Preclinical drug development in pancreatic cancer

进度报告使用变构小分子抑制剂AZD6244对70个胰腺癌细胞系进行了MEK抑制敏感性分析，该药物目前处于II期临床试验中。大约40%的胰腺癌细胞系表现出明显的MEK敏感性，根据它们的一半生长抑制浓度（GI50）小于1微摩尔和曲线上的活性面积。在敏感细胞系中，对MEK抑制的总体反应构成了细胞抑制生长阻滞效应，而不是像其他实体器官癌症中MEK治疗那样诱导细胞死亡。为了提高未来MEK治疗胰腺癌的疗效，已经取得了以下进展：a.高通量siRNA筛选鉴定了胰腺癌中介导MAPK通路抑制的耐药基因；为了确定胰腺癌细胞使用或开启的细胞内信号通路和靶点以逃避MAPK通路阻断和MEK抑制，由RNAi筛选中心、NIH化学基因组学中心、NIH转化治疗中心、NHGRI/NIH对AZD6244抑制的YAPC细胞系进行了合成致死性药物致敏筛选。在两个独立的二级筛选中证实了介导MEK抑制的靶点是CNKSR1 （Ras 1激酶抑制因子的连接增强子）、WNK2 （WNK赖氨酸缺陷蛋白激酶2）、MAP3K8 （MAPK途径的COT激酶）和RPS6KA5（核糖体蛋白S6激酶，90kDa，多肽5）。CNKSR1的基因表达水平似乎与对MAPK抑制的抗性直接相关，这使得CNKSR1既是一个有吸引力的靶标，也是胰腺癌对MEK抑制反应的可能预测因子。b.下一代最常参与癌症的基因基因组测序确定了与MAPK通路抑制反应相关的遗传变异。有趣的是，在超过85%的胰腺癌中发生突变的RAS癌基因的几种特异性亚型中，它与MEK敏感性相关。突变异构体G12R仅在MEK敏感细胞系中发现，可能是预测MEK抑制反应的一种容易适用的临床标志物。在三分之一的胰腺癌细胞系中，Erk小分子抑制剂VTX-11e （NCGC00242487-01）诱导细胞凋亡，而不是细胞周期阻滞，这意味着与MEK抑制相比，其治疗策略更优越。MEK-和Erk抑制同样有效地抑制MAPK通路信号传导，这是由phospho Elk水平的降低确定的。与VTX-11e治疗后细胞周期阻滞的细胞相比，VTX-11e治疗后细胞死亡的细胞系显示Erk靶蛋白p90-RSK（核糖体蛋白S6激酶，90kDa，多肽1）磷酸化的减少幅度更大。基因表达谱分析发现，在Erk耐药细胞系中，多个涉及胚胎学途径的新基因上调，可能维持p90-RSK磷酸化，尽管Erk抑制，但p90-RSK磷酸化介导了存活。2. 用PI3K/mTOR双抑制剂BEZ235治疗的70种胰腺癌细胞系，在低纳摩尔范围内的GI50值判断，几乎所有的胰腺癌细胞系都表现出明显的敏感性。当对PI3K/mTOR抑制诱导细胞死亡进行测试时，约20%的细胞系显示出超过2.5倍的凋亡诱导，并被归类为对PI3K- akt抑制敏感。通过建立敏感和耐药胰腺癌细胞系的异位异种移植模型，在体内证实了PI3K/mTOR双抑制的体外活性。突变检测结果显示，在PI3K和akt基因的内含子区域，以及对PI3K/mTOR抑制敏感的细胞系中，存在新的单核苷酸变异，目前的研究正在研究这些变异对PI3K信号传导的功能影响。敏感系和抗性系的基因表达谱显示，两组之间的激酶和磷酸酶表达存在差异。为了对这种可能的生物标志物进行前瞻性验证，已经建立了来自NCI外科分部手术患者的人类胰腺癌标本的异种样本库。3. 在针对胰腺癌干细胞样细胞的高通量药理学筛选中，发现“ITK”抑制剂NCGC-00188382是活性最高的化合物。胰腺癌细胞和原位激酶筛选显示，该化合物抑制多种激酶（CDK7、IRAK4、CLK1、CLK2、CaMMK2、TAOK3、aurora B、Ephrin受体B2）。这些新靶点在二次筛选中得到了验证。siRNA沉默研究探索了这种新化合物的多药理学作用机制，发现了其中一些靶点的“内在协同作用”。该化合物在体内进行了测试，显示出很强的抗转移表型。目前的研究重点是提高这一有前景的新分子的选择性和药代动力学特征，并进一步了解其作用机制。4. 肿瘤环境对胰腺导管腺癌抗癌治疗效果的影响微环境的作用尚不能在体外细胞系统中进行研究。为了评估各种细胞成分之间复杂的相互作用，作为胰腺癌新治疗策略的可能靶点，需要一个体内模型：胰腺癌的转基因/基因敲除小鼠是研究可能的致癌调节剂的成熟模型。这些模型包含Kras癌基因的条件敲入突变，该突变存在于85%的胰腺癌中，并结合敲除常见的肿瘤抑制基因CDKN2A （p16）和Smad4，这些基因在50%的病例中丢失。这些基因工程小鼠模型类似于人类胰腺癌的基因组图谱，95%的病例是由其中一个基因的改变引起的。转基因Kras p16敲除胰腺癌动物模型Pdx-Cre的治疗LSLKrasG12D;Ink4a/Arflox/lox与TGFRbeta抑制剂ly2109761相比，通过增加葡聚糖灌注和肿瘤内吉西他滨的测量，使胰头肿瘤的灌注增加了数倍。此外，与对照组相比，新型白细胞介素10肽类似物10N与吉西他滨的组合导致肿瘤体积急剧减少，并且治疗动物的生存优势。这种抗肿瘤作用不依赖于吉西他滨向肿瘤输送的增加，这表明10A和吉西他滨之间存在一种新的、尚未发现的协同效应。目前的研究旨在确定抗白细胞介素10和吉西他滨靶向的细胞区室，以及两者联合治疗后表达不同的基因。这些积极的发现正在推广到其他转基因动物模型中，以探索基因型导向的抗微环境治疗策略，将抗基质与抗癌治疗结合起来治疗胰腺癌。其目的是将吉西他滨联合治疗的阳性结果转化为目前正在进行的RECLAP试验。

Progress Report 1. Improving anti-MAPK pathway therapy in pancreatic cancer A panel of 70 pancreatic cancer cell lines was profiled for sensitivity to MEK inhibition using the allosteric small molecule inhibitor AZD6244 which is currently in phase II clinical trials. About 40 percent of profiled pancreatic cancer lines exhibit marked MEK sensitivity according to their half growth inhibitory concentration (GI50) of less than 1 micromolar and activity area over the curve. Overall response to MEK inhibition in sensitive cell lines constitutes a cytostatic growth arrest effect rather than induction of cell death as described for MEK therapy in other solid organ cancers. To improve efficacy of future MEK treatment in pancreas cancer the following progress has been made: a. A high-throughput siRNA screen identified genes mediating resistance to MAPK pathway inhibition in pancreas cancer: to identify intracellular signaling pathways and targets which are used, or switched on, by pancreas cancer cells to escape MAPK pathway blockade and MEK inhibition a synthetic lethality drug sensitization screen in the cell line YAPC inhibited with AZD6244 has been carried out by the RNAi Screening Center, NIH Chemical Genomics Center, NIH Center for Translational Therapeutics, NHGRI/NIH. Targets validated in two independent secondary screens mediating resistance to MEK inhibition are CNKSR1 (connector enhancer of kinase suppressor of Ras 1), WNK2, (WNK lysine deficient protein kinase 2), MAP3K8 (COT kinase of the MAPK pathway), and RPS6KA5 (ribosomal protein S6 kinase, 90kDa, polypeptide 5). Gene expression levels of CNKSR1 appear to directly correlate with resistance to MAPK inhibition making CNKSR1 both an attractive target as well as possible predictor of response to MEK inhibition in pancreas cancer. b. Next generation genomic sequencing of genes most commonly involved in cancer identified genetic variants associated with response to MAPK pathway inhibition. Interestingly, among several, specific isoforms of the RAS oncogene, which is mutated in more than 85 percent of pancreas cancers, are associated with MEK sensitivity. The mutation isoform G12R is exclusively identified in MEK sensitive cell lines and might represent a readily applicable clinical marker predicting response to MEK inhibition. c. The Erk2 inhibitor VTX-11e (NCGC00242487-01) is superior to MEK inhibition in a subset of pancreatic cancers The Erk small molecule inhibitor VTX-11e (NCGC00242487-01) induces apoptosis rather than cell cycle arrest in a third of pancreatic cancer cell lines implying a superior treatment strategy compared to MEK inhibition. Both MEK- and Erk inhibition equally effectively inhibit MAPK pathway signaling as determined by reduction of phospho Elk levels. Cell lines undergoing cell death following treatment with VTX-11e show a greater reduction of phosphorylation of the Erk target p90-RSK (ribosomal protein S6 kinase, 90kDa, polypeptide 1) than cells undergoing cell cycle arrest upon VTX-11e treatment. Gene expression profiling identified multiple novel genes involved in embryological pathways upregulated in Erk-resistant cell lines possibly maintaining p90-RSK phosphorylation which mediates survival despite Erk inhibition. 2. Targeting the PI3K-Akt pathway in pancreas cancer Nearly all of 70 pancreatic cancer lines treated with the dual PI3K/mTOR inhibitor BEZ235 displayed marked sensitivity when judged on their GI50 values in the low nanomolar range. When tested for induction of cell death upon PI3K/mTOR inhibition, about 20 percent of cell lines showed a greater than 2.5-fold induction of apoptosis and were classified as sensitive to PI3K-Akt inhibition. In vitro activity of dual PI3K/mTOR inhibition was confirmed in vivo using a heterotopic xenotransplant models established from sensitive and resistant pancreas cancer cell lines. Results of mutation testing revealed novel single nucleotide variants in intronic regions of both the PI3K andAKT genes, and others, in cell lines sensitive to PI3K/mTOR inhibition and current studies are examining the functional impact on PI3K signaling of these variants. Gene expression profiling of sensitive and resistant lines showed a number of kinases and phosphatases differently expressed between the two groups. For prospective validation of such a possible biomarker a xenobank from human pancreas cancer specimens from patients operated on at the Surgery Branch/NCI has been established. 3. Preclinical evaluation of the ITK inhibitor NCGC-00188382 in pancreas cancer In a high-throughput pharmacological screen against pancreas cancer stem-like cells, the 'ITK' inhibitor NCGC-00188382 was found to be the most active compound. Cell-based and in situ kinase screens in the pancreas cancer line Panc1 showed that the compound inhibits a number of kinases (CDK7, IRAK4, CLK1, CLK2, CaMMK2, TAOK3, aurora B, Ephrin receptor B2). These novel targets were validated in secondary screens. siRNA silencing studies probing into the polypharmacological mechanism of action of this novel compound identified 'intrinsic synergism' of some of these targets. The compound was tested in vivo and showed a strong anti-metastatic phenotype. Current studies focus on improving the selectivity and pharmacokinetic profile of this promising new molecule, and to further understand its mechanism of action. 4. The impact of the tumor environment on the efficacy of anticancer therapy in ductal adenocarcinoma of the pancreas The role of the microenvironment cannot be studied in an in vitro cell system. To evaluate the complex interactions of the various cellular components as possible targets for novel treatment strategies in pancreas cancer requires an in vivo model: Transgenic/knockout mice who develop pancreatic cancer are well-established models for studying possible modulators of carcinogenesis. These models contain conditional knock-in mutations of the Kras oncogene which is present in 85% of pancreas cancer in combination with knock-outs of the common tumor suppressor genes CDKN2A (p16) and Smad4 which are lost in 50% of cases. These genetically engineered mouse models resemble the human genomic landscape of pancreas cancer which is driven by alterations in one of these genes in 95% of cases. Treatment of the transgenic Kras p16 knockout pancreas cancer animal model Pdx-Cre; LSLKrasG12D; Ink4a/Arflox/lox with the TGFRbeta inhibitor LY2109761increases perfusion of pancreatic head tumors several-fold compared to control as measured by increased dextran perfusion and intratumoral gemcitabine. Additionally, the combination of the novel interleukin 10 peptide analogue 10N with gemcitabine led to a dramatic decrease in tumor volume compared to control as well as to a survival advantage in the treated animals. This anti-tumor effect is independent of increased delivery of gemcitabine to the tumor and suggests a novel, yet undioscovered synergistic effect between 10A and gemcitabine. Current studies are aiming to identify the cellular compartment targeted by anti-interleukin 10 and gemcitabine as well as gene differently expressed upon treatment with this combination. These positive findings are in the process of being extended to other transgenic animal models to probe into genotype-directed anti-microenvironment treatment strategies of combining anti-stroma with anti-cancer treatments in pancreas cancer. It is aimed to translate positive findings of gemcitabine combinations into the currently ongoing RECLAP trial.