Genome-scale functional characterization of Drosophila developmental enhancers in vivo


Transcriptional enhancers are crucial regulators of gene expression and animal development1 and the characterization of their genomic organization, spatiotemporal activities and sequence properties is a key goal in modern biology2,3,4,5,6,7,8. Here we characterize the in vivo activity of 7,705 Drosophila melanogaster enhancer candidates covering 13.5% of the non-coding non-repetitive genome throughout embryogenesis. 3,557 (46%) candidates are active, suggesting a high density with 50,000 to 100,000 developmental enhancers genome-wide. The vast majority of enhancers display specific spatial patterns that are highly dynamic during development. Most appear to regulate their neighbouring genes, suggesting that the cis-regulatory genome is organized locally into domains, which are supported by chromosomal domains, insulator binding and genome evolution. However, 12 to 21 per cent of enhancers appear to skip non-expressed neighbours and regulate a more distal gene. Finally, we computationally identify cis-regulatory motifs that are predictive and required for enhancer activity, as we validate experimentally. This work provides global insights into the organization of an animal regulatory genome and the make-up of enhancer sequences and confirms and generalizes principles from previous studies1,9. All enhancer patterns are annotated manually with a controlled vocabulary and all results are available through a web interface (, including the raw images of all microscopy slides for manual inspection at arbitrary zoom levels.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.

Figure 1: Enhancers display highly diverse and dynamic activity patterns across Drosophila development.
Figure 2: The organization of the Drosophila cis-regulatory genome.
Figure 3: Intragenic enhancers in the Drosophila genome.
Figure 4: Prediction and validation of cis-regulatory motif requirements for tissue-specific enhancer activities.


  1. 1

    Levine, M. Transcriptional enhancers in animal development and evolution. Curr. Biol. 20, R754–R763 (2010)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  2. 2

    The modENCODE Consortium. Identification of functional elements and regulatory circuits by Drosophila modENCODE. Science 489, 57–74 (2010)

  3. 3

    ENCODE Project Consortium. An integrated encyclopedia of DNA elements in the human genome. Nature 489, 57–74 (2012)

  4. 4

    Pennacchio, L. A. et al. In vivo enhancer analysis of human conserved non-coding sequences. Nature 444, 499–502 (2006)

    CAS  PubMed  PubMed Central  Article  ADS  Google Scholar 

  5. 5

    Jenett, A. et al. A GAL4-driver line resource for Drosophila neurobiology. Cell Rep. 2, 991–1001 (2012)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  6. 6

    Manning, L. et al. A resource for manipulating gene expression and analyzing cis-regulatory modules in the Drosophila CNS. Cell Rep. 2, 1002–1013 (2012)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  7. 7

    Jory, A. et al. A survey of 6,300 genomic fragments for cis-regulatory activity in the imaginal discs of Drosophila melanogaster. Cell Rep. 2, 1014–1024 (2012)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  8. 8

    Zinzen, R. P., Girardot, C., Gagneur, J., Braun, M. & Furlong, E. E. M. Combinatorial binding predicts spatio-temporal cis-regulatory activity. Nature 462, 65–70 (2009)

    CAS  ADS  Google Scholar 

  9. 9

    Yáñez-Cuna, J. O., Kvon, E. Z. & Stark, A. Deciphering the transcriptional cis-regulatory code. Trends Genet. 29, 11–22 (2013)

    PubMed  Article  CAS  Google Scholar 

  10. 10

    Visel, A. et al. ChIP-seq accurately predicts tissue-specific activity of enhancers. Nature 457, 854–858 (2009)

    CAS  PubMed  PubMed Central  Article  ADS  Google Scholar 

  11. 11

    Shen, Y. et al. A map of the cis-regulatory sequences in the mouse genome. Nature 448, 116–120 (2012)

    Article  ADS  CAS  Google Scholar 

  12. 12

    Zeitlinger, J. & Stark, A. Developmental gene regulation in the era of genomics. Dev. Biol. 339, 230–239 (2010)

    CAS  PubMed  Article  Google Scholar 

  13. 13

    Rubin, G. M. & Lewis, E. B. A brief history of Drosophila’s contributions to genome research. Science 287, 2216–2218 (2000)

    CAS  PubMed  Article  Google Scholar 

  14. 14

    Tomancak, P. et al. Systematic determination of patterns of gene expression during Drosophila embryogenesis. Genome Biol. 3, RESEARCH0088 (2002)

    PubMed  PubMed Central  Article  Google Scholar 

  15. 15

    Gallo, S. M. et al. REDfly v3.0: toward a comprehensive database of transcriptional regulatory elements in Drosophila. Nucleic Acids Res. 39, D118–D123 (2011)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  16. 16

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

    CAS  PubMed  Article  ADS  PubMed Central  Google Scholar 

  17. 17

    Campos-Ortega, J. A. & Hartenstein, V. The Embryonic Development of Drosophila melanogaster (Springer, 1997)

    Google Scholar 

  18. 18

    Graveley, B. R. et al. The developmental transcriptome of Drosophila melanogaster. Nature 471, 473–479 (2011)

    CAS  Article  ADS  Google Scholar 

  19. 19

    Bonn, S. et al. Tissue-specific analysis of chromatin state identifies temporal signatures of enhancer activity during embryonic development. Nature Genet. 44, 148–156 (2012)

    MathSciNet  CAS  Article  Google Scholar 

  20. 20

    Thomas, S. et al. Dynamic reprogramming of chromatin accessibility during Drosophila embryo development. Genome Biol. 12, R43 (2011)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  21. 21

    Sexton, T. et al. Three-dimensional folding and functional organization principles of the Drosophila genome. Cell 148, 458–472 (2012)

    CAS  Article  Google Scholar 

  22. 22

    Nègre, N. et al. A comprehensive map of insulator elements for the Drosophila genome. PLoS Genet. 6, e1000814 (2010)

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  23. 23

    Drosophila 12 Genomes Consortium Evolution of genes and genomes on the Drosophila phylogeny. Nature 450, 203–218 (2007)

    Article  CAS  Google Scholar 

  24. 24

    Hong, J.-W., Hendrix, D. A. & Levine, M. S. Shadow enhancers as a source of evolutionary novelty. Science 321, 1314 (2008)

    CAS  PubMed  PubMed Central  Article  ADS  Google Scholar 

  25. 25

    Calhoun, V. C., Stathopoulos, A. & Levine, M. Promoter-proximal tethering elements regulate enhancer-promoter specificity in the Drosophila Antennapedia complex. Proc. Natl Acad. Sci. USA 99, 9243–9247 (2002)

    CAS  PubMed  Article  ADS  Google Scholar 

  26. 26

    Jack, J., Dorsett, D., Delotto, Y. & Liu, S. Expression of the cut locus in the Drosophila wing margin is required for cell type specification and is regulated by a distant enhancer. Development 113, 735–747 (1991)

    CAS  PubMed  Google Scholar 

  27. 27

    Maeda, R. K. & Karch, F. Gene expression in time and space: additive vs hierarchical organization of cis-regulatory regions. Curr. Opin. Genet. Dev. 21, 187–193 (2011)

    CAS  PubMed  Article  Google Scholar 

  28. 28

    Marinić, M., Aktas, T., Ruf, S. & Spitz, F. An integrated holo-enhancer unit defines tissue and gene specificity of the Fgf8 regulatory landscape. Dev. Cell 24, 530–542 (2013)

    PubMed  Article  CAS  Google Scholar 

  29. 29

    Yáñez-Cuna, J. O., Dinh, H. Q., Kvon, E. Z., Shlyueva, D. & Stark, A. Uncovering cis-regulatory sequence requirements for context-specific transcription factor binding. Genome Res. 22, 2018–2030 (2012)

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  30. 30

    Liang, H.-L. et al. The zinc-finger protein Zelda is a key activator of the early zygotic genome in Drosophila. Nature 456, 400–403 (2008)

    CAS  PubMed  PubMed Central  Article  ADS  Google Scholar 

  31. 31

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

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  32. 32

    Weiszmann, R., Hammonds, A. S. & Celniker, S. E. Determination of gene expression patterns using high-throughput RNA in situ hybridization to whole-mount Drosophila embryos. Nature Protocols 4, 605–618 (2009)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  33. 33

    Lécuyer, E., Parthasarathy, N. & Krause, H. M. Fluorescent in situ hybridization protocols in Drosophila embryos and tissues. Methods Mol. Biol. 420, 289–302 (2008)

    PubMed  Article  Google Scholar 

  34. 34

    Kvon, E. Z., Stampfel, G., Yáñez-Cuna, J. O., Dickson, B. J. & Stark, A. HOT regions function as patterned developmental enhancers and have a distinct cis-regulatory signature. Genes Dev. 26, 908–913 (2012)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  35. 35

    Fisher, W. W. et al. DNA regions bound at low occupancy by transcription factors do not drive patterned reporter gene expression in Drosophila. Proc. Natl Acad. Sci. USA 109, 21330–21335 (2012)

    CAS  PubMed  Article  ADS  Google Scholar 

  36. 36

    Frise, E., Hammonds, A. S. & Celniker, S. E. Systematic image-driven analysis of the spatial Drosophila embryonic expression landscape. Mol. Syst. Biol. 6, 345 (2010)

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  37. 37

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

    CAS  PubMed  PubMed Central  Article  ADS  Google Scholar 

  38. 38

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

    CAS  PubMed  Article  Google Scholar 

  39. 39

    Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842 (2010)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  40. 40

    Small, S., Blair, A. & Levine, M. Regulation of even-skipped stripe 2 in the Drosophila embryo. EMBO J. 11, 4047–4057 (1992)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  41. 41

    Visel, A., Minovitsky, S., Dubchak, I. & Pennacchio, L. A. VISTA Enhancer Browser–a database of tissue-specific human enhancers. Nucleic Acids Res. 35, D88–D92 (2007)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  42. 42

    Hartenstein, V. Atlas of Drosophila development (1993)

    Google Scholar 

  43. 43

    Erives, A. & Levine, M. Coordinate enhancers share common organizational features in the Drosophila genome. Proc. Natl Acad. Sci. USA 101, 3851–3856 (2004)

    CAS  PubMed  Article  ADS  Google Scholar 

  44. 44

    Perry, M. W., Boettiger, A. N. & Levine, M. Multiple enhancers ensure precision of gap gene-expression patterns in the Drosophila embryo. Proc. Natl Acad. Sci. USA 108, 13570–13575 (2011)

    CAS  PubMed  Article  ADS  Google Scholar 

  45. 45

    Gajewski, K. et al. Pannier is a transcriptional target and partner of Tinman during Drosophila cardiogenesis. Dev. Biol. 233, 425–436 (2001)

    CAS  PubMed  Article  Google Scholar 

  46. 46

    Harrison, M. M., Li, X.-Y., Kaplan, T., Botchan, M. R. & Eisen, M. B. Zelda binding in the early Drosophila melanogaster embryo marks regions subsequently activated at the maternal-to-zygotic transition. PLoS Genet. 7, e1002266 (2011)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  47. 47

    Nien, C.-Y. et al. Temporal coordination of gene networks by Zelda in the early Drosophila embryo. PLoS Genet. 7, e1002339 (2011)

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  48. 48

    Nambu, J. R., Lewis, J. O., Wharton, K. A. & Crews, S. T. The Drosophila single-minded gene encodes a helix-loop-helix protein that acts as a master regulator of CNS midline development. Cell 67, 1157–1167 (1991)

    CAS  PubMed  Article  Google Scholar 

  49. 49

    Ochoa-Espinosa, A. et al. The role of binding site cluster strength in Bicoid-dependent patterning in Drosophila. Proc. Natl Acad. Sci. USA 102, 4960–4965 (2005)

    CAS  PubMed  Article  ADS  Google Scholar 

  50. 50

    Baylies, M. K. & Bate, M. twist: a myogenic switch in Drosophila. Science 272, 1481–1484 (1996)

    CAS  PubMed  Article  ADS  Google Scholar 

  51. 51

    Nguyen, H. T., Bodmer, R., Abmayr, S. M., McDermott, J. C. & Spoerel, N. A. D-mef2: a Drosophila mesoderm-specific MADS box-containing gene with a biphasic expression profile during embryogenesis. Proc. Natl Acad. Sci. USA 91, 7520–7524 (1994)

    CAS  PubMed  Article  ADS  Google Scholar 

Download references


We thank members of the Dickson laboratory VT project for cloning the candidate regions and generating transgenic flies and the VDRC ( for their maintenance and distribution. We are grateful to the IMP/IMBA Scientific Services, in particular BioOptics, Genomics, and IT for help and to C. H. Lampert (IST Austria) for advice. We thank M. Levine (UC Berkeley) and V. Hartenstein (UCLA) for their permission to reproduce figures. The Stark group is supported by a European Research Council (ERC) Starting Grant from the European Community’s Seventh Framework Programme (FP7/2007-2013)/ERC grant agreement no. 242922 awarded to A.S. and by the Austrian Science Fund (FWF, F4303-B09). Generation of transgenic lines was supported in part by an ERC Advanced Investigator Grant to B.J.D. Basic research at the IMP is supported by Boehringer Ingelheim GmbH.

Author information




E.Z.K. and A.S. conceived the study and wrote the paper. E.Z.K., G.S., M.P. and K.S. performed the screen, E.Z.K. annotated all patterns and performed validation experiments, E.Z.K. and T.K. performed the imaging, T.K. and G.S. analysed the image data, T.K. developed, J.O.Y.-C. performed all sequence analyses, B.J.D. provided transgenic flies prior to publication, E.Z.K., T.K., G.S., J.O.Y-C. and A.S. analysed data. A.S. supervised the project.

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 A high-throughput in vivo enhancer screen in Drosophila embryos.

a, The distribution of tested candidate fragments (blue) across the D. melanogaster genome (fragments in heterochromatic regions of the chromosomes are not shown). b, Schematic representation of the candidate fragments in the VT library used. The cartoon displays a genomic locus (top) with genes (black) and enhancer candidates (blue). The library contains each candidate fragment in a constant transcriptional GAL4 reporter (middle) integrated in a constant genomic landing site. c, Coverage of the genome by the VT library. Shown is the distribution of the tested DNA fragments in the Drosophila genome with respect to the distance to the closest gene TSS. d, Distribution of the tested DNA fragments in the genome with respect to the expression of the closest gene at relevant developmental stages (2–24 hours post fertilization (h.p.f.)). Each region was uniquely assigned to the closest gene (sixth and eighth columns), and the fraction of genes which are expressed (RPKM values higher than 1), highly expressed (top 20% of genes according to their RPKM values), or active in various tissues of the stages 13–16 embryo is indicated (*embryonic developmental time course RNA-seq data18; **RNA in situ hybridization data from BDGP14; 65,911 fragments cover the whole non-coding, non-repetitive genome and serve as baseline). The deviation from the genome average is typically within the range of 0.8- to 1.2-fold. The few exceptions are genes active in CNS (1.8–2.0-fold enriched) and salivary glands (0.36-fold). e, Pipeline for the assessment of enhancer activities in Drosophila embryos. We collected about 400 embryos representing all stages of embryogenesis per transgenic VT strain and performed whole-mount 96-well in situ hybridization with an antisense GAL4 probe. Positively stained embryos were imaged and their enhancer activity pattern was annotated manually with a controlled vocabulary. A representative enhancer (VT4454) active in a subset of the dorsal epidermis is shown. All images and annotations are available online (

Extended Data Figure 2 Validation of the screening pipeline and context-independence of enhancer activities.

a, Reporter activity of the eve stripe 2 minimal enhancer40 in the context of longer genomic fragments (the average size of the fragments tested in this study is 2,022 base pairs (bp)). The minimal 484 bp eve stripe 2 enhancer40 was cloned in different positions within longer fragments corresponding to the enhancer’s extended genomic context: minimal element alone (first row), in the middle (second row) or towards one side of a 2 kb fragment (third row), each in two different orientations (forward, second column; reverse, third column). In all cases the endogenous eve stripe 2 activity40 was reproduced. b, Recovery of known Drosophila enhancers. Twenty-eight previously published non-redundant enhancers (first column, enhancers with available images from the REDfly database15) fully overlapped with tested VT fragments (third column). For 24 of them, the VT fragment fully or partially recapitulated the published pattern. Of the four REDfly enhancers, for which the activity was not fully reproduced, three partially matched to the pattern of the VT fragment and only one was inactive in our screen. c, Activity of the human DNA sequences in the Drosophila embryos. Only two of 12 human ultra-conserved enhancer sequences41 that we tested in the screen’s reporter setup were very weakly active, below the threshold we used to defined active enhancers. The inactivity of all 12 non-Drosophila fragments serves as a measure of specificity and demonstrates that less than 1 in 12 (8%) of random fragments are expected to be active. d, The majority of enhancers show identical activity patterns in the context of two different transcriptional reporter systems (different reporter vector and genomic position/landing site). Shown are representative enhancer activities of the fragments using the GAL4 reporter integrated at the attP2 landing site (VT strains, left column) or using a LexAGAD reporter integrated at the attP40 landing site (VTL strains, right column; Supplementary Information section 1). We re-tested a total of 112 fragments in VTL strains, of which 34 were negative and 78 positive in the original screen in VT strains. e, Fraction (y axis) and total number (numbers inside bars) of fragments that were negative (right bar) or positive (left bar) in VT embryos that in VTL embryos displayed identical (purple), more narrow (dark blue), broader (dark red) and weaker or no activity (grey). The activities measured in both independent reporter systems agreed very well (Supplementary Table 2) and according to our experience, most differences likely stem from differences in in situ sensitivities rather than enhancer function. f, Systematic comparison of enhancer activities to the expression patterns of the neighbouring genes for enhancers for which the expression patterns of all respective neighbours are available14. Shown is the fraction of enhancers (bar heights), which fully (black) or partially (grey) matched to the first (left) and first or second (right) degree neighbouring genes.

Extended Data Figure 3 Enhancers display diverse tissue and cell-type-specific activity patterns.

Shown are representative A–P, D–V, composite and other activity patterns observed in early embryos (a, stages 4–6, 1.5–3.3 h.p.f.) and in various tissues/cell types of late embryos (b, stages 13–14, 9.5–11.3 h.p.f.). The number of VT fragments per pattern class is indicated in parenthesis (see Extended Data Fig. 4a, b for detailed numbers on all patterns).

Extended Data Figure 4 Spatial activity patterns trace tissue fatemap and development.

a, b, Total number of VT fragments active in a given anatomical structure (a, stages 4–6; or b, stages 13–14). c, Co-regulated domains resemble the embryo fate map. The schematic embryo on top shows the fate map of the Drosophila blastoderm (reproduced from ref. 42 with permission) and the one below shows co-regulated domains determined by unbiased reverse clustering of the raw microscope images for 429 early enhancers. These domains correspond to the following presumptive germ layers (colour coded and manually annotated, from left to right): anterior endoderm (corresponds to es/pv. (oesophagus/proventriculus), eph (epipharynx) and hy (hypopharynx) of the fate map on the top; dark blue), anterior midgut (corresponds to amg; red), head mesoderm (corresponds to ms (mesoderm)), head neuroectoderm (head NE; corresponds to pNR (procephalic neurogenic region); purple), ventral ectoderm (corresponds to vNR (ventral neurogenic region); purple and dark blue), trunk mesoderm (corresponds to ms; green), dorsal ectoderm (corresponds to dEpi (dorsal epidermis); light blue and dark blue), amnioserosa (corresponds to as; grey) hindgut (corresponds to hg; dark blue) and posterior midgut (corresponds to pmg; red). d, Enhancer activities during embryogenesis follow the presumptive tissue fate map. Columns represent enhancers active in major presumptive cell types or ubiquitously in all cell types of the early Drosophila embryo. Rows show the corresponding most strongly enriched annotation terms at later stages (rows for which all corrected enrichment values were below 20.25 (1.19) were excluded; rows and columns are clustered). eg, The same as in d but now going from late to early stages. For enhancers active in the late embryo (stages 13–16, columns) specifically in the gut (e), CNS (f) or ubiquitously (g), the most enriched terms for earlier stages are shown (stages 4–10; rows). Only enhancers that are active in both early and late stages were considered and rows and columns were clustered.

Extended Data Figure 5 Enhancer activities are highly dynamic during Drosophila embryogenesis.

a, The majority of all enhancers are active only during specific time points of embryo development. Heatmap representation of 2,563 strong enhancers, clustered on the basis of their temporal activity profiles during six-stage intervals of Drosophila embryogenesis. Seven major temporal groups of enhancers are indicated on the right, together with representative transgenic embryos for each of the groups. b, c, Enhancers display activity patterns that are temporally and spatially sparser than gene expression patterns (statistical significance was estimated by binomial P values shown above the bars). b, The fraction of genes that are ubiquitously expressed at all stages (20.5%) was >25 fold higher than the fraction of ubiquitously active enhancers (0.8%). c, Similarly, the percentage of continuously expressed genes (41.1%) was >5 fold higher than the percentage of continuously active enhancers (8.1%). We considered stages 4 through 16 (1.5–14 h.p.f.) from the Berkeley Drosophila Genome Project (BDGP)14.

Extended Data Figure 6 The dynamics of enhancer activities are reflected in the dynamics of the chromatin landscape.

a, H3K4me1, H3K9me3, H3K9ac, H3K4me3 and K3K27me3 histone marks2 of early (E), middle (M), late (L) and continuous (C) enhancers (data from whole embryos). b, c, ChIP and DHS signals help to refine minimal enhancer elements. UCSC Genome Browser screen shots showing the twist (twi; b) and ventral nervous system defective (vnd; c) genomic loci including stage 5 DNA accessibility20 and 0–4 h.p.f. ChIP data for H3K4me1, H3K27ac and CBP/p300 (ref. 2). Displayed are genomic fragments tested in this study (purple), which could be refined to smaller elements (RE, refined elements; grey) that coincide with known minimal enhancers (ME; blue15,19). The corresponding transgenic embryos show the respective enhancer activities during stage 5 in mesoderm (b) or neuroectoderm (c; right image is reproduced from ref. 43 with permission, copyright (2004) National Academy of Sciences, USA; ME and LacZ added to image for clarity). See Supplementary Table 3 for the full list of refined elements. d, Tissue-specific 6–8 h.p.f. ChIP signals for histone marks, the Mef2 transcription factor and PolII binding (data from Batch isolation of tissue-specific chromatin for immunoprecipitation (BiTS-ChIP)19) on early (E), middle (M) and late (L) mesodermal enhancers and on enhancers active outside the mesoderm (analysis as in ref. 19 but evaluated for enhancers from this study). e, The corresponding ChIP signals from whole embryos2 for comparison.

Extended Data Figure 7 The cis-regulatory organization of the Drosophila genome.

a, We inspected the intergenic regions (IRs) between intergenic enhancers (purple) and their assigned target genes (blue) and the neighbouring gene that was not assigned (red) for the location of chromosomal domain boundaries as determined by Hi-C21 (Fig. 2b), breakpoints during evolutionary chromosomal rearrangements (between 12 Drosophila species23), and insulator protein binding sites22. We restricted this analysis to the 151 intergenic enhancers for which both immediately flanking genes were characterized and that were uniquely assigned to one of these first-degree neighbours (see also Fig. 2a). b, Breakpoints during evolutionary genome rearrangements (from 12 Drosophila species23) are significantly reduced between enhancers and their target genes (blue) compared to enhancers and genes they do not regulate (red). Shown are different score cutoffs to define breakpoints with increasing stringency (left to right), which all show a consistent trend and an increasing difference with increasing stringency. c, The location of insulator protein binding sites correlate with enhancer–target gene assignments. Bar plots show the fraction of IRs between enhancers and their assigned target genes (blue) or non-target genes (red) that contain at least one binding site for one of the insulator proteins CTCF, Suppressor of Hairy wing (Su(Hw)), Boundary element-associated factor of 32kD (BEAF-32), Modifier of mdg4 (Mod(mdg4)), Centrosomal protein 190kD (Cp190), or the transcriptional activator Trl as a control (1% false discovery rate regions from ref. 22). Statistical significance in b, c was estimated by binomial P values (above the bars). d, An enhancer located in the Ultrabithorax (Ubx) intron does not fully match the Ubx expression pattern (overview of the Ubx locus on top). The VT42746 fragment drives a characteristic gap gene expression pattern in the early stage 6 embryo (upper right embryo), which recapitulates the known Ubx pattern at that stage. During stage 15, however, it displays strong activity in salivary glands (lower right embryo) which does not match Ubx expression14. e, Genes that appear to be regulated by more than one enhancer with identical or overlapping activity patterns (shadow enhancers24) at stages 4–6, including known shadow enhancers24,44. f, g, Examples of shadow enhancers in the tailup (tup, f) and senseless (sens, g) gene loci. Displayed are genomic fragments (purple) that act as enhancers with overlapping activity patterns. The corresponding transgenic embryos highlighting enhancer activity during stage 5 in dorsal ectoderm and amnioserosa (f) or during stage 14 in salivary glands (g) together with wild-type embryos stained against tup (f) or sens (g) mRNA14 are shown below.

Extended Data Figure 8 The cis-regulatory organization of the SoxN locus.

a, b, The regulatory landscape of the SoxN locus (a, upstream; b, downstream). UCSC Genome Browser screenshot including DNA accessibility data as determined by DHS-seq20 and genomic fragments tested in this study (top: positive fragments are purple, negative grey). For strong enhancers, the corresponding GAL4-stained transgenic embryos are shown below at six time points of embryo development. The boxed embryos show the in situ hybridization against SoxN mRNA for each developmental stage.

Extended Data Figure 9 Predicting cis-regulatory motifs requirements across Drosophila tissues and cell types.

a, Schematic overview of our approach. We trained a support vector machine (SVM) to distinguish between functionally similar enhancers (black solid blocks) and control fragments (grey blocks: Neg, negative regions; Pos, enhancers with different activity patterns) solely on the basis of their motif content. We excluded each fragment in turn for testing, trained the SVM on the remaining fragments, and predicted the test fragment as described before29. b, Representative enhancers active in the midgut (stages 13–15), broad CNS (stages 15–16) and A–P system (stages 4–6) and the corresponding receiver-operating-characteristic (ROC) curves together with area under the curve (AUC) values for predictions of each of the groups (red curve, versus inactive control regions; blue curve, versus other enhancers) and for controls, for which we randomized the enhancers’ class assignments (black curve, versus inactive control regions; grey curve, versus other enhancers). The predictions were not successful when we shuffled the enhancers’ assignments between classes, demonstrating that the predictive signals reside within the enhancer sequences rather than stemming merely from the computational procedure per se29 (see Supplementary Table 5 for more enhancer groups and technical details). The most discriminative transcription factor motifs together with their enrichments over control regions are shown in columns 3–5. c, Motif-to-tissue associations revealed by motif-enrichment analyses. The heatmap columns represent a subset of all known Drosophila transcription factor motifs29,37 which were discriminative during supervised machine learning and the rows show different enhancer classes. Each matrix cell shows the enrichment of the corresponding motif in enhancers of the corresponding enhancer class (log2); only rows and columns for which at least one matrix cell has enrichment values ≥ 20.7 are shown and the matrix rows and columns are sorted by bi-clustering. Further highlighted are broad CNS and ubiquitous enhancers enriched in Trl (GAGA) and CAC(N)NCAC-like motifs (region 1); midgut tube enhancers enriched in GATA-like motifs (region 2); embryonic heart enhancers enriched in motifs for Tinman and Pannier45 (region 3); early A–P and D–V enhancers enriched in Zelda motifs30,46,47 (regions 4, 6a and 7a); ventral midline enhancers enriched in motifs for Single-minded48 (Sim) (region 5); anterior endoderm, procephalic ectoderm and A–P system enhancers enriched in motifs for Bicoid (Bcd)49 and Ttk (regions 6b and 7b); early mesoderm enhancers enriched in Twi motifs50 (region 8) and late somatic muscle enhancers enriched in Mef2 motifs51 (region 9).

Extended Data Figure 10 Validation of cis-regulatory motif requirements for enhancer activity.

am, Wild-type (WT; top cartoon) and motif-mutated (mutant; bottom cartoon) enhancers were tested by the transcriptional reporter gene assay used here. Shown are the representative embryos for WT (top) and mutant (bottom) enhancers. The activity of 10 out of 11 enhancers (al) was abolished or strongly reduced as we quantified directly from the strengths of the in situ signal in the original microscopy images. Histogram (middle) shows the distribution of in situ staining intensities in different parts of the embryo for WT (blue) and corresponding mutant (green) enhancer averaged over all independent embryos of the same genotype. Right, quantification of inferred enhancer strengths for WT (blue) and mutant (green) enhancers in transgenic reporter tests. Shown are the in situ staining levels for individual reporter-bearing embryos (dots), mean (horizontal lines) and standard deviations (P values according to Kolmogorov–Smirnov). The activity of four different broad CNS enhancers was dependent on presence of Trl (GAGA) motifs (ad), the activity of three different midgut enhancers on the presence of GATA motifs (eg), and the activity of three out of four tested early A–P enhancers was dependent on the presence of AGGAC (Ttk) motifs (hl). n, position weight matrixes of the transcription factor motifs tested in am.

Supplementary information

Supplementary Information

This file contains Supplementary Notes 1-6, legends for Supplementary Tables 1-6 (see separate zipped file) and Supplementary References. (PDF 243 kb)

Supplementary Tables

This zipped file contains Supplementary Tables 1-6 (see Supplementary Information file for legends). (ZIP 344 kb)

PowerPoint slides

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kvon, E., Kazmar, T., Stampfel, G. et al. Genome-scale functional characterization of Drosophila developmental enhancers in vivo. Nature 512, 91–95 (2014).

Download citation

Further reading


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.


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