ERG (ETS-related gene) encodes a member of the ETS family of transcription factors. ERG contains an N-terminal SAM/Pointed (PNT) domain involved in protein-protein interactions and a C-terminal ETS DNA-binding domain that mediates sequence-specific binding to the core consensus 5'-GGA(A/T)-3' (ETS binding site). ERG functions as both a transcriptional activator and repressor depending on context. In vascular endothelium, ERG is a key regulator of endothelial homeostasis, directly activating genes such as VE-cadherin (CDH5) and CLDN5 while repressing inflammatory adhesion molecules like E-selectin (SELE). ERG is essential for vascular development (Erg knockout is embryonic lethal in mice) and lymphatic vessel formation. ERG also interacts with chromatin-modifying enzymes including SETDB1 and KDM4A. Clinically, ERG is highly significant in oncology due to the recurrent TMPRSS2-ERG fusion found in approximately 50% of prostate cancers, where aberrant ERG expression drives oncogenic transcriptional programs. ERG fusions are also found in Ewing sarcoma (EWSR1-ERG) and acute myeloid leukemia (FUS-ERG, ELF4-ERG). Loss-of-function variants in ERG cause primary lymphedema (LMPHM14).
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0000981 DNA-binding transcription factor activity, RNA polymerase II-specific | IBA GO_REF:0000033 | ACCEPT | Summary: ERG is a well-established ETS-family transcription factor with RNA polymerase II-specific activity. IBA annotation is well supported by phylogenetic conservation across the ETS family and extensive experimental evidence from multiple publications demonstrating ERG binds specific DNA sequences and activates/represses Pol II-dependent transcription [PMID:18195090, PMID:23093599]. Reason: Core function of ERG as an ETS-family transcription factor is unambiguously established. The IBA annotation is at the appropriate level of specificity. Supporting Evidence: PMID:18195090 Erg binds to the VE-cadherin promoter. Furthermore, Erg was found to enhance VE-cadherin promoter activity in a transactivation assay. PMID:23093599 The minimal EBS core is the consensus 5β²-GGA(A/T)-3β² known to be recognized by the ETS family of transcription factors through their highly conserved winged helix-turn-helix DNA-binding domain (ETS-domain) (13,14) |
| GO:0005634 nucleus | IBA GO_REF:0000033 | ACCEPT | Summary: ERG is a nuclear transcription factor. Nuclear localization is essential for its DNA-binding and transcriptional activity. PMID:8502479 describes ERG-2 as a "nuclear phosphoprotein." UniProt also confirms nuclear localization. Reason: Nuclear localization is a core aspect of ERG function as a transcription factor, well supported by IBA and multiple experimental studies. Supporting Evidence: PMID:8502479 ERG-2 is a nuclear phosphoprotein and binds to purine-rich sequences |
| GO:0006357 regulation of transcription by RNA polymerase II | IBA GO_REF:0000033 | ACCEPT | Summary: ERG regulates Pol II-dependent transcription of multiple target genes including VE-cadherin (CDH5), CLDN5, E-selectin, Wnt genes, and YAP1 [PMID:18195090, PMID:22235125, PMID:23913826, PMID:27109047]. This is a core biological process for ERG. Reason: ERG's role in regulation of Pol II transcription is its primary biological process. The IBA annotation is at an appropriate level. Supporting Evidence: PMID:18195090 Erg binds to the VE-cadherin promoter. Furthermore, Erg was found to enhance VE-cadherin promoter activity in a transactivation assay. |
| GO:0030154 cell differentiation | IBA GO_REF:0000033 | ACCEPT | Summary: ERG plays roles in cell differentiation, particularly endothelial and hematopoietic differentiation. ERG expression is enriched in endothelial cells and restricted to specific cell types including early myeloid cells [PMID:8502479]. ERG regulates endothelial differentiation programs and lymphatic endothelial identity. This is a broad term but appropriate for IBA. Reason: Cell differentiation is a well-supported biological role for ERG across endothelial and hematopoietic lineages. The IBA annotation is appropriate as a general term capturing the ETS family's conserved role in differentiation. Supporting Evidence: PMID:8502479 The expression of ERG-2 protein is restricted to few cell types and is high in early myeloid cells, indicating that it may function at an early stage of hematopoietic lineage determination. |
| GO:0003677 DNA binding | IEA GO_REF:0000043 | ACCEPT | Summary: ERG binds DNA through its ETS domain. This IEA annotation from UniProt keyword mapping is correct but less specific than the IBA and IDA annotations for sequence-specific DNA binding. Reason: Correct but general. More specific terms (GO:0000978, GO:1990837) are also present. As an IEA annotation it is acceptable to keep alongside more specific experimental annotations. |
| GO:0003700 DNA-binding transcription factor activity | IEA GO_REF:0000002 | ACCEPT | Summary: ERG is a DNA-binding transcription factor. This IEA annotation from InterPro domain mapping is correct. It is slightly less specific than the Pol II-specific annotation (GO:0000981) but acceptable as an IEA. Reason: Correct annotation from InterPro mapping of the ETS domain. The more specific Pol II-specific term is also present via IBA and ISA evidence. |
| GO:0005634 nucleus | IEA GO_REF:0000120 | ACCEPT | Summary: Duplicate of the IBA nucleus annotation. IEA annotation is consistent with known nuclear localization of ERG. Reason: Correct. Duplicates with different evidence codes are acceptable. ERG is clearly a nuclear protein. |
| GO:0005737 cytoplasm | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt records that ERG localizes to cytoplasm, specifically in cytoplasmic mRNP granules containing untranslated mRNAs [PMID:17289661]. This is not the primary localization but is supported by the IMP1 RNP granule study. Reason: While ERG is primarily nuclear, its presence in cytoplasmic mRNP granules has been documented by mass spectrometry [PMID:17289661]. The IEA annotation from UniProt subcellular location mapping is therefore correct. Supporting Evidence: PMID:17289661 We isolated the IMP1-containing RNP granules and found that they represent a unique RNP entity |
| GO:0006355 regulation of DNA-templated transcription | IEA GO_REF:0000002 | ACCEPT | Summary: ERG regulates DNA-templated transcription as an ETS-family transcription factor. This IEA from InterPro is correct and is a parent term of the more specific Pol II regulation annotation. Reason: Correct but more general than the Pol II-specific annotation. Acceptable as an IEA annotation. |
| GO:0006357 regulation of transcription by RNA polymerase II | IEA GO_REF:0000002 | ACCEPT | Summary: Duplicate of the IBA annotation for this term. IEA from InterPro domain mapping is consistent with known function. Reason: Correct. Duplicates with different evidence codes are acceptable. |
| GO:0043565 sequence-specific DNA binding | IEA GO_REF:0000002 | ACCEPT | Summary: ERG binds DNA in a sequence-specific manner via its ETS domain, recognizing the core 5'-GGA(A/T)-3' motif [PMID:8502479, PMID:23093599]. IEA from InterPro mapping is correct. Reason: Well supported by experimental evidence. ERG binds purine-rich sequences with the consensus (C/G)(C/a)GG-AA(G/a)T [PMID:8502479]. Supporting Evidence: PMID:8502479 ERG-2 is a nuclear phosphoprotein and binds to purine-rich sequences (C/G)(C/a)GG-AA(G/a)T |
| GO:1990837 sequence-specific double-stranded DNA binding | IEA GO_REF:0000117 | ACCEPT | Summary: ERG binds double-stranded DNA in a sequence-specific manner. This IEA from ARBA machine learning is consistent with the IDA annotation for the same term from PMID:28473536. Reason: Correct annotation. The same term is also supported by direct experimental evidence (IDA from PMID:28473536). |
| GO:0005515 protein binding | IPI PMID:20478527 An integrated network of androgen receptor, polycomb, and TM... | MODIFY | Summary: PMID:20478527 demonstrates ERG interaction with androgen receptor (AR, P10275) in prostate cancer. ChIP-seq showed ERG binds to and inhibits AR activity at gene-specific loci. The term "protein binding" is uninformative; a more specific term describing the functional interaction would be preferable. Reason: The interaction between ERG and AR is functionally relevant in prostate cancer biology. However, "protein binding" is too vague. ERG acts as a transcriptional co-regulator with AR. A more informative term would be appropriate. Proposed replacements: transcription corepressor activity Supporting Evidence: PMID:20478527 ERG disrupts androgen receptor (AR) signaling by inhibiting AR expression, binding to and inhibiting AR activity at gene-specific loci |
| GO:0005515 protein binding | IPI PMID:21575865 Mechanistic rationale for inhibition of poly(ADP-ribose) pol... | MODIFY | Summary: PMID:21575865 reports ERG interaction with PARP1 (P09874). ETS gene fusions including ERG interact with PARP1 and induce DNA damage. The interaction with PARP1 is relevant to the prostate cancer oncogenic context of TMPRSS2-ERG fusion. "Protein binding" is uninformative. Reason: The ERG-PARP1 interaction has functional significance. However, "protein binding" is too vague. A more specific term would better capture the biology. Proposed replacements: molecular condensate scaffold activity Supporting Evidence: PMID:21575865 Recurrent fusions of ETS genes are considered driving mutations in a diverse array of cancers |
| GO:0005515 protein binding | IPI PMID:22531786 A transcriptional repressor co-regulatory network governing ... | MODIFY | Summary: PMID:22531786 demonstrates ERG interaction with AR (P10275) via ChIP-seq in prostate cancer cells. ERG, HDACs, and EZH2 form a transcriptional repressor network modulating AR-regulated transcription. "Protein binding" is uninformative. Reason: ERG functions as a transcriptional co-regulator with AR. "Protein binding" is not informative enough. Proposed replacements: transcription corepressor activity Supporting Evidence: PMID:22531786 ERG, HDACs, and EZH2 are directly involved in androgen-regulated transcription and wired into an AR centric transcriptional network |
| GO:0005515 protein binding | IPI PMID:22722839 The mutational landscape of lethal castration-resistant pros... | MODIFY | Summary: PMID:22722839 is a genomic landscape study of castration-resistant prostate cancer. The IPI annotation for ERG-AR interaction is from a large-scale genomics study context. "Protein binding" is uninformative. Reason: While the ERG-AR interaction is real, "protein binding" does not capture the functional significance. Better annotated with a more specific term. Proposed replacements: transcription corepressor activity Supporting Evidence: PMID:22722839 ETS gene family fusions, PTEN loss and androgen receptor (AR) amplification, which drive |
| GO:0005515 protein binding | IPI PMID:27109047 ETS transcription factor ERG cooperates with histone demethy... | MODIFY | Summary: PMID:27109047 demonstrates that ERG directly binds KDM4A (O75164), a histone H3K9 demethylase. ERG and KDM4A cooperate to upregulate YAP1 transcription. GST pulldown showed direct binding. "Protein binding" is uninformative. Reason: The ERG-KDM4A interaction is functionally characterized - ERG recruits KDM4A to modify chromatin at target promoters. "Protein binding" should be replaced with a more informative term. Proposed replacements: transcription coactivator activity Supporting Evidence: PMID:27109047 ERG can directly bind to KDM4A (also known as JMJD2A), a histone demethylase that particularly demethylates lysine 9 on histone H3. ERG and KDM4A cooperated in upregulating the promoter of Yes-associated protein 1 (YAP1) |
| GO:0003682 chromatin binding | IEA GO_REF:0000107 | ACCEPT | Summary: ERG binds chromatin at regulatory elements across the genome. ChIP-seq studies show ERG occupancy at approximately 50% of promoters and 44% of active enhancers in endothelial cells. ERG also recruits chromatin-modifying enzymes like SETDB1 and KDM4A. This IEA from Ensembl ortholog transfer is well supported. Reason: Chromatin binding is well supported by ChIP-seq data from multiple studies including PMID:20478527 and deep research evidence showing extensive ERG chromatin occupancy. Supporting Evidence: PMID:27109047 ERG expression reduced histone H3 lysine 9 trimethylation at the YAP1 gene promoter, consistent with its epigenetic regulation through the ERG interaction partner, KDM4A |
| GO:0005654 nucleoplasm | IDA GO_REF:0000052 | ACCEPT | Summary: ERG localizes to the nucleoplasm as demonstrated by immunofluorescence data curated by HPA. As a nuclear transcription factor, nucleoplasm localization is expected and consistent with its function. Reason: Nucleoplasm localization is consistent with ERG's role as a nuclear transcription factor. IDA from HPA immunofluorescence data. |
| GO:0005829 cytosol | IDA GO_REF:0000052 | ACCEPT | Summary: HPA immunofluorescence data indicates ERG presence in the cytosol. While ERG is primarily nuclear, some cytoplasmic presence has been noted, and ERG has been identified in cytoplasmic mRNP granules [PMID:17289661]. However, cytosol localization may represent incomplete nuclear import or antibody cross-reactivity. Reason: Supported by HPA immunofluorescence data and consistent with the cytoplasmic mRNP granule finding from PMID:17289661. While not the primary localization, the evidence is present. |
| GO:1990837 sequence-specific double-stranded DNA binding | IDA PMID:28473536 Impact of cytosine methylation on DNA binding specificities ... | ACCEPT | Summary: PMID:28473536 is a systematic analysis of DNA binding specificities of 542 human TFs using methylation-sensitive SELEX. ERG was included in this study and its sequence-specific double-stranded DNA binding was confirmed. This is consistent with the well-established ETS domain-mediated DNA binding. Reason: IDA evidence from a high-quality systematic study of TF binding specificities. Confirms ERG's sequence-specific dsDNA binding activity. Supporting Evidence: PMID:28473536 By analysis of 542 human TFs with methylation-sensitive SELEX (systematic evolution of ligands by exponential enrichment) |
| GO:0001228 DNA-binding transcription activator activity, RNA polymerase II-specific | IMP PMID:18195090 Transcription factor Erg regulates angiogenesis and endothel... | ACCEPT | Summary: PMID:18195090 demonstrates that ERG directly activates the VE-cadherin (CDH5) promoter. ChIP showed ERG binding to the VE-cadherin promoter, and transactivation assays confirmed ERG enhances promoter activity. Inhibition of ERG decreased VE-cadherin expression. Reason: ERG acts as a transcriptional activator of VE-cadherin and other endothelial genes. This is a core molecular function of ERG, directly demonstrated by ChIP and reporter assays. Supporting Evidence: PMID:18195090 Using chromatin immunoprecipitation, we showed that Erg binds to the VE-cadherin promoter. Furthermore, Erg was found to enhance VE-cadherin promoter activity in a transactivation assay. |
| GO:0000785 chromatin | ISA GO_REF:0000113 | ACCEPT | Summary: ERG localizes to chromatin as expected for a DNA-binding transcription factor. ISA from TFClass database annotation is consistent with ChIP-seq data showing extensive ERG chromatin occupancy. Reason: Consistent with ERG's role as a chromatin-bound transcription factor. Well supported by multiple ChIP studies. |
| GO:0000981 DNA-binding transcription factor activity, RNA polymerase II-specific | ISA GO_REF:0000113 | ACCEPT | Summary: ISA annotation from TFClass database. Duplicates the IBA annotation for the same term. Consistent with ERG's established function. Reason: Correct annotation. Duplicates with different evidence codes are acceptable. |
| GO:0000978 RNA polymerase II cis-regulatory region sequence-specific DNA binding | IDA PMID:23093599 Targeting the DNA-binding activity of the human ERG transcri... | ACCEPT | Summary: PMID:23093599 demonstrates ERG binding to specific DNA sequences using EMSA, DNase I footprinting, and cellular luciferase assays. ERG binds the ETS binding site (5'-GGA(A/T)-3') in cis-regulatory regions including the osteopontin promoter and synthetic EBS constructs linked to the SV40 minimal promoter. This confirms ERG binds Pol II cis-regulatory regions in a sequence-specific manner. Reason: Strong direct experimental evidence (IDA) demonstrating ERG binds cis-regulatory regions in a sequence-specific manner. Core molecular function of ERG. Supporting Evidence: PMID:23093599 ERG, an ETS-family transcription factor, is commonly over-expressed or translocated in leukaemia and prostate carcinoma. In this work, we selected the di-(thiophene-phenyl-amidine) compound DB1255 as an ERG/DNA binding inhibitor |
| GO:0045944 positive regulation of transcription by RNA polymerase II | IDA PMID:18195090 Transcription factor Erg regulates angiogenesis and endothel... | ACCEPT | Summary: PMID:18195090 shows ERG directly activates VE-cadherin transcription. ChIP demonstrated ERG binding to the VE-cadherin promoter, and transactivation assays confirmed positive regulation. Reason: Direct experimental evidence of ERG positively regulating Pol II transcription at the VE-cadherin locus. Core function. Supporting Evidence: PMID:18195090 Using chromatin immunoprecipitation, we showed that Erg binds to the VE-cadherin promoter. Furthermore, Erg was found to enhance VE-cadherin promoter activity in a transactivation assay. |
| GO:0045944 positive regulation of transcription by RNA polymerase II | IMP PMID:18195090 Transcription factor Erg regulates angiogenesis and endothel... | ACCEPT | Summary: Same study as above (PMID:18195090). IMP evidence from mutant phenotype: inhibition of ERG expression resulted in decreased VE-cadherin expression, supporting its role as a positive regulator. Reason: IMP evidence complements the IDA evidence from the same study. ERG knockdown reduces VE-cadherin expression, confirming positive transcriptional regulation. Supporting Evidence: PMID:18195090 Inhibition of Erg expression in human umbilical vein endothelial cells (HUVECs), using antisense oligonucleotides, resulted in detachment of cell-cell contacts and increased cell death. Inhibition of Erg expression by antisense in HUVECs also lowered expression of the adhesion molecule vascular endothelial (VE)-cadherin |
| GO:0045944 positive regulation of transcription by RNA polymerase II | IMP PMID:22235125 ETS-related gene (ERG) controls endothelial cell permeabilit... | ACCEPT | Summary: PMID:22235125 identifies CLDN5 (claudin 5) as a downstream target of ERG in endothelial cells. ERG knockdown results in reduced CLDN5 expression and increased EC permeability, supporting ERG as a positive transcriptional regulator. Reason: Additional IMP evidence for ERG's role as a positive transcriptional regulator, here at the CLDN5 locus. Consistent with ERG's endothelial regulatory functions. Supporting Evidence: PMID:22235125 ERG knockdown results in marked increases in EC permeability. This is associated with a significant increase of stress fiber and gap formation in EC. Furthermore, we identify CLDN5 as a downstream target of ERG in EC. |
| GO:0045944 positive regulation of transcription by RNA polymerase II | IDA PMID:23913826 ERG is a critical regulator of Wnt/LEF1 signaling in prostat... | KEEP AS NON CORE | Summary: PMID:23913826 demonstrates ERG directly activates Wnt/LEF1 signaling in prostate cancer. ERG bound to promoters of Wnt genes and LEF1, directly increasing their expression. Reason: While ERG clearly activates transcription of Wnt pathway genes, this occurs in the context of aberrant TMPRSS2-ERG fusion expression in prostate cancer rather than normal ERG physiology. The positive transcriptional regulation by ERG is core, but the specific Wnt/LEF1 activation in prostate cancer is a disease context. Supporting Evidence: PMID:23913826 ERG activates Wnt/LEF1 signaling cascade through multiple mechanisms. ERG bound to the promoters of various Wnt genes to directly increase ligand expression. |
| GO:1990904 ribonucleoprotein complex | IDA PMID:17289661 Molecular composition of IMP1 ribonucleoprotein granules. | KEEP AS NON CORE | Summary: PMID:17289661 identified ERG as a component of IMP1-containing mRNP granules by mass spectrometry. ERG was found in cytoplasmic ribonucleoprotein granules containing untranslated mRNAs. This is not a core function of ERG as a transcription factor; ERG mRNA (or protein) may be present in these granules for regulatory purposes. Reason: ERG was identified in RNP granules by mass spectrometry in a large-scale proteomics study. While the finding is real, it likely reflects ERG mRNA being transported in mRNP granules or incidental co-purification, rather than a core functional role of ERG in ribonucleoprotein complexes. This is peripheral to ERG's primary function as a transcription factor. Supporting Evidence: PMID:17289661 We isolated the IMP1-containing RNP granules and found that they represent a unique RNP entity distinct from neuronal hStaufen and/or fragile X mental retardation protein granules, processing bodies, and stress granules. |
| GO:0003677 DNA binding | TAS PMID:8502479 Human ERG-2 protein is a phosphorylated DNA-binding protein-... | ACCEPT | Summary: PMID:8502479 directly demonstrates that ERG-2 is a DNA-binding protein using random oligonucleotide selection and EMSA. ERG-2 binds purine-rich sequences with consensus (C/G)(C/a)GG-AA(G/a)T. This is a core molecular function. Reason: Well-supported TAS annotation from a primary study demonstrating ERG's DNA-binding activity. While more specific terms exist, this is acceptable as legacy evidence. Supporting Evidence: PMID:8502479 ERG-2 is a nuclear phosphoprotein and binds to purine-rich sequences (C/G)(C/a)GG-AA(G/a)T |
| GO:0005634 nucleus | TAS PMID:8502479 Human ERG-2 protein is a phosphorylated DNA-binding protein-... | ACCEPT | Summary: PMID:8502479 identifies ERG-2 as a nuclear phosphoprotein. TAS evidence for nuclear localization. Reason: Consistent with all other evidence for ERG nuclear localization. Additional evidence code for the same correct annotation. Supporting Evidence: PMID:8502479 ERG-2 is a nuclear phosphoprotein |
| GO:0006468 protein phosphorylation | TAS PMID:8502479 Human ERG-2 protein is a phosphorylated DNA-binding protein-... | REMOVE | Summary: MISANNOTATION: ERG is a SUBSTRATE of protein phosphorylation, not an enzyme that catalyzes phosphorylation. PMID:8502479 explicitly states "ERG-2 is a nuclear phosphoprotein" and that "ERG-2 protein is phosphorylated by activation of protein kinase C." ERG has no kinase activity; it is a transcription factor whose activity is regulated by phosphorylation. Reason: ERG does not catalyze protein phosphorylation. The annotation confuses ERG being a phosphorylation substrate with ERG being involved in the process of phosphorylation. The GO term "protein phosphorylation" (GO:0006468) means the process of phosphorylating a protein, which requires kinase activity that ERG does not possess. ERG is phosphorylated BY protein kinase C [PMID:8502479]. Supporting Evidence: PMID:8502479 ERG-2 is a nuclear phosphoprotein and binds to purine-rich sequences (C/G)(C/a)GG-AA(G/a)T. ERG-2 protein, with a half-life of 21 h, is considerably more stable than the short-lived ETS-1 or ETS-2 proteins. Its phosphorylation is stimulated by phorbol myristate acetate (PMA) |
| GO:0007165 signal transduction | TAS PMID:8502479 Human ERG-2 protein is a phosphorylated DNA-binding protein-... | MARK AS OVER ANNOTATED | Summary: PMID:8502479 suggests ERG-2 is involved in signal transduction based on its phosphorylation by PKC and its role in downstream gene regulation. However, ERG is a downstream effector (transcription factor) rather than a signal transduction molecule per se. The annotation is overly broad. Reason: While ERG is regulated by signal transduction (phosphorylated by PKC) and its transcriptional targets include signaling pathway components, ERG itself is not a signal transduction molecule. It is a transcription factor that responds to signaling inputs. The GO term "signal transduction" is too broad and over-represents ERG's actual role, which is transcriptional regulation downstream of signaling pathways. Supporting Evidence: PMID:8502479 ERG-2 protein is phosphorylated by activation of protein kinase C, suggesting their involvement in distinct signal transduction mechanisms. |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: What are the relative contributions of ERG's transcriptional activation versus repression activities in endothelial homeostasis?
Q: Does ERG have distinct target gene programs in blood vascular versus lymphatic endothelial cells?
Q: Are there isoform-specific functional differences among the six known ERG splice variants?
Experiment: ChIP-seq comparison of ERG binding sites in blood endothelial cells versus lymphatic endothelial cells to define tissue-specific regulatory programs.
Experiment: Isoform-specific expression and functional analysis of ERG splice variants to determine if different isoforms have distinct transcriptional activities or target gene preferences.
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)