The involvement of lncRNAs in the development and progression of pancreatic cancer

ABSTRACT

Pancreatic cancer is one of the most malignant tumors that are difficult to diagnose at its early stage and there is no effective therapy. Recent studies uncovered that many non-protein-coding RNAs including the class of long noncoding RNAs (lncRNAs) are differentially expressed in various types of tumors and they are potent regulators of tumor progression and metastasis. LncRNA can mediate tumor initiation, proliferation, migration and metastasis through modulating epigenetic modification, alternative splicing, transcription, and protein translation. In this review, we discuss the molecular mechanism of lncRNAs in the involvement of tumor growth, survival, epithelial-mesenchymal transition, tumor microenvironment, cancer stem cells and chemoresistance in pancreatic ductal adenocarcinoma (PDAC).

KEYWORDS:

• Cancer Stem Cells
• Chemoresistance
• EMT
• Hypoxia
• Long-non-coding RNAs (lncRNAs)
• Pancreatic cancer
• Pancreatic Ductal Adenocarcinoma (PDAC)
• Tumor Microenvironment

Introduction

Pancreatic cancer is one of the most devastating human tumors. According to the most recent statistical data, the 5-y survival of pancreatic cancer is 8%, which is very low in contrast to the steady increase in survival in other types of cancers.¹ The mortality rate of pancreatic cancer continued to increase (by 0.3% per year) in men. Recent advances revealed that only <2% transcripts of human genome code for proteins and the remaining 98% transcripts encode different classes of non-coding RNAs.^2,3 Understanding the potential functions of the noncoding transcriptome is of great importance. An important feature of lncRNAs is that their genomic location is transcribed in relation to protein-coding genes. On this basis, lncRNAs can be broadly classified as 1) intronic, where they originate from within coding gene loci; 2) cis-antisense, where they are transcribed from the opposite stand to other transcripts, or 3) intergenic, where they are transcribed outside of, or between, coding genes.⁴ LncRNAs can also be transcribed from other genomic loci including bi-directional regions (either head-to-head or tail-to-tail) with other transcripts,⁵ 3′ from coding genes (uaRNAs)⁶ and from promoter regions (promoter-associated).⁷

Though lncRNAs are generally defined as RNA transcripts longer than 200 nucleotides (nt) with no protein-coding potential, they are involved in diverse biological processes, including cell growth, differentiation and proliferation.^8–10 Accumulating evidence indicates that lncRNAs are frequently aberrantly expressed in cancers.^11,12 Some of them play significant roles in oncogenic or tumor-suppressive pathways and have diagnostic and prognostic values.^13–15

Notably, increasing data suggests that lncRNAs play essential roles in the progression of pancreatic ductal adenocarcinoma (PDAC). For instance, lncRNA HULC serves as a promoting factor in pancreatic cancer development and may become a candidate prognostic biomarker for pancreatic cancer.¹⁶ Expression of gas5 is significantly decreased in pancreatic cancer tissues and its overexpression in pancreatic cancer cells inhibits cell proliferation.¹⁷ Although genomic analyses have discovered a growing list of lncRNAs,¹⁸ the mechanisms of lncRNAs in pancreatic cancer development remain largely unknown. In this review, we address the recent advances in our understanding of the contribution of lncRNAs related to microRNA, EMT, pancreatic cancer stem cells (PCSCs), epigenetics, hypoxia, chemoresistance and tumor microenvironment during pancreatic cancer progression. The function of several important lncRNAs in pancreatic cancer is summarized in Table 1.

Table 1. The biologic functions of lncRNAs in pancreatic cancer (PC).

LncRNA	Expression change in PC	Relevant target genes	Related phenotype of PC	Known molecular mechanisms	Reference
ROR	up	Nanog, p53, ZEB1	cancer pluripotency, epithelial-mesenchymal transition	ceRNA	^24,52
H19	up	HMGA2	epithelial-mesenchymal transition	ceRNA	^28
NEAT1	up	unknown	proliferation,cell cycle arrest and apoptosis	ceRNA	^31
MIR31HG	up	CCND1, Mcl-1, NT5E, KRAS, uPA, ETS1	proliferation, invasion, cell cycle and apoptosis	ceRNA	^33
NUTF2P3–001	up	KRAS	viability, proliferation and invasion	ceRNA	^38
HOTAIR	up	Notch3,miR-34a, miR-663b, GDF15	cancer pluripotency, proliferation, and apoptosis	ceRNA, chromatin modification	^39,77,78,79
HOTTIP	up	HOXA13	epithelial-mesenchymal transition, chemoresistance	unknown	^46
MALAT-1	up	E-cadherin,other EMT-markers, Sox2	epithelial-mesenchymal transition, cancer pluripotency and chemoresistance	chromatin modification	^48,49
LOC389641	up	TNFRSF10A, E-cadherin	epithelial-mesenchymal transition, and invasion	unknown	^51
ENST00000480739	down	OS-9, HIF-1α	invasion and metastasis	unknown	^68
IRAIN	up	KLF2, P15	apoptosis and cell cycle	chromatin modification	^82
PVT1	up	unknown	chemoresistance	unknown	^88
JHDM1D-AS1	up	HGF, FGF1	angiogenesis	unknwon	^109

LncRNAs and microRNA

MicroRNAs (miRNAs) are a small non-coding RNAs shorter than 22 nucleotides that differentially expressed in various types of tumors including pancreatic cancer.¹⁹ It exerts either oncogenic or tumor suppressive roles depending on its downstream targets. Increasing evidence indicates that lncRNAs can function as a miRNA sponge to regulate the expression of specific genes targeted by miRNA which involved in progression of cancer.²⁰ For instance, Qu L et al. identified lncRNA ARSR as a mediator of sunitinib resistance in renal cell carcinoma by acting as a competing endogenous RNA(ceRNA) for miR-34 and miR-449.²¹ Moreover, lncRNA UCA1 contributes to the progression of hepatocellular carcinoma through inhibiting miR-216b and activating FGFR1/ERK signaling pathways.²² Thus, lncRNAs exhibited their regulatory function with different miRNAs.

LincROR which was originally identified in the induction of pluripotent stem cells (iPS), functions to maintain stem cell pluripotency.²³ In the study of Song G et al.,²⁴ they observed that ROR was substantially overexpressed in pancreatic cancer tissues and acted as an endogenous sponge of miR-145, thereby activating the core transcription factor Nanog. Consequently, the up-regulated Nanog significantly increased the tumourigenicity of pancreatic cancer stem cells (PCSCs).

Another example, H19, as an imprinted gene, is highly expressed from the early stages of embryogenesis to fetal life in many organs. However, H19 barely expressed after birth.²⁵ Emerging evidences have shown that H19 was overexpressed in many types of cancers including bladder and breast cancer.^26,27 It was reported²⁸ that up-regulated expression of H19 in pancreatic cancer promoted cell invasion and metastasis partially by increasing HMGA2-mediated epithelial-mesenchymal transition (EMT) through sponging let-7(a miRNA). This work provided a positive correlation between H19 and HMGA2 and shed a light on the potential therapeutic target in pancreatic cancer.

NEAT1(nuclear enriched abundant transcript 1), a newly identified nuclear-restricted lncRNA, acts as a transcriptional regulator for numerous genes.²⁹ Previous study showed that NEAT1 dysregulation promoted tumorigenesis in a variety of human cancers.³⁰ Huang B et al.³¹ observed that NEAT1 could promote pancreatic cancer progression through binding to miR-506–3p, a tumor suppressor reported in hepatocellular carcinoma.³²

Moreover, in the study of Yang H et al.[33], they proved that the lncRNA, MIR31HG, was upregulated in PDAC through microarray gene-expression profile and qRT–PCR assay. Subsequently, they showed that the tumor suppressor miR-193b^34,35 negatively regulated MIR31HG through binding to target sequence, a mechanism similar to that of miRNA-mediated silencing of target mRNAs. In addition, MIR31HG also functioned as endogenous decoy for miR-193b to affect its distribution on specific targets without inducing miR-193b destabilization, as the level of miR-193b was not affected following MIR31HG knockdown or overexpression. Therefore, effect of MIR31HG on PDAC cell proliferation and invasion could be explained, in part by its function as a molecular sponge of miR-193b.

KRAS gene mutation was found at high rates in pancreatic carcinoma, and KRAS was commonly considered as an oncogene in pancreatic cancer tumorigenesis.³⁶ In addition to the mutation of KRAS, some studies also demonstrated that KRAS overexpression was related to pancreatic tumorigenesis.³⁷ Xiang L and colleagues³⁸ found that lncRNA NUTF2P3–001 was significantly increased in pancreatic cancer and it up-regulated KRAS expression by depriving the inhibition of miR-3923 on KRAS. Similarly, another study³⁹ found that the HOTAIR-miR-613-notch3 axis may be a promising therapeutic target for pancreatic cancer. Taken together, lncRNAs could affect development and progression of pancreatic cancer through lncRNA-miRNA interaction.

LncRNAs and epithelial to mesenchymal transition

Epithelial to mesenchymal transition (EMT) is a developmental process during which polarized epithelial cells loss tight intercellular adhesions and acquire a mesenchymal cell phenotype with the capacity to spread to distant locations.⁴⁰ In recent years, EMT was considered playing a critical role in the progression of cancers. EMT documented by the loss of E-cadherin expression and the acquisition of mesenchymal markers, such as N-cadherin or vimentin, led to increased cell migratory and invasive properties.^41,42 Recent studies have indicated that lncRNAs can regulate cancer epithelial to mesenchymal transition. It was shown that inhibiting the expression of linc-ROR could not only increase the sensitivity of cancer cells to docetaxel and reduce the capacity of cancer cell proliferation, but also reverse EMT process in docetaxel resistant lung adenocarcinoma (LAD) cells.⁴³ Another study showed that LncRNA T-Cell Factor-7 (LncTCF7) was highly expressed in HCC and mediated IL6-induced EMT causing cancer cell invasion.⁴⁴

The HOTTIP lncRNA which located at the 5′ end of the HOXA cluster, was recently functionally characterized. HOTTIP significantly expressed in human fibroblasts.⁴⁵ In 2015, Zhihua L et al. identified that HOTTIP contributed to EMT in pancreatic cancer cell, and HOTTIP achieved its the regulatory function in PDAC, by controlling HOXA13.⁴⁶

Metastasis-associated lung adenocarcinoma transcript-1 (MALAT-1), an evolutionarily highly conserved lncRNA with a length of 8,700 nucleotides, has been linked to several human tumor types since its discovery in non-small cell lung cancer.⁴⁷ Jiao F et al. demonstrated that the expression of MALAT-1 was higher in pancreatic cancer cells compared to human pancreatic ductal epithelial (HPDE) cells. Suppressing MALAT-1 decreased expression of genes MMP-2 and MMP-9. Suppressing MALAT-1 also down-regulated EMT biomarkers including Snail, Slug, N-cadherin, and vimentin, and significantly upregulated the expression of E-cadherin. Furthermore, morphological change was observed in CFPAC-1 after MALAT-1 down-regulation.⁴⁸ Another work indicated that EZH2 could be recruited by MALAT-1 and increased H3K27me3 at the E-cadherin promoter, thereby suppressing E-cadherin expression and leading to a malignant phenotype with enhanced tumor migration and invasion.⁴⁹ EZH2, which functions predominately as a transcriptional repressor that silences tumor suppressor gene, has been found to mediate E-cadherin repression in tongue squamous cell carcinoma.⁵⁰

Shangyou Z et al. utilized microarrays to explore the expression profiles of mRNAs and lncRNAs in PDAC(GSE61166). Numerous lncRNAs differentially expressed in PDAC and non-tumorous tissues. Among them a novel lncRNA LOC389641 was remarkably increased in PDAC tissues. Additionally, LOC389641 suppressed the expression of EMT-related epithelial marker E-cadherin and induce the mesenchymal markers Vimentin and Snail. The loss of E-cadherin expression directly triggered EMT. The underlying mechanism of LOC389641-suppressed E-cadherin might be related to TNFRSF10A.⁵¹

Despite linc-ROR functions as a microRNA sponge to influence progression of PDAC, Han-xiang Z and colleagues proposed that linc-ROR-induced EMT in pancreatic cancer cells may be partly through activation of ZEB1 and inhibition of p53 expression.^24,52

Based upon, the LncRNAs have shown significant function in epithelial to mesenchymal transition in PDAC. As most of the pancreatic cancer have metastasized to other organs or locally progressed when diagnosed, the LncRNAs may predict the early metastasis and help making therapeutic decision in the future.

LncRNAs and pancreatic cancer stem cells (PCSCs)

Recently it was demonstrated that the presence of a minor population of cells, termed cancer stem cells (CSCs), are capable of self-renewing, differentiating into other cell types that associated with tumor initiation, metastasis, chemoresistance and recurrence.⁵³ Studies have shown that lncRNAs play pivotal roles in mediating the pluripotency of the embryonic stem cells and differentiation of cancer stem cells.^54,55 Peng F et al.⁵⁶ found that ectopic overexpression of lncRNA H19 significantly promote breast cancer cell colony formation, migration and sphere-forming ability. Furthermore, it was reported that lncRNA RBM5-AS1 contributed to the stem-like characteristics of colon cancer cells in the Wnt pathway via physical interactions with β-catenin.⁵⁷

Pancreatic cancer stem cells (PCSCs) have been identified not only based on the expression of CD24, CD44, and epithelial-specific antigen (ESA), but also several other cell-surface markers, such as ALDH, CD133.^58,59,60 Li et al.⁶¹ observed that CD44⁺CD24⁺ESA⁺ cells had a higher tumorigenic ability in immunodeficient mice than PDAC cells that had those markers negative.

In a previous study of Jiao F⁴⁸, it was found that induced expression of MALAT-1 served as an oncogenic lncRNA in pancreatic cancer by upregulating expression of CSCs markers. However, the detailed molecular mechanisms of MALAT-1 regulating pancreatic CSCs and tumor progression remained unknown. MALAT-1 upregulated the expression of a self-renewal factor Sox2 to give rise to the stem cell-like phenotypes of pancreatic cancer in PDAC cells.⁶² It was revealed that high level of MALAL-1 was expressed in CD133+ cell and CD133 expression was reduced after MALAT-1 downregulation. However, the detailed molecular mechanisms of MALAT-1 upregulating Sox2 still need to be clarified. On the contrary, other researchers documented that Linc-ROR can function as a competing endogenous RNA to sponge miR-145, leading to an increase in expression of a miR-145 target, the transcription factor Nanog, which effectively induce the malignant tumor characteristics of PCSCs.²⁴ Collectively, the above data suggested that dysregulating lncRNAs in pancreatic cancer is related to its stem cell-like properties.

LncRNAs and Hypoxia

The presence of intratumoral hypoxia is a negative prognostic indicator for many patients as it has been associated with aggressive tumor cell phenotypes and distant metastasis.⁶³ High level of hypoxia often occurred in pancreatic cancers.⁶⁴ Hypoxia-inducible factor- 1α (HIF-1α) is the best characterized biomarker among the transcriptional regulators under hypoxia.⁶⁵ In addition, the overexpression of HIF-1α was reported to play a critical role in pancreatic cancer under hypoxia and promoted invasion and metastasis.⁶⁶ LncRNAs have been reported to function in diverse regulatory mechanisms in the hypoxia process. Sun YW et al. found that hyaluronan synthase 2 antisense 1 (HAS2-AS1) increased in oral squamous cell carcinoma and in cells cultured under hypoxia. Moreover, the expression of hypoxia-induced HAS2-AS1 is dependent on HIF-1α which directly binds to and activates the transcription of HAS2-AS1.⁶⁷

Recent research showed⁶⁸ that ENST00000480739 functioned as an important anti-metastatic lncRNA, which downregulated HIF-1α by upregulating OS-9 and served as an independent risk factor for the overall survival of pancreatic cancer patients (Fig. 1) after surgery. Meanwhile, HIF-1α could transcriptionally upregulated lncRNA-NUTF2P3–001 by binding to hypoxia-response elements (HRE) of promoter area, which significantly promoted proliferation and invasive ability of pancreatic cancer.³⁸ However, more explorations are needed to demonstrate how lncRNAs participate in this process in pancreatic cancer.

Figure 1. The molecular mechanism of lncRNAs in the progression of pancreatic cancer. A. lncRNAs bind to specific microRNAs to activate the expression of targeted genes which involve in carcinoma progression, B. MALAT-1 promotes stem cell-like phenotypes of pancreatic cancer through upregulating self-renewal factor Sox2 and induces EMT. C. HOTTIP contributes to EMT through upregulating HOXA13 in pancreatic cancer cells, D. ROR induces EMT through inhibiting p53 and ZEB1 expression, E. LOC389641 triggers EMT through suppressing E-cadherin expression, F. ENST00000480739 downregulates HIF-1 by upregulating OS-9 resulted in facilitating pancreatic cancer invasion and metastasis, G. HIF-1α transcriptionally upregulates lncRNA-NUTF2P3–001 under hypoxia which promotes invasive ability of pancreatic cancer.

lncRNA-mediated modulation of epigenetic modifications

lncRNAs participated in genomic transcriptional control through association with epigenetic modifier complexes.⁶⁹ Polycomb repressive complex 2 (PRC2) is one of the main histone modification complexes tethered by lncRNAs since a substantial fraction of lncRNAs was found to interact with this complex.^70,71 Enhancer of zeste homolog 2 (EZH2) acts as a catalytic subunit of PRC2, which is overexpressed in a wide range of tumors and functions predominately as a transcriptional repressor that silences tumor suppressor gene targets via methylation of lysine 27 of histone 3 (H3K27).^72–74 Accumulating evidence indicated that long noncoding RNA contributed to the epigenetic regulation of tumor initiation. Terashima M and his coworkers revealed that MEG3 regulated the recruitment of JARID2 and EZH2 and histone H3 methylation on the regulatory regions of CDH1 and microRNA-200 family genes for transcriptional repression in lung cancer cells.⁷⁵ Another study showed that lncRNA HOXA11-AS can also function as a scaffold to recruit EZH2 along with the histone demethylase LSD1 or DNMT1 to serve as critical effectors in tumorigenesis and progression of gastric cancer.

Similarity, EZH2 was overexpressed in pancreatic cancer promoted tumor cell proliferation, migration and invasion via epigenetic repression of target genes. Intriguingly, it was reported that EZH2 recruitment by MALAT-1 played a key role in the pathogenesis of pancreatic cancer through subsequent involvement of the complex in repressing E-cadherin. In addition, RNA immunoprecipitation (RIP) assay confirmed that EZH2 directly bound to MALAT-1.⁴⁹

Another example, the first HOX-associated lncRNA, HOTAIR, was initially identified as a scaffold RNA associated with histone modification complexes, namely polycomb repressive complex (PRC2) and the LSD1/CoREST/REST complex.⁷⁶ Chi-Han L and colleagues⁷⁷ demonstrated that HOTAIR could guide EZH2 to affect the level of the tumor suppressor miR-34a in PDAC cells, thereby promoting cancer cell proliferation. Apart from the conventional repressing mechanism of H3K27 trimethylation, EZH2 could also induce heterochromatin formation at the promoter of miR-34a. Similarity, evidences showed that HOTAIR also suppressed another tumor suppressor miR-663b in pancreatic cancer cells by modifying histone methylation on miR-663b promoter.⁷⁸ However, another study⁷⁹ showed that silencing HOTAIR decreased interaction of EZH2 with the growth inhibitory/pro-apoptotic gene GDF15 promoter region in Panc1 and L3.6pL cells, thus activating derepression of GDF15 and promoting apoptosis of tumor cells.

Since there was evidence that HOTTIP interacted with both chromatin-modifying complexes, PRC2 and MLL1/WDR5⁴⁵, Yating C et al.⁸⁰ downregulated HOTTIP and the other two complexes. They performed microarray analysis and found that <40% of all HOTTIP-regulated genes were coregulated by HOTTIP and the two complexes. Though the regulation of gene expression by HOTTIP in pancreatic cancer cells involved interaction with complexes in addition to PRC2 and MLL1/WDR5, the interactions need more explorations.

Meanwhile, studies have found lncRNA IRAIN, which is 5359 nt in length and located in chromosome 15q26.3, may contribute to the progression of non-small cell lung cancer (NSCLC).⁸¹ Also, its overexpression was significant and correlated with advanced pathological stage, larger tumor size and lymph node metastasis in patients with pancreatic cancer.⁸² Yifan L and coworkers found IRAIN can directly bind to EZH2 and LSD1(another important role in the epigenetic regulation of gene transcription) complexes and repress two key tumor suppressors, KLF2 and P15, responsible for cell apoptosis and cell cycle control (Fig. 2).^83,84

Figure 2. LncRNAs contribute to the epigenetic regulation of pancreatic cancer tumorigenesis, A. EZH2(one unit of PRC2 complex) recruited by MALAT-1 and suppresses E-cadherin expression through increasing H3K27me3 at the E-cadherin promoter, B. HOTAIR increases interaction of EZH2 with GDF15 promoter region, C. IRAIN binds to EZH2 and LSD1 complexes and repress two key tumor suppressors, KLF2 and P15. D. HOTAIR guides EZH2 to target the tumor suppressor miR-34a, E. HOTAIR suppresses tumor suppressor miR-663b in pancreatic cancer cells by LSD1 and PRC2 modification on H3K4me3 or H3K27me3 of miR-663b promoter.

Although lncRNAs showed a critical role in epigenetic modulation of pancreatic cancer, more complex cases of epigenetic regulation need to be clarified. For instance, the radiation-induced PARTICLE represses tumor suppressive gene MAT2A in breast cancer cell via forming a DNA-lncRNA triplex within the 456 bp upstream of the CpG island in the MAT2A promoter. Besides, it could also induce methylation of the MAT2A promoter through interaction with transcription-repressive complex proteins G9a and SUZ12 (subunit of PRC2).⁸⁵

LncRNAs and Chemoresistance

Recent evidence indicated a close link between lncRNAs and the acquisition of chemoresistance in various carcinoma cells. It was found that decreased expression of linc-ROR effectively reversed EMT in docetaxel-resistant lung adenocarcinoma cells and sensitized them to chemotherapy.[43] However, lncRNA RP11–838N2.4 could enhances the cytotoxic effects of temozolomide to glioblastomas cells both in vivo and in vitro.⁸⁶

PDAC is an intractable disease as a result of its rapid dissemination and innate or acquired resistance to chemotherapy. Recent findings provide novel information on lncRNAs expression profiles that might be related to chemoresistance in PDAC. Considering the effect of HOTTIP on tumor cell proliferation, cell cycle, and EMT identified above, Zhihua L et al.⁴⁶ further investigated whether down-regulation of HOTTIP affected the resistance of PDAC cells to gemcitabine. Interestingly, SW1990-shHOTTIP knockdown cells exhibited much slower growth and a lower IC50 for gemcitabine. Moreover, combined treatment with gemcitabine and HOTTIP knockdown led to an even further reduction in tumor volume in vivo.

In a previous study, Feng J et al.⁴⁸ shown that MALAT-1 increased the proportion of pancreatic CSCs. They investigated the anti-proliferative effects of gemcitabine in MALAT-1-nc and MALAT-1-si cells, and found that the resistance index(RI) of gemcitabine in AsPC-1/M-nc and CFPAC-1/M-nc was 2.96 and 4.29 times higher than that of M-si group, respectively.⁶² Nevertheless, the increased chemosensitivity may be a reflection of the slower growth rate of the MALAT-1 knockdown. Resistance to chemotherapy is a property that distinguished pancreatic CSCs from other cancer cells.⁸⁷

Meanwhile, through a piggyBac transposon-based genome-wide mutagenesis strategy, Lei Y et al.⁸⁸ identified that functional inactivation of gene PVT1 led to increased sensitivity of gemcitabine in human pancreatic ASPC-1 cells. PVT1, a long non-coding RNA (lncRNA), was observed to play a key role in tumorigenesis through DNA rearrangements or amplifications in non-small cell lung cancer.⁸⁹ Furthermore, PVT1 was expressed significantly higher in gastric cancer patients and might be a novel biomarker for predicting gastric cancer.⁹⁰ Xian-wen Z et al.⁹¹ concluded that up-regulation of PVT1 promoted the development of multidrug resistance in gastric cancer via increasing the expression of multidrug resistant related gene (MDR1, MRP, mTOR and HIF-1a). However, the exact mechanism of PVT1 on the pancreatic cancer chemoresistance needs further study.

Although comprehensive therapies have been used for the treatment of pancreatic cancer, it is urgent to identify reliable biomarkers for predicting therapeutic efficacy and for choosing individualized therapeutic strategies in patients with pancreatic cancer.

LncRNAs and tumor microenvironment

The complex microenvironment surrounding tumor cells includes inflammatory cells, fibroblasts, extracellular matrix, and small blood and lymphatic vessels. Emerging evidence supported that tumor stroma or tumor microenvironment influences the aggressiveness and drug resistance of cancer cells, thus accelerating the progression of carcinomas.^92–94

The desmoplastic stroma of pancreatic cancer results from the proliferation and differentiation of fibroblasts. The increase in fibrosis within the stroma associates with invasiveness and poor prognosis.^95–97 Cancer-associated fibroblast (CAF), which is one of the major components in the tumor stroma, plays a critical role in tumor growth and proliferation.⁹⁸ A new data in bladder cancer showed that lncRNA ZEB2NAT was upregulated by CAF-CM(Cancer-associated fibroblast condition medium) through secreted TGFβ1, and the knockdown of ZEB2NAT reversed CAF-CM induced ZEB2 protein level, which was a major transcription factor involved in EMT.^99,100 In another important study, CAF-secreted CXCL14 caused lncRNA LINC00092 expression alteration in ovarian cancer cells to promote cancer progression.¹⁰¹ LINC00092 modulated glycolysis to promote ovarian cancer metastasis through directly interaction with PFKFB2, which was a key enzyme of glycolysis in cancer.¹⁰² More interestingly, PFKFB2-induced glycolytic phenotype of ovarian cancer cells sustained the CAFs-like features of fibroblasts in turn, indicating a reciprocal feedback loop between CXCL14-positive CAFs and ovarian cancer cells. As for pancreatic cancer, evidence provided that expression of SHH(Sonic Hedgehog) or TGFβ which derived from cancer cells may contribute to the formation of desmoplasia that impeded proper drug delivery.^103,104

Tumor neovascularization is a another critical tumor-supportive element for the stroma, where tumor cells release proangiogenic signals to recruit endothelial cells to form the neovasculature, thereby obtaining nutrients and oxygen or evacuating metabolic wastes during tumor growth and metastasis.^105,106 lncRNAs have been evidenced to regulate angiogenic process recently. For example, Jia P and coworkers uncovered that H19 was up-regulated in microvessels from glioma tissues and over-expression of H19 promoted glioma-associated endothelial cell proliferation, migration and tube formation.¹⁰⁷ Other research found that the upregulated H19 CD90+ liver cancer cells could modulate endothelial cell phenotype through the release of exosomes, resulting in angiogenic phenotype.¹⁰⁸ Remarkably, a xenograft study revealed¹⁰⁹ that nutrient starvation-responsive lncRNA JHDM1D-AS1 overexpression increased pancreatic cancer growth in vivo, accompanied by elevated blood vessel formation and macrophage infiltration. Through investigating the expression of major pro- and anti-angiogenic factors by quantitative real-time PCR, they found that several pro-angiogenic factors such as HGF and FGF1 were significantly up-regulated in tumor xenografts comprising hJHDM1D-AS1-expressing PC cells compared to xenografts comprising of control cells. Although overexpression of JHDM1D-AS1 had minor effects on cell growth in vitro, the significant increase of tumor growth in vivo demonstrated that JHDM1D-AS1 affects the tumor microenvironment rather than the cancer cells themselves. These results provided us a new insight into the therapeutic approach for pancreatic cancer in the future.

It is proposed that the tumor microenvironment with selective inflammatory immune cells play important roles in promoting the malignant progression of many types of carcinomas.^110,111 In addition, tumor immune evasion is a hallmark of cancer progression.¹¹² There are many immunosuppressive mechanisms in the tumor microenvironment, programmed death receptor(PD-1) is found to be a major factor related to tumor immunity.^113,114 PD-1 is mainly expressed in tumor infiltrating lymphocytes (TILs) which inactivates tumor-reactive CTLs that directly target tumor cells for apoptosis.^115,116 A study demonstrated that the expression of lncRNA AFAP1-AS1 was positively correlated with PD-1 in nasopharyngeal carcinoma(NPC), and the co-expression of AFAP1-AS1 and PD-1 predicted high incidence of recurrence or metastasis among patients. It was reported that AFAP1-AS1 may indirectly regulate PD-1 expression and the AFAP1-AS1-PD-1 might be ideal candidates for future clinical trials of anti-PD-1 therapy.¹¹⁷ In addition, long non-coding RNA AFAP1-AS1 could predict pancreatic cancer progression within 6 months and 1 y when the areas under receiver operating characteristic (ROC) curves of AFAP1-AS1 were 0.8669 and 0.9370 respectively.¹¹⁸ Although researchers have reported that PD-1 was expressed in 51.2% to 52.1% of pancreatic tumor–infiltrating cytotoxic T lymphocytes,¹¹⁹ the correlations between the PD-1 and AFAP1-AS1 or other tumor associated-lncRNAs in PDAC warranted in-depth study in the future.

Conclusion

Pancreatic cancer is known to have a poor prognosis, mostly due to lack of effective diagnosis at the early stage of tumor development and effective therapy. Although surgical resection of pancreatic cancers diagnosed at early stage remains the potentially curative option for PDAC, it is only possible in a small subset of patients.

More and more lncRNAs with differential expression in PDAC are discovered and most of them are major regulators in the multi-step process of cancer development from tumor initiation to metastasis. Some prominent functions of lncRNAs (such as HOTAIR, HOTTIP, MALAT1, and ROR) are associated with clinicopathologic characteristics of pancreatic cancer and may be potential tumor biomarkers of pancreatic cancer in the future. In-depth investigation on the molecular mechanisms of lncRNA activity is required to better understand the diverse function of lncRNAs. Tumor microenvironment is a driving force in pancreatic cancer progression. It is critical to elucidate the emerging pathophysiologic roles of lncRNA in the crosstalk between pancreatic cancer cells and tumor stroma. We believe that understanding the roles of lncRNAs would provide novel perspectives for developing new therapeutic agents and diagnostic biomarkers.