RNA Binding Proteins (RBPs) and their role in DNA Damage and radiation response in cancer

Abstract

Traditionally majority of eukaryotic gene expression is influenced by transcriptional and post-transcriptional events. Alterations in the expression of proteins that act post transcriptionally can affect cellular signaling and homeostasis. RNA binding proteins (RBPs) are a family of proteins that specifically bind to RNAs and are involved in post-transcriptional regulation of gene expression and important cellular processes such as cell differentiation and metabolism. Deregulation of RNA-RBP interactions and any changes in RBP expression or function can lead to various diseases including cancer. In cancer cells, RBPs play an important role in regulating the expression of tumor suppressors and oncoproteins involved in various cell-signaling pathways. Several RBPs such as HuR, AUF1, RBM38, LIN28, RBM24, tristetrapolin family and Musashi play critical roles in various types of cancers and their aberrant expression in cancer cells makes them an attractive therapeutic target for cancer treatment. In this review we provide an overview of i). RBPs involved in cancer progression and their mechanism of action ii). the role of RBPs, including HuR, in breast cancer progression and DNA damage response and iii). explore RBPs with emphasis on HuR as therapeutic target for breast cancer therapy.

1. Introduction

Breast cancer (BC), a heterogenous complex disease characterized by dysregulation of multiple genes and associated regulatory proteins, is the leading cause of cancer-related death among women. Current treatment modalities for BC include surgery, radiation therapy, chemotherapy, hormone therapy and novel targeted therapy. Although tremendous progress has been made in redefining treatment approaches for BC, a complete understanding of the molecular mechanisms underlying its origin and progression is still lacking [1]. Six major subtypes of breast cancer have been identified by molecular profiling and each of them exhibit different survival outcomes [2]. Current research aims to identify novel therapeutic targets and biomarkers for tailoring patient specific treatment options and develop a comprehensive approach towards effective disease management of BC patients.

It is now widely accepted that post-transcriptional regulation of gene expression is central to cancer progression. Studies over the last two decades have irrevocably established the role of RBPs in post-transcriptional regulation of gene expression, and its impact in modulating cellular processes such as cell growth, differentiation and metabolism in eukaryotic cells [3]. RBPs can bind to single or double stranded RNA in a sequence specific manner and regulate mRNA life cycle including RNA splicing, polyadenylation, mRNA stability and localization [4, 5]. The binding of RBPs to structural motifs or RNA sequences occurs through limited group of structurally well-defined RNA-binding domains (RBDs) that result in the formation of ribonucleoprotein (RNP) complexes. RBPs can be grouped into conventional/canonical and non-conventional/non-canonical RBPs. Conventional RBPs are grouped based on their RBD and comprise of RNA recognition motif (RRM), cold-shock domain (CSD), K homology (KH) domain, DEAD/DEAH helicase and zinc-finger domains [6]. Figure 1 summarizes basic domains of various RBPs belonging to the RRM family, heterogeneous nuclear ribonucleoprotein family (hnRNP), Tristetrapolin (TTP), Musashi1 (MSI1), Lin28 and HuR. Non-conventional RBPs lack RBDs and bind through intrinsically disordered regions [7, 8]. Such RNA binding regions have been identified in metabolic enzymes that are involved in glycolysis and TCA cycle, heat shock proteins and transcription factors [8, 9]. Genome-wide screening and mass spectrometry studies have identified about 1500 RBPs and their RNA targets accounting for about 7.5% of the total protein-encoding genes in human genome [3]. RBPs, in addition to binding to RNA, can also bind to DNA [9]. Hudson et al., conducted gene ontology analysis of several DNA- and RNA-binding proteins (DRBPs) and indicated that DRBPs comprise about 2% of the human proteome and play a role in regulating transcription, mRNA processing, DNA replication as well as the DNA damage response (DDR) [9, 10]. Given the fact that RBPs critically control the expression of various genes, any perturbation in RBP expression levels leads to neurodegenerative diseases, kidney diseases, diabetes, cardiovascular disorders and cancer [6, 11-13]

Figure 1.

Functional RNA-binding domains of some RBPs implicated in cancer progression. These includes the RNA-recognition motif (RRM)-the most common identified RNA-binding domain, zinc finger motifs CCCH (Cys3His) or CCHC, arginine-glycine-glycine (RGG) motif and CSD (cold-shock domain). The scale on top of the schematic indicates amino acid length. Images were created with Biorender.com

2. RBPs in Breast Cancer

RBPs are vital players in the process of post-transcriptional gene regulation, and hence any dysregulation or mutation in RBPs can potentially lead to incorrect binding of RBPs to target RNA molecules further impacting gene regulation resulting in cancer. RBPs are found to be dysregulated in various cancer tissues such as lung, liver, pancreas and breast (Table 1) compared to the healthy tissues [14]. The online RBPDB database and the Cancer Genome Atlas (TCGA) provides extensive information on various RBP domains and differential expression levels of RBPs in cancer and normal samples [5, 15]. TCGA and integrated bioinformatics analysis identified various RBPs involved in BC signaling and tumorigenesis. Two hundred and five RBPs were identified in the study by Wang et.al., of which 90 were upregulated and 115 were downregulated in BC samples [16]. Lan et al., demonstrated that expression of a panel of RBPs namely MRPL13, IGF2BP1, BRCA1, and MAEL were implicated in predicting patient prognosis in BC [17]. In the study by Wang et al., overexpression of RBPs DCAF13, EZR, and MRPL13 correlated with lower survival whereas overexpression of APOBEC3C and eIF4E3 correlated with higher survival in BC patients [16]. Thus, alteration in the RBP expression can be utilized to assess disease progression and predict patient survival in BC.

RBP	Tumor type	Expression levels	Targets	Role in RNA regulation	Biological functions in cancer	References
LIN28A/B	Colon, Breast, Brain, cervical, Esophageal, Head and Neck, Liver, Lung, Renal, Prostate, Ovarian	upregulated	let-7 family members, HMGA1, PD-L1	miRNA processing, mRNA translation	Sustained proliferation, invasion, migration, metastasis, angiogenesis, EMT	[3, 6, 242, 243]
CPEB1	Breast, Liver	upregulated	BUB3, MMP9, VEGF, ZO-1, HIF-1α	mRNA translation, polyadenylation	EMT, angiogenesis, proliferation, invasion, metastasis	[6, 244-246]
CPEB2	Breast	upregulated	HIF-1α	polyadenylation	Metastasis	[246]
CPEB4	Breast, Liver, Pancreas, Melanoma, Brain	upregulated	MITF, RAB72A, TPA, VEGF	mRNA translation, polyadenylation	Angiogenesis, proliferation	[245, 247-249]
ESRP1 and ESRP2	Breast, Colon, Lung, Pancreas, Renal, Head and Neck, Melanoma, Prostate, Ovarian	upregulated or downregulated	FGFR2, CTNND1, Snail, CD44, ENO1, ENAH, Myc, PKM2, CCND1	Alternative splicing, mRNA translation	Migration, invasion, EMT, metastasis	[6, 250-255]
MSI1 and MSI2	Breast, Brain, Colon, Lung, Bladder, Ovarian, Pancreatic, Gastric, Liver, Leukemia	upregulated	CDKN1A, NUMB, HMGA2, BRD4, p27, PTEN, c-met	mRNA translation, mRNA stability	EMT, proliferation, invasion, apoptosis, metastasis	[3,154,211, 256, 257]
SRSF1 (SF2/ASF1)	Breast, Colon, Lung, Kidney, Prostate	upregulated	S6K1, RON, BIN1, MKNK2, Bcl-x, Mcl-1, CASP2, CASP9, TEAD1, VEGF, RPL-MDM2, p53	Alternative splicing	EMT, migration, proliferation, senescence, angiogenesis	[3, 258-262]
SRSF3	Breast, Colon, Lung, Kidney, Brain, Stomach, Thyroid	upregulated	FoxM2, HIPK2, ERN1, HMGCS1, DHCR7, SCAP, LIFR, Epb4.1l5, ERBB2, Myo1b, CTNND1, GIT2, SLK, SP4	Alternative splicing	EMT, apoptosis, proliferation, metastasis	[3, 129, 263, 264]
SRSF6	Lung, Colon	upregulated	MKNK2, INSR, DLG1	Alternative splicing	Proliferation	[265]
hnRNP A1	Lung, Liver, Brain, Breast, Colon	upregulated	miR-18a, let-7a	miRNA processing	Proliferation	[266-271]
hnRNP A2/B1	Breast, Brain, Lung	upregulated	CFLAR, WWOX, BIN1, CASP9, CD44, TP53IP2, E-cadherin, Twist, Snail1	Alternative splicing	EMT, proliferation, metastasis, apoptosis	[268, 270, 272-274]
hnRNP D (AUF1)	Breast, Lung, Liver, Colon, Pancreas, Renal, Thyroid	upregulated	Bax, p21, p16, Bcl-2, CyclinD1, CASP2, MMP9, FGF9 FOS, JUN, Myc, NEAT1, Gadd45a	mRNA stability, mRNA translation	EMT, migration, invasion, senescence	[6, 21, 33, 275, 276]
hnRNP E1/2 (PCBP1/2)	Breast, Lung, Colon, Pancreas, Liver, Brain, Prostate, Cervical, Esophageal	upregulated or downregulated	GSL2, Dab2, ILEI, PRL-3, p73, Inhibin βA	mRNA stability, mRNA translation	EMT, senescence, invasion, metastasis	[3, 129, 277-279]
hnRNP M	Breast	upregulated	CD44	Alternative splicing	Invasion, metastasis, EMT	[96, 154, 251]
hnRNP H/F	Breast, Brain	upregulated	ErbB2, RON, Mcl-1	Alternative splicing	Proliferation, invasion, migration	[280-282]
hnRNP I (PTB)	Breast, Brain, Colon, Ovarian	upregulated	USP5, CyclinD3, Cyclin D1, PKM, FGFR-1, PTEN	Alternative splicing	Proliferation, invasion	[3, 268, 270, 283, 284]
hnRNP K	Breast, Head and Neck, Lung, Liver, Bladder, Melanoma, Colon, Prostate, Gastric	upregulated or downregulated	MAP1B-LC1, TNFR2, Cyclin D1, ERCC4, MRPL33	mRNA transcription, mRNA translation	EMT, proliferation	[3, 239, 285-288]
RBM 4	Breast, Liver, Lung, Colon, Ovarian, Pancreatic, Gastric	downregulated	BCL-X, TEAD4, CD44, PKM2, FGFR2	Alternative splicing	Proliferation, migration, apoptosis	[6, 239]
RBM 9 (Rbfox2)	Breast, Ovarian, Colon	downregulated	RB1	Alternative splicing	Invasion, EMT, Metastasis, migration	[6, 289]
RBM 10	Liver, Lung, Pancreas, Thyroid	downregulated or mutated	NUMB, p53, CREBBP	Alternative splicing	Proliferation, apoptosis	[3, 6, 290]
RBM 38, 39	Breast, Lymphoma, Colon, Liver, AML, Kidney	upregulated or downregulated	p53, PTEN, E cadherin, Vimentin, Mdm2, ZO-1, CDKN1A	mRNA translation, mRNA stability, Alternative splicing	Proliferation, apoptosis, EMT, Migration, Invasion	[6, 25, 96, 126, 219, 291-294]
RBM 47	Breast, Colon, Lung	downregulated	DKK1, Nrf2	mRNA stability	Proliferation, metastasis, EMT	[295-297]
KHSRP	Breast, Brain, Bladder, Colon, Lung, Cervical, Liver, Pancreatic, Ovarian, Prostate, gall bladder	upregulated or downregulated	miR-26a/b, let-7a, miR-23a, miR-192-5p, miR-130b, miR-30, miR-21	miRNA processing, mRNA stability	Metastasis, EMT, proliferation, invasion	[96, 298-306]
HuR	Breast, Bladder, Ovarian, Pancreatic	upregulated in most cancers	IL-1β, IL-2, IL-8, TNF-α, c-fos, p21, p16, p27, MMP9, HIF-1α, VEGF, SIRT1, Snail, c-Myc, Wnt5a, COX-2, Cyclin (A, B1, D1, and E), TIN2, MSI1, Bcl-2,BclxL, Mcl-1, ProTα, lncRNA NEAT1, lncRNA-HGBC, OIP5-AS1, LincRNA-p21, lncRNA HOTAIR, ProT , p53, MSI1, HIF-1, XIAP	mRNA stability, mRNA translation	Angiogenesis, apoptosis, proliferation, senescence, invasion, metastasis	[3, 6, 33, 41, 44, 54, 59, 60, 65, 184, 185, 235, 248, 307-315]
PUM1 and PUM2	Cervical, Head and Neck, Liver, Leukemia	downregulated or upregulated	E2F3, JUN, N-Ras, lncRNA NORAD, VEGF, vimentin	mRNA translation	Proliferation, Genomic imbalance, EMT, Invasion	[3, 6, 316, 317]
LARP1	Breast, Brain, Prostate	upregulated	Bax, Bcl-2, Bik, Mdm2, 5'TOP mRNAs, XIAP	mRNA stability, mRNA translation	Apoptosis, Proliferation	[3, 318, 319]
LARP4B	Breast	downregulated	Bax, CDKN1A, p53	mRNA stability	Proliferation, invasion	[6, 320, 321]
LARP6	Breast, Cervical, Gastric	upregulated	VEGF, MMP9	mRNA translation	Angiogenesis, migration, invasion	[6, 321, 322]
LARP7	Breast, Colon, Lymphoma, Cervical, Melanoma, Bladder, Liver, Lung, Gastric, Prostate, Head and Neck	downregulated	P-TEFb, 7SK snRNP	snRNA stability	Metastasis, EMT, invasion	[6, 323-326]
elF4E	Breast, Renal, Colon, Leukemia, Prostate	upregulated	Bcl-2, Bcl-xL, VEGF, FGF2, MMP2, MMP9	mRNA translation	Apoptosis, EMT, angiogenesis, metastasis	[3, 6, 161, 327-330]
SAM68	Breast, Leukemia, Lung, Colon	upregulated	Bcl-x, Cyclin D1, CD44	Alternative splicing	EMT, invasion, metastasis, proliferation	[3, 6, 331-336]
TIA-1/TIAR	Breast, Lung, Colon, Melanoma, Ovary, Liver, Skin	upregulated	BRCA1, VEGF, CDKN1A, PDCD4, myc, elF4E2, GADD45A, GADD45B, Fas	mRNA stability, mRNA translation, Alternative splicing	Angiogenesis, apoptosis, proliferation	[3, 80, 337, 338]
IGF2BP1 (IMP1/ZBP1)	Breast, Brain, Leukemia, Lung	upregulated	β-TrCP1, CD44, lncRNA HULC, β-actin, E-cadherin, α-actinin, Arp-16	mRNA stability, mRNA translation, mRNA localization	EMT, invasion, metastasis, proliferation	[3, 6, 17, 339]
IGF2BP2 (IMP2)	Breast, Leukemia, Lung, Colon, Pancreas, Liver, Renal	upregulated	NDUFS3, COX7b, RAF1, PR	mRNA stability, mRNA localization	EMT, invasion, metastasis	[3, 6, 243]
IGF2BP3 (IMP3)	Breast, Colon, Leukemia, Lung, Ovary, Renal, Pancreas	upregulated	HMGA2, LIN28, Myc, PR	mRNA stability, mRNA translation	EMT, invasion, metastasis, proliferation	[3, 6, 339]

The following sections summarize some of the well-studied RBPs and describe their profound roles in BC progression. The readers are informed that the information provided on RBPs is not comprehensive nor exclusive for BC and are therefore directed to refer to recent reviews [18].

AUF1:

The AU-rich element RNA-binding protein 1 (AUF1), also known as heterogeneous nuclear ribonucleoprotein D (hnRNP D), is an RBP that can both stabilize and destabilize various cancer-related gene transcripts. AUF1 expression levels are increased in several tumors including BC [19-21]. Overexpression of AUF1 in multiple cancer types activates ERK1/2 and Akt pathways leading to cancer progression and metastasis [19]. Correspondingly, AUF1 silencing inhibited cell proliferation, migration and invasion, and increased apoptosis [19, 20]. Though this information is available for several cancer types, there is a dearth of first-hand information on AUF1 expression and its correlation with patient demographics in BC samples. However, a preliminary report has linked AUF1 expression via regulation of epithelial-mesenchymal transition (EMT) factors ZEB1, TWIST1 and SNAIL1 as a contributor to stemness in triple negative breast cancer (TNBC) cell lines [22]. Future studies, centered on analyzing the role of AUF1 in BC, are needed to help better understand its involvement in BC.

RRM family:

RBM38 and RBM3, members of the RNA recognition motif (RRM) family of RBPs, play a significant role in post-transcriptional regulation, mRNA splicing, stabilization and translation of gene expression and are overexpressed in numerous cancer types including BC [23, 24]. In BC, RBM38 functions as a tumor suppressor partially by increasing PTEN mRNA stability thereby leading to an increased expression of the tumor suppressor protein PTEN [25]. A study by Li et al., showed overexpression of RBM38 in MCF-7 BC cells suppressed c-myc expression demonstrating the possible involvement of RBM38-c-myc loop in growth suppression of BC [26]. RBM38 was shown to bind to adenylate and uridylate (AU)-rich elements (AREs) in the 3’UTR of ZO-1 mRNA and stabilize its expression that played a role in transforming growth factor β (TGF-β) induced EMT in BC [26]. Xue et al., demonstrated that overexpression of RBM38 diminished the migration and invasion of BC cells by inhibiting mutant p53-induced EMT leading to cell cycle arrest.

RNA binding motif 3 (RBM3), originally discovered to be a cold shock response protein, is essential for cell proliferation and protects the cells from adverse conditions such as hyperthermia, serum deprivation, hypoxia, radiation or drug treatment [27-29]. Although data pertaining to the expression profile and function of RBM3 across cancer types is conflicting, overexpression of RBM3 has been consistently reported in BC [28, 30]. A report by Jogi et al., showed an association of nuclear RBM3 expression with low-grade BC tumors and decreased Ki67 positivity [23]. The authors also claimed nuclear RBM3 expression to be a prognostic factor resulting in improved overall survival in estrogen receptor (ER) positive breast tumors [23]. A positive correlation between expression of RBM3 and RBM10 with the pro-apoptotic gene Bax in BC was also established in the study by Arribas et al. [31]. These studies collectively demonstrate that RBM3 is a promising prognostic marker that can be used to predict BC survival.

hnRNP family:

Heterogeneous nuclear ribonucleoprotein family of RBPs (hnRNP A1-U) play a role in various cellular processes including cell cycle and cell proliferation [32, 33]. Numerous studies have independently explored the expression of hnRNP family of proteins in BC. hnRNP AB is frequently upregulated in BC tissues and correlates with poor survival in patients and its knockdown halted cell proliferation and induced a G2/M cell cycle arrest [34]. Similarly, high expression levels of hnRNP A2/B1 and hnRNP C have been reported in BC cells in comparison to normal cells [35, 36]. hnRNP A1 is also reported to influence CD44 splicing in BC [37]. Given the prominent role of CD44 splicing in the maintenance of cancer stem cell traits and EMT process in BC, delineating the interaction between hnRNP A1 and CD44 would help better understand the EMT process as well as design better anti-cancer therapies in BC [38]. A study by Loh et al., demonstrated that knockdown of hnRNP A1 and hnRNP C decreased cell viability, proliferation and invasive abilities of MDA-MB-231, MCF-7 and T47D breast cancer cells respectively, thereby demonstrating them as candidate targets for BC treatment [37]. Finally, hnRNP U is also reportedly involved in breast cancer invasion and the EMT process [39].

TTP:

Tristetrapolin (TTP), belonging to the TPA-inducible sequence 11 (TIS11) family of RBPs, regulates mRNA stability by binding to specific sequences located in the 3′ UTR of the target mRNA. Recent studies have identified TTP as a tumor suppressor in various cancers including BC [40-42]. Decreased TTP expression correlated with higher tumor grade and poor prognosis of BC patients [42]. Similarly, a decreased expression of TTP has been reported in BC cell line (MDA-MB-231) compared to normal breast cell lines supporting its role in tumor cell invasion [43]. In another study, presence of TTP inhibited tumor cell proliferation and induced an S phase cell cycle arrest via the c-Jun/Wee 1 axis [44]. Additionally, TTP deficient BC cells and spheroids showed elevated levels of interleukin 16 (IL-16) expression and higher infiltration of monocytes/macrophages suggesting a promising role for TTP in immune response in BC [45].

Musashi-1:

Expression of Musashi family of RBPs – Musashi1 (MSI1), in BC cells and tissue specimens is associated with poor prognosis and survival, higher lymph node metastasis, and a highly metastatic and stem-cell like tumor cell population [46]. However, expression of Musashi-2 (MSI2) is reportedly dependent on the hormone status of BC. MSI2 expression was found to be high in ER-positive breast cancer compared to ER-negative BC tumors [47]. Involvement of MSI-2 in regulation of estrogen receptor 1 (ESR1) expression further substantiates the finding of higher MSI-2 expression observed in ER positive tumors [47]. In line with this, Li et al., demonstrated that downregulation of MSI2a, a canonical isoform of MSI2 protein, was associated with cancer progression and poor patient survival in TNBC tissues [48]. Overexpression of MSI2a in MDA-MB-231 and MDA-MB-468 cells resulted in inhibition of ERK1/2 activity and an increase in TP53INP1 mRNA (tumor protein p53 inducible nuclear protein 1) which correlated with diminished breast tumor growth and improved overall survival [49].

Lin28:

Overexpression of Lin28 is reported in BC as well as a positive association of Lin28 with hormone receptors i.e. HER2, estrogen and progesterone receptors have been documented [50]. Lin28 plays a major role in the regulation of cancer cell proliferation, metastasis, radioresistance, chemoresistance, and stemness in BC [51, 52]. Lin28, through repression of let-7a, induces EMT, enhances sphere formation and ALDH activity and is associated with increased invasion and migration in BC cells [53].

HuR:

Human antigen R (HuR) is a ubiquitously expressed protein that belongs to the ELAV family of RBPs and is involved in post-transcriptional gene regulation. HuR is located in the nucleus of a cell and binds to the ARE regions of target mRNAs in the nucleus. However, in response to stress signals in the cell, HuR-mRNA complex translocates from the nucleus to cytoplasm with the help of HuR nucleocytoplasmic shuttling domain (HNS) and stabilizes the target mRNAs by binding through three classic RRMs [6, 54, 55]. Since translocation of HuR is important for its activity, it was hypothesized that cytoplasmic HuR expression in cancer cells could serve as a prognostic marker [54, 56]. Several studies show that high cytoplasmic HuR expression is associated with large tumor size, angiogenesis and poor survival in multiple cancers including ovarian, oral, colorectal, breast and lung carcinoma [54, 56-58]. HuR stabilizes various mRNAs by binding to the AREs in their 5’ and 3’ UTRs and several ARE containing cancer-related transcripts, including mRNAs for proto-oncogenes, cytokines, growth factors, and invasion factors involved in regulation of tumor growth, metastasis, angiogenesis and resistance to therapy (Figure 2) [8]). Apart from stabilization of target mRNAs, HuR upregulates translation of several target mRNAs that are involved in cancer growth and metastasis [e.g. hypoxia inducible factor-1α (HIF1α), COX-2, Bcl-2, VEGF] and also represses some target mRNAs that encode for proteins involved in cell cycle (p27) [59-61]. Multiple reports have established high HuR levels in atypical ductal hyperplasia (ADH), ductal carcinoma in situ (DCIS) and ductal invasive carcinoma (DIC) compared to normal healthy tissues [57, 62]. Similarly, elevated HuR expression was seen in BC cell lines compared to the non-tumorigenic cell lines [62].

Figure 2.

HuR regulates cancer traits and markers of cancer progression by stabilizing multiple target mRNAs that encode proteins involved in angiogenesis, DNA damage response, oxidative stress, tumor growth, metastasis and invasion. Images were created with Biorender.com

Studies have also reported that increased cytoplasmic localization of HuR is associated with high tumor grade, ER/PR status and invasiveness in breast carcinomas [58, 63]. Elevated HuR levels correlated with an increase in cyclin E1, IL-8, TSP1 and ER, and a large tumor size [64-66]. The increased stability and overexpression of HuR mRNA in TNBC was also linked to TTP deficiency in BC cells [67, 68]. Cytoplasmic expression of HuR in breast tumor tissues is positively associated with increased expression of COX-2, an important carcinogenesis related factor involved in the synthesis of prostaglandins [69]. GATA3, a transacting T-cell-specific transcription factor, co-expresses with ER alpha in BC cells. Studies have reported that HuR stabilizes GATA3 mRNA in a manner similar to ER mRNA, and silencing HuR leads to a decrease in GATA3 mRNA and protein levels, further inhibiting cell growth in ER positive BC cells [70]. HuR expression correlates with distinct subsets of mRNAs and differentially regulates various target genes such as CD9 and Calmodulin 2 (CALM2) in ER+ and ER− BC [71]. Additionally, HuR post transcriptionally regulates the expression of tumor suppressor p21WAF1 and the non-canonical and non-transforming Wnt protein, Wnt-5a [72, 73].

HuR plays an important role in tumor angiogenesis and cell adhesion by regulating the expression of multiple genes including VEGF-A, HIF1α, and thrombospondin 1 (TSP1). HuR binds to and stabilizes VEGF mRNA resulting in increased VEGF protein expression [74]. Interestingly in orthotopic mouse models injected with TNBC cells, HuR overexpression leads to an antiangiogenic effect and tumor growth inhibition via increasing TSP1 expression and decreasing VEGF [75]. Downregulation of HuR in BC cells in vitro and in vivo reduced the expression of HIF-1alpha, decreased angiogenesis and led to tumor growth inhibition [76]. HuR stabilizes TSP1 mRNA in many cancers including breast carcinoma but in MCT-1 overexpressing aggressive BC cells a decrease in the HuR-TSP1 mRNA association leads to a reduction in TSP1 protein expression [74]. The homeobox protein Hox-A5 (HOXA5) involved in angiogenesis, apoptosis and growth suppression is expressed at low levels in 60% of the primary breast carcinomas. HuR, along with miR-130, induces HOXA5 in response to retinoic acid (RA) treatment and plays an important role in RA-induced cell death [77]. Several proteins that regulate cell proliferation, migration and metastasis in breast cancer such as platelet-derived growth factor (PDGF)-C and matrix metalloproteinase-9 (MMP-9) are also positively regulated by HuR [78, 79]. IL-8, produced in response to inflammatory cytokines such as IL-1β, is associated with angiogenesis, metastasis and malignant phenotype not only in BC but in several other tumors as well. HuR binds to the proximal 3′-UTR of IL-8 mRNA and increases IL-8 levels in IL-1β stimulated BC cells in a time dependent manner [80]. The proto-oncogene c-fms, is overexpressed in BC and is associated with metastasis and poor survival in patients. Interestingly, as the 3'UTR in c-fms is not AU-rich, HuR regulates its gene expression by binding to a 69-nt fragment of the 3′UTR of c-fms mRNA containing five ‘CUU’ motifs. HuR silencing leads to a decrease in c-fms RNA expression whereas overexpression of HuR increases c-fms levels [81].

Since chemoresistance is a major limiting factor in cancer treatment, several studies have focused on ascertaining the role of HuR in governing the response to chemotherapeutic drugs. In MCF-7 BC cells, doxorubicin treatment led to shuttling of HuR from the nucleus to cytoplasm in its phosphorylated form. In addition, results from the same study, using MCF-7/doxR model identified downregulation of HuR contributed to doxorubicin resistance [82]. Similar results were obtained in MCF-7 BC wherein treatment with tamoxifen increased cytoplasmic HuR levels. However, in this context, HuR upregulation was identified as a contributor to tamoxifen resistance [83]. Therefore, HuR localization and its impact on tumor promoting activities might differ based on the cell type and additional studies are warranted to delineate its role in chemotherapeutic drug response.

HuR also plays a role in cell cycle regulation and DDR in BC cells. Through its interactions with cyclin-dependent kinase inhibitor 1 (p21), cyclin-dependent kinase 1 (CDK1), cyclin-dependent kinase 7 (CDK7), and DNA repair protein RAD51 homolog 1 (RAD51) mRNA, it plays a role in tumorigenesis [84, 85]. Using the ribonomics approach, in a breast epithelial cell model, it was determined that HuR/AUF1 plays an important role in post-transcriptional regulation of cancer related genes involved in angiogenesis (TWEAK, TSP1), apoptosis (BAX, CASP2), cell cycle (CDK7, CDK1), translation regulation (eIF4EBP2), signal transduction (RAB2, STAT3) and DDR (RAD51) [66]. These studies indicate a significant role for HuR in modulating several cancer-related pathways and mediating resistance towards therapeutic agents.

3. RBPs in DNA damage response

Extrinsic factors such as exposure to radiation (UV and ionizing radiation), chemicals, and chemotherapeutic drugs causes DNA double stranded breaks (DSBs) and leads to DNA damage. Additionally, intrinsic factors such as replication errors and oxidative metabolism contribute to the generation of mismatches and single stranded breaks (SSBs) and DSBs in the cells [86, 87]. The DDR is an inherent protective mechanism occurring in the cells to rescue its genomic errors. Dysregulation in the DDR also correlates with multiple diseases including cancer and various DDR inhibitors are currently being explored as a part of cancer therapy [88]. Based on the type of DNA damage, cells go through DDR via one of the two main pathways involved in the DSB repair: Non-homologous end joining (NHEJ) and Homologous recombination (HR) (Figure 3). The NHEJ pathway is an error-prone repair pathway whereas the HR pathway is error free, with both the pathways operating together throughout the cell cycle. NHEJ is active mostly in the G0/G1 phases whereas HR is initiated during S and G2 phases of the cell cycle [89]. The early DDR relies on activation of sensor proteins like the Rad9-Rad1-Hus1 complex (9-1-1 complex) and Ku70/Ku80 heterodimer, which together recognize the DSB ends and recruit DNA-dependent protein kinase (DNA-PKcs). Subsequently, ataxia telangiectasia mutated (ATM), ATM and Rad3 related (ATR) and tumor protein p53 binding protein 1 (53BP1) activate CHK1 and CHK2 protein kinases which diminishes CDK activity and weakens cell cycle progression giving the cells enough time for DNA repair [9, 86, 90].

Figure 3.

RBPs in DDR: Schematic representation of RBPs involved in the DDR following induction of DNA damage. Initiation of DSB signaling involves RBPs such as WRAP53, FUS and CIRBP that regulate the choice of repair pathway: A) HR-directed pathway involves RBPs such as hnRNPUL1/2, GEMIN2, SFPQ and FUS B) RBPs involved in the NHEJ-directed pathway include PRPF19, RBMX, RBM14, SFPQ/NONO and FUS. Images were created with Biorender.com

Accumulating evidence supports the role of RBPs in the regulation of DDR genes. As shown in Figure 3, in response to DNA damage, RBPs such as CIRBP (Cold-inducible RNA-binding protein, hnRNP A18), FUS (fused in sarcoma, hnRNP P2) and WRAP53 (WD40-encoding RNA antisense to p53) are recruited to the damage site and act in early DSB response [91, 92]. These RBPs, referred to as DNA-damage response RNA-binding proteins (DDRBPs), play a role at the post-transcriptional level and are involved in early and/or late DNA repair response [86, 87]. The genomic stability following DNA damage is potently influenced by RBPs and noncoding RNAs (primarily microRNAs), which bind to mRNAs and regulate selective expression of DDR genes [93]. hnRNP, FET (FUS, EWS), DEAD-box helicase family of RBPs [94, 95] directly bind to sites of DNA damage and interact with DNA and repair proteins [9]. In the following section, we describe various RBPs that play a role in the double-stranded (ds) DDR and provide a detailed review of HuR and its implication in DDR.

CIRBP:

CIRBP, participates in multiple RNA processes such as mRNA maturation, translation and stabilization. Evidence supporting a role for CIRBP in DDR comes from its interaction with poly (ADP ribose) polymerase-1 (PARP-1), ATM kinase and MRN complex. Reduction in the levels of CIRBP lead to a diminished DNA repair capacity of cells, both by HR and NHEJ pathways [91, 92, 96]. Several studies have revealed CIRBP’s involvement in inflammatory diseases and various types of cancers including breast [97, 98]. Interestingly, CIRBP has been reported to play a role as both an oncogene and a tumor suppressor gene in various tumor types [97, 99]. CIRBP positively regulates HuR and co-regulates cyclin E1 levels leading to increased proliferation and decreased cancer cell apoptosis in BC cells [97]. Indacochea et al., showed that CIRBP contributes to the tumorigenic properties of MCF-7 BC cells by downregulating Cystatin C (CST3) levels. Further, a reduction in CIRBP levels in MCF-7 cells led to a decrease in proliferation rates [98]. Analysis of BC patient samples indicated that CIRBP up-regulation corresponded to poor disease-free survival.

WRAP53:

WRAP53 is another RBP that is recruited to the site of the DSB where it regulates the interaction between E3 ligase RNF8 and MDC1, causing an accumulation of RNF8 at the damage site. MDC1, binds to H2AX upon DNA damage, where it facilitates recruitment of the E3 ligase RNF8 to the site of DSBs [86, 100]. WRAP53 generates an antisense p53 transcript required for p53 induction upon DNA damage and encodes proteins that direct regulatory complexes to nuclear Cajal bodies and maintain genomic integrity. Cells overexpressing WRAP53 show rapid DNA repair with enhanced HR and NHEJ efficiencies and less H2AX foci formation resulting in cancer cell survival [101, 102]. WRAP53 protein is localized both in the cytoplasm and nucleus and BC patients with positive WRAP53 nuclear staining in tumor cells showed better survival than patients lacking nuclear WRAP53, making it a potential prognostic biomarker in cancer [103]. In head and neck cancer patients, the subcellular localization of WRAP53 in the nucleus correlates to increased survival and enhanced radiosensitivity [104].

hnRNP FUS/TLS:

Another RBP that plays a significant role in both HR and NHEJ repair pathways is the RNA binding protein fused in sarcoma/translated in liposarcoma (hnRNP FUS/TLS), that gets recruited to the site of DSBs downstream of PARP, independent of ATM kinase, and enhances histone H2AX phosphorylation followed by activation of DDR signaling proteins [105]. FUS deprived cells exhibit slower DDR, through both HR and NHEJ, and delayed early damage response compared to cells expressing FUS-WT [105-107]. Studies show that FUS interacts with HDAC1 and regulates neuronal DDR and DNA repair [107]. Like CIRBP, FUS also gets phosphorylated by both ATM and DNA-PKcs leading to its removal from the damaged chromatin [91, 108]. FUS binds both ss-and ds-DNA and plays a role in D-loop formation, an important step in HR during stand exchange, leading to DNA annealing [109, 110]. Although FUS has been shown to play a role in DNA homologous pairing, further research needs to be conducted to define its role in HR mediated repair. FUS/TLS overexpression has been reported in BC cells where it interacts with nuclear paraspeckle assembly transcript 1 (NEAT1) and plays a role in prevention of cancer cell apoptosis [111]. Overexpression of FUS/TLS has also been reported in non-small cell lung cancer (NSCLC) tissues and is associated with poor prognosis in patients. A negative correlation between FUS/TLS and E-cadherin expression was observed in these NSCLC tissue cohorts leading to cancer progression [112].

RMBX:

RNA binding motif protein, X chromosome (RBMX) or hnRNP G, binds to DSBs, guards it from the exonucleases, and stops DNA end resection. RBMX plays an important role in mRNA processing and splicing of DDR related genes and activates ATR during replication stress response. Due to its multiple roles in DDR, deficiency of RBMX leads to sensitization of cells to DNA damaging agents including ionizing radiation [113]. RBMX is expressed at high levels in hepatocellular carcinoma (HCC) patients and stabilizes a variety of proteins, including BLACAT1, which leads to chemoresistance, making it a therapeutic target in HCC cells [114]. On the contrary, in the study by Yan et al., low levels of RBMX in bladder cancer patients correlated with poor prognosis and advanced tumor stage making it a tumor suppressor. They demonstrated that RBMX inhibits aerobic glycolysis by interacting competitively with hnRNPA1 and reducing PKM2 levels in bladder cancer cells [115]. In BC, overexpression of RBMX positively correlated with the pro-apoptotic gene Bax, and inversely with the angiogenesis associated gene CD105 [31].

PRPF19/PSO4:

Other RBP’s such as Pre-mRNA-processing factor 19 (PRPF19/PSO4) and FUS help recruit NHEJ factors such as HDAC1, SETMAR (Metnase), and Ku to the DNA damage site. PRPF19/PSO4, is a ubiquitin ligase and an important regulator of various pathways including DDR, DNA interstrand crosslinks and replication stress response [116, 117]. Treatment with DNA damaging agents, increases expression of PRPF19/PSO4 and promotes NHEJ by recruiting DSB repair factor SETMAR to the damage site. SETMAR, a histone methyl transferase causes dimethylation of H3K36 (histone H3 lysine 36) and improves the binding of early DNA repair proteins such as NBS1 and Ku70 [118, 119]. Loss of PRPF19 in mammalian cells leads to reduced DNA repair and increased apoptosis and cell death following DNA damage. In response to DNA damage induced as well as replication stress, PRPF19 acts as a sensor of DNA damage and ubiquitylates RPA coated-single stranded (ss) DNA complex along with another ubiquitin ligase, RFWD3 [120, 121]. This ubiquitylation of RPA promotes HR and ATR-interacting protein (ATRIP) accumulation with activation of ATR leading to the repair of stalled replication forks. Inhibition of PRPF19 leads to diminished ATR phosphorylation and progression of replication forks on the damaged DNA [120, 122]. Moreover, mutants lacking PRPF19 ubiquitin ligase activity show diminished RPA-ssDNA ubiquitylation and hence impede the HR pathway [121]. The role of PRPF19 in cancer however has not been well studied. It has been shown that high expression of PRPF19 in HCC tissues correlates with advanced stage of cancer, poor survival prognosis and decreased immune cell infiltration leading to an immunosuppressive tumor microenvironment and resistance to immunotherapy [123]. However, no studies delineating the role of PRPF19 are available in BC.

RBM14:

RNA-binding protein 14 (RBM14) is an hnRNP family member that plays a crucial role in NHEJ by assisting in the recruitment of XRCC4-DNA ligase IV complex to the DNA damage site. RBM14 binds to other co-activators and regulates transcription and alternative splicing. It also plays a role in the phosphorylation of DNA-PKcs upon DNA damage and interacts with both DNA-PKcs and Ku proteins [124]. Recruitment of RBM14 to the damage site was found to be dependent on Ku, PARP1 and RNA polymerase II (RNAPII). Accumulation of RBM14 at the damage site was decreased upon inhibition and knockdown of PARP1 expression [90, 124]. A strong interaction between RBM14 and Ku proteins was observed which was required for the generation of RNA: DNA hybrids at the DSB site. Hence, RBM14 acts as a platform to connect Ku, DNA-PKcs and XRCC4-DNA ligase IV for DNA repair. In glioblastoma multiforme (GBM), knockdown of RBM14 decreased clonogenic survival and enhanced radiosensitivity in GBM spheres. Additionally, inhibition of RBM14 in the GBM spheres slows down the DSB repair process as indicated by persistent γ-H2AX foci compared to the untreated GBM spheres. While, it has been established that RBM14 plays a role in regulating cancer stem-cell like properties and NHEJ DSB repair in brain tumor cells [124], its significance in the context of BC is currently unknown.

TDP43:

TAR DNA-binding protein 43 (TDP43), a member of the hnRNP family, accumulates at the DSB and regulates DNA ligation by acting as a scaffold to recruit XRCC4-DNA ligase IV complex to repair DNA [107]. While TDP43, like FUS, is involved in neurodegenerative diseases including amyotrophic lateral sclerosis (ALS), its role in cancer progression is being gradually recognized. Overexpression of TDP43 in TNBC is associated with poor prognosis and promotes tumor metastasis [125]. Knockdown of TDP43 on the other hand causes a reduction in cell proliferation, induces cell apoptosis and G1 phase cell cycle arrest in TNBC cell lines [125]. In association with serine/arginine-rich splicing factor 3 (SRSF3), TDP43 regulates the splicing events involved in TNBC progression. Furthermore, in vivo studies showed that knockdown of TDP43 or SRSF3 in MDA-MB-231 resulted in reduced tumor growth and lung metastasis in mice [125]. Chen et al., examined the role of TDP43 in miRNA processing and showed that downregulation of TDP43 in SH-SY-5Y neuroblastoma cells leads to alterations in expression of various miRNAs including miR-423-3p and miR-500a-3p. The study also demonstrated that knockdown of TDP43 causes a reduction in lung cancer cell migration by regulating miR-423-3p expression [126].

SFPQ-NONO:

Splicing factor proline/glutamine rich (SFPQ) and Non-POU domain-containing octamer-binding protein (NONO) belong to the Drosophila behavior/human splicing (DBHS) family of proteins that are involved in RNA processing, transcriptional regulation and DNA repair [127, 128]. SFPQ-NONO heterodimer work together in DSB repair and knockdown of either protein leads to delayed DSB repair and increased radiosensitivity in the cells [128-130]. SFPQ-NONO accumulates and then releases from the damage site very rapidly in a fashion similar to CIRPB and RBMX. SFPQ-NONO co-ordinates with other repair proteins and leads to autophosphorylation of DNA-PKcs [131]. The SFPQ-NONO heterodimer associates with Ku proteins and promotes DNA end joining by acting as a scaffold in stabilizing the XRCC4-DNA ligase IV complex [131, 132]. A study by Lino et al., demonstrated NONO as a poor prognostic factor associated with shorter survival rate in breast cancer patients. In the same study, authors showed that knockdown of NONO suppressed breast cancer cell proliferation by post transcriptional modification of cell proliferation-related genes, SKP2 and E2F8 [133]. Besides playing a role in NHEJ, RBP SFPQ also regulates HR-mediated DNA repair, though without the binding partner NONO. SFPQ directly interacts with Rad51 and stimulates homologous pairing and strand-exchange reactions under low Rad51 concentrations. It directly interacts with Rad51 recombinase and Rad51D and promotes the formation of D-loop structures between recombining DNA molecules [134]. Downregulation of SPFQ affects the DNA damage repair and leads to increased chromosomal aberrations as it also functions in the maintenance of sister chromatid cohesion [134]. Cells deficient in Rad51D followed by inhibition of SFPQ showed decreased cell viability and reduced HR activity. Similarly, knockdown of SFPQ in Rad51D deficient mouse embryonic fibroblasts leads to sensitization to the DNA damaging agents such as cisplatin [134, 135].

hnRNP UL:

Heterogeneous nuclear ribonucleoprotein U-like (hnRNP UL) proteins 1 and 2 play important roles in DDR. Upon DNA damage, cells recruit hnRNP UL1 and −2 which then bind to the Nijmegen breakage syndrome (NBS1) component of the MRN complex (consisting of Mre11, Rad50 and Nbs1) and stimulate DNA end resection. hnRNP UL1 and −2 act downstream of MRN and CtIP and enhance ATR signaling and HR repair [136]. Inhibition of hnRNP UL1 and 2, using siRNA, enhanced the sensitivity of U2OS and HeLa cells to DNA damaging agents including ionizing radiation (IR) and camptothecin (CPT) suggesting its role in mediating resistance to therapy. Molecular studies showed that siRNA-mediated depletion of both hnRNP UL1 and 2 altered the phosphorylation of Chk1 while having no effect on the ATM target Chk2, indicating an involvement of ATR signaling following DNA damage [136]. Both hnRNP UL1 and 2 are released within minutes from the DSBs after induction of DNA damage and associate with the bromodomain (BD)-containing protein BRD7 linking it to chromatin events [136-138].

hnRNP C:

HR- directed repair is also promoted by hnRNP C, which is composed of two isoforms: hnRNP C1 and hnRNP C2. It binds to pre-mRNAs and is a core component of the 40S ribonucleoprotein promoting splicing, RNA processing and nuclear export [139]. hnRNP C is part of the chromatin bound PALB2 nucleoprotein complex composed of breast cancer suppressor proteins such as BRCA1/2. Downregulation of hnRNP C using siRNA causes significant decrease in the expression of repair proteins BRCA1/2 and Rad51, diminishes overall DNA repair efficiency by HR, and leads to persistent γH2AX foci in the cells post radiation [139]. The importance of hnRNP C protein in NHEJ remains unclear but it has been shown to be phosphorylated by DNA-PKcs and modulate RNA processing and transport [140]. It is reported to be overexpressed in multiple cancers including breast cancer. Interestingly knockdown of hnRNP C in breast cancer cells leads to an increase in endogenous dsRNA, triggering the expression of interferon stimulated genes (ISGs), including the type I interferon IFNβ which suppresses tumor growth and cell proliferation [36].

GEMIN2:

The Gem-associated protein 2 (GEMIN2) interacts with Rad51 and amplifies Rad51-DNA complex formation by preventing the removal of Rad51 from damaged DNA and promoting HR mediated DNA repair [141]. GEMIN2 is part of the survival of motor neuron (SMN) complex and is associated with the biosynthesis of spliceosomal small nuclear ribonucleoproteins (snRNPs) [142]. The SMN-GEMIN2 interaction not only plays a role in mRNA splicing but also in HR-mediated DNA repair by binding to Rad51 and enhancing the homologous pairing and strand exchange in the HR process [141, 143]. GEMIN2 was found to be essential for embryonic viability and cell proliferation in vivo. GEMIN2 deficient cells displayed reduced Rad51 foci formation but did not affect the Rad51 protein levels upon induction of DNA damage [143].

MSI1:

MSI1, is a stem cell marker that is frequently upregulated in many cancers and has been shown to play a role in DDR by stabilizing DNA-PKcs, which promotes NHEJ repair pathway in cells [144-146]. In glioblastoma cells, radiation treatment increased MSI1 protein levels, whereas knockdown of MSI1 led to decreased expression of DNA-PKcs, enhanced DNA damage and increased sensitivity to radiation [145]. Thus, targeting MSI1 in radioresistant glioblastoma tumors can prove to be an efficient therapeutic strategy [145]. Similarly, in a study by Wang et al., MSI1 expression was detected in 55% of the 20 breast cancer cell lines and 68% of the 140 primary breast tumors. Knockdown of MSI1 led to reduced expression of stem cell markers (CD133, Bmi1, SOX-2 and Oct4) and a decreased spheroid size in BC cells. Furthermore, a decrease in tumor growth was observed following inhibition of MSI1 in vivo thereby making it a selective therapeutic target for BC [46].

HuR:

In response to DNA damaging stress, HuR translocates from the nucleus to the cytoplasm and stabilizes various mRNAs that encode proteins involved in DDR (p53, cyclins A, B1 and D1, p21 and p27) [60, 147]. As shown in Figure 4, during DDR, ATM/ATR get phosphorylated and activate the checkpoint kinases, Chk1 which controls HuR translocation and Chk2 which regulates HuR-RNA interaction. Chk1, upon phosphorylation, inactivates dual specificity phosphatase Cdc25 which in turn inactivates Cdk1 which is responsible for HuR phosphorylation at S202. This decrease in HuR phosphorylation at S202 enhances cytoplasmic accumulation of HuR and its interaction with target mRNAs that encode DDR proteins. Additional kinases, such as PKC and p38 MAPK, also regulate HuR function in DDR by inactivating Cdk1 and enhancing cytoplasmic accumulation of HuR. Furthermore, in response to radiation p38 MAPK phosphorylates HuR at Thr118, leading to the cytoplasmic accumulation of HuR and stabilization of p21 mRNA causing a G1/S cell cycle arrest [148]. Activation of Chk2 through ATM/ATR leads to the phosphorylation of effector proteins BRCA1, p53, Cdc25 and Cdc25A leading to cell cycle arrest and apoptosis thereby allowing the damaged cells to undergo DNA repair [149-151]. Activated Chk2 phosphorylates HuR at S88, S100, T118 residues. HuR phosphorylation at S100 leads to dissociation of SIRT1 mRNA from HuR causing a decrease in SIRT1 mRNA and protein expression. SIRT1 is a pro-survival protein that promotes cell survival and enhances the DDR by causing deacetylation of NBS1 which recruits ATM to the damage site and initiates DNA repair [93, 152, 153]. Thus, HuR binds to SIRT1 mRNA and enhances SIRT1 expression which in turn deacetylates NBS1 leading to the activation of ATM and DNA repair. In order to ensure that the repair is not indefinite, a negative feedback loop is induced whereby ATM activates Chk2 which phosphorylates HuR at S100, thereby segregating the HuR-SIRT1 mRNA binding [152, 153] Studies from our laboratory demonstrated that HuR silencing in BC cells, in combination with radiation, led to a decrease in the expression of cell survival proteins such as survivin, COX-2, SIRT-1 and DNA repair (ATM, Ku80, and Rad51) (Figure 5 A, B) resulting in diminished cancer cell migration and enhanced radiosensitization. Additionally, siRNA-mediated knockdown of HuR in TNBC cells increased ROS levels upon radiation along with a concomitant decrease in the activity of the antioxidant thioredoxin reductase (TrxR) in BC cells [68]. This decrease in TrxR activity was associated with a decrease in the thioredoxin protein, further impairing the capacity of the cells to resist oxidative stress leading to cell death. In addition, inhibition of HuR led to increased expression and persistence of radiation-induced γH2AX foci, suggesting an inhibition of the DSB repair pathway. Thus, HuR silencing increases ROS activity and DNA damage resulting in radiosensitization of TNBC cells [68]. In a parallel study, we observed that HuR binds to AT-rich interactive domain 1A SWI-like (ARID1A) 3′-UTR and stabilizes the mRNA resulting in increased ARID1A expression in breast cancer cells [154]. ARID1A is a key player in cellular resistance to DNA damage by promoting NHEJ repair and its inhibition leads to a decrease in the expression of repair proteins resulting in enhanced radiosensitivity in cancer cells [155]. A significant radiosensitization of TNBC cells was observed upon silencing ARID1A (Figure 6 A, B, C). On the contrary, enhanced radioresistance was observed following ARID1A overexpression (Figure 6 D, E, F) indicating a role for ARID1A axis in resistance to therapy by enhancing DDR in cancer cells. Genetic inhibition of HuR in BC cells led to a greater accumulation of DSBs and DNA damage in combination with radiation, as evident by the formation of longer comet tails (Figure 6 G). Furthermore, treatment with empty vector had no effect on the levels of DSBs, whereas overexpression of ARID1A significantly reduced the DNA damage after radiation (Figure 6 G), thus demonstrating that ARID1A promotes DNA repair and plays an important role in HuR-dependent radioresistance [154]. HuR has also been shown to promote drug resistance in many cancers including pancreatic ductal adenocarcinoma (PDAC). Lal et al., demonstrated that HuR positively regulated Wee1 expression and provided PDAC cells with a protective mechanism against DNA damage [156]. HuR-Wee1 binding increased γ-H2AX levels, promoted DDR and enhanced chemoresistance in PDAC cells [156]. In a study by Chand et al., targeted silencing of HuR in combination with PARPi monotherapy showed a greater reduction in tumor growth in PDAC mouse xenograft models compared to PARPi monotherapy alone [157].

Figure 4.

Role of HuR in DDR. Regulation of HuR by ATM/ATR axis upon DNA damage: ATM/ATR activated Chk1/Chk2 regulates phosphorylation of HuR at different phospho sites, which controls its export to the cytoplasm and association with its target mRNAs. Images were created with Biorender.com

Figure 5.

A) HuR silencing in combination with radiation downregulates cell survival and DNA repair proteins in MDA-MB-231 cells. Cells treated with siScr and siHuR were irradiated, harvested 2h post radiation and analyzed by Western blot analysis. B) Fluorescent images of MDA-MB-231cells treated with siScr and siHuR at different time points post radiation demonstrate that inhibition of HuR leads to prolonged expression of radiation induced γH2AX foci. Quantification of γ-H2AX foci plotted as average number of foci per cell. Blue stain: DAPI (nuclei). Green stain: γ-H2AX foci. Image modified and reproduced from Mehta et al., 2016 [68]. Creative Commons CC BY 4.0 licence.

Figure 6.

Inhibition of ARID1A using siRNA radiosensitizes breast cancer cells A) MDA-MB-231 B) MDA-MB-468 C) SUM-159PT. D, E) Overexpression of ARID1A in MDA-MB-231 and SUM-159PT leads to radioresistance. ARID1A can rescue HuR silencing mediated radiation sensitivity in MDA-MB-231 cells as assessed by F) clonogenic assay and G) by neutral comet assay. Image modified and reproduced from Andrade et al., 2019 [154]. Creative Commons CC BY 4.0 licence.

Results from these numerous studies described above demonstrate that RBPs, including HuR, play a role in DDR. Thus, RBPs serve as molecular targets for cancer therapy, and incorporating combinatorial therapies will result in increased anticancer activities.

4. RBPs as therapeutic targets

RBPs, by regulating multiple mRNAs, modulate processes such as cell proliferation, invasion, angiogenesis, and immune response that are crucial for tumor growth, metastasis and drug resistance. Given their profound importance in cancer, RBPs have become attractive targets for therapy [3]. Various small molecule inhibitors, anti-sense oligonucleotides, peptides, siRNA, and clustered regularly interspaced short palindromic repeats (CRISPR)-based-therapies (Figure 7) are being explored to target RBP functions with some being clinically evaluated for their anticancer effects [3, 5]. Given the paucity of information on novel approaches targeting RBPs selective to BC, this section summarizes information on therapeutic targeting of RBPs available across multiple cancer types.

Figure 7.

Strategies used to target RBPs for therapy include small molecule inhibitors, antisense oligonucleotides, aptamers, siRNA, CRISPR/Cas9 and nanomedicine-based approaches. Images were created with Biorender.com

4.1. Small molecule inhibitors

Small molecule inhibitors targeted towards oncoproteins (PI3K/Akt/mTOR, EGFR, ALK, MET, RET, KRAS) have demonstrated huge success resulting in a paradigm shift in cancer treatment. This has resulted in many small molecule inhibitors receiving FDA approval for cancer therapy [158]. Similarly, significant advancements have been made in the identification of small molecule inhibitors targeting RBPs and these are discussed in the following sections. A review by Wu et al., compiled a timeline of the recent advancements in targeting RBPs using several small molecule inhibitors [159]. Various small molecule inhibitors, peptides and antisense oligonucleotides that have been tested against RBPs are described below and compiled in Table 2.

RBP	Therapeutic approach	Compounds	Mechanism of action	References
HuR	Small molecule inhibitors	MS-444, okicenone, dehydromutactin	Inhibiting HuR dimerization and disrupting HuR-target mRNA interaction	[181, 182, 314, 340]
		Tanshinone and its derivatives (6a, 6n), Dihydrotanshinone-I (DHTS-1)	Disrupting HuR-target mRNA interaction	[180, 187, 341]
		AZA-9		[342]
		Triptolide, MPT0B09 (a novel indoline-sulfonamide compound)	Inhibiting HuR shuttling from nucleus to cytoplasm	[293, 302]
		CMLD-2	Disrupting HuR-target mRNA interaction	[55, 183, 184]
	siRNA		Downregulation of HuR	[68, 234-236, 239, 240]
MSI	Small molecule inhibitors	(−)-gossypol Oleic acid	Inhibiting MSI1 -target RNA binding	[170, 172]
		Ro 80-2750 (Ro)	Disrupting MSI2 interacting with RNA	[174]
	Antisense oligonucleotide (ASO)	ASO	Targets MSI1 and inhibits its expression	[211]
	siRNA		Downregulation of MSI	[219]
eIF4E	Small molecule inhibitors	4Ei-1	4Ei1 gets converted to 7Bn-GMP and antagonizes eIF4E cap binding and inhibits translation initiation in cancer cells	[166]
		Ribavirin	Guanosine analogue binds to eIF4E at 7-methyl guanosine mRNA cap site and inhibits eIF4E mediated translation of oncogenes	[162]
		MnK inhibitors	Suppresses eIF4E phosphorylation and cell proliferation and induces apoptosis in vitro and in vivo	[169]
	Antisense oligonucleotide (ASO)	ISIS 183750	Inhibits production of eIF4E protein and inhibits cell proliferation in combination with irinotecan	[210]
		LY2275796	Blocks expression of eIF4E	[343]
		ASO	Causes destruction of eIF4E mRNA and represses eIF4E regulated proteins like VEGF, cyclin D1, survivin, c-myc and induces apoptosis	[208]
	siRNA/shRNA		Inhibits eIF4E expression and decreases cancer cell growth in combination with Rapamycin, enhances chemo-sensitivity to cisplatin	[223, 344]
LIN28	Small molecule inhibitors	Compound 1632	inhibit interaction of Lin-28 with let-7	[174][178]
		TPEN	destabilizes zinc-knuckle domain (ZKD) of LIN28 and elevates levels of let-7	[179]
		LI71	binds and modulates the cold shock domain (CSD) and inhibits LIN28 mediated oligouridylation	[179]
		SB1301, KCB3602, KCB3613, SBZW0065, 6-hydroxy-DL-DOPA	targets LIN28-let7 interaction	[184, 345-347]

eIF4E:

Elevated levels of Eukaryotic translation initiation factor 4E (eIF4E) are associated with tumor progression and aggressiveness in various cancer types [160]. Since overexpression of eIF4E is reported across multiple cancer types, various inhibition strategies are currently being pursued to target eIF4E in cancers [161]. Ribavirin, a guanosine ribonucleoside analogue (1-β-D-ribofuranosyl-1,2,4-triazole-3-carboxamide) is an antiviral that interacts with the 5’cap of eIF4E mRNA and competes with eIF4E:m7G mRNA cap binding. This disrupts the functioning of eIF4E-regulated oncogenes by blocking its transport and translation in vitro and in vivo [162]. Ribavarin has been reported to inhibit tumor growth in breast, prostate, glioma, ovarian and leukemia cancer cell lines [163, 164] and has demonstrated potent clinical activity in a clinical trial for treating acute myeloid leukemia (AML) (NCT00559091). However, no conclusive results were obtained when Ribavarin was tested in metastatic BC patients expressing high levels of eIF4E (NCT01056757). Although Ribavarin demonstrated antitumor activity in preclinical studies and in limited clinical study, its use has been limited due to side effects. Hence, current research aims at assessing the efficacy of next generation Ribavarin analogs in cancer. Wambecke et al., using a synthetic Ribavarin analogue (SRO-91) demonstrated tumor cell cytotoxicity and inhibition of cell migration in ovarian cancer cells [165].

Phosphoramidate, N-7 benzyl guanosine monophosphate tryptamine phosphoramidate prodrug (4Ei-1) is another small molecule inhibitor of cap-dependent translation of eIF4E. Treatment of breast and lung cancer cells with 4Ei-1 sensitized the cells to gemcitabine therapy and decreased their colony forming abilities [166]. A similar observation was made in mesothelioma cell lines wherein use of 4Ei-1 enhanced sensitivity to the chemotherapeutic drug pemetrexed [167]. Additional small molecule inhibitors [4EGI-1, 4E1RCat, and 4E2RCat] have been reported that disrupt the interaction of eIF4E with eIF4G and eIF4A and inhibit cap-dependent translation, and reverse chemoresistance to doxorubicin in vitro and in vivo [160, 168]. Lastly, targeting eIF4E phosphorylation with small molecule MnK (MAPK interacting kinases such as CGP57380, cercosporamide) inhibitors also represents a promising class of anti-cancer therapeutics and prevents cancer metastasis in lymphoma, colon and lung cancer models, both in vitro and in vivo [169].

MSI proteins:

Several small molecule inhibitors obtained from plant and microbial sources have been assayed for their activity against MSI group of proteins. Lan et al., upon screening for small molecules capable of inhibiting the binding of MSI1 to its RNA binding site, identified (−)-gossypol, a toxin from cottonseed having anti-tumor activities against colon cancer cells. Gossypol inhibited the Notch/Wnt signaling pathway and reduced cell proliferation and tumor growth both, in vitro and in vivo [170]. The efficacy of Gossypol has also been tested in clinical studies to treat patients with GBM (NCT00540722), small cell lung cancer (SCLC) (NCT00773955), and primary adrenocortical cancer (NCT00848016). Similarly, combination therapies using gossypol in combination with other chemotherapeutics and radiation therapies (NCT00390403) have also been assessed in GBM patients.

Another MSI1 inhibitor, Aza-9, is a secondary metabolite identified from the fungi Aspergillus nidulans that directly binds to MSI1 and exhibits anti-cancer activity [171]. Clingman et al., discovered that oleic acid interacts with RRM1 of MSI1 and changes the shape of the protein such that it prevents RNA binding and deregulates the expression of the MSI1 target genes [172]. Luteolin, a polyphenolic compound present in fruits and vegetables, was identified to play a role in decreasing the expression of MSI1 target genes namely PDGFRα, IGF-IR, EGFR, CCND1 and CDK. Luteolin inhibited proliferation, colony formation, migration and invasion of glioblastoma cells [173]. Small molecule inhibitors are also currently available for MSI2. Ro 08-2750 (Ro) is a small molecule inhibitor that binds selectively to MSI2, leading to apoptosis in mouse and human myeloid leukemia cells [174]. Molecular docking studies identified Largazole, a compound isolated from marine cyanobacterium Symploca sp to be a promising MSI2 inhibitor. Use of Largazole suppressed downstream targets of MSI2 (mTOR pathway), decreased proliferation and induced apoptosis in NSCLC and chronic myeloid leukemia (CML) [175]. Another MSI2 inhibitor, Palmatine, was shown to exhibit anticancer activity against colon cancer cells in vitro and in vivo [176].

Lin28:

Small molecule inhibitors against LIN28A/B have been identified. Roos et al., showed N-methyl-N-[3-(3-methyl[1,2,4]triazolo[4,3-b]pyridazin-6-yl)phenyl]acetamide (1632) inhibited the interaction of Lin28 with let-7, and induced differentiation of embryonic stem cells. Use of 1632 compound decreased tumor cell proliferation and sphere formation in 22Rv1 and Huh7 cancer cell lines [177]. In another study, Chen et al., reported that 1632, when used at high doses, restored let-7 expression and suppressed LIN28-mediated programmed death ligand-1 (PD-L1) expression leading to reduced cell proliferation and tumor growth in vitro and in vivo [178]. High-throughput screening of 101,017 compounds using fluorescence polarization assay identified several LIN28 inhibitors, of which two have been investigated in detail including LI71 and TPEN. LI71 interacts with the N-terminal CSD of LIN28 and competes with let-7 binding whereas TPEN inhibits the zinc-knuckle domain (ZKD) of LIN28. These inhibitors induce conformational changes and impair LIN28-mediated oligouridylation of let-7 thereby restoring the expression of let-7 and inhibition of cancer cell survival [179].

HuR:

Numerous small molecule inhibitors (MS-444, CMLD-2, dehydromutactin, okicenone, dihydrotanshinone-I (DHTS)) have been identified to target HuR (Figure 8) [180]. Each of these HuR-targeted inhibitors has been shown to exhibit potent antitumor activity in preclinical studies. For example, MS-444, one of the first HuR inhibitors discovered in a high throughput screen using a fluorescence-based RNA binding assay, reportedly interferes with HuR dimerization and its binding to the target mRNAs thereby affecting HuR cytoplasmic localization [181]. Treatment of colorectal (CRC) cells with MS-444 attenuated COX-2 expression, increased apoptosis, and decreased the growth of CRC tumors in xenograft models [182]. Similarly, in malignant glioblastoma cells, MS-444 enhanced apoptosis, impaired angiogenesis and the invasion capacity of tumor cells [181].

Figure 8.

Strategies applied to target HuR for cancer therapeutics include using small interfering RNA (siRNA)-mediated inhibition of HuR, and small-molecule inhibitors that disrupt the stability and interaction of HuR with its target mRNA. Images were created with Biorender.com.

CMLD2, a coumarin-derived compound, is another HuR-targeted inhibitor that has been shown to exhibit anticancer activity against a broad-spectrum of human cancer cells including thyroid, CRC, lung and pancreatic cancer [55, 183, 184]. CMLD-2 exerts its anticancer activity by competitively binding to HuR and disrupting interaction with ARE-containing target mRNAs. Treatment of thyroid cancer cell lines with CMLD-2 led to an antitumor effect by downregulating mitotic arrest deficient-2 (MAD2) protein. Similarly, in HCT-116 colon cancer cell line, CMLD-2 significantly disrupted the interaction between HuR and ARE region of RBP MSI-1, a target of HuR. CMLD-2 treatment not only shortened the half- lives of MSI1, Bcl-2 and XIAP mRNA accompanied with a decrease in their protein levels, but also induced caspase-3, PARP cleavage and conversion of LC3I to LC3II suggesting that CMLD-2 destabilizes and decreases the expression of HuR target mRNAs and regulates autophagy and apoptotic cancer cell death [184, 185]. Results from our laboratory in lung cancer cells, showed CMLD-2 treatment led to a reduction in the expression of HuR targeted proteins such as Bcl-2, Cyclin E and Bcl-XL and induced G1 cell cycle arrest. Mitochondrial perturbation and apoptotic cell death was also seen upon treatment of lung tumor cells with CMLD-2 [183].

The FDA-approved inhibitor, pyrvinium pamoate, was shown to block HuR nucleo-cytoplasmic translocation through indirect activation of AMPK pathway and inhibition of CDK1 pathway in urothelial carcinoma of the bladder [186]. Treatment of glioma cells with 15, 16-dihydrotanshinone-I, a HuR inhibitor, led to reduced tumor cell survival and proliferation [187]. Initial investigation by Zhang et al., incorporated a combinational therapy regimen with KH3, a small molecule HuR inhibitor, and PD-1 antibody. Use of the combination treatment mitigated tumor growth resulting in extended survival rates in EMT6 orthotopic BC mouse models [188]. While the above-described small molecule inhibitors are widely used in various studies, other less familiar HuR inhibitors are also available and are currently being explored. The readers are informed that compiling an entire list of available HuR inhibitors is beyond the scope of this review and are directed to recent reviews for additional information on available HuR inhibitors and their mechanism of action [181][189].

4.2. Oligonucleotide based therapies

4.2.1. Aptamers

Aptamers are short DNA or RNA oligonucleotides generated using the systematic evolution of ligands by exponential enrichment (SELEX) technique. The ability of aptamers to form 3D structures and accurately bind to molecular targets has been exploited in the field of cancer diagnosis, imaging and therapy. The target moiety chosen for aptamers is generally an extracellular, surface molecule overexpressed in cells or tissues of interest [190]. Albeit few differences, aptamers largely work in a similar fashion to the conventional antibodies and are referred to as “synthetic” or “chemical” antibodies [191]. Unique attributes, such as their short size (20-80 nt), low immunogenic properties, ease of synthesis and modification make aptamers an attractive option for cancer therapy. Various modifications, such as incorporation of phosphorothioate and sugar linkages, and locked nucleic acid modification are incorporated into aptamer sequences to improve stability, resist nuclease degradation and aid in targeted delivery [191]. Aptamers can be conjugated to other functional moieties such as drugs, siRNA, peptides, nanomaterials, and fluorophore tags to specifically deliver therapeutic agents to cell/tissue of interest [191, 192]. Following Macugen, the first FDA approved aptamer for therapeutic use, various other aptamers are being explored in clinical trials and have been reviewed extensively by Shigdar et al [193]. Despite a large number of aptamers in clinical testing, there are few FDA approved aptamers for cancer therapy thereby limiting the application of RBP-targeted aptamers for cancer.

Mochizuki et al., developed two RNA aptamers (aptamer 1 and 2) towards eIF4E and identified Aptamer 1 to inhibit cap-dependent in vitro translation of eIF4E. However, aptamer 1 could not inhibit the interaction of eIF4E to eIF4E-BP1 [194]. Guo et al., in their study designed and characterized DNA aptamers capable of inhibiting eIF4E translation. Use of eIF4E aptamer reportedly reduced the growth of Hela and HEK 293 cells [195]. Another RBP that is overexpressed in various cancer cells and serves as a candidate for anticancer therapy is Nucleolin. AS1411, a 26 nt aptamer against nucleolin, led to growth inhibition of various cancer cell lines [196, 197]. In the context of BC, AS1411 inhibited the binding of nucleolin to Bcl2 mRNA and exhibited greater and selective inhibitory activity against MCF-7 BC cells compared to control MCF-10A cells [197]. Ghahremani et al., conjugated AS1411 to gold nanoclusters (Apt-GNC) and assessed the ability of the conjugates to enhance sensitivity to radiation therapy in 4T1 BC cells. Treatment of Apt-GNC to BC cells augmented the effect of radiation, supporting the role of Apt-GNC as a radiosensitizer [198]. Additional approaches conjugating AS1411 to nanoparticles, liposomes and drugs, have been evaluated for BC therapy and are extensively reviewed in Yazdian-Robati et al [196]. The efficacy of AS1411 has also been assayed in clinical trials and Rosenberg et al., observed a striking 84% tumor reduction in AS1411 treated patients with metastatic renal cell carcinoma. Mild to moderate adverse events were observed indicating lower toxicity in patients [199]. Given these positive outcomes of AS1411 in clinical trials, future studies centered on the use of AS1411 in combination with other conventional or targeted therapies are anticipated to benefit BC.

4.2.2. Antisense oligonucleotides (ASOs)

Antisense oligonucleotides (ASOs) are single stranded DNA (ssDNA) of 18-30 bp length that are highly specific and function via RNase H-mediated RNA degradation, mRNA splicing and repress translation of target genes via complementary binding to target mRNAs. Various modifications made to the chemical structure of ASOs, such as phosphodiester backbone, improve the stability and protein binding properties of ASOs [200, 201]. Significant advances have been made in the field of ASOs and they have been used extensively as disease modifying therapies in various diseases including neurodegenerative disorders [202, 203], inflammatory diseases [204], diabetes [205] and more importantly in the context of this review, cancer [206, 207]. Graff et al., developed four ASOs against eIF4E (ASO1-4) and tested their potential as anti-cancer therapeutics in BC, NSCLC, prostate and head and neck cancer cell lines. Among these sequences, ASO4 showed greater efficacy in reducing levels of key oncogenic proteins such as VEGF, Survivin, c-Myc, Cyclin D1, and Bcl-2. Transfection of eIF4E ASO4 further induced apoptosis in BC and NSCLC cells and reduced the tube forming capacity of human umbilical vascular endothelial cells (HUVEC). Additionally, intravenous administration of ASO4 in human BC xenograft mice models notably suppressed eIF4E expression further inhibiting tumor growth [208]. Another study from the same group showed that treatment of NSCLC cells with eIF4E-ASO4 reduced tumor cell proliferation and increased sensitivity to gemcitabine [209]. The efficacy of eIF4E ASOs has also been investigated in clinical trials. A phase I/II clinical trial evaluated the efficacy of a second generation eIF4E ASO - ISIS 183750 in combination with irinotecan in CRC patients. Combinatorial treatment with ISIS 183750 decreased eIF4E levels in peripheral blood and tumor tissues of CRC patients [208, 210]. Similar clinical studies encompassing ISIS 183750 with widely used chemotherapeutic drugs such as carboplatin, paclitaxel, docetaxel and prednisone have been attempted for NSCLC (NCT01234038) and castration resistant prostate cancer (NCT01234025), respectively. While the last two clinical trials haven’t provided results, the data summarized above from preclinical and clinical studies supports the propensity of eIF4E knockdown via aptamers or ASOs to decrease tumor growth and possibly sensitize cells to chemotherapy. A similar approach exploited for BC may provide new avenues for mitigating disease progression in BC. Interestingly, ASOs have been designed towards other key RBPs such as MSI1 and HuR. For instance, ASO specific inhibition of MSI1 resulted in inhibition of pancreatic tumor growth [211]. Similarly, use of ASO-based HuR knockdown reduced spinal HuR protein and proved efficacious in relieving neuropathic pain [212].

4.2.3. siRNA and shRNA

RNA interference (RNAi) is a powerful strategy aimed at silencing the expression of disease-causing genes using complementary small interfering RNA (siRNA) or short hairpin RNA (shRNA) [213]. These RNAi approaches, due to the selective and specific regulation of gene expression, present an attractive option for cancer therapy. Currently, several clinical trials take advantage of RNAi approach for cancer therapy and have been reviewed in detail in other publications [214, 215]. In many cases, RNAi approach is combined with conventional cancer therapy or novel targeted therapy such as immunotherapy, to obtain enhanced therapeutic outcomes. siRNA and shRNA-based inhibition of RBP expression has been used across various cancer types and a few of them have been described in brief in the following paragraphs [51, 157, 216-218].

siRNA mediated MSI1 knockdown in ovarian cancer cells decreased tumor growth by reducing proliferation, invasion and migration. Additionally, the decreased MSI1 expression via activation of ERK signaling pathways increased the levels of p-Bcl2 to reverse observed resistance to paclitaxel [219]. Likewise, shRNA-based intervention of MSI2 in NSCLC decreased TGF-β signaling, further retarding tumor invasion and metastasis [220]. While these studies individually assayed genetic inhibition of either of the two MSI proteins (MSI1/MSI2), a more recent report involving siRNA mediated double knockdown of MSI proteins – MSI1 and MSI2 in ovarian cancer cells showed reduced expression of cancer stem cell markers and sensitized cells to both chemotherapy and radiation therapy [221]. Similar results were obtained in the context of BC, wherein siRNA mediated MSI knockdown decreased stemness and radioresistance in TNBC cells. However, this study also had a pitfall in the context of reporting enhanced cell migration and invasion obtained as a result of MSI knockdown [222]. siRNA approaches have also been tested for other RBPs such as eIF4E and AUF1. Delivery of eIF4E siRNA reduced cyclin D1, Bcl2 and Bcl-xL protein expression and repressed tumor cell growth in a panel of BC cell lines [223]. Treatment of BC cells with AUF1 siRNA inhibited EMT, reduced stemness and sensitized the cells to cisplatin [22].

Studies describing the effect of siRNA/shRNA mediated HuR knockdown across various cancer types and their significant outcome on several cancer traits have been collated in our recent review [224]. Few studies have also been conducted in the framework of HuR targeting leading to sensitization of BC cells to further treatment in preclinical models. Studies conducted in our own laboratory have shown that siRNA mediated silencing of HuR in TNBC cells sensitized them to gamma radiation and enhanced radiation-induced DNA damage [68]. We have further shown that HuR-mediated ARID1A expression in TNBC cells increased radioresistance and promoted DNA repair, making HuR-ARID1A axis an important therapeutic target for TNBC [154]. On a similar note, Zhang et al., investigated the potential application of HuR knockdown strategy to enhance immunotherapy response in BC. HuR knockdown decreased the expression of PD-L1, a key molecule implicated in immunotherapy, and decreased tumor growth in BC [188].

Although the use of RNAi-based gene knockdown is proven to be an attractive strategy, few challenges in the form of siRNA instability, innate immune stimulation and identified off-target effects exist that prevent reaping the full benefits of this technology. Overall, designing a safe, stable, suitable and effective delivery system is imperative for the successful translation of RNAi therapeutics into the clinic. Recent advances in the targeted delivery of siRNA in vivo involves the use of ligands, glycosylated molecules, peptides, proteins and antibodies to further enhance their recognition and internalization by target tissues. In this context, the RBPs themselves have been proposed as carriers for siRNA packaging and targeted delivery. RBP recognizes an siRNA independent of charge, making it a favorable carrier with specific structure and drug ratio. Most widely used RBPs as siRNA carriers, include double-strand RNA binding domain, a truncated protein kinase R (PKR), p19 and U1A RBD [225]. Additionally, there is a growing interest in using nanomedicine based (nanocarrier) delivery system to deliver siRNA and shRNA.

4.3. Nanomedicine based targeting of RBPs

The term “nanomaterial” largely refers to particles in the 10-100 nm size range synthesized via top-down synthesis or bottom-up synthesis approaches [226]. Accumulating studies support the use of nanomaterials to deliver chemotherapeutic drugs, peptides, and imaging agents. Significant advancements have also been made in the design of nanomaterials that can effectively deliver siRNA to the tissue of interest. The advantages of nanocarrier based siRNA delivery include the ease of synthesis and surface modification, improved stability of encapsulated biological molecules, biocompatibility, lower cytotoxicity and reduced immunogenicity, precise targeted delivery to desired site of action with minimal off-target effects on normal cells. Additionally, the nanocarrier can penetrate tissues and achieve tumor-selective accumulation through the enhanced permeability and retention effect, as well as overcome limitations associated with the systemic administration of siRNA. Consequently, various types of nanomaterials such as polymeric nanoparticles, metallic nanoparticles, dendrimers, lipid-based delivery systems (liposomes, solid lipid nanoparticles, micelles) and exosomes are being used for siRNA delivery [214, 215]. In the section below, we discuss few nanocarrier-based delivery systems that have been evaluated by several research groups, including our own, for targeting and delivery of RBPs to cancer cells.

As described in earlier sections, the RBP eIF4E is an attractive target overexpressed across cancer types and its knockdown reportedly sensitizes cancer cells to chemotherapy. Teng et al., utilized a novel theranostics system containing up-conversion nanoparticle (UCNP) carrier system loaded with both eIF4E siRNA and Pt (IV) for sensitizing cancer cells to chemotherapy with platinum agents. UCNP + eIF4E siRNA + Pt (IV) complex showed greater anti-tumor activity towards Hep-2 cells. In vivo studies using BALB/C mice showed UCNP + eIF4E siRNA + Pt (IV) complex treatment produced greater tumor inhibitory properties compared to controls. Furthermore, in vivo bio-imaging results showed effective penetration of complexes into tumor tissue to produce the desired therapeutic outcome [227].

Liposomes comprising of a phospholipid bilayer with hydrophilic and/or aqueous inner compartment are increasingly recognized as vehicles for siRNA delivery. Various liposomal formulations have been exploited for clinical use due to their improved bioavailability, controlled biological characteristics, ability to carry multiple and large payloads and target specificity [228] [229]. Positive or neutral charged siRNA nano-complexes are effectively taken up by the cells and hence cationic lipids such as DOTAP (1,2-dioleoyl-3-trimethylammonium-propane), DOTMA ((N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethyl-ammonium methyl sulfate), neutral lipids i.e. DOPC (1,2-dioleoyl-sn-glycero-3- phosphatidylcholine) DOPE (dioleoylphosphatidylethanolamine) are widely used to form complexes with siRNA. Additionally, other types of liposomal systems namely solid lipid nanoparticles and lipoid nanoparticles have been found beneficial upon testing [230, 231]. Sureban et al., utilized a DOPC liposomal preparation to deliver MSI-1 siRNA to colon adenocarcinoma xenograft models and showed decreased tumor growth via decreased proliferation, induction of cellular apoptosis and a G2/M phase cell cycle arrest [232]. Use of MSI-1 siRNA in combination with radiation therapy led to increased cellular apoptosis when compared to radiation or siRNA alone [232]. Among lipid-based nanoparticles, DOTAP is the preferred choice for delivery of siRNA. When stabilized by small amounts of cholesterol, DOTAP has been used to form complexes with negatively charged siRNA Use of cholesterol reportedly increases stability of siRNA, improves cellular uptake and protects from RNase mediated degradation [233].

Adequate precautions need to be taken to deliver the siRNA to the target site and avoid its premature release into the system. Targeting ligands towards integrins or other cell surface receptors can be attached to the free ends of the nanoparticle to achieve a selective targeting of cancer cells overexpressing the molecule of interest. Utilizing this strategy, we conjugated DOTAP:Cholesterol-HuR siRNA nanoparticles with Folic acid (HuR-FNP) and transferrin (HuR-TFr NP) on the surface with the help of short PEG linkers and tested their efficacy in lung cancer models. HuR siRNA delivery using HuR-FNP and HuR-TFrNP resulted in targeted knockdown of HuR s with minimal toxicity to the normal cells. HuR-FNP induced apoptosis and led to decreased cell migration and invasion of lung cancer cells in vitro [234]. HuR-TFrNP treatment inhibited lung tumor growth both in vitro and in vivo. Another advantage of these tumor-targeted siRNA delivery strategies is that the siRNA encapsulated in the lipid bilayer is protected from degradation thereby increasing stability and efficacy [234, 235].

An additional method used to deliver HuR siRNA to cancer cells is through a DNA dendrimer derivatized with FA [236, 237]. Dendrimers are nanocarriers, synthesized as well-defined spherical structures, ranging from 1-10 nm in diameter. A variety of dendrimers have been synthesized based on the core structure, but the ones derived from polyamidoamine (PAMAM) with ethylenediamine core are widely used in cancer [237, 238]. This method, when applied to target HuR siRNA in ovarian tumor-bearing mice, suppressed tumor growth and ascites development [236]. Similarly, a FA-conjugated polyamidoamine dendrimer based nanocarrier used to co-deliver HuR siRNA and chemotherapeutic drug, cis-diamine platinum (CDDP), resulted in increased therapeutic efficacy and reduced cytotoxicity in lung cancer cells [239]. Dendrimer-antibody conjugates use antibodies to direct dendrimer associated therapeutic agents to antigen-bearing in vivo tumors. Studies have also demonstrated that CDDP encapsulated in dendrimer polymers shows increased efficacy and less toxicity [238]. Lipid based poly (ethylene glycol)-dioleoyl phosphatidyl ethanolamine (PEG-DOPE) modified G (4)-PAMAM nanocarriers were synthesized for drug-siRNA co-delivery and can prove beneficial in multi-drug resistant cancers. PAMAM served as a siRNA complexing moiety and DOPE creates hydrophobic cores to encapsulate small hydrophobic molecules, such as doxorubicin. Another recently developed approach to deliver siRNA is by loading siRNAs (siHuR) into extracellular vesicles (EVs). EVs are heterogenous membranous nanoparticles naturally produced by cells and regulate intercellular communication between cells by transferring biomolecules, including proteins and RNA. One of the most common methods used to load siRNA into EVs, apart from transfection, is electroporation [240, 241]. However, the difficulty with this approach is that it can cause precipitation and aggregation of the siRNA. Recent methods using hydrophobically modified siRNA, such as cholesterol conjugated HuR siRNA loaded into EVs, have been successful in silencing the target gene (HuR) in HEK293 cells up to 168 hrs [240]. While this study was performed in non-cancerous HEK293 cells, similar methods of incorporating EVs based siRNA delivery for targeting RBPs in cancer models may be beneficial.

5. Conclusions

Accumulating evidence supports RBPs as critical drivers of the tumorigenic process by modulating cell proliferation, apoptosis, invasion, drug response and metastasis. RBPs are also currently implicated in the DDR in cancer cells. The mechanism of DDR is pivotal in determining the cellular response to chemotherapy and radiation therapy. Following DNA damage, RBPs are recruited to the site where they post-transcriptionally regulate the function of genes involved in DDR. However, the molecular signaling mechanisms underlying sensing of DNA damage by RBPs which in turn leads to the regulation of DDR associated proteins is unknown in cancer and needs further investigation. Similarly, the contribution of DNA damage to alterations in the localization patterns of RBPs and binding to target proteins is poorly understood.

Given the pronounced role of RBPs in regulation of numerous cancer related genes, including the ones responsible for DDR, deems them attractive targets for designing anti-cancer therapeutics across multiple cancer types. Various inhibitory strategies, including small molecule inhibitors and oligonucleotide-based therapies (aptamers, ASO siRNAs, shRNAs, CRISPR-Cas9) have proven successful under in vitro and in vivo experimental settings and a few have reaped favorable outcomes in clinical trials. Despite this success, few setbacks such as non-specific binding of oligonucleotides leading to off target effects and their poor bio-distribution warrants the need for a suitable carrier moiety to safely deliver the payloads to specific disease sites with minimal cytotoxicity to normal cells. Emerging studies from our laboratory and others advocate the use of nanomedicine-based delivery systems for siRNA targeting of RBPs such as HuR and Musashi and have achieved significant outcomes. The findings, however, need to be further validated in a large preclinical cohort before application to the clinic. Favorable outcomes from these studies would help incorporate combination therapies encompassing nanomedicine based RBP targeting along with the conventional chemotherapy and radiation therapy to increase therapeutic efficacy and offer significant outcomes in cancer treatment.

List of Abbreviations

53BP1

Tumor Protein P53 Binding Protein 1

ADH

Atypical Ductal Hyperplasia

ALS

Amyotrophic Lateral Sclerosis

ARID1A

AT-Rich Interactive Domain 1A SWI-Like

ASOs

Antisense Oligonucleotide

ATM

Ataxia Telangiectasia Mutated

ATR

ATM and Rad3 Related

ATRIP

ATR-Interacting Protein

AU

Adenylate And Uridylate

AUF1

Au-Rich Element RNA-Binding Protein 1

BC

Breast Cancer

BRCA1

Breast Cancer Gene 1

CALM2

Calmodulin 2

CDK

Cyclin-Dependent Kinase

CIRBP

Cold-Inducible RNA-Binding Protein

CSD

Cold-Shock Domain

CST3

Cystatin C

DCIS

Ductal Carcinoma In Situ

DDR

DNA Damage Response

DDRBPs

DNA-Damage Response RNA-Binding Proteins

DIC

Ductal Invasive Carcinoma

DNA-PKcs

DNA-Dependent Protein Kinase

DRBPs

DNA- And RNA-Binding Proteins

ds

Double-Stranded

DNA-DSBs

DNA Double Stranded Breaks

eIF4E

Eukaryotic Translation Initiation Factor 4E

EMT

Epithelial Mesenchymal Transition

ER

Estrogen Receptor

ESR1

Estrogen Receptor 1

FUS

Fused In Sarcoma

FUS/TLS

Fused In Sarcoma/Translated In Liposarcoma

GEMIN2

Gem-Associated Protein 2

GBM

Glioblastoma multiforme

HCC

Hepatocellular Carcinoma

HIF1α

Hypoxia Inducible Factor-1A

hnRNP

Heterogeneous Nuclear Ribonucleoprotein

hnRNP UL

Heterogeneous Nuclear Ribonucleoprotein U-Like

HNS

HuR Nucleocytoplasmic Shuttling Domain

HOXA5

Homeobox Protein Hox-A5

HR

Homologous Recombination

HuR

Human Antigen R

IGF2BP1

Insulin-Like Growth Factor 2 Mrna-Binding Protein 1

IL-16

Interleukin 16

IL-8

Interleukin 8

IR

Ionizing Radiation

KH

K Homology Domain

miRNA

Micro RNAs

MMP

Matrix Metalloproteinase

MRPL13

Mitochondrial Ribosomal Protein L3

MSI1

Musashi-1

MSI2

Musashi-2

NHEJ

Non-Homologous End Joining

NONO

Non-Pou Domain-Containing Octamer-Binding Protein

NSCLC

Non-small cell lung cancer

PARP-1

Poly ADP Ribose Polymerase-1

PDAC

Pancreatic Ductal Adenocarcinoma

PDGF

Platelet-Derived Growth Factor

PD-L1

Programmed Death Ligand −1

PKC

Protein Kinase C

PRPF19/PSO4

Pre-mRNA-Processing Factor 19

RA

Retinoic Acid

RBDs

RNA-Binding Domains

RBM

RNA Binding Motif

RBMX

RNA Binding Motif Protein, X

RBPs

RNA Binding Proteins

RNAPII

RNA Polymerase II

RNP

Ribonucleoprotein Complexes

ROS

Reactive Oxygen Species

RRM

RNA Recognition Motif

SFPQ

Splicing Factor Proline/Glutamine Rich

shRNA

Short hairpin RNA

siRNA

Small Interfering RNA

ssDNA

Single stranded DNA

TCGA

The Cancer Genome Atlas

TDP43

Tar DNA-Binding Protein 43

TGF-β

Transforming Growth Factor B

TIS11

TPA-Inducible Sequence 11

TNBC

Triple Negative Breast Cancer

TP53INP1

Tumor Protein P53 Inducible Nuclear Protein 1

TrxR

Thioredoxin Reductase

TSP1

Thrombospondin 1

TTP

Tristetrapolin

VEGF

Vascular Endothelial Growth Factor

WRAP53

Wd40-Encoding RNA Antisense To P53