ZFX / ZFY
ZFX and its Y-linked paralogue ZFY are thirteen-finger C2H2 zinc-finger activators that bind a short GC-rich site containing a GGCCT core. Their placement is the interesting part: they bind inside CpG-island promoters a couple of hundred base pairs downstream of the transcription start site, between the open region at the TSS and the first downstream nucleosome, and they activate the promoters they occupy. Together with the related ZNF711, they occupy the majority of active CpG-island promoters in a cell. ZFX loss slows proliferation, which is why the gene recurs in growth-based CRISPR screens, and ZFX supports embryonic and hematopoietic stem-cell self-renewal. Germline ZFX variants cause an X-linked neurodevelopmental disorder.
The factors
ZFY was cloned in 1987 from the sex-determining region of the human Y chromosome and proposed as the testis-determining factor 3690661, a role later taken by SRY 1695712. Its X-linked paralogue ZFX has a similar gene structure, escapes X inactivation 2500252, and was recognized early as a probable transcriptional activator with alternative isoforms 2512506. The third member, ZNF711, is also X-linked and has an overlapping binding pattern 32406922. All three are ubiquitously expressed, which is unusual among C2H2 factors and consistent with a housekeeping role at promoters.
The proteins have an acidic N-terminal region and thirteen C2H2 zinc fingers. Mutational dissection shows that fingers 11 to 13 are necessary and sufficient for DNA binding, and that these fingers together with the N-terminal region constitute a functional transactivator 32406922; the remaining ten fingers are not required for site recognition. In the databases behind this family, only ZFX has matrices, and the four of them agree closely (cohesion 0.91), so this is one of the more internally consistent zinc-finger families despite the short motif.
The motif
The consensus is GC-rich and short. The core that all models share is GGCCT, and the fuller form used by JASPAR and CIS-BP is GCCBMGGCCT, which is also the data consensus for this family; the HOCOMOCO ChIP-derived version, CCAGGCCBCG, is the same core with different flanks. There is no crystal or NMR structure of the ZFX fingers on DNA, so which bases the three DNA-binding fingers contact is inferred from the standard C2H2 recognition rules rather than observed. With roughly five to eight informative positions, the motif is one of the least specific in routine use, and it is intrinsically GC-rich, so matches are dense wherever GC content is high. Reported analyses note that the GGCCT motif occurs throughout the genome but at higher density inside CpG islands, which is where occupancy is found 29429977.
ZFX, ZFY and ZNF711 are not separable by sequence in practice, and their genomic binding patterns at CpG-island promoters are essentially the same. ZNF711 also has its own reported motif, and where the two are distinguished, ZFX appears to be the dominant of the pair, with ZNF711 playing a subordinate role at shared promoters 32406922.
Where it acts
This family is a promoter family, and the position of binding is its defining feature. ChIP in four cancer cell lines recovered roughly 9,000 ZFX sites, most of them in CpG-island promoters, and genes with ZFX-bound promoters were expressed at higher levels than unbound ones 29429977; knocking ZFX down down-regulated many of them, so ZFX is an activator. The binding position is about 240 base pairs downstream of the transcription start site of responsive promoters, and nucleosome-occupancy and methylome sequencing places it between the nucleosome-free region at the TSS and the first downstream nucleosome 29429977. Taking ZFX, ZFY and ZNF711 together, the union of their bound promoters covers on the order of 10,000 CpG-island promoters, a large majority of the active CpG-island promoters in a cell 32406922. Mechanistically, reducing ZFX lowers histone H4 acetylation and increases RNA polymerase II pausing at target promoters 38726870, which is consistent with a role in promoter architecture and early elongation rather than in enhancer-driven, cell-type-specific regulation.
Biology and disease
Zfx-null mice are viable but smaller, with reduced self-renewal of embryonic and hematopoietic stem cells 17448993, so the gene supports proliferative capacity rather than specifying a lineage. That fits the cancer literature, in which ZFX expression correlates with proliferation and outcome across several tumor types, and in which Zfx facilitates tumorigenesis driven by Hedgehog-pathway activation 25164012. Because ZFX contributes to growth in many cell types, bi-allelic ZFX knockouts reduce cell growth and combined ZFX/ZNF711 knockouts cause major proliferation defects 32406922, which is why ZFX turns up repeatedly as a hit in growth-based genome-wide CRISPR knockout screens; readers should be aware that a ZFX hit in such a screen usually reflects this general proliferative requirement rather than a pathway-specific dependency. In human genetics, germline ZFX variants cause an X-linked neurodevelopmental disorder with a recurrent facial gestalt 38325380. ZFY, on the Y chromosome, has a distinct role in male germ-cell development 37354519.
Reading a hit
A GGCCT-containing match is weak evidence by itself: the motif is short, GC-rich, and denser in CpG islands, so hits accumulate in exactly the sequence people usually submit. Position is the most useful filter. A match a couple of hundred base pairs downstream of a CpG-island TSS is a plausible ZFX site; the same match in an intergenic or AT-poor enhancer is probably not. The hit identifies the family rather than the protein: ZFX, ZFY and ZNF711 bind the same sequence, and which one matters depends on the sex and expression of the cells. Note also that the CIS-BP human ZFX motif in this collection is transferred from mouse Zfx rather than measured on the human protein, though at 99% DNA-binding-domain identity that transfer is harmless. Finally, because ZFX sits at a large fraction of all active promoters, a ZFX-like motif recovered from another factor's ChIP data is more likely to reflect promoter co-occupancy than that factor's own specificity.
How this was worked out
A zinc-finger gene, ZFY, is cloned from the sex-determining region of the human Y chromosome and proposed as the testis-determining factor.
Still open then: No DNA-binding site was known, and the assignment rested on map position.
The X-linked paralogue ZFX is shown to have the same gene structure, to escape X inactivation, and to encode alternative isoforms with the features of a transcriptional activator, making the pair a dosage-related gene family rather than a male-specific switch.
SRY is identified as the testis-determining factor, removing the role that had motivated interest in ZFY and leaving the family without an assigned function for a decade.
Zfx-null mice are viable but small, with reduced self-renewal of embryonic and hematopoietic stem cells, giving the family a proliferative rather than a lineage-specifying role.
- 2018mediumChIP-seq in four cell lines, knockdown with RNA-seq, nucleosome-occupancy and methylome sequencing29429977
ChIP in human tumor cell lines places roughly 9,000 ZFX sites overwhelmingly in CpG-island promoters, about 240 base pairs downstream of the start site, between the nucleosome-free region and the first downstream nucleosome, and knockdown lowers expression of bound genes, so ZFX is an activator positioned by promoter architecture rather than by enhancer context.
Still open then: Whether ZFY and ZNF711 occupy the same promoters was untested.
- 2020mediumCRISPR bi-allelic knockouts in HEK293T, RNA-seq, structure-function analysis of ZFX mutants32406922
Treating ZFX, ZFY and ZNF711 as one family shows their bound promoters cover on the order of 10,000 CpG islands, that zinc fingers 11-13 are necessary and sufficient for DNA binding, and that losing ZFX and ZNF711 together causes major proliferation defects.
- 2024mediumDegron and knockdown of ZFX with chromatin profiling and nascent transcription assays38726870
Reducing ZFX lowers histone H4 acetylation and increases RNA polymerase II pausing at target promoters, which recasts the family as acting on promoter chromatin and early elongation rather than on recruitment of the polymerase.
Germline ZFX variants are shown to cause an X-linked neurodevelopmental disorder with a recurrent facial gestalt, establishing that a factor sitting at most active CpG-island promoters is dosage-sensitive in human development.
Still unresolved
- How ZFX is positioned a couple of hundred base pairs downstream of the start site is unexplained: the GGCCT core is short and common in CpG islands, so either an unidentified feature of promoter architecture selects the bound copy, or the position reflects nucleosome placement rather than sequence.
- Whether ZFX, ZFY and ZNF711 differ at all in intrinsic specificity is untested; their measured motifs are the same and their promoter occupancy overlaps, so the apparent dominance of ZFX over ZNF711 at shared promoters has no sequence explanation.
- No structure of the ZFX fingers on DNA has been determined, so which bases fingers 11-13 contact is inferred from the standard C2H2 recognition rules rather than observed, and the GCCBMGGCCT and CCAGGCCBCG forms in the databases have not been reconciled structurally.
- Whether the transcriptional effects of ZFX loss are direct or follow from a general proliferation defect is hard to separate, which is also why ZFX recurs as a hit in growth-based CRISPR screens.
The motif is short (roughly five to eight informative positions) and GC-rich, so hits are dense in CpG islands and GC-rich sequence generally, and a match alone means very little. Position carries more information than score: credible sites lie a couple of hundred base pairs downstream of a CpG-island transcription start site. ZFX, ZFY and ZNF711 read the same sequence and cannot be separated, and only ZFX has matrices here; the CIS-BP entry is transferred from mouse Zfx, benign at 99% DNA-binding-domain identity. No structure of these fingers on DNA exists, so the base contacts are inferred. Because ZFX occupies most active CpG-island promoters, a ZFX-like motif in another factor's ChIP data usually means co-occupancy.
References
- Page DC, Mosher R, Simpson EM, et al. The sex-determining region of the human Y chromosome encodes a finger protein. Cell 1987;51:1091-104. PMID 3690661
- Schneider-Gädicke A, Beer-Romero P, Brown LG, et al. ZFX has a gene structure similar to ZFY, the putative human sex determinant, and escapes X inactivation. Cell 1989;57:1247-58. PMID 2500252
- Schneider-Gädicke A, Beer-Romero P, Brown LG, et al. Putative transcription activator with alternative isoforms encoded by human ZFX gene. Nature 1989;342:708-11. PMID 2512506
- Rhie SK, Yao L, Luo Z, et al. ZFX acts as a transcriptional activator in multiple types of human tumors by binding downstream from transcription start sites at the majority of CpG island promoters. Genome Res 2018;28:310-320. PMID 29429977
- Ni W, Perez AA, Schreiner S, et al. Characterization of the ZFX family of transcription factors that bind downstream of the start site of CpG island promoters. Nucleic Acids Res 2020;48:5986-6000. PMID 32406922
- Hsu E, Hutchison K, Liu Y, et al. Reduction of ZFX levels decreases histone H4 acetylation and increases Pol2 pausing at target promoters. Nucleic Acids Res 2024;52:6850-6865. PMID 38726870
- Galan-Caridad JM, Harel S, Arenzana TL, et al. Zfx controls the self-renewal of embryonic and hematopoietic stem cells. Cell 2007;129:345-57. PMID 17448993
- Palmer CJ, Galan-Caridad JM, Weisberg SP, et al. Zfx facilitates tumorigenesis caused by activation of the Hedgehog pathway. Cancer Res 2014;74:5914-24. PMID 25164012
- Shepherdson JL, Hutchison K, Don DW, et al. Variants in ZFX are associated with an X-linked neurodevelopmental disorder with recurrent facial gestalt. Am J Hum Genet 2024;111:487-508. PMID 38325380
- Holmlund H, Yamauchi Y, Ruthig VA, et al. Return of the forgotten hero: the role of Y chromosome-encoded Zfy in male reproduction. Mol Hum Reprod 2023;29. PMID 37354519
- Sinclair AH, Berta P, Palmer MS, et al. A gene from the human sex-determining region encodes a protein with homology to a conserved DNA-binding motif. Nature 1990;346:240-4. PMID 1695712
4 motifs in this family — JASPAR 1 · HOCOMOCO 1 · CIS-BP 1 · Vierstra 1
| motif | database | consensus | bits |
|---|---|---|---|
| M08100_2.00Mus musculus 0.99 | CIS-BP | GCCbmGGCCT | 13.1 |
| Zfx (MA0146.3) | JASPAR | GCCbmGGCCT | 12.4 |
| ZFX.H14CORE.0.P.BPB | HOCOMOCO | ccAGGCCbcg | 11.7 |
| AC0329:ZFX/ZNF:C2H2_ZF | Vierstra | GGCCT | 9 |
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.