Estrogen-related receptors
The estrogen-related receptors ERRalpha (ESRRA), ERRbeta (ESRRB) and ERRgamma (ESRRG) are orphan nuclear receptors with no established natural ligand; their activity is set by coactivators such as PGC-1alpha rather than by a hormone. Unlike most nuclear receptors they bind DNA as monomers, on a single AGGTCA half-site carrying a three-base 5' extension: TNAAGGTCA, usually seen as TCAAGGTCA. The extension is read by a C-terminal extension of the DNA-binding domain in the minor groove, and it is what allows a monomer to bind with high affinity. The three paralogues share one motif and are not distinguishable from sequence. ERR sites sit at nuclear-encoded mitochondrial and oxidative-metabolism genes, at pluripotency enhancers, and in the placenta and inner ear.
The factors
Three paralogues make up subfamily NR3B: ESRRA (ERRalpha, NR3B1), ESRRB (ERRbeta, NR3B2) and ESRRG (ERRgamma, NR3B3). They were found by homology to the estrogen receptor but are not activated by estradiol; no endogenous ligand is established, the ligand-binding pocket is small and largely filled, and transcriptional output is controlled by coactivator availability, chiefly PGC-1alpha and PGC-1beta (Giguere 2008) 18664618. Architecturally they resemble other nuclear receptors - a C4 double-zinc-finger DNA-binding domain (DBD) and a ligand-binding domain - but they bind DNA as monomers 12654265, and can also homodimerize. ERRalpha is broadly expressed with highest levels in heart, kidney, skeletal muscle and brown adipose tissue 9271417; ERRgamma is enriched in heart, brain and kidney; ERRbeta is the most restricted, appearing in extraembryonic ectoderm and trophoblast, the inner ear and pluripotent stem cells 9285590,18179891,23040477. Ten motif models fall here (JASPAR 3, HOCOMOCO 3, CIS-BP 3, Vierstra 1), covering all three genes.
The motif
The ERR response element is an extended half-site, not a repeat. Binding-site selection with ERRalpha defined it as TNAAGGTCA: the canonical nuclear receptor hexamer AGGTCA preceded by three bases, of which the two immediately 5' are almost always AA (Sladek et al. 1997) 9271417. In practice the models here read TCAAGGTCA or CAAGGTCA, and the family's data consensus is CAAGGTCA.
The 5' extension is functional, not decorative. The solution structure of the ERRbeta DBD bound to its site showed that monomeric high-affinity binding requires base-specific minor-groove contacts from a modified guanosine-binding AT-hook in the C-terminal extension (CTE) that flanks the core DBD; the CTE is stabilized by packing against the DBD surface that other receptors use for dimerization, a pseudo-dimer interface (Gearhart et al. 2003) 12654265. That is the structural reason why ERRs need three extra 5' bases while a GR or ER dimer needs a second hexamer instead. Dimeric arrangements do occur: the CIS-BP ESRRG model M03412 is two extended half-sites in tandem (AAGGTCAttTcAAGGTCA), the highest-information model in the family at 22 bits.
Spacing and orientation are what separate this family from its neighbours, and they separate it imperfectly. An inverted AGGTCA repeat with a 3 bp spacer is an ERE; direct repeats with 1-5 bp spacers are RXR-heterodimer elements; a single 5'-extended half-site is an ERR element - but it is also the SF-1/LRH-1 response element (SFRE), and the Vierstra archetype in this family is labelled ESRRA/NR5A for exactly that reason. ERRalpha and ERRbeta also bind and activate through consensus EREs, while ERbeta does not reciprocate (Vanacker et al. 1999) 10428965. The family data marks one CIS-BP model labelled ESRRA (M07988) as a full ERE belonging with ESR. ESRRA, ESRRB and ESRRG themselves are indistinguishable: their models are 0.96-0.99 similar to one another.
Where it acts
ERR sites are found at both promoters and distal enhancers, with a strong bias towards nuclear-encoded mitochondrial genes: oxidative phosphorylation subunits, TCA cycle enzymes and fatty-acid beta-oxidation genes, where ERRalpha works with PGC-1alpha 15100410,18664618. The founding example was an ERRE in the promoter of the medium-chain acyl-CoA dehydrogenase gene (Sladek et al. 1997) 9271417. Genome-wide mapping in heart showed ERRalpha and ERRgamma occupying an overlapping program of cardiac metabolic and contractile genes (Dufour et al. 2007) 17488637. At such promoters ERR elements co-occur with NRF1 17488637, GABPA 15100410 and YY1 sites. In mouse embryonic stem cells Esrrb sits at pluripotency enhancers with OCT4, SOX2 and NANOG, and is itself a direct Nanog target that can substitute for Nanog (Festuccia et al. 2012) 23040477.
Biology and disease
The ERRs are best understood as regulators of oxidative capacity. Esrra-null mice are lean, resistant to diet-induced obesity and impaired in cold adaptation; ERRalpha and ERRgamma together are required for the perinatal switch of the heart to fatty-acid oxidation 18664618. Esrrb-null embryos die from placental failure with abnormal chorion development (Luo et al. 1997) 9285590. In humans, recessive ESRRB mutations cause nonsyndromic hearing loss DFNB35 (Collin et al. 2008) 18179891. ERRalpha expression is associated with poor outcome in breast cancer, where it supports the metabolic and biosynthetic demands of proliferation and interacts with ERBB2 signalling (Deblois and Giguere 2013) 23192231. ERRgamma has been studied in prostate cancer and hepatic gluconeogenesis, and inverse agonists of ERRalpha are in preclinical development.
Reading a hit
The practical warning is length. TCAAGGTCA is a 9-mer with roughly eight informative positions, so an exact match occurs on the order of 10^4 times in the haploid human genome by chance alone, and near-matches many times more. A hit therefore carries very little weight on its own, and far less than a 13-15 bp dimeric element would. Measure the 5' extension explicitly: without the AA at positions -2 and -3 you have a bare half-site that anything in the superfamily could touch, and with an inverted second hexamer three bases downstream you have an ERE instead. A match cannot distinguish ESRRA, ESRRB and ESRRG, nor reliably distinguish ERRs from NR5A1/NR5A2, whose element is the same 9-mer. All three HOCOMOCO models here are grade A with combined ChIP, HT-SELEX and methyl-SELEX evidence (PSM), so direct binding to this sequence is well supported; the uncertainty is about which family member and which site is used, not about the sequence.
How this was worked out
Screening with the estrogen receptor DNA-binding domain as a probe isolates two cDNAs (ERR1 and ERR2) encoding receptor-like proteins with no known hormone, founding the estrogen-related receptors as orphan receptors.
Still open then: Their DNA target and any activating ligand were unknown.
Binding-site selection with ERRalpha defines its element as a single extended half-site, TNAAGGTCA, and identifies one in the MCAD promoter, linking the receptor to fatty-acid oxidation.
Still open then: What activates ERRalpha, since it did not activate the MCAD promoter on its own.
Esrrb-null embryos die at about E10.5 with failed chorion development, too many trophoblast giant cells and too little diploid trophoblast, a defect rescued by wild-type tetraploid extraembryonic tissue.
ERRalpha and ERRbeta bind and activate through both the classical ERE and the extended SF-1 response element, sharing targets with ERalpha but not with ERbeta.
The solution structure of the ERR2 (ERRbeta) DNA-binding domain on its site shows that monomeric high-affinity binding needs minor-groove contacts from an AT-hook-like C-terminal extension that packs against the surface other receptors use to dimerize, explaining the required 5' extension.
- 2004mediumExpression profiling, computational motif analysis, cellular assays with a synthetic inhibitor15100410
Motif discovery in PGC-1alpha-induced genes identifies ERRalpha and GABPA sites as key regulators of oxidative phosphorylation genes, and a synthetic ERRalpha inhibitor blunts PGC-1alpha effects on gene expression and respiration: the two factors are DNA-binding partners through which PGC-1alpha drives this program.
Genome-wide location analysis in adult heart shows ERRalpha and ERRgamma occupying a common set of promoters for substrate uptake, ATP production and transport, calcium handling and contraction, with NRF-1, CREB and STAT3 sites enriched alongside.
- 2012mediumInducible Nanog alleles, overexpression and deletion in mouse ES cells, reprogramming assays23040477
Esrrb is a direct Nanog target that sustains cytokine-independent self-renewal of embryonic stem cells even without Nanog, placing the ERR element in the pluripotency network.
Still unresolved
- Whether ERRalpha has an endogenous ligand is disputed: the receptor is constitutively active and regulated mainly through PGC-1 coactivators, while cholesterol has been proposed as an endogenous agonist.
- The oligomeric state of ERRs on DNA is unsettled: the DNA-binding-domain structure explains recognition of the extended half-site by a monomer, but ERRalpha and ERRgamma have been described as non-obligatory heterodimers in heart, and the highest-information model in this family (CIS-BP M03412, labelled ESRRG) is a tandem pair of extended half-sites.
- Whether ERRs compete with ER at endogenous EREs is not established: ERRalpha and ERRbeta activate consensus EREs in reporter assays, yet every genome-derived ERR model here is an extended half-site rather than an inverted repeat.
The ERR element is a 9 bp extended half-site, which is short: expect on the order of 10^4 chance matches per haploid genome, so a hit alone is weak evidence. It does not distinguish ESRRA, ESRRB and ESRRG, whose models are 0.96-0.99 similar, and it does not distinguish ERRs from NR5A1/NR5A2, which use the same TCAAGGTCA element - the Vierstra archetype here is labelled ESRRA/NR5A. Cross-recognition with the estrogen receptor is real, not just a labelling artefact: ERRs activate through consensus EREs, and one CIS-BP model labelled ESRRA in this family is a full ERE flagged as belonging with ESR. Require the 5' AA extension before calling an ERR site.
References
- Sladek R, Bader JA, Giguère V. The orphan nuclear receptor estrogen-related receptor alpha is a transcriptional regulator of the human medium-chain acyl coenzyme A dehydrogenase gene. Mol Cell Biol 1997;17:5400-9. PMID 9271417
- Luo J, Sladek R, Bader JA, et al. Placental abnormalities in mouse embryos lacking the orphan nuclear receptor ERR-beta. Nature 1997;388:778-82. PMID 9285590
- Vanacker JM, Pettersson K, Gustafsson JA, et al. Transcriptional targets shared by estrogen receptor- related receptors (ERRs) and estrogen receptor (ER) alpha, but not by ERbeta. EMBO J 1999;18:4270-9. PMID 10428965
- Gearhart MD, Holmbeck SM, Evans RM, et al. Monomeric complex of human orphan estrogen related receptor-2 with DNA: a pseudo-dimer interface mediates extended half-site recognition. J Mol Biol 2003;327:819-32. PMID 12654265
- Dufour CR, Wilson BJ, Huss JM, et al. Genome-wide orchestration of cardiac functions by the orphan nuclear receptors ERRalpha and gamma. Cell Metab 2007;5:345-56. PMID 17488637
- Collin RW, Kalay E, Tariq M, et al. Mutations of ESRRB encoding estrogen-related receptor beta cause autosomal-recessive nonsyndromic hearing impairment DFNB35. Am J Hum Genet 2008;82:125-38. PMID 18179891
- Giguère V. Transcriptional control of energy homeostasis by the estrogen-related receptors. Endocr Rev 2008;29:677-96. PMID 18664618
- Festuccia N, Osorno R, Halbritter F, et al. Esrrb is a direct Nanog target gene that can substitute for Nanog function in pluripotent cells. Cell Stem Cell 2012;11:477-90. PMID 23040477
- Deblois G, Giguère V. Oestrogen-related receptors in breast cancer: control of cellular metabolism and beyond. Nat Rev Cancer 2013;13:27-36. PMID 23192231
- Giguère V, Yang N, Segui P, et al. Identification of a new class of steroid hormone receptors. Nature 1988;331:91-4. PMID 3267207
- Mootha VK, Handschin C, Arlow D, et al. Erralpha and Gabpa/b specify PGC-1alpha-dependent oxidative phosphorylation gene expression that is altered in diabetic muscle. Proc Natl Acad Sci U S A 2004;101:6570-5. PMID 15100410
10 motifs in this family — JASPAR 3 · HOCOMOCO 3 · CIS-BP 3 · Vierstra 1
| motif | database | consensus | bits |
|---|---|---|---|
| M03412_2.00 | CIS-BP | AAGGTCAttTcAAGGTCA | 22.3 |
| ESRRB (MA0141.4) | JASPAR | TCAAGGTCAt | 16.2 |
| ERR2.H14CORE.0.PSM.APSMA | HOCOMOCO | tCAAGGTCA | 13.7 |
| M07988_2.00looks like ESR | CIS-BP | aGGTCAccnTGACCt | 13.5 |
| Esrrg (MA0643.2) | JASPAR | CAAGGTCA | 13 |
| ERR3.H14CORE.0.PSM.APSMA | HOCOMOCO | tCAAGGTCA | 13 |
| ESRRA (MA0592.4) | JASPAR | tCAAGGTCA | 12.9 |
| M05595_2.00 | CIS-BP | CAAGGTCA | 12.8 |
| ERR1.H14CORE.0.PSM.APSMA | HOCOMOCO | CAAGGTCA | 12.3 |
| AC0121:ESRRA/NR5A:Nuclear_receptor | Vierstra | CAAGGTCA | 10.8 |
HOCOMOCO tags read as assay + quality grade: P ChIP-seq, S HT-SELEX, M methyl-HT-SELEX, G GHT-SELEX, I SMiLE-seq, B PBM; grade A means two or more assay types agreed. A P-only matrix is a record of where the protein was crosslinked, which need not be a sequence it binds itself.
The factors, elsewhere
These entries are written from the primary literature and every citation is checked against PubMed, but this one can still be out of date, too confident, or simply wrong — especially where a motif’s published site and its database matrices disagree. Corrections and additions are welcome, and they are the fastest way to make this page better for everyone.