ZBTB33 (Kaiso)
ZBTB33, better known as Kaiso, is a BTB/POZ zinc-finger repressor with only three C2H2 fingers and two distinct ways of reading DNA. It binds the non-methylated consensus TCCTGCNA - the Kaiso binding site, or KBS, whose minimal core is CTGCNA - and, separately, methylated CpG dinucleotides, with higher affinity for the KBS of the two. Crystal structures show both modes use the same first two fingers, with methylcytosine read through CH-to-O hydrogen bonds. Kaiso was found as a partner of p120-catenin, which blocks its DNA binding, and it links adhesion to transcription. Note that all four database matrices for ZBTB33 describe a CGCG-containing palindrome, TCTCGCGAGA, not the KBS.
The factors
Kaiso is a single-gene family. ZBTB33 encodes a roughly 670-residue protein with an N-terminal BTB/POZ dimerization and repression domain and, unusually for a BTB-ZF factor, only three C2H2 zinc fingers at the C terminus plus a short C-terminal extension. It was isolated in a yeast two-hybrid screen for partners of the Armadillo-family adhesion protein p120-catenin 10207085, and was later identified as the DNA-binding component of a mammalian methyl-CpG-binding activity that acts as a methylation-dependent repressor 11445535; it recruits the N-CoR corepressor complex 14527417, and its Xenopus orthologue is required to keep genes silent before the mid-blastula transition 15548582. p120-catenin binding inhibits Kaiso's DNA binding in both of its modes 12087177, which is the mechanistic basis for the claim that Kaiso couples cadherin-based adhesion to transcriptional output: when p120 is released from junctions it can sequester Kaiso off DNA. Repression is delivered through recruited corepressors including N-CoR and the MTG16/CBFA2T3 corepressor 11445535,23251453; Kaiso is also regulated by phosphorylation, with modification at Thr606 driving cytoplasmic accumulation and reduced repression of CDH1 35851744. Kaiso is broadly expressed and is not restricted to a lineage.
The motif
This is the family where the phrase "dual specificity" is literal, and where the databases and the literature describe different things. Site-selection experiments defined a sequence-specific consensus, TCCTGCNA, with a minimal sufficient core of CTGCNA; two copies occur in the human and mouse matrilysin (MMP7) promoters 12087177. The same work showed that Kaiso separately possesses a methyl-CpG-dependent binding activity, requiring consecutive methylated CpGs and biochemically distinct from KBS binding, and that affinity for the KBS is the higher of the two 11445535,12087177. Only zinc fingers 2 and 3 are needed for sequence-specific binding 12087177. Crystal structures of the Kaiso zinc-finger domain bound to the KBS and to a symmetrically methylated sequence from the E-cadherin promoter explain how one small domain does both jobs: bases are read in the major groove by residues in the first two fingers, using classical hydrogen bonds for the KBS and methyl CH-to-O hydrogen bonds for methylcytosine, while the C-terminal extension past finger 3 binds the opposing minor groove and is required for high affinity 22949637. That extension is disordered in the free protein and folds on binding. Now the discrepancy: all four matrices in this family - the JASPAR ENCODE-derived MA0527.2 23693142, both HOCOMOCO KAISO models (evidence P, grade B), and CIS-BP M08094 - read TCTCGCGAGA or a close variant. That palindrome is the CGCG element found at TATA-less, CpG-island-associated bidirectional promoters 30332484. It is CpG-containing, so it is compatible with the methyl-CpG mode, but it is not the KBS, and it is not what site selection returned. In other words, the collections capture one Kaiso mode and miss the other entirely. Genome-wide work with Kaiso-knockout cells as a control reports both: CGCG and CTGCNAT are the two principal enriched motifs, with about 60 percent of sites containing both 39702290.
Where it acts
Kaiso sites are overwhelmingly promoter-proximal and CpG-island-associated. ChIP in Kaiso-knockout-controlled experiments places Kaiso at CpG islands with a preference for methylated ones and for islands with partial methylation at their 5' or 3' boundaries, including a class showing wave-like methylation 39702290. Canonical KBS-containing targets include MMP7 (matrilysin) and Wnt-pathway genes 12087177,23251453; the methylation-dependent mode has been studied at the CDH1 promoter. Because the CGCG element is a general bidirectional-promoter feature, Kaiso motif hits cluster where CpG-island promoters cluster rather than at tissue-specific enhancers. Named partners are p120-catenin (negative), N-CoR, MTG16, and, at Wnt targets, the TCF/LEF-beta-catenin machinery.
Biology and disease
Kaiso-deficient mice are viable and fertile, and on an Apc(Min) background they show reduced intestinal tumor burden and longer survival 16354691, which put Kaiso forward as a tumor-promoting repressor rather than a suppressor. Much of the disease literature concerns Kaiso's subcellular distribution in carcinoma: nuclear versus cytoplasmic Kaiso correlates with p120-catenin status and with prognosis in several tumor types 30419310, and Thr606 phosphorylation shifts the balance toward the cytoplasm, releasing CDH1 35851744. Kaiso has also been linked to clonal hematopoiesis and myelodysplastic phenotypes 39702290, and to DNA-damage responses in breast cancer cells. The tumor-biology claims are more contested than the biochemistry: reported effects of Kaiso depletion vary between cell systems and between the two binding modes, and the field has not settled how much of Kaiso's function depends on methyl-CpG reading versus KBS binding.
Reading a hit
First decide which motif you matched. A hit to a database ZBTB33 matrix is a hit to the TCTCGCGAGA CGCG element, which is a common CpG-island promoter feature also matched by other GC- and CpG-reading factors; on its own it is closer to evidence of "this is a CpG-island bidirectional promoter" than of Kaiso occupancy. A hit to TCCTGCNA is the literature KBS, but it is short, degenerate at one position, and therefore frequent by chance. Second, note that every matrix here is ChIP-derived: JASPAR MA0527.2 is ENCODE ChIP-seq and both HOCOMOCO models carry evidence code P at grade B, with no HT-SELEX or methyl-HT-SELEX model in this family's data 28473536,37971293. That is exactly the situation in which indirect binding cannot be excluded, and published Kaiso ChIP-seq datasets have been noted to conflict with the in vitro binding data 23693142,39702290. Third, methylation state is part of the specificity, and no PWM encodes it: a CpG-containing match that is unmethylated in your cell type may not be a Kaiso site at all, and vice versa. The practical recommendation is to pair a motif hit with methylation data and, where possible, with a knockout-controlled occupancy dataset.
How this was worked out
Kaiso was cloned in a two-hybrid screen for p120-catenin partners as a BTB/POZ protein with only three C2H2 fingers, placing an adhesion cofactor in contact with a transcriptional repressor.
Still open then: What sequence, if any, Kaiso recognizes.
Kaiso was identified as a methylation-dependent repressor that requires at least two symmetrically methylated CpGs, preferentially in the sequence CGCG, and as a component of the MeCP1 methyl-CpG binding activity.
Still open then: Whether Kaiso also reads unmethylated sequence specifically.
- 2002mediumrandom-oligonucleotide site selection, EMSA with finger deletions, competition assays12087177
Site selection defined the unmethylated Kaiso binding site TCCTGCNA with the minimal core CTGCNA, showed that fingers 2 and 3 suffice for it, established that the KBS is bound with higher affinity than methyl-CpG sites, and showed that p120-catenin blocks both modes.
Still open then: How one three-finger domain can read two unrelated kinds of site.
Kaiso-null mice were found to be viable and fertile with no overt developmental or expression phenotype, but crossing them to Apc(Min) mice delayed intestinal tumorigenesis, which framed Kaiso as tumor-promoting rather than as an essential repressor.
Crystal structures of the Kaiso zinc-finger domain on the KBS and on a symmetrically methylated E-cadherin sequence showed that the same first two fingers read both, using classical hydrogen bonds for the KBS and methyl CH-to-O bonds for methylcytosine, with a C-terminal extension that is disordered until it folds into the opposing minor groove.
Analysis of ENCODE Kaiso ChIP-seq together with methylation maps reported that Kaiso occupies unmethylated, highly active promoters in vivo rather than methylated DNA, contradicting the model built from the in vitro methyl-binding work and leaving the two literatures in open conflict.
Still open then: Whether the discrepancy reflects antibody quality, chromatin accessibility or a genuinely different in vivo mode.
- 2018mediummotif analysis of CpG-island promoters, nascent-transcript mapping, reporter assays30332484
The TCTCGCGAGA palindrome that all the ZBTB33 matrices describe was characterized in its own right as a conserved, methylation-sensitive element that drives bidirectional transcription at TATA-less CpG-island promoters, showing that the catalogued Kaiso motif is also a general core-promoter feature.
- 2024mediumknockout-controlled ChIP-seq, methylation profiling, binding-deficient mutant comparison39702290
Kaiso ChIP in Caki-1 cells with a Kaiso-knockout line as the negative control recovered both motifs - CGCG and CTGCNAT, with about 60 percent of sites carrying both - and placed Kaiso at CpG islands with a preference for methylated and partially methylated ones, partially reconciling the in vitro and in vivo pictures.
Still unresolved
- Whether Kaiso prefers methylated or unmethylated DNA in cells is still contested: an ENCODE-based analysis concluded that it occupies unmethylated, actively transcribed promoters, while knockout-controlled ChIP in kidney carcinoma cells reports enrichment at methylated and partially methylated CpG islands, and the crystal structures show that the protein is capable of both.
- Which of the two binding modes carries Kaiso's biological function is unresolved, since no study has separated KBS-dependent from methyl-CpG-dependent target genes in the same cells with separation-of-function mutants beyond the single E535A allele.
- It is unclear how much of a match to the catalogued TCTCGCGAGA matrix says anything about Kaiso at all, because the same element independently drives bidirectional transcription at TATA-less CpG-island promoters and would be recovered in any ChIP experiment centered on such promoters.
- Whether Kaiso promotes or restrains tumor growth in humans is unsettled: loss delays intestinal tumorigenesis in Apc(Min) mice, whereas the correlative human data depend on whether the protein is nuclear or cytoplasmic and point in both directions across tumor types.
Kaiso has two binding modes and the databases represent only one. All four ZBTB33 matrices here read the CGCG palindrome TCTCGCGAGA - a generic CpG-island bidirectional-promoter element - rather than the site-selection consensus TCCTGCNA, so the same gene symbol can mean two different motifs depending on the source. Every matrix in this family is ChIP-derived (ENCODE JASPAR; HOCOMOCO evidence P, grade B), with no in vitro model, and published Kaiso ChIP datasets are known to conflict with in vitro data, so indirect binding is a live possibility. Methylation state is part of the specificity and no position weight matrix captures it; interpret hits alongside methylation data.
References
- Daniel JM, Reynolds AB. The catenin p120(ctn) interacts with Kaiso, a novel BTB/POZ domain zinc finger transcription factor. Mol Cell Biol 1999;19:3614-23. PMID 10207085
- Daniel JM, Spring CM, Crawford HC, et al. The p120(ctn)-binding partner Kaiso is a bi-modal DNA-binding protein that recognizes both a sequence-specific consensus and methylated CpG dinucleotides. Nucleic Acids Res 2002;30:2911-9. PMID 12087177
- Prokhortchouk A, Hendrich B, Jørgensen H, et al. The p120 catenin partner Kaiso is a DNA methylation-dependent transcriptional repressor. Genes Dev 2001;15:1613-8. PMID 11445535
- Buck-Koehntop BA, Stanfield RL, Ekiert DC, et al. Molecular basis for recognition of methylated and specific DNA sequences by the zinc finger protein Kaiso. Proc Natl Acad Sci U S A 2012;109:15229-34. PMID 22949637
- Prokhortchouk A, Sansom O, Selfridge J, et al. Kaiso-deficient mice show resistance to intestinal cancer. Mol Cell Biol 2006;26:199-208. PMID 16354691
- Starshin A, Abramov P, Lobanova Y, et al. Dissecting the Kaiso binding profile in clear renal cancer cells. Epigenetics Chromatin 2024;17:38. PMID 39702290
- Barrett CW, Smith JJ, Lu LC, et al. Kaiso directs the transcriptional corepressor MTG16 to the Kaiso binding site in target promoters. PLoS One 2012;7:e51205. PMID 23251453
- Tian W, Yuan H, Qin S, et al. Kaiso phosphorylation at threonine 606 leads to its accumulation in the cytoplasm, reducing its transcriptional repression of the tumour suppressor CDH1. Mol Oncol 2022;16:3192-3209. PMID 35851744
- Pierre CC, Hercules SM, Yates C, et al. Dancing from bottoms up - Roles of the POZ-ZF transcription factor Kaiso in Cancer. Biochim Biophys Acta Rev Cancer 2019;1871:64-74. PMID 30419310
- Mahpour A, Scruggs BS, Smiraglia D, et al. A methyl-sensitive element induces bidirectional transcription in TATA-less CpG island-associated promoters. PLoS One 2018;13:e0205608. PMID 30332484
- Yin Y, Morgunova E, Jolma A, et al. Impact of cytosine methylation on DNA binding specificities of human transcription factors. Science 2017;356. PMID 28473536
- 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
- Blattler A, Yao L, Wang Y, et al. ZBTB33 binds unmethylated regions of the genome associated with actively expressed genes. Epigenetics Chromatin 2013;6:13. PMID 23693142
- Yoon HG, Chan DW, Reynolds AB, et al. N-CoR mediates DNA methylation-dependent repression through a methyl CpG binding protein Kaiso. Mol Cell 2003;12:723-34. PMID 14527417
- Ruzov A, Dunican DS, Prokhortchouk A, et al. Kaiso is a genome-wide repressor of transcription that is essential for amphibian development. Development 2004;131:6185-94. PMID 15548582
4 motifs in this family — JASPAR 1 · HOCOMOCO 2 · CIS-BP 1
| motif | database | consensus | bits |
|---|---|---|---|
| M08094_2.00 | CIS-BP | TCTCGCGAGabc | 14.3 |
| ZBTB33 (MA0527.2) | JASPAR | TCTCGCGAGa | 13.1 |
| KAISO.H14CORE.1.P.BPB | HOCOMOCO | gatctCGCGAGA | 12.2 |
| KAISO.H14CORE.0.P.BPB | HOCOMOCO | ctCGCGAGA | 11.9 |
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.