C/EBP
The C/EBP family - CEBPA, CEBPB (NF-IL6), CEBPD, CEBPE, CEBPG, with DDIT3 (CHOP) as a heterodimer partner - binds the palindromic element TT(G/A)CGCAA, most often written as the 10-mer ATTGCGCAAT. These are basic region-leucine zipper proteins; the leucine zipper itself was first described in C/EBP, and the family's distinctive DNA contacts centre on an arginine that is conserved in C/EBPs but not in other bZIPs. Family members control hepatic and adipose metabolism, the acute-phase response, and granulocyte maturation, and CEBPA and CEBPE are mutated in human myeloid disease. The paralogues are indistinguishable from sequence. The boundary with the CREB/ATF family is genuinely blurred: CHOP-containing dimers and ATF4 read a hybrid half-C/EBP, half-CRE site, and CpG methylation converts a CRE into a C/EBP-alpha site.
The factors
Five C/EBP proteins bind DNA as this family: CEBPA, CEBPB (also NF-IL6 or LAP/LIP), CEBPD, CEBPE and CEBPG. All are basic region-leucine zipper (bZIP) factors, and the leucine zipper was in fact first proposed from the C/EBP sequence 3289117, with the scissors-grip model of DNA recognition advanced for the same family a year later 2683088. CEBPA, CEBPB, CEBPD and CEBPE have N-terminal activation domains; CEBPG has essentially none and works mainly through heterodimers. Expression differs sharply even though the motif does not: CEBPB and CEBPG are broad, CEBPA is concentrated in liver, adipose tissue, lung and granulocytes, CEBPD is stress- and cytokine-inducible, and CEBPE is restricted to the granulocyte and eosinophil lineages. Alternative translation initiation produces short LIP isoforms of CEBPA and CEBPB that dimerize and occupy sites without activating, so the same motif can be read as an activator or a competitive repressor depending on isoform ratio. DDIT3 (CHOP/GADD153) is included here because it is an obligate C/EBP partner, but its basic region is substituted at positions that matter for C/EBP-site contact, so CHOP-C/EBP heterodimers bind a different, hybrid sequence rather than the C/EBP palindrome 8657121. C/EBPs also pair outside the family, notably with ATF4 and with NF-kappa-B subunits.
The motif
The consensus is the palindrome TT(G/A)CGCAA, usually presented with flanks as ATTGCGCAAT 8632009. It is an inverted repeat of the half-site TTGC, one half per subunit. The C/EBP-alpha crystal structure resolved the contacts across ATTGCGCAAT and identified Arg289 as the keystone residue: it hydrogen-bonds to an adenine in the half-site, to the DNA backbone, and to Asn292, the asparagine invariant across bZIPs 12578822. A second residue explains the family boundary. Val296, which is alanine in most other bZIPs, discriminates against purines and restrains the invariant Arg300; mutating it to alanine strongly increased C/EBP-alpha binding to CRE sites while leaving C/EBP-site affinity intact 12578822. In other words a single side chain, not a different reading frame, separates the C/EBP site from the cyclic-AMP response element. Common variants replace the central CG: ATTGCACAAT and ATTGCGTAAT are both well represented among the 25 models here, as are half-degenerate forms in which only one TTGC half is intact. Methylation adds a twist in the same direction: CpG methylation of the CRE TGACGTCA, and of the half-sites CGTCA and CGCAA, increases C/EBP-alpha binding, and methylated promoters bound by C/EBP-alpha lose that binding when demethylated 21059933. The paralogues are not distinguishable - JASPAR CEBPB MA0466.4 and CEBPE MA0837.3 are 0.997 similar, and every CEBPA, CEBPD and CEBPG model here falls within 0.95 of the family medoid. What does look different is the CHOP-partnered and ATF4-adjacent set: JASPAR Ddit3::Cebpa, HOCOMOCO DDIT3 and JASPAR CEBPG MA1636.2 all read ATGATGCAAT-like sequences, which is the C/EBP-ATF composite element (CARE), half CRE and half C/EBP 10085237.
Where it acts
C/EBP sites occur at both promoters and enhancers, with less CpG-island bias than the CRE. In macrophages and adipocytes they are among the motifs that lineage-determining factors use when priming enhancers, and C/EBP-beta co-occupies acute-phase promoters with NF-kappa-B while C/EBP-alpha and C/EBP-beta act with PPARG at adipocyte enhancers and with RUNX1 and SPI1 at myeloid enhancers. The named composite element is the CARE, TTGCATCA or TGATGCAAT, at integrated-stress-response genes including DDIT3 itself, TRIB3, ASNS and ATF3; there C/EBP-family proteins and ATF4 occupy the same element 10085237. On frequency: the exact palindrome TTGCGCAA is expected of order 10^4 times in 3.1 Gb before allowing for the genome's CpG depletion, while its non-CpG relatives TTGCACAA and TTGTGCAA are several-fold more common, so scans at usual thresholds return large numbers of C/EBP-like matches.
Biology and disease
Cebpa-null mice die soon after birth from hypoglycemia, with no hepatic glycogen and disrupted adipose tissue 7652557. Cebpb-null mice survive but their macrophages fail at bacterial killing and tumor cytotoxicity 7530603. Cebpe-null mice have impaired granulopoiesis, myelodysplasia and early lethality 9371821, and in humans loss-of-function CEBPE mutation causes neutrophil-specific granule deficiency, a disorder of defective secondary granules and recurrent bacterial infection 10359588. Dominant-negative CEBPA mutations that truncate the N-terminus define a recurrent subtype of acute myeloid leukemia 11242107, and germline CEBPA variants cause familial AML 15575056. C/EBP-beta and C/EBP-delta are the transcriptional arm of the IL-6 acute-phase response. DDIT3/CHOP, acting with ATF4, raises protein synthesis during sustained ER stress to the point of cell death 23624402, which makes the CARE element a control point for stress-induced apoptosis.
Reading a hit
A full ATTGCGCAAT palindrome is among the more distinctive bZIP matches: TTGC half-sites are not shared with AP-1, CREB/ATF, MAF or NFE2/BACH, so the family call is comparatively safe. The member call is not available at all. CEBPA, CEBPB, CEBPD, CEBPE and CEBPG give one motif, and which of them is bound is a question about expression, not sequence. A lone TTGC half-site is not interpretable. The real ambiguity is with the C/EBP-ATF composite: if your match reads ATGATGCAAT or TTGCATCA, it is a CARE, and ATF4 in the CREB/ATF family accounts for it as well as any C/EBP does; the DDIT3 and CEBPG matrices here are the same element. Methylation cuts both ways: a methylated CRE can be a genuine C/EBP-alpha site, and a CGCAA-containing hit may require methylation to be occupied. None of the 25 models is flagged as belonging elsewhere, and the in vitro (HT-SELEX and methyl-SELEX) models agree with the ChIP-derived ones, which is the reassuring case for a ChIP-heavy family.
How this was worked out
A periodic array of leucines every seventh residue in C/EBP, also present in Fos, Jun, Myc and GCN4, is proposed to form an interdigitating dimerization helix, the leucine zipper.
Still open then: The zipper was a hypothesis, and how the adjacent region contacts DNA was not addressed.
The scissors-grip model places the basic regions of a zipper-linked dimer in the major groove of a dyad-symmetric site, explaining why bZIP sites are abutted half-sites.
Still open then: No atomic structure of a bZIP-DNA complex yet existed.
Knockouts separate member functions despite a shared motif: Cebpa-null mice die of neonatal hypoglycemia, Cebpb-null macrophages fail to kill bacteria, and Cebpe-null mice (1997) fail to make functional neutrophils.
Random site selection shows the C/EBP family members share one consensus, RTTGCGYAAY, so the paralogues are not separable by sequence preference.
CHOP-C/EBP heterodimers cannot bind classic C/EBP sites but select a distinct core, PuPuPuTGCAAT(A/C)CCC, giving CHOP a dual role as inhibitor and direct activator.
Still open then: How much CHOP's own DNA binding contributes in cells, as against sequestering C/EBP partners, was unresolved.
At the Gadd153/CHOP promoter, C/EBP-beta, ATF4 and ATF3 bind in sequence to one C/EBP-ATF composite site during stress, defining the half-CRE, half-C/EBP element shared with the CREB/ATF family.
The C/EBP-alpha bZIP crystal structure on ATTGCGCAAT shows Arg289 as the keystone contact and Val296 as a residue that excludes CRE binding; V296A gains CRE affinity without losing C/EBP sites.
- 2010highDNA-binding assays, transfection, ChIP in primary keratinocytes, 5-azacytidine and DNMT1 depletion21059933
Reversing the usual rule that CpG methylation blocks binding, methylation of the CRE and of CGTCA and CGCAA half-sites creates C/EBP-alpha sites, and demethylation removes C/EBP-alpha from methylated tissue-specific promoters.
Still open then: The genome-wide fraction of C/EBP sites that depend on methylation, and whether other paralogues share the preference, were not measured.
Still unresolved
- How much of C/EBP-alpha's genomic binding depends on methylated CpG is unclear: methylation of CGTCA and CGCAA half-sites creates binding at some tissue-specific promoters, but whether this is a minor class of sites or a general feature, and whether the other C/EBP paralogues share it, has not been established.
- Whether the CEBPG matrices describe CEBPG or its partners is unresolved in the databases: the ChIP-derived HOCOMOCO model and JASPAR MA1636.2 read as the C/EBP-ATF composite ATTGCATCAT, while the HT-SELEX and methyl-HT-SELEX models read as the C/EBP palindrome, consistent with a protein that works mainly as a heterodimer.
- How CHOP acts at DNA is not settled: CHOP-C/EBP heterodimers are excluded from classic C/EBP sites and select a distinct core in vitro, while genome-wide ChIP-seq places CHOP together with ATF4 at shared directly induced genes, and the relative weight of inhibition by sequestration and direct activation in cells is unknown.
C/EBP is one of the more trustworthy family calls in this atlas: the motif is cohesive across all four databases, its in vitro and ChIP-derived models agree, and the TTGC half-site is not shared with the other bZIP families. Do not read a member into it - the five paralogues are identical at this resolution, and the biological consequence differs between them and between full-length and LIP isoforms of the same protein. Two caveats are specific. Matches of the form ATGATGCAAT are C/EBP-ATF composites and are equally well explained by ATF4. And because methylated CpG creates C/EBP-alpha sites at half-CRE and half-C/EBP sequences, a scan run without methylation data will both miss real sites and over-call unmethylated ones.
References
- Landschulz WH, Johnson PF, McKnight SL. The leucine zipper: a hypothetical structure common to a new class of DNA binding proteins. Science 1988;240:1759-64. PMID 3289117
- Vinson CR, Sigler PB, McKnight SL. Scissors-grip model for DNA recognition by a family of leucine zipper proteins. Science 1989;246:911-6. PMID 2683088
- Tanaka T, Akira S, Yoshida K, et al. Targeted disruption of the NF-IL6 gene discloses its essential role in bacteria killing and tumor cytotoxicity by macrophages. Cell 1995;80:353-61. PMID 7530603
- Wang ND, Finegold MJ, Bradley A, et al. Impaired energy homeostasis in C/EBP alpha knockout mice. Science 1995;269:1108-12. PMID 7652557
- Osada S, Yamamoto H, Nishihara T, et al. DNA binding specificity of the CCAAT/enhancer-binding protein transcription factor family. J Biol Chem 1996;271:3891-6. PMID 8632009
- Ubeda M, Wang XZ, Zinszner H, et al. Stress-induced binding of the transcriptional factor CHOP to a novel DNA control element. Mol Cell Biol 1996;16:1479-89. PMID 8657121
- Yamanaka R, Barlow C, Lekstrom-Himes J, et al. Impaired granulopoiesis, myelodysplasia, and early lethality in CCAAT/enhancer binding protein epsilon-deficient mice. Proc Natl Acad Sci U S A 1997;94:13187-92. PMID 9371821
- Fawcett TW, Martindale JL, Guyton KZ, et al. Complexes containing activating transcription factor (ATF)/cAMP-responsive-element-binding protein (CREB) interact with the CCAAT/enhancer-binding protein (C/EBP)-ATF composite site to regulate Gadd153 expression during the stress response. Biochem J 1999;339 ( Pt 1):135-41. PMID 10085237
- Lekstrom-Himes JA, Dorman SE, Kopar P, et al. Neutrophil-specific granule deficiency results from a novel mutation with loss of function of the transcription factor CCAAT/enhancer binding protein epsilon. J Exp Med 1999;189:1847-52. PMID 10359588
- Pabst T, Mueller BU, Zhang P, et al. Dominant-negative mutations of CEBPA, encoding CCAAT/enhancer binding protein-alpha (C/EBPalpha), in acute myeloid leukemia. Nat Genet 2001;27:263-70. PMID 11242107
- Miller M, Shuman JD, Sebastian T, et al. Structural basis for DNA recognition by the basic region leucine zipper transcription factor CCAAT/enhancer-binding protein alpha. J Biol Chem 2003;278:15178-84. PMID 12578822
- Smith ML, Cavenagh JD, Lister TA, et al. Mutation of CEBPA in familial acute myeloid leukemia. N Engl J Med 2004;351:2403-7. PMID 15575056
- Rishi V, Bhattacharya P, Chatterjee R, et al. CpG methylation of half-CRE sequences creates C/EBPalpha binding sites that activate some tissue-specific genes. Proc Natl Acad Sci U S A 2010;107:20311-6. PMID 21059933
- Han J, Back SH, Hur J, et al. ER-stress-induced transcriptional regulation increases protein synthesis leading to cell death. Nat Cell Biol 2013;15:481-90. PMID 23624402
25 motifs in this family — JASPAR 7 · HOCOMOCO 11 · CIS-BP 5 · Vierstra 2
| motif | database | consensus | bits |
|---|---|---|---|
| CEBPE.H14CORE.1.SM.BSMB | HOCOMOCO | gTTgCGCAAT | 16.4 |
| CEBPB (MA0466.4) | JASPAR | ATTGCGCAAT | 16.1 |
| CEBPE (MA0837.3) | JASPAR | ATTGCGCAAT | 15.9 |
| CEBPG.H14CORE.2.P.BPB | HOCOMOCO | ATTGCATCAT | 15.9 |
| CEBPB.H14CORE.1.SM.BSMB | HOCOMOCO | TTgCGCAAc | 15 |
| CEBPG (MA0838.1) | JASPAR | ATTgCGCAAt | 14.7 |
| CEBPG.H14CORE.0.P.BPB | HOCOMOCO | ATTGCgtAAt | 14.2 |
| CEBPG.H14CORE.1.SM.BSMB | HOCOMOCO | ATTGCGtAA | 14.2 |
| CEBPG (MA1636.2) | JASPAR | ATGATGCAAT | 13.3 |
| M08076_2.00 | CIS-BP | aTTgTGCAAT | 13.3 |
| CEBPD.H14CORE.1.SM.BSMB | HOCOMOCO | aTTGCGtAA | 13.2 |
| M07830_2.00 | CIS-BP | aTTgTGCAAt | 12.9 |
| CEBPA.H14CORE.0.P.BPB | HOCOMOCO | gTTgcGCAAT | 12.4 |
| CEBPB.H14CORE.0.P.BPB | HOCOMOCO | gTTgcGCAAT | 12.4 |
| CEBPA (MA0102.5) | JASPAR | ATTGCACAAT | 12.1 |
| M04380_2.00 | CIS-BP | TTgCGCAAc | 12 |
| Ddit3::Cebpa (MA0019.2)dimer | JASPAR | aTGCAATccC | 11.5 |
| CEBPD (MA0836.3) | JASPAR | TTGCACAA | 11.5 |
| CEBPE.H14CORE.0.P.BPB | HOCOMOCO | TtgtGCAA | 10.8 |
| M04303_2.00 | CIS-BP | TTaCGCAA | 10.7 |
| CEBPD.H14CORE.0.P.BPB | HOCOMOCO | ATTGCacAAt | 10.4 |
| AC0628:CEBPB/CEBPE:bZIP | Vierstra | TTgcGCAA | 9.7 |
| DDIT3.H14CORE.0.P.BPB | HOCOMOCO | tgaTGCAAT | 9.5 |
| M04009_2.00 | CIS-BP | TTaCGCAA | 7.9 |
| AC0156:CEBPA:bZIP | Vierstra | TtacCAaA | 6.6 |
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.