Only a subset of the genes in a cell is active at any time. Gene activation or ‘transcription’ is controlled by specific DNA sequences called promoters and enhancers. Promoters are found next to genes and recruit the protein machinery needed to transcribe the gene. Enhancers are located further away from genes and interact with the promoter to increase gene transcription.

Some genes have multiple enhancers with overlapping activities, typically including a primary enhancer and secondary or ‘shadow’ enhancer that is even more distant. Multiple enhancers are used to control gene transcription in different types of cells. For example, shortly after fertilization around 1000 enhancers in the embryo of the fruit fly Drosophila regulate several hundred genes that are important for development. These activities establish which cell type any given cell in the embryo will become. Many of these developmental genes have shadow enhancers, which are thought to make gene activity more precise and reliable. It is not known, however, how several enhancers interact with the same promoter at the same time. For example, do the enhancers' individual effects add together? Or can the combined effects of multiple enhancers be more (or even less) than the sum of their parts?

Bothma, Garcia et al. have now examined how multiple enhancers regulate the activity of three developmental genes (called hunchback, knirps, and snail) in early Drosophila embryos. The experiments showed that the individual effects of the knirps enhancers add together as one might expect. On the other hand, the snail enhancers interfere with each other, which means that their combined effect on transcription is less than the sum of the two individual effects. Furthermore, other experiments revealed that the combined effect of the hunchback enhancers depends upon whether another component that is needed for gene activation is in short supply.

To understand these observations, Bothma, Garcia et al. then developed a mathematical model. The model proposes that behavior of enhancers depends upon how strongly they interact with the target promoter. Since ‘weak’ enhancers do not interact very often, their effects can easily add together. However, the effects of ‘strong’ enhancers do not add together because they often compete to interact with the promoter. These findings show how multiple enhancers can work in a complex manner to control gene transcription.

There is emerging evidence that metazoan genes occur in a complex regulatory landscape encompassing numerous enhancers (Jeong et al., 2006; Hong et al., 2008; Frankel et al., 2010; Perry et al., 2011; Rada-Iglesias et al., 2011; Buecker and Wysocka, 2012; Lagha et al., 2012; Levine et al., 2014). For example, the mouse Sonic Hedgehog gene is regulated by at least 20 different enhancers scattered over a distance of ∼1 Mb (Jeong et al., 2006). Individual enhancers mediate expression in a variety of different tissues, including the brain, floorplate, and limb buds. Multiple enhancers with overlapping regulatory activities are also used to control gene expression within individual cell types. For example, the transcriptional activation of the gap gene hunchback (hb) in the early Drosophila embryo is mediated by both a proximal enhancer and distal ‘shadow’ enhancer that independently mediate activation in response to high levels of the Bicoid activator gradient (Buecker and Wysocka, 2012). Despite overwhelming evidence for multiple enhancers regulating the same gene it is unknown whether they simultaneously interact with the same promoter in a given cell. Here, we use a combination of quantitative live imaging and theoretical modeling to investigate the function of multiple enhancers for the regulation of a common target gene within a single cell type.

The fate map of the adult fly is established by ∼1000 enhancers that regulate several hundred patterning genes during the 1-hr interval between two and three hrs after fertilization (Levine, 2010; Nien et al., 2011). As many as half of these genes contain ‘shadow’ enhancers with overlapping spatiotemporal activities that are thought to improve the precision and reliability of gene expression (Hong et al., 2008; Frankel et al., 2010; Perry et al., 2010, 2011; Lagha et al., 2012; Miller et al., 2014). For example, the hb shadow enhancer helps produce a sharp boundary of activation by the Bicoid gradient, while its snail (sna) counterpart helps ensure reliable activation under stressful conditions such as high temperatures (Perry et al., 2010, 2011). There is emerging evidence that shadow enhancers are used pervasively in a variety of developmental processes, in both invertebrates and vertebrates (Lagha et al., 2012; Arnold et al., 2013; Miller et al., 2014; Lam et al., 2015).

The underlying mechanisms by which two enhancers with extensively overlapping regulatory activities produce coordinated patterns of gene expression are uncertain. It is possible that they augment the levels of gene expression above the minimal thresholds required to execute appropriate cellular processes (Gregor et al., 2014). However, there is currently only limited experimental evidence for enhancers acting in an additive fashion (Arnold et al., 2013; Lam et al., 2015). An alternative view is that shadow enhancers suppress transcriptional noise and help foster uniform expression among the different cells of a population (Buecker and Wysocka, 2012). To explore these and other potential mechanisms, we examined the timing and levels of gene activity using bacterial artificial chromosomes (BAC) transgenes containing individual enhancers and combinations of primary and shadow enhancers in the early Drosophila embryo.

BAC transgenes containing three key patterning genes, hb, knirps (kni), and sna, were examined in living precellular embryos. Quantitative analyses suggest that shadow enhancers mediate different mechanisms of transcriptional activity. For kni, we observe additive, even super-additive, activities of the primary and shadow enhancer pairs. In contrast, the hb enhancers function sub-additively in anterior regions containing saturating levels of the Bicoid activator but function additively in regions where there are diminishing levels of the Bicoid gradient. Strikingly, sub-additive behavior is also observed for sna, in that removal of the proximal enhancer causes a significant increase in gene expression. These observations suggest that the levels of enhancer activity determine the switch between additive and non-additive behaviors.

Using theoretical modeling, we suggest that these behaviors can be understood in the context of enhancers competing or cooperating for access to the promoter. Weak enhancers work additively due to infrequent interactions with the target promoter, whereas strong enhancers are more likely to impede one another due to frequent associations. Our results highlight the potential of combining quantitative live imaging and modeling in order to dissect the molecular mechanisms responsible for the precision of gene control in development (Gregor et al., 2014) and provide a preview into the complex function of multiple enhancers interacting with the same promoter.

Previous live-imaging studies have relied on simple gene fusions containing a single enhancer attached to a reporter gene with MS2 stem loops inserted in either the 5′ or 3′ UTR (Garcia et al., 2013; Lucas et al., 2013; Bothma et al., 2014). Detection depends on the binding of a maternal mRNA-binding fusion protein (MCP::GFP) expressed throughout the early embryo. In order to examine the interplay between multiple enhancers, we created a series of BAC transgenes containing complete regulatory landscapes (summarized in Figure 1). The BAC transgenes contain an MS2-yellow reporter gene in place of the endogenous transcription units (Figure 1A). For each locus, hb, kni, and sna, we examined a series of three BAC transgenes: containing both primary and shadow enhancers, as well as derivatives lacking individual enhancers (Figure 1B,C). As expected, the BAC transgenes containing both enhancers produce robust expression of the MS2 reporter gene which recapitulate endogenous patterns previously measured using mRNA FISH and immunostaining (Perry et al., 2010, 2011) (Figure 1D–I, Videos 1–3).

Figure 1 with 1 supplement see all

Download asset Open asset

Video 1

Download asset

Video 2

Download asset

Video 3

Download asset

Enhancer ‘deletions’ were created by substituting native sequences with neutral sequences of similar sizes (see ‘Materials and methods’). These substitutions remove most of the critical sequences identified by ChIP-Seq assays (Harrison et al., 2011; Nien et al., 2011; Perry et al., 2011). It is nonetheless possible that critical flanking sequences persist within the transgenes. However, removal of both kni enhancers eliminates detectable transcripts in abdominal regions of early embryos (Figure 1—figure supplement 1), suggesting that any remaining flanking sequences are insufficient to mediate expression.

Qualitative inspection of the hb and kni expression videos suggests that removal of either the primary or shadow enhancer does not cause a dramatic alteration in the overall patterns of gene activity. In order to identify more nuanced changes, we quantified the transcriptional activities of the complete series of BAC transgenes (Figure 2). The fluorescence intensities of active transcription foci were measured during nuclear cleavage cycles (nc) 13 and 14 at different positions across the anterior–posterior (AP) axis. These intensities were converted into an absolute number of elongating Pol II molecules by calibrating with internal standards (see Garcia et al., 2013). Several embryos were analyzed for each time point, and the data were merged to determine the average behavior as a function of AP position and time.

Figure 2

Download asset Open asset

Hb expression was examined during the ∼15 min interphase of nc 13 when both the primary and shadow enhancers are active, but before the onset of later-acting ‘stripe’ enhancers during nc 14 (Perry et al., 2011, 2012). We measured the transcriptional activity of all three hb BAC transgenes (Figure 1D,F,H). Contrary to simple expectations suggested by previous studies (Arnold et al., 2013), we find that the two hb enhancers do not function in an additive fashion in anterior regions (20–40% egg length, EL) of the embryo (e.g., Figure 2A,B). Indeed, the levels produced by the wild-type transgene fall far short of the additive levels predicted by simply summing the levels of expression produced by the transgenes containing either the shadow or primary enhancer alone (Figure 2A,B). Moreover, the removal of the shadow enhancer has no effect on the levels of transcription in anterior regions, which is consistent with the original conception of the shadow enhancer as a ‘back-up’ in the event of stress (Hong et al., 2008).

A very different scenario is observed in central regions of the embryo (40–50% EL) where hb expression switches from ‘on’ to ‘off’ to form a sharp border (Gregor et al., 2007). In this region, the wild-type transgene produces significantly higher levels of expression than either of the transgenes driven by a single enhancer. In fact, these levels correspond to the values predicted by simply adding the activities of the single-enhancer transgenes. Thus, the two enhancers transition from sub-additive to additive behavior in the region of the embryo where there are diminishing levels of the Bicoid activator gradient. We therefore suggest that the hb enhancers function additively only when they are operating below peak capacity (see below).

To further explore the activities of multiple enhancers, we examined the kni gene, which is regulated by an intronic enhancer and a distal 5′ enhancer (Perry et al., 2011). We focus our analysis on central regions of the abdominal expression pattern since previous studies suggest the occurrence of long-range repressive interactions that establish the borders of the ‘stripe’ (Perry et al., 2011). During early periods of nc 14, the two enhancers function super-additively (Figure 2C). That is, the wild-type kni BAC transgene produces higher levels of expression than the predicted sum of the two transgenes containing either enhancer alone. During later stages of development, there is a twofold reduction in the expression levels of the endogenous gene, and at this time the two enhancers work in a simple additive manner (Figure 2D). Note that the maximum number of elongating polymerase (Pol II) complexes falls short of that seen for hb (compare with Figure 2A,B).

Understanding the stark difference in the behaviors of the hb and kni enhancer pairs necessitates measuring the absolute strengths of the different enhancers. Using absolute counts of mRNA molecules (Little et al., 2013), we calibrated our live fluorescence intensity traces to determine the average numbers of actively elongating Pol II transcription complexes (Garcia et al., 2013). The kni transgenes containing single enhancers exhibit as little as fourfold lower levels of expression as compared with the corresponding hb transgenes (Figure 3A,B). At peak activity, the proximal hb enhancer induces ∼50 transcribing Pol II complexes across the yellow reporter gene. By contrast, individual kni enhancers produce an average of only ∼15 elongating Pol II complexes. We propose that the additive and super-additive behaviors of the two kni enhancers reflect their inherently ‘weaker’ activities as compared with the ‘stronger’ proximal hb enhancer (see ‘Discussion’). Note that, despite these differences, the overall output of transcripts and the overall rate of transcript production are essentially identical for all gap genes (Little et al., 2013).

Figure 3

Download asset Open asset

To test the proposed anti-correlation between enhancer strength and additivity, we analyzed the expression of sna, which is essential for delineating the invaginating mesoderm during gastrulation. sna transgenes containing either the proximal or distal enhancer produce peak transcriptional activities of ∼40 actively transcribing Pol II complexes across the yellow reporter gene, similar to the numbers seen for the proximal hb enhancer (Figure 3C). Thus, both sna enhancers are strong and they exhibit strikingly sub-additive behaviors. In particular, the wild-type transgene displays significantly lower levels of expression than the mutant transgene containing only the shadow enhancer. Thus, strong enhancers not only fail to function additively but also interfere with one another, leading to sub-additive expression levels. This observation is also consistent with an earlier study, which suggested that the weaker proximal enhancer attenuates the activities of the stronger distal shadow enhancer (Dunipace et al., 2011).

In an effort to understand how multiple enhancers might function additively or sub-additively, we developed a mathematical model for dynamic enhancer–promoter interactions. In this model, a single enhancer interacts with its promoter via a forward rate k_on and a backward rate k_off (Figure 4A). The relative values of the forward and reverse rates determine the strength of the enhancer–promoter interaction by controlling what fraction of time the two are bound. When the enhancer and promoter interact, the promoter is in the ON state and initiates transcription at a rate r. This rate can be interpreted as the efficiency of enhancer-mediated transcriptional initiation upon enhancer–promoter interaction. Hence, the observable rate of mRNA production depends on the interaction strength given by the ratio k_on/k_off, and the efficiency r with which transcription is initiated upon interaction (Figure 4B). This scheme can be generalized to include two enhancers (A and B) interacting with the same promoter (Figure 4C and see ‘Materials and methods’ for details of the mathematical analysis). In this model only one enhancer can interact with the promoter at a given time.

Figure 4

Download asset Open asset

When the individual enhancers interact infrequently with the promoter (k_on/k_off << 1) they are unlikely to attempt to engage the promoter simultaneously. In this regime, the enhancers will work additively, and the rate of mRNA production of enhancers A and B will simply be the sum of the production rate of A and B alone as shown in Figure 4D. However, as the strength of promoter–enhancer interactions increases, the combined activity of both enhancers is less than the sum of each individual enhancer. When we model enhancers of different strengths (Figure 4E), the amount of mRNA production is reduced as the enhancer with the weaker transcriptional efficiency interacts more frequently with the promoter. This occurs because the two enhancers compete for access to the promoter, effectively inhibiting one another. Thus, weak enhancers might work additively due to infrequent interactions with the target promoter, whereas strong enhancers interfere with one another due to more frequent interactions (see below).

Our quantitative analysis of hb and kni expression provides seemingly opposing results. For kni, we observe additive, sometimes even super-additive, action of the two enhancers within the presumptive abdomen. In contrast, the two hb enhancers do not function in an additive fashion in anterior regions but are additive only in central regions where expression abruptly switches from ‘on’ to ‘off’. We propose that ‘weak’ enhancers function additively or even super-additively, whereas ‘strong’ enhancers can impede one another (Figure 5).

Figure 5

Download asset Open asset

Additional support for this view is provided by the analysis of sna. We found that the removal of the proximal enhancer significantly augments expression, consistent with the occurrence of enhancer interference within the native locus. It is also conceivable that a single strong enhancer (e.g., hb proximal or sna distal) already mediates maximum binding and release of Pol II at the promoter, and additional enhancers are therefore unable to increase the levels of expression. However, the increase in the levels of sna expression upon removal of the primary enhancer is inconsistent with this explanation. Perhaps, the proximity of the proximal enhancer to the sna promoter gives it a ‘topological advantage’ in blocking access of the distal enhancer (Dunipace et al., 2011). The proximal enhancer might mediate less efficient transcription than the distal enhancer and thereby reduce the overall levels of expression (see Figure 4E). We do not believe that this proposed difference is due to differential rates of Pol II elongation since published (Garcia et al., 2013) and preliminary studies suggest that different enhancers and promoters lead to similar elongation rates (∼2 kb/min; T Fuyaka and M Levine, unpublished results). A nonexclusive alternative possibility is that deletion of the proximal enhancer removes associated sna repression elements (MacArthur et al., 2009), thereby augmenting the efficiency of the distal enhancer.

A minimal model of enhancer–promoter associations provides insights into potential mechanisms. In the parameter regime where such interactions are infrequent the two enhancers display additive behavior. However, in the regime of frequent interactions, enhancers compete for access to the promoter resulting in sub-additive behavior. Enhancer–promoter interaction parameters are likely to vary not only between different enhancers but also as the input patterns are modulated in time and space during development (Rushlow and Shvartsman, 2012; Kok and Arnosti, 2015).

This simple model explains the switch from sub-additive to additive enhancer activities for hb and sna. However, in order to explain the super-additive behavior of the kni enhancers, it would be necessary to incorporate an additional state in the model, whereby both enhancers form an active complex with the same target promoter. Such a complex would have a more potent ability to initiate transcription than individual enhancer–promoter interactions.

In summary, we propose that enhancers operating at reduced activities (‘weak enhancers’) can function in an additive manner due to relatively infrequent interactions with their target promoters. In contrast, ‘strong’ enhancers might function sub-additively due to competition for the promoter (Figure 4E). For hb, this switch between competitive and additive behavior occurs as the levels of Bicoid activator diminish in central regions where the posterior border of the anterior Hb domain is formed. Similarly, stress might reduce the performance of the sna enhancers to foster additive behavior under unfavorable conditions such as increases in temperature (Perry et al., 2010). Our study highlights the complexity of multiple enhancers in the regulation of gene expression. They need not function in a simple additive manner, and consequently, their value may be revealed only when their activities are compromised.

1. 1. Arnold CD
   2. Gerlach D
   3. Stelzer C
   4. Boryń ŁM
   5. Rath M
   6. Stark A

   (2013) Genome-wide quantitative enhancer activity maps identified by STARR-seq

   Science 339:1074–1077.

   https://doi.org/10.1126/science.1232542

   • Google Scholar
2. 1. Bothma JP
   2. Garcia HG
   3. Esposito E
   4. Schlissel G
   5. Gregor T
   6. Levine M

   (2014) Dynamic regulation of eve stripe 2 expression reveals transcriptional bursts in living Drosophila embryos

   Proceedings of the National Academy of Sciences of USA 7:1–6.

   https://doi.org/10.1073/pnas.1410022111

   • Google Scholar
3. 1. Buecker C
   2. Wysocka J

   (2012) Enhancers as information integration hubs in development: lessons from genomics

   Trends in Genetics pp. 1–9.

   https://doi.org/10.1016/j.tig.2012.02.008

   • Google Scholar
4. 1. Di Talia S
   2. Wieschaus EF

   (2012) Short-term integration of Cdc25 dynamics controls mitotic entry during Drosophila gastrulation

   Developmental Cell 22:763–774.

   https://doi.org/10.1016/j.devcel.2012.01.019

   • Google Scholar
5. 1. Dunipace L
   2. Ozdemir A
   3. Stathopoulos A

   (2011) Complex interactions between cis-regulatory modules in native conformation are critical for Drosophila snail expression

   Development 4084:4075–4084.

   https://doi.org/10.1242/dev.069146

   • Google Scholar
6. 1. Frankel N
   2. Davis GK
   3. Vargas D
   4. Wang S
   5. Payre F
   6. Stern DL

   (2010) Phenotypic robustness conferred by apparently redundant transcriptional enhancers

   Nature 466:490–493.

   https://doi.org/10.1038/nature09158

   • Google Scholar
7. 1. Garcia HG
   2. Tikhonov M
   3. Lin A
   4. Gregor T

   (2013) Quantitative imaging of transcription in living Drosophila embryos links polymerase activity to patterning

   Current Biology 23:2140–2145.

   https://doi.org/10.1016/j.cub.2013.08.054

   • Google Scholar
8. 1. Gregor T
   2. Wieschaus EF
   3. Tank DW
   4. Bialek W

   (2007) Probing the limits to positional information

   Cell 130:153–164.

   https://doi.org/10.1016/j.cell.2007.05.025

   • Google Scholar
9. 1. Gregor T
   2. Garcia HG
   3. Little SC

   (2014) The embryo as a laboratory: quantifying transcription in Drosophila

   Trends in Genetics 30:364–375.

   https://doi.org/10.1016/j.tig.2014.06.002

   • Google Scholar
10. 1. Harrison MM
    2. Li X-Y
    3. Kaplan T
    4. Botchan MR
    5. Eisen MB

    (2011) Zelda binding in the early Drosophila melanogaster embryo marks regions subsequently activated at the maternal-to-zygotic transition

    PLOS Genetics 7:e1002266.

    https://doi.org/10.1371/journal.pgen.1002266

    • Google Scholar
11. 1. Hocine S
    2. Raymond P
    3. Zenklusen D
    4. Chao JA
    5. Singer RH

    (2013) Single-molecule analysis of gene expression using two-color RNA labeling in live yeast

    Nature Methods 10:119–121.

    https://doi.org/10.1038/nmeth.2305

    • Google Scholar
12. 1. Hong JW
    2. Hendrix DA
    3. Levine MS

    (2008) Shadow enhancers as a source of evolutionary novelty

    Science 321:1314.

    https://doi.org/10.1126/science.1160631

    • Google Scholar
13. 1. Jeong Y
    2. El-Jaick K
    3. Roessler E
    4. Muenke M
    5. Epstein DJ

    (2006) A functional screen for sonic hedgehog regulatory elements across a 1 Mb interval identifies long-range ventral forebrain enhancers

    Development 133:761–772.

    https://doi.org/10.1242/dev.02239

    • Google Scholar
14. 1. Kok K
    2. Arnosti DN

    (2015) Dynamic reprogramming of chromatin: paradigmatic palimpsests and HES factors

    Frontiers in Genetics 6:29.

    https://doi.org/10.3389/fgene.2015.00029

    • Google Scholar
15. 1. Lagha M
    2. Bothma JP
    3. Levine M

    (2012) Mechanisms of transcriptional precision in animal development

    Trends in Genetics 28:1–8.

    https://doi.org/10.1016/j.tig.2012.03.006

    • Google Scholar
16. 1. Lam DD
    2. de Souza FS
    3. Nasif S
    4. Yamashita M
    5. López-Leal R
    6. Otero-Corchon V
    7. Meece K
    8. Sampath H
    9. Mercer AJ
    10. Wardlaw SL
    11. Rubinstein M
    12. Low MJ

    (2015) Partially redundant enhancers cooperatively naintain mammalian pomc expression above a critical functional threshold

    PLOS Genetics 11:1–21.

    https://doi.org/10.1371/journal.pgen.1004935

    • Google Scholar
17. 1. Levine M
    2. Cattoglio C
    3. Tjian R

    (2014) Looping back to leap forward: transcription enters a new era

    Cell 157:13–25.

    https://doi.org/10.1016/j.cell.2014.02.009

    • Google Scholar
18. 1. Levine M

    (2010) Transcriptional enhancers in animal development and evolution

    Current Biology 20:R754–R763.

    https://doi.org/10.1016/j.cub.2010.06.070

    • Google Scholar
19. 1. Little SC
    2. Tikhonov M
    3. Gregor T

    (2013) Precise developmental gene expression arises from globally stochastic transcriptional activity

    Cell 154:789–800.

    https://doi.org/10.1016/j.cell.2013.07.025

    • Google Scholar
20. 1. Liu F
    2. Morrison AH
    3. Gregor T

    (2013) Dynamic interpretation of maternal inputs by the Drosophila segmentation gene network

    Proceedings of the National Academy of Sciences of USA 110:6724–6729.

    https://doi.org/10.1073/pnas.1220912110

    • Google Scholar
21. 1. Lucas T
    2. Ferraro T
    3. Roelens B
    4. De Las Heras Chanes J
    5. Walczak AM
    6. Coppey M
    7. Dostatni N

    (2013) Live imaging of bicoid-dependent transcription in Drosophila embryos

    Current Biology 23:2135–2139.

    https://doi.org/10.1016/j.cub.2013.08.053

    • Google Scholar
22. 1. MacArthur S
    2. Li X-Y
    3. Li J
    4. Brown JB
    5. Chu HC
    6. Zeng L
    7. Grondona BP
    8. Hechmer A
    9. Simirenko L
    10. Keränen SV
    11. Knowles DW
    12. Stapleton M
    13. Bickel P
    14. Biggin MD
    15. Eisen MB

    (2009) Developmental roles of 21 Drosophila transcription factors are determined by quantitative differences in binding to an overlapping set of thousands of genomic regions

    Genome Biology 10:R80.

    https://doi.org/10.1186/gb-2009-10-7-r80

    • Google Scholar
23. 1. Miller SW
    2. Rebeiz M
    3. Atanasov JE
    4. Posakony JW

    (2014) Neural precursor-specific expression of multiple Drosophila genes is driven by dual enhancer modules with overlapping function

    Proceedings of the National Academy of Sciences of USA 111:17194–17199.

    https://doi.org/10.1073/pnas.1415308111

    • Google Scholar
24. 1. Nien C-Y
    2. Liang H-L
    3. Butcher S
    4. Sun Y
    5. Fu S
    6. Gocha T
    7. Kirov N
    8. Manak JR
    9. Rushlow C

    (2011) Temporal coordination of gene networks by Zelda in the early Drosophila embryo

    PLOS Genetics 7:e1002339.

    https://doi.org/10.1371/journal.pgen.1002339

    • Google Scholar
25. 1. Perry MW
    2. Boettiger AN
    3. Bothma JP
    4. Levine M

    (2010) Shadow enhancers foster robustness of Drosophila gastrulation

    Current Biology 20:1562–1567.

    https://doi.org/10.1016/j.cub.2010.07.043

    • Google Scholar
26. 1. Perry MW
    2. Boettiger AN
    3. Levine M

    (2011) Multiple enhancers ensure precision of gap gene-expression patterns in the Drosophila embryo

    Proceedings of the National Academy of Sciences of USA 108:13570–13575.

    https://doi.org/10.1073/pnas.1109873108

    • Google Scholar
27. 1. Perry MW
    2. Bothma JP
    3. Luu RD
    4. Levine M

    (2012) Precision of hunchback expression in the Drosophila embryo

    Current Biology 22:2247–2252.

    https://doi.org/10.1016/j.cub.2012.09.051

    • Google Scholar
28. 1. Rada-Iglesias A
    2. Bajpai R
    3. Swigut T
    4. Brugmann SA
    5. Flynn RA
    6. Wysocka J

    (2011) A unique chromatin signature uncovers early developmental enhancers in humans

    Nature 470:279–283.

    https://doi.org/10.1038/nature09692

    • Google Scholar
29. 1. Rushlow C
    2. Shvartsman SY

    (2012) Temporal dynamics, spatial range, and transcriptional interpretation of the Dorsal morphogen gradient

    Current Opinion in Genetics & Development 22:542–546.

    https://doi.org/10.1016/j.gde.2012.08.005

    • Google Scholar
30. 1. Venken KJT
    2. Carlson JW
    3. Schulze KL
    4. Pan H
    5. He Y
    6. Spokony R
    7. Wan KH
    8. Koriabine M
    9. de Jong PJ
    10. White KP
    11. Bellen HJ
    12. Hoskins RA

    (2009) Versatile P[acman] BAC libraries for transgenesis studies in Drosophila melanogaster

    Nature Methods 6:431–434.

    https://doi.org/10.1038/nmeth.1331

    • Google Scholar

Author details

1. Jacques P Bothma

   Department of Molecular and Cell Biology, University of California, Berkeley, Berkeley, United States

   Contribution

   JPB, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article, Contributed unpublished essential data or reagents

   Contributed equally with

   Hernan G Garcia

   Competing interests

   The authors declare that no competing interests exist.

2. Hernan G Garcia

   Department of Physics, Princeton University, Princeton, United States

   Present address

   Department of Molecular and Cell Biology, Department of Physics, and Biophysics Graduate Group, University of California, Berkeley, Berkeley, United States

   Contribution

   HGG, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article, Contributed unpublished essential data or reagents

   Contributed equally with

   Jacques P Bothma

   Competing interests

   The authors declare that no competing interests exist.

3. Samuel Ng

   Department of Molecular and Cell Biology, University of California, Berkeley, Berkeley, United States

   Contribution

   SN, Acquisition of data, Analysis and interpretation of data, Contributed unpublished essential data or reagents

   Competing interests

   The authors declare that no competing interests exist.

4. Michael W Perry

   Department of Molecular and Cell Biology, University of California, Berkeley, Berkeley, United States

   Present address

   Department of Biology, New York University, New York, United States

   Contribution

   MWP, Conception and design, Contributed unpublished essential data or reagents

   Competing interests

   The authors declare that no competing interests exist.

5. Thomas Gregor

   1. Department of Physics, Princeton University, Princeton, United States
   2. Lewis-Sigler Institute for Integrative Genomics, Princeton University, Princeton, United States

   Contribution

   TG, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   tg2@princeton.edu

   Competing interests

   The authors declare that no competing interests exist.

6. Michael Levine

   Department of Molecular and Cell Biology, University of California, Berkeley, Berkeley, United States

   Present address

   Department of Molecular Biology, and Lewis-Sigler Institute for Integrative Genomics, Princeton University, Princeton, United States

   Contribution

   ML, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   mlevine@berkeley.edu

   Competing interests

   The authors declare that no competing interests exist.

• 5,753

  views

• 1,222

  downloads

• 160

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.