WT1
WT1 is the Wilms tumor suppressor, a four-fingered C2H2 zinc-finger protein whose last three fingers are closely related to those of EGR1 and which binds essentially the same GC-rich site. In vitro selection also found a slightly different high-affinity site, GCGTGGGAGT, called the WTE. Alternative splicing inserts three amino acids (lysine-threonine-serine) between fingers three and four, and the +KTS form binds DNA poorly and localizes to splicing-factor compartments, so WT1 is a DNA-binding protein only in part. Expression is restricted to developing kidney and gonad, podocytes, mesothelium and epicardium. The WT1 matrices in motif databases are ChIP-derived and G-rich rather than EGR-like, which matters when reading a hit.
The factors
WT1 was isolated as a zinc-finger gene at the Wilms tumor locus on human chromosome 11p13, spanning about 50 kb (Call et al. 1990) 2154335. The protein has four Cys2-His2 fingers; fingers 2-4 align with the three fingers of EGR1, and finger 1 adds further contacts 17716689. Two alternative splice events generate four main isoforms: insertion or omission of exon 5 (17 amino acids) in the N-terminal region, and insertion or omission of the tripeptide lysine-threonine-serine (KTS) between fingers 3 and 4. The KTS insertion is the consequential one. The -KTS protein localizes to nuclear compartments occupied by transcription factors, while the +KTS protein concentrates in splicing-factor domains (Larsson et al. 1995) 7736591, WT1, unlike EGR1, also binds RNA through the same zinc fingers (Caricasole et al. 1996) 8755514, in a region that overlaps the DNA-binding surface (Bardeesy and Pelletier 1998) 9512553. Which isoform does so is unsettled: RNA selection found +KTS isoforms unable to bind the selected ligands 9512553, while an NMR study found -KTS and +KTS binding a stem-loop RNA with similar affinity, the KTS insertion instead abolishing high-affinity DNA binding 32955251. So one gene produces a DNA-binding transcriptional regulator and an RNA-binding protein from the same finger array, and the ratio between them is under splice control. WT1 expression is restricted: nephron progenitors and podocytes in kidney, Sertoli and granulosa cells in gonad, mesothelium, epicardium and spleen, with additional roles in heart and nervous system (Hastie 2017) 28811308. WT1 has no close paralogue; its nearest relatives by DNA-binding domain are the EGR proteins.
The motif
WT1 binds the EGR-1 consensus, which was how its sequence specificity was first demonstrated: a bacterially expressed WT1 finger region bound the EGR-1 site (Rauscher et al. 1990) 2244209. A whole-genome PCR selection later identified a distinct preferred 10-mer, GCGTGGGAGT, termed the WTE, which is similar to but not the same as the EGR-1 consensus GCG(G/T)GGGCG (Nakagama et al. 1995) 7862142. The difference is the penultimate triplet: GGGAGT in the WTE against GGGCG in the EGR site. The four-finger domain has been solved on DNA by X-ray crystallography and NMR on both 14-bp and 17-bp duplexes, which resolved the position of finger 1 (Stoll et al. 2007) 17716689. Here the atlas data and the literature agree only in part. Both HOCOMOCO models (WT1.H14CORE.0.P.B, GGGGGAGGA; WT1.H14CORE.1.P.B, GCGGGGGAGG) and the JASPAR profile (MA1627.2, CCTCCCCCAC, which is GTGGGGGAGG on the other strand) are G-rich runs without the EGR signature CACGC, whose nearest neighbors outside the family are EGR, SP/KLF, ZNF281 and ZBTB17 models 37971293. The longer model is not a bare G-run: GCGGGGGAGG matches the biochemically defined WTE, GCGTGGGAGT, at 8 of 10 positions, including the adenine at position 8 that contributes most to WTE affinity, and differs only at position 4 and at the 3' thymine 7862142. Family cohesion is nonetheless low, 0.57. The remaining two matrices carry no usable signal: the Vierstra archetype AC0095:WT/FOXC is an AT-rich model (ATATATT) whose nearest neighbors are K50-class homeodomain motifs, and the CIS-BP entry M00243_2.00 has 0.1 bits of information. In the other direction, a HOCOMOCO matrix labelled ZBTB17/MIZ1 (ZBT17.H14CORE.0.P.B, GTGGGGGAGGGG) has its nearest neighbor across all four databases in this family - a sign of how much these ChIP-derived models share with the general G-rich promoter element.
Where it acts
WT1 sites sit in GC- and GA-rich promoter-proximal sequence, and the genes WT1 controls are concentrated in the programs it is expressed in: nephron and podocyte differentiation, gonadal development, epicardial and mesothelial identity (Hastie 2017) 28811308. Because WT1 and EGR1 recognize overlapping sequences, they compete at shared sites 7862142, and which one occupies a site is determined by which protein is present - EGR1 only transiently after stimulation, WT1 only in the restricted lineages that express it. The G-rich matrices in the databases, by contrast, point at CpG-island promoters generally, where MAZ-, ZNF281- and SP-family sites are also abundant.
Biology and disease
Wt1-null mice fail at an early step of kidney development, with defective ureteric bud outgrowth and no metanephros, together with gonadal and mesothelial defects (Kreidberg et al. 1993) 8395349. In humans, germline WT1 mutations cause two distinct syndromes that map onto the protein's two functions. Denys-Drash syndrome arises from germline missense mutations concentrated in the zinc fingers and combines urogenital malformation, nephropathy and Wilms tumor (Pelletier et al. 1991) 1655284. Frasier syndrome arises from donor splice-site mutations in intron 9 that specifically abolish the +KTS isoform, producing gonadal dysgenesis and progressive glomerulopathy without the same Wilms tumor risk (Barbaux et al. 1997) 9398852 - direct evidence that the KTS isoform ratio, not just total WT1, matters in vivo. Somatic WT1 mutations occur in a minority of Wilms tumors, and WT1 is also reported to be highly expressed in acute leukemias 28811308.
Reading a hit
Be careful here, because the label and the matrix disagree. The structurally and biochemically characterized WT1 site is EGR-like (the EGR-1 consensus, or the WTE), but every informative WT1 matrix in these databases is a ChIP-derived G-rich model at grade B 37971293. A hit to one of those models tells you the sequence belongs to the GGGAGGG family of G-rich promoter elements - shared with MAZ, PATZ1, ZNF281, ZNF148, VEZF1 and at least one ZBTB17 matrix - and does not on its own establish that WT1 binds there, though the longer HOCOMOCO model is 8 of 10 matched to the WTE. Conversely, a strong hit to an EGR matrix is also a candidate WT1 site, and one Vierstra archetype is labelled AC0519:EGR/WT precisely because the two could not be separated. Ignore the two uninformative entries: AC0095:WT/FOXC is an AT-rich homeodomain-like model, and M00243_2.00 carries essentially no information. Two further constraints are biological rather than sequence-based. WT1 is expressed in few cell types, so in most samples there is no WT1 available to bind. And even in those cell types, roughly half the WT1 protein may be the +KTS isoform, which binds DNA poorly because the three-residue insertion sits between fingers 3 and 4 32955251.
How this was worked out
WT1 is cloned at the 11p13 Wilms tumor locus as a four-zinc-finger gene, and site selection with its finger region recovers sequences similar to the EGR-1 consensus; a finger mutation from a Wilms tumor patient abolishes binding.
Still open then: Whether WT1 prefers a site of its own, distinct from EGR1's, was not known.
Germline point mutations in the zinc fingers of one WT1 copy, nine of ten in finger 3, are found in Denys-Drash syndrome and directly affect DNA recognition, tying the DNA-binding domain to urogenital development.
Wt1-null mouse embryos fail to form kidneys and gonads: the metanephric blastema undergoes apoptosis and the ureteric bud does not grow out, with additional mesothelial, heart and lung defects.
Whole-genome PCR selection identifies the WTE, GCGTGGGAGT, which WT1 binds with 20- to 30-fold higher affinity than the EGR-1 site; the A at position 8 and the 3' T contribute most, fingers 2-4 are required for recognition, and finger 1 raises affinity about tenfold.
Still open then: Whether native WT1 targets in kidney use WTE-type sites rather than EGR-type sites was not established.
- 1995mediumImmunolocalization, immunoprecipitation; RNA binding assays; in vitro RNA selection with isoform comparison7736591 8755514 9512553
The KTS splice choice decides whether WT1 sits mainly with splicing factors or in transcription-factor domains, and WT1, unlike EGR1, binds specific Igf-2 RNA sequences through its fingers (1996), proposing a post-transcriptional role. RNA selection in 1998, however, found that the +KTS isoforms could not bind the selected RNAs.
Still open then: Which isoform carries the RNA-binding function was left contested.
Frasier syndrome is caused by donor splice-site mutations in intron 9 that are predicted to remove the +KTS isoform, giving gonadal dysgenesis and progressive glomerulopathy without Wilms tumor: direct evidence that the KTS isoform ratio matters in vivo.
X-ray and NMR structures of the four-finger domain on 14- and 17-bp DNA show fingers 2-4 deep in the major groove making base-specific contacts, while finger 1 follows the groove in a non-canonical orientation without base contacts, anchoring WT1 and raising affinity but adding little specificity.
The ChIP-derived WT1 models in HOCOMOCO v14 are G-rich (GCGGGGGAGG, GGGGGAGGA) and lack the EGR signature CACGC, sitting among the SP/KLF, ZNF281 and EGR models; the longer one still matches the biochemically defined WTE at 8 of 10 positions, so the models depart from the EGR-1 site without being unrelated to the selected WT1 site.
Still open then: Whether these models reflect direct WT1 binding or co-occupancy of G-rich promoters is untested.
Still unresolved
- Whether the ChIP-derived WT1 matrices are indirect or a genuine WT1 preference is unresolved: they cluster with the MAZ and ZNF281 G-rich motifs, yet the more informative one, GCGGGGGAGG, matches the biochemically defined WTE (GCGTGGGAGT) at 8 of 10 positions, including the A at position 8 that contributes most to WTE affinity.
- Which WT1 isoform binds RNA is contested: the +KTS form co-localizes with splicing factors and has been described as the RNA-binding isoform, but in vitro RNA selection found that +KTS isoforms could not bind the selected RNAs while the -KTS isoforms could, and a later NMR study found the two isoforms binding a stem-loop RNA with similar affinity.
- Whether WT1 acts as an oncogene or a tumor suppressor in leukemia is unsettled: it is highly expressed in many acute leukemias, while the somatic WT1 mutations found in acute myeloid leukemia are mostly frameshifts expected to inactivate the protein.
The WT1 label and the WT1 matrices point in different directions. Biochemistry and structure say WT1 binds an EGR-like GC-rich site (the EGR-1 consensus, or the WTE GCGTGGGAGT); all informative database matrices here are ChIP-derived, grade B, and G-rich (GGGGGAGGA), sitting among the SP/KLF, ZNF281 and EGR models, although the longer one (GCGGGGGAGG) does match the WTE at 8 of 10 positions. Family cohesion is only 0.57, one matrix has 0.1 bits, and one Vierstra archetype labelled WT is an AT-rich homeodomain model. A hit therefore identifies a G-rich promoter element and does not establish WT1 occupancy. Also check that WT1 is expressed at all, and remember that the +KTS isoform binds DNA poorly.
References
- Call KM, Glaser T, Ito CY, et al. Isolation and characterization of a zinc finger polypeptide gene at the human chromosome 11 Wilms' tumor locus. Cell 1990;60:509-20. PMID 2154335
- Rauscher FJ 3rd, Morris JF, Tournay OE, et al. Binding of the Wilms' tumor locus zinc finger protein to the EGR-1 consensus sequence. Science 1990;250:1259-62. PMID 2244209
- Pelletier J, Bruening W, Kashtan CE, et al. Germline mutations in the Wilms' tumor suppressor gene are associated with abnormal urogenital development in Denys-Drash syndrome. Cell 1991;67:437-47. PMID 1655284
- Kreidberg JA, Sariola H, Loring JM, et al. WT-1 is required for early kidney development. Cell 1993;74:679-91. PMID 8395349
- Nakagama H, Heinrich G, Pelletier J, et al. Sequence and structural requirements for high-affinity DNA binding by the WT1 gene product. Mol Cell Biol 1995;15:1489-98. PMID 7862142
- Larsson SH, Charlieu JP, Miyagawa K, et al. Subnuclear localization of WT1 in splicing or transcription factor domains is regulated by alternative splicing. Cell 1995;81:391-401. PMID 7736591
- Caricasole A, Duarte A, Larsson SH, et al. RNA binding by the Wilms tumor suppressor zinc finger proteins. Proc Natl Acad Sci U S A 1996;93:7562-6. PMID 8755514
- Barbaux S, Niaudet P, Gubler MC, et al. Donor splice-site mutations in WT1 are responsible for Frasier syndrome. Nat Genet 1997;17:467-70. PMID 9398852
- Bardeesy N, Pelletier J. Overlapping RNA and DNA binding domains of the wt1 tumor suppressor gene product. Nucleic Acids Res 1998;26:1784-92. PMID 9512553
- Stoll R, Lee BM, Debler EW, et al. Structure of the Wilms tumor suppressor protein zinc finger domain bound to DNA. J Mol Biol 2007;372:1227-45. PMID 17716689
- Hastie ND. Wilms' tumour 1 (WT1) in development, homeostasis and disease. Development 2017;144:2862-2872. PMID 28811308
- Vorontsov IE, Eliseeva IA, Zinkevich A, et al. HOCOMOCO in 2024: a rebuild of the curated collection of binding models for human and mouse transcription factors. Nucleic Acids Res 2024;52:D154-D163. PMID 37971293
- Nishikawa T, Wojciak JM, Dyson HJ, et al. RNA Binding by the KTS Splice Variants of Wilms' Tumor Suppressor Protein WT1. Biochemistry 2020;59:3889-3901. PMID 32955251
5 motifs in this family — JASPAR 1 · HOCOMOCO 2 · CIS-BP 1 · Vierstra 1
| motif | database | consensus | bits |
|---|---|---|---|
| WT1.H14CORE.1.P.BPB | HOCOMOCO | GcGgGGGAGG | 14.5 |
| Wt1 (MA1627.2) | JASPAR | CCTCCCCCAC | 13 |
| WT1.H14CORE.0.P.BPB | HOCOMOCO | GgGGGAGGa | 12 |
| AC0095:WT/FOXC:C2H2_ZFlooks like HD_Q50 | Vierstra | aTATATT | 6.7 |
| M00243_2.00 | CIS-BP | n | 0.1 |
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.