Background
Pancreatic cancer (PC) is a frequent cause of cancer death worldwide [
1]. While advances in clinical treatments, including chemotherapy and surgery, have improved the prognosis of PC in the past decades, the early detection of PC remains quite difficult. Thus, the prognosis of PC remains poor, even when using advanced imaging techniques such as computed tomography or positron emission tomography. Carbohydrate antigen 19–9 (CA19–9) is the most sensitive diagnostic marker for PC, but it is not useful for diagnosing early PC [
2]. Therefore, new biomarkers that can detect the initial stages of PC are urgently needed.
Intraductal papillary mucinous neoplasm (IPMN) is a pre-cancerous lesion, and 1–2% of IPMN cases progress to PC each year [
3]. IPMN progress from a non-invasive to an invasive lesion [
4], and the postoperative prognosis of patients with invasive IPMN appears to be considerably worse than that of patients with non-invasive IPMN [
5]. This evidence suggests that biological markers able to distinguish invasive IPMN from non-invasive IPMN can improve the survival of PC patients. While the utility of CA19–9 and MUC5AC as serum markers of malignant IPMN has been reported, their sensitivities were not high enough to be indicative factors for resection [
6]. Indeed, even when novel imaging procedures are utilized, it is difficult to predict the malignant potential of IPMN [
4]. Novel indicators that can predict the malignant potential of IPMN are therefore eagerly awaited.
MicroRNAs (miRs), which are small RNAs that regulate approximately 30% of human genes [
7], are secreted into the blood and body fluids [
8]. Recent studies have shown that the abnormal expression of extracellular circulating miRs (CmiRs) in serum or plasma was correlated with the prognosis of PC, suggesting that CmiRs may be potential diagnostic or prognostic markers for advanced PC [
9].
miR-21 was reported that proportionally increased during the progression from IPMN to PC, but no other miRs have been identified as markers for the detection of IPMN as well as the progression of PC.
miRs have been reported to be stably contained within vesicles called exosomes [
10]. Exosomes are small (40–100 nm diameter) vesicles composed of a lipid bilayer and secreted by cells to interact with distant tissues; they may be found in all body fluids, including the serum and plasma [
10‐
12]. miRs and mRNAs were found to be enclosed in exosomes, stabilized from RNase and highly enriched compared to the serum [
11,
12], and the expressions of these exosomal microRNAs (ExmiRs) were dysregulated in several types of cancer patients [
10]. ExmiRs are therefore expected to be useful as non-invasive diagnostic biomarkers in cancer patients.
We herein assessed for the first time the expression of ExmiRs in patients with IPMN and PC using a next-generation sequencing analysis, and revealed that three ExmiRs were upregulated in IPMN and PC. In addition, the expression of these ExmiRs was correlated with poor prognosis in PC patients and the high-risk cases in the IPMN group, respectively.
Methods
Patients
Thirty-two patients with newly diagnosed PC and 29 with IPMN (no prior treatment) at Asahikawa Medical University Hospital from April 2013 to December 2015 were respectively enrolled in the PC group and IPMN group in this study. Twenty-two patients without malignant or neoplastic lesions were registered in the control group; these patients were recruited from patients who visited the Division of Gastroenterology and Hematology/Oncology in Asahikawa Medical University during the study period. The characteristics of the patients in the control group are shown in Table
1. Six cases complaining of abdominal pain and 1 case complaining of nausea were included. Patients with other cancers or neoplasms were excluded from this study. All patients with PC and IPMN underwent enhanced computed tomography from the chest to the abdominal region for tumor staging, according to either the TNM criteria or the IPMN guidelines. Informed consent was obtained from all of the participants regarding the use of their blood samples in this study. The study was approved by the Medical Ethics Committee of Asahikawa Medical University.
Table 1
Characteristics of the control, IPMN, and PC groups
Sex, n (%) | Female | 8 (36.4) | 16 (55.2) | 15 (46.9) |
Male | 14 (63.6) | 13 (44.8) | 17 (53.1) |
Age (mean ± SD) | 57.5 ± 15.3 | 73.8 ± 7.8 | 64.0 ± 10.1 |
Stage (UICC) I / Ila / llb / III / IV | – | – | 2 / 7 / 4 / 5 / 14 |
Fukuoka criteria FN / WF / HRS | – | 14 / 11 / 4 | – |
Clinical information | GBP 4 | | |
Chronic gastritis 3 |
Gallbladder stone 2 |
ADM 2 |
Liver cyst 1 |
IBS 1 |
Accessory spleen 1 |
Only symptom 7 |
Serum samples
A blood examination and sampling were performed before treatment, which included surgery, chemotherapy, and radiotherapy. The peripheral blood from patients was collected and then centrifuged at 5000 rpm (rpm) for 10 min at 4 °C. The serums were then transferred to fresh tubes and stored at − 80 °C. Before analysis, the serum samples were filtrated through a 0.45-μm pore membrane (Millipore, Billerica, MA, USA). The amount of serum used in all of this study was unified in 250 μl according to the Manufacture.
Isolation of the exosomes from the serum and MicroRNA isolation from the exosomes
Exosomes were collected from the serum using ExoQuick Exosome Precipitation Solution (System Biosciences, Mountain View, CA, USA) in accordance with the manufacturer’s instructions. Exosomal RNAs were isolated by using Trizol (Invitrogen, Grand Island, NY, USA) and purified using a mirVana miRNA isolation kit (Life Technologies, Carlsbad, CA, USA). The purity and concentration of all RNA samples were quantified spectrophotometrically using the NanoDrop ND-1000 system (NanoDrop, Wilmington, DE, USA). Exosomes were quantified using a CD63 ExoELISA kit (System Biosciences) in accordance with the manufacturer’s instructions.
Selection of MicroRNA in the exosome using a next-generation sequencer
Five patients were randomly selected from each groups to examine the expression of their exosomal miR. The volumes of the RNA samples (collected from 250-μl serum samples) was normalized. RNA libraries were generated using an Ion Total RNA-Seq Kit v2 (Life Technologies) in accordance with the manufacturer’s instructions. The RNA libraries were then processed for the emulsion PCR using an Ion OneTouchTM system and an Ion OneTouch 200 Template kit v2 (Life Technologies). Template-positive Ion SphereTM particles were enriched and purified for the sequencing reaction with an Ion OneTouchTM ES system (Life Technologies). The template-positive Ion SphereTM Particles were then applied to Ion PI™ Chips (Life Technologies), and the next-generation sequencing reaction was carried out using an Ion Proton™ Semiconductor sequencer (Life Technologies). All of the sequencing data were mapped on a miR sequence using the CLC Genomics Workbench software program (CLC Bio, Aarhus, Denmark), and an expression analysis was performed for each sample.
MicroRNA detection by quantitative real-time polymerase chain reaction
miRs were reverse-transcribed, and their expressions were determined by quantitative real-time polymerase chain reaction (qRT-PCR) using TaqMan microRNA assay kits in accordance with the manufacturer’s instructions (Applied Biosystems, Foster City, CA, USA). The Ct values were used in the analysis of the qRT-PCR data.
Statistical analysis
The expression of miR and CD63 in serum samples was compared using the Mann-Whitney U test (for two groups) or the Kruskal-Wallis test followed by Dunn’s test (for three groups). There was no adjustment for multiple comparisons in the subgroup or multiple miRs analysis. The diagnostic performance was confirmed by Receiver Operating Characteristic (ROC) curve analysis. The cutoff point was determined by the following formula: Distance = (1-sensitivity)2 + (1-specificity)2.
In survival analyses, the probability of overall survival (OS) was determined by the Kaplan-Meier method with a log-rank test and Cox’s proportional-hazards regression model. The statistical analysis was performed using the Graph Pad PRISM (Version 5.0a; GraphPad Software, Inc. La Jolla, CA, USA), SPSS and R software programs. The level of significance was set at p < 0.05.
Discussion
The present study analyzed for the first time the serum ExmiRs in PC and IPMN patients using a next-generation sequencing, resulting that ExmiR-191, ExmiR-21, and ExmiR-451a were identified as candidate miRs which are dysregulated in IPMN and PC patients. The qRT-PCR confirmed that the expressions of ExmiR-191, ExmiR-21, and ExmiR-451a were increased in the patients with PC and IPMN.
Previous reports have suggested that CmiRs are useful for detecting or determining the prognosis of PC, invasive IPMNs, and other cancers [
12,
13]. In the present study, we showed that the expressions of ExmiR-191, ExmiR-21 and ExmiR-451a were significantly up-regulated in PC and IPMN. However, of note: the expressions of CmiR-191, CmiR-21, and CmiR-451a were not markedly changed between the control, IPMN, and PC groups, illustrating the utility of ExmiRs as detection makers of PC and IPMN over CmiRs. CmiRs have been reported to be stabilized in vesicles such as exosomes [
14], and the exosomes in serum are highly enriched in miRs [
11]. Tanaka et al. also showed that circulating miR-21 originated from exosomes, as the miR-21 expression was significantly higher in exosomes than in the serum remaining after exosome extraction [
12]. These present and previous findings therefore suggest that ExmiRs are more useful as markers for tumor detection than CmiRs.
It should be noted that the current established tumor markers were elevated in the advanced cancers, but not in IPMN, while the ExmiRs were upregulated in both IPMN and PC, including both early and advanced phases. The diagnostic performance estimated by the ROC curve analysis favorable AUC and accuracy as compared to CEA and CA19–9 in IPMN and early stage of PC, suggesting the levels of ExmiR-191, ExmiR-21, and ExmiR-451a can thus serve as early diagnostic markers of pancreatic neoplasms.
miR-191 has been reported to be up-regulated in a wide range of human cancers, including PC [
15]. miR-191 might be responsible for the abnormal expression of many target genes such as CDK9, NOTCH2, and RPS6KA3 [
16], as these genes have been reported to be direct targets of miR-191 and regulators of proliferation. In addition, miR-191 was found to regulate cell invasion and differentiation, facilitate extracellular matrix formation, and encourage metastasis [
15]. miR-191 was also found to up-regulate p53 deletion and inhibit the expression of the tumor-suppressive mRNA, C/EBPβ, thereby enhancing the tumor progression in colorectal cancer [
17]. Taken together, the findings from these previous studies show the oncogenic features of miR-191, which supports our finding that ExmiR-191 is a candidate diagnostic marker of pancreatic neoplasms.
miR-21 is also considered an oncogenic miR because it is up-regulated in various cancers and targets the tumor-suppressive mRNAs [
18]. Previous reports of basic studies have stated that overexpression of miR-21 promoted cellular proliferation, survival, and invasion and migration of cancer cells, including PC cells [
19,
20]. The overexpression of miR-21 down-regulated the expression of tumor suppressors such as PDCD4 and TIMP3 and promoted cell proliferation, leading to the development of PC [
21,
22]. Conversely, the suppression of miR-21 reduced cancer cell survival and tumor growth in a murine xenograft model [
23]. Indeed, clinical studies have shown that miR-21 was up-regulated in invasive IPMN compared with noninvasive IPMN [
24] and was up-regulated in noninvasive IPMN compared to normal pancreatic tissue. In the present study, ExmiR-21 was identified as an early diagnostic marker of pancreatic neoplasms, a finding which is consistent with those of previous studies. Recent studies have indeed reported that miR-21 reduced the sensitivity of cancer cells to anticancer drugs such as gemcitabine and 5-FU-based chemotherapy [
25], and the suppression of miR-21 increased sensitivity to gemicitabine and induced apoptosis in PC patients [
26]. Furthermore, the overexpression of miR-21 correlates with a poor prognosis after PC resection, independent of other clinicopathologic factors [
27]. Interestingly, ExmiR-21 was also identified as a chemo-sensitive marker and a prognostic factor for the overall survival in this study. Taken together, these present and previous findings suggest that miR-21 plays a role in pancreatic carcinogenesis.
miR-451a is located on chromosome 17q11.2 in humans and has been reported to suppress cell proliferation and colony formation by targeting the Ywhaz (14–3-3zeta) gene, RAB14 protein, and activating transcription factor 2 (ATF2). It is therefore considered to be a tumor-suppressive miR in several human malignancies [
28,
29]. In addition, a clinical study showed that miR-451a expression was down-regulated in cancer tissue, and low miR-451a expression tends to be associated with metastasis and shorter survival duration [
30]. These findings suggest that miR-451a acts as a tumor suppressor. In our analysis, ExmiR-451a expression was significantly increased in both the IPMN and PC groups. A high expression of ExmiR-451a might be induced by a positive feedback system due to the progression of the pancreatic tumor. Further analyses of the miR-451a expression in IPMN and PC cells will be needed to fully clarify the role of miR-451a in the development of PC.
The present study was associated with some limitations. We investigated biomarkers using serum samples from 32 patients with PC, 29 patients with IPMN and 22 healthy volunteers as a discovery cohort. To validate the efficacy of three ExmiRs, further studies should be conducted using large cohorts. In addition, the present study focused on IPMN and PC, and did not include cases with other pancreatic cystic neoplasms. A further analysis that includes patients with cystic neoplasms as well as those with non-neoplasms should be performed to identify biomarkers that can be used to distinguish IPMN from other cystic lesions.