Limits...
A bioinformatics approach reveals novel interactions of the OVOL transcription factors in the regulation of epithelial - mesenchymal cell reprogramming and cancer progression.

Roca H, Pande M, Huo JS, Hernandez J, Cavalcoli JD, Pienta KJ, McEachin RC - BMC Syst Biol (2014)

Bottom Line: Focusing on the target genes for these four TFs plus the OVOLs, we produced the OI-MET-TF sub-model, which shows even greater enrichment for these annotations, plus significant evidence of cooperation among these five TFs.Reflecting these results back to the OI-MET model, we found that binding motifs for the TF pair AP1/MYC are more frequent than expected and that the AP1/MYC pair is significantly enriched in binding in cancer models, relative to non-cancer models, in these promoters.This effect is seen in both MET models (solid tumors) and in non-MET models (leukemia).

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Computational Medicine and Bioinformatics, University of Michigan, Ann Arbor, MI, USA. kpienta1@jhmi.edu.

ABSTRACT

Background: Mesenchymal to Epithelial Transition (MET) plasticity is critical to cancer progression, and we recently showed that the OVOL transcription factors (TFs) are critical regulators of MET. Results of that work also posed the hypothesis that the OVOLs impact MET in a range of cancers. We now test this hypothesis by developing a model, OVOL Induced MET (OI-MET), and sub-model (OI-MET-TF), to characterize differential gene expression in MET common to prostate cancer (PC) and breast cancer (BC).

Results: In the OI-MET model, we identified 739 genes differentially expressed in both the PC and BC models. For this gene set, we found significant enrichment of annotation for BC, PC, cancer, and MET, as well as regulation of gene expression by AP1, STAT1, STAT3, and NFKB1. Focusing on the target genes for these four TFs plus the OVOLs, we produced the OI-MET-TF sub-model, which shows even greater enrichment for these annotations, plus significant evidence of cooperation among these five TFs. Based on known gene/drug interactions, we prioritized targets in the OI-MET-TF network for follow-on analysis, emphasizing the clinical relevance of this work. Reflecting these results back to the OI-MET model, we found that binding motifs for the TF pair AP1/MYC are more frequent than expected and that the AP1/MYC pair is significantly enriched in binding in cancer models, relative to non-cancer models, in these promoters. This effect is seen in both MET models (solid tumors) and in non-MET models (leukemia). These results are consistent with our hypothesis that the OVOLs impact cancer susceptibility by regulating MET, and extend the hypothesis to include mechanisms not specific to MET.

Conclusions: We find significant evidence of the OVOL, AP1, STAT1, STAT3, and NFKB1 TFs having important roles in MET, and more broadly in cancer. We prioritize known gene/drug targets for follow-up in the clinic, and we show that the AP1/MYC TF pair is a strong candidate for intervention.

Show MeSH

Related in: MedlinePlus

STAT3 sub-network. The STAT3 network, including STAT3 and its MeSH annotated targets. All of the nodes are connected, though MeSH association does not necessarily indicate direct binding.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4008156&req=5

Figure 3: STAT3 sub-network. The STAT3 network, including STAT3 and its MeSH annotated targets. All of the nodes are connected, though MeSH association does not necessarily indicate direct binding.

Mentions: The AP1, STAT3, and STAT1 networks each include all of the input genes in a very simple, parsimonious, network (Figures 2, 3, and 4). This is consistent with what was expected for the AP1 network because the genes in this set are annotated as being direct AP1 binding targets in TransFac annotation. Note that Additional file 3 is the key for interpreting GeneGo graphics and that the icon labeled “AP1 (FOS/JUN)” represents the dimer of FOS and JUN gene family members in a single icon. Genes in the STAT1 and STAT3 networks are found in MeSH annotation and, while all the genes are in the network, they are not all direct targets of the TF. The NFKB1 network (Figure 5), also derived from MeSH annotation, illustrates that the annotation does not necessarily indicate direct interaction with the TF. Rather, using the same parameter settings as for the other networks, NGFR, CARD6, and NALP3 are disconnected genes. Also, this network includes NFKBIA, which interacts closely with, but is distinct from the NFKB1 dimer. Note that GeneGo used two icons for NFKB1, but we collapsed them into a single rectangular icon in this graphic. It’s possible that a more complex (less parsimonious) NFKB1 network would connect all the genes in the NFKB1 set, but our hypothesis is that these four TFs work together in regulating the genes differentially expressed in OI-MET. Therefore, we developed the network for the combined set of genes targeted by the four TFs using the parameter settings for the parsimonious network. The network we found (Figure 6) is consistent with this hypothesis; it connects all the genes and includes only one gene that was not part of the input set (the aforementioned NFKBIA).


A bioinformatics approach reveals novel interactions of the OVOL transcription factors in the regulation of epithelial - mesenchymal cell reprogramming and cancer progression.

Roca H, Pande M, Huo JS, Hernandez J, Cavalcoli JD, Pienta KJ, McEachin RC - BMC Syst Biol (2014)

STAT3 sub-network. The STAT3 network, including STAT3 and its MeSH annotated targets. All of the nodes are connected, though MeSH association does not necessarily indicate direct binding.
© Copyright Policy - open-access
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4008156&req=5

Figure 3: STAT3 sub-network. The STAT3 network, including STAT3 and its MeSH annotated targets. All of the nodes are connected, though MeSH association does not necessarily indicate direct binding.
Mentions: The AP1, STAT3, and STAT1 networks each include all of the input genes in a very simple, parsimonious, network (Figures 2, 3, and 4). This is consistent with what was expected for the AP1 network because the genes in this set are annotated as being direct AP1 binding targets in TransFac annotation. Note that Additional file 3 is the key for interpreting GeneGo graphics and that the icon labeled “AP1 (FOS/JUN)” represents the dimer of FOS and JUN gene family members in a single icon. Genes in the STAT1 and STAT3 networks are found in MeSH annotation and, while all the genes are in the network, they are not all direct targets of the TF. The NFKB1 network (Figure 5), also derived from MeSH annotation, illustrates that the annotation does not necessarily indicate direct interaction with the TF. Rather, using the same parameter settings as for the other networks, NGFR, CARD6, and NALP3 are disconnected genes. Also, this network includes NFKBIA, which interacts closely with, but is distinct from the NFKB1 dimer. Note that GeneGo used two icons for NFKB1, but we collapsed them into a single rectangular icon in this graphic. It’s possible that a more complex (less parsimonious) NFKB1 network would connect all the genes in the NFKB1 set, but our hypothesis is that these four TFs work together in regulating the genes differentially expressed in OI-MET. Therefore, we developed the network for the combined set of genes targeted by the four TFs using the parameter settings for the parsimonious network. The network we found (Figure 6) is consistent with this hypothesis; it connects all the genes and includes only one gene that was not part of the input set (the aforementioned NFKBIA).

Bottom Line: Focusing on the target genes for these four TFs plus the OVOLs, we produced the OI-MET-TF sub-model, which shows even greater enrichment for these annotations, plus significant evidence of cooperation among these five TFs.Reflecting these results back to the OI-MET model, we found that binding motifs for the TF pair AP1/MYC are more frequent than expected and that the AP1/MYC pair is significantly enriched in binding in cancer models, relative to non-cancer models, in these promoters.This effect is seen in both MET models (solid tumors) and in non-MET models (leukemia).

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Computational Medicine and Bioinformatics, University of Michigan, Ann Arbor, MI, USA. kpienta1@jhmi.edu.

ABSTRACT

Background: Mesenchymal to Epithelial Transition (MET) plasticity is critical to cancer progression, and we recently showed that the OVOL transcription factors (TFs) are critical regulators of MET. Results of that work also posed the hypothesis that the OVOLs impact MET in a range of cancers. We now test this hypothesis by developing a model, OVOL Induced MET (OI-MET), and sub-model (OI-MET-TF), to characterize differential gene expression in MET common to prostate cancer (PC) and breast cancer (BC).

Results: In the OI-MET model, we identified 739 genes differentially expressed in both the PC and BC models. For this gene set, we found significant enrichment of annotation for BC, PC, cancer, and MET, as well as regulation of gene expression by AP1, STAT1, STAT3, and NFKB1. Focusing on the target genes for these four TFs plus the OVOLs, we produced the OI-MET-TF sub-model, which shows even greater enrichment for these annotations, plus significant evidence of cooperation among these five TFs. Based on known gene/drug interactions, we prioritized targets in the OI-MET-TF network for follow-on analysis, emphasizing the clinical relevance of this work. Reflecting these results back to the OI-MET model, we found that binding motifs for the TF pair AP1/MYC are more frequent than expected and that the AP1/MYC pair is significantly enriched in binding in cancer models, relative to non-cancer models, in these promoters. This effect is seen in both MET models (solid tumors) and in non-MET models (leukemia). These results are consistent with our hypothesis that the OVOLs impact cancer susceptibility by regulating MET, and extend the hypothesis to include mechanisms not specific to MET.

Conclusions: We find significant evidence of the OVOL, AP1, STAT1, STAT3, and NFKB1 TFs having important roles in MET, and more broadly in cancer. We prioritize known gene/drug targets for follow-up in the clinic, and we show that the AP1/MYC TF pair is a strong candidate for intervention.

Show MeSH
Related in: MedlinePlus