Aberrant regulation of CXCR4 in cancer via deviant microRNA-targeted interactions

ABSTRACT

CXCR4 is involved in many facets of cancer, including being a major player in establishing metastasis. This is in part due to the deregulation of CXCR4, which can be attributed to many genetic and epigenetic mechanisms, including aberrant microRNA–CXCR4 interaction. MicroRNAs (miRNAs) are a type of small non-coding RNA that primarily targets the 3’ UTR of mRNA transcripts, which in turn suppresses mRNA and subsequent protein expression. In this review, we reported and characterized the many aberrant miRNA–CXCR4 interactions that occur throughout human cancers. In particular, we reported known target sequences located on the 3’ UTR of CXCR4 transcripts that tumour suppressor miRNAs bind and therefore regulate expression by. From these aberrant interactions, we also documented affected downstream genes/pathways and whether a particular tumour suppressor miRNA was reported as a prognostic marker in its respected cancer type. In addition, a limited number of cancer-causing miRNAs coined ‘oncomirs’ were reported and described in relation to CXCR4 regulation. Moreover, the mechanisms underlying both tumour suppressor and oncomir deregulations concerning CXCR4 expression were also explored. Furthermore, the miR-146a-CXCR4 axis was delineated in oncoviral infected endothelial cells in the context of virus-causing cancers. Lastly, miRNA-driven therapies and CXCR4 antagonist drugs were discussed as potential future treatment options in reported cancers pertaining to deregulated miRNA–CXCR4 interactions.

KEYWORDS:

• MicroRNA
• oncomir
• CXCR4
• tumour suppressor
• oncoviruses
• microRNA mimics

Introduction

Chemokines are immunomodulatory cytokines (8–10 kDa peptides) involved in directional migration of immune cells and play important role in preserving the homoeostasis of the immune system (adaptive and innate immune system) [1-3]. Chemokines are classified based on the position of first two cysteine residues into four families (two major types: CC and CXC and two minor types: C and CX3C chemokines) [4,5]. Normal immune function requires a balance between the differential and stably regulated expression of chemokines and its receptors. Imbalanced and aberrant expression of chemokines and chemokine receptors results in various immunological disorders, especially autoimmune diseases, and cancer [6,7]. In addition, chemokines are also utilized by tumour cells to actively progress and evade immunosurveillance. Chemokines control multistage carcinogenic process by affecting survival, proliferation, and metastatic spread of malignant cells as well as indirect modulation of neovascularization or angiogenesis [8,9] or in recruitment of immunosuppressive cells that cause escape from immunosurveillance [10–13].

CXCR4, also known as Fusin/LESTR and cluster of differentiation-184 (CD-184) [14,15] is one of the best studied chemokine receptors because of its role as co-receptor for HIV entry [16] and mediator of metastasis in a variety of cancers [17]. Though CXCR4 expression is low or absent in many normal tissues, it is the most widely expressed chemokine receptor in almost 23 human cancers, including melanoma, ovary, breast, prostate, and colorectal cancer (CRC) [1,3,18]. CXCR4 is a 352 amino acid chemokine receptor and functions as rhodopsin-like G protein-coupled receptor (GPCR) [5,19]. CXCR4 can exist as monomer, dimer, oligomer, or nanoclusters in the plasma membrane [20]. It has N-terminus outside the cell and C-terminus in the cytoplasm. Moreover, there are three extracellular and three intracellular loops with seven transmembrane helices [21]. While chemokine receptors can bind multiple ligands, CXCR4, selectively binds CXC chemokine stromal cell factor 1 (SDF-1) also referred to as CXCL12 [19,22]. However, it can also bind to cytokine MIF (macrophage migration inhibitory factor) as well as ubiquitin [23,24]. CXCR4-CXCL12 is the best identified axis involved in homing and retention of haematopoietic stem cells in the bone marrow [25] and involved in tumour growth, invasion, angiogenesis, and metastasis in various cancer types [26–28]. Binding of CXCR4 to CXCL12 activates multiple G protein-dependent (AKT, PI3K, mTOR and EGFR) as well as G protein-independent signalling pathways (JAK/STAT, p53 MAPK and ERK) [29–31], resulting in downstream cellular phenotypic effects that are pertinent to several essential biological processes including haematopoiesis, gestational development, vascularization, and organogenesis [32–37]. Moreover, CXCR4 deregulated overexpression frequently associates with promoting cell migration [38–45], proliferation [31,43,46–48], invasion [43,49–53], inflammation [45], angiogenesis [44], and metastasis [3,42,54–57] in numerous cancer types. CXCR4 expression is also altered in human oncovirus-infected progenitor blood cells. For example, CXCR4 overexpression is found in HIV+ patients and functions as a coreceptor for HIV and therefore promotes viral integration in haematopoietic stem cells (HSC), which serve as reservoirs of latent HIV [58]. CXCR4 is regulated by several different biological mechanisms including transcription factor binding, G protein signalling, and various epigenetic processes. In the context of T-cell-tropic HIV-1 entry, a bimodal pattern of CXCR4 expression is controlled through promoter binding of basal transcription factors Nuclear Respiratory Factor-1 (NRF-1) as a positive regulator and Ying Yang (YY1) as a negative regulator [59,60]. In addition, intracellular second messengers also regulate transcriptional control of CXCR4. For instance, second messengers like cyclic AMP [61], calcium (Ca²⁺) [59], transmembrane proteins CD3 and CD28 [59], and several cytokine interleukins promote CXCR4 expression [59, 62–64], while cytokines related to inflammation IL-1β, INF-γ, and TNF-α reduce CXCR4 expression [63–66]. Interestingly, G protein signalling can be initiated in the absence of ligand-bound induced CXCR4 activation. Alternatively, it has been shown that CXCR4 can form homodimers in T-cells and potentially illicit a downstream signal, which may have implications associated with HIV-1 entry [67].

Aberrant epigenetic modifications also affect CXCR4 expression in cancer. For example, CXCR4 promoter hyper/hypo methylation deregulates CXCR4 expression observed in several cancers and contributes to metastasis. For instance, prior studies have linked increased CXCR4 promoter 5-methylcytosine (5mC) to decreased CXCR4 protein expression in primary breast cancer [54], melanoma cells [68], and primary cervical cancer [69]. On the other hand, decreased promoter methylation and CXCR4 overexpression is an individual prognostic factor associated with a poorer outcome in patients with glioblastoma multiforme [70].

First identified in C. elegans in the early 1990s [71], microRNAs (miRNAs) are small oligonucleotides that function as an epigenetic regulator that plays a vital role in post-transcriptional gene regulation as it pertains to essential biological processes, but are commonly deregulated in cancer, which can lead to tumourigenesis. Extraordinarily, miRNAs are a type of small non-coding RNA ranging between only 21–22 nucleotides in length by which they negatively regulate genes through imperfect complementary binding of the 3'UTR sequence on mRNA transcripts [72]. It is well known that miRNAs primarily play an instructive role in cell development and differentiation. Quintessentially, ectopic overexpression of miRNAs miR-9 and miR-124 in human fibroblasts leads to the activation of neurogenic related pathways that drastically change the cellular morphology of fibroblasts, which are then ultimately converted into neuronal cells that exhibit neuronal functions and may therefore have future therapeutic implications in treating neurological diseases [73]. In cancer, many miRNAs are deregulated, therefore compromising their ability to target and dismantle oncogenic mRNA transcripts like CXCR4. In addition, some miRNAs can enhance oncogenic expression and are typically upregulated in cancer.

There are several reviews in the past focusing on clinical role of chemokines in cancer [74], regulation of CXCR4 signalling [30] or role of CXCR4 signalling in cell proliferation, invasion, and metastasis of cancer [29,75–77], and mechanism of miRNAs action in cancers [78,79]. However, none of them are focused on miRNA interactions with CXCR4, particularly in cancers. This is the first review, where we summarized the current knowledge pertaining to deregulated miRNA-CXCR4 targeted interaction in the context of human cancers and explored the utility of miRNAs as a potential therapeutic drug in targeting CXCR4 in cancer.

Characterization of tumour suppressor miRNAs and their clinical significance in deregulating CXCR4 in cancer

In cancer, tumour suppressor miRNA loss of function or repression can disable their ability to degrade oncogenic mRNA through 3'UTR-miRNA targeted interaction and therefore perturbs their ability to regulate genes involved in molecular processes like cell proliferation, differentiation, and apoptosis. For example, numerous studies have reported that decreased expression of tumour suppressor miRNA-let-7 leads to unregulated protein overexpression of its target oncogene RAS and therefore contributes to the tumour phenotype observed in lung cancer [80,81].

Given CXCR4's established reputation as an oncogene augmented with metastatic capabilities, it comes to no surprise that the majority of aberrantly expressed miRNAs targeting CXCR4 are downregulated, enabling overexpression of CXCR4. Thus, currently, the majority of the reported miRNAs that interact with CXCR4 in cancer function as tumour suppressors in their respective cell types. A total of 24 tumour suppressor miRNAs targeting CXCR4 are downregulated in various cancers ranging from solid tumours like colon cancer: miR-126 [82–85], miR-133b [86], miR-193a-5p [87], miR-622 [88] and breast cancer: miR-146a [89], miR-302a [90], miR-381[91], miR-494 [92] to blood cancers such as leukaemia of either myeloid: miR-146a [93–95], miR-335 [93], miR-622 [93], miR-9 [96] or lymphoid: miR-139 [97] of origin (Table 1).

Table 1. Deregulated tumour suppressor (n = 24) or oncogenic (n = 3) miRNAs identified as CXCR4 targets in human cancers with corresponding downstream target genes and pathways.

MicroRNA	Type of cancer	MicroRNA function	3'UTR of CXCR4 & miRNA binding	Binding targets and downstream genes/pathways	Prognostic marker	Patient survival	Reference
miR‐126	Colorectal cancer	Tumour suppressor	5’-GGTAGGA-3’ 3’-CCAUGCU-5’	CXCR4, AKT & ERK1/2, IRS-1, and RhoA signaling	Yes	Available	[82–85]
miR-622	Colorectal cancer	Tumour suppressor	5’-AGACUG-3’ 3’-UCUGAC-5’	CXCR4-VEGFA signaling	No	Not done	[88]
miR-133b	Colorectal cancer	Tumour suppressor	5’-GACCAA-3’ 3’-CUGGUU-5’	CXCR4, VEGF, and MMP-9	Yes	Not done	[86]
miR-193a-5p	Colorectal cancer	Tumour suppressor	Not found	EMT, Vimentin CXCR4	No	Not done	[87]
miR-302a	Breast Cancer	Tumour suppressor	5’-GCACUUA-3’ 3’-CGUGAAU-5’	CXCR4, AKT & RAD 52	No	Not done	[90]
miR-494	Breast cancer	Tumour suppressor	5'AGTTTTCAGGAGTGGGTT-3’ 3'UCCAAAGGGCACAUACAA-5’	CXCR4, Wnt/β catenin signaling, LEF1, CD44, cyclin D1	No	Not done	[92]
miR-146a	Breast cancer	Tumour suppressor	5'TGGTTTTTCAGTTTTC-3’ 3'ACCUUAAGUCAAGAG-5’	CXCR4, TRAF6, EGFR, IRAK1, MTA2 IL-6, IL-8, VEGF, NF-κB pathway	No	Not done	[89]
miR-381	Breast Cancer	Tumour suppressor	5’-CUUGUAU-3’ 3’-GAACAUA-5’	CXCR4, EMT, MMP-2, MMP-9, Vimentin, N-cadherin, and E-cadherin	No	Not done	[91]
miR-9	Acute myeloid leukemia	Tumour suppressor	5’-ACCAAAG-3’ 3’-UGGUUUC-5’	CXCR4, Cyclin D1, MMP-2, MMP-9, Bcl-2 & Bax	No	Available	[96]
miR-146a	Acute myeloid leukemia	Tumour suppressor	5'TGGTTTTTCAGTTTTC-3’ 3'ACCUUAAGUCAAGAG-5’	CXCR4, HIF-1α or −2α, CXCR4	No	Not done	[93–95]
miR-622	Acute myeloid leukemia	Tumour suppressor	5'GUACUGGUCUCAGCAGAUUG3’ 3'CGAGGUUGGAGUCGUCUGAC5’ *	CXCR4 and uPAR	No	Not done	[93]
miR-335	Acute myeloid leukemia	Tumour suppressor	Not found	CXCR4 and uPAR	No	Not done	[93]
miR-206	Gastric Cancer	Tumour suppressor	5’-ACAUUCCA-3’ 3’-UGUAAGGU-5’	CXCR4, MAP-, PI3K-AKT, mTOR, TGF-β, HIF-1, and Hippo pathway	No	Available	[121]
miR-139	Gastric cancer	Tumour suppressor	5'CUUAAGACGUGAU UUUGCUGUAGA-3’ 3'GACCUCUGUGC–ACGUGACAUCU-5’	HER2 signaling, CD44, CXCR4	Yes	Not done	[106]
miR-155-5p	Gastric cancer	Tumour suppressor	Not found	SMAD1, STAT1, CAB39, CXCR4, CA9	No	Not Done	[98]
miR-122	Hepatocellular carcinoma	Tumour suppressor	Not found	NUSAP1, CXCR4, CXCR2, ITGAV	No	Not Done	[99]
miR-622	Hepatocellular carcinoma	Tumour suppressor	5'TTCAGATATAAAAGACTG-3’ 3'AGGUUGGAGUCGUCUGAC-5’	CXCR4, SDF1 (CXCL12) EZH2	Yes	Available	[108]
miR-494-3p	Prostate cancer	Tumour suppressor	5’-UGUUUC-3’ 3’-ACAAAG-5’	CXCR4, SDF-1 (CXCL12)	Yes	Not done	[110]
miR-338-3p	Prostate cancer	Tumour suppressor	5’-AUGCUGG-3’ 3’-UACGACC-5’ **	CXCR4, CXCR7, CXCL1	Yes	Available	[109]
miR-1	Thyroid carcinoma	Tumour suppressor	5’-ACAUUCCA-3’ 3’-UGUAAGGU-5’	CXCR4, Cyclin D2, SDF-1α	No	Not done	[100]
miR-126	Thyroid Cancer	Tumour suppressor	5’-GGTAGGA-3’ 3’-CCAUGCU-5’	CXCR4	No	Not done	[101]
miR-9	Oral squamous cell carcinoma	Tumour suppressor	5’-ACCAAAG-3’ 3’-UGGUUUC-5’	CXCR4, Wnt/β-catenin signaling, Bcl-2, c-myc, c-Jun, and p-c-Jun	No	Not done	[119]
miR-143	Oral squamous cell cancer	Tumour suppressor	Not found	CXCR4, K-Ras, MMP9, C-Myc, ADAMTS	Yes	Not done	[111]
miR-494	Pancreatic cancer	Tumour suppressor	5’-UGUUUC-3’ 3’-ACAAAG-5’ ** (miR494-3p)	ZEB1, CXCR4, HIF1A, SDC1, TWIST1	Yes	Available	[112]
miR-494-3p	Synovial sarcoma	Tumour suppressor	5’-UGUUUC-3’ 3’-ACAAAG-5’ **	CXCL12, CXCR4	No	Not done	[116]
miR-613	Osteosarcoma	Tumour suppressor	5’-ACAUUCC-3’ 3’-UGUAAGG-5’	SDF-1, CXCR4	No	Available	[102]
miR-301a	Esophageal cancer	Tumour suppressor	Not found	CXCR4, Wnt/β-catenin and NF-κB signaling	No	Not done	[122]
miR-302b	Esophageal cancer	Tumour suppressor	5’-GCACUU-3’ 3’-CGUGAA-5’ **	ERBB4, IRF2 & CXCR4	No	Not done	[103]
miR-139	T-cell acute lymphoblastic leukemia	Tumour suppressor	5’-CUGUAGA-3’ 3’-GACAUCU-5’	CXCR4	Yes	Available	[97]
miR-139	Laryngeal squamous cell carcinoma	Tumour suppressor	5’-CUGUAGA-3’ 3’-GACAUCU-5’	CXCR4	Yes	Not done	[107]
miR-9	Head and Neck Squamous cell carcinoma	Tumour suppressor	5’-CUUUGGU-3’ 3’-GAAACCA-5’	CXCR4, CXCL12	Yes	Not done	[113]
miR-9	Nasopharyngeal carcinoma	Tumour suppressor	5’-ACCAAAG-3’ 3’-UGGUUUC-5’	CXCR4, CXCL12, p38 MAPK pathway	Yes	Not done	[114]
miR140	Nasopharyngeal carcinoma	Tumour suppressor	5’-CCUGUGG-3’ 3’-GGACACC-5’	CXCR4, Caspase-3, Bcl-2, and PARP	No	Not done	[104]
miR-663	Glioblastoma	Tumour suppressor	5'TGGATCCCTGCC-3’ 3'GCCGCGGGGCGG5’	CXCR4, PIK3CD	Yes	Available	[115]
miR-1246	Lung cancer	Tumour suppressor	Not found	CXCR4, JAK/STAT/PI3K/AKT signaling pathways, E-cadherin, N-cadherin, Vimentin, ZEB1, and Snail	No	Not done	[105]
miR-125b	Colorectal cancer	OncomiR or Tumour promoter	Not found	CXCR4, Wnt/β-catenin signaling, APC gene, ZEB1, Snail, Vimentin, and E-cadherin	No	Not done	[129]
miR-301a	Osteosarcoma	OncomiR or Tumour promoter	Not found	CXCR4, Wnt/β-catenin and NF-kB signaling, Bcl-2, Bax, caspase-3, and caspase-9	No	Not done	[131]
miR-199a-5p	Melanoma	OncomiR or Tumour promoter	Not found	CXCR4, Myc, Tnfsf10, cd44, Cxcr4, Mmp3, Rb1, Tcf20, Hprt1, Actb1, Rpsa, Etv4, Fxyd5 and Cdh1	No	Not done	[130]
Let-7 f	Breast cancer	OncomiR or Tumour promoter	Not found	CXCL12	No	Not done	[127]
miR-410	Non-small cell lung cancer	OncomiR or Tumour promoter	5'TTATAT3’ 3'AAUAUA5’ *	CXCR4, Sox2, Wnt/β-catenin signaling	No	Not done	[128]

*miRDB, ** TargetScanHuman Release 7.2.

Indeed, many of the deregulated tumour suppressor miRNAs targeting CXCR4 in cancer hold prognostic value in terms of predicting patient outcome. These include miR-126, miR-133b in colorectal cancer [84,86]; miR-139 in gastric cancer, T-cell acute lymphoblastic leukaemia, and laryngeal squamous cell carcinoma [97,106,107]; miR-622 in hepatocellular carcinoma [108]; miR-494-3p, miR-338-3p in prostate cancer [109,110]; miR-143 in oral squamous cell cancer [111]; miR-494 in pancreatic cancer [112]; miR-9 in head and neck squamous cell and nasopharyngeal carcinoma [113,114]; and miR-663 in glioblastoma [115]. In particular, deregulated miR-494-CXCR4 interactions have been reported in a myriad of cancers. Examples include breast, prostate, pancreatic, and synovial cancers, in which tumour suppressor miR-494 is downregulated, leaving CXCR4 unchecked and overexpressed [92,110,112,116]. Although, it is important to note that like many other miRNAs, miR-494 can act as an oncogene in other forms of cancer. For example, in hepatocellular carcinoma and colorectal cancers, miR-494 is upregulated and functions as an oncogene by targeting and subsequently downregulating important tumour suppressors like PTEN, CDKN1B, PUMA and APC [117,118]. Other notable underexpressed miRNAs targeting CXCR4 include miR-9, which lacks tumour suppressor function in several forms of head/neck-related cancers including oral squamous cell and nasopharyngeal carcinomas, emphasizing the importance of the miR-9-CXCR4 axis as a potential therapeutic target in head and neck cancers [113,114,119].

Tumour suppressor microRNAs targeting CXCR4 are downregulated in cancer via epigenetic regulation

Little is known of the mechanism(s) contributing to tumour suppressor miRNA underexpression themselves, which can result in CXCR4 overexpression in many cancer forms. Currently, two general processes of miRNA underexpression in CXCR4-targeted regulation have been described. The first epigenetic mechanism reflects a fairly recent hypothesis of competitive endogenous RNAs (ceRNAs) that function as a microRNA ‘sponge’ through RNA–RNA mediated interaction leading to subsequent miRNA degradation and therefore inhibiting the miRNA-3'UTR interactivity of its oncogenic target [120]. For example, tumour suppressors miR-206 and miR-494 are neutralized by overexpressed circular RNAs circ_0056618 and circ-UBAP2 that function as ceRNAs in gastric and pancreatic cancers, respectively [112,121]. Therefore, the counter activity imposed by ceRNAs on miR-206 and miR-494 leads to CXCR4 upregulation and subsequent tumourigenesis. Similarly, lncRNA growth arrest-specific transcript 5 (GAS5) is overexpressed in oesophageal cancer cells and behaves like a ceRNA by sponging CXCR4 targeted miR-301a transcripts [122]. Interestingly, in the case of oesophageal cancer, Li et al. reported that CXCR4 itself functions as a tumour suppressor, where induced overexpression of CXCR4 promotes apoptosis. The authors demonstrated that miR-301a functions as a positive regulator of CXCR4. Therefore, similar to [miR-206 and miR-494], miR-301a also functions as a tumour suppressor, but instead as a positive regulator of CXCR4.

The second epigenetic mechanism in CXCR4 driven cancers describes a decrease of tumour suppressor miRNA function by the means of promoter hypermethylation coupled with histone modifications that leads to an inactive (silenced) heterochromatin conformational state of miRNAs. For instance, the promoter of tumour suppressor miR-9 is hypermethylated and silenced in nasopharyngeal carcinomas, while the promoter of miR-622 is hypermethylated and transcriptionally repressed by EZH2-induced post-translational histone modification H3K27 trimethylation in hepatocellular carcinoma [108,114]. In addition, miR-139 undergoes histone deacetylation, thus impeding its tumour-suppressor function in gastric cancer [106]. One could therefore hypothesize that at least some of the unknown mechanisms downregulating the remaining tumour suppressor miRNAs targeting CXCR4, consists of some form of RNA–RNA mediated interaction or downregulation by DNA hypermethylation or modifications of amino acid residues on histone tails, residing in the promoter region. For this reason, epigenetic-targeted therapies may apply to deregulated miRNA-CXCR4 related cancers.

CXCR4 is deregulated by overexpressed oncomirs in cancer

Overexpressed miRNAs in cancers are typically categorized as oncogenes, coined ‘oncomirs,’ which are less well described than tumour suppressor miRNAs. Oncomirs contribute to tumour development by either negatively regulating tumour suppressor genes through miRNA–3'UTR interaction and subsequent mRNA degradation or by positively regulating target oncogenes through mechanisms not completely understood. Interestingly, clusters of oncomirs can exist as a single polycistronic transcript within the human genome. For instance, the existence of the miR-17-92 miRNA cluster in the human genome has been well described in cancer and is the first to be identified in human developmental diseases [123]. The miR-17-92 cluster is comprised of 7 miRNAs: miR-17-5p, miR-17-3p, miR-18a, miR-19a, miR-19b-1, miR-20a, miR-92-1 that play multifaceted roles essential to normal embryonic and skeletal development [123]. Indeed, deletion of the cluster was found to be perinatal lethal in mice, emphasizing its importance in mammalian development. As an oncomir, overexpression of the miR-17-92 cluster is a common feature observed in lymphoma and lung cancer [123,124]. In addition, their molecular interplay and subsequent downregulation of notable tumour suppressor PTEN and antiapoptotic regulator BIM was shown to increase lymphocyte hyper-proliferation and autoimmune diseases in mice [125,126].

A mere total of 5 miRNAs were documented as oncomirs targeting CXCR4 in various cancer types (Table 1) [127–131]. In the case of colorectal cancer, osteosarcoma, melanoma cells, breast cancer, and non-small cell lung cancer: miR-125b, miR-301a, and miR-199a-5p, Let-7 f, miR-410, respectively, are upregulated and function as oncomirs by positively regulating CXCR4 and thereby promoting tumourigenesis and metastasis [127–131]. For example, one of the mechanisms contributing to miR-125b upregulation in colorectal cancer cells includes an established positive feedback loop leading to increased CXCR4 expression and ultimately triggering epithelial–mesenchymal transition (EMT) and cell invasion. In modest detail, activation of the (SDF-1/CXCL12)/CXCR4 axis induces miR-125b expression in colorectal cancer cells, which in turn enhances CXCR4 expression, thus completing the positive feedback loop [129]. In another example, non-coding RNA gene UCA1 is speculated to positively regulate the expression of miR-301a in some manner leading to CXCR4 upregulation as demonstrated in UCA1 knockdown studies in osteosarcoma cells. However, the mechanistic underpinnings describing oncomir driven upregulation of CXCR4 and other oncogenes in general are lacking and require further study. A recent novel mechanism reported by Place et al. describes miR-373 positively regulating the expression of its target genes through complementary binding of promoter sequences and thereby inducing transcription, thus expression [132]. Future studies analysing oncomir interactions in the promoter regions of CXCR4 are therefore warranted.

CXCR4 is targeted by miR-146a in oncovirus infected endothelial cells

In addition to tumour cells, the miRNA-CXCR4 axis is altered in oncoviral infected endothelial cells carrying, for example, the human herpesvirus 8 (HHV8) observed in Kaposis sarcoma or the Epstein–Barr virus (EBV) involved in lymphomas like Burkitt lymphoma and Hodgkin/non-Hodgkin's lymphomas [133,134]. Endothelial cells harbouring oncoviruses HHV8 and EBV produce viral proteins which directly bind and upregulate their miRNA target, which in turn downregulates CXCR4. For example, miR-146a is upregulated by viral protein LMP1 and functions like a tumour suppressor in EBV+ human endothelial cells by targeting CXCR4, which inhibits cell growth and promotes apoptosis [134]. Paradoxically, Wang et al. suggests that the biological effects of CXCR4 targeting by miR-146a results in protecting the oncovirus in latently infected EBV+ endothelial cells by decreasing arrest of the G0/G1 phase cell growth. Therefore, the LMP1-miR-146a-CXCR4 axis serves an important function in EBV+ endothelial cells by delicately shifting between regulating cell growth and evading anti-tumour immune responses.

In HHV8 latent infected endothelial cells, viral protein K13 induces miR-146a expression, which in turn targets and downregulates CXCR4. Punj et al. hypothesizes that CXCR4 downregulation may result in premature release of progenitor endothelial cells infected with HHV8 into circulation, considering increased CXCR4 surface expression is a marker of developing immature endothelial cells that are maintained in the bone marrow [133]. In summary, the miR-146a-CXCR4 axis plays a crucial role in oncoviral infected endothelial cells and may therefore serve as an important therapeutic target in the future.

MicroRNA ‘mimics’ as a therapeutic approach in targeting CXCR4 in cancer

Given that RNA transcripts vastly outnumber protein-coding genes, the potential use of small RNA drugs in a clinical setting remains largely untapped [111,135]. The future use of miRNA-targeted therapy may hold a unique advantage in terms of eliminating chemo-resistant cancer cells. Indeed, miRNAs are pleiotropic in nature, meaning several hundred or even thousands of gene transcripts may be targeted. This notion implies that entire deregulated biological pathways may be targeted in cancer [136]. However, the potential unwanted off-target biological effects remains a potential issue. Although, ectopic expression of miRNA ‘mimics’ are typically designed to conserve most of their endogenous counterpart miRNAs sequence with the intent to mitigate off-targeting.

Countless studies have demonstrated the potential therapeutic utility of overexpressing tumour suppressor miRNA ‘mimics’ to target and downregulate CXCR4 and thereby decreasing migration, invasion, and/or proliferation of cancer cells. For instance, ectopic overexpression of the entire miR-302-367 cluster suppressed the stemness and tumourigenic properties of glioblastoma cells, in part by targeted downregulating CXCR4 [137]. The same research team again utilized glioma cells to express the miR-302-367 cluster and determined that glioma cells expel the miRNA cluster via exosome secretion, which are then internalized by neighbouring glioma cells [138]. Indeed, their glioblastoma in vivo model demonstrated that brain tumour development was inhibited in rats treated with miR-302-367 compared to untreated controls. Therefore, miRNA ‘mimics’ remain a viable option in targeting CXCR4 and potentially a broad span of other oncogenes involved in cancer.

CXCR4 antagonists as a therapeutic approach to restore endogenous expression of CXCR4 targeted miRNAs in cancer

Administration of CXCR4 antagonist not only successfully inhibited CXCR4 at the protein level, but the reported downstream effects included restoration of miRNA tumour suppressor or anti-arthritic function. Previously, we have discussed the use of CXCR4-CXCL12 axis in the therapy of Gastrointestinal cancers both in pre-clinical and clinical trials and its limitations [139]. Multiple antagonists of CXCR4 (bicyclams and T22) and peptide analogues designed for the N-terminal of CXCL12 (TN14003, CTCE-9908 and ALX40-4C) are in use to target CXCR4 in multiple cancers [140–143]. AMD3100 (plerixafor or mozobil) is the most frequently used CXCR4 antagonist drug in clinical trials targeting the CXCR4-CXCL12 axis [144]. Immunotherapy targeting of the CXCR4-CXCL12 axis has provided encouraging results in ovarian cancer [145], multiple myeloma [146], acute myeloid leukaemia [147], and cervical cancer [148] as well as relapsed or refractory multiple myeloma [149].

Targeted therapies involving CXCR4-CXCL12 axis in combination with silencing or inhibiting expression of oncogenes by small interfering RNA (siRNA) and microRNA (miR) could be another approach for cancer therapy [150]. SiRNA-based drugs patisiran [151], givosiran [152], and lumasiran [153] are approved by FDA for phase 3 clinical trials. However, only a few miRNA-based therapeutics have been in phase 1 or 2 clinical trials, and most of them have been discontinued because of off-target effects resulting in adverse events/death [154,155].

CXCR4 antagonist drug TN14003 upregulated CXCR4 target miR-146a in leukaemic cells [156]. A similar effect was also seen in colorectal cancer cells upon antimicrobial peptide (AMP) treatment, where CXCR4 was downregulated and its tumour suppressor miRNA target, miR-663a was upregulated and therefore functionally revived [157]. Taken together, the prior miRNA-driven studies suggest a bright future for miRNA-mediated therapy as an additional treatment option in cancer. Future miRNA-driven therapies may therefore include overexpressing tumour suppressor miRNAs that target CXCR4 transcripts. Moreover, CXCR4 antagonist drugs not only inhibited CXCR4 protein function, but also alleviated previously stifled tumour suppressor miRNA function.

Conclusion

This is the first review where we have identified and characterized many deregulated miRNAs that target CXCR4 in various cancers, where most of these miRNAs serve as tumour suppressors in their respective cell types. In addition to tumour suppressor miRNAs, but to a much lesser extent, oncomirs targeting CXCR4 were also identified and their complex and not fully understood mechanisms described. Furthermore, the miR-146a-CXCR4 axis was explored in oncoviral infected endothelial cells. Recent evidence suggests that there are many undiscovered miRNAs, potentially thousands [158]. Therefore, it is highly likely that CXCR4 is regulated by many other miRNAs that would presumably succumb to deviant expression, which may then deregulate CXCR4 in cancer. For instance, using the microRNA prediction software tool ‘TargetScanHuman 7.2’ and searching for broadly conserved microRNAs that target CXCR4 in humans, we found 4 miRNAs, 3 of which were already identified as tumour suppressors in this review: miR-140, miR-1, and miR-9 (Figure 1) [159]. Intriguingly, miR-655 has not been investigated as a potential target of CXCR4 in any biological context and may be worth exploring, given its tumour suppressor role in many cancers [160–162]. Furthermore, identifying miR-CXCR4 complements may hold the key to unlocking novel therapies in cancer, which may include tumour suppressor miRNA ‘mimics’ and other miR-driven forms of treatment.

Figure 1. UCSC Genome Browser (hg19 build) depiction of 4 conserved putative microRNA binding sites within the 3’ UTR (gold bar) of CXCR4 (blue and black bars) as predicted by in silico analysis using TargetScanHuman 7.2. Site type prediction: Red miR = exact match to positions 2–8 of the mature miRNA, Purple miR = an exact match to positions 2–8 of the mature miRNA followed by an ‘A.’

Key highlights

• CXCR4 transcripts are targeted by tumour suppressor microRNAs and oncomirs in cancer.

• The CXCR4-miRNA axis plays an important role in oncoviral function within the cell.

• MiRNA ‘mimics’ and CXCR4 targeted antagonist therapeutics help in combating tumour progression.

Data availability

Figure 1 was created using the UCSC Genome Browser.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was supported by VA merit award 2I01BX000824 and National Cancer Institute RO3 awards CA249487; CA216160 to S.K.