i.) Abstract Nanog is one of several core transcription factors that maintain embryonic stem cells in a pluripotent, dedifferentiated state. Recent data indicate that expression of NANOG or its nearly identical retrogene NANOGP8 is increased in human gastric, hepatocellular and colon cancer. Our postulate is that NANOG/NANOGP8 is critical for the functions of putative CSC in colorectal carcinoma (CRC): self-renewal, pluripotency, spherogenicity and resistance to chemotherapy. Five human CRC lines were used in this study (HCC 2998, KM-12c, Clone A, Mip-101, CX-1). Anoikis (programmed cell death in suspension culture) was measured after 96 hours of suspension culture by TUNEL analysis of cells sorted for expression of ALDH1, CD44 or CD133 - alone or in combination-. No consistent pattern of resistance to anoikis was identified among the CRC lines. However, NANOG and/or Oct4 expression was increased in suspension culture and associated with resistance to anoikis. qRT-PCR analysis of spheroids of CX-1 and Clone A cultured in suspension in serum free medium indicated that NANOG transcript levels were increased by 43- and 12-fold respectively relative to monolayer cultures while the relative increase in transcript levels of Oct4, Sox2, Klf4, ABCG2, ABCA3, ALDH1A1, c-myc, CD44, CD133, CD166 and β-catenin was less than 5-fold except for Sox2 in Clone A which was 10-fold. CRC stably transfected with a GFP NANOG promoter were low expressors in monolayer culture, but expressed GFP as they transitioned to vertical growth and formed spheroids. Stable transfection with lentiviral NANOG shRNA decreased NANOG transcript levels and spherogenicity of CX-1 by 65% compared to vector controls in a single cell spherogenicity assay in serum-free medium. Finally, fewer NANOG shRNA transfected cells survived exposure to 0.3 M Topotecan for 48 than vector controls or parental cells. In summary, our data suggest that expression of NANOG/NANOGP8 occurs in human CRC and may be critical for such functions of putative CSC as spherogenicity and resistance to chemotherapy. ii.) Progress We developed an unbiased approach to isolating putative CSC by culturing cells in serum free medium supplemented with EGF and bFGF (Neurocult, Stem Cell Technologies) under low attachment conditions. We have primarily used two human CRC lines for this work: Clone A, a poorly differentiated colon cell line that is tumorigenic in the sub cutis but weakly metastatic after intrasplenic injection into athymic nude mice, and CX-1, a variant of HT-29 that is moderately differentiated, highly tumorigneic and metastatic. Initial attempts to reproducibly develop a flow cytometry based method of isolating a side population by cell size and dye exclusion were unsuccessful because the transporter function of the CRC was too high. A single cell spherogenic assay was developed in which single cells are placed in individual wells of a 96 well microtiter plate by limiting dilution in serum-free medium and then the % of cell forming spheroids are assessed at 1 - 2 weeks. CX-1 was more aggressive than Clone A in this assay with 33% of single cells forming spheroids compared to less than 8% of Clone A cells. No difference in plating efficiency between the CRC lines but p<0.0004 for spherogenicity between the 2 CRC lines. Spheroids formed by the cell lines Clone A and CX-1 increase the gene expression of the core pluripotent TFs, especially NANOG that is increased 12 - 43-fold relative to the expression in monolayer culture. Increases in gene expression of transporter, cell membrane proteins or other core pluripotent TFs were not to the same degree. Western blots have been performed that demonstrate commensurate changes in expression of transporter and cell adhesion proteins. However, the antibodies for NANOG do not work well in western blot and are not presented. As an alternative we created stable constructs in Clone A and CX-1 of a NANOG promoter-driven GFP to confirm upregulation of NANOG expression in spheroids (described in Project 2). Once we confirmed that gene expression of NANOG was upregulated in spheroids, we then began to develop a lentiviral shRNA approach to silencing NANOG. Using the open source shRNA we isolated stably transfected clones of Clone A and CX-1 that were had at least a 70% decrease in NANOG gene expression. When we compared these clones to vector controls and the parental cell lines in the single cell spherogenicity assay, inhibition of NANOG expression decreased spherogenicity in CX-1. Work is in progress with Clone A. The effect of NANOG expression on resistance to chemotherapy has been preliminarily studied. The agent that has been used here is Topotecan (TOPO), a camptothecin that is similar to irinotecan that is used in colorectal carcinomas in patients. Preliminary experiments indicate that inhibition of NANOG expression increases sensitivity to TOPO in CRC. A complicating issue is that inhibition of NANOG causes S phase arrest and future aspects of this project will determine how inhibition of cell cycle affects sensitivity to Topotecan. iii.) Future Directions For the immediate future we are planning to assess the effect of NANOG on both tumorigienicity in the sub cutis and on metastatic potential after intrasplenic injection in NOD/SCID mice. Cells will be injected in a dilution series and experiments will be repeated at least once. Malignant growth assays in vivo will be carried out without matrigel so as to establish the role of NANOG in unmodified tumor growth. In addition, lentiviral shRNA to Oct4 and SOX2 the other major core pluripotent TFs will be stably transfected into the CRC and their effect on CSC assessed and compared to that obtained by silencing NANOG. After the biology of NANOG and other core pluripotent TFs is established in these CRC lines, we will begin aim 3 by creating an assay that may be used to screen compounds that inhibit NANOG binding to its DNA sequences in promoters. If successful, we will propose to the NExT program that the assay be used to screen chemical libraries for inhibitory agents and then if an agent is found it will be tested for ability to inhibit CSC function both in vitro and in vivo.
