Embryonic Transcription Factor Function in Human Colorectal Cancer Stem Cells

Nanog调控其他干细胞转录因子Nanog在胚胎早期发育过程中调控Oct4和SOX2的表达。我们发现，用等位基因特异性shRNA Np8-1抑制NanogP8可抑制Oct4和SOX2的表达。因此，靶向一个因子可能抑制其他因子，因此单一药物靶向可能影响多个下游途径。此外，我们已经证明Nanog和NanogP8通过降低结直肠癌细胞（CRC）的球形性、致瘤性和转移潜力来影响干细胞。我们现在已经发现，等位基因特异性抑制NanogP8减少了侧种群的大小，而Nanog或NanogP8的过表达增加了侧种群的大小。因此，Nanog和NanogP8在调节CRC的干性方面起着重要作用。从NIH临床中心治疗的患者中切除的10个去鉴定的结直肠癌样本进行了Nanog和NanogP8转录物和蛋白的表达分析。RT-PCR后用限制性内切酶AlwNI切割cDNA，该酶特异性切割NanogP8而非Nanog。7例肝转移灶和3例相邻肝标本均含有nanog相关转录本。7份标本中6份含有NanogP8, 1份仅含有Nanog。相邻的肝脏标本中有一个含有NanogP8，而另外两个只表达NanogP8转录本。所有结果均经Sanger测序证实。免疫组织化学（IHC）也证实了Nanog蛋白存在于含有可检测的Nanog或NanogP8转录本的肿瘤细胞的细胞质中，而不存在于缺乏这些转录本的CRC细胞中。此外，样品9的免疫组化表明，在样品中只鉴定出NanogP8转录物，而NanogP8转录物中存在Nanog蛋白。因此，基于这个需要确认和扩大的小样本，大约70%的临床结直肠癌表达Nanog家族成员，60%表达NanogP8。与NCI-Frederick的Veenstra实验室合作，我们尝试使用串联质谱/质谱法鉴定结直肠癌细胞系中的NanogP8蛋白。免疫沉淀结直肠癌细胞系提取物，SDS-PAGE分离，凝胶胰消化，MS/MS。在过表达NanogP8的克隆A的提取物中鉴定出4个nanog相关肽（KTWFQNQRM、KYLSLQQMQELSNILNLSYKQ、KKEDKVPVKK和KGKQPTSAENSVAKK）。有趣的是，最后一个肽是NanogP8独有的，它包括在密码子82处从NanogP8的Lys (K)到NanogP8的Asn (N)的转变。Ambady(2)也报道了这种氨基酸变化，与我们的基因测序数据一致。在未转染过表达的CRC细胞系提取物中未发现Nanog/NanogP8蛋白，可能是因为CRC细胞系表达低水平的内源性蛋白。NanogP8启动子活性在三维生长过程中增加，与NanogP8基因表达一致。我们之前已经证明，在无血清培养基（SFM）中培养的CRC细胞系克隆A和CX-1的球状体中，NanogP8转录物增加。我们测试了这样的假设：当CRC细胞从二维单层培养转变为三维球形培养时，NanogP8基因的启动子活性增加。从Jeter等人(3)获得了一个慢病毒NanogP8启动子报告子。克隆A和CX-1的单层培养比克隆A和CX-1的球形培养表达水平低。对克隆a和CX-1的GFP阳性细胞计数显示，球状体中GFP阳性细胞的数量分别比单层增加了6至14倍。这类似于qRT-PCR评估的从二维单层到三维球体转变过程中转录本数量的增加。慢病毒shRNA转导到Nanog或NanogP8的作用机制和抑制3-D生长要想成为一种成功的治疗方法，shRNA抑制Nanog必须在临床前模型中抑制体内已建立的肿瘤的生长。我们从国防部获得了一笔拨款来支持这种临床前工作，但不幸的是，这项工作还没有开始。然而，我们一直忙于用我们的等位基因特异性shrna进行体外实验，以及它们对已建立的三维球体和悬浮细胞的消退的影响，以确定引起肿瘤消退所需的感染多样性（MOI）。我们之前的实验表明，单层结直肠癌的处理可诱导膜联蛋白V的表达，作为细胞凋亡的指示。现在，我们首先将带有shRNA的LV带到NanogP8 （shNp8-1）、Nanog （shNG-1）和对照shRNA （shNEG）上，转导CX-1和克隆a细胞，然后在单层中培养24小时，然后过渡到SFM。我们首先证明LVshNp8-1转导在悬浮培养72小时后显著降低细胞质量。然后我们开始在ULLA板上诱导悬浮在SFM中的CRC生长。由于我们的LV制剂表达GFP，我们首先确定了悬浮液中转导的细胞比例，发现40%至75%的CRC细胞稳定转导。然后，我们评估了与未治疗的CRC细胞或用对照shNEG处理的细胞相比，LV shRNA转导到Nanog或NanogP8是否抑制了细胞的生长。超过7个不同实验的累积结果表明，在三维生长中治疗CRC，在72小时内显著抑制了50%的短期生长。然后，我们试图确定LV shRNA如何在3d培养中影响细胞凋亡。我们之前证明悬浮培养引起的anoikis细胞凋亡依赖于Caspase 8的激活(4)。在LV shRNA处理的克隆A和CX-1悬浮培养4天后，LV处理3天后，我们测量了Caspase 3、8和9的活性。在CX-1和克隆A中，通过shRNA对Nanog和NanogP8的处理，Caspase 3（刽子手Caspase）的活性比单独悬浮培养时增加。shRNA Np8-1分别提高了克隆A和CX-1中Caspase 8和Caspase 9的活性。有趣的是，与转导细胞相邻的非转导细胞与转导细胞具有相似水平的caspase 9活性。最后，至少在CX-1中。Caspase 9及其通过线粒体扩增参与细胞凋亡的内在途径对shNp8-1和shNG-1介导的细胞凋亡具有重要意义。我们的研究表明，1)抑制Nanog或NanogP8可抑制其他核心胚胎tf SOX2和Oct4, 2) NanogP8转录物可被翻译，使其成为一个功能基因，3)LV shRNA对Nanog或NanogP8的转导细胞在3- d中生长，4)MOI 8足以抑制结直肠癌的短期生长，5)LV shRNA对生长的抑制是通过激活caspase介导的，尤其是caspase 9和caspase 3。当国防部IDEA拨款获得资助后，体内实验将继续进行。最后，目前尚不清楚为什么一些肿瘤只产生Nanog转录本，而另一些肿瘤既产生NanogP8也产生NanogP8，或者只产生NanogP8。然而，通过产生针对这两个基因的等位基因特异性抑制shRNA，可以针对其中一个或两个基因，并提供个性化治疗的可能性。虽然我们还没有确定在体外悬浮培养模型中，这两种等位基因特异性抑制shrna的组合是否比单独使用更好，但我们很快就会这样做。这一点很重要，特别是因为抑制任何一种基因都是通过细胞凋亡的内在途径起作用的，这与对疾病很重要的外在途径是不同的。

Nanog Regulates Other Stem Cell Transciption Factors Nanog controls the expression of Oct4 and SOX2 during early embryogenesis. We have found that inhibition of NanogP8 with allele specific shRNA Np8-1 inhibits expression of Oct4 and SOX2. Thus, targeting one factor may inhibit the others so that single agent targeting may effect multiple downstream pathways. In addition, we have shown that Nanog and NanogP8 affect stemness by reducing spherogenicity, tumorigenicity and metastatic potential in colorectal carcinoma cells (CRC). We have now found that allele-specific inhibition of NanogP8 reduces the size of the side population, while overexpression of Nanog or NanogP8 increases the side population. Thus, Nanog and NanogP8 are instrumental in modulating the stemness of CRC. Nanog and NanogP8 are expressed in clinical liver CRC metastases Ten de-identified samples of colorectal carcinoma resected from patients treated at the NIH Clinical Center were analyzed for expression of Nanog and NanogP8 transcripts and protein. RT-PCR was followed by cutting the cDNA with the restriction endonuclease AlwNI that specifically cuts NanogP8 but not Nanog. Seven liver metastases contained a Nanog-related transcript as did 3 adjacent liver specimens. Six of the 7 specimens contained NanogP8 with and one only Nanog. One of the adjacent liver specimens contained NanogP8 while the 2 others only expressed Nanog transcripts. All results were confirmed by Sanger sequencing. Immunohistochemistry (IHC) was also performed and confirmed that Nanog protein was present in the cytoplasm of the tumor cells that contained detectable Nanog or NanogP8 transcripts but not in CRC that lacked such transcripts. In addition, IHC of sample 9 demonstrated the presence of Nanog protein in a sample where only NanogP8 transcripts were identified. Thus, based on this small sample that needs confirmation and expansion approximately 70% of clinical CRC express a Nanog family member and 60% express NanogP8. NanogP8 Can be Translated into Protein In collaboration with the Veenstra laboratory at NCI-Frederick, we attempted to use tandem MS/MS to identify the NanogP8 protein in CRC cell lines. Extracts of CRC cell lines were immunoprecipitated, isolated by SDS-PAGE, subjected to in gel tryptic digestion and then MS/MS. Four Nanog-related peptides (KTWFQNQRM, KYLSLQQMQELSNILNLSYKQ, KKEDKVPVKK, and KGKQPTSAENSVAKK) were identified in extracts from Clone A overexpressing NanogP8. Interestingly, the last peptide is unique to NanogP8, which includes the shift from Lys (K) in Nanog to Asn (N) in NanogP8 at codon 82. This amino acid change was also reported by Ambady (2) and is consistent with our gene sequencing data. Nanog/NanogP8 proteins were not identified in extracts of CRC cell lines without overexpression by transfection presumably because CRC cell lines express low levels of endogenous proteins. NanogP8 Promoter Activity is increased during 3-D growth consistent with gene expression of NanogP8 We have previously shown that NanogP8 transcripts are increased in spheroids of the CRC cell lines Clone A and CX-1 cultured in serum-free medium (SFM). We tested the postulate that the promoter activity of the NanogP8 gene increases as CRC cells transition from a two-dimensional monolayer culture to a three-dimensional spheroid culture. A lentiviral NanogP8 promoter reporter was obtained from Jeter et al (3). Monolayer cultures of Clone A and CX-1 had low levels of GFP expression compared to expression in spheroids of Clone A and CX-1. Counting GFP positive cells demonstrated a 6- to 14-fold increase in GFP+ cells in spheroids compared to monolayers for Clone A and CX-1, respectively. This is similar to the increase in transcript number assessed by qRT-PCR during transition from 2-D monolayers to 3-D spheroids. Mechanism of Action and Inhibition of 3-D Growth By Transduction with Lentiviral shRNAs to Nanog or NanogP8 To be a successful therapy, inhibition of Nanog by shRNA must inhibit the growth of established tumors in vivo in preclinical models. We obtained a grant from the Department of Defense to support such preclinical work but unfortunately that has not yet started. However, we have been busy performing experiments in vitro with our allele-specific shRNAs and their effects on regression of established 3-D spheroids and cells in suspension to determine the multiplicity of infection (MOI) needed to cause tumor regression. Our previous experiments demonstrated that treatment of CRC in monolayer induced Annexin V expression as an indication of apoptosis. Now we started by taking LV with shRNA to NanogP8 (shNp8-1), Nanog (shNG-1) and a control shRNA (shNEG) and transducing CX-1 and Clone A cells that were then cultured for 24 hr in monolayer and then transitioned to SFM. We first demonstrated that LVshNp8-1 transduction significantly reduced the mass of cells after 72 hr of suspension culture. We then began to transduce CRC growing in suspension in SFM on ULLA plates. Since our LV preparations express GFP, we first determined the fraction of cells transduced in suspension and found that 40% to 75% of CRC cells were transduced stably. We then assessed whether transduction with LV shRNA to Nanog or NanogP8 inhibited growth compared to either untreated CRC cells or cells treated with the control shNEG. The cumulative results of more than 7 different experiments indicated that treatment of CRC in 3-D growth significantly inhibits short term growth by 50% by 72 hr. We then sought to determine how the LV shRNA effects apoptosis in 3-D culture. We previously demonstrated that anoikis apoptosis caused by suspension culture is dependent on activation of Caspase 8 (4). We measured Caspase 3, 8 and 9 activity in Clone A and CX-1 treated with LV shRNA after 4 days of suspension culture, 3 days after LV treatment. Activity of Caspase 3, the executioner caspase, was increased in both CX-1 and Clone A by shRNA to Nanog and NanogP8 beyond what occurs with just suspension culture alone. Activity of Caspase 8 or 9 was increased in Clone A or CX-1, respectively, by shRNA Np8-1. Interestingly, non-transduced cells cells adjacent to transduced cells had similar levels of caspase 9 activity as transduced cells did. Finally, at least in CX-1. Caspase 9 and the intrinsic pathway that uses mitochondrial amplification in apoptosis is important for apoptosis mediated by shNp8-1 and shNG-1. SUMMARY Our work suggests that 1) inhibition of Nanog or NanogP8 inhibits the other core embryonic TFs SOX2 and Oct4, 2) the NanogP8 transcript can be translated so that the retrogene is a functional gene, 3) LV shRNA to Nanog or NanogP8 transduce cells growing in 3-D, 4) an MOI of 8 is sufficient to inhibit CRC short term growth, and 5) inhibition of growth by LV shRNA is mediated by activation of caspases, especially caspase 9 and 3. In vivo experiments will go forward when the DOD IDEA grant is funded. Finally, it is not clear why some tumors produce only Nanog transcripts while others produce either both NanogP8 and Nanog or only NanogP8. However, by producing allele-specific inhibitory shRNA to both genes it is possible to target either or both and offers the possibility of personalizing therapy. While we have not yet determined whether the combination of both allele specific inhibitory shRNAs is better than either alone in the in vitro suspension culture model, we will do this soon. This is important especially since inhibition of either gene works through the intrinsic pathway of apoptosis that is distinct from the extrinsic pathway that is important for anoikis.