YY1 / YY2
YY1 is a ubiquitously expressed zinc-finger protein that binds a short CCAT-centered site, usually written CGCCATNTT or, on the other strand, AANATGGCGGC. It sits at or just inside the transcription start site of a large fraction of CpG-island promoters, works as an activator at some and a repressor at others, and is found on both the enhancer and the promoter side of enhancer-promoter contacts. Two retrogenes derived from YY1, YY2 and ZFP42/REX1, share the same zinc-finger domain; YY2 binds the same consensus more weakly, and REX1 prefers divergent sites that still contain the CCAT core. Because the motif is short and concentrated in CpG islands, a YY1 match is one of the most common hits in promoter sequence.
The factors
YY1 (Yin Yang 1) has four C2H2 zinc fingers at its C terminus and belongs to the GLI-Kruppel class; its Drosophila counterpart is Pleiohomeotic (Pho). It was described independently in the same period as a repressor whose activity is relieved by adenovirus E1A 1655281 and as an initiator-binding protein that directs and activates transcription in vitro 1720509, and it also appears in the older literature as NF-E1, delta factor and UCRBP. It binds DNA as a monomer and is expressed in essentially all cell types.
The family in these databases has three human members. YY2 and ZFP42 (REX1) are both retrogenes generated from YY1 in placental mammals 17478514, and their zinc-finger domains remain highly similar to YY1's. Motif measurements show that YY2 binds the same consensus as YY1 but with much lower affinity, while REX1 binds sequences divergent from YY1's that nevertheless share the CCAT core 17478514. ZFP42/REX1 is best known as a pluripotency marker in embryonic stem cells. All three are assigned here to a single family because their matrices are not separable.
The motif
The literature core is CCAT. The short consensus most often quoted is CGCCATNTT, and the reverse-strand form AANATGGCGGC is what several databases show, including the form used in the ontology entry for this family. Selection experiments give a longer high-affinity version, GCCGCCATTTTG, and the endogenous sites at Peg3 (GGCGCCATNTT) and Xist (CCGCCATNTT) are longer than the classic motif 18950698. Substituting three residues in the first zinc finger, to mimic Drosophila Pho, collapses specificity to a bare GCCAT, showing that the four fingers contribute jointly rather than modularly and that the flanking information depends on finger one 18950698. The data consensus here, CCGCCATYTT, is the same site.
The cocrystal structure of the YY1 zinc fingers on the adeno-associated virus P5 initiator explains an unusual property. Upstream of the start sites the protein contacts both strands, whereas downstream the contacts are essentially restricted to the template strand, leaving the non-template strand free 8942976. That asymmetry is what allows YY1 to sit over an initiator, support TBP-independent initiation with a fixed direction, and bind a preformed transcription bubble or a DNA/RNA hybrid 8942976. Functionally the site is orientation-sensitive in context: in the c-fos promoter YY1 bends DNA and its effect depends on which way the site points 8276234. Sequence alone does not distinguish YY1 from YY2, and only partially separates ZFP42.
Where it acts
YY1 is a promoter factor first. Its motif is strongly enriched at CpG-island promoters, where it co-occurs with a characteristic set of partners at active promoters, including HCFC1, ZNF143, THAP11 and GABP 23539139, and it binds within a few tens of base pairs of the transcription start site far more often than at a distance. It is also present at active enhancers, and the observation that it occupies both sides of many enhancer-promoter pairs led to the proposal that YY1 acts as a general structural contributor to enhancer-promoter loops 29224777, in loose analogy to CTCF at structural loops. YY1 is additionally reported as a CTCF cofactor at the X-inactivation choice 17218270, and as the mammalian Pho counterpart that helps recruit Polycomb complexes, an activity that requires both DNA binding and interaction with YAF2 24285299. Because the informative site is short and CpG islands are the preferred context, matches are extremely dense in promoter sequence: a YY1 hit near a TSS is close to the default expectation rather than a discovery.
Biology and disease
YY1 is dosage-sensitive in humans: de novo heterozygous variants and deletions cause Gabriele-de Vries syndrome, an intellectual-disability syndrome with movement abnormalities in which patient cells show transcriptional and chromatin dysfunction 28575647, consistent with partial loss of a general promoter factor rather than a tissue-restricted one. Mouse YY1 mutants show homeotic transformations of the axial skeleton and interact genetically with the Polycomb gene Ring1 16624538, placing YY1 in the PcG-dependent regulation of Hox genes. At the molecular level YY1 has been assigned both activating and repressing roles at different promoters, often through the cofactor it recruits (HDACs and Polycomb components on one side, coactivators and the basal machinery on the other), which is why the same motif cannot be read as a sign. YY1 is frequently reported as overexpressed in tumors, though the causal claims there are much less settled than the developmental genetics.
Reading a hit
A YY1 match is common enough that on its own it is close to uninformative, particularly inside a CpG island. Treat it as a family-level call: YY1, YY2 and ZFP42 are not separable by sequence, and YY2 in particular binds the same consensus. A hit tells you nothing about direction of effect, because YY1 activates at some promoters and represses at others. Position matters more than score: a site within tens of base pairs of a TSS, in the orientation seen at initiators, is a more credible YY1 element than a stronger match in intergenic DNA. Finally, YY1 has been used as a positive control in so many ChIP experiments that the motif is over-represented in derived models generally; if a ChIP-derived matrix for some unrelated factor looks like AANATGGCGGC, indirect or co-occupancy-driven enrichment at active promoters is the first explanation to consider, not a new specificity.
How this was worked out
- 1991mediumPromoter deletion and reporter assays, protein purification, in vitro transcription1655281 1720509
Two groups describe the same protein from opposite directions: as a repressor of the adeno-associated virus P5 promoter whose repression is converted to activation by adenovirus E1A, and as an initiator-binding protein that sets the start site and activates transcription in vitro.
Still open then: Why one protein gave activation in one assay and repression in the other.
At the c-fos promoter YY1 bends the DNA, and whether it represses depends on the orientation of its site, showing the element is read in context rather than as a free-standing word.
The cocrystal structure of the four YY1 zinc fingers on the AAV P5 initiator shows contacts to both strands upstream of the start site but essentially template-strand-only contacts downstream, explaining how YY1 can sit over an initiator and leave the non-template strand accessible.
Still open then: The structure covered one site; the contribution of each finger to flanking specificity was not tested.
Phylogenetic analysis identifies YY2 and ZFP42/REX1 as retroposed copies of YY1 that arose in placental mammals; YY2 keeps the YY1 consensus at much lower affinity while REX1 has diverged, retaining only the CCAT core.
- 2009mediumGel-shift assays on endogenous sites, binding-site selection, zinc-finger mutagenesis18950698
The functional YY1 site is longer than the classic CGCCATNTT: the Peg3 and Xist elements and an independent selection give GCCGCCATTTTG with higher affinity, and swapping three residues of finger 1 for the Drosophila Pho version collapses specificity to GCCAT, so the four fingers act jointly rather than modularly.
Genome-wide occupancy places YY1 in a fixed company of factors at active CpG-island promoters, coinciding with HCFC1, ZNF143, THAP11 and GABP, which sets the baseline expectation that a YY1 site near a TSS is unremarkable.
- 2017mediumExome sequencing in a patient cohort, patient-cell transcriptomics and chromatin profiling28575647
De novo heterozygous YY1 variants and deletions are shown to cause Gabriele-de Vries syndrome, establishing that this broadly expressed promoter factor is dosage-sensitive in human development.
- 2017mediumChIP-seq, chromosome-conformation assays, site deletion and degron-based protein depletion in mouse embryonic stem cells29224777
YY1 is proposed to be a general structural regulator of enhancer-promoter loops, by analogy to CTCF at larger loops: it binds both enhancers and promoter-proximal elements, and deleting its sites or depleting the protein disrupts looping and expression.
Still open then: Whether loss of looping is a direct structural effect or a consequence of losing transcription at the same promoters was not separated.
Still unresolved
- Whether YY1 is a general structural component of enhancer-promoter loops is unsettled: the proposal rests on co-occupancy at loop anchors plus site deletion and protein depletion in a small number of cell lines, and an alternative reading is that YY1 marks active CpG-island promoters, so the contacts it is credited with follow from transcription rather than cause it.
- Nothing in the sequence predicts whether YY1 will activate or repress a given promoter; the recruited cofactor (Polycomb and HDACs versus coactivators and the basal machinery) is the usual explanation, but no context feature has been shown to determine the choice.
- Whether ZFP42/REX1 belongs with YY1 as a motif family member is debatable: selection experiments say its sites have diverged from YY1's apart from the CCAT core, yet its HOCOMOCO model (GGCAGCCATTTT) is close enough to the YY1 model to be pooled here.
- What YY2 does that YY1 does not is unresolved, since it binds the same consensus with much lower affinity and no distinct target class has been established for it.
The site is short, CpG-island-associated and present at a large share of active promoters, so a hit is expected rather than informative and should be weighted by position relative to the TSS. YY1, YY2 and ZFP42/REX1 all descend from the same zinc-finger domain and cannot be told apart from sequence; YY2 binds the identical consensus with lower affinity. A match says nothing about activation versus repression, since YY1 does both depending on the cofactor recruited. And because YY1 sits at essentially all active promoters, an AANATGGCGGC-like motif recovered from another factor's ChIP data is more likely promoter co-occupancy than that factor's own specificity.
References
- Shi Y, Seto E, Chang LS, et al. Transcriptional repression by YY1, a human GLI-Krüppel-related protein, and relief of repression by adenovirus E1A protein. Cell 1991;67:377-88. PMID 1655281
- Seto E, Shi Y, Shenk T. YY1 is an initiator sequence-binding protein that directs and activates transcription in vitro. Nature 1991;354:241-5. PMID 1720509
- Houbaviy HB, Usheva A, Shenk T, et al. Cocrystal structure of YY1 bound to the adeno-associated virus P5 initiator. Proc Natl Acad Sci U S A 1996;93:13577-82. PMID 8942976
- Kim Jd, Kim J. YY1's longer DNA-binding motifs. Genomics 2009;93:152-8. PMID 18950698
- Kim JD, Faulk C, Kim J. Retroposition and evolution of the DNA-binding motifs of YY1, YY2 and REX1. Nucleic Acids Res 2007;35:3442-52. PMID 17478514
- Natesan S, Gilman MZ. DNA bending and orientation-dependent function of YY1 in the c-fos promoter. Genes Dev 1993;7:2497-509. PMID 8276234
- Michaud J, Praz V, James Faresse N, et al. HCFC1 is a common component of active human CpG-island promoters and coincides with ZNF143, THAP11, YY1, and GABP transcription factor occupancy. Genome Res 2013;23:907-16. PMID 23539139
- Weintraub AS, Li CH, Zamudio AV, et al. YY1 Is a Structural Regulator of Enhancer-Promoter Loops. Cell 2017;171:1573-1588.e28. PMID 29224777
- Basu A, Wilkinson FH, Colavita K, et al. YY1 DNA binding and interaction with YAF2 is essential for Polycomb recruitment. Nucleic Acids Res 2014;42:2208-23. PMID 24285299
- Donohoe ME, Zhang LF, Xu N, et al. Identification of a Ctcf cofactor, Yy1, for the X chromosome binary switch. Mol Cell 2007;25:43-56. PMID 17218270
- Gabriele M, Vulto-van Silfhout AT, Germain PL, et al. YY1 Haploinsufficiency Causes an Intellectual Disability Syndrome Featuring Transcriptional and Chromatin Dysfunction. Am J Hum Genet 2017;100:907-925. PMID 28575647
- Lorente M, Pérez C, Sánchez C, et al. Homeotic transformations of the axial skeleton of YY1 mutant mice and genetic interaction with the Polycomb group gene Ring1/Ring1A. Mech Dev 2006;123:312-20. PMID 16624538
14 motifs in this family — JASPAR 3 · HOCOMOCO 5 · CIS-BP 3 · Vierstra 3
| motif | database | consensus | bits |
|---|---|---|---|
| M04647_2.00 | CIS-BP | CCGCCATTTTG | 19.8 |
| TYY2.H14CORE.1.SM.BSMB | HOCOMOCO | AAAATGGCGG | 16.6 |
| M07855_2.00 | CIS-BP | GCCGCCATcTTg | 15.9 |
| ZFP42 (MA1651.2) | JASPAR | AAaATGGCTGCC | 15.3 |
| TYY2.H14CORE.0.PS.APSA | HOCOMOCO | AAgATGGCGGC | 14.6 |
| M08348_2.00 | CIS-BP | GCaGCCATTTT | 14.6 |
| AC0510:YY/MZF:C2H2_ZF | Vierstra | CCATtTTtgrcnnwnhGTG | 14.6 |
| TYY1.H14CORE.0.PSM.APSMA | HOCOMOCO | AAgATGGCGGC | 14.5 |
| ZFP42.H14CORE.0.PM.APMA | HOCOMOCO | gGCaGCCATTTT | 14.4 |
| TYY1.H14CORE.1.S.BSB | HOCOMOCO | AATGGCG | 11.6 |
| Yy1 (MA0095.4) | JASPAR | CAAAATGG | 11.4 |
| AC0511:YY/ZFP:C2H2_ZF | Vierstra | CcGCCATyTT | 11.2 |
| YY2 (MA0748.3) | JASPAR | ATGGCGG | 10.4 |
| AC0512:YY:C2H2_ZF | Vierstra | CCATnTt | 7.5 |
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.