchymal transition (EMT) of cancer cells (155–157). In 2014, to the best of our knowledge, Xue et al (156) reported for the first time that RBM38 was downregulated at the mRNA and protein expression levels in multiple breast cancer cell lines, as well as in 121 pairs of human breast cancer and adjacent normal tissues. RBM38 is expressed as two isoforms: RBM38a and RBM38b. In vitro, RBM38a overexpression has been reported to inhibit the proliferation, migration and invasion of ER+ MCF-7 and triple-negative MDA-MB-231 cells by arresting cells in G1 phase (156). In a clinical sense, RBM38 has been shown to be positively correlated with long-term relapse-free and overall survival (158).

RBM38 has been shown to induce G1-phase arrest by binding and stabilizing p21 transcripts (159) (Fig. 6; Table I), and its expression may be negatively correlated with C-Myc at the protein level (160). C-Myc is a well-known oncogene that is overexpressed in numerous types of human cancer and promotes G1/S cell cycle progression (161). Although C-Myc is generally considered a transcriptional activator, it binds the E-box region of the RBM38 promoter to inhibit its transcription, which, in turn, results in the destabilization of C-Myc mRNA in ER+ breast cancer cells (160). RBM38 also has negative effects when regulating the transcription of its target (such as binding the mRNAs of P63, MDM2 and p53), mediating the instability of its mRNAs and attenuating their translation. In addition, RBM38A protein expression has been shown to be significantly positively correlated with that of ERα (162) and PR (163). RBM38 can stabilize the mRNA expression levels of PR and ERα, but not ERβ, in ER+/PR+ breast cancer cells. However, ERα has been reported to negatively regulate RBM38 expression in response to estrogen stimulation (162). RBM38 may serve a role in tumor suppression by partially enhancing PTEN expression. PTEN, a well-characterized tumor suppressor, is an inhibitor of the PI3K/Akt pathway, which is frequently overactivated in several types of cancer and is associated with metastasis and CSCs (164–167). RBM38 can positively affect the expression and activity of PTEN in breast cancer cells by acting on the PTEN 3′-UTR (168).

Figure 6.

RBM38, its target genes and its modulators in cellular pathways. Arrows indicate activation and blunted lines indicate inhibition.

EMT alters the polarity of epithelial cells and stem cell properties, thus contributing to metastasis and drug resistance (169). Notably, during TGF-β-induced EMT in ER+ breast cancer cells, RBM38 expression is lost via Snai1-mediated transcriptional suppression through the E-box element in the RBM38 promoter (170). Biologically, RBM38 blocks the EMT by upregulating the mRNA stability of the epithelial marker ZO-1 (170). In addition, RBM38 overexpression has been shown to decrease the expression of Mutp53 protein in breast cancer. Mutp53 could induce partial EMT-like transitions as reflected in the increased suppression of E-cadherin synthesis (171); thus, Mutp53 may be involved in the RBM38-regulated EMT process. Notably, RBM38 may promote competing endogenous RNA (ceRNA) interactions among STARD13, CDH5, HOXD10 and HOXD1 (STARD13-correlated ceRNA network) in breast cancer tissues; by promoting the expression of these four genes in breast cancer cells, breast cancer cell metastasis is inhibited and Adriamycin resistance is attenuated (158).

In summary, the genomic characterization and expression pattern of RBM38 provide evidence for its tumor-suppressive effect in breast cancer and highlight the diversity of its mechanisms in different biological contexts. These data also suggest the potential value of RBM38 with respect to future targeted tumor therapies.

SAM68

SAM68 was originally identified as a substrate for Src-associated in mitosis and is a member of the signal transduction and RNA activation family (172,173). SAM68 is considered to be an RBP that links extracellular signal transduction and RNA processing (174,175), and has an hnRNP K homologous domain, which is required for binding to RNA with high specificity and high affinity. It has been reported that SAM68 can nonspecifically interact with the poly(U) and poly(A) chains of RNA molecules, and can also specifically interact with UAAA or UUUA structures (174). Selective splicing of multiple genes, such as Bcl-XL, CD44, SGCE, cardiophilin and cyclin D1, has been shown to be regulated by SAM68 (176,177) (Table I). In addition, SAM68 has been reported to interact with numerous signaling proteins, such as Src, BRK, P59fyn, PI3K, PRMT, FBP21 and FBP309, through their SRC homologous (SH)2, SH3 and WW domains, suggesting that SAM68 may also be involved in various biological processes as an adaptor protein for signal transduction (172,173,178–181).

SAM68 contains six proline-rich sequences and a tyrosine-rich region at the C-terminus, which form docking sites for signaling proteins containing SH2 and SH3 domains (180–183). Notably, the tyrosine phosphorylation of Src-associated kinases has been reported to impair SAM68 homodimerization (173), and its RNA affinity in vitro (181,184) and in vivo (185). Other post-translational modifications have also been reported to affect the function of this RBP. SAM68 binds and is methylated by the arginine methyltransferase PRMT1 (186), thus influencing SAM68 and SH3 domain interactions (187) and its nuclear localization (186) (Fig. 7). Furthermore, SAM68 acetylation by the acetyltransferase CBP in oncogenic breast cancer cell lines (188) has been shown to increase the binding of SAM68 to RNA. In addition, SAM68 can be sumoylated by the SUMO E3 ligase PIAS1, thereby enhancing its transcriptional inhibitory activity (189). Thus, post-translational modifications greatly influence the biochemical properties of SAM68 and fine-tune its subcellular localization, interactions with signaling proteins, and RNA binding affinity.

Figure 7.

SAM68, its target genes and its modulators in cellular pathways. Arrows indicate activation and blunted or crossed-out lines indicate inhibition.

The tyrosine phosphorylation of SAM68 may also affect its role in breast cancer cells. The breast tumor kinase BRK is an overexpressed nonreceptor tyrosine kinase in human breast cancer cells (190), which can promote proliferation and anchorage-independent growth (191). SAM68 was identified as one of the first substrates of BRK, and its phosphorylated tyrosine residue overlaps with the NLS of SAM68. BRK-dependent phosphorylation has been reported to induce transient subcellular relocation of SAM68 in cells stimulated with EGF to induce proliferation (192). Because tyrosine phosphorylation reduces the RNA-binding activity of SAM68 while enhancing its interaction with signaling proteins (193), it is likely that BRK activation leads to functional reprogramming of SAM68 activity in breast cancer cells. Notably, SAM68 and BRK have been reported to be upregulated in breast cancer, supporting their role in cell proliferation and invasion (190,191,194).

SAM68 can also promote polarization movements and cell migration independent of its RNA-binding activity (195). The positive role of SAM68 in tumor transformation has been demonstrated in other types of human cancer (196–198). It has been demonstrated that SAM68-deficient cells exhibit intron 5 retention in mTOR mRNA, resulting in an early termination codon and a reduction in the mTOR protein levels (199). Notably, mTOR is a key effector in the cellular signaling pathway of human cancer, and the overexpression of related components in the PI3K/AKT/mTOR pathway has been demonstrated to induce malignant transformation (200). Notably, the loss of SAM68 can reduce the incidence of breast cancer, suggesting that SAM68 activation may also regulate the expression of PI3K downstream mTOR in breast cancer cells (201). SAM68 deficiency in breast cancer cells may also impair cell proliferation and tumorigenic properties by upregulating the expression of the cyclin-dependent kinase inhibitors p21 (Cip1) and p27 (Kip1). Thus, in this case, the loss of SAM68 may lead to reductions in Akt phosphorylation and the subsequent activation of the FOXO factor, thereby promoting the expression of p21 and p27 (194). This finding is consistent with the role of SAM68 in breast tumorigenesis in vivo. SAM68-knockout mice have been reported to be insensitive to tumor formation in vivo (202), and SAM68 haploid insufficiency can delay the onset of breast tumors and reduces the spread of metastases (201). These results suggested that high levels of SAM68 are necessary for cell transformation in vivo and support the role of SAM68 as a proto-oncogene. In addition, in patients with breast cancer, the expression and cytoplasmic localization of SAM68 have been shown to be significantly correlated with their clinical characteristics, including clinical stage, tumor-node-metastasis grade, histological grade and ER expression status (194). Thus, SAM68 may be considered an attractive target for breast cancer treatment; however, a better understanding of its functions in breast cancer cells is required to develop tools to interfere with its activity.

Conclusions

Numerous studies have shown that disruptions in RBPs occur in different subtypes of breast cancer and affect every step of breast cancer development. RBPs regulate gene expression to induce or reduce the expression levels of genes associated with breast cancer. As described in this review, some RBPs have been found to regulate multiple genes involved in breast cancer development simultaneously, leading to different changes in cancer progression. With the development of research techniques, such as single-cell analysis, and cross-linking and immunoprecipitation, several new RBPs and their partners have been discovered. However, in general, the complex regulatory network of RBPs is still not fully understood, as much remains to be discovered with respect to the role of RBPs in breast cancer biology.

Based on the results of previous studies, researchers have attempted to target RBPs and/or their partners in clinical and nonclinical studies using siRNA, antisense-oligonucleotides and small molecules. However, for most of the RBPs described in the present review, no drugs are currently being tested, let alone clinically available. More research should be conducted to expand the understanding of RBP interaction networks and to develop specific ways to target RBPs in cancer therapy without affecting adjacent normal cells.

In conclusion, the present review highlights the regulatory role of RBPs in the occurrence and progression of breast cancer. Considering the number of RBPs whose functions are still not completely understood, current knowledge about breast cancer-related RBPs is in its infancy, thus reinforcing the need for future studies.