A 2021 systematic survey by Chaudhri and Singh used the computational pipeline CEseek to discover more than 20,000 composite transcriptional elements (CEs) — stereospecific pairings of two transcription factor binding motifs — in open chromatin of diverse immune cells, and classified them as 'digital' (stringently spaced) or 'fuzzy' (variable spacing) configurations. The authors' discussion raises five unresolved issues that jointly define what remains to be done to establish CEs as a second layer of the cis-regulatory code: (1) most discovered CEs lack molecular and functional validation; (2) the survey covered only immune and hematopoietic cells, so the CE universe in other mammalian tissues is uncharted; (3) the molecular basis of cooperative DNA binding at fuzzy CEs — for which a flexible-scaffold mechanism involving coactivators such as p300/CBP was proposed but never tested — is unknown; (4) whether the bottom-up binary-motif approach extends to higher-order (ternary and beyond) regulatory syntax is undetermined; and (5) the anti-cooperative, mutually exclusive occupancy documented at many digital CEs with partially overlapping motifs — which the survey found is not necessarily a consequence of steric hindrance — lacks a mechanistic explanation. As of August 2026 each of these issues remains open or only partially advanced: the binary CE universe has been greatly expanded in vitro and functional/deep-learning frameworks have grown (Xie et al. 2025; Fan et al. 2025; Hu et al. 2025), but no direct test of the scaffold hypothesis for fuzzy CEs, no structural mechanism for anti-cooperative exclusivity, and no systematic ternary grammar exist.

Background

In mammals, the regulatory DNA sequences that control when and where genes are active (the 'cis-regulome', i.e., the collection of promoters and enhancers) are built from short recognition sequences, called transcription factor (TF) binding motifs, that are read by DNA-binding proteins called transcription factors. A single motif is short and degenerate, so it cannot by itself confer the specificity observed in gene regulation. One way specificity increases is through composite elements (CEs): two simple motifs juxtaposed at a defined spacing and orientation, so that the two TFs bind the DNA cooperatively (each TF's binding stabilizes the other's) or anti-cooperatively (binding of one excludes the other), implementing AND-like or OR-like — and, for exclusion, XOR-like — logic at the DNA level. Classic examples include the EICE (Ets–Irf composite element) and the NFAT–AP-1 composite sites of immune-receptor genes. Before 2021, CEs were known mostly from such one-by-one biochemical characterizations. High-throughput in-vitro methods existed — notably CAP-SELEX (consecutive-affinity-purification systematic evolution of ligands by exponential enrichment; Jolma et al., Nature 2015), which selects DNA ligands that bind two TFs together and so reveals cooperative motif pairs, and ChIP-seq (immunoprecipitation of TF-bound DNA followed by sequencing), which maps where TFs bind in the genome — but no systematic, genome-wide, in-vivo survey of CE diversity existed. Chaudhri and Singh (bioRxiv 2021.11.26.470154) filled this gap with CEseek, a pipeline that scanned open chromatin regions (regions of accessible, nucleosome-free DNA identified by DNase/ATAC-type assays) catalogued by the ImmGen consortium across many immune cell types and states, searching for motif pairs whose spacing and orientation were constrained beyond chance. They discovered over 20,000 CEs ('ImmCEs'), validated subsets against CAP-SELEX data, ChIP-seq co-occupancy, and STARR-seq (a massively parallel reporter assay, MPRA, in which thousands of candidate regulatory sequences are barcoded and their ability to drive transcription is quantified in bulk), and observed a striking bimodality: 'digital' CEs, with stringent spacing/orientation constraints, often involving structurally diverse TF pairs and sometimes partially overlapping motifs with anti-cooperative binding; and 'fuzzy' CEs, in which the spacing between the two sub-motifs varies widely (up to about 10 bp) yet binding is cooperative. Fuzzy CEs are selectively enriched at p300-associated enhancers — p300/CBP being large, flexible histone acetyltransferase coactivator proteins that contact many TFs — which motivated the authors' untested proposal that a flexible protein scaffold (p300/CBP or the Mediator complex) bridges the two DNA-bound TFs across variable distances. The unresolved points dispersed through the discussion — together with the survey's results on anti-cooperative occupancy — constitute the five issues of this problem, and a structural-biology literature on cooperative TF pairs (reviewed by Morgunova and Taipale, 2017) provides the conceptual backdrop for the mechanistic questions among them.

Problem Statement

The discussion and results of Chaudhri and Singh's (2021) CEseek survey raise the five interlinked unresolved issues below; settling them would establish composite elements as a general second layer of the mammalian cis-regulatory code: (1) Functional validation — a substantial fraction of the ~20,000 discovered composite elements have not been molecularly or functionally tested; comprehensive validation could be achieved by high-throughput approaches such as protein-binding microarrays (arrays on which thousands of DNA sequences are probed for TF binding), massively parallel reporter assays, and extensive profiling of TF binding regions across diverse cell contexts. (2) Tissue generality — the analysis covered only immune and hematopoietic cells; the CE universe is expected to expand considerably when the same discovery approach is applied to additional mammalian cell types and tissues. (3) Mechanism of fuzzy-CE cooperativity — the molecular basis of cooperative DNA binding at fuzzy CEs, which exhibit highly variable spacing between their sub-motifs, remains unclear; the proposed mechanism involving flexible scaffolds such as p300/CBP (or Mediator) has been suggested but never tested directly. (4) Higher-order syntax — whether the bottom-up, binary-motif approach extends to ternary and higher-order levels of regulatory syntax remains to be determined; in the authors' view, both unbiased deep machine learning and hierarchical analyses based on the structural properties of DNA-protein interactions will be necessary to thoroughly delineate the regulatory code. (5) Mechanism of anti-cooperativity — the anti-cooperative mode of TF binding documented at many digital CEs, often with partially overlapping motifs, remains without a mechanistic explanation; detailed mechanistic study is needed to explain how TF pairs achieve mutually exclusive occupancy that the survey found is not necessarily a consequence of direct steric hindrance. All five issues are mandatory and must be addressed jointly — resolving any single one alone does not settle the problem. The deliverable is a proposed resolution of each issue rather than an executed determination: for each issue, state the proposed mechanism, strategy, or determination, the discriminating experiment or analysis that would test it, and the predicted outcomes that would confirm or refute it.

References

  1. Chaudhri VK, Singh H. 'Diverse digital and fuzzy composite transcriptional elements are prevalent features of mammalian cis-regulomes.' bioRxiv preprint (2021). DOI: 10.1101/2021.11.26.470154. https://doi.org/10.1101/2021.11.26.470154
  2. Jolma A, Yin Y, Nitta KR, Dave K, Popov A, Taipale M, Enge M, Kivioja T, Morgunova E, Taipale J. 'DNA-dependent formation of transcription factor pairs alters their binding specificity.' Nature 527, 384–388 (2015). DOI: 10.1038/nature15518. https://doi.org/10.1038/nature15518
  3. Xie Z, Sokolov I, Osmala M, Yue X, Bower G, Pett JP, Chen Y, Wang K, Cavga AD, Popov A, Teichmann SA, Morgunova E, Kvon EZ, Yin Y, Taipale J. 'DNA-guided transcription factor interactions extend human gene regulatory code.' Nature (2025). DOI: 10.1038/s41586-025-08844-z. https://doi.org/10.1038/s41586-025-08844-z
  4. Fan J, Chaudhri VK, Bisht D, Pease NA, Hall DR, Gerges P, Yi VF, Kales S, Ho C, Das J, Ray JP, Tewhey R, Meyer CA, Sahni N, Singh H. 'Expanding the DNA Motif Lexicon of the Transcriptional Regulatory Code.' bioRxiv preprint (2025). DOI: 10.1101/2025.07.09.662874. https://doi.org/10.1101/2025.07.09.662874
  5. Hu Y, Horlbeck MA, Zhang R, Ma S, Shrestha R, Kartha VK, Duarte FM, Hock C, Savage RE, Labade A, Kletzien H, Meliki A, Castillo A, Durand NC, Mattei E, Anderson LJ, Tay T, Earl AS, Shoresh N, Epstein CB, Wagers AJ, Buenrostro JD. 'Multiscale footprints reveal the organization of cis-regulatory elements.' Nature (2025). DOI: 10.1038/s41586-024-08443-4. https://doi.org/10.1038/s41586-024-08443-4
  6. Rao S, Ahmad K, Ramachandran S. 'Cooperative binding between distant transcription factors is a hallmark of active enhancers.' Molecular Cell 81, 1651–1665.e4 (2021). DOI: 10.1016/j.molcel.2021.02.014. https://doi.org/10.1016/j.molcel.2021.02.014
  7. Wang X, Xie C, Shen K, Li D, Xie XS. 'Quantification and potential functional relevance of binding cooperativity of adjacent transcription factors on DNA.' Proceedings of the National Academy of Sciences USA 122 (2025). DOI: 10.1073/pnas.2422555122. https://doi.org/10.1073/pnas.2422555122
  8. Savage RE, McRoberts Amador CD, Hock CT, Zhang R, Kuo HC, Gao AR, Horlbeck MA, Gersbach CA, Buenrostro JD. 'Transcription factor collaboration enables precise T cell state engineering.' bioRxiv preprint (2026). DOI: 10.64898/2026.04.20.718569. https://doi.org/10.64898/2026.04.20.718569
  9. Frömel R, Rühle J, Bernal Martinez A, Szu-Tu C, Pacheco Pastor F, Martinez-Corral R, Velten L. 'Design principles of cell-state-specific enhancers in hematopoiesis.' Cell (2025). DOI: 10.1016/j.cell.2025.04.017. https://doi.org/10.1016/j.cell.2025.04.017
  10. Batool F, Shireen H, Malik MF, Abrar M, Abbasi AA. 'The combinatorial binding syntax of transcription factors in forebrain-specific enhancers.' Biology Open (2025). DOI: 10.1242/bio.061751. https://doi.org/10.1242/bio.061751
  11. Morgunova E, Taipale J. 'Structural perspective of cooperative transcription factor binding.' Current Opinion in Structural Biology 47, 1–8 (2017). DOI: 10.1016/j.sbi.2017.03.006. https://doi.org/10.1016/j.sbi.2017.03.006
  12. Deng Q, Lakra P, Gou P, Yang H, Meydan C, Teater M, Chin C, Zhang W, Dinh T, Hussein U, Li X, Rojas E, Liu W, Reville PK, Kizhakeyil A, Barisic D, Parsons S, Wilson A, Henderson J, Scull B, Gurumurthy C, Vega F, Chadburn A, Cuglievan B, El-Mallawany NK, Allen C, Mason C, Melnick A, Green MR. 'SMARCA4 is a haploinsufficient B cell lymphoma tumor suppressor that fine-tunes centrocyte cell fate decisions.' Cancer Cell 42, 605–622.e11 (2024). DOI: 10.1016/j.ccell.2024.02.011. https://doi.org/10.1016/j.ccell.2024.02.011