Figure 1.
MYC, ATAD2 and 8q24 associations.
(a) MYC expression and (b) MYC signaling are both increased among endometrial cancers with 8q24 amplification. (c) Variations in MYC expression only explain a small proportion of the variation in MYC signaling. Linear fits are shown in red, yellow, and green for the primary investigation series, internal validation series, and external validation series, respectively. (d) The lengths of the amplifications that contain MYC are significantly larger than expected compared to amplifications observed elsewhere in these cancers. (e) Among 26 genes in the 8q24 peak with corresponding expression data, expression of ATAD2 is most strongly and significantly associated with amplification. Blue bars show the percent increase in gene expression and red bars show the p-values. The significance threshold is Bonferroni-corrected for multiple hypotheses. (f) Expression of ATAD2 is highly correlated with MYC signaling strength. Linear fits are shown as in panel c.
Table 1.
Associations between other MYC activation gene sets and ATAD2- and MYC expression.
Figure 2.
Amplification of 8q24, ATAD2 overexpression and increased MYC signaling are associated with poor prognosis.
FISH probes against 8q24 (red) and the chromosome 8 centromere (green) in a primary tumor and the paired metastasis show amplification only in the latter (a) (b) Among 399 patients assessed by FISH, those with 8q24 amplifications have worse outcome. In the primary investigation series, tumors among the highest quartiles of (c) ATAD2 expression and (d) MYC signaling strength also had increased risk of disease-specific death. (e) Estrogen receptor negative (ER−) tumors with ATAD2 expression in the top quartile were also associated with a high risk of disease-specific death; the risk was much lower among estrogen receptor positive (ER+) tumors with ATAD2 expression in the bottom quartile. (f) ATAD2 expression and (g) MYC signaling are both higher among metastases than primary tumors in the internal validation series.
Figure 3.
3-D-plots showing ATAD2 expression and -copy-number and E2F1 expression.
(a) Endometrial cancer, (b) breast cancer, (c) ovarian cancer, and (d) glioblastoma. Yellow dots represent the samples and the blue plate is the predicted 3-D fit. The green and red lines are the distance between the predicted fit and the actual observations for samples above and below the 3D-fit plate, respectively.
Figure 4.
Correlation between effects of ATAD2 and MYC knockdown.
Western blots for (a) ATAD2 and (b) MYC indicate extent of knockdown with six shRNAs against ATAD2 and three shRNAs against MYC, respectively. ATAD2 experiments were performed in KLE cells and MYC experiments were performed in TE9 cells infected with GFP control and MYC vectors. Subsequent experiments used ATAD2 shRNAs a and e, and MYC shRNAs a and b. Reductions in cell viability among seven endometrial cancer cell lines (c) and 21 breast cancer cell lines (d) were highly correlated after knockdown of ATAD2 or MYC and after knockdown of MYC and treatment with the HDAC inhibitor Trichostatin-A (1.25 µM) (e–f).