{
 "key": "MEF2",
 "name": "MEF2",
 "class": "MADS",
 "aka": [
  "MEF2 site",
  "myocyte enhancer factor 2 site",
  "A/T-rich element",
  "MEF2 A/T box",
  "RSRF site",
  "CTA(A/T)4TAG",
  "MADS-box (MEF2 type)"
 ],
 "definition": "AT-rich CTA(A/T)4TAG. MEF2A/B/C/D.",
 "consensus": "CTA(A/T)4TAG",
 "dataConsensus": "CTATTTTTAG",
 "motifCount": 18,
 "bySource": {
  "JASPAR": 4,
  "HOCOMOCO": 7,
  "CIS-BP": 5,
  "Vierstra": 2
 },
 "symbols": [
  "MEF2A",
  "MEF2B",
  "MEF2C",
  "MEF2D"
 ],
 "representative": "MEF2A.H14CORE.0.P.B",
 "cohesion": 0.86,
 "short": "MEF2 sites are bound by the four MADS-box factors MEF2A, MEF2B, MEF2C and MEF2D, which act as obligate dimers. The element is an A/T-rich decamer, CTA(A/T)4TAG, and is easy to confuse with ordinary AT-rich sequence: the information sits almost entirely in the flanking CTA and TAG. MEF2 and SRF share the MADS fold but not the site - SRF requires the CC and GG of a CArG box where MEF2 requires TA and TA. The four paralogues have the same sequence preference and overlapping expression, so which one acts at a given element is decided by cell type and by chromatin rather than by DNA. MEF2 factors drive muscle differentiation, cardiac morphogenesis and activity-dependent programmes in neurons.",
 "long": "## The factors\n\nMEF2A, MEF2B, MEF2C and MEF2D each carry a MADS box - a 56-58 residue domain shared with SRF and with the plant and yeast MADS proteins [PMID:7744019] - immediately followed by a MEF2-specific extension of about 30 residues. The two regions together form the DNA-binding and dimerisation surface; MEF2 proteins bind only as homodimers or as heterodimers among themselves [PMID:1748287]. The crystal structure of the MEF2A core on DNA at 1.5 A resolution and an NMR structure of the same complex showed that MEF2 lacks the N-terminal extension SRF uses, and that the MEF domain C-terminal to the MADS box adopts a conformation substantially different from the equivalent region in SRF and yeast MCM1 - the structural basis for a different sequence preference and a different degree of DNA bending within one fold [PMID:10715212,10835359]. All four genes are broadly expressed with different emphases: MEF2A and MEF2D are near-ubiquitous, MEF2C is prominent in brain, cardiac and skeletal muscle and in B lymphocytes, MEF2B in germinal-centre B cells [PMID:23974956]. Activity is regulated post-translationally rather than by abundance: class IIa histone deacetylases (HDAC4, HDAC5, HDAC7, HDAC9) bind the MADS/MEF2 domain and repress it until calcium-CaMK signalling drives their nuclear export, and p38 MAPK, ERK5, sumoylation and the corepressor CABIN1 tune the same surface [PMID:17959722,23974956].\n\n## The motif\n\nThe consensus came with the cloning of the three human SRF-related proteins, later named MEF2, which recognize CTA(A/T)4TAG, distinct from the SRF site CC(A/T)6GG [PMID:1748287]. Site selection with anti-MEF2A antibodies and skeletal muscle, heart and brain extracts later returned the same core, 5'-CT(A/T)(A/T)AAATAG-3' [PMID:7559475]; it is now written CTA(A/T)4TAG or YTA(A/T)4TAR, a 10-bp near-palindrome. Each MADS monomer reads one half, contacting the bases through an antiparallel coiled coil of two alpha-helices laid in the widened major groove, with basic residues gripping the flanking phosphates [PMID:10715212,10835359]. Because the centre is a run of A and T, essentially all the sequence information lives in the outer CTA and TAG - which is why an A/T-rich stretch with the wrong flanks scores poorly, and why MEF2 sites are easy to over-call in AT-rich genomes. That same selection experiment found that MEF2 complexes from brain extracts imposed additional constraints outside the decamer, a tissue effect that has never been resolved into a distinct PWM [PMID:7559475]. The four paralogues are not separable: in these databases the JASPAR and CIS-BP models for MEF2A, MEF2B, MEF2C and MEF2D are nearly identical (family cohesion 0.86), and no in vitro study has reported a usable difference [PMID:23332764]. One real complication is visible in the data. Alongside the canonical CTATTTTTAG models sits a second flavour - CCAAATATGG, CCNAATTTGGAA and similar - which comes almost entirely from HT-SELEX and methyl-HT-SELEX experiments and which is, read literally, a CArG box: CC plus an A/T core plus GG. These entries have lower label agreement (0.6-0.7) and their nearest non-MEF2 neighbours are MADS-box motifs from SRF. Whether this is a genuine secondary MEF2 specificity or an artefact of selection is unsettled; either way, scanning with those models partly scans for SRF.\n\n## Where it acts\n\nMEF2 motifs are enhancer elements, not promoter elements, and are depleted from CpG islands - unsurprising for an A/T-rich site. In differentiating muscle they mark the same enhancers as the myogenic bHLH factors MYOD1 and MYF5, and MEF2-E-box pairs are the canonical muscle enhancer architecture [PMID:17959722]; MEF2 does not have a named composite element of its own but the MEF2-plus-E-box arrangement is close to one. In neurons MEF2 sites mark activity-regulated enhancers; in cardiomyocytes they co-occur with GATA4, NKX2-5, TBX20 and SRF sites. Paralogue occupancy is largely redundant where two MEF2 proteins are co-expressed, and which paralogue compensates at a given site depends on the local chromatin environment [PMID:31722213].\n\n## Biology and disease\n\nMEF2 factors are required for differentiation rather than for lineage specification. Mef2c-null mouse embryos fail cardiac morphogenesis, with a single ventricular chamber and absent right ventricle, and die around E9.5 [PMID:9162005]. In neurons, activity-dependent MEF2 activation restricts excitatory synapse number, so MEF2 acts in synapse elimination rather than formation [PMID:16484497]. In humans, MEF2C haploinsufficiency - by 5q14.3 microdeletion or point mutation - causes severe intellectual disability with stereotypic movements, epilepsy and absent speech [PMID:19592390]. MEF2B is somatically mutated in roughly 11% of diffuse large B-cell lymphomas and 12% of follicular lymphomas; the mutations increase MEF2B activity, either by disrupting its interaction with the corepressor CABIN1 or by escaping inhibitory phosphorylation and sumoylation, and deregulate the proto-oncogene BCL6 [PMID:23974956]. MEF2D fusions define a subtype of B-cell acute lymphoblastic leukemia [PMID:27824051], and a 21-bp MEF2A deletion was reported to cause an autosomal dominant form of coronary artery disease [PMID:14645853], a link not supported by later resequencing: in about 300 patients with premature disease no causative mutations were found, the same deletion appeared in unaffected controls, and it did not cosegregate with disease [PMID:15841183].\n\n## Reading a hit\n\nThe main risk here is the base composition of the genome, not the paralogue problem. An A/T-rich decamer is common in intergenic and intronic DNA, so scan with a model that actually constrains the CTA and TAG flanks and check the flanks in the match itself; a hit on a poly-A or poly-T stretch with soft flanks is almost certainly noise. A hit identifies MEF2, not MEF2A versus MEF2B, MEF2C or MEF2D - there is no reported sequence-level discrimination, and expression plus chromatin decide the occupant. Be explicit about which model you used, because the collections contain two incompatible flavours: canonical CTA(A/T)4TAG models, and CArG-like CCAAATATGG models from in vitro selection whose nearest neighbours are SRF motifs. If a putative MEF2 hit came from the second kind, consider SRF as an equally good explanation. The ChIP-derived MEF2A and MEF2C models here are grade B and reflect occupancy in specific cell types; because MEF2 sits beside E-boxes at muscle enhancers, E-box information can leak into ChIP-derived MEF2 models and vice versa.",
 "caveats": "Two distinct problems. First, base composition: the site is an A/T-rich decamer whose only real information is the CTA and TAG flanks, so AT-rich sequence throws false positives unless the flanks are genuinely constrained - inspect them in every hit. Second, the databases carry two incompatible model flavours for this family: canonical CTA(A/T)4TAG, and a CArG-like CCAAATATGG form from HT-SELEX whose nearest neighbours are SRF motifs. A hit from the second kind is as consistent with SRF as with MEF2. No model separates MEF2A, MEF2B, MEF2C and MEF2D, so report MEF2-family and let expression and chromatin decide the member.",
 "references": [
  {
   "cite": "Pollock R, Treisman R. Human SRF-related proteins: DNA-binding properties and potential regulatory targets. Genes Dev 1991;5:2327-41.",
   "pmid": "1748287",
   "url": "https://pubmed.ncbi.nlm.nih.gov/1748287/"
  },
  {
   "cite": "Shore P, Sharrocks AD. The MADS-box family of transcription factors. Eur J Biochem 1995;229:1-13.",
   "pmid": "7744019",
   "url": "https://pubmed.ncbi.nlm.nih.gov/7744019/"
  },
  {
   "cite": "Andrés V, Cervera M, Mahdavi V. Determination of the consensus binding site for MEF2 expressed in muscle and brain reveals tissue-specific sequence constraints. J Biol Chem 1995;270:23246-9.",
   "pmid": "7559475",
   "url": "https://pubmed.ncbi.nlm.nih.gov/7559475/"
  },
  {
   "cite": "Lin Q, Schwarz J, Bucana C, et al. Control of mouse cardiac morphogenesis and myogenesis by transcription factor MEF2C. Science 1997;276:1404-7.",
   "pmid": "9162005",
   "url": "https://pubmed.ncbi.nlm.nih.gov/9162005/"
  },
  {
   "cite": "Santelli E, Richmond TJ. Crystal structure of MEF2A core bound to DNA at 1.5 A resolution. J Mol Biol 2000;297:437-49.",
   "pmid": "10715212",
   "url": "https://pubmed.ncbi.nlm.nih.gov/10715212/"
  },
  {
   "cite": "Huang K, Louis JM, Donaldson L, et al. Solution structure of the MEF2A-DNA complex: structural basis for the modulation of DNA bending and specificity by MADS-box transcription factors. EMBO J 2000;19:2615-28.",
   "pmid": "10835359",
   "url": "https://pubmed.ncbi.nlm.nih.gov/10835359/"
  },
  {
   "cite": "Flavell SW, Cowan CW, Kim TK, et al. Activity-dependent regulation of MEF2 transcription factors suppresses excitatory synapse number. Science 2006;311:1008-12.",
   "pmid": "16484497",
   "url": "https://pubmed.ncbi.nlm.nih.gov/16484497/"
  },
  {
   "cite": "Potthoff MJ, Olson EN. MEF2: a central regulator of diverse developmental programs. Development 2007;134:4131-40.",
   "pmid": "17959722",
   "url": "https://pubmed.ncbi.nlm.nih.gov/17959722/"
  },
  {
   "cite": "Le Meur N, Holder-Espinasse M, Jaillard S, et al. MEF2C haploinsufficiency caused by either microdeletion of the 5q14.3 region or mutation is responsible for severe mental retardation with stereotypic movements, epilepsy and/or cerebral malformations. J Med Genet 2010;47:22-9.",
   "pmid": "19592390",
   "url": "https://pubmed.ncbi.nlm.nih.gov/19592390/"
  },
  {
   "cite": "Jolma A, Yan J, Whitington T, et al. DNA-binding specificities of human transcription factors. Cell 2013;152:327-39.",
   "pmid": "23332764",
   "url": "https://pubmed.ncbi.nlm.nih.gov/23332764/"
  },
  {
   "cite": "Ying CY, Dominguez-Sola D, Fabi M, et al. MEF2B mutations lead to deregulated expression of the oncogene BCL6 in diffuse large B cell lymphoma. Nat Immunol 2013;14:1084-92.",
   "pmid": "23974956",
   "url": "https://pubmed.ncbi.nlm.nih.gov/23974956/"
  },
  {
   "cite": "Majidi SP, Reddy NC, Moore MJ, et al. Chromatin Environment and Cellular Context Specify Compensatory Activity of Paralogous MEF2 Transcription Factors. Cell Rep 2019;29:2001-2015.e5.",
   "pmid": "31722213",
   "url": "https://pubmed.ncbi.nlm.nih.gov/31722213/"
  },
  {
   "cite": "Gu Z, Churchman M, Roberts K, et al. Genomic analyses identify recurrent MEF2D fusions in acute lymphoblastic leukaemia. Nat Commun 2016;7:13331.",
   "pmid": "27824051",
   "url": "https://pubmed.ncbi.nlm.nih.gov/27824051/"
  },
  {
   "cite": "Wang L, Fan C, Topol SE, et al. Mutation of MEF2A in an inherited disorder with features of coronary artery disease. Science 2003;302:1578-81.",
   "pmid": "14645853",
   "url": "https://pubmed.ncbi.nlm.nih.gov/14645853/"
  },
  {
   "cite": "Weng L, Kavaslar N, Ustaszewska A, et al. Lack of MEF2A mutations in coronary artery disease. J Clin Invest 2005;115:1016-20.",
   "pmid": "15841183",
   "url": "https://pubmed.ncbi.nlm.nih.gov/15841183/"
  }
 ],
 "timeline": [
  {
   "year": 1991,
   "finding": "Three human SRF-related MADS-box proteins (RSRFs, later MEF2) are cloned and shown to bind CTA(A/T)4TAG as dimers, a specificity distinct from SRF's CC(A/T)6GG that maps to the basic N-terminal part of the MADS domain; they dimerise with one another but not with SRF.",
   "methods": "cDNA cloning; DNA-binding assays with recombinant proteins; reporter transfection",
   "confidence": "medium",
   "pmids": [
    "1748287"
   ],
   "open": "Why broadly expressed proteins act mainly on muscle-specific and growth-factor-regulated genes."
  },
  {
   "year": 1995,
   "finding": "Site selection with anti-MEF2A antibodies and skeletal muscle, heart and brain extracts gives the same core, CT(A/T)(A/T)AAATAG, in all three tissues, but brain-derived complexes impose additional flanking constraints.",
   "methods": "Binding-site selection from degenerate oligonucleotides with tissue extracts",
   "confidence": "medium",
   "pmids": [
    "7559475"
   ],
   "open": "What causes the brain-specific constraint - a paralogue, an isoform or a partner protein."
  },
  {
   "year": 1997,
   "finding": "Mef2c-null mouse embryos fail cardiac looping and lack a right ventricle, showing that MEF2 is required for differentiation and morphogenesis of the heart.",
   "methods": "Mouse gene targeting",
   "confidence": "medium",
   "pmids": [
    "9162005"
   ]
  },
  {
   "year": 2000,
   "finding": "Crystal and NMR structures of the MEF2A core on CTA(A/T)4TAG show MADS-box contacts in both grooves and explain, through differences from SRF and MCM1, why MEF2 has a different sequence preference and bends DNA about 15 degrees rather than about 70; the MEF2-specific domain has a fold unlike SRF's.",
   "methods": "X-ray crystallography; NMR spectroscopy",
   "confidence": "high",
   "pmids": [
    "10715212",
    "10835359"
   ]
  },
  {
   "year": 2006,
   "finding": "Activity-dependent MEF2 activation in neurons restricts excitatory synapse number, extending MEF2 function from muscle differentiation to synapse elimination.",
   "methods": "Cultured neurons with RNAi knockdown and activity manipulation",
   "confidence": "medium",
   "pmids": [
    "16484497"
   ]
  },
  {
   "year": 2013,
   "finding": "MEF2B, mutated in about 11% of diffuse large B-cell lymphomas and 12% of follicular lymphomas, directly activates BCL6 in germinal-centre B cells; the mutations raise its activity by disrupting CABIN1 binding or inhibitory phosphorylation and sumoylation.",
   "methods": "ChIP, reporter and knockdown assays in germinal-centre B cells and lymphoma lines; analysis of tumour mutations",
   "confidence": "medium",
   "pmids": [
    "23974956"
   ]
  },
  {
   "year": 2019,
   "finding": "In cerebellar granule neurons, loss of MEF2D raises MEF2A occupancy at a subset of MEF2D sites concentrated in open chromatin, so paralogue compensation at a given site is set by chromatin rather than by the motif.",
   "methods": "Conditional single and double knockouts; genome-wide occupancy and transcriptome analysis",
   "confidence": "medium",
   "pmids": [
    "31722213"
   ]
  }
 ],
 "openQuestions": [
  "Whether the CArG-like CCAAATATGG models from HT-SELEX and methyl-HT-SELEX are a genuine secondary MEF2 specificity or a selection artefact is unsettled: they sit alongside canonical CTA(A/T)4TAG models for the same proteins, and their nearest neighbours outside the family are SRF motifs.",
  "The extra flanking constraints that brain-derived MEF2 complexes imposed in 1995 selection experiments have never been resolved into a distinct motif, and it is unknown whether they reflect a partner protein, a paralogue or a splice isoform."
 ],
 "members": [
  {
   "id": "MEF2D (MA0773.1)",
   "source": "JASPAR",
   "consensus": "CTAtAAATAGa",
   "bits": 18.5
  },
  {
   "id": "M03343_2.00",
   "source": "CIS-BP",
   "consensus": "CTAtAAATAGa",
   "bits": 18.5,
   "prov": {
    "inferred": false
   }
  },
  {
   "id": "MEF2B (MA0660.1)",
   "source": "JASPAR",
   "consensus": "gCTAtAAATAGc",
   "bits": 18.4
  },
  {
   "id": "M03344_2.00",
   "source": "CIS-BP",
   "consensus": "gCTAtAAATAGc",
   "bits": 18.4,
   "prov": {
    "inferred": false
   }
  },
  {
   "id": "MEF2B.H14CORE.1.SM.B",
   "source": "HOCOMOCO",
   "consensus": "GTTACCAtATttGG",
   "bits": 15.5,
   "prov": {
    "evidence": "SM",
    "grade": "B"
   }
  },
  {
   "id": "MEF2C (MA0497.2)",
   "source": "JASPAR",
   "consensus": "CtAAAAATAG",
   "bits": 13.4
  },
  {
   "id": "MEF2D.H14CORE.0.PS.A",
   "source": "HOCOMOCO",
   "consensus": "gCTATTTTTAG",
   "bits": 13.4,
   "prov": {
    "evidence": "PS",
    "grade": "A"
   }
  },
  {
   "id": "AC0295:MEF2A/MEF2C:MADS_box",
   "source": "Vierstra",
   "consensus": "CTAAAAATAG",
   "bits": 13.2
  },
  {
   "id": "MEF2A.H14CORE.0.P.B",
   "source": "HOCOMOCO",
   "consensus": "CTATTTTTAG",
   "bits": 13.1,
   "prov": {
    "evidence": "P",
    "grade": "B"
   }
  },
  {
   "id": "M07979_2.00",
   "source": "CIS-BP",
   "consensus": "CcAAAAATAG",
   "bits": 12.8,
   "prov": {
    "inferred": false
   }
  },
  {
   "id": "M07980_2.00",
   "source": "CIS-BP",
   "consensus": "CcAAAAATAG",
   "bits": 12.5,
   "prov": {
    "inferred": false
   }
  },
  {
   "id": "MEF2A (MA0052.5)",
   "source": "JASPAR",
   "consensus": "CTAAAAATAG",
   "bits": 11.8
  },
  {
   "id": "MEF2C.H14CORE.1.SM.B",
   "source": "HOCOMOCO",
   "consensus": "CCvnATTTGGaA",
   "bits": 11.3,
   "prov": {
    "evidence": "SM",
    "grade": "B"
   }
  },
  {
   "id": "MEF2D.H14CORE.1.M.C",
   "source": "HOCOMOCO",
   "consensus": "CCnaATTtGGaA",
   "bits": 10.5,
   "prov": {
    "evidence": "M",
    "grade": "C"
   }
  },
  {
   "id": "MEF2C.H14CORE.0.P.B",
   "source": "HOCOMOCO",
   "consensus": "CTATTTttGG",
   "bits": 10.1,
   "prov": {
    "evidence": "P",
    "grade": "B"
   }
  },
  {
   "id": "AC0297:MEF2B/MEF2C:MADS_box",
   "source": "Vierstra",
   "consensus": "CCaaATaTGG",
   "bits": 9.9
  },
  {
   "id": "M05557_2.00",
   "source": "CIS-BP",
   "consensus": "CcaaATwtGG",
   "bits": 8.8,
   "prov": {
    "inferred": false
   }
  },
  {
   "id": "MEF2B.H14CORE.0.PS.A",
   "source": "HOCOMOCO",
   "consensus": "CTaTTTttGg",
   "bits": 7.8,
   "prov": {
    "evidence": "PS",
    "grade": "A"
   }
  }
 ],
 "url": "https://motif.zhoulab.io/families/mef2/",
 "generated": "2026-10-21",
 "provenance": "Entry text drafted by a language model reading the primary literature; every cited PMID verified against PubMed. Computed fields derive from JASPAR 2024, HOCOMOCO v14, CIS-BP and the Vierstra archetypes. See https://motif.zhoulab.io/families/methods/"
}