Cross-platform expression profiling demonstrates that SV40 small tumor antigen activates Notch, Hedgehog, and Wnt signaling in human cells

Abstract

Background: We previously analyzed human embryonic kidney (HEK) cell lines for the effects that simian virus 40 (SV40) small tumor antigen (ST) has on gene expression using Affymetrix U133 GeneChips. To cross-validate and extend our initial findings, we sought to compare the expression profiles of these cell lines using an alternative microarray platform.

Methods: We have analyzed matched cell lines with and without expression of SV40 ST using an Applied Biosystems (AB) microarray platform that uses single 60-mer oligonucleotides and single-color quantitative chemiluminescence for detection.

Results: While we were able to previously identify only 456 genes affected by ST with the Affymetrix platform, we identified 1927 individual genes with the AB platform. Additional technical replicates increased the number of identified genes to 3478 genes and confirmed the changes in 278 (61%) of our original set of 456 genes. Among the 3200 genes newly identified as affected by SV40 ST, we confirmed 20 by QRTPCR including several components of the Wnt, Notch, and Hedgehog signaling pathways, consistent with SV40 ST activation of these developmental pathways. While inhibitors of Notch activation had no effect on cell survival, cyclopamine had a potent killing effect on cells expressing SV40 ST.

Conclusions: These data show that SV40 ST expression alters cell survival pathways to sensitize cells to the killing effect of Hedgehog pathway inhibitors.

Figure 1

(A) Venn diagram of the number of genes identified as changed between HEK-TERST and HEK-TERV cells using the AB and Affymetrix platforms and an equivalent number of replicates. Approximately 40% of previously identified genes were verified with equal replicates and 4.2 times as many genes were significant on the AB platform. (B) Venn diagram of the number of genes identified including two additional technical replicates for each cell line. Approximately 61% of previously identified genes were verified with these data, and 7.6 times as many genes were significant on the AB platform. (C) Venn diagram of the 548 re-analyzed Affymetrix genes and the 1927 AB genes without technical replicates. (D) Venn diagram of the 548 re-analyzed Affymetrix genes and the 3478 AB genes with technical replicates.

Figure 2

(A) Scatterplot comparison of the log2 ratio of gene expression data between HEK-TERST and HEK-TERV cells. Log2 ratios from the AB platform are plotted on the y-axis and from the Affymetrix platform on the x-axis. The genes plotted include any gene identified as significantly changed on either platform. (B) Scatterplot comparison similar to panel (A), except that only the 278 genes significantly altered on both platforms are shown. (C) Scatterplot of the log2 ratio of HEK-TERST and HEK-TERV cells comparing different technical replicates on the AB platform. (D) Scatterplot comparison as in panel (C), except comparing biological replicates on the AB platform. (E) Scatterplot comparison of the log2 transformed absolute signal from HEK-TERST cells from two technical replicates on the AB platform. (F) Scatterplot comparison of the log2 transformed absolute signal from HEK-TERST cells from two biological replicates on the AB platform.

Figure 3

(A) QRTPCR validation of the AB Expression Array System measured Fold Change of gene expression data between HEK-TERST and HEK-TERV cells. Fold change values from the AB platform are plotted on the x-axis and TaqMan assay based QRTPCR fold change is plotted on the y-axis. Data for the same set of genes are shown in all panels (A-F). (B) QRTPCR validation of the Affymetrix measured Fold Change of gene expression data between HEK-TERST and HEK-TERV cells. Fold change values from the Affymetrix platform are plotted on the x-axis and TaqMan assay based QRTPCR fold change is plotted on the y-axis. (C) Comparison of SYBR Green reagent based QRTPCR data to the AB Expression Array System fold change data. (D) Comparison of SYBR Green regent based QRTPCR data to the Affymetrix fold change data. (E) Comparison of the SYBR Green regent based QRTPCR data and the TaqMan assay based QRTPCR data. (F) Scatterplot comparison of the Affymetrix measured and AB Expression Array System measured Fold Change of gene expression data between HEK-TERST and HEK-TERV cells.

Figure 4

(A) Plot of the average frequency of HES1 sites in genes upregulated by ST > 2-fold and in a random control list of 200 genes. (B) Sequence logo of the position weight matrix for HES1 sites based on the TRANSFAC database.(C) Immunoblot of whole cell lysates prepared from HEK-TERST and HEK-TERV cells and probed with antibodies specific for the activated human Notch-2. Activated Notch-2 is equally present in both cell lines (right), although longer exposures (left) reveal a shorter isoform that is more abundant in cells expressing SV40 ST.

Figure 5

Impact of SV40 ST on the Notch, Wnt, and Hedgehog Pathways. Schematics of the Notch, Wnt, and Hedgehog pathways are shown. Genes with significantly increased mRNA's in HEK-TERST cells compared to HEK-TERV cells are shown in red, while genes significantly decreased are shown in green. Sets of downstream targets of the Wnt pathway are grouped in boxes, as are sets of c-myc downstream targets. Activation of the Hedgehog pathway due to increased HIP, SMO, and GLI2 levels results in increased Wnt ligand expression. Activation of Wnt signaling increases c-myc, activating and repressing multiple downstream target genes. Wnt pathway activation also feeds back to activate expression of Notch ligands such as DLL1.

Figure 6

Antitumor activities of cyclopamine and γ-secretase inhibitor against HEKTER-LUK and HEKTER-ST cell lines. (A) Survival of HEK-TERV and HEK-TERST cells determined by MTT assay after treatment with vehicle, 2.4μM Cyclopamine or 1 nM γ-secretase inhibitor for 72 hrs. Mean and standard error are shown for each treatment. Percent cell survival was computed relative to untreated cells. (B) Survival of HEK-TERST cells determined by Trypan Blue exclusion assay after treatment of 1 × 10⁵ cells with vehicle, 2.4μM Cyclopamine or 1 nM γ-secretase inhibitor for 72 hrs.

Figure 7

Effect of SV40 ST on Death Receptor Pathways. The diagram is colored as in Figure 5. SV40 ST downregulates pro-apoptotic components of all death receptor pathways and upregulates anti-apoptotic components such as survivin.