ABT-263 br B Schematic summary of CSC CIC
(B) Schematic summary of CSC/CIC heterogeneity.
Genes with high expression levels in LL clone ABT-263 compared to the levels in sphere clone, serum-cultured cells and sphere-culture cells.
Gene name Locus LL clone FPKM S clone FPKM Serum cultured FPKM Sphere cultured FPKM reference sequence
are related to tumorigenicity (Ginestier et al., 2007; Kuroda et al., 2013; Mariya et al., 2016). Our results showed that LL clone cells express higher expression level of SOX2 and higher ratio of ALDHhigh cells. In contrast, S clone cells showed no of tumorigenicity despite having a higher expression level of SOX2 and higher ratio of ALDHhigh cells than
those of parental serum-cultured cells. These results indicate that SOX2 and ALDHhigh are not only factors that define tumorigenicity. The cells
were isolated from patient primary tumor who received chemotherapy before surgical resection. The tumor showed partial response (PR) to chemotherapy, thus remaining cancer cells should be resistant to che-motherapy. Therefore, chemo-resistant but low tumorigenic S clone cells might be isolated in this analysis.
Genes with high expression levels in S clone and sphere-cultures cells compared to LL clone and serum-cultured cells.
Gene name Locus LL clone FPKM S clone FPKM Serum culture FPKM Sphere culture FPKM Reference sequence
FPKM (Fragments Per Kilobase of exon per Million fragments mapped) indicates the number of fragments per sample which were read by the reference sequence.
Carboplatin (CBDCA) inhibits DNA synthesis and Paclitaxel (PTX) inhibits the microtubule assembly, and suppress cell proliferation (Jiang et al., 2015). Therefore, these drugs target fast growing cancer cells well, and LL clones, which have thigh proliferation ability, are very sensitive to chemotherapeutic reagents. Since S clone I cells are very slow-growing cells in a dormant state, this is one factor that might explain the chemo-resistance of S clone I cells. However, there was no significant diﬀerence between the cell proliferation ability of S clone I cells and that of sphere-cultured cells, whereas S clone I cells showed greater resistance than sphere-cultured cells to CBDCA and PTX. We previously reported that MMP10 is involved in CSCs/CICs and is related to chemo-resistance (Mariya et al., 2016). Therefore, a high expression level of MMP10 in S clone I cells might the reason for chemo-resistance. Interestingly, a high expression level of SOX2 is not related to chemo-resistance in this study.
Estimated enriched pathway analysis revealed the diﬀerences be-tween features of genes with high expression levels in S clone cells and
LL clone cells. Molecules involved in tissue development pathways (such as embryo, skeletal, sensory organ and mesoderm development) are enriched in S clone cells, and S clone cells have a tendency to have stem cell features. On the other hand, pathways related to cell pro-liferation are enriched in LL clone cells. Second-messenger signaling, which is related to cell proliferation and diﬀerentiation is enriched in LL clone cells. Recently, a strategy for cancer therapy targeting calcium signaling which is second-messenger signaling has been reported (Cui et al., 2017). MAPK signaling has been reported to be related to tumor progression and it can be a target of cancer therapy (Cheng et al., 2013). Genes related to the MAPK pathway are enriched in LL clone cells. These pathways might have a role in the higher tumorigenicity of LL clone cells.
CSCs/CICs have been reported to have high levels of tumorigenicity and chemo-resistance (Hirohashi et al., 2010; Murase et al., 2009; Park et al., 2009; Visvader and Lindeman, 2008). However, we could not explain these features of CSCs/CICs by single gene or one clone cells, even by analysis of the sample from the patient. We think that the degree of malignancy and the prognosis for a patient are determined not by one gene or one clone cells but by multiple genes or multiple phenotypical CSCs/CICs. In this study, the phenotypes as carcinoma were shared by multiple clones, including LL clone cells as tumor-in-itiating clones and S clones as chemo-resistant cells (Fig. 5B), and the gene expression profiles were quite diﬀerent among the clones. Therefore, analysis of a sample from one patient at clonal level, might
be nucessary to elucidate the complexity of CSC/CIC phenotypes.
In summary, we established several clone cells from a primary en-dometrial adenocarcinoma sample and analyzed their functions, and we found that heterogenous CSC/CIC includes tumorigenic clone cells and chemo-resistant clone cells.
The authors have no conflict of interest.
Author contribution statement