Table 1.
Candidate HKG genes.
Fig 1.
Characterization of sarcospheres obtained from MG-63.
Representative images of adherent native cells and CSC floating spheres observed by inverted optical microscope for the different cell lines (scale bar 100 μm, left panel), and gene expression of stem cell markers by qRT-PCR (right panel). Normalization to ACTB. For cscMG-63, c-Myc **p = 0.0019, KLF4 p = ns, Nanog **p = 0.0010, OCT3/4 *p = 0.0265. For cscMDA-MB-231, c-Myc **p = 0.0011, KLF4 *p = 0.0130, Nanog **p = 0.0011, OCT3/4 *p = 0.0325. For cscACHN, c-Myc p = ns, KLF4 *p = 0.0130, Nanog *p = 0.0130. For adherent ACHN, OCT3/4 *p = 0.0130. (n = 6–12).
Fig 2.
Transcription profiling of the selected reference genes.
(A) Heat map showing the relative expression of the selected genes in native cells and CSC from MG-63, HOS, and SAOS-2. (B). Transcriptional profile of Cycle threshold (Ct) values of candidate HKG genes in CSC and native cell lines. Boxes represent lower and upper quartiles of cycle threshold range with the median indicated, vertical bars represent the 10th and 90th percentiles. In both CSC and adherent cell lines, 18S rRNA was the most highly expressed gene (lower Ct value), whereas G6PD was the least expressed gene (highest Ct value).
Table 2.
Transcriptome data from deep sequencing analysis of CSC and native cells of osteosarcoma.
Table 3.
Ct values of candidate HKG genes in CSC and native cell lines.
Table 4.
Ranking of the stability of the expression of candidate reference genes by NormFinder, geNorm, and CV analyses.
Fig 3.
GeNorm evaluation of the minimal number of genes and validation of the identified top-ranked HKG genes for qRT-PCR normalization.
The minimal number of genes required for data normalization was evaluated by pairwise variation (Vn/n +1) in (A) CSC, (B) native cells, (C) in the CSC and native cell lines from all tumors, (D) from sarcoma and (C) from carcinomas. A variation coefficient (V) below 0.15 indicates the optimal number of genes required for data normalization. V2/3<0.15 indicates that 2 genes are required for data normalization.
Table 5.
Ranking of the stability of the expression of candidate reference genes by NormFinder, geNorm, and CV analyses in CSC and native cells from carcinoma and sarcoma tumors.
Fig 4.
Validation of the identified housekeeping genes.
The effect of gene expression normalization with optimal HKG was investigated on CSC and native cells from MG-63 and ACHN. (A) For the osteosarcoma cell line, the expression of the stem cell markers Nanog and cMyc were evaluated and normalized to the geometric mean of GAPDH and YWHAZ. ***p = 0.0007 for Nanog, ***p = 0.0003 for c-Myc. (B) For the ranal carcinoma cell line, the expression of Nanog and cMyc was normalized to the geometric mean of PPIA and HMBS. **p = 0.0043 for Nanog. (n = 6).