A.) Aim 1 was completed in year 1 B.) Aim 2 Using our Single Cell Spherogenicity Assay (SCSA), we completed our analysis of spherogenicity in our human CRC lines and found that spherogenicity generally parallels metastatic potential: LS 174T greater than CX-1 = Mip-101 greater than Clone A = KM-12c and HCC 2998. Expression of the core embryonic genes Nanog, OCT4, and SOX2 was analyzed in spheroids and monolayer cultures by qRT-PCR, normalized by the expression of GAPDH and expression in spheroids compared to the expression in monolayer cultures of each cell lines as a relative expression and indicates that Nanog gene expression is consistently increased in spheroids compared to the expression in that lines monolayer culture while the expression of the other core embryonic genes OCT4 and SOX2 is more variable. OCT4 is a tightly regulated gene in embryonic stem cells and this does not seem to change in these CRC lines since multiple qRT-PCR experiments indicate that the level of OCT4 transcripts in spheroids only ranges between 5- and 9-fold increased over basal levels in monolayer cultures of each CRC line. However, the levels of Nanog gene transcripts ranges from 8-fold increase over monolayer in Clone A to 300-fold in KM-12c when normalized to GAPDH. SOX2 transcript levels also were quite different among the CRC with CX-1 and HCC 2998 expressing low levels by normalized qRT-PCR in monolayer and spheroid, Clone A had a 10-fold change in SOX2 expression and MIP-101 and LS 174T had a 50 - to 60-fold increase in SOX2 levels compared to monolayer cultures. These data suggest that the relative of Nanog, SOX2 and OCT4 may be critical determinants of the malignant properties of human CRC. With this knowledge about the variability of expression of the 3 embryonic genes, we then began to focus on Clone A, CX-1 and LS 174T for subsequent research. This was done by specific gene knock down with lentiviral vector mediated shRNA to each of the 3 genes. Stable transductants were created for proof of principle experiments in which expression of each target gene was inhibited by 60% or more by qRT-PCR. Then transductants and vector controls were assessed first in the SCSA in vitro and then in vivo in NOD/SCID mice. in the SCSA, shRNA to Nanog consistently reduced spherogenicity by 94% in Clone A, 70% in CX-1 and 52% in LS 174T. In contrast, shRNA to OCT4 or SOX2 actually increased spherogenicity by over 200% in CX-1 and 500% in Clone A. However, in LS174T that had elevated levels of SOX2 gene expression, shRNA to OCT4 or SOX2 reduced spherogenicity by 76% and 52% respectively. These novel findings are real since shRNA to OCT4 or SOX2 significantly shortened time to appearance of tumors after injection subcutaneously in NOD/SCID mice with CX-1 and Clone A (with shRNA to SOX2). However, shRNA to OCT4 or SOX2 prolonged the mean time to appearance of tumors with LS 174T by nearly 50% compared to the parental controls. shRNA to Nanog significantly reduced tumorigenicity by prolonging time to appearance or reducing tumor takes on limiting dilution analysis in all 3 CRC tumors. In addition shRNA completely prevented liver colony formation after intrasplenic injection of CX-1 in an experimental metastasis assay. Thus, shRNA to Nanog has a consistent inhibition of malignant properties of CRC whereas shRNA to OCT4 or SOX2 do not. Also analysis of transductants demonstrates that shRNA to Nanog inhibits OCT4 and SOX2 gene expression by 50% but that inhibition of OCT4 or SOX2 does not affect Nanog gene expression. Finally, we have developed a screening assay to detect which of the Nanog family members is activated in human CRC. In order to identify which gene product is expressed in human CRC, we sought to identify a site in NanogP8 around one of the 5 single nucleotide changes that might be amenable to endonuclease digestion. A search of databases revealed that the only specific commercially available endonuclease that was predicted to cut NanogP8 cDNA but not Nanog was 143 nt 3 downstream to the ATG by AlwNI which cuts at a 5'... C A G N N N C T G ... 3' 3'... G T C N N N G A C ... 5'sequence. Primers were designed to amplify both Nanog and NanogP8 and then RT-PCR was performed on total RNA isolated from Clone A and CX-1 cells growing under basal conditions in monolayer with serum-containing medium and in spheroids in ULLA plates and serum-free medium. The assay produces two cleavage products on a gel when NanogP8 is cleaved by AlwNI but no cleavage products when only Nanog is present. Monolayer cultures of CRC express both Nanog and NanogP8 and rarely NanogP7 but the spheroids express predominantly NanogP8. This was confirmed by direct sequencing of 20 30 clones of individual, separate cDNAs from CX-1 and Clone A that demonstrated only NanogP8 was expressed in spheroid cultures whereas 10 -12.5% of monolayer cDNAs contained Nanog or NanogP7 transcripts. With this we began to focus on therapy that would specifically target only NanogP8 and not Nanog. This is important for clinical studies since the function of Nanog and the other core embryonic genes occur during pre- and early implantation before the woman can know she is pregnant. B.) Aim 3 - To develop therapies that inhibit NANOG expression in vivo. We originally had planned to develop an assay to develop drugs that would target Nanog. However, the uncertainty of the mechanism of action of Nanog in carcinoma cells and now the realization that the active gene is NanogP8 has caused us to rethink our original plan of attack. As described below, the amount of Nanog or NanogP8 gene transcripts even in cells enriched for stemness is quite low. As a result, gene therapy with lentiviral vector delivered shRNA may inhibit Nanog sufficiently inhibit malignant potential. We have designed allele-specific shRNA that knocks down NanogP8 and not Nanog and are in the process of determining whether NIH will support a patent application for this. the specific shRNA to NanogP8 inhibits spherogenicity of Clone A, CX-1 and even the aggressive LS 174T. Our plan for the future is to determine whether lentiviral vector inhibits growth of spheroids that have been created in serum-free medium under ULLA conditions and then if it does (as a model of inhibition of solid tumor masses) we will test the ability of this shRNA to inhibit established liver colonies of CX-1 and LS 174T in NOD/SCID mice. We have also developed luciferase-expressing constructs of all three CRC that will be used to facilitate both the in vivo testing as well as to make a high throughput screen for several other shRNAs that we have in the pipeline. Several manuscripts are in preparation and will be submitted in the coming year once the status of the patent application is clear.
