Recent trends in targeting miRNAs for cancer therapy

Introduction

The small noncoding RNAs that regulate the gene expression post-transcriptionally are called microRNAs (miRNAs).[1] They are endogenous to the cell[2] and are typically 20–22 nucleotides long.[3] Through various experimental studies, it is now established that miRNA dysregulation has a profound effect on the development and progression of cancer in humans. They control a wide array of biological processes, including carcinogenesis through the transcriptional activation of oncogenes. miRNAs act on numerous target RNAs by identifying the complementary region in the 3′ untranslated region (UTR) and through this mechanism regulate various biological processes such as cell differentiation, proliferation and apoptosis,[4] via feedback mechanisms.[5]

Ambros et al. discovered miRNA in Caenorhabditis elegans (C. elegans) called lin-4, which regulates lin-14 protein expression for the first time and thus laid the foundation for miRNA research.[6] Thereafter, Reinhart et al.[7] showed that let-7 negatively regulates lin-14 protein expression by RNA–RNA sequence-specific interaction on the 3′ UTR of the heterochronic gene lin-41. This study has revealed that these small noncoding RNAs bind to a sequence-specific target mRNA and alters its expression. Deregulation of a single or a small subset of miRNAs has shown to have a profound effect on the expression pattern of a huge pool of mRNAs.[8, 9] Consequently, several other studies have showcased that miRNAs are highly conserved across the domains of life, suggesting miRNAs have a general regulatory role post-transcriptionally.[10-12] All this research has established miRNAs as a key regulator of cellular functions by specific interaction with epigenetic modifiers, proteins, transcription factors and RNP complexes.[13-15]

Calin et al.[16] showed for the first time the role of miRNA in human cancer by carrying out the studies on B-cell chronic lymphocytic leukaemia (CLL) cells. Two miRNA genes, miR-15a and miR-16-1, are very frequently found to be deleted in CLL cells. Furthermore, it was found that mir-15 and mir-16-1 genes act as a tumour suppressor by repressing bcl-2 protein, thereby inducing apoptosis.[16] This study established the role of miRNAs in cancer. Since then, several other groups have revealed the importance of miRNA in the development and progression of cancer.[17, 18]

The advent of next-generation sequencing and miRNA profiling methods has greatly facilitated our understanding of miRNAs for the purposes of cancer identification, classification, diagnosis and prognosis. In this review, we describe the role of miRNAs in cancer and the emerging role of miRNAs as therapeutic targets. Finally, we discuss about the challenges in miRNA research and its clinical applications.

miRNA biogenesis/biosynthesis and mechanism of regulation

The biogenesis of miRNAs is highly conserved. miRNAs are encoded into the genome in different ways, either by clusters of multiple precursors or by expression from intergenic transcripts, each encoding only a single strand of pre-miRNA (which then adopts a hairpin-like secondary structure).[19] Transcription of pri-miRNA is carried out typically by RNA polymerase II; however, some are processed by RNA polymerase III.[20, 21] pri-miRNA is then translocated within the nucleus where DGCR8 (RNA-binding protein) and DROSHA (a type III RNase) like endonuclease enzymes cut the transcribed sequence and result into an 80–100 nucleotide long pre-miRNA sequence.[22, 23] The Ran/GTP/Exportin-5 complex then exports the pre-miRNA from the nucleus to the cytoplasm.[24] In the cytoplasm, the cytoplasmic ribonuclease (RNase III) enzyme called Dicer is present and cleaves the pre-miRNA into a double-stranded mature miRNA strand.[25]

As a result of this processing method, the released mature single strand miRNA binds to Argonaute 2 (AGO 2) resulting in a complex called the RNA-induced silencing complex (RISC). This complex has an intrinsic capacity of binding to typical 3′UTRs that are specific to their cytosolic mRNA targets. Binding to mRNAs is based on the complementarity between the base-pairing at the 5′ end of mature miRNA or open reading frame and the cytosolic mRNA molecule, with the binding site known as seed region, being about 6–8 bp long from the 5′ end of the miRNA. The short length of the binding site enables the miRNA to target a large number of different mRNAs.[26-28] miRNA biogenesis is regulated by methyltransferase like 3, which by methylating pri-miRNAs, marks them, and enables DGCR8-based identification and processing, eventually resulting in a mature miRNA.[29] (Figure 1).

image miRNA biosynthesis and regulation. microRNA is transcribed by RNA polymerase II to yield pri-miRNAs which is cleaved by a complex of Drosha and DGCR8 leading to the formation of a hairpin-like structure called pre-miRNA. The exportin-5-Ran-GTP exports this structure from the nucleus to the cytoplasm. Here, the multiprotein complex of TRBP (trans-activation-responsive RNA-binding protein) and the RNase Dicer cleaves it to form a mature microRNA sequence. This mature miRNA strand is incorporated into the RISC (RNA-induced silencing complex), which consisting of AGO2 (Argonaute 2) and GW182. Facilitated by this complex, it partially binds to complementary sequences in the 3′ UTR of target mRNAs, thereby controlling mRNA translational repression or degradation.). [Colour figure can be viewed at wileyonlinelibrary.com]

To have a deeper understanding of how miRNA-mRNA base-pairing regulates gene expression, Helwak et al. used an unbiased technique called CLASH. They found additional noncanonical binding cluster which was independent of the seed region and interaction complexity. Once the interaction of mRNA and miRNA is formed, imperfect complementarity leads to translational repression, whereas perfect complementarity leads to mRNA degradation.[30, 31]

miRNA in tumours have shown to act as a ligand, upregulating various types of signalling pathways. Toll-like receptor I was found to be affected by miRNA in natural killer cells by modulation of a nuclear factor-kB signalling pathway.[32] For example, the miR-21/miR-29a was secreted by the tumour cells and signalled to immune cells by binding TLR8, inducing a pro-metastatic inflammatory response, which might contribute towards tumour growth and tumour metastasis.[33] Thus, in many ways, any alteration to miRNA biogenesis significantly influences various cancer-related mechanisms and pathways.

Role of miRNA in cancer

In the last few years of research, miRNAs have been established as a novel cell component differentially expressed in pathological and normal cells.[34] Recently, advances have demonstrated the importance of miRNAs in cancer biology through their regulation of gene expression. miRNA acts as a helper in facilitating tumour invasion, growth, immune invasion and angiogenesis.[35, 36] These findings have highlighted possible miRNA-based biomarkers associated with cancer that can be detected in various body fluids and would allow for less invasive detection and monitoring of cancer.[37]

The first example of alteration of miRNA levels in cancer was reported in CLL when a cluster of miR-15 and miR-16 was identified at 13q14.3, which is frequently deleted in CLL.[16] Hanahan and Weinberg[38] have established the role of miRNA as a hallmark in several different types of cancer by studying the ‘tumour microenvironment’. Different types of tumours show specific miRNA signatures which help in the discrimination of various cancer types.[39] Through multiple studies, various cancer-associated targets and their respective miRNAs have been well-characterized (see Table 1).

Table 1. Regulatory role of miRNA in different cancer types Type of cancer MicroRNA Effect on CSCs property References Lung miR-34a Inhibitory effect by targeting CD44 [179 ] miRNA-200b Inhibition of HDAc 1 and Suz-12 [180 ] Breast Let-7 Inhibits self-renewal and dedifferentiation by targeting RAS and HMGA2 [181 ] miR-200 family Inhibits EMT, self-renewal and mammosphere formation

[182 ]

[183 ]

[64 ]

miR-22 Reduces expression of miR-200 family [184 ]

Leukaemia (AML and MDS)

AML – acute myelogenous leukaemia

MDS – myelodysplastic syndrome

miR-22 Promotes self-renewal [184 ] Prostate miR-34a Inhibits self-renewal and metastasis by targeting CD44 [185 ] miR-320 Inhibits Wnt signalling pathway [186 ] miR-25 Inhibitory effect by targeting cytoskeleton αv- and α6-integrin [187 ] Liver miRNA-150 Inhibitory effect by targeting c-Myb [188 ] Pancreas miR-200c Inhibitory effect by targeting ZEB1 and E-cadherin [189 ] Brain miR-17 Promotes cell proliferation of CD133+ [190 ] Colon miR-451 Inhibits tumorigenicity and self-renewal bt targeting COX-2 [191 ]

Regulation by miRNAs is mainly carried out by two different functions: (1) the homeostatic maintenance of gene regulation, which is highly cell-type dependent and (2) cell fate specification and the preservation of cell identity through feedback mechanisms.[34] In response to stress, changes in miRNAs assist cells in adapting to the altered conditions in their microenvironment.[19] This has been observed in the case of glioblastoma, wherein low miR-451 levels correlate with low glucose levels. miR-451 regulates the AMP-activated protein kinase pathway activation and suppression, which in turn regulates the cell survival and mammalian target of rapamycin-activated cell proliferation.[40] microRNA-specific genetic alterations are observed in cancer cells leading to a modification in target binding, processing and post-transcriptional changes in 3′UTR of mRNA.[41]

microRNA regulation of mRNA is lost in cancer cells during mRNA splicing due to deletion of 3′UTR, single nucleotide polymorphism and mutations.[42] Mutations causing a reduction in efficiency of miRNA processing machinery lead to a significant reduction in the total amount of mature miRNA in the cell. Often, low levels of mature miRNA are observed in tumours,[39] which may be caused by genetic loss, epigenetic silencing and changes in the biogenesis pathway or through transcriptional repression.[43, 44] The same can also be observed in some microsatellite unstable cancers. Here, mutations in exportin-5 (XPO5) lead to trapping of pre-miRNAs inside the nucleus, preventing further processing of miRNAs.[45]

Reduced levels of DICER expression have been found in various human carcinomas like lung cancer, ovarian cancer and CLL.[46-48] Binding of BCDIN3D (Bicoid-interacting 3, domain-containing) regulates O-methylation of 5'monophosphate leading to an alteration in miRNA processing, as that methyl mark is required for efficient cleavage by DICER, and therefore negatively regulates miRNAs.[49] For example, in ovarian cancer reduced Dicer expression has a direct correlation with drug resistance marker and poor drug therapy outcome.[50] In contrast to these findings, overexpression of DICER has been implicated in prostate cancer progression.[51] Furthermore, the amplification of the Drosha locus is observed in oesophageal cancer.[52] These suggest that it is important to establish, in multiple cancer types, both the frequency of these mutations and alterations in the miRNA expression signature.

Mechanisms of miRNA dysregulation in cancer

In human malignancies, high irregularity in the miRNA expression level is observed in cancerous cells as compared to the normal cells. The major causes of these alterations of miRNA expression in the cancer are summarized in Figure 2:

image Different mechanisms of miRNA deregulation in cancer. [Colour figure can be viewed at wileyonlinelibrary.com] Amplification or deletion of genes encoding miRNAs

The alteration of miRNA expression levels in malignant cells is thought to be caused by gene amplification, deletion or translocation. For instance, amplification of gene representing miR-17-92 clusters is observed in lung cancer and B-cell lymphomas.[17, 18] Conversely, in B-cell CLL patients, there is a loss of miR-15a/16-1-related genes at 13q14 chromosome.[16] Similarly, the deletion of miR-143 and miR-145 is observed at 5q33 region in lung cancer patients.[53] However, translocations are also observed, as in the T-cell acute lymphoblastic leukaemia translocation of miRNA-17-92 causes overexpression of this miRNA.[54]

These data have been further confirmed by array-based comparative genomic hybridization technique for 227 specimens representing human breast cancer, ovarian cancer and other melanomas.[55] Furthermore, whole-genome sequencing of these samples showcased that a high amount of miRNA genes is located in cancer-associated genomic regions, that is tumour suppressor genes, oncogene or common breakpoint regions. Thus, specific regions in the genome are responsible for the altered miRNA expression profiles due to deletion, amplification and translocation of certain specific genomic sites.

Dysregulated epigenetic change

Abnormal epigenetic modifications like tumour suppressor genes hypermethylation, variation in histone modification pattern and global DNA hypomethylation are the characteristic features of cancer cells.[56] A high proportion of miRNA loci is associated with CpG islands, indicating the role of DNA methylation-based epigenetic regulation of miRNA expression.[57] One example is the epigenetic silencing of miR-223 expression by AML1/ETO (AML fusion protein) via CpG methylation.[58] Seventeen miRNAs were upregulated by more than threefold in DNA methylation and histone acetylation inhibitor-treated T24 bladder cancer cells. Of these 17 miRNAs, miR-127 (embedded in CpG Island) was highly upregulated in treated cancer cells as compared to normal cells, simultaneously downregulating the proto-oncogene BCL6. These results suggest that miRNA expression-based tumour suppression can be achieved by the use of DNA methylation and histone acetylation inhibitor treatment.[59] Furthermore, DNA hypomethylation-mediated upregulation of potential oncogenic miRNA has been exhibited through various studies.[60, 61]

The miRNA and epigenetic mechanisms have been shown to have a strong relationship with cancer as miR-29 expression inhibit the expression of DNMT3A and DNMT3B,[62] genes required for regulating the DNA methylation. Restoration of miR-29 levels in NSCLC (nonsmall-cell lung cancer) caused derepression of CpG island methylation-silenced tumour suppressor genes. EZH2 a type of histone methyltransferase is targeted by miR-101 leading to target gene silencing and regulates cancer cell’s survival and metastasis.[63] SUZ12 a polycomb repressor complex 2 component is targeted by miR-200 family, having an ability of cancer stem cell (CSC) formation. Loss of miR-200 expression consequently leads to increased expression and binding of SUZ12, H3-K27 tri-methylation and E-cadherin gene repression.[64]

The miR-148a and miR-34b/c were discovered by Lujambio et al.[65] as a hypermethylation-specific silencer of cancer cells with decreased tumour growth and metastasis formation. All these examples demonstrate the role and importance of epigenetic regulation by miRNAs and its ability to alter DNA methylation and histone acetylation levels of the described genes, thus, showcasing its utility as cancer diagnostic or prognostic biomarkers.

Transcriptional control of miRNA

Almost half of the genes representing miRNAs are present in the introns of protein-coding genes or the long noncoding RNA genes and have their associated promoters and enhancers.[66] Transcription of genes that solely encode miRNAs is performed by RNA polymerase II.[21] miRNAs are mainly transcribed as a polycistronic message since the miRNA gene is present in a clustered form. A plethora of RNA polymerase II-associated transcription factor governs several miRNA genes by a single factor, generally via a complex circuit of feedback and feed-forward loops.

It is evident from multiple studies that transcription factors like c-Myc and p53 govern the expression of miRNAs in different cancers. Generally, c-Myc is upregulated because of miR-17-92 cluster activation, regulating apoptosis and cell proliferation of malignant cells.[67] Furthermore, c-Myc downregulates the transcriptional activity of miR-15a, miR-26, miR-29, miR-30 and let-7 families of tumour-suppressive miRNAs.[43] This is thought to be a result of the feedback loop where c-Myc regulates miR-122 by binding to its promoter, whereas Tfdp2 and E2f1 are indirectly inhibited by miR-122, thus inhibiting c-Myc transcription. Thus, showcasing the importance of this feedback loop in the development of carcinoma.[68] In nonsmall-lung cancer, expression of miR-221/miR-222 clusters is controlled by hepatocyte growth factor receptor c-MET, which in turn controls AP1 and ELK-1 transcriptional factors, initiating a negative feedback loop with miR-27a.[69, 70] A similar loop is observed where miR-148a-5p/miR-363-3p gene promoter is directly targeted by c-Myc, repressing their expression. These also promote the progression of cell cycle, specifically from G1 to S phase. As a response, c-Myc expression is directly inhibited by miR-148a-5p and destabilized by miR-363-3p via direct targeting of ubiquitin-specific protease.[71]

A synergistic type of loop is observed between p53 and miR-34, imparting a tumour-suppressive activity.[72] It is shown that p53 directly binds with mir-34a gene promoter and triggers the apoptosis process.[73, 74] As a feedback response miR-34a directly targets SIRT1 and downregulates it, in turn SIRT1via deacetylation negatively regulates p53 and prevents transcriptional dependent apoptosis by p53. However, an increase in transcriptionally independent p53-mediated apoptosis is observed.[75] The expression of miR-107,[76] miR-605[77] and miR-1246[78] is also regulated by p53. p63, a p53 family member, is capable of regulating Dicer1 transcription. In tumours with a p63 deficiency, very low Dicer1 expression level is observed, leading to levels of low mature miRNAs, the consequence of which is an increased tendency for metastasis.[79]

Defects in miRNA biogenesis machinery

Dysregulation of enzymes and/or cofactors, like Dicer, Drosha, DGCR8 and exportin 5 that are involved in the biogenesis pathways, significantly affects the overall mature miRNA levels. As evident in both in-vitro and in-vivo models, when Dicer1 and Drosha were partially deleted, faster tumorigenesis was observed in different types of tumours.[80] Drosha processing has emerged to be a critical step in the regulation of miRNAs in both cancers and in embryonic development.[81] Similar results can be observed in Dicer dysregulation, as in colorectal cancer cells Dicer1 impairment has led to higher tumour metastasis and initiation capacity.[82] Furthermore, increased median survival has been witnessed in ovarian cancer patients with high mRNA levels of Dicer and Drosha.[83] Conversely, the reduced patient survival rate can be correlated with decreased Dicer expression levels.[47, 84]

Argonaute proteins (AGO) play a central role in RNA-silencing, and their dysregulation can have serious implications in cancer. The loss of human EIF2C1/hAgo1 gene has been observed in Wilms’ tumour of the kidney.[85] Low AGO2 expression has been evident in melanomas as compared to primary melanocytes.[86] On the other hand, a high AGO2 expression has been observed in primary gastric cancer patients.[87] Lin28, a highly conserved RNA-binding protein that modulates the processing of miRNA let-7, has been implicated in oncogenesis, cell pluripotency and developmental timing.[88] Exportin 5 (XPO5) is a dsRNA-binding protein that is responsible for the export of pre-miRNA from the nucleus to the cytoplasm. A truncated version of the XPO5 gene is unable to export pre-miRNA from the nucleus, and as a result, pre-miRNA is trapped in the nucleus, leading to low mature miRNA processing.[45] Interestingly, XPO5 function restoration normalizes miRNA processing and also provides tumour suppressor activity. It is noteworthy that various other miRNAs are capable of regulating miRNA processing. In aggressive breast cancers, the miR-103/107 family of miRNAs targets DICER and thus reduces the level of the global miRNA. In summary, the key mechanisms linking miRNAs to cancer are chromosomal abnormalities, transcriptional changes, nuclear receptors and defects in miRNA biogenesis.

Altered miRNA expression in tumours

Tumours acquire the ability to resist apoptosis, dodge growth suppressors, maintain proliferative signalling, empower replicative immortality, provoke angiogenesis and initiate invasion and metastasis.[38] miRNA profiling of these tumours has shown abnormal expression as compared to the normal tissues and hence is believed that dysregulated miRNAs function as either tumour suppressor genes or oncogenes depending on the gene target, affecting any of the above-mentioned hallmarks. The balance between extracellular signalling molecules and intracellular processes controls the cell cycle progression. Through different studies, it has been apparent that miRNAs are integrated into multiple cell proliferation pathways, therefore sustaining proliferation and evades growth suppression in cancerous cells.

The E2F proteins, in a cell cycle-dependent fashion, are key cell proliferation regulators, which are in turn regulated by miRNAs. In the G1 to S transition period, E2F1-mediated induction of gene transcription has been observed.[89] Several different types of cancer were observed in E2F1-/- mice, suggesting the role of E2F1 as a tumour suppressor. E2F1 translation is inhibited by miR-17-92 cluster post c-Myc activation.[67] E2F2 and E2F3 translation are also regulated by the miR-17-92 cluster.[90] A feedback mechanism regulates the expression of miR-17-92 cluster and E2F to achieve cell cycle progression in normal cells.[91] In tumorous conditions, disruption in the feedback loop can be observed due to miR-17-92 cluster overexpression, leading to cell proliferation.[92] miRNAs also regulate cyclins, cyclin-dependent kinases (CDKs) and CDKs inhibitors on whom the cell cycle progression is dependent.

Dicer-1 knockout in germline stem cells of Drosophila blocked the transition from G1 to S phase, demonstrating the importance of miRNAs in this transition.[93] Furthermore, in this context increased expression levels of CDK inhibitors (Dacapo) of the p21/27 family were also observed, suggesting that downregulation of the protein by miRNAs would boost cell cycle progression. Cdk inhibitor p27kip1 is directly targeted by miR-221/222 in glioblastoma cells.[94] In cancerous cells, high expression of miR-221/222 speeds up cell proliferation, and its low expression causes G1 cell cycle arrest. These data are well correlated with both primary tumour samples and cancer cell line studies.[95-97] Moreover, the upregulation of the miR-221/222 is observed in various human tumours, confirming the findings that Cdk inhibitor p27kip1 regulation is a part of an oncogenic programme. The miRNA family of miR-302, miR-663 and miR-24 regulates the p21CIP1 and p16INK4a other than p27Kip1.[98, 99] miR-663 and p21CIP1 form a loop at the molecular level and are responsible for cell proliferation in nasopharyngeal cancer.[100] miRNAs also regulate the expression of cyclins and Cdks, as the expression levels of CDK4 and cyclin D1 are decreased by miRNA-545 in lung cancer cells as a consequence of cell cycle arrest.[101] miRNAs also regulate a variety of signalling pathways thereby affecting cell proliferation. For example, miR-486 affects cell proliferation and migration by targeting p85α, IGF1 and IGF1R of phosphoinositide-3-kinase (PI3K) and insulin growth factor (IGF) signalling pathways.[102] Thus, concurrent and extensive indications of altered miRNAs have been implicated in cancer, representing as a candidate target for treating cancer.

Targeting key cancer-related pathways Cell cycle and cell proliferation as targets

It has been well-established that miRNAs have a key role in controlling cell proliferation, altering various regulatory pathways, and, hence, have a profound effect on carcinogenesis. Oncogenic miRNAs are typically overexpressed and act as a facilitator for cancerous cells to enter and progress through the cell cycle. miRNAs that suppress tumour growth are typically lost during cancer and, hence, normally help in inducing cell cycle arrest.[103]

The retinoblastoma (pRb) pathway has a significant effect on the regulation of the cell cycle and is affected in a variety of human cancers.[104,105] It acts by repressing the transcription factor family E2F, which governs the gene expression of genes essential for cell cycle progression.[106] Cyclin-dependent kinases mediated phosphorylation of pRb leads to activation of transcription of genes by E2Fs. Specific kinases and cyclins form complex with active CDKs and aid in the progression of the cell cycle through its sequential phases.[107] These important cell cycle components (i.e. CDKs and cyclins) are targeted by growth-restricting miRNAs, acting on growth diminishing pathways such as p53 or by growth-enhancing mitogenic pathways such as RAS/RAF/MAPK.[107, 108] For example, miR-20a, miR-125b and miR-17-92 clusters possess a tumour-suppressing function by targeting the E2F transcription factor.[67,109, 110]

The miRNAs regulate cell cycle inhibitors, which negatively regulates the CDKs, as shown by the CDK inhibitors from the cip/kip family. miR-106b and miR-17-92 families act upon p21, which is a potent CDK inhibitor and a primary mediator of the downstream cell cycle’s G1 phase arrest of the p53 gene. Wu et al. experimentally demonstrated that about 28 miRNAs have the potential to target the 3'UTR region of p21 mRNA by a luciferase assay.[111] Similarly, p27 and p57 are controlled post-transcriptionally by miRNAs. In particular, p57 is controlled by the miR221/222 cluster.[96, 112] Thus, miRNAs have a significant impact on cancerous cell entry and progression through the cell cycle.

Senescence as target

Senescence is an irreversible exit from the cell cycle. It is mainly of two different types, replicative and premature senescence. The replicative senescence occurs due to the shortening of telomeres, and premature senescence occurs due to higher oxidative stress levels, DNA damage signalling or increased oncogene expression levels.[113] miRNAs negatively regulate cell cycle progression and hence plays a role in the induction of senescence. For example, the senescence inducers p16 and p19 are repressed by HMGA2, which is in turn a primary target of miRNA let-7.[114-

留言 (0)

沒有登入
gif