Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Transcriptional regulators form diverse groups with context-dependent regulatory functions


One of the most important questions in biology is how transcription factors (TFs) and cofactors control enhancer function and thus gene expression. Enhancer activation usually requires combinations of several TFs1, indicating that TFs function synergistically and combinatorially2,3. However, while TF binding has been extensively studied, little is known about how combinations of TFs and cofactors control enhancer function once they are bound. It is typically unclear which TFs participate in combinatorial enhancer activation, whether different TFs form functionally distinct groups, or if certain TFs might substitute for each other in defined enhancer contexts. Here we assess the potential regulatory contributions of TFs and cofactors to combinatorial enhancer control with enhancer complementation assays. We recruited GAL4-DNA-binding-domain fusions of 812 Drosophila TFs and cofactors to 24 enhancer contexts and measured enhancer activities by 82,752 luciferase assays in S2 cells. Most factors were functional in at least one context, yet their contributions differed between contexts and varied from repression to activation (up to 289-fold) for individual factors. Based on functional similarities across contexts, we define 15 groups of TFs that differ in developmental functions and protein sequence features. Similar TFs can substitute for each other, enabling enhancer re-engineering by exchanging TF motifs, and TF–cofactor pairs cooperate during enhancer control and interact physically. Overall, we show that activators and repressors can have diverse regulatory functions that typically depend on the enhancer context. The systematic functional characterization of TFs and cofactors should further our understanding of combinatorial enhancer control and gene regulation.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Enhancer complementation assays for 474 TFs.
Figure 2: TFs have diverse regulatory functions.
Figure 3: Transcriptional cofactors can be sufficient for activation and repression and have context-dependent regulatory functions.
Figure 4: TF–cofactor assignments through functional similarities.

Accession codes

Primary accessions

Sequence Read Archive

Data deposits

All data are available at The next-generation sequencing data have been deposited at the NCBI Sequence Read Archive (SRA) under the accession SRS806429. The cofactor Gateway entry clones and other plasmids are available from Addgene (


  1. Arnone, M. I. & Davidson, E. H. The hardwiring of development: organization and function of genomic regulatory systems. Development 124, 1851–1864 (1997)

    CAS  PubMed  Google Scholar 

  2. Han, K., Levine, M. S. & Manley, J. L. Synergistic activation and repression of transcription by Drosophila homeobox proteins. Cell 56, 573–583 (1989)

    Article  CAS  Google Scholar 

  3. Yáñez-Cuna, J. O. et al. Dissection of thousands of cell type-specific enhancers identifies dinucleotide repeat motifs as general enhancer features. Genome Res. 24, 1147–1156 (2014)

    Article  Google Scholar 

  4. Ptashne, M. & Gann, A. Transcriptional activation by recruitment. Nature 386, 569–577 (1997)

    Article  CAS  ADS  Google Scholar 

  5. Cheng, J. X., Gandolfi, M. & Ptashne, M. Activation of the Gal1 gene of yeast by pairs of ‘non-classical’ activators. Curr. Biol. 14, 1675–1679 (2004)

    Article  CAS  Google Scholar 

  6. Keung, A. J., Bashor, C. J., Kiriakov, S., Collins, J. J. & Khalil, A. S. Using targeted chromatin regulators to engineer combinatorial and spatial transcriptional regulation. Cell 158, 110–120 (2014)

    Article  CAS  Google Scholar 

  7. Yuasa, Y. et al. Drosophila homeodomain protein REPO controls glial differentiation by cooperating with ETS and BTB transcription factors. Development 130, 2419–2428 (2003)

    Article  CAS  Google Scholar 

  8. Papadopoulos, D. K. et al. Functional synthetic Antennapedia genes and the dual roles of YPWM motif and linker size in transcriptional activation and repression. Proc. Natl Acad. Sci. USA 108, 11959–11964 (2011)

    Article  CAS  ADS  Google Scholar 

  9. Arnosti, D. N., Barolo, S., Levine, M. & Small, S. The eve stripe 2 enhancer employs multiple modes of transcriptional synergy. Development 122, 205–214 (1996)

    CAS  PubMed  Google Scholar 

  10. Brodu, V., Mugat, B., Fichelson, P., Lepesant, J. A. & Antoniewski, C. A UAS site substitution approach to the in vivo dissection of promoters: interplay between the GATAb activator and the AEF-1 repressor at a Drosophila ecdysone response unit. Development 128, 2593–2602 (2001)

    CAS  PubMed  Google Scholar 

  11. Hens, K. et al. Automated protein-DNA interaction screening of Drosophila regulatory elements. Nature Methods 8, 1065–1070 (2011)

    Article  CAS  Google Scholar 

  12. Shlyueva, D. et al. Hormone-responsive enhancer-activity maps reveal predictive motifs, indirect repression, and targeting of closed chromatin. Mol. Cell 54, 180–192 (2014)

    Article  CAS  Google Scholar 

  13. Zabidi, M. A. et al. Enhancer–core-promoter specificity separates developmental and housekeeping gene regulation. Nature 518, 556–559 (2015)

    Article  CAS  ADS  Google Scholar 

  14. Gerber, H. P. et al. Transcriptional activation modulated by homopolymeric glutamine and proline stretches. Science 263, 808–811 (1994)

    Article  CAS  ADS  Google Scholar 

  15. Galant, R. & Carroll, S. B. Evolution of a transcriptional repression domain in an insect Hox protein. Nature 415, 910–913 (2002)

    Article  CAS  ADS  Google Scholar 

  16. Lam, K. C. et al. The NSL complex regulates housekeeping genes in Drosophila. PLoS Genet. 8, e1002736 (2012)

    CAS  PubMed  Google Scholar 

  17. Wright, K. J., Marr, M. T. & Tjian, R. TAF4 nucleates a core subcomplex of TFIID and mediates activated transcription from a TATA-less promoter. Proc. Natl Acad. Sci. USA 103, 12347–12352 (2006)

    Article  CAS  ADS  Google Scholar 

  18. Martens, J. A. & Winston, F. Recent advances in understanding chromatin remodeling by Swi/Snf complexes. Curr. Opin. Genet. Dev. 13, 136–142 (2003)

    Article  CAS  Google Scholar 

  19. Wang, Y. et al. IXL, a new subunit of the mammalian Mediator complex, functions as a transcriptional suppressor. Biochem. Biophys. Res. Commun. 325, 1330–1338 (2004)

    Article  CAS  Google Scholar 

  20. Mittler, G. et al. A novel docking site on Mediator is critical for activation by VP16 in mammalian cells. EMBO J. 22, 6494–6504 (2003)

    Article  CAS  Google Scholar 

  21. Bryant, G. O. & Ptashne, M. Independent recruitment in vivo by Gal4 of two complexes required for transcription. Mol. Cell 11, 1301–1309 (2003)

    Article  CAS  Google Scholar 

  22. Ihry, R. J. & Bashirullah, A. Genetic control of specificity to steroid-triggered responses in Drosophila. Genetics 196, 767–780 (2014)

    Article  CAS  Google Scholar 

  23. Kim, T. W. et al. MED16 and MED23 of Mediator are coactivators of lipopolysaccharide- and heat-shock-induced transcriptional activators. Proc. Natl Acad. Sci. USA 101, 12153–12158 (2004)

    Article  CAS  ADS  Google Scholar 

  24. Gan, M., Moebus, S., Eggert, H. & Saumweber, H. The Chriz-Z4 complex recruits JIL-1 to polytene chromosomes, a requirement for interband-specific phosphorylation of H3S10. J. Biosci. 36, 425–438 (2011)

    Article  CAS  Google Scholar 

  25. Guruharsha, K. G. et al. A protein complex network of Drosophila melanogaster. Cell 147, 690–703 (2011)

    Article  CAS  Google Scholar 

  26. Nibu, Y., Zhang, H. & Levine, M. Interaction of short-range repressors with Drosophila CtBP in the embryo. Science 280, 101–104 (1998)

    Article  CAS  ADS  Google Scholar 

  27. Giot, L. et al. A protein interaction map of Drosophila melanogaster. Science 302, 1727–1736 (2003)

    Article  CAS  ADS  Google Scholar 

  28. Rhee, D. Y. et al. Transcription factor networks in Drosophila melanogaster. Cell Rep. 8, 2031–2043 (2014)

    Article  CAS  Google Scholar 

  29. Murali, T. et al. DroID 2011: a comprehensive, integrated resource for protein, transcription factor, RNA and gene interactions for Drosophila. Nucleic Acids Res. 39, D736–D743 (2011)

    Article  CAS  Google Scholar 

  30. Amelio, A. L. et al. A coactivator trap identifies NONO (p54nrb) as a component of the cAMP-signaling pathway. Proc. Natl Acad. Sci. USA 104, 20314–20319 (2007)

    Article  CAS  ADS  Google Scholar 

  31. St Pierre, S. E., Ponting, L., Stefancsik, R. & McQuilton, P. FlyBase Consortium. FlyBase 102–advanced approaches to interrogating FlyBase. Nucleic Acids Res. 42, D780–D788 (2014)

    Article  CAS  Google Scholar 

  32. Finn, R. D. et al. Pfam: the protein families database. Nucleic Acids Res. 42, D222–D230 (2014)

    Article  CAS  Google Scholar 

  33. Arnold, C. D. et al. Genome-wide quantitative enhancer activity maps identified by STARR-seq. Science 339, 1074–1077 (2013)

    Article  CAS  ADS  Google Scholar 

  34. Massouras, A., Decouttere, F., Hens, K. & Deplancke, B. WebPrInSeS: automated full-length clone sequence identification and verification using high-throughput sequencing data. Nucleic Acids Res. 38, W378–W384 (2010)

    Article  CAS  Google Scholar 

  35. Pfeiffer, B. D. et al. Tools for neuroanatomy and neurogenetics in Drosophila. Proc. Natl Acad. Sci. USA 105, 9715–9720 (2008)

    Article  CAS  ADS  Google Scholar 

  36. Schneider, I. The culture of cells from insects and ticks. I. Cultivation of dipteran cells in vitro. Curr. Top. Microbiol. Immunol. 55, 1–12 (1971)

    CAS  PubMed  Google Scholar 

  37. Ui, K. et al. Newly established cell lines from Drosophila larval CNS express neural specific characteristics. In Vitro Cell. Dev. Biol. Anim. 30, 209–216 (1994)

    Article  Google Scholar 

  38. Echalier, G. & Ohanessian, A. Isolation, in tissue culture, of Drosophila melangaster cell lines. C. R. Acad. Sci. Hebd. Seances Acad. Sci. D 268, 1771–1773 (1969) [transl.]

    CAS  PubMed  Google Scholar 

  39. Saito, K. et al. A regulatory circuit for piwi by the large Maf gene traffic jam in Drosophila. Nature 461, 1296–1299 (2009)

    Article  CAS  ADS  Google Scholar 

  40. Hampf, M. & Gossen, M. A protocol for combined Photinus and Renilla luciferase quantification compatible with protein assays. Anal. Biochem. 356, 94–99 (2006)

    Article  CAS  Google Scholar 

  41. R Development Core Team. R: A language and environment for statistical computing. (R Foundation for Statistical Computing, 2011)

  42. Bailey, T. L. et al. MEME SUITE: tools for motif discovery and searching. Nucleic Acids Res. 37, W202–W208 (2009)

    Article  CAS  Google Scholar 

  43. Eddy, S. R. Profile hidden Markov models. Bioinformatics 14, 755–763 (1998)

    Article  CAS  Google Scholar 

  44. Dinkel, H. et al. ELM–the database of eukaryotic linear motifs. Nucleic Acids Res. 40, D242–D251 (2012)

    Article  CAS  Google Scholar 

  45. Ashburner, M. et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nature Genet. 25, 25–29 (2000)

    Article  CAS  Google Scholar 

  46. Tomancak, P. et al. Systematic determination of patterns of gene expression during Drosophila embryogenesis. Genome Biol. 3, (2002)

  47. Zhu, L. J. et al. FlyFactorSurvey: a database of Drosophila transcription factor binding specificities determined using the bacterial one-hybrid system. Nucleic Acids Res. 39, D111–D117 (2011)

    Article  CAS  Google Scholar 

  48. Stark, A. et al. Discovery of functional elements in 12 Drosophila genomes using evolutionary signatures. Nature 450, 219–232 (2007)

    Article  CAS  ADS  Google Scholar 

  49. Yu, S. X. & Shi, J. Multiclass spectral clustering. In ICCV, 313–319 (IEEE, 2003)

  50. Pedregosa, F. et al. Scikit-learn: machine learning in Python. J. Machine Learn. Res. 12 2825–2830 (2011)

    MathSciNet  MATH  Google Scholar 

  51. von Luxburg, U. Clustering stability: an overview. Foundations and Trends in Machine Learning 2, 235–274 (2010)

    MATH  Google Scholar 

  52. van der Maaten, L. & Hinton, G. Visualizing data using t-SNE. J. Machine Learn. Res. 9, 2579–2605 (2008)

    MATH  Google Scholar 

  53. The modENCODE Consortium et al. Identification of functional elements and regulatory circuits by Drosophila modENCODE. Science 330, 1787–1797 (2010)

  54. Pfeiffer, B. D. et al. Refinement of tools for targeted gene expression in Drosophila. Genetics 186, 735–755 (2010)

    Article  CAS  Google Scholar 

Download references


We are grateful to K. Hens and B. Deplancke for sharing the TF entry clones, J. O. Yáñez-Cuna for help designing the enhancer contexts, and O. Bell, J. Brennecke, L. Cochella and S. Westermann for comments on the manuscript. We thank IMP/IMBA services, especially H. Scheuch, R. Heinen and Z. Dzupinkova, for technical support, and A. Posekany and A. Aszódi for advice on multiple-testing correction. Deep sequencing was performed at the CSF Next-Generation Sequencing Unit ( The Stark group is supported by a European Research Council (ERC) Starting Grant (no. 242922) awarded to A.S., Boehringer Ingelheim GmbH, and the Austrian Research Promotion Agency (FFG).

Author information

Authors and Affiliations



G.S. and A.S. conceived the project. G.S. and O.F. cloned the cofactors and the GAL4–DBD fusions. G.S. and F.R. performed the luciferase assays in Drosophila cells and S.W. in HeLa cells. T.K. and G.S. conducted the bioinformatics analyses. G.S., T.K. and A.S. wrote the manuscript.

Corresponding author

Correspondence to Alexander Stark.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Extended data figures and tables

Extended Data Figure 1 24 regulatory contexts.

Tested contexts included 19 motif-mutant enhancer contexts which were designed by replacing 43 occurrences of 15 different motif types in 11 previously characterized enhancers with broad (‘Ubi-1’ to ‘Ubi-3’) or cell type-specific (‘S2-1’ to ‘S2-3’: S2 cell-specific; ‘OSC’: ovarian somatic cell (OSC)-specific) activities (all from ref. 3) or hormone-inducible enhancers (from ref. 12.). We also designed five synthetic contexts consisting of UAS sites with or without Trl sites and a developmental core promoter (dCP; Drosophila synthetic core promoter (DSCP) derived from the transcription factor gene Eve54) or with a housekeeping core promoter (hkCP; derived from the ribosomal gene RpS1213). Shown are schemes of the luciferase reporter constructs used for the targeted recruitment of GAL4-DBD–TF/cofactor fusion proteins to UAS sites (the luciferase gene is not drawn to scale). Motif names denote the motifs (as named by refs 3, 12) that have been replaced by UAS sites (blue boxes) to create the enhancer context. Note that TF-to-motif assignments are not unique and typically several TFs can bind each of the motifs (see Extended Data Table 1).

Extended Data Figure 2 TF clustering is robust and reproducible.

a, Cluster assignment is robust during bootstrapping (474 rounds of removing 10 randomly selected TFs). The cluster label stability denotes the fraction (out of 474 trials) a given TF was assigned to the same cluster as in the original clustering (node layout shown is identical to Fig. 2a). The vast majority of TFs were assigned to the same group in ≥90% of the cases (histogram). b, Cluster assignment for individual TFs is reproducible for biological replicates. We repeated the clustering six times, each time using only two out of the four biological replicates (six corresponds to all possibilities to choose two out of the four replicates). The cluster stability denotes the number of times (out of six) a given TF is assigned to the same cluster as in the original clustering. The majority of TFs were assigned to the same cluster, independent of which pair of biological replicates was used to generate the clustering (histogram).

Extended Data Figure 3 Cluster activity profiles.

Normalized luciferase values for all TFs assigned to each of the 15 clusters across all 24 contexts. Shown are median and quartiles as boxes, and the tenth and ninetieth percentiles as whiskers for each of the 24 contexts. Boxes are coloured according to the median activity in each context (see colour legend).

Extended Data Figure 4 Complete feature enrichment analysis of 15 groups of TFs.

Features analysed include eukaryotic linear motifs44 (ELM), homopolymeric amino acid repeat motifs discovered using MEME42, protein domains as annotated by Pfam32, Gene Ontology45 (GO), and gene expression patterns as annotated by IMAGO46. Red and blue shadings denote enrichments and depletions (log2-transformed) with an empirical FDR of at most 10% per feature type and cluster (others are white).

Extended Data Figure 5 TFs behave consistently across Drosophila cell types.

We tested 171 of the 474 TFs (36.1%) in 6 of the 24 contexts in Kc167, BG3 and OSC cell lines, which are derived from embryos, larvae and adult, respectively. Shown are normalized luciferase values and the Pearson correlation coefficients (PCC; P < 1 × 10−3 for all comparisons). We tested synthetic contexts containing an array of UAS sites upstream of a developmental and a housekeeping core promoter13 (4×UAS dCP and hkCP), three contexts derived from cell-type-specific enhancers3 (OSC dCP, S2-1 Pal dCP, S2-1 CGCG dCP), and one context derived from a broadly active enhancer3 (Ubi-1 Trl dCP). The latter showed the highest similarities (PCCs of 0.94, 0.94 and 0.92 for Kc167, BG3 and OSC cells, respectively) while the lowest PCCs for the non-embryonic BG3 and OSC cells (0.72 for BG3 and 0.55 for OSC) were obtained for S2-1 Pal dCP, derived from an enhancer active only in the embryonic S2 and Kc167 cells, presumably because the corresponding wild-type enhancer sequence is inactive in larval and adult cells3,12 such that combinatorial effects between the tethered TF and other enhancer-bound TFs may be less effective or lack entirely. Enhancer complementation presumes (and the results throughout this study confirm this presumption) that the regulatory functions of the tethered TFs are revealed (or altered) by other enhancer-bound factors; that is, factors that are bound to the enhancer in S2 cells (in which the corresponding enhancer is active3) but not in the other cell types (in which the enhancer is not active). This emphasizes the value of enhancer complementation for the study of regulatory activities and the importance of contexts derived from active enhancers. Error bars denote standard deviation (n = 4, biological replicates).

Extended Data Figure 6 Drosophila TFs and cofactors retain their activating functions in human HeLa cells.

We expressed GAL4-DBD fusion proteins for 107 of the 812 Drosophila factors (90 TFs and 17 cofactors) under the control of a constitutively active CMV promoter in human HeLa cells (see Methods). Shown are normalized luciferase values for the tested proteins recruited to the synthetic 4×UAS-dCP context. The values are remarkably similar quantitatively, with an overall Pearson correlation coefficient (PCC) of 0.74 (P < 1 × 10−3). The activation domain of the human TF P65 was used as a positive control. Error bars denote standard deviation (n = 4, biological replicates).

Extended Data Figure 7 Regulatory activities of selected cofactor complexes or protein domain families.

Heat maps of normalized luciferase values for sets of proteins annotated as being part of the same complex by Gene Ontology45 (GO) or containing a chromodomain or SIR2 domain as annotated by Pfam32.

Extended Data Figure 8 Regulatory activities of uncharacterized TFs and cofactors.

Heat maps of normalized luciferase values for all ‘CG genes’ among the tested TFs and cofactors, which activate or repress in at least one context (≥1.5-fold compared to GFP; P < 0.05 FDR-corrected for 24 × 474 and 24 × 338 tests for TFs and cofactors, respectively).

Extended Data Figure 9 Consistent effects of varying the amounts of plasmid DNA for TF and cofactor expression.

The effects of using 1 ng, 2 ng, 3 ng, 4 ng and 5 ng of GAL4-DBD–TF/cofactor fusion protein expressing plasmids on luciferase assays in S2 cells suggest that reporter activity is robust to variation in TF levels. Shown are normalized luciferase values of GAL4-DBD N-terminally fused to six activating and six repressing TFs and cofactors of different strengths targeted to the synthetic 4×UAS-dCP context. The amount of plasmid expressing the GAL4-DBD–TF/cofactor fusion proteins was 3 ng for all factors throughout this study. Error bars denote standard deviation (n = 4, biological replicates).

Extended Data Table 1 TF recovery analysis for S2 cell enhancer contexts

Supplementary information

Supplementary Table 1: Activities of 812 tested TFs and cofactors

Activities across all 24 contexts and the assigned cluster for each of the 474 TFs and 338 cofactors. For each context, shown are log2-transformed luciferase activities (median fold-changes compared to targeted recruitment of GFP) and t-test p-values that are not multiple-testing corrected. The multiple-testing correction depends on the number of experiments (see main text), which is 19,488 experiments when considering all 812 factors across all 24 contexts. (XLSX 558 kb)

Supplementary Table 2: Feature enrichment analysis for the 15 different clusters of TFs

Fold changes underlying the heatmaps in Figures 2d and Extended Data Figure 4. Shown are significant (empirical FDR≤0.1, see Methods) log2-transformed fold-enrichment and depletion values when comparing the following features for TFs in each of the clusters with all TFs (Fig. 2d and Extended Data Fig. 4 show a subset): de novo discovered motifs using MEME42, Eukaryotic Linear Motifs44 (ELMs), Pfam domains32, GO annotations45, and IMAGO gene expression patterns in the Drosophila embryo46 (FDR≤0.1 suggests that 10% of the reported enrichments might occur by chance). (XLSX 53 kb)

Supplementary Table 3: List of 338 cofactors tested

Details of the cofactors tested including the criterion we used to categorize each protein as transcriptional cofactor, the Pfam domain content, and the primer sequences used to amplify the coding sequence from cDNA (Gateway attB sites not included in sequence; see Methods). The column criterion might be ‘Complex’ (protein is annotated as being part of a cofactor complex), ‘Domain’ (protein contains at least one Pfam domain typically found in cofactors), ‘Flybase’ (protein is annotated to act as coactivator or corepressor in FlyBase), or ‘Homology’ (mouse or human orthologue of protein acts as cofactor). (XLSX 83 kb)

Supplementary Data 1: Tested TF and Cofactor cDNA sequences

cDNA sequence for each of the 812 tested proteins in FASTA format. For the clones obtained from ref. 11 we included the coding sequences from the reference genome. For the clones we generated and validated ourselves we included the actual experimentally determined sequences as assembled by PrInSes-C34 (see Methods). (TXT 1391 kb)

PowerPoint slides

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Stampfel, G., Kazmar, T., Frank, O. et al. Transcriptional regulators form diverse groups with context-dependent regulatory functions. Nature 528, 147–151 (2015).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

This article is cited by


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing