MiR-497/Pim-1信号通路对胰腺癌恶性表型调控作用及诊断、预后价值研究

【作者】 徐建威；

【导师】 张太平；

【作者基本信息】 北京协和医学院 ， 普通外科学， 2014， 博士

【摘要】 研究背景早期诊断困难、化疗耐药、缺乏有效的预后指标是目前胰腺癌研究领域的重要难题,限制了胰腺癌患者的总体预后。寻找新的诊断及预后标记物,深入了解胰腺癌发生、发展机制,逆转化疗耐药是改善胰腺癌患者预后的重要手段。MicroRNAs (miRNAs)是一类长约22个核苷酸的非编码单链小分子RNA,在肿瘤发生、发展、增殖、侵袭、转移及化疗耐药等多个方面发挥重要作用。我们前期对胰腺癌耐药株和亲本株的miRNAs表达谱进行了高通量筛选及验证,发现miR-497在耐药株中表达显著下调,可能在胰腺癌耐药表型调控中发挥重要作用。MiR-497在结直肠癌、宫颈癌、乳腺癌等肿瘤中表达降低,参与多种肿瘤生物学表型的调控。但是其在胰腺癌中的表达水平、调控作用及机制尚不明确。因此,检测胰腺癌中miR-497的表达水平,深入研究miR-497的调控作用及机制,对于明确其在胰腺癌鉴别诊断、判断预后及逆转化疗耐药等方面的价值具有重要的意义。研究目的探索miR-497对胰腺癌细胞株增殖、凋亡、细胞周期、侵袭及迁移、化疗耐药、干细胞标志物等恶性表型的调控作用。明确miR-497的调控靶基因及信号通路。对其中一个靶基因Pim-1的功能及调控机制进行探索。检测胰腺癌组织及血浆miR-497、Pim-1的表达水平,评估其诊断与预后价值。研究方法通过CCK8法检测细胞增殖及化疗敏感性；流式细胞术检测细胞周期和细胞凋亡；transwell法检测细胞侵袭和迁移能力；实时定量PCR法检测miRNA及mRNA的表达水平；western blot检测蛋白表达水平；建立稳转miR-497的细胞株,裸鼠体内观察miR-497对肿瘤生长的影响；双萤光素酶报告基因实验验证miR-497的靶基因；分别采用原位杂交及免疫组化法检测组织miR-497及Pim-1蛋白的表达水平,实时定量PCR法及酶联免疫吸附测定法(ELISA)检测血浆miR-497及Pim-1蛋白表达水平；单因素及多因素生存分析评估miR-497及Pim-1的预后价值；受试者工作特征曲线(ROC曲线)及约登指数(Youden index)评估血浆miR-497及Pim-1的诊断价值。研究结果1.MiR-497对胰腺癌细胞恶性表型的调控作用在两株胰腺导管腺癌细胞株SW1990及MiaPaCa-2中上调miR-497的表达水平,显著抑制胰腺癌细胞增殖、增加吉西他滨诱导的Cleaved caspase-3及PARP的表达水平、增加吉西他滨及厄洛替尼化疗敏感性、抑制细胞迁移和侵袭能力,减少S期细胞百分比,并抑制干细胞标志物ABCG2和EZH2的表达水平；反之,下调miR-497表达水平,观察到相反的实验结果。体内实验发现,稳定表达miR-497的实验组的肿瘤生长速度显著慢于对照组。(P<0.05)。2.MiR-497调控的靶基因及信号通路通过构建包含种子区的野生型和突变型双萤光素酶报告基因载体,并分别与miR-497mimics或mimics control共转染293A细胞,发现共转染mimics与野生型载体的细胞的荧光素酶活性显著低于共转染mimics与突变型载体,mimics control与野生型载体,及mimics control与突变型载体的细胞(P<0.05),提示IGF-1R、FGF2、FGFR1、Kras及Pim-1是miR-497的直接靶基因。蛋白水平的检测发现,上调miR-497的表达水平显著抑制IGF-1R、FGF2、FGFR1、Kras及Pim-1的表达水平,反之,下调miR-497的表达水平显著上调IGF-1R、FGF2、 FGFR1、Kras及Pim-1的表达水平。对重要信号通路的关键信号分子进行检测,发现上调miR-497显著抑制p-AKT及p-ERK等蛋白分子的表达；下调miR-497显著增加,p-AKT及p-ERK等蛋白分子的表达。(P<0.05)。3.Pim-1对胰腺癌细胞恶性表型的调控作用及机制作为miR-497的一个靶基因,Pim-1的调控作用和机制仍不明确。通过siRNA下调Pim-1的表达水平,显著抑制胰腺癌细胞增殖、促进细胞凋亡、增加吉西他滨及厄洛替尼化疗敏感性及诱导的细胞凋亡、抑制细胞迁移和侵袭能力,减少S期细胞百分比,并抑制干细胞标志物ABCG2和EZH2的表达水平(P<0.05)；通过SGI-1776下调Pim-1的表达水平,显著抑制胰腺癌细胞增殖、促进细胞凋亡、增加吉西他滨及厄洛替尼化疗敏感性及诱导的细胞凋亡、抑制细胞迁移和侵袭能力,减少S期细胞百分比(P<0.05)。对Pim-1调控机制的研究发现,Pim-1与EGFR通路形成一个调控环路,通过siRNA下调Pim-1的表达水平,显著抑制p-EGFR及EGFR的表达水平；而通过厄洛替尼抑制EGFR通路活性,显著抑制Pim-1蛋白的表达水平,通过siRNA下调EGFR表达水平,也观察到相同效应(P<0.05)。4.组织miR-497、Pim-1蛋白的表达水平检测及预后价值评估对90例胰腺癌组织及相应的癌旁组织(90例)进行了miR-497原位杂交及Pim-1蛋白免疫组化检测。胰腺癌组织中miR-497的表达水平显著低于癌旁组织(P=0.000),而癌组织Pim-1蛋白表达水平显著高于癌旁组织(P=0.041)。胰腺癌组织miR-497表达水平与Pim-1表达水平存在负相关性(P=0.006)。组织miR-497,Pim-1表达水平与患者性别、年龄、肿瘤部位、分化程度、T分期、N分期、TNM分期、是否合并糖尿病及是否存在周围神经侵犯等临床病理参数并无显著相关性。单因素生存分析发现,TNM分期、Pim-1蛋白表达水平、miR-497表达水平与患者预后相关(P值分别为P=0.027;P=0.039；P=0.027)。多因素生存分析发现,TNM分期(Ⅱ/Ⅲ/Ⅳ)及组织miR-497表达水平(Low)是胰腺癌患者预后不良的独立危险因素(分别为P=0.016,HR=1.97,95%可信区间(CI):1.13-3.42;P=0.01,HR=2.76,95%CI:1.159-6.58).5.血浆miR-1497、Pim-1蛋白的表达水平检测及诊断价值评估通过检测78例血浆样本,发现miR-497在血浆中的表达水平很低,难于有效检测。胰腺癌患者(90例)血浆标本Pim-1蛋白表达水平(29.8±47.7ng/ml)显著高于健康志愿者(20例,0.21±0.31ng/ml).慢性胰腺炎患者(19例,3.11±5.2ng/ml).胰腺其他肿瘤(29例,8.75±6.6ng/ml)及胰腺内分泌肿瘤患者(20例,15.7±8.9ng/ml)的血浆标本的Pim-1蛋白表达水平(P值分别为P=0.000；P=0.000;P=0.000;P=0.01);且胰腺癌患者血浆Pim-1蛋白表达水平显著高于非癌对照病例(29.8±47.7ng/ml VS7.21±8.3ng/ml,P=0.000).血浆Pim-1表达水平具有鉴别胰腺癌患者与健康志愿者的价值(曲线下面积(AUC)为0.984,95%CI为0.963-1.006,P=0.000),诊断敏感性及特异性分别为95.6%和100%。血浆Pim-1表达水平具有鉴别胰腺癌患者与慢性胰腺炎患者的价值(AUC=0.895,95%CI为0.812-0.977,P=0.000),敏感性及特异性分别为87.8%和77.8%。血浆Pim-1表达水平具有鉴别胰腺癌患者与胰腺其他肿瘤患者的价值(AUC=0.706,95%CI为0.611～0.802,P=0.001),敏感性及特异性分别为51.1%和86.2%。血浆Pim-1表达水平具有鉴别胰腺癌患者与非胰腺癌对照病例的价值(AUC=0.768,95%CI为0.701～0.836,P=0.000),敏感性及特异性分别为73.3%和66.7%。未发现血浆Pim-1表达水平鉴别胰腺癌患者与PNET患者的价值(曲线下面积(AUC)为0.529,95%CI为0.411-0.647,P=0.684)。6.血浆Pim-1蛋白表达水平与临床病理参数及预后的相关性通过与临床病理参数的相关性研究,发现血浆Pim-1表达水平与患者年龄、肿瘤部位、TNM分期相关(P值分别为P=0.031,P=0.000,P=0.013)。高表达Pim-1的患者具有更高TNM分期的肿瘤。单因素生存分析发现M分期、TNM分期、切除情况、Pim-1蛋白表达水平与患者预后相关(P=0.000；P=0.000；P=0.002；P=0.026)。多因素生存分析发现,TNM分期(advanced)及血浆Pim-1表达水平(High)是胰腺癌患者预后不良的独立危险因素(分别为P=0.000,HR=1.88,95%CI:1.17.2.57;P=0.037,HR1.87,95%CI:1.04-3.35).结论MiR-497通过负向调控IGF-1R、FGF2、FGFR1、Kras及Pim-1表达水平,参与胰腺癌细胞增殖、凋亡、细胞周期、侵袭及迁移、吉西他滨及厄洛替尼化疗耐药、干细胞标志物等恶性表型的调控。作为miR-497的一个靶基因,Pim-1与EGFR通路形成一个调控环路,参与调控胰腺癌细胞恶性表型。MiR-497、Pim-1可能成为新的胰腺癌分子治疗靶标。组织miR-497和Pim-1及血浆Pim-1表达水平具有一定的预后价值,可能成为新的胰腺癌预后指标。血浆Pim-1表达水平展现出一定的诊断价值,可能用于胰腺癌的早期或鉴别诊断。更多还原

【Abstract】 BackgroundThe main obstacles in the field of pancreatic cancer research, which we must face, are difficulty in early diagnosis, chemoresistance and lack of effective biomarkers for prognosis. Identification of new biomarkers for early detection and predicting prognosis, disclosing the mechanisms of tumorigenesis and development, and reversal of multidrug resistance are effective strategies to improve overall survival of patients with pancreatic cancer.MicroRNAs (miRNAs) are short, single-stranded RNA molecules with22nucleotides, which involve in tumorigenesis, cancer development, proliferation, metastasis and chemoresistance. Our previous study showed miR-497was significant downregulation in gemcitabine resistant cell lines, compared with that in parent cell lines, which meant that miR-497might play an important role in the regulation of drug resistance in pancreatic cancer. MiR-497decrease has been illustrated in breast, colorectal and cervical cancers and contributes to the malignancy of multiple tumors. However, the expression levels and regulatory roles of miR-497in pancreatic cancer are still unclear. It is necessary to investigate the regulatory roles and mechanisms of miR-497in pancreatic cancer and assess its diagnostic and prognostic values.ObjectiveThis study aimed to investigate the effects of miR-497on regulating proliferation, apoptosis, cell cycle, migration and invasion, chemosensitivity, biomarkers of cancer stem cells in pancreatic cancer cells. Then we studied the mechanisms of miR-497in regulating the progression of pancreatic cancer. We also investigated the regulatory roles and mechanisms of Pim-1, which was a target of miR-497. In addition, we detected the expression levels of miR-497and Pim-1in tissues and plasma, and then assessed their diagnostic and prognostic values. MethodsFor the in vitro experiments, proliferation was analyzed using a cell count kit (CCK-8). The effects on the chemosensitivity of gemcitabine and erlotinib were evaluated by CCK-8assay. Cell-cycle and apoptosis analysis were performed by fluorescence-activated cell sorting (FACS) flow-cytometry. Real-time PCR was used to detect the expression levels of miRNA and mRNA. Protein levels were detected by western blot. For the in vivo experiment, SW1990cells steadily expressing miR-497were established. Subcutaneous transplantation models of pancreatic cancer in nude mice were established to observe the tumor growth. Luciferase reporter assays were used to verify the candidate genes. The expression levels of miR-497and Pim-1in tissues were detected by in situ hybridization (ISH) and immunohistochemistry (IHC), respectively. Enzyme linked immunosorbent assay (ELISA) were used to detect the levels of plasma Pim-1. The prognostic values of miR-497and Pim-1were assessed by univariate and multivariate survival analysis. The diagnostic values of plasma pim-1levels were evaluated using receiver operating characteristic curve (ROC curve).Results1. The effects of miR-497on the malignant phenotype of pancreatic cancer cellsUpregulation of miR-497in SW1990and MiaPaCa-2cells significantly suppressed proliferation, increased sensitivity to gemcitabine and erlotinib, promoted expression of cleaved caspase-3and PARP activated by gemcitabine, inhibited migration and invasion capacities, decreased the percentage of S phase cells, and attenuated the expression levels of ABCG2and EZH2mRNA, which were biomarkers of pancreatic cancer stem cells. Opposite effects were observed after downregulation of miR-497. For the in vivo experiment, overexpression of miR-497significantly suppressed tumor growth, compared with control group.(P<0.05).2. The mechanisms of miR-497in regulating the progression of pancreatic cancer. Dual-luciferase reporter assays were performed to confirm the targets of miR-497. We found that luciferase activities were significantly decreased after co-transfection of miR-497mimics and vectors expressing wild type target sequence in293A cells, compared with that in cells co-transfected with mimics and vectors expressing mutated target sequence (P<0.05), which meant IGF-1R, FGF2, FGFR1, Kras and Pim-1were direct targets of miR-497. We also detected proteins levels, which showed upregulation of miR-497significantly decreased the expression of IGF-1R, FGF2, FGFR1, Kras and Pim-1. On the contrary, inhibition of miR-497increased the levels of these proteins. Additionally, we investigated the levels of the key molecules in the important signaling pathways. We observed that the expression levels of p-AKT and p-ERK decreased significantly after overexpression of miR-497. On the other hand, inhibition of miR-497increased the expression levels of p-AKT and p-ERK.(P<0.05).3. The effects of Pim-1on the malignant phenotype of pancreatic cancer cellsAs a target of miR-497, the regulatory roles of Pim-1in pancreatic cancer were still unknown. We showed that decrease of Pim-1levels by siRNA significantly suppressed proliferation, induced apoptosis, increased sensitivity to gemcitabine and erlotinib, inhibited migration and invasion capacities, decreased the percentage of S phase cells, and attenuated the expression levels of ABCG2and EZH2mRNA. Similar effects were observed after blocking Pim-1activity by SGI-1776. In addition, we investigated the mechanisms. We found that Pim-1and EGFR pathway formed a feedback loop.Knockdown of Pim-1by siRNA decreased the expression of EGFR and p-EGFR. Whereas blocking EGFR pathway by erlotinib or EGFR siRNA suppressed the expression of Pim-1.(P<0.05).4. The expression levels and prognostic values of miR-497and Pim-1in tissuesThe expression levels of miR-497were significantly decreased in pancreatic cancer tissues, compared with that In tumor-adjacent tissues (P=0.000).Whereas Pim-1protein levels were significantly increased in cancer tissues, compared with that in tumor-adjacent tissues (P=0.041). There was a negative correlation between the expression levels of miR-497and Pim-1protein in pancreatic cancer tissues (P=0.006). There was no association between the expression levels of miR-497/Pim-1and clinicopathological parameters, including sex, age, tumor locations, differential degree, TNM staging, diabetes, and perineuronal invasion. Univariate analysis showed TNM staging, Pim-1levels and miR-497levels were the potential prognostic factors of pancreatic cancer (P=0.027, P=0.039, and P=0.027, respectively). Multivariate analysis indicated that TNM staging (Ⅱ/Ⅲ/Ⅳ) and miR-497expression (Low) were the independent adverse prognostic factor (P=0.016, HR=1.97,95%CI:1.13-3.42, P=0.01, HR=2.76,95%CI:1.159-6.58, respectively).5. The expression levels and diagnostic values of miR-497and Pim-1in plasma78plasma samples were collected for detection of miR-497. Depressingly, miR-497could not be effectively detected for its extremely low expression in plasma. The plasma levels of Pim-1in patients with pancreatic cancer (29.8±47.7ng/ml) were significantly higher than that in healthy volunteers (0.21±0.31ng/ml), chronic pancreatitis (3.11±5.2ng/ml), other pancreatic tumors (8.75±6.6ng/ml), pancreatic neuroendocrine pancreatic patients (PNET,15.7±8.9ng/ml)(P=0.000; P=0.000; P=0.000; P=0.01, respectivley). In addition, the plasma levels of Pim-1in patients with pancreatic cancer were significantly higher than that in all controls (29.8±47.7ng/ml VS7.21±8.3ng/ml,P=0.000).The plasma levels of Pim-1displayed diagnostic values for distinguishing patients with pancreatic cancers from healthy volunteers, chronic pancreatitis, other pancreatic tumors patients and all controls (AUC=0.984,95%CI:0.963～1.006, P=0.000; AUC=0.895,95%CI:0.812～0.977, P=0.000; AUC=0.706,95%CI:0.611～0.802, P=0.001, AUC=0.768,95%CI:0.701～0.836, P=0.000, respectively). The sensitivity and specificity for distinguishing pancreatic cancer from healthy volunteers, chronic pancreatitis, other pancreatic tumors patients and all controls were95.6%and100%,87.8%and77.8%,51.1%and86.2%,73.3%and66.7%, respectively.6. The correlation between plasma Pim-1levels and clinicopathological parameters and survival analysisPlasma Pim-1levels were associated with age, tumor locations and TNM staging (P=0.031,P=0.000, and P=0.013, respectively). Patients with high levels of Pim-1had advanced tumors. Univariate analysis demonstrated that M staging, TNM staging, R status and Pim-1levels were the potential prognostic factors of pancreatic cancer (P=0.000; P=0.000; P=0.002; P=0.026). Multivariate analysis indicated that TNM staging (advanced) and plasma Pim-1expression (High) were the independent adverse prognostic factor (P=0.000, HR=1.88,95%CI:1.17-2.57, P=0.037, HR=1.87,95%CI:1.04-3.35, respectively).ConclusionMiR-497involved in regulation of proliferation, apoptosis, cell cycle, migration and invasion, chemosensitivity, biomarkers of cancer stem cells in pancreatic cancer cells through negatively regulating the expression levels of IGF-1R, FGF2, FGFR1, Kras and Pim-1. As a target of miR-497, Pim-1contributed to the malignancy of pancreatic cancer. MiR-497and Pim-1could be new therapeutic targets in pancreatic cancer. The levels of miR-497and Pim-1in tissues and the plasma levels of pim-1displayed prognostic values, and could be new biomarkers in predicting the prognosis of pancreatic cancer. Plasma Pim-1levels presented a potential value for distinguishing pancreatic lesions, and could be also a new plasma biomarker for early detection of pancreatic cancer.更多还原