- Email: [email protected]
The Bicoid Morphogen System
Current Biology 20, R249–R254, March 9, 2010 ª2010 Elsevier Ltd All rights reserved
The Bicoid Morphogen System Aude Porcher and Nathalie Dostatni*
Several fundamental concepts of developmental biology have emerged from studies on the early development of the Drosophila melanogaster embryo. In the late 1980s, studies on Bicoid provided the first solid experimental evidence for the existence of morphogenetic gradients and their implication in axial patterning. Bicoid has since stimulated further research, bringing together developmental and cell biologists, physicists and theoreticians to address fundamental biological questions. These include mechanistic aspects of transcriptional and translational control, molecular and functional aspects of evolution and, more recently with the development of quantitative approaches, the robustness of axial patterning in a systems biology view. However, recent studies provide data which lead to contradictory interpretations. Here, we discuss these recent observations, highlighting the data helping to understand how anterior patterning is achieved under the control of Bicoid and point to novel challenges for future studies. Introduction The bicoid (bcd) gene was initially identified in a screen for maternal genes involved in the development of the anterior-posterior axis of the Drosophila embryo. As embryos from bcd mutant females did not develop anterior structures [1] and as injections of the bcd RNA were able to induce the development of anterior structures at ectopic sites [2], bcd appeared both necessary and sufficient for anterior patterning. The Bicoid (Bcd) protein was shown to be expressed in a concentration gradient with the highest Bcd concentration at the anterior end of the embryo and decreasing towards the posterior [3]. Moreover, Bcd is able to determine distinct cell fates in a dose-dependent manner [4]. Bcd thus represented the first solid instance of a morphogenetic gradient essential for the development of a polarized axis in an organism (Box 1). At the molecular level, the Bcd protein was shown to possess two very different activities, both contributing to anterior-posterior patterning. First, Bcd is a homeo-domain containing transcription factor able to activate [5,6], in synergy with the zinc finger protein Hunchback (Hb) [7], many target genes essential for anterior patterning. Second, Bcd was also shown to bind RNA through its homeo-domain and to down-regulate the translation of maternal Caudal [8,9], a protein involved in the transcriptional control of posterior development [10,11]. Despite its key role in anterior patterning in the fruit fly, the Bcd protein is surprisingly poorly conserved during evolution. Search for sequence conservation indicates that the bcd gene is a recent acquisition in long-germ dipterans and arises from a duplication of the Hox3 homologue zerknu¨llt [12]. This lack of evolutionary conservation is particularly striking, when compared to the other maternal determinants, Hb or Caudal, which are dispensable for anteriorposterior patterning in the fruit fly but more conserved than
Institut Curie, Centre de Recherche, and CNRS, UMR218, Paris, F-75248 France. *E-mail: [email protected]
DOI 10.1016/j.cub.2010.01.026
Minireview
Bcd through evolution [10,13]. Given the transcriptional synergy between maternal Hb and Bcd, it was first proposed that Hb was the ancestral anterior morphogen progressively replaced by the newly acquired Bcd in long-germ dipterans [7]. Recently, the search for anterior-posterior patterning factors in related species, with no bcd-like sequences but in which hunchback is conserved, shed light on how the bcd gene could have emerged during evolution. In the flour beetle Tribolium and in the long-germ wasp Nasonia, anterior patterning was shown to be controlled by Hb and another homeo-domain containing protein with the same DNA-binding specificity as Bcd, Orthodenticle (Otd) [14,15]. In the hover fly Episyrphus, a long-germ dipteran where no bcd gene has yet been identified, a maternal posterior gradient of the Caudal homolog is essential for posterior development. In this species, anterior patterning involves two distinct factors: a first factor, required for transcriptional control of most anterior zygotic genes, which are under Bcd control in the fruit fly, and a second one, mostly mediating the translational down-regulation of Caudal [16]. None of these factors has yet been identified but it has been proposed that these two functions were taken over by a single gene with the emergence of bcd in related species [16]. In parallel to the evolutionary interest in Bcd, this simple system has recently become approachable to multi-scale quantitative biology in 4D, owing to the development of in situ cytochemistry and live cell imaging. In this minireview, we will discuss how these recent developments help to understand the function of the Bcd morphogen with particular emphasis on the questions concerning the establishment and precision of its concentration gradient, the precision of its transcriptional response and how this response could, in conjunction with other factors, lead to robust patterning along the axis. Precision of the Bicoid Concentration Gradient The early development of the Drosophila embryo takes place in a syncytium, a single cell in which nuclei divide rapidly thirteen times without cytokinesis. When the Bcd protein gradient was first discovered, the syncytial nature of the embryo provided an obvious cellular context to explain how the gradient could be formed [3]. The bcd mRNA was shown to be transcribed maternally during oogenesis and anchored at the anterior pole of the oocyte [17] through a process involving its 30 untranslated region (UTR) and the maternal proteins, Exuperentia, Swallow and Staufen [18]. Translation of Bcd was shown to be activated upon egglaying [3] but, unlike its mRNA, the protein was presumably free to diffuse along the anterior-posterior axis (Figure 1A,B). The establishment of the Bcd gradient was proposed to occur by passive diffusion of the protein in the syncytium away from its localized source [3] according to the ‘Synthesis-Diffusion-Degradation’ model [19]. This model assumes that the morphogen is synthesized at a constant rate at the anterior pole, that it freely diffuses along the AP axis and that it is uniformly degraded across the embryo. The first doubts concerning the formation of the Bcd gradient were raised in 2002, with attempts to precisely quantify Bcd concentration along the anterior-posterior axis [20]. The degree of fluctuation in Bcd concentration was compared among several embryos, in order to understand how, despite naturally occurring variation in the overall length of
Current Biology Vol 20 No 5 R250
Box 1 Morphogen gradients. Morphogen gradients are used by various organisms to establish polarity along embryonic axes or within organ systems. In these systems, positional information is provided by the concentration of the morphogen which is detected by each cell in the target tissue and which gives rise to differential expression of the various target genes responsible for the determination of cell identities. Although the critical role of morphogens in axial patterning is now well established, important mechanistic questions concerning their mode of action remain unanswered: how do morphogens become gradually distributed in a field of cells; how are they able to differentially control the expression of different target genes at different threshold levels; and finally to what extent do they contribute, directly or through the activity of their target genes products, to the precision of patterning. The Drosophila transcription factor Bicoid was the first morphogen identified and it has since been used as a model to address these questions.
the embryos, simple diffusion of Bcd could generate reproducibly positioned target gene expression patterns [20]. Using fluorescent immuno-detection, it was observed that the Bcd gradient was rather variable among embryos [20]. This variability stood in contrast to the response to the gradient which appeared to be scaled with the embryo size, when deduced from expression of Bcd’s principal target gene hb. It was proposed that the variability of the Bcd gradient must be filtered to produce such a precise output [20]. As the size of the Hb expression domain was more variable in genetic backgrounds mutant for staufen which did not apparently modify the variability of the Bcd gradient, Staufen was proposed to be part of this filter [20]. However, the variability of the Bcd gradient was subsequently questioned with the use of synthetic reporters and gradients of Gal4-derived transcription factors artificially expressed in the embryo (Figure 2) [21]. These artificial transcription factor gradients produced a precise transcriptional response, indicating that such a precise read-out could be obtained from anterior gradients of bona fide transcription factors and that the precision of the Bcd response only requires Bcd’s ability to activate transcription [21]. The expression of these reporters was also less precise in the staufen mutant background [21], indicating that the putative filter [20], transforming the fluctuating Bcd input into a precise output, was acting upstream or at the level of the transcription process [21]. Finally, recent quantifications using fluorescently-tagged Bcd avoiding immuno-detection showed that the Bcd gradient profiles were in fact highly reproducible among embryos and scaled with embryo size, with a variability of only 2–3% along the egg length [22]. The precision of the Bcd gradient was subsequently confirmed using immuno-detection and careful normalization procedure to quantify Bcd [22,23]. The variability of the Bcd gradient was shown to be higher in the staufen mutant background, indicating that the variability of the response in this genetic background was a direct consequence of the variability of the gradient [23]. Altogether, these recent studies clearly indicate that the Bcd gradient is precisely established in no more than 90 minutes and that it scales with embryo length. Interestingly, longer embryos express slightly higher levels of Bcd than shorter embryos [23]. The molecular mechanism allowing the precise control of Bcd concentration levels in correlation with embryo size is likely to be established maternally [24] but has yet to be elucidated. Establishment of the Bicoid Gradient Given the syncytial nature of the Drosophila embryo, the simplest and most widely considered model for the
establishment of the Bcd gradient is by far the simple diffusion or Synthesis-Diffusion-Degradation model relying on free diffusion of the protein from the anterior pole. Given the stable establishment of the gradient in 90 minutes observed with the fluorescently-tagged versions of Bcd [22] and the average exponential decay length of the gradient along the axis, a simple prediction from the Synthesis-Diffusion-Degradation model is that the Bcd protein should diffuse throughout the embryo with a diffusion constant larger than 2 mm2/s. Fluorescently-tagged gradients of Bcd [22] offer the possibility to directly observe their formation in live embryos and measure the mobility of the protein to directly test the Synthesis-Diffusion-Degradation model. These observations showed that Bcd concentration is remarkably constant in interphase nuclei, at a given position along the anterior-posterior axis from nuclear cycle 10 to 14 [22]. This suggests that the gradient is steadily established in the cytoplasm and that a precise nucleo-cytoplasmic ratio of Bcd is recovered after each mitosis. The identification of the molecular effectors of this process will allow the determination of whether it is resulting from protein retention in nuclear traps [25] or from a tightly regulated nuclear import of Bcd, which is acting both as a transcription factor in the nucleus and as a translational regulator in cytoplasm. Measurements of Bcd mobility using fluorescence recovery after photobleaching (FRAP) returned a value for the diffusion coefficient of Bcd that was one to two orders of magnitude too low to explain the rapid establishment of the gradient by simple diffusion [22]. As this observation is quite surprising, it is important to keep in mind several reasons arguing that the Bcd protein could move faster than observed by FRAP: first, inert fluorescent molecules of a similar size as Bcd are much more mobile than Bcd in the embryo [26]; second, the FRAP experiments [22] provided an estimate of the diffusion coefficient at the scale of the size of the bleached volume (w1 mm). They did not directly address the question of the Bcd diffusion at the scale of the whole embryo and it is also possible that Bcd moves faster in the center of the egg, where the fluorescent proteins are more difficult to visualize than in the cortical cytoplasm; third, FRAP requires that the photobleaching pulse is short enough to ensure that the movements of molecules during the pulse are negligible. As technically there is a lower limit for reducing the duration of the pulse, FRAP is not always adapted for fast moving molecules [27]. Therefore, before abandoning the simple diffusion model to explain the formation of the Bcd gradient, it is essential to confirm the observed slow diffusion of Bcd using alternative quantitative approaches of live-cell imaging, such as those provided by engineered photoactivatable proteins
Review R251
A
Bcd 3’UTR bcd bcd RNA
B Bcd concentration
Figure 1. The Bicoid morphogen in the Drosophila blastoderm embryo. (A) The maternal bcd RNA, detected by in situ hybridization, is anchored at the anterior pole through its 30 UTR. (B) Translation of the Bcd protein is induced upon egg laying and this gives rise to a concentration gradient revealed by immuno-detection. The Bcd protein contains a homeodomain (HD) allowing its binding to nucleic acids and several domains involved in transcriptional activation or repression (HP, ST, Q, A and C). (C–E) The Bcd morphogen activates directly the expression of its target genes, orthodenticle (otd), hunchback (hb), and kru¨ppel (kr) in distinct domains, revealed by in situ hybridization. In contrast to the shallow gradient of the Bcd protein, target gene expression domains are limited by sharp and precisely positioned posterior borders. According to the French Flag model, each target gene is expressed for concentrations of Bcd (dotted line) above a specific threshold. Figure adapted from [21].
N
HP
HD ST
Q A
C
C
Bcd protein
C
D
otd RNA
E
hb RNA
Kr RNA
A 100
P Anterior-posterior axis of the embryo (% of egg length)
0 Current Biology
[28] or by fluorescent correlation spectroscopy, more appropriate than FRAP to detect rapidly moving proteins and recently used to analyze the diffusion of the Fgf8 morphogen gradient in zebrafish embryos [29]. If true, the surprising low mobility of the Bcd protein in the embryo is not compatible with the stable establishment of the gradient in 90 minutes by simple diffusion and raises the problem of understanding how the gradient could be formed. It has since inspired a number of models proposing alternative mechanisms to simple diffusion of the Bcd protein, including active [22] or advective [30] transport of the Bicoid protein, or an underlying mRNA gradient [31]. The latter of these studies [31] revisited the dynamics and redistribution of the bcd mRNA in the cortex, which in late blastoderm embryos resemble the Bcd protein gradient, and proposed that this bcd mRNA gradient could be involved in the establishment of the Bcd protein gradient [31]. Although these observations certainly call for a more quantitative analysis at early stages, they raise the question of understanding when and how this bcd mRNA gradient is established and also leave open the possibility that it could contribute to the steady maintenance of the Bcd protein gradient from cycle 10 to cycle 14. Obviously, as none of these models has been tested yet, the question of how exactly the Bcd gradient is established remains largely open. Precision of the Bicoid Response Concerning the transcriptional response to the Bcd gradient, two opposite models are recurrently proposed. On the one hand, the rapid establishment of a precise Bcd concentration gradient [22] raises the question of to what extent the Bcd protein itself could directly produce a precise response. On the other hand, it is proposed that the initial response to the Bcd gradient is not precise and that subsequent refinement by feedback involving the Bcd target proteins themselves is required for this precision [32,33]. The most intriguing feature of the Bcd response is that, while the Bcd protein is distributed as a smooth gradient (Figure 1B), expression domains of its target genes display fairly sharp
posterior borders at nuclear cycle 14 (Figure 1C–E). These sharp borders are also observed for a simple Bcd reporter gene only containing a minimal promoter with three Bcd DNA-binding sites and for a Gal4 reporter expressed from an artificially produced Gal4-derived gradient (Figure 2) [21]. Assuming that these reporters correctly recapitulate endogenous transcription, the sharp borders of their expression domains indicate that the transcriptional machinery does not respond gradually to the Bcd gradient but instead almost in an ON/OFF fashion, with clearly different responses despite only small differences of Bcd concentration. As the Bcd target genes contain in general more than a single binding site for Bcd, it has been proposed that the formation of these sharp borders results from a cooperative binding of Bcd molecules to DNA [21,34,35]. Given the sharp border of the synthetic lacZ reporter (Figure 2C), three strong binding sites for Bcd are sufficient for this sharp response. It should be noted that the natural boundary of the hb target gene (which is more posterior than the border of the lacZ reporters) is sharp at a lower concentration of Bcd. Therefore, the period required for the Bcd protein to find its DNA binding site is likely larger at these lower concentrations of Bcd. Thus, the sharp response observed for hb (Figure 1D) requires either a longer period to be established [34] and/or additional mechanisms involving posterior repressors [33]. The recent measurements of Bcd’s mobility [22] allowed the estimation of the period required for nearby nuclei to express different levels of the Hb protein, using a theoretical model based on statistical mechanics assuming that the limiting step of this process was the random arrival of Bcd molecules at their DNA-binding sites [34]. Given the slow mobility of Bcd, the period required for the system to respond accurately was too long and not compatible with the rapid development of the early fruit fly embryo. Therefore, it was proposed that the initial transcriptional response to Bcd is noisy and that the Bcd response is subsequently refined to produce sharp borders of gene expression by specific mechanisms, which do not directly involve Bcd [34]. This study has since inspired models explaining how
Current Biology Vol 20 No 5 R252
A
D
Gal4-DB GCN4-Act
lacZ
Bcd-3’ UTR
lacZ
Bcd BS
UAS
B
E
Anchored bcd mRNA
Artificially anchored Gal4-3GCN4 encoding mRNA lacZ
lacZ
UAS
Bcd BS
Artificial Gal4-3GCN4 anterior-posterior gradient
Bcd anterior-posterior gradient
C
F
lacZ
lacZ Current Biology
a noisy initial response to Bcd can be refined as the embryo develops, including feedback and cross-regulation by the proteins encoded by the Bcd target genes [33] or slow mobility of these target proteins in the embryo [36]. However, neither autoregulation nor the involvement of other transcription factors are plausible explanations for the sharp expression borders (Figure 2) of the synthetic reporters, as these only contain binding sites for Bcd or Gal4 and no binding sites for other transcription factors [21]. More importantly, the notion that the initial transcriptional response to Bcd was noisy has emerged — assuming that the Bcd transcriptional response was accurately measured by target protein expression [33,34] — from a theoretical model based on the slow mobility of the protein determined by FRAP [22]. Given that the response to Bcd has only been analyzed in cycle 14 embryos, i.e. much after the initiation of zygotic transcription, a direct analysis of the Bcd transcriptional response at earlier cycles is lacking and will certainly shed light on this question. Positional Information Downstream of the Bicoid Gradient Even though increased bcd gene dosage in females shifts the entire pattern of the embryo towards the posterior [4], the question of whether the Bcd gradient is acting as a morphogen is often debated. Importantly, the shift of the expression domains of Bcd targets induced by increased bcd dosage are coordinated and occur without interfering with their order along the anterior-posterior axis [4,6]. This indicates that the Bcd gradient provides positional information along the axis in a dose-dependent manner and efforts have been made to understand how this could be achieved. As the Bcd protein encodes a DNA-binding transcription factor, it was initially proposed that the thresholds of Bcd concentration required for the expression of its target genes
Figure 2. The precision and sharp borders of Bcd target gene expression domains. Bcd transcriptional activity is analyzed using synthetic reporters and artificial gradients of transcription factors. (A) A bcd3-lacZ transgene places lacZ under the control of a naive promoter containing three binding sites for Bcd (Bcd BS). (B) In embryos carrying the transgene, the Bcd protein binds to the three Bcd binding sites and this leads to activation of lacZ expression in an anterior domain detected by in situ hybridization. (C) The position of the posterior border of the lacZ expression domain is sharp and precisely positioned relative to the embryo length. (D) A transgenic fly carrying both the UAS-lacZ transgene, placing lacZ under the control of a naive promoter containing UAS binding sites for Gal4, and a transgene allowing maternal expression and anterior anchoring of mRNAs encoding the Gal4-3GCN4 transcriptional activator. (E) This second transgene allows expression, in a Bcd-like maternal gradient, of the Gal4 DNA-binding domain fused to three copies of the GCN4 activation domain. The binding of Gal4-3GCN4 to the UAS binding sites allows expression of lacZ. F: lacZ expression activated by Gal4-3GCN4 is detected by in situ hybridization. LacZ expression domain harbors a precisely positioned and a sharp posterior border (F). Data in (C) and (F) were taken from [21].
depend on the number and on the affinity of Bcd binding sites found in their regulatory regions [5,6]. However, it is now clear that other elements in target gene promoters and the integration of positive and negative transcriptional inputs from proteins bound to these elements are major determinants for the interpretation of positional information along the anterior-posterior axis [37]. Proteins binding with Bcd to regulatory regions of Bcd target genes include Caudal, a major translational target of Bcd (Figure 3A), as well as the gap-proteins Hb, Kru¨ppel, Knirps, Giant and their downstream targets [37,38]. Most of the proteins binding with Bcd in the regulatory regions of the Bcd target genes are themselves Bcd target proteins. Therefore, an increase in bcd gene dosage in females will not only increase Bcd concentration in the embryo but it will also modify the expression domains and the levels of expression of these Bcd target proteins and this will contribute to coordinate the shift of expression of most Bcd target genes along the axis. Such a downstream cascade could therefore in part explain how the system is able to respond by a coordinate shift of the entire gene expression pattern without interfering with their order along the axis. Anterior Positional Information Independent of Bicoid The coordinated shift of expression of the Bcd targets observed upon an increase in bcd dosage is more difficult to understand if some of the proteins binding with Bcd in regulatory regions of its target genes are not themselves Bcd targets. If a protein is not itself a Bcd target, its expression domain and its level of expression will not be modified upon an increase in bcd gene dosage. Therefore, if this protein is binding in the regulatory regions of some of the Bcd target genes, and contributes differently to the expression of two given target genes, it remains possible that an increase in the bcd gene dosage without a change in the
Relative concentration/activity (arbitrary units)
Review R253
F C
B D E
A
Anterior Bicoid Caudal
Posterior Nanos Maternal Hunchback
Torso Repressors downregulated by Torso (Cic, Gr) Current Biology
Figure 3. Maternal contributions to anterior-posterior patterning. Expression or activity of proteins represented by dashed lines is under control of proteins represented by plain lines. In red, the Bicoid protein is expressed as a concentration gradient along the anterior-posterior axis of the embryo [22] and it inhibits the translation of the Caudal protein in the anterior half, creating the concentration gradient of Caudal (A) [8,9]. In blue, the Nanos protein, expressed as a posterior to anterior gradient [48], represses translation of maternal Hunchback, allowing its expression as an anterior-posterior gradient [49] (B). In green, activation of the Torso receptor at the anterior pole represses expression of the Bcd target genes at the anterior pole [39,50] (C). The activation potential of Bcd is enhanced by its phosphorylation by the Torso pathway [43] (D) and/or by the inhibition of ubiquitous repressor [45] (E) such as Capicua (Cic) or Groucho (Gr), which are repressed by Torso activation at the poles [46,47] (F).
expression of this particular partner of Bcd will lead to the inversion of the expression domain of these two genes along the axis. In this regard, two systems established independently of Bcd and providing anterior information have been described (Figure 3). The first system involves the maternal contribution of the Hb protein, which is expressed in an anterior-posterior gradient and repressed in the posterior by the major posterior determinant and translational repressor Nanos (Figure 3B) [13]. Most Bcd targets are also positively regulated by maternal Hb and the presence of Hb-binding sites in a Bcd-dependent promoter induces a posterior shift of its expression domain [7]. This synergy between maternal Hb and Bcd is important for anterior patterning but the fact that it gives rise to the zygotic expression of hb itself has obscured the role of maternal Hb in anterior patterning [7]. The second system providing anterior information independently of Bcd is responsible for the formation of the terminal structures of the embryo and is determined by the signal transduction cascade acting downstream of the Torso receptor tyrosine kinase. The interaction between Bcd and Torso, revealed by the expression of the Bcd targets, appears to be both antagonistic and synergic. First, activation of Torso represses expression of most Bcd targets at the anterior pole (Figure 3C) [39]. The repression of hb, induced by Torso at the anterior pole, has been shown to mediate most terminal patterning of the non-segmented part of the head [40]. Second, several studies have shown that Torso activation induces a small shift of the expression domains of Bcd targets towards the posterior [41–43]. This synergy likely explains the partial rescue of the torso anterior phenotype by increasing amounts of Bcd [44] and the mirror
image expression of anterior Bcd targets in embryos expressing an almost flat gradient of Bcd [45]. At the molecular level, the positive interaction between Bcd and Torso is only partially dependent on phosphorylation of Bcd by the MAPkinase, the nuclear effector of Torso (Figure 3D) [43]. As proposed recently [45], the synergy between Torso and Bcd could also involve local posterior to anterior gradients of ubiquitous maternal repressors (Figure 3E), such as those known to be repressed by Torso activation at the poles (Figure 3F) [46,47]. Whether this occurs by a direct inhibition of Bcd activity or indirectly through the binding to DNA of transcription factors regulated by these repressors has to be clarified. Outlook Quantitative approaches will allow determination of the exact contribution of maternal Hb and the Torso cascade to the Bcd thresholds required for target gene expression. Importantly, the uneven distribution and activity of these sources of positional information (Figure 3), with their highest concentration/activity at the anterior pole gradually decreasing towards the posterior pole, can explain why they do not interfere with the order of expression of the Bcd target genes along the anterior-posterior axis upon increases in bcd dosage, and why the Bcd threshold required for the expression of a given target gene could be different, when artificially positioned elsewhere along the axis, as for instance when increasing the bcd gene dosage [20] or when modifying the shape of the Bcd and maternal Hb gradients [45]. At the molecular, cellular and wholeembryo levels, the Bcd protein is not functioning alone but with many partners and this is not surprising. Is this a valid reason for demoting it from its qualification of morphogen? Bcd’s ability to provide positional information along the axis in a dose-dependent manner makes it unique and how this is achieved is really what we need to understand. Acknowledgements We thank Genevie`ve Almouzni, Baptiste Roelens, Vale´rie Borde and the anonymous reviewers for constructive suggestions and apologize to those whose important work could not be cited due to space constraints. Research in the Dostatni lab is supported by the CNRS and the Institut Curie. N.D. is supported by INSERM and anterior-posterior by the University Rene´ Diderot and the Institut Curie. References 1.
Frohnho¨fer, H.G., and Nu¨sslein-Volhard, C. (1986). Organization of anterior pattern in the Drosophila embryo by the maternal gene bicoid. Nature 324, 120–125.
2.
Driever, W., Siegel, V., and Nu¨sslein-Volhard, C. (1990). Autonomous determination of anterior structures in the early Drosophila embryo by the bicoid morphogen. Development 109, 811–820.
3.
Driever, W., and Nu¨sslein-Volhard, C. (1988). A gradient of bicoid protein in Drosophila embryos. Cell 54, 83–93.
4.
Driever, W., and Nu¨sslein-Volhard, C. (1988). The bicoid protein determines position in the Drosophila embryo in a concentration-dependent manner. Cell 54, 95–104.
5.
Driever, W., and Nu¨sslein-Volhard, C. (1989). The bicoid protein is a positive regulator of hunchback transcription in the early Drosophila embryo. Nature 337, 138–143.
6.
Struhl, G., Struhl, K., and Macdonald, P.M. (1989). The gradient morphogen bicoid is a concentration-dependent transcriptional activator. Cell 57, 1259– 1273.
7.
Simpson-Brose, M., Treisman, J., and Desplan, C. (1994). Synergy between the hunchback and bicoid morphogens is required for anterior patterning in Drosophila. Cell 78, 855–865.
8.
Dubnau, J., and Struhl, G. (1996). RNA recognition and translational regulation by a homeodomain protein. Nature 379, 694–699.
Current Biology Vol 20 No 5 R254
9.
Rivera-Pomar, R., Niessing, D., Schmidt-Ott, U., Gehring, W.J., and Jackle, H. (1996). RNA binding and translational suppression by bicoid. Nature 379, 746–749.
10.
Macdonald, P.M., and Struhl, G. (1986). A molecular gradient in early Drosophila embryos and its role in specifying the body pattern. Nature 324, 537–545.
38.
Rivera-Pomar, R., Lu, X., Perrimon, N., Taubert, H., and Jackle, H. (1995). Activation of posterior gap gene expression in the Drosophila blastoderm. Nature 376, 253–256.
39.
12.
Stauber, M., Jackle, H., and Schmidt-Ott, U. (1999). The anterior determinant bicoid of Drosophila is a derived Hox class 3 gene. Proc. Natl. Acad. Sci. USA 96, 3786–3789.
40.
13.
Irish, V., Lehmann, R., and Akam, M. (1989). The Drosophila posterior-group gene nanos functions by repressing hunchback activity. Nature 338, 646– 648.
41.
14.
Lynch, J.A., Brent, A.E., Leaf, D.S., Pultz, M.A., and Desplan, C. (2006). Localized maternal orthodenticle patterns anterior and posterior in the long germ wasp Nasonia. Nature 439, 728–732.
15.
Schroder, R. (2003). The genes orthodenticle and hunchback substitute for bicoid in the beetle Tribolium. Nature 422, 621–625.
16.
Lemke, S., and Schmidt-Ott, U. (2009). Evidence for a composite anterior determinant in the hover fly Episyrphus balteatus (Syrphidae), a cyclorrhaphan fly with an anterodorsal serosa anlage. Development 136, 117–127.
44.
17.
Berleth, T., Burri, M., Thoma, G., Bopp, D., Richstein, S., Frigerio, G., Noll, M., and Nu¨sslein-Volhard, C. (1988). The role of localization of bicoid RNA in organizing the anterior pattern of the drosophila embryo. EMBO J. 7, 1749–1756.
45.
18.
St. Johnston, D., Driever, W., Berleth, T., Richstein, S., and Nu¨sslein-Volhard, C. (1989). Multiple steps in the localization of bicoid RNA to the anterior pole of the Drosophila oocyte. Development 107, 13–19.
19.
Crick, F. (1970). Diffusion in embryogenesis. Nature 225, 420–422.
20.
Houchmandzadeh, B., Wieschaus, E., and Leibler, S. (2002). Establishment of developmental precision and proportions in the early Drosophila embryo. Nature 415, 798–802.
11.
21.
Crauk, O., and Dostatni, N. (2005). Bicoid determines sharp and precise target gene expression in the Drosophila embryo. Curr. Biol. 15, 1888–1898.
22.
Gregor, T., Wieschaus, E.F., McGregor, A.P., Bialek, W., and Tank, D.W. (2007). Stability and nuclear dynamics of the bicoid morphogen gradient. Cell 130, 141–152.
23.
He, F., Wen, Y., Deng, J., Lin, X., Lu, L.J., Jiao, R., and Ma, J. (2008). Probing intrinsic properties of a robust morphogen gradient in Drosophila. Dev. Cell 15, 558–567.
24.
Lott, S.E., Kreitman, M., Palsson, A., Alekseeva, E., and Ludwig, M.Z. (2007). Canalization of segmentation and its evolution in Drosophila. Proc. Natl. Acad. Sci. USA 104, 10926–10931.
25.
Coppey, M., Berezhkovskii, A.M., Kim, Y., Boettiger, A.N., and Shvartsman, S.Y. (2007). Modeling the bicoid gradient: Diffusion and reversible nuclear trapping of a stable protein. Dev. Biol. 312, 623–630.
26.
Gregor, T., Bialek, W., van Steveninck, R.R.d.R., Tank, D.W., and Wieschaus, E.F. (2005). Diffusion and scaling during early embryonic pattern formation. Proc. Natl. Acad. Sci. USA 102, 18403–18407.
27.
Weiss, M. (2004). Challenges and artifacts in quantitative photobleaching experiments. Traffic 5, 662–671.
28.
Patterson, G.H., and Lippincott-Schwartz, J. (2002). A photoactivatable GFP for selective photolabeling of proteins and cells. Science 297, 1873–1877.
29.
Yu, S.R., Burkhardt, M., Nowak, M., Ries, J., Petrasek, Z., Scholpp, S., Schwille, P., and Brand, M. (2009). Fgf8 morphogen gradient forms by a source-sink mechanism with freely diffusing molecules. Nature 461, 533–536.
30.
Hecht, I., Rappel, W.-J., and Levine, H. (2009). Determining the scale of the Bicoid morphogen gradient. Proc. Natl. Acad. Sci. USA 106, 1710–1715.
31.
Spirov, A., Fahmy, K., Schneider, M., Frei, E., Noll, M., and Baumgartner, S. (2009). Formation of the bicoid morphogen gradient: an mRNA gradient dictates the protein gradient. Development 136, 605–614.
32.
Bergmann, S., Sandler, O., Sberro, H., Shnider, S., Schejter, E., Shilo, B.Z., and Barkai, N. (2007). Pre-steady-state decoding of the Bicoid morphogen gradient. PLoS Biol. 5, e46.
33.
ManuSurkova, S., Spirov, A.V., Gursky, V.V., Janssens, H., Kim, A.-R., Radulescu, O., Vanario-Alonso, C.E., Sharp, D.H., Samsonova, M., et al. (2009). Canalization of gene expression in the Drosophila blastoderm by gap gene cross regulation. PLoS Biol. 7, e1000049.
34.
Gregor, T., Tank, D.W., Wieschaus, E.F., and Bialek, W. (2007). Probing the limits to positional information. Cell 130, 153–164.
35.
Lebrecht, D., Foehr, M., Smith, E., Lopes, F.J.P., Vanario-Alonso, C.E., Reinitz, J., Burz, D.S., and Hanes, S.D. (2005). Bicoid cooperative DNA binding is critical for embryonic patterning in Drosophila. Proc. Natl. Acad. Sci. USA 102, 13176–13181.
36.
Okabe-Oho, Y., Murakami, H., Oho, S., and Sasai, M. (2009). Stable, precise, and reproducible patterning of bicoid and hunchback molecules in the early Drosophila embryo. PLoS Comput. Biol. 5, e1000486.
37.
42.
43.
46.
47.
48. 49.
50.
Ochoa-Espinosa, A., Yucel, G., Kaplan, L., Pare, A., Pura, N., Oberstein, A., Papatsenko, D., and Small, S. (2005). The role of binding site cluster strength in Bicoid-dependent patterning in Drosophila. Proc. Natl. Acad. Sci. USA 102, 4960–4965. Segal, E., Raveh-Sadka, T., Schroeder, M., Unnerstall, U., and Gaul, U. (2008). Predicting expression patterns from regulatory sequence in Drosophila segmentation. Nature 451, 535–540. Ronchi, E., Treisman, J., Dostatni, N., Struhl, G., and Desplan, C. (1993). Down-regulation of the Drosophila morphogen Bicoid by the Torso receptor- mediated signal transduction cascade. Cell 74, 347–355. Janody, F., Reischl, J., and Dostatni, N. (2000). Persistence of Hunchback in the terminal region of the Drosophila blastoderm embryo impairs anterior development. Development 127, 1573–1582. Gao, Q., Wang, Y., and Finkelstein, R. (1996). Orthodenticle regulation during embryonic head development in Drosophila. Mech. Dev. 56, 3–15. Grossniklaus, U., Cadigan, K.M., and Gehring, W.J. (1994). Three maternal coordinate systems cooperate in the patterning of the Drosophila head. Development 120, 3155–3171. Janody, F., Sturny, R., Catala, F., Desplan, C., and Dostatni, N. (2000). Phosphorylation of bicoid on MAP-kinase sites: contribution to its interaction with the torso pathway. Development 127, 279–289. Schaeffer, V., Killian, D., Desplan, C., and Wimmer, E.A. (2000). High Bicoid levels render the terminal system dispensable for Drosophila head development. Development 127, 3993–3999. Ochoa-Espinosa, A., Yu, D., Tsirigos, A., Struffi, P., and Small, S. (2009). Anterior-posterior positional information in the absence of a strong Bicoid gradient. Proc. Natl. Acad. Sci. USA 106, 3823–3828. Jimenez, G., Guichet, A., Ephrussi, A., and Casanova, J. (2000). Relief of gene repression by Torso RTK signaling: role of capicua in Drosophila terminal and dorsoventral patterning. Genes Dev. 14, 224–231. Paroush, Z., Wainwright, S.M., and Ish-Horowicz, D. (1997). Torso signalling regulates terminal patterning in Drosophila by antagonising Groucho-mediated repression. Development 124, 3827–3834. Wang, C., Dickinson, L.K., and Lehmann, R. (1994). Genetics of nanos localization in Drosophila. Dev. Dynamics 199, 103–115. Cho, P.F., Gamberi, C., Cho-Park, Yoon A., Cho-Park, I.B., Lasko, P., and Sonenberg, N. (2006). Cap-dependent translational inhibition establishes two opposing morphogen gradients in Drosophila embryos. Curr. Biol. 16, 2035–2041. Janody, F., Sturny, R., Schaeffer, V., Azou, Y., and Dostatni, N. (2001). Two distinct domains of Bicoid mediate its transcriptional downregulation by the Torso pathway. Development 128, 2281–2290.
The Bicoid Morphogen System Aude Porcher and Nathalie Dostatni*
Several fundamental concepts of developmental biology have emerged from studies on the early development of the Drosophila melanogaster embryo. In the late 1980s, studies on Bicoid provided the first solid experimental evidence for the existence of morphogenetic gradients and their implication in axial patterning. Bicoid has since stimulated further research, bringing together developmental and cell biologists, physicists and theoreticians to address fundamental biological questions. These include mechanistic aspects of transcriptional and translational control, molecular and functional aspects of evolution and, more recently with the development of quantitative approaches, the robustness of axial patterning in a systems biology view. However, recent studies provide data which lead to contradictory interpretations. Here, we discuss these recent observations, highlighting the data helping to understand how anterior patterning is achieved under the control of Bicoid and point to novel challenges for future studies. Introduction The bicoid (bcd) gene was initially identified in a screen for maternal genes involved in the development of the anterior-posterior axis of the Drosophila embryo. As embryos from bcd mutant females did not develop anterior structures [1] and as injections of the bcd RNA were able to induce the development of anterior structures at ectopic sites [2], bcd appeared both necessary and sufficient for anterior patterning. The Bicoid (Bcd) protein was shown to be expressed in a concentration gradient with the highest Bcd concentration at the anterior end of the embryo and decreasing towards the posterior [3]. Moreover, Bcd is able to determine distinct cell fates in a dose-dependent manner [4]. Bcd thus represented the first solid instance of a morphogenetic gradient essential for the development of a polarized axis in an organism (Box 1). At the molecular level, the Bcd protein was shown to possess two very different activities, both contributing to anterior-posterior patterning. First, Bcd is a homeo-domain containing transcription factor able to activate [5,6], in synergy with the zinc finger protein Hunchback (Hb) [7], many target genes essential for anterior patterning. Second, Bcd was also shown to bind RNA through its homeo-domain and to down-regulate the translation of maternal Caudal [8,9], a protein involved in the transcriptional control of posterior development [10,11]. Despite its key role in anterior patterning in the fruit fly, the Bcd protein is surprisingly poorly conserved during evolution. Search for sequence conservation indicates that the bcd gene is a recent acquisition in long-germ dipterans and arises from a duplication of the Hox3 homologue zerknu¨llt [12]. This lack of evolutionary conservation is particularly striking, when compared to the other maternal determinants, Hb or Caudal, which are dispensable for anteriorposterior patterning in the fruit fly but more conserved than
Institut Curie, Centre de Recherche, and CNRS, UMR218, Paris, F-75248 France. *E-mail: [email protected]
DOI 10.1016/j.cub.2010.01.026
Minireview
Bcd through evolution [10,13]. Given the transcriptional synergy between maternal Hb and Bcd, it was first proposed that Hb was the ancestral anterior morphogen progressively replaced by the newly acquired Bcd in long-germ dipterans [7]. Recently, the search for anterior-posterior patterning factors in related species, with no bcd-like sequences but in which hunchback is conserved, shed light on how the bcd gene could have emerged during evolution. In the flour beetle Tribolium and in the long-germ wasp Nasonia, anterior patterning was shown to be controlled by Hb and another homeo-domain containing protein with the same DNA-binding specificity as Bcd, Orthodenticle (Otd) [14,15]. In the hover fly Episyrphus, a long-germ dipteran where no bcd gene has yet been identified, a maternal posterior gradient of the Caudal homolog is essential for posterior development. In this species, anterior patterning involves two distinct factors: a first factor, required for transcriptional control of most anterior zygotic genes, which are under Bcd control in the fruit fly, and a second one, mostly mediating the translational down-regulation of Caudal [16]. None of these factors has yet been identified but it has been proposed that these two functions were taken over by a single gene with the emergence of bcd in related species [16]. In parallel to the evolutionary interest in Bcd, this simple system has recently become approachable to multi-scale quantitative biology in 4D, owing to the development of in situ cytochemistry and live cell imaging. In this minireview, we will discuss how these recent developments help to understand the function of the Bcd morphogen with particular emphasis on the questions concerning the establishment and precision of its concentration gradient, the precision of its transcriptional response and how this response could, in conjunction with other factors, lead to robust patterning along the axis. Precision of the Bicoid Concentration Gradient The early development of the Drosophila embryo takes place in a syncytium, a single cell in which nuclei divide rapidly thirteen times without cytokinesis. When the Bcd protein gradient was first discovered, the syncytial nature of the embryo provided an obvious cellular context to explain how the gradient could be formed [3]. The bcd mRNA was shown to be transcribed maternally during oogenesis and anchored at the anterior pole of the oocyte [17] through a process involving its 30 untranslated region (UTR) and the maternal proteins, Exuperentia, Swallow and Staufen [18]. Translation of Bcd was shown to be activated upon egglaying [3] but, unlike its mRNA, the protein was presumably free to diffuse along the anterior-posterior axis (Figure 1A,B). The establishment of the Bcd gradient was proposed to occur by passive diffusion of the protein in the syncytium away from its localized source [3] according to the ‘Synthesis-Diffusion-Degradation’ model [19]. This model assumes that the morphogen is synthesized at a constant rate at the anterior pole, that it freely diffuses along the AP axis and that it is uniformly degraded across the embryo. The first doubts concerning the formation of the Bcd gradient were raised in 2002, with attempts to precisely quantify Bcd concentration along the anterior-posterior axis [20]. The degree of fluctuation in Bcd concentration was compared among several embryos, in order to understand how, despite naturally occurring variation in the overall length of
Current Biology Vol 20 No 5 R250
Box 1 Morphogen gradients. Morphogen gradients are used by various organisms to establish polarity along embryonic axes or within organ systems. In these systems, positional information is provided by the concentration of the morphogen which is detected by each cell in the target tissue and which gives rise to differential expression of the various target genes responsible for the determination of cell identities. Although the critical role of morphogens in axial patterning is now well established, important mechanistic questions concerning their mode of action remain unanswered: how do morphogens become gradually distributed in a field of cells; how are they able to differentially control the expression of different target genes at different threshold levels; and finally to what extent do they contribute, directly or through the activity of their target genes products, to the precision of patterning. The Drosophila transcription factor Bicoid was the first morphogen identified and it has since been used as a model to address these questions.
the embryos, simple diffusion of Bcd could generate reproducibly positioned target gene expression patterns [20]. Using fluorescent immuno-detection, it was observed that the Bcd gradient was rather variable among embryos [20]. This variability stood in contrast to the response to the gradient which appeared to be scaled with the embryo size, when deduced from expression of Bcd’s principal target gene hb. It was proposed that the variability of the Bcd gradient must be filtered to produce such a precise output [20]. As the size of the Hb expression domain was more variable in genetic backgrounds mutant for staufen which did not apparently modify the variability of the Bcd gradient, Staufen was proposed to be part of this filter [20]. However, the variability of the Bcd gradient was subsequently questioned with the use of synthetic reporters and gradients of Gal4-derived transcription factors artificially expressed in the embryo (Figure 2) [21]. These artificial transcription factor gradients produced a precise transcriptional response, indicating that such a precise read-out could be obtained from anterior gradients of bona fide transcription factors and that the precision of the Bcd response only requires Bcd’s ability to activate transcription [21]. The expression of these reporters was also less precise in the staufen mutant background [21], indicating that the putative filter [20], transforming the fluctuating Bcd input into a precise output, was acting upstream or at the level of the transcription process [21]. Finally, recent quantifications using fluorescently-tagged Bcd avoiding immuno-detection showed that the Bcd gradient profiles were in fact highly reproducible among embryos and scaled with embryo size, with a variability of only 2–3% along the egg length [22]. The precision of the Bcd gradient was subsequently confirmed using immuno-detection and careful normalization procedure to quantify Bcd [22,23]. The variability of the Bcd gradient was shown to be higher in the staufen mutant background, indicating that the variability of the response in this genetic background was a direct consequence of the variability of the gradient [23]. Altogether, these recent studies clearly indicate that the Bcd gradient is precisely established in no more than 90 minutes and that it scales with embryo length. Interestingly, longer embryos express slightly higher levels of Bcd than shorter embryos [23]. The molecular mechanism allowing the precise control of Bcd concentration levels in correlation with embryo size is likely to be established maternally [24] but has yet to be elucidated. Establishment of the Bicoid Gradient Given the syncytial nature of the Drosophila embryo, the simplest and most widely considered model for the
establishment of the Bcd gradient is by far the simple diffusion or Synthesis-Diffusion-Degradation model relying on free diffusion of the protein from the anterior pole. Given the stable establishment of the gradient in 90 minutes observed with the fluorescently-tagged versions of Bcd [22] and the average exponential decay length of the gradient along the axis, a simple prediction from the Synthesis-Diffusion-Degradation model is that the Bcd protein should diffuse throughout the embryo with a diffusion constant larger than 2 mm2/s. Fluorescently-tagged gradients of Bcd [22] offer the possibility to directly observe their formation in live embryos and measure the mobility of the protein to directly test the Synthesis-Diffusion-Degradation model. These observations showed that Bcd concentration is remarkably constant in interphase nuclei, at a given position along the anterior-posterior axis from nuclear cycle 10 to 14 [22]. This suggests that the gradient is steadily established in the cytoplasm and that a precise nucleo-cytoplasmic ratio of Bcd is recovered after each mitosis. The identification of the molecular effectors of this process will allow the determination of whether it is resulting from protein retention in nuclear traps [25] or from a tightly regulated nuclear import of Bcd, which is acting both as a transcription factor in the nucleus and as a translational regulator in cytoplasm. Measurements of Bcd mobility using fluorescence recovery after photobleaching (FRAP) returned a value for the diffusion coefficient of Bcd that was one to two orders of magnitude too low to explain the rapid establishment of the gradient by simple diffusion [22]. As this observation is quite surprising, it is important to keep in mind several reasons arguing that the Bcd protein could move faster than observed by FRAP: first, inert fluorescent molecules of a similar size as Bcd are much more mobile than Bcd in the embryo [26]; second, the FRAP experiments [22] provided an estimate of the diffusion coefficient at the scale of the size of the bleached volume (w1 mm). They did not directly address the question of the Bcd diffusion at the scale of the whole embryo and it is also possible that Bcd moves faster in the center of the egg, where the fluorescent proteins are more difficult to visualize than in the cortical cytoplasm; third, FRAP requires that the photobleaching pulse is short enough to ensure that the movements of molecules during the pulse are negligible. As technically there is a lower limit for reducing the duration of the pulse, FRAP is not always adapted for fast moving molecules [27]. Therefore, before abandoning the simple diffusion model to explain the formation of the Bcd gradient, it is essential to confirm the observed slow diffusion of Bcd using alternative quantitative approaches of live-cell imaging, such as those provided by engineered photoactivatable proteins
Review R251
A
Bcd 3’UTR bcd bcd RNA
B Bcd concentration
Figure 1. The Bicoid morphogen in the Drosophila blastoderm embryo. (A) The maternal bcd RNA, detected by in situ hybridization, is anchored at the anterior pole through its 30 UTR. (B) Translation of the Bcd protein is induced upon egg laying and this gives rise to a concentration gradient revealed by immuno-detection. The Bcd protein contains a homeodomain (HD) allowing its binding to nucleic acids and several domains involved in transcriptional activation or repression (HP, ST, Q, A and C). (C–E) The Bcd morphogen activates directly the expression of its target genes, orthodenticle (otd), hunchback (hb), and kru¨ppel (kr) in distinct domains, revealed by in situ hybridization. In contrast to the shallow gradient of the Bcd protein, target gene expression domains are limited by sharp and precisely positioned posterior borders. According to the French Flag model, each target gene is expressed for concentrations of Bcd (dotted line) above a specific threshold. Figure adapted from [21].
N
HP
HD ST
Q A
C
C
Bcd protein
C
D
otd RNA
E
hb RNA
Kr RNA
A 100
P Anterior-posterior axis of the embryo (% of egg length)
0 Current Biology
[28] or by fluorescent correlation spectroscopy, more appropriate than FRAP to detect rapidly moving proteins and recently used to analyze the diffusion of the Fgf8 morphogen gradient in zebrafish embryos [29]. If true, the surprising low mobility of the Bcd protein in the embryo is not compatible with the stable establishment of the gradient in 90 minutes by simple diffusion and raises the problem of understanding how the gradient could be formed. It has since inspired a number of models proposing alternative mechanisms to simple diffusion of the Bcd protein, including active [22] or advective [30] transport of the Bicoid protein, or an underlying mRNA gradient [31]. The latter of these studies [31] revisited the dynamics and redistribution of the bcd mRNA in the cortex, which in late blastoderm embryos resemble the Bcd protein gradient, and proposed that this bcd mRNA gradient could be involved in the establishment of the Bcd protein gradient [31]. Although these observations certainly call for a more quantitative analysis at early stages, they raise the question of understanding when and how this bcd mRNA gradient is established and also leave open the possibility that it could contribute to the steady maintenance of the Bcd protein gradient from cycle 10 to cycle 14. Obviously, as none of these models has been tested yet, the question of how exactly the Bcd gradient is established remains largely open. Precision of the Bicoid Response Concerning the transcriptional response to the Bcd gradient, two opposite models are recurrently proposed. On the one hand, the rapid establishment of a precise Bcd concentration gradient [22] raises the question of to what extent the Bcd protein itself could directly produce a precise response. On the other hand, it is proposed that the initial response to the Bcd gradient is not precise and that subsequent refinement by feedback involving the Bcd target proteins themselves is required for this precision [32,33]. The most intriguing feature of the Bcd response is that, while the Bcd protein is distributed as a smooth gradient (Figure 1B), expression domains of its target genes display fairly sharp
posterior borders at nuclear cycle 14 (Figure 1C–E). These sharp borders are also observed for a simple Bcd reporter gene only containing a minimal promoter with three Bcd DNA-binding sites and for a Gal4 reporter expressed from an artificially produced Gal4-derived gradient (Figure 2) [21]. Assuming that these reporters correctly recapitulate endogenous transcription, the sharp borders of their expression domains indicate that the transcriptional machinery does not respond gradually to the Bcd gradient but instead almost in an ON/OFF fashion, with clearly different responses despite only small differences of Bcd concentration. As the Bcd target genes contain in general more than a single binding site for Bcd, it has been proposed that the formation of these sharp borders results from a cooperative binding of Bcd molecules to DNA [21,34,35]. Given the sharp border of the synthetic lacZ reporter (Figure 2C), three strong binding sites for Bcd are sufficient for this sharp response. It should be noted that the natural boundary of the hb target gene (which is more posterior than the border of the lacZ reporters) is sharp at a lower concentration of Bcd. Therefore, the period required for the Bcd protein to find its DNA binding site is likely larger at these lower concentrations of Bcd. Thus, the sharp response observed for hb (Figure 1D) requires either a longer period to be established [34] and/or additional mechanisms involving posterior repressors [33]. The recent measurements of Bcd’s mobility [22] allowed the estimation of the period required for nearby nuclei to express different levels of the Hb protein, using a theoretical model based on statistical mechanics assuming that the limiting step of this process was the random arrival of Bcd molecules at their DNA-binding sites [34]. Given the slow mobility of Bcd, the period required for the system to respond accurately was too long and not compatible with the rapid development of the early fruit fly embryo. Therefore, it was proposed that the initial transcriptional response to Bcd is noisy and that the Bcd response is subsequently refined to produce sharp borders of gene expression by specific mechanisms, which do not directly involve Bcd [34]. This study has since inspired models explaining how
Current Biology Vol 20 No 5 R252
A
D
Gal4-DB GCN4-Act
lacZ
Bcd-3’ UTR
lacZ
Bcd BS
UAS
B
E
Anchored bcd mRNA
Artificially anchored Gal4-3GCN4 encoding mRNA lacZ
lacZ
UAS
Bcd BS
Artificial Gal4-3GCN4 anterior-posterior gradient
Bcd anterior-posterior gradient
C
F
lacZ
lacZ Current Biology
a noisy initial response to Bcd can be refined as the embryo develops, including feedback and cross-regulation by the proteins encoded by the Bcd target genes [33] or slow mobility of these target proteins in the embryo [36]. However, neither autoregulation nor the involvement of other transcription factors are plausible explanations for the sharp expression borders (Figure 2) of the synthetic reporters, as these only contain binding sites for Bcd or Gal4 and no binding sites for other transcription factors [21]. More importantly, the notion that the initial transcriptional response to Bcd was noisy has emerged — assuming that the Bcd transcriptional response was accurately measured by target protein expression [33,34] — from a theoretical model based on the slow mobility of the protein determined by FRAP [22]. Given that the response to Bcd has only been analyzed in cycle 14 embryos, i.e. much after the initiation of zygotic transcription, a direct analysis of the Bcd transcriptional response at earlier cycles is lacking and will certainly shed light on this question. Positional Information Downstream of the Bicoid Gradient Even though increased bcd gene dosage in females shifts the entire pattern of the embryo towards the posterior [4], the question of whether the Bcd gradient is acting as a morphogen is often debated. Importantly, the shift of the expression domains of Bcd targets induced by increased bcd dosage are coordinated and occur without interfering with their order along the anterior-posterior axis [4,6]. This indicates that the Bcd gradient provides positional information along the axis in a dose-dependent manner and efforts have been made to understand how this could be achieved. As the Bcd protein encodes a DNA-binding transcription factor, it was initially proposed that the thresholds of Bcd concentration required for the expression of its target genes
Figure 2. The precision and sharp borders of Bcd target gene expression domains. Bcd transcriptional activity is analyzed using synthetic reporters and artificial gradients of transcription factors. (A) A bcd3-lacZ transgene places lacZ under the control of a naive promoter containing three binding sites for Bcd (Bcd BS). (B) In embryos carrying the transgene, the Bcd protein binds to the three Bcd binding sites and this leads to activation of lacZ expression in an anterior domain detected by in situ hybridization. (C) The position of the posterior border of the lacZ expression domain is sharp and precisely positioned relative to the embryo length. (D) A transgenic fly carrying both the UAS-lacZ transgene, placing lacZ under the control of a naive promoter containing UAS binding sites for Gal4, and a transgene allowing maternal expression and anterior anchoring of mRNAs encoding the Gal4-3GCN4 transcriptional activator. (E) This second transgene allows expression, in a Bcd-like maternal gradient, of the Gal4 DNA-binding domain fused to three copies of the GCN4 activation domain. The binding of Gal4-3GCN4 to the UAS binding sites allows expression of lacZ. F: lacZ expression activated by Gal4-3GCN4 is detected by in situ hybridization. LacZ expression domain harbors a precisely positioned and a sharp posterior border (F). Data in (C) and (F) were taken from [21].
depend on the number and on the affinity of Bcd binding sites found in their regulatory regions [5,6]. However, it is now clear that other elements in target gene promoters and the integration of positive and negative transcriptional inputs from proteins bound to these elements are major determinants for the interpretation of positional information along the anterior-posterior axis [37]. Proteins binding with Bcd to regulatory regions of Bcd target genes include Caudal, a major translational target of Bcd (Figure 3A), as well as the gap-proteins Hb, Kru¨ppel, Knirps, Giant and their downstream targets [37,38]. Most of the proteins binding with Bcd in the regulatory regions of the Bcd target genes are themselves Bcd target proteins. Therefore, an increase in bcd gene dosage in females will not only increase Bcd concentration in the embryo but it will also modify the expression domains and the levels of expression of these Bcd target proteins and this will contribute to coordinate the shift of expression of most Bcd target genes along the axis. Such a downstream cascade could therefore in part explain how the system is able to respond by a coordinate shift of the entire gene expression pattern without interfering with their order along the axis. Anterior Positional Information Independent of Bicoid The coordinated shift of expression of the Bcd targets observed upon an increase in bcd dosage is more difficult to understand if some of the proteins binding with Bcd in regulatory regions of its target genes are not themselves Bcd targets. If a protein is not itself a Bcd target, its expression domain and its level of expression will not be modified upon an increase in bcd gene dosage. Therefore, if this protein is binding in the regulatory regions of some of the Bcd target genes, and contributes differently to the expression of two given target genes, it remains possible that an increase in the bcd gene dosage without a change in the
Relative concentration/activity (arbitrary units)
Review R253
F C
B D E
A
Anterior Bicoid Caudal
Posterior Nanos Maternal Hunchback
Torso Repressors downregulated by Torso (Cic, Gr) Current Biology
Figure 3. Maternal contributions to anterior-posterior patterning. Expression or activity of proteins represented by dashed lines is under control of proteins represented by plain lines. In red, the Bicoid protein is expressed as a concentration gradient along the anterior-posterior axis of the embryo [22] and it inhibits the translation of the Caudal protein in the anterior half, creating the concentration gradient of Caudal (A) [8,9]. In blue, the Nanos protein, expressed as a posterior to anterior gradient [48], represses translation of maternal Hunchback, allowing its expression as an anterior-posterior gradient [49] (B). In green, activation of the Torso receptor at the anterior pole represses expression of the Bcd target genes at the anterior pole [39,50] (C). The activation potential of Bcd is enhanced by its phosphorylation by the Torso pathway [43] (D) and/or by the inhibition of ubiquitous repressor [45] (E) such as Capicua (Cic) or Groucho (Gr), which are repressed by Torso activation at the poles [46,47] (F).
expression of this particular partner of Bcd will lead to the inversion of the expression domain of these two genes along the axis. In this regard, two systems established independently of Bcd and providing anterior information have been described (Figure 3). The first system involves the maternal contribution of the Hb protein, which is expressed in an anterior-posterior gradient and repressed in the posterior by the major posterior determinant and translational repressor Nanos (Figure 3B) [13]. Most Bcd targets are also positively regulated by maternal Hb and the presence of Hb-binding sites in a Bcd-dependent promoter induces a posterior shift of its expression domain [7]. This synergy between maternal Hb and Bcd is important for anterior patterning but the fact that it gives rise to the zygotic expression of hb itself has obscured the role of maternal Hb in anterior patterning [7]. The second system providing anterior information independently of Bcd is responsible for the formation of the terminal structures of the embryo and is determined by the signal transduction cascade acting downstream of the Torso receptor tyrosine kinase. The interaction between Bcd and Torso, revealed by the expression of the Bcd targets, appears to be both antagonistic and synergic. First, activation of Torso represses expression of most Bcd targets at the anterior pole (Figure 3C) [39]. The repression of hb, induced by Torso at the anterior pole, has been shown to mediate most terminal patterning of the non-segmented part of the head [40]. Second, several studies have shown that Torso activation induces a small shift of the expression domains of Bcd targets towards the posterior [41–43]. This synergy likely explains the partial rescue of the torso anterior phenotype by increasing amounts of Bcd [44] and the mirror
image expression of anterior Bcd targets in embryos expressing an almost flat gradient of Bcd [45]. At the molecular level, the positive interaction between Bcd and Torso is only partially dependent on phosphorylation of Bcd by the MAPkinase, the nuclear effector of Torso (Figure 3D) [43]. As proposed recently [45], the synergy between Torso and Bcd could also involve local posterior to anterior gradients of ubiquitous maternal repressors (Figure 3E), such as those known to be repressed by Torso activation at the poles (Figure 3F) [46,47]. Whether this occurs by a direct inhibition of Bcd activity or indirectly through the binding to DNA of transcription factors regulated by these repressors has to be clarified. Outlook Quantitative approaches will allow determination of the exact contribution of maternal Hb and the Torso cascade to the Bcd thresholds required for target gene expression. Importantly, the uneven distribution and activity of these sources of positional information (Figure 3), with their highest concentration/activity at the anterior pole gradually decreasing towards the posterior pole, can explain why they do not interfere with the order of expression of the Bcd target genes along the anterior-posterior axis upon increases in bcd dosage, and why the Bcd threshold required for the expression of a given target gene could be different, when artificially positioned elsewhere along the axis, as for instance when increasing the bcd gene dosage [20] or when modifying the shape of the Bcd and maternal Hb gradients [45]. At the molecular, cellular and wholeembryo levels, the Bcd protein is not functioning alone but with many partners and this is not surprising. Is this a valid reason for demoting it from its qualification of morphogen? Bcd’s ability to provide positional information along the axis in a dose-dependent manner makes it unique and how this is achieved is really what we need to understand. Acknowledgements We thank Genevie`ve Almouzni, Baptiste Roelens, Vale´rie Borde and the anonymous reviewers for constructive suggestions and apologize to those whose important work could not be cited due to space constraints. Research in the Dostatni lab is supported by the CNRS and the Institut Curie. N.D. is supported by INSERM and anterior-posterior by the University Rene´ Diderot and the Institut Curie. References 1.
Frohnho¨fer, H.G., and Nu¨sslein-Volhard, C. (1986). Organization of anterior pattern in the Drosophila embryo by the maternal gene bicoid. Nature 324, 120–125.
2.
Driever, W., Siegel, V., and Nu¨sslein-Volhard, C. (1990). Autonomous determination of anterior structures in the early Drosophila embryo by the bicoid morphogen. Development 109, 811–820.
3.
Driever, W., and Nu¨sslein-Volhard, C. (1988). A gradient of bicoid protein in Drosophila embryos. Cell 54, 83–93.
4.
Driever, W., and Nu¨sslein-Volhard, C. (1988). The bicoid protein determines position in the Drosophila embryo in a concentration-dependent manner. Cell 54, 95–104.
5.
Driever, W., and Nu¨sslein-Volhard, C. (1989). The bicoid protein is a positive regulator of hunchback transcription in the early Drosophila embryo. Nature 337, 138–143.
6.
Struhl, G., Struhl, K., and Macdonald, P.M. (1989). The gradient morphogen bicoid is a concentration-dependent transcriptional activator. Cell 57, 1259– 1273.
7.
Simpson-Brose, M., Treisman, J., and Desplan, C. (1994). Synergy between the hunchback and bicoid morphogens is required for anterior patterning in Drosophila. Cell 78, 855–865.
8.
Dubnau, J., and Struhl, G. (1996). RNA recognition and translational regulation by a homeodomain protein. Nature 379, 694–699.
Current Biology Vol 20 No 5 R254
9.
Rivera-Pomar, R., Niessing, D., Schmidt-Ott, U., Gehring, W.J., and Jackle, H. (1996). RNA binding and translational suppression by bicoid. Nature 379, 746–749.
10.
Macdonald, P.M., and Struhl, G. (1986). A molecular gradient in early Drosophila embryos and its role in specifying the body pattern. Nature 324, 537–545.
38.
Rivera-Pomar, R., Lu, X., Perrimon, N., Taubert, H., and Jackle, H. (1995). Activation of posterior gap gene expression in the Drosophila blastoderm. Nature 376, 253–256.
39.
12.
Stauber, M., Jackle, H., and Schmidt-Ott, U. (1999). The anterior determinant bicoid of Drosophila is a derived Hox class 3 gene. Proc. Natl. Acad. Sci. USA 96, 3786–3789.
40.
13.
Irish, V., Lehmann, R., and Akam, M. (1989). The Drosophila posterior-group gene nanos functions by repressing hunchback activity. Nature 338, 646– 648.
41.
14.
Lynch, J.A., Brent, A.E., Leaf, D.S., Pultz, M.A., and Desplan, C. (2006). Localized maternal orthodenticle patterns anterior and posterior in the long germ wasp Nasonia. Nature 439, 728–732.
15.
Schroder, R. (2003). The genes orthodenticle and hunchback substitute for bicoid in the beetle Tribolium. Nature 422, 621–625.
16.
Lemke, S., and Schmidt-Ott, U. (2009). Evidence for a composite anterior determinant in the hover fly Episyrphus balteatus (Syrphidae), a cyclorrhaphan fly with an anterodorsal serosa anlage. Development 136, 117–127.
44.
17.
Berleth, T., Burri, M., Thoma, G., Bopp, D., Richstein, S., Frigerio, G., Noll, M., and Nu¨sslein-Volhard, C. (1988). The role of localization of bicoid RNA in organizing the anterior pattern of the drosophila embryo. EMBO J. 7, 1749–1756.
45.
18.
St. Johnston, D., Driever, W., Berleth, T., Richstein, S., and Nu¨sslein-Volhard, C. (1989). Multiple steps in the localization of bicoid RNA to the anterior pole of the Drosophila oocyte. Development 107, 13–19.
19.
Crick, F. (1970). Diffusion in embryogenesis. Nature 225, 420–422.
20.
Houchmandzadeh, B., Wieschaus, E., and Leibler, S. (2002). Establishment of developmental precision and proportions in the early Drosophila embryo. Nature 415, 798–802.
11.
21.
Crauk, O., and Dostatni, N. (2005). Bicoid determines sharp and precise target gene expression in the Drosophila embryo. Curr. Biol. 15, 1888–1898.
22.
Gregor, T., Wieschaus, E.F., McGregor, A.P., Bialek, W., and Tank, D.W. (2007). Stability and nuclear dynamics of the bicoid morphogen gradient. Cell 130, 141–152.
23.
He, F., Wen, Y., Deng, J., Lin, X., Lu, L.J., Jiao, R., and Ma, J. (2008). Probing intrinsic properties of a robust morphogen gradient in Drosophila. Dev. Cell 15, 558–567.
24.
Lott, S.E., Kreitman, M., Palsson, A., Alekseeva, E., and Ludwig, M.Z. (2007). Canalization of segmentation and its evolution in Drosophila. Proc. Natl. Acad. Sci. USA 104, 10926–10931.
25.
Coppey, M., Berezhkovskii, A.M., Kim, Y., Boettiger, A.N., and Shvartsman, S.Y. (2007). Modeling the bicoid gradient: Diffusion and reversible nuclear trapping of a stable protein. Dev. Biol. 312, 623–630.
26.
Gregor, T., Bialek, W., van Steveninck, R.R.d.R., Tank, D.W., and Wieschaus, E.F. (2005). Diffusion and scaling during early embryonic pattern formation. Proc. Natl. Acad. Sci. USA 102, 18403–18407.
27.
Weiss, M. (2004). Challenges and artifacts in quantitative photobleaching experiments. Traffic 5, 662–671.
28.
Patterson, G.H., and Lippincott-Schwartz, J. (2002). A photoactivatable GFP for selective photolabeling of proteins and cells. Science 297, 1873–1877.
29.
Yu, S.R., Burkhardt, M., Nowak, M., Ries, J., Petrasek, Z., Scholpp, S., Schwille, P., and Brand, M. (2009). Fgf8 morphogen gradient forms by a source-sink mechanism with freely diffusing molecules. Nature 461, 533–536.
30.
Hecht, I., Rappel, W.-J., and Levine, H. (2009). Determining the scale of the Bicoid morphogen gradient. Proc. Natl. Acad. Sci. USA 106, 1710–1715.
31.
Spirov, A., Fahmy, K., Schneider, M., Frei, E., Noll, M., and Baumgartner, S. (2009). Formation of the bicoid morphogen gradient: an mRNA gradient dictates the protein gradient. Development 136, 605–614.
32.
Bergmann, S., Sandler, O., Sberro, H., Shnider, S., Schejter, E., Shilo, B.Z., and Barkai, N. (2007). Pre-steady-state decoding of the Bicoid morphogen gradient. PLoS Biol. 5, e46.
33.
ManuSurkova, S., Spirov, A.V., Gursky, V.V., Janssens, H., Kim, A.-R., Radulescu, O., Vanario-Alonso, C.E., Sharp, D.H., Samsonova, M., et al. (2009). Canalization of gene expression in the Drosophila blastoderm by gap gene cross regulation. PLoS Biol. 7, e1000049.
34.
Gregor, T., Tank, D.W., Wieschaus, E.F., and Bialek, W. (2007). Probing the limits to positional information. Cell 130, 153–164.
35.
Lebrecht, D., Foehr, M., Smith, E., Lopes, F.J.P., Vanario-Alonso, C.E., Reinitz, J., Burz, D.S., and Hanes, S.D. (2005). Bicoid cooperative DNA binding is critical for embryonic patterning in Drosophila. Proc. Natl. Acad. Sci. USA 102, 13176–13181.
36.
Okabe-Oho, Y., Murakami, H., Oho, S., and Sasai, M. (2009). Stable, precise, and reproducible patterning of bicoid and hunchback molecules in the early Drosophila embryo. PLoS Comput. Biol. 5, e1000486.
37.
42.
43.
46.
47.
48. 49.
50.
Ochoa-Espinosa, A., Yucel, G., Kaplan, L., Pare, A., Pura, N., Oberstein, A., Papatsenko, D., and Small, S. (2005). The role of binding site cluster strength in Bicoid-dependent patterning in Drosophila. Proc. Natl. Acad. Sci. USA 102, 4960–4965. Segal, E., Raveh-Sadka, T., Schroeder, M., Unnerstall, U., and Gaul, U. (2008). Predicting expression patterns from regulatory sequence in Drosophila segmentation. Nature 451, 535–540. Ronchi, E., Treisman, J., Dostatni, N., Struhl, G., and Desplan, C. (1993). Down-regulation of the Drosophila morphogen Bicoid by the Torso receptor- mediated signal transduction cascade. Cell 74, 347–355. Janody, F., Reischl, J., and Dostatni, N. (2000). Persistence of Hunchback in the terminal region of the Drosophila blastoderm embryo impairs anterior development. Development 127, 1573–1582. Gao, Q., Wang, Y., and Finkelstein, R. (1996). Orthodenticle regulation during embryonic head development in Drosophila. Mech. Dev. 56, 3–15. Grossniklaus, U., Cadigan, K.M., and Gehring, W.J. (1994). Three maternal coordinate systems cooperate in the patterning of the Drosophila head. Development 120, 3155–3171. Janody, F., Sturny, R., Catala, F., Desplan, C., and Dostatni, N. (2000). Phosphorylation of bicoid on MAP-kinase sites: contribution to its interaction with the torso pathway. Development 127, 279–289. Schaeffer, V., Killian, D., Desplan, C., and Wimmer, E.A. (2000). High Bicoid levels render the terminal system dispensable for Drosophila head development. Development 127, 3993–3999. Ochoa-Espinosa, A., Yu, D., Tsirigos, A., Struffi, P., and Small, S. (2009). Anterior-posterior positional information in the absence of a strong Bicoid gradient. Proc. Natl. Acad. Sci. USA 106, 3823–3828. Jimenez, G., Guichet, A., Ephrussi, A., and Casanova, J. (2000). Relief of gene repression by Torso RTK signaling: role of capicua in Drosophila terminal and dorsoventral patterning. Genes Dev. 14, 224–231. Paroush, Z., Wainwright, S.M., and Ish-Horowicz, D. (1997). Torso signalling regulates terminal patterning in Drosophila by antagonising Groucho-mediated repression. Development 124, 3827–3834. Wang, C., Dickinson, L.K., and Lehmann, R. (1994). Genetics of nanos localization in Drosophila. Dev. Dynamics 199, 103–115. Cho, P.F., Gamberi, C., Cho-Park, Yoon A., Cho-Park, I.B., Lasko, P., and Sonenberg, N. (2006). Cap-dependent translational inhibition establishes two opposing morphogen gradients in Drosophila embryos. Curr. Biol. 16, 2035–2041. Janody, F., Sturny, R., Schaeffer, V., Azou, Y., and Dostatni, N. (2001). Two distinct domains of Bicoid mediate its transcriptional downregulation by the Torso pathway. Development 128, 2281–2290.