miRNA and Gene Expression in Pancreatic Ductal Adenocarcinoma

Abstract

Pancreatic ductal adenocarcinoma (PDAC) remains a challenging disease that is mostly diagnosed late in the course of the illness. Unlike other cancers in which measurable successes have been achieved with traditional chemotherapy, targeted therapy, and, recently, immunotherapy, PDAC has proved to be poorly responsive to these treatments, with only marginal to modest incremental benefits using conventional cytotoxic therapy. There is, therefore, a great unmet need to develop better therapies based on improved understanding of biology and identification of predictive and prognostic biomarkers that would guide therapy. miRNAs are small noncoding RNAs that regulate the expression of some key genes by targeting their 3′-untranslated mRNA region. Aberrant expression of miRNAs has been linked to the development of various malignancies, including PDAC. A series of miRNAs have been identified as potential tools for early diagnosis, prediction of treatment response, and prognosis of patients with PDAC. In this review, we present a summary of the miRNAs that have been studied in PDAC in the context of disease biology.

Pancreatic ductal adenocarcinoma (PDAC) is the main histologic type (>90%) of cancer arising from the pancreas. A total of 55,440 new cases of PDAC are estimated to be diagnosed in 2018, whereas 44,330 deaths are estimated to be caused by PDAC in the United States.¹ In the United States, PDAC is currently the third leading cause of cancer-related mortality; however, it is projected to be the second leading cause of cancer-related death by 2020.1, 2 Progress in PDAC treatment has been slow. The 5-year overall survival rate has only improved to 9% during the period 2006 to 2012 in the United States, from 3% for the period 1975 to 1977.³ Modern cancer chemotherapy showed modest incremental improvement in patient outcomes, especially in patients who have advanced disease. The median overall survival of patients with metastatic PDAC treated with newer regimens has improved to approximately 11 months compared with the historic benchmark of 5 to 6 months.4, 5 Because of rapid clinical decline, most patients do not receive further chemotherapy after first-line therapy has failed.⁶ Patients who receive second-line treatment have been shown to live for >1 year.7, 8 Despite poor specificity and sensitivity, CA19-9 remains the only biomarker in routine clinical use as a surrogate for prognosis and response to therapy.9, 10, 11, 12 Clinical trials in PDAC that were completed in the past few years have shown no therapeutic benefit from newer agents that include drugs aimed at intracellular molecular targets and the immune system.¹³

miRNAs and Pancreatic Ductal Adenocarcinoma

miRNAs are noncoding, single-strand RNAs of approximately 22 nucleotides that interact with untranslated regions of a target mRNA, resulting in their reduced translation or a rapid degradation. More than 700 miRNAs have been described so far. miRNAs represent approximately 3% of the human genome¹⁴ and control the expression of >60% of the protein-coding genes in humans.¹⁵ They serve as regulators of gene expression, thereby playing a key role in the regulation of a wide range of cellular processes, including growth, differentiation, proliferation, and apoptosis. They may also play oncogenic or tumor-suppressive functions and have been profiled in various cancers, including breast, lung, colorectal, pancreas, and liver cancers.¹⁶

The differential expression of miRNA in malignancies has led to their evaluation as biomarkers of diagnostic and therapeutic importance. In addition, description and targeting of oncogenic miRNAs and those with tumor suppressor functions has been a major research focus. Multiple genetic abberations involved in the development of PDAC are shown to be influenced by miRNAs.¹⁷ Table 1 summarizes a few examples of aberrant expression of some of the miRNAs described in malignancies.18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33

Table 1.

Examples of Dysregulated miRNAs with Their Potential Targets and Suggested Biological Functions

miRNAs	Expression	Potential targets	Biological function	References
miR-34a	Down-regulation	Bcl-2/Notch1/Notch2	Inhibits proliferation and invasion and induces apoptosis	18
miR-221/222	Up-regulation	MMP-2 and MMP-9	Cancer cell invasion	19
miR-320a	Up-regulation	PDCD4	Reduces apoptosis and promotes EMT	20
miR-365	Up-regulation	SHC1/BAX	Reduces apoptosis	21
miR-21	Up-regulation	PI3K/AKT/PTEN, PDCD4, and Bcl-2/FasL	Reduces apoptosis and promotes proliferation	22, 23, 24
miR-29a	Up-regulation	Wnt/β-catenin	Regulates transcription factors	25
miR-210	Up-regulation	Pancreatic stellate cells	Promotes invasion and EMT	26
miR-100	Down-regulation	FGFR3	Inhibits proliferation	27
miR-146a	Down-regulation	MAPK/Kras/EGFR/IRAK/MTA-2	Inhibits invasion	28
miR-145/143	Down-regulation	KRAS	Inhibits tumor growth	29
miR-33a	Down-regulation	AKT/β-catenin and Pim-kinase	Inhibits proliferation	30
miR-101-3p	Down-regulation	RRM1	Interferes with DNA synthesis	31
Let-7	Down-regulation	N-cadherin/ZEB1	Reverses EMT and inhibits invasion	32
miR-211	Down-regulation	RRM2	Reduces drug targets	33

EGFR, epidermal growth factor receptor; EMT, epithelial-to-mesenchymal transformation; FGFR3, fibroblast growth factor receptor 3; IRAK, interleukin-1 receptor-associated kinase 1; KRAS, kirsten rat sarcoma virus; MAPK, mitogen-activated protein kinase; MMP, matrix metalloproteinase; MTA-1, metastasis-associated protein 1; PDCD4, programmed cell death 4; PI3K, phosphatidylinositol 3-kinase; PTEN, phosphatase and tensin homolog; RRM1, ribonucleoside-diphosphate reductase large subunit; SHC1, SHC transformic protein 1; ZEB1, zinc finger E-box-binding homeobox 1.

The interaction between miRNA and mRNA molecules was investigated in pancreatic cancer using bioinformatics techniques. There were 17,401 relevant miRNA-mRNA interactions estimated to occur in pancreatic cancer, highlighting the complexity of miRNA-mRNA interactions in PDAC and the challenge to identify the most relevant target for therapeutic development.³⁴

miRNAs and Molecular Changes in Pancreas Ductal Adenocarcinoma

Oncogene Activation and miRNAs

Activating KRAS mutations are the earliest genetic changes in the malignant transformation of the pancreatic ductal epithelial cells to adenocarcinoma and are seen in nearly all cases of PDAC at diagnosis. Activation of the RAS family oncogenes leads to cellular proliferation, survival, and invasion through the stimulation of several effector pathways.³⁵ miRNAs targeting KRAS and its downstream genes have been identified. Down-regulation of miR-96, miR-126, and miR-217 was shown in KRAS mutant PDAC, unlike in normal pancreatic tissues.36, 37, 38 KRAS activity was suppressed by the reexpression of miR-96 and miR-217, with resultant reduced cell migration and invasion, highlighting a tumor suppressor–like function.37, 39 Overexpression of miR-217 also influences AKT, downstream to RAS proteins, with effects on cell survival and proliferation.³⁹ The loss of miR-143/145 cluster expression is seen in KRAS mutated PDAC, and its reexpression has been shown to mitigate carcinogenesis and inhibit RAS-responsive element–binding protein, mitigating the effects of KRAS activation in preclinical models.²⁹ The expression of the miRNA let-7 has been shown to suppress KRAS expression and mitogen-activated protein kinase activation, whereas its loss plays a role in KRAS-activating mutations.⁴⁰

miRNAs and Tumor Suppressor Gene Inactivation

Tumor suppressor genes play a key role in triggering cell cycle arrest or apoptosis when there is irreparable DNA damage. Their inactivation leads to the loss of their regulatory function, thereby favoring the growth and sustenance of malignant cells. The loss of tumor suppressor genes Ink4A (p16/CDKN2A), TP53, SMAD4, and BRCA2 has been well described in PDAC.⁴¹ Ink4A acts as a checkpoint that controls the continued cellular proliferation from KRAS activation and leads to cellular senescence. The functional loss of Ink4A, seen in up to 95% of PDAC, plays a key role in the perpetual growth of malignant PDAC cells by overriding the cellular senescence induced by activated KRAS.42, 43

The inactivation of TP53 is seen in >50% of PDAC, mostly in the later stage of carcinogenesis.³⁵ The rate of genomic aberrations increases dramatically after loss of the p53 function, leading to tumor heterogeneity and resistance to multiple chemotherapeutic agents.⁴⁴ Mutant p53 affects the transcription of miR-15a, which is suppressed in PDAC. The overexpression of miR-15a down-regulates WNT3A and FGF7, resulting in reduced cellular proliferation and survival.⁴⁵ A functional P53 induces the overexpression of miR-200, which leads to the down-regulation of Jag1, a target of Zeb1 and a ligand of the Notch pathway. The miRNAs Zeb1 and Zeb2 activate epithelial-to-mesenchymal transformation (EMT) and facilitate metastases of cancer cells.⁴⁶ The expression of Zeb1, on the other hand, has a negative feedback inhibition on the levels of miR-200.⁴⁷ Hence, loss of p53 function indirectly facilitates EMT through a complex interaction of miRNAs. On the other hand, miR-491-5p inhibited the expression of TP53, Bcl-XL genes, in addition to other effects.⁴⁸ In human PDAC cell lines with p53 mutations, the reexpression of miR-34 restores tumor suppressor function of p53.¹⁸

MDM2 is a proto-oncogene that results in the functional inactivation of tumor suppressor proteins, like p53. MDM2 overexpression induces PDAC cellular proliferation, migration, and invasion. At the post-transcriptional level, miR-509-5p negatively regulates MDM2, reversing the effects of MDM2 overexpression: cell proliferation, migration, and invasion. miR-509-5p is down-regulated in pancreatic cancer tissues.⁴⁹

The inactivation of SMAD4 [deleted in pancreatic cancer (DPC) 4] is seen in approximately 30% to 50% of PDAC, mostly in a later stage of carcinogenesis. The loss of SMAD4 leads to loss of transforming growth factor (TGF)-β–mediated growth inhibition and subsequent tumor progression by influencing tumor-stroma interactions.³⁵ miR-137 targets SMD4A mRNA during Ras-induced senescence, resulting in the activation of tumor suppressors p53 and retinoblastoma. The expression of miR-137 is significantly reduced in PDAC, resulting in SMAD4-mediated loss of tumor suppressor functions. Restoring its expression restores the tumor suppressor functions lost in these pancreatic cancer cells.⁵⁰ In another study, miR-421 and miR-483-3p promoted PDAC progression through directly regulating the tumor suppressor SMAD4.51, 52

Overexpression of miR-192 leads to a reduction in the expression of SMAD-interacting protein 1.³⁹ miR-323-3p inhibited TGF-β signaling, via direct suppression of SMAD2 and SMAD3, leading to decreased metastasis.⁵³ In irradiated pancreatic cancer tumor models, a high expression of miR-193a was seen, which promotes pancreatic cancer cell repopulation and subsequent metastases in preclinical models. miR-193a inhibited TGFβ/SMAD/c-Myc signaling and disrupted normal intercellular junctions with metastases promotion. Pancreatic cancer repopulation and metastasis after radiation were blocked by the restoration of TGF-β/SMAD signaling or suppression of miR-193a.⁵⁴

P16, otherwise known as cyclin-dependent kinase inhibitor 2A, functions as a tumor suppressor gene. Abberations of p16 expression have been well described in PDAC.⁵⁵ Studies have shown the inhibitory role of miR-10b and miR-24 on the expression of p16 in malignancies other than pancreatic cancer.56, 57 Both miR-10 and miR-24 were overexpressed in pancreatic cancer.58, 59

Down-regulation of miR-7-5p in PDAC correlates with poor survival outcomes in patients. In vitro cell function assays demonstrated that miR-7-5p inhibited proliferation, migration, and invasion of PANC-1 cells by inversely regulating SOX18. Silencing of SOX18 leads to the suppression of the gp130/JAK2/STAT3 signaling pathway in PANC-1 cells. Hence, miR-7-5p likely has a tumor suppressor role, and its down-regulation may lead to enhanced gp130/JAK2/STAT3 signaling.⁶⁰

miRNAs in PDAC Stemness

Accumulating evidence shows that most solid tumors have a heterogeneous cell mix, some of which are a cancer stem cell (CSC) population that is generally resistant to conventional cytotoxic chemotherapy. Several research groups have identified the overexpressed molecular markers that are relatively specific to the PDAC CSCs. These include CD24, CD44, CD133, epithelial cell adhesion molecule (EPCAM), epithelial-specific antigen (ESA), tyrosine-protein kinase met (c-met), aldehyde dehydrogenase (ALDH1), leucine-rich repeat-containing receptor (lgr5), and serine/threonine-protein kinase (Dclk1). Because these cancer stem cells have self-renewal capabilities, they can repopulate tumor cells that possibly carry newer genetic mutations, if the bulk of cancer cells have been eliminated by treatment.61, 62 This characteristic of cancer stem cells may be influenced by specific miRNAs.⁶³ A differential expression of miRNAs in cancer stem cells has been described, including miR-99a, miR-100, miR-125b, miR-192, and miR-429.⁶⁴

Various miRNAs have been implicated in promoting stemness. In PDACs, the loss of miR-34 was seen in CSCs, and its restoration could suppress the spheroid-forming ability through suppression of notch and bcl-2 and restoration of p53.¹⁸ The EMT-activator ZEB1 (a zinc finger E-box binding homeobox 1) represses the expression of stemness-inhibiting miR-203. The candidate targets of miR-200 family members are also stem cell factors, such as sex determining region Y–box 2 (Sox2) and Kruppel-like factor 4 (Klf4).⁶⁵ These studies show that the ZEB1-miR-200 axis may become a novel therapeutic target against PDAC CSCs. Larger-scale differential expression studies performed using microarray analysis, in sphere culture methods enriching the CSC, identified 210 miRNAs and 258 stem cell–associated mRNAs that were differentially expressed in the PDAC CSCs.⁶⁴ Several other studies have identified a series of miRNAs with prominent roles in promoting EMT and cancer stemness in PDAC (Table 2).66, 67, 68, 69, 70, 71, 72, 73, 74, 75

Table 2.

Prominent miRNAs with Confirmed Roles in PDAC Stemness

miRNAs	Status	Stemness markers regulated	References
miR-1246	Up-regulation	CCNG2	66
miR-200a	Down-regulation	N-cadherin, ZEB1, and vimentin	67
miR-1181	Down-regulation	SOX2 and STAT3	68
Mir-29c	Down-regulation	Wnt	69
miR-21 and miR-221	Up-regulation	CDK6, C5ORF41, EFNA1, IRAK3, KLF12, MAPK10, NRP1, SMAD7, SOCS6, and ZBTB41	70
miR-17-92	Up-regulation	NODAL/ACTIVIN/TGF-β1	71
miR-200c	Down-regulation	ZEB1	72
miR-744	Up-regulation	SFRP1, GSK3β, and TLE3 (Wnt)	73
miR-30a, miR-30b, and miR-30c	Up-regulation	EMT and CD133	74
miR-106a	Up-regulation	EMT	75

C5ORF41, chromosome 5 open reading frame 41; CCNG2, cyclin G-2 protein; CDK6, cyclin-dependent kinase 6; EFNA1, ephrin A1; EMT, epithelial-to-mesenchymal transformation; GSK3β, glycogen synthase kinase-3 beta; IRAK3, IL-1 receptor-associated kinase 3; MAPK, mitogen-activated protein kinase; NODAL, nodal growth differentiation factor; NRP1, neuropilin-1; PDAC, pancreatic ductal adenocarcinoma; SFRP1, secreted frizzled related protein 1; SOCS6, suppressor of cytokine signaling 6; SOX2, sex determining region Y–box 2 (alias SOY); TGF-β1, transforming growth factor-β1; TLE, transducin-like enhancer protein; ZBTB41, zinc finger and BTB domain containing 41; ZEB1, zinc finger E-box binding homeobox 1.

miRNA Epigenetic Regulation in PDAC

Several different tumor suppressor genes in human cancers are silenced by chromatin alterations, including promoter methylation and histone deacetylation. Likewise, miRNAs have also been shown to be regulated epigenetically.⁷⁶ Such epigenetic alterations may be used in the diagnosis and treatment of PDAC. However, the research on the impact of epigenetic modulation of miRNAs in PDAC is scarce.

Overexpression of several miRNAs that are generally suppressed, miR-29a, miR-29b, miR-103, miR-107, and miR-320, was shown in PDAC cell lines as a result of treatment with the histone deacetylase inhibitor trichostatin A or the hypomethylating agent decitabine. The enforced expression of miR-107 in MiaPACA-2 and PANC-1 cells inhibited their in vitro growth and suppressed the putative miR-107 target, cyclin-dependent kinase 6, thereby providing a functional basis for the epigenetic inactivation of this miRNA in PDAC.⁷⁷

A study of epigenetic regulation of different miRNAs in a broader panel of tumors (including mammary tumors, colorectal tumors, pancreatic tumors, and soft tissue sarcomas) showed consistent CpG methylation of miR-34a and miR-34-b/c in the tumor tissues.⁷⁸ Inducing overexpression of miR34a by treatment with decitabine or the histone deacetylase inhibitor vorinostat resulted in cell cycle arrest, reversal of stemness, suppression of EMT and invasion, and PDAC growth inhibition.⁷⁹ CpG island methylation was found to lead to the underexpression of the tumor suppressor miR-1247.⁸⁰ Contrary to noncancerous tissues, PDAC contains highly methylated miR-124-1, miR-124-2, and miR-124-3. Hypermethylation of these genes was linked to worse survival of PDAC patients.⁸¹ Functional studies showed that miR-124 inhibited cell proliferation, invasiveness, and metastasis through KRAS downstream RhoGTPase effector Rac1, highlighting its tumor suppressor property, often silenced epigenetically in PDAC.

Mortality factor 4 (MORF4)-related gene-binding protein (MRGBP) binds directly to the histone acetyltransferase complex, components of tat-interacting protein 60, and histone deacetylase complexes. The loss of MRGBP suppresses the replication and growth of pancreatic cancer cells, promotes apoptosis, and inhibits migration and invasiveness. The overexpression of MRGBP has been documented in malignancies, including pancreas cancer. MRGBP expression is down-regulated by miR-137, its loss leading to overexpression of MRGBP. Adverse tumor characteristics were seen in patients with PDAC in whom there was loss of miR-137. The reexpression of miR-137 significantly blocked migration and invasion of pancreatic cancer cells. Hence, the loss of miR-137 may function as a novel tumor promoter by facilitating the overexpression of MRGBP in PDAC.⁸²

Higher serum levels of miR-21 as a result of promoter histone acetylation are seen in patients with gemcitabine-resistant PDAC compared with subjects with gemcitabine-sensitive tumors. In PDAC cells, after treatment with gemcitabine, the miR-21 levels increased significantly, causing a marked increase in invasion and metastatic potential through activation of Akt and reduction in phosphatase and tensin homolog (PTEN). These changes were reversible with anti–miR-21 transfection.⁸³

Given that several epigenetically silenced miRNAs have a role in promoting EMT, the link to miRNA epigenetic alterations and stemness has also been postulated. A genome-wide study of the DNA methylation profile in PDAC CSC investigated the impact of DNA methyltransferases in PDAC CSC maintenance and tumorigenicity. CSCs had higher levels of DNA methylation because of higher DNA (cytosine-5)-methyltransferase 1 (DNMT1) levels that lead to the suppression of the miR-17-92 cluster. In functional studies, pharmacologic or genetic targeting of DNMT1 blocked the self-renewal and in vivo tumorigenic potential of CSCs through the reexpression of miR-17-92.⁸⁴

The restoration of miR-377 levels resulted in a reduction of the expression of DNMT1 and reactivation of target genes BNIP3 and SPARC via promoter demethylation, leading to inhibition of cell proliferation and induction of apoptosis.⁸⁵ Epigenetic reprogramming of miR-663a and miR-4787-5p by synthetic histone methylation reversal agents attenuated TGF-β1–induced EMT features in human PDAC.⁸⁶ The levels of miR-663b are lower in PDAC than in normal tissue. Expression of miR-663b is suppressed epigenetically (on H3K4me3 and H3K27me3 promoters) by the long noncoding RNA, homeobox (HOX) transcript antisense RNA (HOTAIR). Overexpression of miR-663b resulted in inhibition of cell proliferation, invasion, and migration and induced apoptosis. Reexpression of miR-663b or silencing of HOTAIR blocked tumor growth by targeting insulin-like growth factor 2 in PDAC.⁸⁷ These proof-of-concept studies validate the concept of epigenetic modifications targeting miRNAs and may be explored further in the development of therapeutics for PDAC.

miRNA Deregulation Supporting the PDAC Microenvironment

Desmoplastic reaction, which is characterized by a dense concentration of extracellular matrix proteins, activated pancreatic stellate cells, and immune cells surrounding tumor, has been a well-described feature of pancreatic adenocarcinoma. This has been considered a barrier to the delivery of anticancer drugs, as one explanation for poor treatment outcome.⁸⁸ In experiments in which human pancreatic stellate cells were co-cultured with pancreatic adenocarcinoma cells, there was increased expression of fibrosis-related genes along with reduction in the expression of miRNA let-7d. The inhibition of the miRNA let-7d resulted in enhanced expression of fibrosis-related genes [α-smooth muscle actin, platelet-derived growth factor receptor-β, and collagen type XXI alpha 1 chain (COL1A1)]. Data from The Cancer Genome Atlas showed the expression of let-7d was consistently reduced in pancreatic cancer compared with normal tissue. In addition, patients with pancreatic ductal adenocarcinoma were shown to have lower serum levels of let-7d compared with healthy controls. Reduced let-7d expression correlated with poor overall survival in patients who were treated with gemcitabine-based chemotherapy.⁸⁹

miRNA Deregulation, Exosomes, and Other Cellular Transport Systems

Exosomes that are alias microvesicles are small vesicular structures that are approximately 40 μm in size.⁹⁰ These vesicular structures harbor several different types of biological molecules, ranging from RNA to proteins and even drug breakdown products.⁹¹ There is compelling evidence suggesting the role of exosomes in cancer-related signaling, due in part to the varied cargoes that they carry.⁹² It is not surprising to note that miRNAs are also transported through exosomes and have been implicated in disease progression and resistance to therapy.⁹³ PDAC is shown to have high exosomal levels of miR-10b, miR-21, miR-30c, and miR-181a and low miR-let7a compared with normal pancreatic tissue.⁹⁴

In patients with PDAC, serum levels of miR-155 and miR-196a were lower, whereas miR-17-5p and miR-21 were higher, when compared with healthy controls. The overexpressed miR-17-5p also correlated with metastasis and poor overall survival.⁹⁵ These miRNAs can be found free or exosome bound in the serum, which may be easily accessed for diagnostics, prognostics, and therapy response in patients with pancreatic cancer.⁹⁶

The connective tissue growth factor and miR-21, expression of which is highly up-regulated in activated pancreatic cancer stellate cells, are seen in pancreatic cancer stellate cell–derived exosomes. Such exosomal secretory miR-21 was shown to generate a positive feedback loop stimulating connective tissue growth factor expression. These exosomes facilitate the delivery of these agents to other pancreatic stellate cells.⁹⁷ Pancreatic cancer–derived exosomes were shown to down-regulate toll-like receptor 4 and downstream cytokines in dendritic cells via miR-203.⁹⁸ In addition, the regulatory factor X–associated protein, an important transcription factor for major histocompatibility complex II, was inhibited by miR-212-3p transferred from PC-secreted exosomes. Such miR-212-3p–mediated suppression of regulatory factor X–associated protein resulted in major histocompatibility complex II suppression and resulting immune tolerance in dendritic cells.⁹⁹

A successful transfection of Panc-1 cells with miR-155 and miR-125b2 caused alterations in their derivative exosome content, resulting in differential communication and reprogramming of the macrophage cells (J774.A1) to an M1 phenotype. This technical success has led to the hypothesis that genetic therapies targeted toward selective manipulation of tumor cell–derived exosome content may become a promising strategy in pancreatic cancer.¹⁰⁰ Table 3 summarizes exosomes described in miRNA aberrant expression in PDAC.101, 102, 103, 104, 105, 106

Table 3.

List of Up-Regulated Exosome-Transported miRNAs Identified in Serum or Tumor Microenvironment of Pancreatic Cancer

Exosome miRNAs	Disease stage and outcome	References
miR-196a	Pancreatic ductal adenocarcinoma	101
miR-1246	Intraductal papillary mucinous neoplasms	101
miR-1246, miR-3976, miR-4306, and miR-4644	Pancreatobiliary cancer diagnosis	102
ExmiR-191, ExmiR-21, and ExmiR-451a	Pancreatic cancer diagnosis	103
miR-451a	Pancreatic cancer recurrence	104
miR-155	Gemcitabine resistance	105
miR-23b-3p	Correlated to carbohydrate antigen 19-9 (CA19-9) levels	106

The human ether-a-go-go-related potassium channel 1 has been identified to play a carcinogenic role in multiple malignancies, including PDAC. High expression of human ether-a-go-go-related potassium channel 1 is associated with proliferation and invasiveness of pancreatic cancer cells. The expression of human ether-a-go-go-related potassium channel 1 has been shown to be inhibited by miR-493 and mi-96. The loss of miR-493 and mi-96 leads to increased expression of human ether-a-go-go-related potassium channel 1 in pancreatic cancer cells, leading to proliferation and invasion.107, 108

miRNA Deregulation and Immunosuppression in Cancers

Tumor-derived exosomes deliver miRNAs to immune system cells that facilitate immune evasion by the tumor cells, by affecting various parts of the immune system. There has been much work published specific to pancreatic cancer. We have summarized the impact of miRNAs on immune responses to various malignancies reported in the literature.

Pancreatic cancer cell–derived exosomes containing miR-212-3p down-regulate the transcription factor regulatory factor X–associated protein when pulsed onto immature dendritic cells, resulting in reduced expression of major histocompatibility complex class II molecules and possibly ineffective antigen presentation abilities of the dendritic cells. Exposing immature dendritic cells to miR-203 resulted in reduced toll-like receptor 4 expression, which may affect subsequent T- and B-cell activation by the dendritic cells.⁹⁹ Aberrant expression of miR-146a in dendritic cells from PDAC patients was observed, along with repression of SMAD4, which resulted in impaired differentiation and antigen presentation by dendritic cells.¹⁰⁹

In patient-derived esophageal adenocarcinoma specimens and cell lines, overexpression of miR125a-5p and miR148a-3p resulted in reduced levels of transporter associated with antigen processing (TAP)2 and major histocompatibility complex-I, which affect tumor cell recognition by specific cytotoxic T lymphocytes.¹¹⁰ Natural killer cells have been shown to be down-regulated by miR-23a carried in exosomes derived from hypoxic tumor cells,¹¹¹ whereas miR-362-5p has been shown to promote natural killer cell activation, which may lead to their overstimulation and subsequent exhaustion.¹¹² Tumor-derived exosomes from nasopharyngeal carcinoma were shown to contain overexpressed miR-24-3p, miR-106a-5p, miR-891a, miR-20a-5p, and miR-1908, which all have been linked to down-regulation of mitogen-activated protein kinase 1 and subsequent transformation of T cells from type 1 and 17 helper T-cell to type 2 helper T-cell and regulatory T-cell phenotypes.¹¹³ Tumor-derived exosomes from non–small-cell lung carcinoma were shown to deliver miR-21 and miR-29a to macrophages entering the tumor microenvironment. These miRNAs activated toll-like receptor 8 (murine toll-like receptor 7) and subsequently nuclear factor κ–light chain enhancer of activated B cell pathway, which culminated in IL-6 and tumor necrosis factor-α secretion, generating a protumor inflammatory environment.¹¹⁴

miRNAs as Diagnostic and Prognostic Tools in Pancreatic Adenocarcinoma

Diagnostic Utility of miRNAs

Differential expression of miRNA profiles has been well described in PDAC, with miRNAs isolated from various patient-derived specimens, including the peripheral blood, pancreatic tissue, and digestive juices. Circulating miRNAs are attractive targets for scientific exploration because of their abundance, stability, and technical ease of their isolation and amplification with inexpensive and noninvasive techniques.¹¹⁵

Several miRNA profiles had differential expression in PDAC tissue, unlike in healthy pancreatic tissue and benign pancreatic pathology. The overexpression of miR-483-3p in the blood differentiates PDAC from intraductal papillary mucinous neoplasms and healthy pancreas tissue with sensitivity of >40%, which is comparable to that of CA19-9 (45%).¹¹⁶ Plasma levels of miR-196a and miR-1246 are significantly higher in patients with early-stage PDAC when compared with healthy volunteers. Higher levels of miR-196a have better specificity to PDAC, whereas miR-1246 is seen with intraductal papillary mucinous neoplasms. In pancreatic neuroendocrine tumors, these miRNAs appear to be unchanged.¹⁰¹

The serum levels of miR-200a identify patients with PDAC from healthy controls with a sensitivity and specificity of >80%. The serum levels of miR-200b identify patients with PDAC from healthy controls with a sensitivity and specificity of 71.1% and 96.9%, respectively.¹¹⁷ Serum miR-1290 identified patients with early-stage pancreatic cancer from controls better than CA19-9, while additionally hinting at the pancreatic cancer cell invasion capability.¹¹⁸ When a panel of miR-16 and miR-196a was used in conjunction with CA19-9, PDAC was identified with a sensitivity and specificity of >90% from healthy controls and chronic pancreatitis.¹¹⁹ In metastatic PDAC, more than twofold down-regulation of miRNA-205 was seen contrary to nonmetastatic disease.⁶³

Diagnostic kits were developed that detect aberrant expression of miRNAs to discriminate malignant from benign pancreatic lesions. One such kit is the miRInform Pancreas (Asuragen, Inc. Austin, TX), which uses miR-196a and miR-217 to differentiate PDAC from other benign conditions with sensitivity and specificity of 95%.¹²⁰ Another kit developed uses miR-21-5p, miR-375, miR-485-3p, and miR-708-5p to differentiate PDAC from intraductal papillary mucinous neoplasms with sensitivity of 95% and specificity of 85%.¹²¹ One of the main challenges in fully exploiting the miRNAs for diagnostic purposes is that, despite their differential expression in cancerous tissues, their mechanistic role in the molecular carcinogenesis is not completely well understood. In addition, because miRNAs control the expression of >60% of the normal protein-coding genes in humans, they may lack the specificity to the oncogenesis in question.¹⁵

miRNAs as Prognostic Biomarkers

Poor survival outcomes have been reported in PDAC patients with lower expression of miR-183 and miR-34a as well as high expression of miR-1290, miR-155, miR-203, miR-222, and miR-10b.118, 122 In a retrospective study of patients who underwent pancreatic resection for PDAC and received adjuvant gemcitabine treatment, the overexpression of miRNA-142-5p and miR-21 were shown to be an independent prognostic marker for longer overall survival.116, 123, 124

From analysis of The Cancer Genome Atlas, 494 miRNAs were evaluated for their impact on overall survival of patients. The top five miRNAs whose overexpression was associated with better overall survival were miR-1301, miR-125a, miR-376c, mi-328, and miR-376b. However, the underlying molecular mechanisms have not been fully investigated.¹²⁵ A differential expression of hsa-miR-as93, hsa-miR-203, hsa-miR-424, hsa-miR-1266, and hsa-miR-4772 was shown to correlate with overall survival in PDAC using The Cancer Genome Atlas data. This was also validated in newly diagnosed patients with PDAC. The target genes for these miRNAs are suggested to involve various pathways, including TGF-β, phosphatidylinositol 3-kinase–Akt, and stem cell signaling pathways.¹²⁶

Tumor analysis of >100 Korean patients with pancreatic ductal adenocarcinoma using microarray analysis yielded 19 miRNAs that can be used as a classifier to predict prognosis. Some these included the following: miRNAs, miR-106b-star, miR-324-3p, and miR-615, which were related to a p53 pathway, whereas miR-324, miR-145-5p, miR-26b-5p, and miR-574-3p were related to a cyclooxygenase (Cox)-2 centered pathway.¹²⁷

The pattern of miRNA expression can also be used as a surrogate for estimation of overall tumor burden and the emergence of resistant clones of cancer cells.¹²⁸ A postoperative decrease in the plasma levels of miR-18a and miR-221 was shown after surgical resection of pancreatic cancer. In a patient with disease recurrence postoperatively, an elevation of miR-18a was shown without a parallel increase in CA19-9.129, 130

Therapeutic Implications of miRNAs

miRNAs and Chemotherapy Resistance in PDAC

Differential expression of miRNA is well described in pancreatic cell lines that are resistant to chemotherapy. Different possible mechanisms of miRNA-related chemotherapy resistance have been postulated. Increased expression of drug efflux pumps by the cancer cell has been described as a mechanism of resistance. The expression of P-glycoprotein, a multidrug efflux pump that confers cancer cell resistance to a broad spectrum of drugs, is modulated by miR-27a and miR-451.¹³¹

Another drug resistance mechanism described with aberrant miRNA expression is evasion of apoptotic signals and cell cycle deregulation. miR-506 enhances chemosensitivity in PDAC by inhibiting cell proliferation and inducing cell cycle arrest at the G₁/S transition. miR-506 down-regulates SPHK1, which participates in Akt/NF-kB–dependent apoptosis, inducing chemoresistance. Reduced expression of miR-506 and overexpression of SPHK1 correlate with poor survival outcomes in PDAC.¹³²

Inhibiting miR-221/222 induces G₁/S arrest and suppresses the proliferation of pancreatic cancer cells, followed by induction of resistance to gemcitabine.¹⁹ The overexpression of antiapoptotic proteins, such as Bcl-2 proteins, induced by miR-21, can confer resistance to chemotherapy, as shown in pancreatic cancer cells.²⁰ miR-301a, which directly controls the expression of the Bim gene (Bcl-2–like protein 11), is overexpressed in PDAC, thereby fostering pancreatic cell proliferation. The reexpression of the Bim gene through depletion of miR-301a leads to reduced pancreatic cancer cell proliferation.¹³³ In in vitro experiments, miR-181b sensitizes pancreatic cancer cells to gemcitabine by down-regulating Bcl-2.¹³⁴ Restoration of miR-34 in pancreatic cancer MiaPaCa2 and BxPC3 cells inhibits the expression of Bcl-2/Notch1/Notch2, sensitizing them to chemotherapy and inducing apoptosis.¹⁸ In pancreatic cell lines, miR-320a increased resistance to fluorouracil through its effects on programmed cell death 4.¹³⁵

Down-regulation of miRNA-200 family expression was seen in gemcitabine-resistant pancreatic cancer cells.²⁴ It was reported that miR-365 induced gemcitabine resistance in pancreatic cancer cells by directly targeting Src homology 2 domain containing 1 (SHC1) and apoptosis promoting protein BAX. The knockdown of both SHC1 and BAX was also shown to increase resistance to gemcitabine.²¹ An in vitro gemcitabine sensitivity was induced by the knockdown of miR-1246 expression in gemcitabine-resistant pancreatic cancer cell lines. This was believed to be due to restoration of cyclin G-2 protein expression, a tumor suppressor gene. A high expression of miR-1246 was associated with low expression of cyclin G-2 protein and poor overall survival in patients.⁶⁶

The transcription factor–activating protein 2 γ is negatively regulated by miR-10a-5p. The overexpression of miR-10a-5p (low levels of transcription factor–activating protein 2 γ) is seen in gemcitabine-resistant PDAC cells and enhances metastatic behavior and gemcitabine resistance in vitro and in vivo. The overexpression of transcription factor–activating protein 2 γ resensitizes PDAC cells to gemcitabine.¹³⁶

The expression of 13 miRNAs that target genes enriched for EMT predicts response to epidermal growth factor receptor inhibition in cancer cell lines and tumors, with distinction between the primary and metastatic tumors. The ectopic expression of miR-200c alters expression of EMT proteins and possibly improves sensitivity to erlotinib.137, 138

Potential of miRNAs as PDAC Therapeutics

miRNA-based interventions have been viewed as possible therapeutic targets because miRNAs affect multiple gene expression and cellular pathways. It has been hypothesized that they may have an advantage over single gene–directed therapeutic interventions in the design of better cancer treatment. Different approaches have been taken to alter the expression of oncogenic and tumor suppressor miRNAs. Transfecting pancreatic CSCs with an miR-200c mimic and miR-200a reduced their colony formation, invasiveness, and resistance to chemotherapy by regulating epithelial mesenchymal transformation.67, 139 The expression of tubulin beta 3 class III (TUBB3) and p21 activated kinase 1 (Pak-1) was reduced by transfection of pancreatic cells resistant to gemcitabine with miRNA-205 and miR-7, respectively. This decreased the CSC population.⁶³

The administration of micelles of gemcitabine and the tumor suppressor miRNA-205 lead to significant inhibition of tumor growth and increased apoptosis in preclinical models.¹⁴⁰ Targeting miR-21 using lentivirus vector was shown to inhibit cell proliferation.¹⁴¹ The suppression of miR-199a and miR-214 in pancreatic stellate cells dampened their protumor effect and decreased tumor growth.¹⁴² The use of synthetic curcumin analog has been shown to down-regulate miR-21 and restore miR-200 and tumor suppressor PTEN, resulting in depletion of the cancer stem cell population and tumor growth suppression.¹⁴³ Reexpression of the miRNA-200 family, along with reversal of EMT and resensitization of pancreatic cancer cells to gemcitabine, was seen as a result of treatment with isoflavone or 3,31-diindolylmethane.³²

Nanodelivery of Anti-miRNAs

Given the significant roles miRNAs play in the biology of PDAC, attempts have been made to use these tiny RNAs as therapeutics in this deadly disease. Nevertheless, these studies remain mostly restricted to preclinical evaluations in cellular and animal models. A poly (D, l-lactide-coglycolide)–based nanoformulation of miR-150 (miR-150-NF), developed by Arora and colleagues,¹⁴⁴ showed PDAC cellular uptake, engagement with its target gene MUC4, and consequent reduction in cellular proliferation. The i.v. delivery with miR-34a and miR-143/145 nanovectors inhibited the growth of MiaPsCa-2 s.c. xenografts in animal models.¹⁴⁵

Similarly, miR-145–based magnetic nanoparticle formulation showed antitumor potential in cellular models of PDAC.¹⁴⁶ Combination approaches in which anti-miRNA oligo miRNAs are loaded in nanoparticles with chemotherapeutics have also been attempted. For example, combined therapy of miR-21 antisense oligonucleotides and gemcitabine using a targeted codelivery nanoparticle carrier was shown to have synergistic antitumor activity in PDAC cells and animal tumor models.¹⁴⁷ In another study, miRNA-mimic to increase miR-34a together with siRNA to silence PLK1 oncogene were loaded in biodegradable amphiphilic polyglutamate amine polymeric nanocarrier. Systemic administration of amphiphilic polyglutamate amine polymeric nanocarrier–miRNA–siRNA polyplexes to orthotopically inoculated PDAC-bearing mice showed no toxicity and accumulated at the tumor, resulting in an enhanced antitumor effect due to inhibition of MYC oncogene, a common target of both miR-34a and PLK1.¹⁴⁸ Despite this promising preclinical work, it is unclear how these different formulations will work in humans. Therefore, more work is needed to justify the use of nanoparticles as delivery vehicles for noncoding RNAs in the treatment of PDAC.

Clinical Updates

Treatments aimed at RNA interference have been of interest for new drug development recently. RNA interference using siRNAs and miRNA inhibitor has been studied for different indications, including pancreatic ductal adenocarcinoma, hepatitis C virus infection, ophthalmic disease, and amyloidosis in clinical trials (phases 1 through 3). Despite the interest, there are some unanswered questions in the development of RNA interference as a therapeutic strategy, including the ideal target, best route of administration, delivery method, and potential toxicities.¹⁴⁹

Currently, there is no ongoing clinical study targeting miRNAs in the treatment of pancreatic ductal adenocarcinoma reported. There are ongoing studies exploring the diagnostic utilities of different combinations of miRNAs. A few clinical trials aimed at RNA interference in patients with pancreatic ductal adenocarcinoma are summarized below to give an overview of such therapeutic interventions.

There is an open-label phase 1/2a study of siRNA against KRAS (G12D) delivered through an intratumoral slow-releasing device in combination with gemcitabine for locally advanced pancreatic adenocarcinoma. The recommended phase 2 dose was then combined with modified 5-fluorouracil, leucovorin, irinotecan and oxaliplatin (FOLFIRINOX). Of the total 15 patients treated, a third experienced serious adverse events, whereas 90% experienced grade 1 to 2 adverse events. Radiographic stable disease was the best response in 83% of the patients, whereas 17% had partial response. The median overall survival was 15.12 months.¹⁵⁰

There is Atu027, a novel liposomal RNA interference agent including an siRNA, which silences expression of protein kinase N3 in the vascular endothelium and has been shown to inhibit local tumor invasion and lymphatic and visceral metastases in animal models. In a first-in-human study, Atu027 was shown to be safe in patients with advanced solid tumors. Stable disease was seen in 41% of patients for at least 8 weeks.¹⁵¹

Conclusions

Accumulating evidence supports that miRNAs may exert a key role in the pathogenesis of PDAC, by affecting important genetic alterations like KRAS, Tp53, and TGFβ/SMAD, cancer stemness, and support for an unfavorable tumor microenvironment. miRNAs represent an appealing therapeutic target and important biomarkers that may be used in the diagnosis, prediction of treatment response, and prognosis of PDAC. Their ubiquitous presence and the incomplete understanding of their effect on the normal cellular mechanisms and intercellular interactions limit their current clinical applicability and need to be explored further. Despite extensive preclinical work, a practical clinical application is still lacking.