BCL11A / BCL11B
BCL11A and BCL11B (originally CTIP1 and CTIP2) are paralogous C2H2 zinc-finger repressors with six fingers, of which a C-terminal cluster binds DNA while the N-terminal region recruits NuRD. They are best known not as motif proteins but as switches: BCL11A silences the fetal gamma-globin genes and is the target of approved gene-editing therapies for sickle cell disease and beta-thalassemia, while BCL11B enforces T-lineage commitment, so that its loss converts T cells into natural-killer-like cells. Reported DNA preferences conflict. In vitro selection on BCL11A gave a GC-box-like GGCCGG core, whereas the site it occupies at the gamma-globin promoter is written TGACCA. The family's matrices are heterogeneous and several look like other factors' motifs.
The factors
BCL11A and BCL11B were isolated as COUP-TF-interacting proteins 1 and 2 (CTIP1 and CTIP2) 10744719 and are also known as EVI9 (BCL11A) and RIT1 (Bcl11b in mouse). Both are large C2H2 zinc-finger proteins with six fingers: an N-terminal region that includes the COUP-TF interaction domains and a short NuRD-interacting motif, and a C-terminal cluster of fingers that carries the DNA-binding activity. Neither is a simple monomeric binder. BCL11A forms a tetramer through a single zinc finger (ZnF0), and tetramer formation is required both for steady-state protein accumulation and for gamma-globin repression, because an engineered monomer fails to engage its corepressor complex 39607926. BCL11A's original in vitro selection likewise recovered binding by an oligomeric complex rather than a monomer 12196208.
Expression is tissue-restricted and largely complementary. BCL11A is expressed in B cells, adult erythroid cells and brain; BCL11B is T-cell-specific among hematopoietic cells and is also expressed in skin, tooth and central nervous system 10744719. Both repress transcription, and the repression seen in reporter assays was not reversed by the class I/II HDAC inhibitor trichostatin A 12196208.
The motif
The literature gives two answers and they do not agree. A selection procedure for high-affinity CTIP1 (BCL11A) sites recovered the core GGCCGG, described at the time as closely related to the canonical GC box 12196208. Separately, the functionally best-characterized BCL11A element is in the HBG1 and HBG2 (gamma-globin) promoters about 115 base pairs upstream of the transcription start site, where BCL11A binds directly and where naturally occurring hereditary-persistence-of-fetal-hemoglobin point mutations disrupt binding and raise gamma-globin expression 29610478; the JASPAR matrix built from that class of data has the consensus CTGACCA. The family data consensus here, TGACC, follows the second form, and we give TGACCA as the headline consensus because it is the site with direct functional validation. But the GGCCGG form is not an error: two of the eight matrices assigned to this family (a CIS-BP BCL11B model with consensus TCCGGCC and a Vierstra archetype with THCGGCC) are GGCCGG-related, and they are the ones that agree with the in vitro selection.
Cohesion in this family is the lowest in the zinc-finger set at 0.50, and the individual matrices are worth listing because they illustrate what goes wrong. Besides the TGACC and GGCCGG groups, a CIS-BP BCL11A model and a Vierstra archetype share the consensus TTTCACTTCCTC, and the archetype is named BCL11A/IRF, reflecting the IRF/ISRE-like TTTC core. One Vierstra BCL11B archetype has a long, high-information consensus (GTGAACGCNDNHGCTACAC) with no close neighbor anywhere in the four databases. And the JASPAR Bcl11B matrix has the consensus AAACCACAA, whose reverse complement, TTGTGGTTT, contains the RUNX core TGTGGT; its nearest neighbors across all four databases are RUNX models rather than BCL11 models, and it is flagged in these data as belonging with RUNX. Since RUNX1 and RUNX3 are dominant occupants of T-cell regulatory elements, where BCL11B ChIP is performed, the simplest reading is indirect binding or co-occupancy at RUNX sites rather than a genuine BCL11B specificity; we are not aware of literature that settles this either way.
Where it acts
BCL11A acts through a small number of intensively studied elements rather than a diffuse cistrome: sites in the beta-globin cluster, its own erythroid-specific +58 intronic enhancer (whose common variants set fetal-hemoglobin level) 24115442, and B-lymphoid and neural regulatory regions. BCL11B occupies T-lineage enhancers and promoters, frequently alongside RUNX, TCF7, GATA3 and ETS factors. Both proteins sit in NuRD-containing complexes, so much of their genomic footprint is complex-driven.
Biology and disease
BCL11A is the developmental-stage-specific repressor of fetal hemoglobin: it is expressed in adult but not fetal erythroid cells, knockdown reactivates HbF, and it occupies discrete sites in the beta-globin cluster 19056937. Because its erythroid enhancer is dispensable outside erythroid cells, disrupting that enhancer raises HbF without compromising B-cell development 24115442, which is the basis of the CRISPR-Cas9 therapy now approved for sickle cell disease and transfusion-dependent beta-thalassemia 33283989. Heterozygous BCL11A microdeletions and variants (2p15-p16.1) cause an intellectual-disability and autism-spectrum phenotype (Dias-Logan syndrome) with persistent fetal hemoglobin but otherwise near-normal blood counts 25938782.
BCL11B is the T-lineage commitment factor. It is first expressed immediately before commitment, suppresses natural-killer-cell and progenitor programs 20595614, and is required for alpha-beta T-cell differentiation and survival 12717433; deleting it converts T cells at all developmental stages into natural-killer-like cells that retain cytotoxic activity 20538915. In disease, BCL11B is targeted by enhancer-hijacking translocations in T-cell and T/myeloid acute leukemias 34103329, and germline heterozygous variants cause immune dysregulation with severe atopic disease 34887873.
Reading a hit
Do not trust a BCL11 motif hit. The family has no single validated consensus: the in vitro selection and the functionally validated gamma-globin element point at different sequences, and both are short (five to seven informative bases). Several matrices in the family are almost certainly other factors' motifs recovered from ChIP data, most clearly the JASPAR Bcl11B matrix that scores as a RUNX site. Nothing distinguishes BCL11A from BCL11B in sequence. And because BCL11A works as a tetramer bound to a corepressor complex and BCL11B works within NuRD, occupancy in vivo depends on partners as much as on sequence, so peaks are a poor guide to intrinsic specificity and motif hits are a poor guide to peaks. Practical advice: for BCL11A, work from the known validated elements (the gamma-globin -115 region and the +58 enhancer) rather than from a genome-wide scan; for BCL11B, check whether your hit is better explained as a RUNX site before attributing it here.
How this was worked out
CTIP1 and CTIP2, later renamed BCL11A and BCL11B, are isolated as COUP-TF-interacting zinc-finger proteins expressed in brain and the immune system and shown to repress transcription.
Still open then: Whether they bound DNA themselves or acted only as COUP-TF cofactors.
In vitro selection gives CTIP1/BCL11A a high-affinity core of GGCCGG, closely related to the GC box, recovered by an oligomeric rather than monomeric complex, and shows repression independent of COUP-TF and insensitive to trichostatin A.
Still open then: The selected GC-box-like site was never matched to a validated genomic element.
Bcl11b is shown to be required for alpha-beta T-cell differentiation and survival in mouse, giving the paralogue a defined developmental role.
BCL11A is identified as the developmental-stage-specific repressor of fetal hemoglobin: it is present in adult but not fetal erythroid cells, its knockdown reactivates gamma-globin, and it occupies discrete sites in the beta-globin cluster.
Still open then: Which sequence BCL11A reads at those sites was not resolved; the occupied regions did not match the GGCCGG selection.
- 2010highStage-specific conditional deletion in mouse, transcriptional profiling, functional assays20595614 20538915
Two studies show Bcl11b is the T-lineage commitment factor: it is required at the commitment checkpoint, and deleting it at any stage converts T cells into natural-killer-like cells.
- 2013mediumFine-mapping of GWAS variants, enhancer dissection with reporter and genome editing24115442
An erythroid-specific enhancer in BCL11A intron 2, containing the common variants that set fetal-hemoglobin level, is identified and shown to be dispensable outside erythroid cells, which converts a transcription factor that cannot be drugged into a cell-type-specific target.
Hereditary-persistence-of-fetal-hemoglobin point mutations at the gamma-globin promoter are shown to act by disrupting direct BCL11A binding, giving the family its one element where a specific sequence, a bound factor and a human phenotype all line up; the matrix from this class of data reads CTGACCA, not the GGCCGG of the in vitro selection.
Still open then: Why the functionally validated site and the selected site are unrelated sequences is unexplained.
- 2024highStructural and biochemical analysis of the oligomerization domain, engineered monomer in erythroid cells39607926
BCL11A is shown to work as a tetramer assembled through a single zinc finger, with tetramerization required for protein accumulation and for gamma-globin silencing, so the unit that engages DNA and the corepressor complex is not a monomer and intrinsic specificity cannot be read off a monomer-based matrix.
Still unresolved
- The family has two incompatible consensus sequences and no accepted reconciliation: in vitro selection on BCL11A gives a GC-box-like GGCCGG core, while the functionally validated gamma-globin element and the matrices derived from that class of data give TGACCA.
- Whether the JASPAR Bcl11B matrix reflects any BCL11B specificity is unresolved: its reverse complement contains the RUNX core TGTGGT and its nearest neighbors in all four databases are RUNX models, which is consistent with indirect binding at RUNX-occupied T-cell elements, but no published experiment settles it.
- How a tetrameric BCL11A engages DNA is unknown, including whether the four subunits contact several copies of a short core or whether only one subunit binds sequence specifically, which determines whether any single-site matrix can describe this family.
- Whether BCL11A and BCL11B differ in intrinsic specificity at all has not been tested; they are pooled here because their matrices are not separable, yet their biological targets barely overlap.
Trust this family least. There is no agreed consensus: in vitro selection on BCL11A gave a GC-box-like GGCCGG core, while the functionally validated site at the gamma-globin promoter is written TGACCA, and both appear among the family's matrices. Cohesion is 0.50. At least three matrices look like other factors: two Vierstra/CIS-BP models with an IRF-like TTTCACTTCC consensus (one archetype is named BCL11A/IRF), and the JASPAR Bcl11B matrix whose AAACCACAA consensus is the RUNX core and whose nearest neighbors in all four databases are RUNX models, most plausibly indirect binding at RUNX sites in T cells. BCL11A and BCL11B are not separable by sequence, and both act within corepressor complexes.
References
- Avram D, Fields A, Pretty On Top K, et al. Isolation of a novel family of C(2)H(2) zinc finger proteins implicated in transcriptional repression mediated by chicken ovalbumin upstream promoter transcription factor (COUP-TF) orphan nuclear receptors. J Biol Chem 2000;275:10315-22. PMID 10744719
- Avram D, Fields A, Senawong T, et al. COUP-TF (chicken ovalbumin upstream promoter transcription factor)-interacting protein 1 (CTIP1) is a sequence-specific DNA binding protein. Biochem J 2002;368:555-63. PMID 12196208
- Sankaran VG, Menne TF, Xu J, et al. Human fetal hemoglobin expression is regulated by the developmental stage-specific repressor BCL11A. Science 2008;322:1839-42. PMID 19056937
- Bauer DE, Kamran SC, Lessard S, et al. An erythroid enhancer of BCL11A subject to genetic variation determines fetal hemoglobin level. Science 2013;342:253-7. PMID 24115442
- Martyn GE, Wienert B, Yang L, et al. Natural regulatory mutations elevate the fetal globin gene via disruption of BCL11A or ZBTB7A binding. Nat Genet 2018;50:498-503. PMID 29610478
- Zheng G, Yin M, Mehta S, et al. A tetramer of BCL11A is required for stable protein production and fetal hemoglobin silencing. Science 2024;386:1010-1018. PMID 39607926
- Frangoul H, Altshuler D, Cappellini MD, et al. CRISPR-Cas9 Gene Editing for Sickle Cell Disease and β-Thalassemia. N Engl J Med 2021;384:252-260. PMID 33283989
- Basak A, Hancarova M, Ulirsch JC, et al. BCL11A deletions result in fetal hemoglobin persistence and neurodevelopmental alterations. J Clin Invest 2015;125:2363-8. PMID 25938782
- Li L, Leid M, Rothenberg EV. An early T cell lineage commitment checkpoint dependent on the transcription factor Bcl11b. Science 2010;329:89-93. PMID 20595614
- Li P, Burke S, Wang J, et al. Reprogramming of T cells to natural killer-like cells upon Bcl11b deletion. Science 2010;329:85-9. PMID 20538915
- Wakabayashi Y, Watanabe H, Inoue J, et al. Bcl11b is required for differentiation and survival of alphabeta T lymphocytes. Nat Immunol 2003;4:533-9. PMID 12717433
- Lu HY, Sertori R, Contreras AV, et al. A Novel Germline Heterozygous BCL11B Variant Causing Severe Atopic Disease and Immune Dysregulation. Front Immunol 2021;12:788278. PMID 34887873
- Montefiori LE, Bendig S, Gu Z, et al. Enhancer Hijacking Drives Oncogenic BCL11B Expression in Lineage-Ambiguous Stem Cell Leukemia. Cancer Discov 2021;11:2846-2867. PMID 34103329
8 motifs in this family — JASPAR 2 · CIS-BP 2 · Vierstra 4
| motif | database | consensus | bits |
|---|---|---|---|
| AC0361:BCL11B:C2H2_ZF | Vierstra | GTGAACGcndnhgCTACAC | 24.4 |
| M07926_2.00 | CIS-BP | TtTCAcTTCCtC | 14.3 |
| AC0677:BCL11A/IRF:C2H2_ZF | Vierstra | TtTCAcTTCCTCt | 11.9 |
| Bcl11B (MA1989.2)looks like RUNX | JASPAR | AAACCACAa | 10.6 |
| BCL11A (MA2324.1) | JASPAR | CTGACCA | 9.3 |
| M00986_2.00 | CIS-BP | TccGGcC | 7.2 |
| AC0139:BCL11A/BCL11B:C2H2_ZF | Vierstra | TGaCC | 6.3 |
| AC0307:BCL11A/BCL11B:C2H2_ZF | Vierstra | ThcGGcC | 5.9 |
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.