The anterior-posterior (A-P) and dorsal-ventral (D-V) axes of the early Drosophila embryo are established by two key maternal morphogens: bicoid (bed) and dorsal (dl), respectively. The bed protein is expressed in a broad concentration gradient along the A-P axis, with peak levels present at the anterior pole, while dl is expressed in a gradient along the D-V axis with peak levels along the ventral surface. The two morphogens are unrelated and their gradients are formed by distinct processes. Nonetheless, we have obtained evidence that they generate sharp on/off stripes of target gene expression through a similar mechanism. Both morphogens establish overlapping patterns of transcriptional activators and repressors in the early embryo. The activators and repressors bind to closely linked sites within short (300 to 500 bp) target promoter elements that have the properties of on/off switches. The activators act in concert with the morphogen to define a broad region where target genes can be initiated. Borders of target gene expression are established by the repressors, resulting in the formation of stripes.

We are interested in understanding how morphogens control development. In particular, how do gradients of morphogens generate sharp patterns of gene expression in the early Drosophila melanogaster embryo? There are two key maternal morphogens in Drosophila, bicoid (bed) and dorsal (dl) (for review, see St Johnston and Nusslein-Vol-hard, 1992). These are expressed during oogenesis and are present in the unfertilized egg. The bed morphogen is important for the specification of cell identity along the anterior-posterior (A-P) axis. The protein is distributed in a broad concentration gradient in early embryos, with peak levels at the anterior pole and progressively lower levels in more posterior regions. In contrast, dl is uniformly expressed along the A-P axis, but is graded along the dorsal-ventral (D-V) axis, with peak levels along the ventral surface and progressively lower levels in more lateral and dorsal regions (Steward et al., 1988). This dl gradient is important for establishing different tissues along the D-V axis.

Although the bed and dl proteins are both sequencespecific transcription factors, they are completely unrelated. bed contains a homeobox (Frigerio et al., 1986; Berleth et al., 1988), while dl contains the REL domain and is related to the mammalian regulatory factor NF-kB and the oncoprotein rel (Steward, 1987; Ghosh et al., 1990; Kieran et al., 1990). In addition, the bed and dl gradients are formed through totally different processes. In the case of bed, the mRNA is trapped at its site of entry at the anterior pole of growing oocytes (Frigerio et al., 1986; Macdonald and Struhl, 1988). This mRNA serves as the source of the morphogen, which is translated in anterior regions and diffuses into more posterior regions (Driever and Nusslein-Volhard,. In contrast to this process of mRNA localization, the dl gradient is formed by regulated nuclear transport (Roth et al., 1989; Rushlow et al., 1989; Steward, 1989). The dl protein is initially detected throughout the cytoplasm of unfertilized eggs and early embryos. About 90 minutes after fertilization, protein in ventral regions enters nuclei while that present in dorsal regions remains in the cytoplasm. Despite these considerable differences, we have obtained evidence that the two morphogens use a similar strategy to produce sharp patterns of gene expression, particularly stripes, in the early embryo.

To understand how the bed and dl morphogens work, we have analyzed the expression of target genes that they directly regulate. We have examined several such genes, but here we will focus on two: the bed target even-skipped (eve) and the dl target rhomboid (rho).

The expression patterns of target genes

eve encodes a homeodomain protein that plays an important role in segmentation: mutant embryos that lack eve protein fail to form segment borders (Nusslein-Volhard et al., 1985; Harding et al., 1986; Macdonald et al., 1986). The eve protein is initially expressed throughout all embryonic nuclei at about 2 h after fertilization. During a short period, just 20 to 30 min, this general staining pattern gives way to a series of 7 transverse stripes (see Fig. IA; Frasch et al., 1987). When they are first formed, the stripes are broad and symmetric, with each including about 5 to 6 nuclei. During another 30 to 60 min, leading to the onset of gastrulation, the stripes are refined to include just 2 or 3 cells. In addition, by this time each stripe shows clear polarity with peak levels of protein at the anterior margin and lower levels in more posterior cells. Genetic studies suggested that the formation of the initial broad stripes depends on the gap class of segmentation genes (Frasch and Levine, 1987; Goto et al., 1989). The subsequent refinement of the stripes involves eve autoregulation (Jiang et al., 1991b). We have analyzed the regulation of several of the eve stripes, but the one that has been examined in the most detail is stripe #2. It is expressed in an anterior region of the embryo, at the future boundary between the posterior head and anterior thorax.

rho encodes a putative transmembrane receptor protein that plays an important role in the differentiation of the ventral epidermis that arises from the neuroectoderm (Mayer and Nusslein-Volhard, 1988). The gene is expressed in a highly dynamic pattern that includes a number of different tissues during embryogenesis (Bier et al., 1990). The initial pattern includes two ventral-lateral stripes that extend along the length of the embryo. Each of these stripes is about 8 to 10 cells wide and covers the presumptive neuroectoderm, which gives rise to the neurons of the CNS and the ventral epidermis.

First we will consider eve stripe 2 and then describe the regulation of the rho lateral stripes. We will then summarize the very striking parallels in the regulation of these two patterns.

eve stripe 2

The first step in the regulation of stripe 2 by bed is the activation of the gap genes. Two of the best characterized gap genes, hunchback (hb) and Kruppel (Kr) (Tautz et al., 1987; Knipple et al., 1985; Preiss et al., 1985). have been shown to encode proteins that are expressed in broad, overlapping domains in precellular embryos (Stanojevic et al., 1989). hb is initially expressed in the anterior half of the embryo, while Kr is expressed in central regions. Both proteins contain zinc fingers (Tautz et al., 1987; Knipple et al., 1985), but with regard to stripe 2 regulation (see below), hb functions as a transcriptional activator while Kr is a repressor, hb and Kr represent just two of the five gap genes, and together they encode a series of transcriptional activators and repressors that span the length of the embyro. We wish to understand how these different permutations of gap gene expression generate on/off stripes of eve expression.

To address this issue we have examined the interaction of gap proteins with defined regions of the eve promoter. These studies have been facilitated by the demonstration that different regions of the eve promoter control the expression of individual stripes (Goto et al., 1989; Harding et al., 1989). For example, a 500-bp fragment located between about −1.6 kb and −1.1 kb is sufficient to direct expression of stripe 2 (Fig. IB), while a different 500-bp fragment between -3.8 kb and −3.3 kb is responsible for stripe 3. We have investigated the genetic regulation of stripe 2 by examining the expression of the eve stripe 2-LacZ fusion gene in a variety of segmentation mutants. These experiments, along with previous studies, have identified four key regulatory genes that control the expression of the stripe. Two of the genes function as activators while two are repressors. The genetic regulation of stripe 2 is summarized in Fig. 2A. The bed morphogen activates the expression of hb. Together, bed and hb define a broad domain in the anterior half of the embryo where stripe 2 can be initiated. The borders of the stripe depend on selective repression: giant (gt) in anterior regions and Kr in posterior regions. The key conclusion of the genetic studies is that repression plays the decisive role in defining the stripe borders (Goto et al., 1989; Frasch et al., 1987; Small et al., 1991). The bed and hb activators merely serve to define a broad domain where the stripe can be expressed.

Fig. 2.

Cis- and trans-control of stripes in the Drosophila blastoderm. (A) Regulation of eve stripe 2. The top schematic shows the relation of stripe 2 to its genetically defined regulators. Gradients of bed and hb activate stripe 2 in a broad anterior domain. The anterior and posterior borders of the stripe are formed through selective repression by the gap genes gt and Kr, respectively. The protein products of all four putative regulators bind to high affinity sites within the 500-bp eve stripe 2 promoter element (bottom). (B) Regulation of rho lateral stripes. The top diagrams represent transverse sections of blastoderm-stage embryos showing the relation of rho stripes to their genetically defined regulators. The dl and twi gradients cooperate to activate rho expression in the ventral half of the embryo. The sna repressor prevents activation in the presumptive mesoderm and thus restricts expression of rho to two lateral stripes in neuroectodermal regions. The 300-bp neuroectoderm element of the rho promoter spans the region from −2.0 kb to − 1.7 kb, and contains a total of 10 binding sites for dl, twi or sna. There are also two E-box sequences (bHLH). Various bHLH proteins are able to bind to these sites with different affinities in vitro.

Fig. 2.

Cis- and trans-control of stripes in the Drosophila blastoderm. (A) Regulation of eve stripe 2. The top schematic shows the relation of stripe 2 to its genetically defined regulators. Gradients of bed and hb activate stripe 2 in a broad anterior domain. The anterior and posterior borders of the stripe are formed through selective repression by the gap genes gt and Kr, respectively. The protein products of all four putative regulators bind to high affinity sites within the 500-bp eve stripe 2 promoter element (bottom). (B) Regulation of rho lateral stripes. The top diagrams represent transverse sections of blastoderm-stage embryos showing the relation of rho stripes to their genetically defined regulators. The dl and twi gradients cooperate to activate rho expression in the ventral half of the embryo. The sna repressor prevents activation in the presumptive mesoderm and thus restricts expression of rho to two lateral stripes in neuroectodermal regions. The 300-bp neuroectoderm element of the rho promoter spans the region from −2.0 kb to − 1.7 kb, and contains a total of 10 binding sites for dl, twi or sna. There are also two E-box sequences (bHLH). Various bHLH proteins are able to bind to these sites with different affinities in vitro.

The importance of repression is illustrated by examining the expression of stripe 2-ZzzcZ fusion genes in early emby-ros (Small et al., 1992). We have shown that by the onset of cellularization they are expressed in tight stripes. However, at earlier stages (nuclear cleavage cycle 12 and 13), these same fusion genes are expressed in very broad patterns covering most of the anterior half of the embryo. During a period of just 20 to 30 min this broad pattern is refined into stripe 2. First the posterior border is established, and then the anterior border. The kinetics of this refinement process correspond quite well with the known time of appearance of the stripe 2 regulators. The bed and hb activators are maternally expressed and present prior to the time when the Kr and gt repressors first appear.

The first indication that bed, hb, gt and Kr might directly regulate stripe 2 expression came from DNA binding studies. All four genes encode sequence-specific DNA binding proteins which interact with multiple sites in the eve stripe 2 promoter element (Stanojevic et al., 1989; Small et al., 1991). There are a total of 12 factor binding sites, 6 activators and 6 repressors (Fig. 2A). Interestingly, there is very close linkage of bed/hb activator sites and gt/Kr repressor sites. This linkage suggests that gt and Kr define the stripe borders through a short range mechanism of transcriptional repression. Several lines of evidence suggest that the stripe 2 element functions as a genetic on/off switch to produce a stripe of gene transcription. In anterior regions, there are high levels of bed and hb, which bind to the six activator sites and turn the stripe 2 element on. In progressively more posterior regions, there are increasing amounts of Kr, which binds to the three Kr repressor sites. This blocks the binding or activities of the bed and hb activators, which leads to repression and the formation of the posterior stripe border. Similarly, in anterior regions containing high levels of gt, the gt repressor sites are filled, leading to the formation of the anterior border.

The most critical test of this switch model involves disrupting individual binding sites in the context of otherwise normal eve-LacZ fusion genes. The first experiments that were done involved the use of a large region of the eve promoter, spanning the first 5.2 kb of 5 • flanking sequence (Stanojevic et al., 1991). This not only directs the expression of stripe 2, but also drives equally intense expression of stripes 3 and 7. Thus, these latter stripes serve as internal controls of even mild disruptions in stripe 2 expression. Disruptions in all three gt binding sites have no effect on stripes 3 and 7, but there is a pronounced anterior expansion of stripe 2, similar to that observed in gt-mutants. This result provides strong evidence that gt directly interacts with the stripe 2 element to repress transcription and form the anterior border of the stripe. In contrast, mutations in 2 of the 6 activator sites (2 bed sites) result in a marked reduction in the levels of stripe 2 expression relative to stripes 3 and 7. This suggests that bed directly interacts with the stripe 2 element to activate transcription. Although mutations in the bed sites reduce expression, the residual stripe is expressed within the normal limits, underscoring the importance of the repressors in defining the stripe borders.

Future studies will focus on understanding how relatively small changes in the ratios of activators and repressors lead to an on/off switch in transcription. Consider the anterior border of the stripe. A cell just anterior to the anterior-most cell in stripe 2 contains about the same levels of the bed and hb activators as its neighbor, yet does not express eve. Co-immunofluorescence staining suggests that there may be no more than a 2-fold difference in the levels of the gt repressor present in these two cells. How does this 2-fold reduction in gt give an on/off switch in eve transcription? Detailed biochemical studies will be required for a complete answer to this question, but recent observations suggest that the key to the on/off switch involves cooperative interactions among the bed and hb activators. There are catastrophic reductions in the levels of expression when individual activator binding sites are disrupted in the context of the minimal 500-bp stripe 2 element. There are 5 bed binding sites in this element and mutations in either of the two right-most bed sites cause nearly a complete loss of expression. Mutations in either the middle bed site or the solo hb site cause a substantial reduction in expression (Small et al., 1992). These results suggest that cooperative interactions among the activators are required for turning the stripe 2 element on. The binding of the gt or Kr repressor would have to interfere with just one or two of the activators in order to effectively repress transcription and form the stripe borders.

rhomboid lateral stripes

We are also studying the regulation of stripes of gene expression along the dorsal-ventral (D-V) axis of the early embryo, which involves the dl morphogen. To contrast crudely the bed and dl morphogens, consider muscle cells stretched along the length of the embryo, bed controls the expression of segmentation genes such as eve, which aré responsible for making the muscles different in the head versus tail. In contrast, dl is responsible for the formation of muscle in the first place. Thus, the regulation of gene expression along the D-V axis is a problem of tissue differentiation.

Genetic studies suggest that the dl morphogen gradient is responsible for establishing three basic embryonic tissues: mesoderm, neuroectoderm and dorsal ectoderm (summarized in Fig. 2B). Peak levels of the protein in ventral regions activate regulatory genes, including twist (twi) and snail (sna), which are important for the formation of the mesoderm (Simpson, 1983; Thisse et al., 1987). Low levels of dl in lateral regions coincide with genes that are expressed within the neuroectoderm, dl also functions as a transcriptional repressor, and restricts the expression of certain genes to dorsal regions where they are important for the differentiation of the dorsal ectoderm (Rushlow et al., 1987; St Johnston and Gelbart, 1987). We have become interested in determining how dl regulates the expression of neuroectodermal regulatory genes in lateral regions of the early embryo.

rho is expressed in two ventral lateral stripes that extend along the length of the embryo (Fig. 1C). Each rho stripe is about 8 to 10 cells in width and encompasses most or all of the presumptive neuroectoderm. A 300-bp region of the rho promoter is sufficient to direct expression of a LacZ reporter gene within the limits of the ventral-lateral stripes (Ip et al., 1992b). Experimental results suggest that the dl morphogen acts in concert with two of its targets, twi and sna, to establish the rho stripes (summarized in Fig. 2B).

Fig. 1.

Expression patterns of bed and dl target genes. (A) An optical sagittal section of a wild-type Drosophila embryo stained with an anti-eve antibody. Anterior is to the left and dorsal is up. eve is expressed in a series of 7 transverse stripes at the cellular blastoderm stage. (B) A cellularizing embryo from a line transformed with a LacZ reporter gene under the control of a 500-bp promoter element spanning -1.6 kb to -1.1 kb upstream of the eve transcription start site. Expression (here and in C) was detected by in situ hybridization with an anti-sense LacZ RNA probe. This construct directs a stripe of expression at the position of the 2nd endogenous eve stripe. (C) A surface view of an embryo from a line transformed with a LacZ fusion gene driven by a 600-bp rho promoter element spanning −2.2 kb to −1.6 kb. This construct directs the expression of two ventral lateral stripes.

Fig. 1.

Expression patterns of bed and dl target genes. (A) An optical sagittal section of a wild-type Drosophila embryo stained with an anti-eve antibody. Anterior is to the left and dorsal is up. eve is expressed in a series of 7 transverse stripes at the cellular blastoderm stage. (B) A cellularizing embryo from a line transformed with a LacZ reporter gene under the control of a 500-bp promoter element spanning -1.6 kb to -1.1 kb upstream of the eve transcription start site. Expression (here and in C) was detected by in situ hybridization with an anti-sense LacZ RNA probe. This construct directs a stripe of expression at the position of the 2nd endogenous eve stripe. (C) A surface view of an embryo from a line transformed with a LacZ fusion gene driven by a 600-bp rho promoter element spanning −2.2 kb to −1.6 kb. This construct directs the expression of two ventral lateral stripes.

twi and sna encode regulatory proteins containing a basic helix-loop-helix (bHLH) and zinc fingers, respectively (Thisse et al., 1988; Boulay et al., 1987). Mutations in either gene block the formation of the ventral furrow and prevent differentiation of the mesoderm. The sna and twi expression patterns overlap, but the dorsal limits of twi extend beyond the sna pattern, since intermediate levels of dl in ventral-lateral regions are sufficient to activate twi but not sna (Jiang et al., 1991a; Ip et al., 1992a). twi functions primarily as a transcriptional activator that is required for the activation of target genes such as nautilus (Michelson et al.,1990), which are expressed in specific cell types in the mesoderm. While twi is expressed in a ventral-to-lateral gradient, with peak levels present in ventral regions, sna is expressed at uniformly intense levels throughout the presumptive mesoderm. Furthermore, the expression of sna is tightly restricted, and the dorsal limits of its expression abruptly end at the boundary between the mesoderm and neuroectoderm. These sharp sna borders represent a key early step in mesoderm differentiation, and are reminiscent of the sharp anterior margins of the pair-rule genes eve and ftz. Double staining studies indicate that a number of neuroectodermal regulatory genes, such as T3 (lethal of scute) and rho, are expressed in patterns that abut, but do not cross, the sna borders. These observations prompted the proposal that sna is responsible for establishing the boundary between mesoderm and neuroectoderm (Kosman et al., 1991; Leptin, 1991).

The first indication that dl, twi and sna might directly regulate the rho lateral stripes came from DNA binding studies. The minimal 300-bp rho promoter element that is sufficient for these stripes, the NEE (neuroectoderm element), contains a total of ten dl, twi and sna binding sites (summarized in Fig. 2B). As was seen for the eve stripe 2 element, there is clustering of the factor binding sites, and close linkage between activator and repressor sites. For example, none of the four dl binding sites maps more than 50 bp away from a sna repressor site. In addition to the four dl activator sites, there are two twi sites, as well as two ‘E box’ sequences that resemble the binding sites for the prototypic E12/E47 and daughterless (da) proteins (Murre et al., 1989). We have obtained evidence that this rho NEE functions as an on/off switch. In the ventral half of the early embryo, there are high levels of dl and twi. These proteins fill the activator sites and switch the NEE on. In ventral regions there are high levels of sna, which fill the repressor sites and switch off expression by interfering with the binding or activities of nearby activators. Repression might involve a competition mechanism, since sna binds to sequences similar to the divergent E box recognized by the bHLH proteins. In fact, at least one of the twi binding sites, t2, is also recognized by the sna repressor (see Fig. 2B).

Evidence that the NEE functions as a switch element was obtained by mutagenizing individual factor binding sites in the context of otherwise normal rho-LacZ fusion genes. Point mutations that disrupt three of the four dl sites virtually abolish expression, suggesting that dl directly interacts with the rho NEE to activate its transcription. Mutations in the two twi sequences cause a substantial reduction in expression. This is consistent with the possibility that twi directly activates the NEE, although it should be noted that bHLH proteins all recognize related E box sequences in vitro, and it is conceivable that other, as yet unidentified, bHLH proteins might contribute to rho expression. The most dramatic result was obtained when the 4 sna repressor sites were mutagenized. This experiment was done in a way that does not interfere with the binding of twi to overlapping or neighboring activator sites. Disrupting the sna binding sites causes a complete derepression of the rho pattern, such that staining extends throughout ventral regions that correspond to the presumptive mesoderm. This result suggests that sna directly represses rho transcription, and provides evidence that it is directly responsible for establishing the boundary between the mesoderm and neuroectoderm (Ip et al., 1992b).

In conclusion, dl and twi act in concert to define a broad region in the ventral half of the embryo where rho can be activated (summarized in Fig. 2B). The ventral limits of the rho lateral stripes depend on selective repression by sna. This situation is similar to the regulation of eve stripe 2, where repressors are crucial for establishing the borders. It is possible that the dorsal limits of the rho stripes are formed by limiting amounts of the dl and/or twi activators.

There are remarkable parallels between the regulation of eve stripe 2 and the rho lateral stripes, as summarized in Fig. 3. In the case of stripe 2, the bed morphogen activates a ‘helper’, hb, and together these genes define a broad region where stripe 2 can be turned on. bed also activates a repressor, gt, which is responsible for defining the anterior border of the stripe. Similarly, dl activates twi, and it appears that these 2 genes are able to initiate the rho NEE in the entire ventral half of the embryo, dl also activates sna, which defines the ventral limit of the rho stripes. In addition to the obvious parallels in the two regulatory circuits, in both cases the encoded proteins interact with short (300 to 500 bp) elements which generate stripes. For the most part the two sets of proteins are distinct, bed contains a homeobox and hb contains zinc fingers, ‘while the gt repressor contains a basic leucine zipper domain. In contrast, dl contains a REL domain and probably interacts with bHLH proteins; the sna repressor contains zinc fingers. Despite these differences the outcome is the same: stripes.

Fig. 3.

bed and dl generate stripes through similar mechanisms. Diagram of the regulatory circuitry involved in the formation of eve stripe 2 (A) and the rho lateral stripes (B). The arrows indicate activation and the lines with T-ends represent repression. For example, bed activates hb and the two proteins cooperatively activate stripe 2, while gt represses the expression.

Fig. 3.

bed and dl generate stripes through similar mechanisms. Diagram of the regulatory circuitry involved in the formation of eve stripe 2 (A) and the rho lateral stripes (B). The arrows indicate activation and the lines with T-ends represent repression. For example, bed activates hb and the two proteins cooperatively activate stripe 2, while gt represses the expression.

We propose that the tight clustering of activator and repressor sites observed for the eve stripe 2 element and the rho NEE reflects weak protein-protein interactions among the regulatory factors active in the early embryo. Studies on gene regulation, primarily in mammalian systems, have identified instances of strong protein-protein interactions, such as jun-fos interactions (Gentz et al., 1989). But we believe that such interactions represent the minority of the cases. Overall, it is likely that the combinatorial control of gene expression depends on weak interactions. Perhaps weak interactions can work to turn promoters on and off coordinately if the interacting factors are bound to closely linked sites. In a sense, the evolution of pattern in the early embryo may involve bringing factor binding sites into close proximity within small promoter elements. This notion can be tested by determining whether random combinations of regulatory prateins present in the early embryo can be made to interact by juxtaposing their binding sites. A strength of the Drosophila system is that many of the regulatory factors active in the early embryo have been identified, and their expression patterns have been determined in precellu-lar embryos. By knowing these expression patterns it is possible to predict the novel patterns that should be obtained when two factors are able to interact. For example, if bed activator sites in the eve stripe 2 element are replaced by dl binding sites, then a hemistripe of expression in the ventral half of the embryo should be observed.

Y.T.I. is a Hoffmann-La Roche fellow of the Life Sciences Research Foundation. S.J.S. is supported by NIH postdoctoral fellowship. This work is funded by NIH grants GM 46638 and GM 34431.

Berleth
,
T
,,
Burri
,
M.
,
Thoma
,
G.
,
Bopp
,
D.
,
Richstein
,
S.
,
Frigerio
,
G.
,
Noll
,
M.
and
Nusslein-Volhard
,
C.
(
1988
).
The role of localization of bicoid RNA in organizing the anterior pattern of the Drosophila embryo
.
EMBO J
.
7
,
1749
1756
.
Bier
,
E.
,
Jan
,
L. Y.
and
Jan
,
Y. N.
(
1990
).
rhomboid, a gene required for dorso-ventral axis establishment and peripheral nervous system development in Drosophila melanogaster
.
Genes Dev
.
4
,
190
203
.
Boulay
,
J.-L.
,
Dennefeld
,
C.
and
Alberga
,
A.
(
1987
).
The Drosophila developmental gene snail encodes a protein with nucleic acid binding fingers
.
Nature
330
,
395
398
.
Driever
,
W.
and
Nusslein-Volhard
,
C.
(
1988
).
A gradient of bicoid protein in Drosophila embryos
.
Cell
54
,
83
93
.
Frasch
,
M.
,
Hoey
,
T.
,
Rushlow
,
C.
,
Doyle
,
H. J.
and
Levine
,
M.
(
1987
).
Characterization and localization of the even-skipped protein of Drosophila
.
EMBO J
.
67
,
749
759
.
Frasch
,
M.
and
Levine
,
M.
(
1987
).
Complementary patterns of even-skipped and fushi-tarazu expression involve their differential regulation by a common set of segmentation genes in Drosophila
.
Genes Dev
.
1
,
981
995
.
Frigerio
,
D.
,
Burri
,
M
,,
Bopp
,
D.
,
Baumgartner
,
S.
and
Noll
,
M.
(
1986
).
Structure of the segmentation gene paired and the Drosophila PRD gene set as part of a gene network
.
Cell
47
,
735
746
.
Gentz
,
R.
,
Rauscher III
,
F.
J.,
Abate
,
C.
and
Curran
,
T.
(
1989
).
Parallel association of Fos and Jun leucine zippers juxtaposes DNA binding domains
.
Science
243
,
1695
1699
.
Ghosh
,
S.
,
Gifford
,
A. M.
,
Riviere
,
L. R.
,
Tempst
,
P.
,
Nolan
,
G. P.
and
Baltimore
,
D.
(
1990
).
Cloning of the p50 DNA binding subunit of NF-kB: homology to rel and dorsal
.
Cell
62
,
1019
1029
.
Goto
,
T.
,
Macdonald
,
P.
and
Maniatis
,
T.
(
1989
).
Early and late periodic patterns of even-skipped expression are controlled by distinct regulatory elements that respond to different spatial cues
.
Cell
57
,
413
422
.
Harding
,
K.
,
Rushlow
,
C.
,
Doyle
,
H. J.
,
Hoey
,
T.
and
Levine
,
M.
(
1986
).
Cross-regulatory interactions among pair-rule genes in Drosophila
.
Science
233
,
953
959
.
Harding
,
K.
,
Hoey
,
T.
,
Warrior
,
R.
and
Levine
,
M.
(
1989
).
Autoregulatory and gap gene response elements of the even-skipped promoter of Drosophila
.
EMBO J
.
8
,
1205
1212
.
Ip
,
Y. T.
,
Park
,
R. E.
,
Kosman
,
D.
,
Yasdanbaksh
,
K.
and
Levine
,
M.
(
1992a
).
dorsal-twist interactions establish snail expression in Drosophila embryos
.
Genes Dev
.
6
,
1518
1530
.
Ip
,
Y. T.
,
Park
,
R. E.
,
Kosman
,
D.
,
Bier
,
E.
and
Levine
,
M.
(
1992b
).
The dorsal gradient morphogen regulates stripes of rhomboid expression in the presumptive neuroectoderm of the Drosophila embryo
.
Genes Dev
.
6
,
17281739
.
Jiang
,
J.
,
Kosman
,
D.
,
Ip
,
Y. T.
and
Levine
,
M.
(
1991a
).
The dorsal morphogen gradient regulates the mesoderm determinant twist in early Drosophila embryos
.
Genes Dev
.
5
,
1881
1891
.
Jiang
,
J.
,
Hoey
,
T.
and
Levine
,
M.
(
1991b
).
Autoregulation of a segmentation gene in Drosophila: combinatorial interaction of the even-skipped homeobox protein with a distal enhancer element
.
Genes Dev
.
5
,
265
277
.
Kieran
,
M.
,
Blank
,
V.
,
Logeât
,
F.
,
Vandekerckhove
,
J.
,
Lottspeich
,
F.
,
Le Bail
,
O.
,
Urban
,
M. B.
,
Kourilsky
,
P-
,
Baeuerle
,
P. A.
and
Israel
,
A.
(
1990
).
The DNA binding subunit of NF-kB is identical to factor KBF1 and homologous to the rel oncogene product
.
Cell
62
,
1007
1018
.
Knipple
,
D. C.
,
Seifert
,
E.
,
Rosenberg
,
U. B.
,
Preiss
,
A.
and
Jackie
,
H.
(
1985
).
Spatial and temporal patterns of Kruppel gene expression in early Drosophila embryos
.
Nature
317
,
40
44
.
Kosman
,
D.
,
Ip
,
Y. T.
,
Levine
,
M.
and
Arora
,
K.
(
1991
).
Establishment of the mesoderm-neuroectoderm boundary in the Drosophila embryo
.
Science
254
,
118
122
.
Leptin
,
M.
(
1991
).
twist and snail as positive and negative regulators during Drosophila mesoderm development
.
Genes Dev
.
5
,
1568
1576
.
Macdonald
,
P. M.
,
Ingham
,
P.
and
Struhl
,
G.
(
1986
).
Isolation, structure, and expression of even-skipped: a second pair-rule gene of Drosophila containing a horneo box
.
Cell
47
,
721
734
.
Macdonald
,
P. M.
and
Struhl
,
G.
(
1988
).
Cis-acting sequences responsible for anterior localization of bicoid mRNA in Drosophila embryos
.
Nature
336
,
595
598
.
Mayer
,
U.
and
Nusslein-Volhard
,
C.
(
1988
).
A group of genes required for pattern formation in the ventral ectoderm of the Drosophila embryo
.
Genes Dev
.
2
,
1496
1511
.
Michelson
,
A. M.
,
Abmayr
,
S. M.
,
Bate
,
M.
,
Martinez Arias
,
A.
and
Maniatis
,
T.
(
1990
).
Expression of a MyoD family member prefigures muscle pattern in Drosophila embryos
.
Genes Dev
.
4
,
2086
2097
.
Murre
,
C.
,
McCaw
,
P. S.
and
Baltimore
,
D.
(
1989
).
A new DNA binding and dimerization motif in immunoglobulin enhancer binding, daughterless, MyoD, and Myc proteins
.
Cell
56
,
777
783
.
Nusslein-Volhard
,
C.
,
Kluding
,
H.
and
Jurgens
,
G.
(
1985
).
Genes affecting the segmental subdivision of the Drosophila embryo
.
Cold Spring Harbor Syrnp. Quant. Biol
.
50
,
145
154
.
Preiss
,
A.
,
Rosenberg
,
U. B.
,
Kienlin
,
A.
,
Seifert
,
E.
and
Jackie
,
H.
(
1985
).
Molecular genetics of Kruppel, a gene required for segmentation of the Drosophila embryo
.
Nature
313
,
27
32
.
Roth
,
S.
,
Stein
,
D.
and
Nusslein-Volhard
,
C.
(
1989
).
A gradient of nuclear localization of the dorsal protein determines dorsoventral pattern in the Drosophila embryo
.
Cell
59
,
1189
1202
.
Rushlow
,
C. A.
,
Frasch
,
M.
,
Doyle
,
H.
and
Levine
,
M.
(
1987
).
Maternal regulation of zerknullt: a homeobox gene controlling differentiation of dorsal tissues in Drosophila
.
Nature
330
,
583
586
.
Rushlow
,
C. A.
,
Han
,
K.
,
Manley
,
J. L.
and
Levine
,
M.
(
1989
).
The graded distribution of the dorsal morphogen is initiated by selective nuclear transport in Drosophila
.
Cell
59
,
1165
1177
.
Simpson
,
P.
(
1983
).
Maternal-zygotic gene interactions during formation of the dorsoventral pattern in Drosophila embryos
.
Genetics
105
,
615
632
.
Small
,
S.
,
Kraut
,
R.
,
Hoey
,
T.
,
Warrior
,
R.
and
Levine
,
M.
(
1991
).
Transcriptional regulation of a pair-rule stripe in Drosophila
.
Genes Dev
.
5
,
827
839
.
Small
,
S.
,
Blair
,
A.
and
Levine
,
M.
(
1992
).
Regulation of even-skipped stripe 2 in the Drosophila embryo
.
EMBO
J11
,
4047
4057
.
Stanojevic
,
D.
,
Hoey
,
T.
and
Levine
,
M.
(
1989
).
Sequence-specific DNA-binding activities of the gap proteins encoded by hunchback and Kruppel in Drosophila
.
Nature
341
,
331
335
.
Stanojevic
,
D.
,
Small
,
S.
and
Levine
,
M.
(
1991
).
Regulation of a segmentation stripe by overlapping activators and repressors in the Drosophila embryo
.
Science
254
,
1385
1387
.
Steward
,
R.
(
1987
).
Dorsal, an embryonic polarity gene in Drosophila is homologous to the vertebrate proto-oncogene, c-rel
.
Science
238
,
692
694
.
Steward
,
R.
,
Zusman
,
S. B.
,
Huang
,
L. H.
and
Schedl
,
P.
(
1988
).
The dorsal protein is distributed in a gradient in early Drosophila embryos
.
Cell
55
,
487
495
.
Steward
,
R.
(
1989
).
Relocalization of the dorsal protein from the cytoplasm to the nucleus correlates with its function
.
Cell
59
,
1179
1188
.
St Johnston
,
R. D.
and
Gelbart
,
W. M.
(
1987
).
decapentaplegic transcripts are localized along the dorsal-ventral axis of the Drosophila embryo
.
EMBO J
.
6
,
2785
2791
.
St Johnston
,
D.
and
Nusslein-Volhard
,
C.
(
1992
).
The origin of pattern and polarity in the Drosophila embryo
.
Cell
68
,
201
219
.
Tautz
,
D.
,
Lehmann
,
R.
,
Schnurch
,
H.
,
Schuh
,
R.
,
Seifert
,
E.
,
Kienlin
,
K.
and
Jackie
,
H.
(
1987
).
Finger protein of novel structure encoded by hunchback, a second member of the gap class of Drosophila segmentation genes
.
Nature
327
,
383
389
.
Thisse
,
B.
,
Stoetzel
,
C.
,
El Messal
,
M.
and
Perrin-Schmitt
,
F.
(
1987
).
Genes of the Drosophila maternal dorsal group control the specific expression of the zygotic gene twist in presumptive mesodermal cells
.
Genes Dev
.
1
,
709
715
.
Thisse
,
B.
,
Stoetzel
,
C.
,
Gorostiza-Thisse
,
C.
and
Perrin-Schmitt
,
F.
(
1988
).
Sequence of the twist gene and nuclear localization of its protein in endomesodermal cells of early Drosophila embryos
.
EMBO J
.
1
,
21752183
.