Identifying New Glioma-Associated Tumor Suppressors and Oncogenes

胶质瘤干细胞是一种能够自我更新的肿瘤细胞亚群，能够在体内原位移植瘤细胞，并在体内和体外产生多种神经元样和胶质样有丝分裂后后代。缺乏PTEN的小鼠表现出增大的、组织结构异常的脑，这是由于细胞增殖增加、细胞死亡减少和细胞大小增大。Pten -/-干细胞/祖细胞也显示出至少部分地对缩短的细胞周期具有更大的增殖能力（Croszer等人，2001年）。成熟星形胶质细胞中p16 INK 4A/ARF敲除和组成型活性EGFR表达的组合导致颅内移植后胶质瘤样病变的形成（Bachoo等人，2002年）。Trp 53和Nf 1（神经特异性神经纤维瘤病1型）双突变小鼠显示出一系列星形细胞瘤阶段，从低级星形细胞瘤到多形性胶质母细胞瘤（GBM）（Reilly et al.，2000年）。最近，常规和基于阵列的CGH（aCGH）分析人类胶质瘤显示了大量的拷贝数改变（CNA），包括增益/扩增，（1 p34 -36、1 q32、3q 26 -28、5 q、7 q31、8 q24、11 q、12 q13、13 q、15 p15、17 q22 -25、19 q、20 p和20 q）和缺失/缺失（3q25-26，4q，6q26-27，9p，10p，10q，11p，12q22，13q，14q13，14q23-31，15q13-21，17p11-13，18q22-23，19q和22q）（Kotliarov et al.，2006; Nigro等人，2005;菲利普斯等人，2006年）。大量的染色体畸变以及其中包含的大量基因，迄今为止使得不可能确定哪些基因部分负责驱动这些肿瘤的生物学。我们使用Affytek 100 K单核苷酸多态性（SNP）阵列芯片和Genechip Human Genome U133 Plus 2.0表达阵列分析了大量胶质瘤样本，以确定复发性CNA的遗传特征（Kotliarov et al. 2006）。基于我们从这些阵列和基因表达谱实验中获得的生物信息学数据，我们发现了神经胶质瘤中经常改变的新基因。我们已经产生了这些基因的序列验证的基因Gateway进入克隆，并将它们克隆到pLenti/UbC/V5表达载体中用于转导各种靶细胞系。我们将使用的靶细胞不仅包括已建立的肿瘤干细胞系，如0308和1228胶质瘤系，而且包括来自各种遗传修饰小鼠系的小鼠胚胎神经干细胞（mNSC），如PTEN loxP/loxP & p16 INK 4A/ARF -/-、PTEN loxP/loxP、PTEN loxP/+、p16 INK 4A/ARF -/-和野生型小鼠。为了产生用于该研究的双敲除mNSC，我们培育了PTEN loxP/loxP（Croszer等人，2001）（The杰克逊实验室，ME）和p16 INK 4A/ARF -/-（Sharpless等人，2001）（NCI-Frederick，MD）小鼠，并产生了PTENloxP/loxP & p16 INK 4A/ARF -/-、PTENloxP/loxP、PTENloxP/+、p16 INK 4A/ARF -/-和野生型小鼠。然后建立E14小鼠胚胎神经干细胞（embryonic neural stem cells，mNSCs），将CRE-EGFP质粒转染E14小鼠胚胎神经干细胞中，进行PTEN基因缺失，用抗生素筛选，获得PTENloxP/loxP和p16 INK 4A/ARF -/-、PTENloxP/loxP、PTENloxP/+ mNSCs。筛选后，用CRE筛选PTEN基因敲除的克隆。我们通过蛋白质印迹分析和基因分型证实了PTEN敲除。利用我们的候选基因构建体和mNSCs，我们将确定候选基因是否以可能与肿瘤发生（即免疫抑制动物中的克隆形成、增殖、凋亡、致瘤潜力）中的作用一致的方式改变这些细胞的生物学。通过这些体外和体内筛选，我们有望发现与胶质瘤发生相关的候选基因，并为治疗干预提供潜在的新靶点。

Glioma stem cell is a tumor subpopulation that can self-renew in culture, perpetuate a tumor in orthotopic transplant in vivo, and generate diversified neuron-like and glia-like postmitotic progeny in vivo and in vitro. Mice lacking PTEN exhibited enlarged, histoarchitecturally abnormal brains, which resulted from increased cell proliferation, decreased cell death, and enlarged cell size. Also Pten -/- stem/progenitor cells display a greater proliferation capacity at least in part, to a shortened cell cycle (Croszer et al., 2001). The combined p16INK4A/ARF knockout and constitutively active EGFR expression in mature astrocytes led to the formation of glioma like lesions following intracranial transplantation (Bachoo et al., 2002). Trp53 and Nf1 (neural-specific neurofibromatosis type 1) double mutant mice show a range of astrocytoma stages, from low-grade astrocytoma to glioblastoma multiforme (GBM) (Reilly et al., 2000). Recently, conventional and array-based CGH (aCGH) profiling of human gliomas have shown a significant number of copy number alterations (CNAs) including gain/amplication (1p34-36, 1q32, 3q26-28, 5q, 7q31, 8q24, 11q, 12q13, 13q, 15p15, 17q22-25, 19q, 20p, and 20q), and deletion/loss ( 3q25-26, 4q, 6q26-27, 9p, 10p, 10q, 11p, 12q22, 13q, 14q13, 14q23-31, 15q13-21, 17p11-13, 18q22-23, 19q, and 22q)(Kotliarov et al., 2006; Nigro et al., 2005; Phillips et al., 2006). The large number of chromosomal aberrations, and the large number of genes contained therein, have to date made it impossible to identify which genes are in part responsible for driving the biology of these tumors. We have analyzed a large number glioma samples for genetic characterization of recurring CNAs using Affymetrix 100K single-nucleotide polymorphism (SNP) array chips and Genechip Human Genome U133 Plus 2.0 Expression array (Kotliarov et al. 2006). Based on our bioinformatics data from these array and gene expression profiling experiments, we have found novel genes frequently altered in gliomas. We have generated sequence-verified gene Gateway entry clones of these genes and cloned them into pLenti/UbC/V5 expression vectors for transduction of various target cell lines. The target cells we will use will include not only established tumor stem cell lines such as 0308 and 1228 glioma lines, but also mouse embryonic neural stem cells (mNSCs) from various genetically modified mice lines such as PTEN loxP/loxP & p16INK4A/ARF -/-, PTEN loxP/loxP, PTEN loxP/+, p16INK4A/ARF -/- and wildtype mice. To generate double knockout mNSCs for this study, we had bred PTEN loxP/loxP (Croszer et al., 2001) (The Jackson Laboratory, ME) and p16INK4A/ARF -/- (Sharpless et al., 2001) (NCI-Frederick, MD) mice and generated PTEN loxP/loxP & p16INK4A/ARF -/-, PTEN loxP/loxP, PTEN loxP/+, p16INK4A/ARF -/- and wildtype mice. Then we had established E14 mouse embryonic neural stem cells (mNSCs) from these mice.To make PTEN loxP/loxP & p16INK4A/ARF -/-, PTEN loxP/loxP, PTEN loxP/+ mNSCs, we had transfected CRE-EGFP plasmid in these E14 mNSCs for PTEN deletion and selected transfected cells with antibiotics during two weeks. After selecting, we screened clones of PTEN knockout by CRE. We confirmed PTEN knockout through Western Blot analysis and genotyping. With our candidate gene constructs and mNSCs, we will identify whether candidate genes change the biology of these cells in such a way that may be consistent with a role in tumorigenesis (i.e. clonogenecity, proliferation, apoptosis, tumorigenic potential in immunosuppressed animals). It is expected that through these in vitro and in vivo screens, we will discover candidate genes which are related to gliomagenesis and which may offer potentially new targets for therapeutic intervention.