The phenotypes of different mutant combinations of teashirt (tsh) and homeotic genes together with their regulatory interactions are described in order to gain insight into tsh gene function. We show that when tsh, Scr, Antp and BX-C genes are missing, the ventral part of the trunk (or thorax and abdomen) is transformed to anterior head identity showing that tsh is a homeotic gene. These genes act synergistically to suppress the expression of the procephalic gene labial (lab) in subsets of cells in each segment of the trunk. Transcripts from the tsh gene always accumulate in segments destined to acquire trunk identities. tsh gene activity is required for the normal function of the Antp and BX-C genes, which modulate in part the expression of tsh. As a whole, our results suggest that tsh plays an essential dual role, during embryogenesis, for determining segmental identity of the trunk. First, tsh is required critically for the identity of the anterior prothorax. Second, tsh is required globally for segmental identity throughout the entire trunk whereas the “classical” homeotic genes have more specific roles. Our results are consistent with the idea that tsh is defining the ground state of the Drosophila trunk region seen in the absence of the Antp and BX-C genes.

The body plan of higher animals, to some extent, has devel-oped along a common theme. The head region, the trunk region and the terminal parts are morphologically distinct in fully differentiated animals. For insects (Fig. 1), these regions or tagma are defined by the head (acron, pro-cephalon and gnathocephalon), trunk (thoracic and abdominal) and tail (abdominal segments posterior to A8 and telson) domains.

Fig. 1.

Summary of the regions where RNA accumulates during embryogenesis from the homeotic genes of the ANT-C, BX-C and the tsh gene. At the top the segmental (S) and parasegmental (P) metameric divisions are indicated. At the bottom the principle morphological parts, or tagmata, of the insect body are indicated. The accumulation of transcripts at the blastoderm (Early) and germ band retraction (Late) stages are shown for each of the genes. Only epidermal expression patterns are given. Abbreviations and references: labial (lab, Diederich et al., 1989), Deformed (Dfd, Chadwick and McGinnis, 1987; Martinez-Arias et al., 1987), Sex combs reduced (Scr, Kuroiwa et al., 1985; Mahaffey and Kaufman, 1987; Martinez-Arias et al., 1987), Antennapedia (Antp, Wirz et al., 1986; Martinez-Arias, 1986), Ultrabithorax (Ubx, Akam and Martinez-Arias, 1985), abdominal-A (abd-A, Harding et al., 1985), Abdominal-B (Abd-B, Kuziora and McGinnis, 1988; Sanchez-Herrero and Crosby, 1988) and teashirt (tsh, Fasano et al., 1991; this study). The ANT-C genes lab, Dfd, Scr and Antp are required for the identity of intercalary, mandibular-maxillary, labial-prothoracic and anterior thoracic segments respectively; the BX-C genes Ubx, abd-A and Abd-B identify posterior thoracic and abdominal segments (Kaufman et al., 1990; Lewis; 1978; Sanchez-Herrero et al., 1985) and tsh is required throughout the trunk region (Fasano et al., 1991).

Fig. 1.

Summary of the regions where RNA accumulates during embryogenesis from the homeotic genes of the ANT-C, BX-C and the tsh gene. At the top the segmental (S) and parasegmental (P) metameric divisions are indicated. At the bottom the principle morphological parts, or tagmata, of the insect body are indicated. The accumulation of transcripts at the blastoderm (Early) and germ band retraction (Late) stages are shown for each of the genes. Only epidermal expression patterns are given. Abbreviations and references: labial (lab, Diederich et al., 1989), Deformed (Dfd, Chadwick and McGinnis, 1987; Martinez-Arias et al., 1987), Sex combs reduced (Scr, Kuroiwa et al., 1985; Mahaffey and Kaufman, 1987; Martinez-Arias et al., 1987), Antennapedia (Antp, Wirz et al., 1986; Martinez-Arias, 1986), Ultrabithorax (Ubx, Akam and Martinez-Arias, 1985), abdominal-A (abd-A, Harding et al., 1985), Abdominal-B (Abd-B, Kuziora and McGinnis, 1988; Sanchez-Herrero and Crosby, 1988) and teashirt (tsh, Fasano et al., 1991; this study). The ANT-C genes lab, Dfd, Scr and Antp are required for the identity of intercalary, mandibular-maxillary, labial-prothoracic and anterior thoracic segments respectively; the BX-C genes Ubx, abd-A and Abd-B identify posterior thoracic and abdominal segments (Kaufman et al., 1990; Lewis; 1978; Sanchez-Herrero et al., 1985) and tsh is required throughout the trunk region (Fasano et al., 1991).

In Drosophila, the identity of segments in these regions depends on the function of a group of selector genes (Garcia-Bellido, 1975) called the homeotic loci (Lewis, 1978; Wakimoto and Kaufman, 1981; Sanchez-Herrero et al., 1985; Duncan, 1987; Kaufman et al., 1990). In this insect, two clusters of genes, the Antennapedia (ANT-C) and Bithorax (BX-C) complexes (Lewis, 1978; Kaufman et al., 1990), control sets of genes necessary for the identity of certain, but not all known, segments of the head (procephalic and gnathocephalic) and trunk (thorax and abdomen) of the larva and adult fly. For example, in embryos, absence of the Antennapedia (Antp) gene causes a replacement of the posterior prothorax, mesothorax and anterior metathorax (i.e. parasegments or PS 4 and 5; see Fig. 1 and Martinez-Arias and Lawrence, 1985, for the definition of parasegments and segments) with posterior labial, prothoracic and anterior mesothoracic (or PS 3 and 4) ones respectively (Kaufman et al., 1990; Schneuwly and Gehring, 1985; Martinez-Arias, 1986).

The ANT-C and BX-C genes have been cloned and their transcription patterns (see Fig. 1) and protein distribution patterns determined (reviewed in Kaufman et al., 1990; Akam, 1987). Each gene possesses a homeobox (McGin-nis et al., 1984; Scott and Weiner, 1984), coding for a 60 amino acid domain required for regulation of downstream genes. A similar clustering of homeobox genes exists in vertebrates (Duboule and Dolle 1989; Graham et al., 1989), including man (Boncinelli et al., 1988), indicating that these clusters may be controlled in part by a common gene network, and may have common targets (McGinnis et al., 1990) across a wide spectrum of animals.

If the Scr, Antp and BX-C (or trunk) genes are removed, all trunk segments exhibit both prothoracic and gnathal identities (Struhl, 1983; Sato et al., 1985) and consequently there is an unknown function that determines the prothoracic identity of the trunk. With respect to function, this prothoracic-gnathal identity defines a ground state, a term initially proposed by Lewis (1963), which is the prototype upon which certain ANT-C and the BX-C genes act to modify segment identity.

Earlier we described a gene that we called teashirt (tsh), which belongs to the homeotic class of genes; it is expressed in the trunk region of the embryo in a pattern similar, but not identical, to the known homeotic genes. However, it is clearly different from the classical homeotic genes: it encodes a zinc finger protein and mutations dis-rupt the entire trunk domain (Fasano et al., 1991). The phenotype of tsh and different homeotic mutations is described in addition to their regulatory interactions to gain insight into the role of tsh in trunk development. We show that the tsh+ gene: (1) is essential for the specific identity of the anterior prothorax, (2) acts independently of the trunk homeotic genes for promoting trunk identity, (3) acts synergistically with the Antp and BX-C genes, to repress anterior head development and head gene activity in the trunk, (4) is modulated and required by the Antp and BX-C genes for their complete function. Together our results show that tsh is a unique homeotic gene essential for global trunk identity. In particular, tsh seems to define the basal segmental identity (or ground state) of the trunk upon which the trunk homeotic genes act to modify segmental identity.

Drosophila stocks, egg collections and cuticle preparations

The alleles of the homeotic genes used were DfdRX1 (Hazelrigg and Kaufman, 1983), labf8 (Merrill et al., 1989), AntpW10 (Riley et al., 1987), ScrW17 (Wakimoto and Kaufman, 1981), tsh8 (Fasano et al., 1991), Df(2L)TW161 (Wright et al., 1976), Df(2L)305 (the kind gift of B. Wakimoto) which removes only the tsh gene in combination with tsh8 (Coré and Kerridge, unpublished data), Df(3R)P9 (Lewis, 1978), Ubx9.22 (Kerridge and Morata, 1982), and salIIA55 (Jürgens, 1988). The ScrC1AntpNs+RC3Df(3R)P9 chromosome is described by Struhl, (1983) and the AntpW10Df(3R)P9 one by Riley et al. (1987). Egg lays were made from trans heterozygotes over a wild-type chromosome where possible. Since Df(3R)P9 requires two copies of the Abd-B gene for fertility, Dp(3;3)P5 (Lewis, 1978), which is a tandem duplication of the BX-C, was used in the relevant crosses. Cuticle preparations were done as in Fasano et al. (1991).

Probes and in situ hybridisation to whole embryos

The in situ hybridisation technique is described in Fasano et al., (1991). The probes used for in situ hybridisation to whole embryos were: lab: a 1.5 kb XhoI-XbaI cDNA fragment (Diederich et al., 1989); Dfd: a 5 kb EcoRI genomic fragment; Scr: a 5 kb HindIII genomic fragment (Martinez-Arias et al., 1987); tsh: a 2.5 kb EcoRI cDNA fragment (Fasano et al.,1991); Antp P1: a 3.6 kb EcoRI genomic fragment; Antp P2: a 1.2 kb HindIII-EcoRI, from the 4.6 kb EcoRI, genomic fragment (Bermingham et al., 1990); Ubx: a 1.44 kb HindIII-EcoRI genomic fragment from the 5′ end of the gene (Irish et al., 1989); abd-A: a 2 kb AvaI fragment (a gift from P. Ingham); and Abd-B: a 2.8 kb PstI fragment common to all three Abd-B transcripts (Kuziora and McGinnis, 1988). The different fragments were PCR amplified and marked with digoxy-genin (DIG)-labelled nucleotides (Boehringer-Mannheim).

Double labelling by whole mount in situ and antibody staining was performed essentially as described by Cohen (1990) with the following modifications: devitillinized eggs were treated for 15 minutes with 3% H2O2 in methanol before rehydration. Proteinase K treatment was for 2 minutes 30 seconds (a suggestion from Dougan and DiNardo). Embryos were then hybridized with DIG probes (see above), washed and blocked with 0.1% bovine BSA, 0.1% Tween in PBS. The eggs were incubated overnight at 4°C with dilutions of anti-invected (the gift of M. Wilcox) and anti-DIG antibodies simultaneously. Following standard washes, embryos were incubated with biotinylated anti-mouse IgG at a dilution of 1:400. All other steps were as described by Cohen (1990).

Heat shock experiments

Embryos, with P elements carrying the Scr (HSS), Ubx (HSU) or Antp (HSA) mini genes under the control of a heat shock pro-moter, were collected following 2 hour egg lay periods. The HSU P element is that described in Gonzalez-Reyes et al. (1990), the HSA one by Gibson and Gehring (1988) and the HSS one was a gift from M. Scott. All are localized on chromosome 3. After aging for 2 or 4 hours, eggs were heat shocked in a water bath at 36°C using one of the following regimes: 1 hour, 2 hours or three 25 minute heat shocks. For multiple heat shocks, eggs were left to recover at 25°C for 11 hours between each heat shock. For in situ hybridization, eggs were aged for 2 or 4 hours and treated as described above; half the embryos were labelled with tsh probe and, in order to verify efficient expression of the respective homeotic gene following heat shock, the other half were probed with Scr, Antp or Ubx probes. For cuticle preparations, eggs were allowed to develop at 25°C for 48 hours and the larvae mounted as described above.

A closer look at the mutant phenotype and transcription pattern of the tsh gene

Fasano et al. (1991) showed that tsh null mutations disrupt all of the ventral trunk region and especially result in the loss of patterns of PS3 or the posterior labium and anterior prothorax of the differentiated larva. Every trunk (thoracic and abdominal) segment is smaller and sclerotic cuticle differentiates in ventro-lateral positions located between the denticle belts (Fig. 2A and B). Internally, a striking phenotype observed in tsh null mutations is that the trunk-specific ventral set of neuron clusters is disrupted suggesting, but not proving, a partial transformation to a gnathal-like arrangement (Fasano et al., 1991). For 10-20% of tsh individuals the prothoracic clusters resemble those found in the labial segment (Figs. 3A and B). Mutations in the tsh gene can therefore be interpreted in two ways; either they partially transform the trunk segments into a gnathal-like identity, and in particular the prothoracic segment into a labial one, or they cause a non-specific change in segmental identity perhaps due to cell death.

Fig. 2.

A comparison of the larval cuticular phenotypes of wild-type (A) and a tsh (B) null mutation. For a description of the wild-type cuticular phenotype see Lohs-Schardin et al. (1979). In A, note the typical array of ventral denticle belts and distinguishing features of the individual thoracic and abdominal segments. In B, the denticle belts throughout the thorax and abdomen are disrupted, the prothoracic denticle belt is missing and the presence of sclerotic cuticle in each trunk segment (arrows). Abbreviations: b: prothoracic beard; PT: prothorax; A1: first abdominal segment. The anterior limits of the prothoracic, mesothoracic, metathoracic and first abdominal segments are indicated with large arrowheads. Bar represents 50 μm.

Fig. 2.

A comparison of the larval cuticular phenotypes of wild-type (A) and a tsh (B) null mutation. For a description of the wild-type cuticular phenotype see Lohs-Schardin et al. (1979). In A, note the typical array of ventral denticle belts and distinguishing features of the individual thoracic and abdominal segments. In B, the denticle belts throughout the thorax and abdomen are disrupted, the prothoracic denticle belt is missing and the presence of sclerotic cuticle in each trunk segment (arrows). Abbreviations: b: prothoracic beard; PT: prothorax; A1: first abdominal segment. The anterior limits of the prothoracic, mesothoracic, metathoracic and first abdominal segments are indicated with large arrowheads. Bar represents 50 μm.

Fig. 3.

The peripheral nervous system of wild-type (A) and a tsh8/Df(2L)TW161 (B) embryos at germ band retraction. The trunk segments of wild-type larvae are made up of three thoracic segments, the prothorax (Pt), mesothorax (Ms) and metathorax (Mt), and 8 abdominal segments (marked with arrowheads: first abdominal segment: A1). The segment immediately anterior to the trunk is called the labial (Lb) segment. The neuron clusters of the peripheral nervous system have been revealed using the antibody 22C10. Each trunk segment has three groups (dorsal, d; lateral, l; and ventral, v; labelled here between the third and fourth abdominal segments only) of neuron clusters which are variations on the same theme. The labial segment has neuron clusters of only dorsal and lateral origin (see Fig. 6 in Ghysen and O’Kane, 1989), which give rise to the hypophysis (hy), the labial sense organ (lso) and a previously undiscovered cluster we call the dorsal labial organ (dlo). The clusters deriving from more head segments are normal and their names can be found in Ghysen et al., (1986). In B, note that the clusters in the prothorax (PT) resemble those of the labial segment (Lb) and have lost their typical trunk characteristics; a small hy, an lso and a dlo differentiate in the prothoracic position (labelled in white). Only two thoracic and the usual 8 abdominal segments remain. Note also that the posterior mesothoracic compartment has a posterior prothoracic identity since a prothoracic specific neuron cluster, Ich 3, sends an axon towards the metathoracic segment (large white arrow in B; a black one in A). Anterior is left and dorsal at the top for each photograph. Bar represents 50 μm.

Fig. 3.

The peripheral nervous system of wild-type (A) and a tsh8/Df(2L)TW161 (B) embryos at germ band retraction. The trunk segments of wild-type larvae are made up of three thoracic segments, the prothorax (Pt), mesothorax (Ms) and metathorax (Mt), and 8 abdominal segments (marked with arrowheads: first abdominal segment: A1). The segment immediately anterior to the trunk is called the labial (Lb) segment. The neuron clusters of the peripheral nervous system have been revealed using the antibody 22C10. Each trunk segment has three groups (dorsal, d; lateral, l; and ventral, v; labelled here between the third and fourth abdominal segments only) of neuron clusters which are variations on the same theme. The labial segment has neuron clusters of only dorsal and lateral origin (see Fig. 6 in Ghysen and O’Kane, 1989), which give rise to the hypophysis (hy), the labial sense organ (lso) and a previously undiscovered cluster we call the dorsal labial organ (dlo). The clusters deriving from more head segments are normal and their names can be found in Ghysen et al., (1986). In B, note that the clusters in the prothorax (PT) resemble those of the labial segment (Lb) and have lost their typical trunk characteristics; a small hy, an lso and a dlo differentiate in the prothoracic position (labelled in white). Only two thoracic and the usual 8 abdominal segments remain. Note also that the posterior mesothoracic compartment has a posterior prothoracic identity since a prothoracic specific neuron cluster, Ich 3, sends an axon towards the metathoracic segment (large white arrow in B; a black one in A). Anterior is left and dorsal at the top for each photograph. Bar represents 50 μm.

The evolution of tsh transcription during embryogenesis suggests that the gene is similar to the classical homeotic ones (Fasano et al., 1991). It is initiated at the blastoderm stage and then is expressed in all trunk segments. Here we analyse in more detail the evolution of tsh transcription during the transition from parasegmental to segmental morphology of the wild-type embryo. Normal embryos were doubly labelled to detect tsh transcripts as well as the invected protein, which provides a marker for the posterior compartments of each segment similar to the engrailed protein (Coleman et al., 1987). In the fully extended germ band (Fig. 4A), the anterior border of tsh transcripts occupies cell for cell the invected stripe localized in PS 3. During retraction of the germ band, transcripts from the tsh gene become segmental in the epidermis, being restricted to the thorax and first eight abdominal segments (Fig. 4B). However, in the central nervous system, tsh mRNAs are found in the posterior part of the labial segment (Figs. 4B and C). Thus in the epidermal cells tsh is expressed in a parasegmental and then a segmental register whereas in the central nervous system its expression remains parasegmental.

Fig. 4.

Wild-type embryos doubly labelled with a probe to the tsh coding sequence and an antibody to the invected protein. Embryos are at the extended germ band stage (A) or during germ band retraction (B, C) when segments are forming. The invected protein (brown) marks the posterior compartments of each segment giving a reference for the position of tsh transcript (blue) accumulation. Overlap between the two gene products is dark brown whereas invected alone is light brown. In A, note that the invected stripe in PS 3 coincides with the anterior border of tsh transcript accumulation, and in B and C that this stripe does not overlap that of tsh messages (arrow) in the epidermis of PS 3 (3). Contrarily, in the central nervous system (C, between arrows), tsh and invected overlap in PS 3. A and B are lateral views and C is a ventral one. Anterior is to the left. Bar represents 50 μm

Fig. 4.

Wild-type embryos doubly labelled with a probe to the tsh coding sequence and an antibody to the invected protein. Embryos are at the extended germ band stage (A) or during germ band retraction (B, C) when segments are forming. The invected protein (brown) marks the posterior compartments of each segment giving a reference for the position of tsh transcript (blue) accumulation. Overlap between the two gene products is dark brown whereas invected alone is light brown. In A, note that the invected stripe in PS 3 coincides with the anterior border of tsh transcript accumulation, and in B and C that this stripe does not overlap that of tsh messages (arrow) in the epidermis of PS 3 (3). Contrarily, in the central nervous system (C, between arrows), tsh and invected overlap in PS 3. A and B are lateral views and C is a ventral one. Anterior is to the left. Bar represents 50 μm

tsh and the homeotic genes act at the same level in the gene hierarchy

The homeotic-like nature of the tsh gene prompted us to analyse its cross-regulatory interactions with the classical homeotic genes. If tsh is a bona fide homeotic gene then it should have three properties with respect to this regulation: first, that its transcription is initiated independently of other homeotic genes, second that maintenance of its transcription depends on the activity of other homeotic genes and third that tsh+ function is required for the maintenance of transcription from other homeotic genes.

For the transcript distributions of the different genes in different mutant situations described below, we focus on the epidermal patterns of expression. No deviation from the wild-type patterns of expression of the homeotic genes or the tsh one (see Fig. 1) have been found before the fully extended germ band stage in any mutant situation analysed. Thus transcription of tsh and the homeotic genes are initiated and evolve independently of each other from the blas-toderm through to the early stages of gastrulation. During later stages of embryogenesis however, specific regulatory interactions have been discovered.

Maintenance of tsh transcription is modulated by the trunk but not the head homeotic genes

First, the expression pattern of tsh in Sex combs reduced (Scr), Deformed (Dfd) and labial (lab) mutant embryos has been followed. No change in the spatial distribution of tsh mRNAs can be detected (data not shown). We have also examined the expression of tsh in embryos carrying the Scr+ gene under the control of a heat shock promoter (HSS). In these embryos, following 1 hour or multiple heat shocks during embryogenesis, tsh expression is unchanged in comparison to the wild-type pattern (not shown). These results show that the activity of these head genes are irrelevant for tsh expression.

In Antennapedia (Antp) null mutants at the extended germ band stage, significantly lower levels of tsh transcripts are seen in PS 4 and 5 (Fig. 5C) compared to wild-type embryos. During germ band retraction (Fig. 5D), tsh transcripts are not found in the cells which are transformed to labial identity (the posterior prothorax). These observations show that Antp+ gene activity is required to maintain a normal level of tsh transcription throughout PS 4 and 5, the epidermal domains of Antp+ gene function (Schneuwly and Gehring, 1985; Martinez-Arias, 1986), and especially for maintaining tsh transcription in the posterior compartment of the prothorax.

Fig. 5.

The distribution of tsh transcripts in different homeotic gene mutations. Genotypes are: wild type (A, B), and homozygotes for AntpW10 (C, D), Ubx9.22 (E), Df(3R)P9 (F) and Scr13AAntpNs+RC3Df(3R)P9 (G, H). In homozygotes for the Antp null allele (C, D), tsh transcripts are weakly expressed, compared to wild type (A) in PS 4 and 5, where ANTPprotein is abundant, especially in the posterior compartments (little arrows). In the abdomen, levels of tsh messages are comparable in wild type (A) and Antp mutants (C). By germ band retraction (B, D), a line of cells in the posterior prothorax lack tsh transcripts in AntpW10 homozygotes (D, arrow). In Ubx embryos (E), tsh transcripts are more abundant in PS 6 compared to wild type (A). In the absence of BX-C genes (F), tsh messages are abundantly expressed in PS 6-13. In embryos lacking products of Scr Antp and BX-C genes (G, H), tsh is not detected in the posterior part of each segment (arrows in G). In H, note that about four cells do not transcribe tsh in each segment; this is the expected size of the posterior compartment. Numbers indicate the identity of parasegments; otherwise segmental designations are shown: labial (Lb), prothoracic (Pt), mesothoracic (Ms) and metathoracic (Mt) segments are labelled. Embryos, in E and F, are at the germ band extension stage, whereas A, C, G and H are during early germ band retraction and B and D at the end of germ band retraction. See Campos-Ortega and Hartenstein (1985) for developmental stages.

Fig. 5.

The distribution of tsh transcripts in different homeotic gene mutations. Genotypes are: wild type (A, B), and homozygotes for AntpW10 (C, D), Ubx9.22 (E), Df(3R)P9 (F) and Scr13AAntpNs+RC3Df(3R)P9 (G, H). In homozygotes for the Antp null allele (C, D), tsh transcripts are weakly expressed, compared to wild type (A) in PS 4 and 5, where ANTPprotein is abundant, especially in the posterior compartments (little arrows). In the abdomen, levels of tsh messages are comparable in wild type (A) and Antp mutants (C). By germ band retraction (B, D), a line of cells in the posterior prothorax lack tsh transcripts in AntpW10 homozygotes (D, arrow). In Ubx embryos (E), tsh transcripts are more abundant in PS 6 compared to wild type (A). In the absence of BX-C genes (F), tsh messages are abundantly expressed in PS 6-13. In embryos lacking products of Scr Antp and BX-C genes (G, H), tsh is not detected in the posterior part of each segment (arrows in G). In H, note that about four cells do not transcribe tsh in each segment; this is the expected size of the posterior compartment. Numbers indicate the identity of parasegments; otherwise segmental designations are shown: labial (Lb), prothoracic (Pt), mesothoracic (Ms) and metathoracic (Mt) segments are labelled. Embryos, in E and F, are at the germ band extension stage, whereas A, C, G and H are during early germ band retraction and B and D at the end of germ band retraction. See Campos-Ortega and Hartenstein (1985) for developmental stages.

In Ultrabithorax (Ubx) mutations at the germ band elongation stage, tsh transcription is similar to wild type but higher levels of transcripts are detected in PS 6 (Fig. 5E), the principle site of Ubx+ gene function (Lewis, 1978). In the absence of the three BX-C genes, Ubx, abdominal-A (abd-A) and Abdominal-B (Abd-B), tsh is expressed more strongly throughout the abdominal region (PS 6-13; Fig. 5F), as in PS 4 of wild-type embryos. The strong expression of tsh observed in these cases is probably mediated by Antp+ gene activity which is expressed strongly in these parasegments (Hafen et al., 1984; Carroll et al., 1986).

In the absence of the Antp and BX-C or Scr, Antp and BX-C gene activities, tsh expression is normal until the end of the elongated germ band stage. During retraction of the germ band, tsh transcripts are no longer detected in the pos-terior compartments of each trunk segment and in the anterior compartments tsh is strongly expressed (Fig. 5G and H). Therefore, the evolution of tsh transcript distribution in this mutant resembles that of PS 3 in the epidermis of wild-type embryos (Figs 4, 5A and B); in the posterior compartment, the absence of tsh mRNA accumulation correlates with labial identity (Struhl, 1983; Sato et al., 1985). These observations indicate that the maintenance of tsh expression, in the posterior compartments of trunk segments is dependent on a common function of the Antp+ and BX-C+ genes, whereas in the anterior compartments it is inde-pendent and correlates with prothoracic identity.

Transcription of some head but not trunk homeotic genes are modulated by the tsh gene

Transcripts from the the lab gene are located in PS-1 at the blastoderm and the germ band elongation stages in wild-type embryos (Fig. 6A; Diederich et al., 1989). Once segments are formed, during retraction of the germ band, lab transcripts occupy the intercalary segment, the optic lobe and dorsal ridge (Fig. 6B; Diederich et al., 1989). Below we describe the patterns of lab mRNAs in different loss of function mutations for tsh, ANT-C and BX-C genes. None affect the wild-type pattern of lab in the head and none have a novel accumulation pattern of lab transcripts before the extended germ band stage.

Fig. 6.

The accumulation of labial (lab) gene transcripts in wild type and in different homeotic mutations during embryogenesis. The wild-type pattern at the extended germ band (A) and retracted germ band (B) stages follows the description of Diederich et al. (1989). Abbreviations: dr: dorsal ridge; ol: optic lobe; Ic: intercalary segment and -1: parasegment -1. In tsh homozygotes at the extended germ band stage (C), lab is expressed in the wild type pattern as well as in small groups of cells in the trunk. In Antp null mutations during germ band retraction (D), lab transcripts accumulate ectopically in the dorsal posterior part of the prothorax (arrow); mg: normal midgut expression. In Scr, Antp, BX-C mutant embryos at the retracted germ band stage (E), lab is ectopically expressed in homologous dorsolateral groups of cells in the thorax and abdomen (arrowheads) and in specific cells of the labial lobe (Lb).

Fig. 6.

The accumulation of labial (lab) gene transcripts in wild type and in different homeotic mutations during embryogenesis. The wild-type pattern at the extended germ band (A) and retracted germ band (B) stages follows the description of Diederich et al. (1989). Abbreviations: dr: dorsal ridge; ol: optic lobe; Ic: intercalary segment and -1: parasegment -1. In tsh homozygotes at the extended germ band stage (C), lab is expressed in the wild type pattern as well as in small groups of cells in the trunk. In Antp null mutations during germ band retraction (D), lab transcripts accumulate ectopically in the dorsal posterior part of the prothorax (arrow); mg: normal midgut expression. In Scr, Antp, BX-C mutant embryos at the retracted germ band stage (E), lab is ectopically expressed in homologous dorsolateral groups of cells in the thorax and abdomen (arrowheads) and in specific cells of the labial lobe (Lb).

In tsh null mutations at the extended germ band stage, the lab gene is expressed ectopically in PS 4 to 13 in a small number of cells in each segment close to the pri-mordia of the tracheal placodes of the trunk (Fig. 6C; see Glaser and Shilo, 1991) and in some cells at the ventral midline (not shown see Fig. 9C).

None of the homeotic mutations of the ANT and BX complexes, alone or in combination, alter the pattern of lab mRNA distributions compared to wild type in embryos until after the extended germ band stage. In Antp embryos during retraction of the germ band, lab expression deviates from the wild-type pattern for the first time; it is expressed ectopically in the posterior dorsal part of the prothoracic segment (Fig. 6D). In BX-C embryos, lab expression is indistinguishable from the wild-type pattern at all stages (not shown). In Antp BX-C embryos however, lab tran-scripts are found in the dorsal parts of each trunk segment (not shown); similarly for Scr, Antp and BX-C null mutants we found the same ectopic pattern of lab gene expression and in addition lab is expressed ectopically in the labial lobe (Fig. 6E). Thus the tsh gene acts early and Scr, Antp and BX-C genes act late to suppress lab gene expression in different subsets of cells within the trunk region.

The expression patterns of the Dfd and Scr genes in tsh and other homeotic mutations has been described previ-ously (Jack et al., 1988; Riley et al., 1987; Fasano et al., 1991). Only the expression of Scr is altered compared to wild type in tsh mutations.

We have examined the patterns of expression of the P1 and P2 transcripts of the Antp gene, plus the patterns of the Ubx, abd-A and Abd-B genes in tsh null embryos to ask whether or not these genes are expressed normally. The spatial distribution of all of these transcripts is as in wild type (data not shown; for wild-type expression patterns of these genes see Bermingham et al., 1990; Akam and Martinez-Arias, 1985; Harding et al., 1985; Kuziora and McGinnis, 1988; Sanchez-Herrero and Crosby, 1988) showing that the expression of these genes is independent of tsh+ gene activity.

In summary, our results show that transcriptional activation of the tsh gene occurs independently of the homeotic genes at the blastoderm stage and that initiation of transcription of the homeotic genes is independent of tsh and other homeotic gene activities. In addition, we show that tsh is at the same level as the homeotic genes in the hierarchy of gene interaction; i.e. tsh as well as the homeotic gene mutations may alter the maintenance of expression of those genes that are initiated in more anterior positions but never those initiated in more posterior domains. The phenotype of tsh mutations can be considered to be a mixture of normal (Antp, Ubx, abd-A and Abd-B) and ectopic (lab) homeotic gene products in the trunk.

The tsh gene is essential for anterior prothoracic identity

From the phenotypic analysis of tsh mutations (Figs 2 and 3) and the analysis of its cross regulatory interactions with homeotic genes (Figs 5 and 6) it seems that tsh+ gene-activity is essential for the development of the prothorax. For example, in the absence of all homeotic gene activity, where all trunk segments have a prothoracic-like identity (Struhl, 1983; Sato et al., 1985), the tsh gene is transcribed (Fig. 5G and H). An apparent paradox arising with this idea is that Scr+ gene activity is known to be essential for identity of the anterior prothorax. Null mutations at the Scr locus cause an incomplete transformation of the prothorax to a mesothoracic segment (Wakimoto and Kaufman 1981; Pat-tatucci et al., 1991) as well as a homeotic transformation of part of the labial segment to a maxillary one. Further-more, ectopic expression of the Scr protein under the control of a heat shock promoter (HSS) induces the transformation of the second and third thoracic segments to prothoracic ones (Gibson et al., 1990).

To analyse whether tsh has a critical role for the iden-tity of the anterior prothorax, the cuticular phenotype of tsh larvae developing with high levels of Scr protein, were examined. Following heat shock treatments of HSS tsh embryos during embryogenesis, two phenotypic differences are observed compared to HSS tsh+ or tsh controls. First, no prothoracic patterns develop and second the numbers of denticles in the trunk segments are reduced (see Figs 7A and 2B). Therefore, in the absence of tsh+ activity, the Scr+ gene cannot promote prothoracic identity and represses trunk (denticle belt) development instead.

Fig. 7.

The cuticular phenotypes of tsh larvae that developed with high levels (A) or in the absence (B) of Scr+ gene activity. Note that when the Scr gene is expressed ectopically, prothoracic denticles do not form in the thorax, as is the case in tsh+ larvae (Gibson et al., 1990). The number of denticles is reduced in each segment compared to tsh larvae (see Fig. 2B). In more extreme examples, the thoracic denticles may be absent whereas in more posterior segments, a reduced number of abdominal denticles always differentiate. Control larvae (not shown) carrying a tsh+ allele caused the transformation of second and third thoracic segments to prothoracic identity as described by Gibson et al., (1990). In the absence of Scr and tsh gene products (B), the prothoracic denticle belt is missing and maxillary sense organs (small arrowheads) are found in the labium due to the loss of Scr gene activity, as well as in the normal position. Note that the dorsal part of the prothorax differentiates (DPT) whereas in tsh larvae it does not (compare with Fig. 2B).

Fig. 7.

The cuticular phenotypes of tsh larvae that developed with high levels (A) or in the absence (B) of Scr+ gene activity. Note that when the Scr gene is expressed ectopically, prothoracic denticles do not form in the thorax, as is the case in tsh+ larvae (Gibson et al., 1990). The number of denticles is reduced in each segment compared to tsh larvae (see Fig. 2B). In more extreme examples, the thoracic denticles may be absent whereas in more posterior segments, a reduced number of abdominal denticles always differentiate. Control larvae (not shown) carrying a tsh+ allele caused the transformation of second and third thoracic segments to prothoracic identity as described by Gibson et al., (1990). In the absence of Scr and tsh gene products (B), the prothoracic denticle belt is missing and maxillary sense organs (small arrowheads) are found in the labium due to the loss of Scr gene activity, as well as in the normal position. Note that the dorsal part of the prothorax differentiates (DPT) whereas in tsh larvae it does not (compare with Fig. 2B).

We have also compared the cuticular phenotypes of tsh and tsh Scr double mutations. The only phenotypic differ-ence in the double compared to single mutants is that the dorsal part of the prothoracic trunk segment is made whereas it is missing in tsh mutations (compare Figs 7B and 2B). This shows that Scr+ gene activity suppresses dorsal trunk development in the prothorax in the absence of tsh+ gene activity.

Mutations in the spalt (sal) gene (Jürgens, 1988) cause anterior prothoracic patterns to differentiate in the labial segment, providing an independent test to ask whether tsh+ activity is linked to prothoracic identity. In sal mutant homozygotes, tsh is expressed ectopically, for the first time at the germ band extension stage as well as at later stages, in the position of the labium (Fig. 8). This same mutant also replaces two tail segments with trunk segments (Jür-gens, 1988) and tsh is transcribed ectopically in these seg-ments (Fig. 8) showing that one normal maintenance function of the sal+ gene is to repress, directly or indirectly, tsh in the labial and tail domains. These observations are consistent with the idea that tsh+ gene activity is always correlated with prothoracic and more generally with trunk identity. Together these results show that tsh+ and Scr+ activities are indispensable for the identity of the anterior prothorax.

Fig. 8.

Ectopic expression of tsh transcripts in sal homozygotes. The tsh transcription pattern at germ band extension (A) and at the retracted germ band (B) stages is shown. Compared to wild type (see Figs. 5 A and B), tsh is expressed ectopically in PS 2 (2) and the labial (Lb) as well as the tail regions (14 and 15). Abbreviations are the same as in the previous Figs.

Fig. 8.

Ectopic expression of tsh transcripts in sal homozygotes. The tsh transcription pattern at germ band extension (A) and at the retracted germ band (B) stages is shown. Compared to wild type (see Figs. 5 A and B), tsh is expressed ectopically in PS 2 (2) and the labial (Lb) as well as the tail regions (14 and 15). Abbreviations are the same as in the previous Figs.

Fig. 9.

The effects of removing tsh+ and Antp+ gene activities on larval cuticular patterns (A) as well as on the Scr (B) and lab (C) gene expression patterns. In Antp tsh double mutants, prothoracic and mesothoracic denticle belts are absent and replaced with head cuticle (arrow in A); a hemi denticle belt differentiates in the position metathorax (arrowhead). In the absence of Antp and tsh functions (B), Scr is expressed ectopically in the anterior part of PS 4. At the same developmental stage, Scr messages in wild type or Antp mutations are restricted to PS 2 and to a small patch of dorsal cells in PS 3 (Riley et al., 1987; data not shown). In tsh mutant homozygotes, Scr shows the wild-type distribution pattern and is ectopically expressed in the ventral part of PS 3 (Fasano et al., 1991). There is no staining of the posterior midgut in this figure. The 2nd, 3rd and 4th parasegments are indicated. In the same genotype, lab is expressed ectopically as in tsh (Fig. 6C) embryos except transcripts are expressed in more cells and more abundantly in parasegments 4 and 5, the principle sites of Antp+ function. Note in this ventro-lateral view that cells in the ventral midline (small arrowheads) accumulate lab transcripts; a similar pattern is seen in tsh homozygotes.

Fig. 9.

The effects of removing tsh+ and Antp+ gene activities on larval cuticular patterns (A) as well as on the Scr (B) and lab (C) gene expression patterns. In Antp tsh double mutants, prothoracic and mesothoracic denticle belts are absent and replaced with head cuticle (arrow in A); a hemi denticle belt differentiates in the position metathorax (arrowhead). In the absence of Antp and tsh functions (B), Scr is expressed ectopically in the anterior part of PS 4. At the same developmental stage, Scr messages in wild type or Antp mutations are restricted to PS 2 and to a small patch of dorsal cells in PS 3 (Riley et al., 1987; data not shown). In tsh mutant homozygotes, Scr shows the wild-type distribution pattern and is ectopically expressed in the ventral part of PS 3 (Fasano et al., 1991). There is no staining of the posterior midgut in this figure. The 2nd, 3rd and 4th parasegments are indicated. In the same genotype, lab is expressed ectopically as in tsh (Fig. 6C) embryos except transcripts are expressed in more cells and more abundantly in parasegments 4 and 5, the principle sites of Antp+ function. Note in this ventro-lateral view that cells in the ventral midline (small arrowheads) accumulate lab transcripts; a similar pattern is seen in tsh homozygotes.

tsh and trunk homeotic genes together promote trunk and suppress head segmental identity

Our results suggest that tsh+ activity is associated with segments destined to differentiate denticle belts (i.e. trunk identity; Figs 4, 5, 6, 8) and never associated with segments destined to make head or other identities. However, the Antp and BX-C genes seem to have the same function since they are transcribed normally in tsh embryos and in this genotype, denticles (although abnormal ones) are made (Fig. 2B). To test whether tsh+ and the trunk homeotic genes act independently for trunk (i.e. denticle belt) identity, we have examined the cuticular phenotypes of tsh larvae in the absence of trunk homeotic gene activities. To test whether these same genes also repress head development, we have monitored the expression of head genes in the same genotypes.

Amorphic Antp mutations cause replacement of the mesothoracic and metathoracic denticle belts with mixed prothoracic-mesothoracic and prothoracic-first abdominal ones respectively (Wakimoto and Kaufman, 1981; Martinez-Arias, 1986). In tsh Antp double mutant embryos the first two thoracic denticle belts are completely absent and replaced with cuticle typical of the head skeleton. In addition, the third thoracic belt is partially or completely deleted (Fig. 9A). Therefore, on the basis of the differen-tiation of denticle belts, the normal function of the tsh+ and Antp+ genes is to suppress head development and to pro-mote normal thoracic identity. We have examined the expression of the lab, Dfd and Scr genes in this genotype. Compared to the expression pattern of these genes in tsh mutations alone, only the Scr (Fig. 9B) and lab (Fig. 9C) genes are expressed differently when Antp+ activity is miss-ing: Scr is expressed ectopically in PS 3 due to loss of tsh+ activity (Fasano et al., 1991) and in the anterior part of PS 4 due to the combined loss of tsh+ and Antp+ products (Fig. 9B); the lab gene is expressed ectopically in PS 4-13 due to loss of tsh+ gene activity (Fig. 6C) and more extensively in PS 4 and 5 (Fig. 9C) due to the absence of Antp+ and tsh+ gene activities.

When Scr, Antp and the three BX-C gene activities are missing, all the segments of the trunk have similar identi-ties: in the anterior compartment a mixture of prothoracic and mesothoracic denticles (that we call prothoracic-like) develop, and in the posterior compartment labial sense organs differentiate (Struhl, 1983; Sato et al., 1985). If tsh+ activity is removed in addition to these genes, the prothoracic-like denticle belts are missing (Fig. 10A) and replaced ventrally with cuticle typically found in the head skeleton; according to Jürgens et al. (1986), this cuticle derives from the procephalon and/or acron (or anterior head; compare Figs 10B and C). On the dorsal side of larvae of this genotype, trunk elements still differentiate indicating that the specification of dorsal trunk patterns is independent of tsh and the trunk homeotic gene functions tested here. In larvae mutant for Antp, BX-C and tsh genes, the Scr+ gene activity suppresses the formation of this head cuticle in the most anterior thoracic region (not shown). No readily recogniz-able patterns (e.g. denticle belt) differentiate in this position, confirming that the Scr+ gene cannot promote trunk (or denticle) identity alone but plays a role in suppressing anterior head identity. In conclusion, the tsh gene is critical for identity of the prothorax and, together with the homeotic genes Antp, Ubx, abd-A and Abd-B, is required for global trunk identity of the larvae.

Fig. 10.

The trunk homeotic genes act co-operatively to repress head development (A) and lab gene expression (D, E) in the embryonic trunk region. In tsh, Scr, Antp, BX-C mutations, the ventral denticle belts are missing and replaced with cuticle normally found in the head (A: below arrowheads). In this mutant combination, the cuticle does not fuse dorsally; upon devitellinisation the cuticle splays out as a sheet. Maxillary sense organs are found in the labium due to the Scr mutation as well as in the normal maxillary position (small arrowheads). In the dorsal position small wild-type trunk segments and pattern elements (dorsal hairs) develop. The head skeleton (hs) differentiates in its normal anterior position as do the anal pads (ap) in their posterior one. B and C are enlargements of the ectopic and normal head cuticle respectively; note the similarity of rippled cuticle found in the ectopic and normal head of the ventral arm (VA), ventral plate (VP), dorsal arm (DA) and dorsal bridge (DBr), all of which derive from the segments of the procephalon or acron (Jürgens et al., 1986). In B an unidentifiable sense organ (arrow) is often observed in each trunk segment. In tsh, Antp, BX-C mutations (D), lab is expressed ectopically in parasegments 4 to 13; here the metameres are labelled with an antibody to the invected protein (light grey stripes; arrowheads). In tsh, Scr, Antp, BX-C mutations (E), lab is ectopically expressed throughout the trunk region especially in PS 4 to 13 at the extended germ band stage and weakly in PS 3 (arrow). For the normal expression of the lab gene at this stage see Fig. 6A.

Fig. 10.

The trunk homeotic genes act co-operatively to repress head development (A) and lab gene expression (D, E) in the embryonic trunk region. In tsh, Scr, Antp, BX-C mutations, the ventral denticle belts are missing and replaced with cuticle normally found in the head (A: below arrowheads). In this mutant combination, the cuticle does not fuse dorsally; upon devitellinisation the cuticle splays out as a sheet. Maxillary sense organs are found in the labium due to the Scr mutation as well as in the normal maxillary position (small arrowheads). In the dorsal position small wild-type trunk segments and pattern elements (dorsal hairs) develop. The head skeleton (hs) differentiates in its normal anterior position as do the anal pads (ap) in their posterior one. B and C are enlargements of the ectopic and normal head cuticle respectively; note the similarity of rippled cuticle found in the ectopic and normal head of the ventral arm (VA), ventral plate (VP), dorsal arm (DA) and dorsal bridge (DBr), all of which derive from the segments of the procephalon or acron (Jürgens et al., 1986). In B an unidentifiable sense organ (arrow) is often observed in each trunk segment. In tsh, Antp, BX-C mutations (D), lab is expressed ectopically in parasegments 4 to 13; here the metameres are labelled with an antibody to the invected protein (light grey stripes; arrowheads). In tsh, Scr, Antp, BX-C mutations (E), lab is ectopically expressed throughout the trunk region especially in PS 4 to 13 at the extended germ band stage and weakly in PS 3 (arrow). For the normal expression of the lab gene at this stage see Fig. 6A.

In embryos lacking the tsh, Antp and the three BX-C gene functions, the expression of the Scr gene is identical to that described for tshAntp embryos (Fig. 9B). On the other hand in the same genotype, lab is expressed as in tsh mutations alone (Fig. 6C) except that this ectopic expression within PS 4-13 is more extensive (Fig. 10D). When the Scr+ gene function is removed in addition to tsh, Antp and BX-C genes, the same deviation from the wild-type pattern is observed, plus the presence of a small patch of lab expression in PS3 (Fig. 10E). Therefore, even though tsh+ activity alone represses lab expression in homologous subsets of cells within the trunk, tsh+ and Scr+ activities repress lab in PS3; tsh+ and Antp+ in PS 4 and 5 and tsh+, Antp+ and BX-C+ genes repress lab in PS4-13. In conclusion, tsh+ activity in combination with that of the other trunk homeotic genes represses the expression of at least one head gene as well as head development in a synergistic manner.

We have also examined the expression of the Dfd gene in the absence of different combinations of homeotic mutations. Dfd expression resembles that of wild type in all cases except when the Scr+ and tsh+ genes are missing. By the extended germ band stage, Dfd transcripts are detected in a small patch of cells of the posterior part of the labial segment and, in some cases, in the posterior part of the pro-thoracic segment (Fig. 11A-C). Thus Scr+ and tsh+ genes suppress Dfd gene activity in subsets of cells of the labium and prothorax, whereas the Antp+ and BX-C+ genes have no detectable role in suppressing Dfd in the trunk.

Fig. 11.

Scr+ and tsh+ gene functions act in combination to suppress the gnathal gene Dfd. The expression of Dfd in wild type (A)and tsh, Scr, Antp BX-C mutant (B, C) embryos, at extended germ band stage, is shown. The Antp and BX-C mutations do not affect Dfd expression patterns. Dfd is normally (A) expressed in the maxillary (Mx) and mandibular (Md) segments at this stage as described by Jack et al. (1988). In tsh, Scr, Antp, BX-C embryos (B, C), a small patch of cells (arrowheads) express Dfd in the posterior part of the labium (B, C) and sometimes a group of cells in the posterior prothorax (C), in addition to the normal regions.

Fig. 11.

Scr+ and tsh+ gene functions act in combination to suppress the gnathal gene Dfd. The expression of Dfd in wild type (A)and tsh, Scr, Antp BX-C mutant (B, C) embryos, at extended germ band stage, is shown. The Antp and BX-C mutations do not affect Dfd expression patterns. Dfd is normally (A) expressed in the maxillary (Mx) and mandibular (Md) segments at this stage as described by Jack et al. (1988). In tsh, Scr, Antp, BX-C embryos (B, C), a small patch of cells (arrowheads) express Dfd in the posterior part of the labium (B, C) and sometimes a group of cells in the posterior prothorax (C), in addition to the normal regions.

In conclusion, tsh as well as the Scr, Antp and BX-C genes have common and independent functions to repress head gene activity and head development in the embryonic trunk. Within this domain in mutants of the trunk homeotic genes, the derepression of the procephalic gene lab is more extensive than that observed for the gnathocephalic genes Scr and Dfd.

Expression of the lab gene in trunk segments of tshembryos is functionally significant

A feature of tsh mutations is that the head gene lab is expressed ectopically in the trunk region (Fig. 6C); there-fore in this region a mixture of head and trunk homeotic gene activities exists. Is the ectopic expression of lab func-tionally significant? To test this idea the cuticular phenotypes of tsh lab double mutants were compared to those of tsh and lab mutants alone. Null mutations in the lab gene delete specific patterns of the anterior head thought to be part of the intercalary segment (Merrill et al., 1989). Null mutations of the lab gene act as partial suppressors of the tsh phenotype (compare Figs 12 and 2B); the number of denticles in each belt of the trunk domain is increased and the sclerotic cuticle, found in the trunk segments of tsh mutations, no longer differentiates. Therefore in the trunk region of tsh mutations, lab+ gene expression is function-ally significant. In addition, these observations suggest that the lab+ gene functions to promote head (sclerotic cuticle) and repress trunk (denticle belt) development.

Fig. 12.

The cuticular phenotype of a tshlab larvae. Compared to tshlab+ cuticular patterns (Fig. 2B), the numbers of individual denticles is increased in each segment and the dark coloured sclerotic cuticle is missing.

Fig. 12.

The cuticular phenotype of a tshlab larvae. Compared to tshlab+ cuticular patterns (Fig. 2B), the numbers of individual denticles is increased in each segment and the dark coloured sclerotic cuticle is missing.

Normal Antp and BX-C gene function requires tsh+ gene activity

The cross regulatory interactions described above suggest that the Antp and BX-C genes are modulating the tran-scription of the tsh gene (Fig. 5C-H) and therefore raise the possibility that tsh+ activity is a requisit for normal trunk homeotic gene function. To analyse this point we have examined tsh transcription in embryos carrying the structural genes coding for the Antp or Ubx proteins fused to the control elements of a heat shock promoter (HSA and HSU respectively). Following heat shock of such embryos, trunk segments differentiate in specific head metameres; the labial, the maxillary and at least one procephalic head segment are transformed to trunk (Gibson and Gehring, 1988; Gonzalez-Reyes and Morata, 1991). HSA and HSU embryos, have been heat shocked during embryogenesis and then tested for tsh transcript distribution (see Materials and methods).

After multiple heat shock treatments, tsh messages can be detected in the normal position as well as ectopically in particular domains of the head in HSA and HSU embryos (compare Figs 5A and 13). tsh transcripts were found more frequently in the procephalic domain than in the maxillary one, which in turn was more frequently labelled than the labial segment (see Table 1). These results correlate with the larval phenotypes observed by Gonzalez-Reyes and Morata (1991) following heat shock treatment of HSU embryos. The head segments transformed are the labial, the maxillary and at least one anterior procephalic head segment including part of the mandibular one. In terms of frequency, the procephalic segment(s) is (are) transformed more frequently than the maxillary one, which in turn, is affected more frequently than the labial one. It is noteworthy that, whereas Ubx or Antp are expressed all over the embryo following heat shock (data not shown), tsh does not respond in the same way in all cells. Ectopic tsh expression is never observed in the tail region. Non-heat shocked HSU or HSA embryos or heat shocked wild type controls always give a normal pattern of tsh expression. These results show that tsh expression is correlated with trunk identity and sup-port the idea that the Antp and Ubx genes act, directly or indirectly, as positive regulators of tsh transcription.

Table 1.

Ectopic tsh expression in different segmental positions of the head in HSA and HSU embryos following heat shock

Ectopic tsh expression in different segmental positions of the head in HSA and HSU embryos following heat shock
Ectopic tsh expression in different segmental positions of the head in HSA and HSU embryos following heat shock

It is clear that tsh transcription depends in part on the normal function of the Antp+ and Ubx+ genes but is this functionally significant? The Antp and BX-C genes are expressed normally in tsh mutations and they can make den-ticle belts which have abnormal morphology (Fig. 2B). Thus the classical trunk homeotic genes are capable of directing cells into the trunk developmental pathway but require tsh+ activity to do this correctly. We have examined the cuticles of HSU and HSA embryos, following heat shock during embryogenesis, in the presence and absence of tsh+ activity to analyse this point. If tsh+ activity is essential for homeotic gene function, abnormal denticle belts should form in the head in the absence of tsh+ gene activity following heat shock.

In tsh+ larvae, our results are in accordance with those described previously except that occasionally, prothoracic (PS 3) patterns can be distinguished in the labial segment of HSU embryos (see Fig. 14A). In the absence of tsh+ activity, ubiquitous expression of Ubx or Antp proteins cause the differentiation of denticle belts in ectopic positions. Following a one hour heat shock treatment, a denti-cle belt appears in the position of the prothorax with about the same frequency as the denticle belt deriving from the procephalic domain (Gonzalez-Reyes and Morata, 1990), in both tsh HSA and tsh HSU larvae (Fig. 14B, D). More extreme heat shock treatments give rise to additional den-ticle belts in the maxillary and less frequently the labial segmental positions, as described by Gonzalez-Reyes and Morata (1990) for wild-type larvae. These ectopic denticle belts resemble those in the trunk domain of tsh null mutations (Fig. 2B), confirming that Antp+ and Ubx+ genes require tsh+ activity for normal function but can direct the development of a trunk segment, independently of tsh+ activity.

Another observation is that high levels of the Antp+ and Ubx+ genes can partially rescue the tsh phenotype in the trunk; the sclerotic cuticle observed in the trunk of tsh muta-tions is not observed and the size of segments seems to be more normal than in tsh mutations alone (compare Figs 2A, B with 14B, D). Surprisingly, following overexpression of the Ubx gene, this rescue can occur even posteriorly to the first abdominal segment, whereas in wild-type embryos, high levels of Ubx protein have no effect in these positions (Gonzalez-Reyes et al., 1990). This observation suggests that, in part, tsh has a function in common with Ubx in the trunk.

In conclusion, these results show that the Antp and Ubx genes act independently of tsh for the determination of trunk identity (denticle belts) although tsh+ function is indispensable for the complete function of the trunk homeotic genes (normal denticle belts).

In Drosophila, specific trunk segmental identity is determined by homeotic genes of the Antennapedia (Scr and Antp) and Bithorax complexes (Kaufman et al., 1990; Lewis, 1978); each of these trunk homeotic genes is expressed and functional within restricted domains of the embryonic trunk (reviewed by Akam, 1987; Ingham, 1988). The pattern of ventral denticle belts of the larval cuticle are morphological markers that allow the specific identities of most segments to be distinguished. The absence of any single homeotic gene results in a change in the identity of a specific segment or segments into another type, giving rise to a typical homeotic transformation. Trunk morphology in the Drosophila embryo also depends on the normal function of the tsh gene, a member of the homeotic class (Fasano et al., 1991). However, the tsh gene is different from the classical homeotic genes in at least two respects: it codes for a zinc finger protein and the analysis of the phenotype of mutations at this locus showed that the entire trunk is affected. In this paper, we present evidence that the tsh+ gene has a specific role in determining the identity of the anterior prothorax (PS 3) as well as a more general one in all trunk segments. We also show that tsh acts in com-bination with the trunk homeotic genes, Scr, Antp, Ubx, abd-A, and Abd-B, to promote specific trunk identity and to repress anterior head development in the trunk. Since tsh+ activity is necessary for segment identity of the entire trunk region it has characteristics in common with the “region specific” homeotic genes spalt and fork head (Jür-gens, 1988; Jürgens and Weigel, 1988).

The tsh+ gene determines the specificity of function of the Scr+ gene for anterior prothoracic identity

Our results argue that tsh+ function is critically required for the establishment of the specific identity of the anterior pro-thorax and probably for PS 3. First, of the homeotic genes expressed in the prothorax, neither Scr+ nor Antp+ gene activities can account fully for this identity; absence of the Scr+ and Antp+ gene functions still give rise to a denticle belt with prothoracic-like characteristics (Struhl, 1983; Sato et al., 1985). Second, we show that tsh expression always correlates with prothoracic identity (Figs 5 and 8). Third, in tsh mutants, the pattern elements specific to the anterior prothoracic segment are absent (Fig. 2B).

The Scr+ gene plays a role in anterior prothoracic identity since mutations cause a partial and weak transforma-tion of this compartment to an anterior mesothoracic one (Wakimoto and Kaufman, 1981; Pattatucci et al., 1991). In tsh embryos, the Scr gene is expressed ectopically in the cells in the position of the anterior prothorax (Fasano et al., 1991) but no prothoracic patterns develop. When the Scr gene is expressed under the control of heat shock promoter, all thoracic segments resemble the prothorax (Gibson et al., 1990); however, in the absence of tsh+ activity, high levels of Scr cannot induce the formation of anterior prothoracic structures (Fig. 7A). These results are consistent with the idea that the tsh+ gene has a critical role for promoting the anterior prothoracic pathway.

The specific identity of the anterior prothorax is probably determined by the simultaneous activities of the tsh and Scr gene products. Whenever tsh+ and Scr+ products coex-ist in a segment prothoracic identity ensues. For example, we show that in HSU embryos prothoracic patterns are sometimes observed in the labial position (Fig. 14A); similarly, in spalt (sal) mutations, the labial segment is transformed to anterior prothorax (Jürgens, 1988). In the labial segment for both of these cases, Scr expression is unaffected (Gonzales-Reyes and Morata, 1990; Casanova, 1989) and tsh transcripts coexist with it (Figs 8 and 13). When an Scr protein is expressed ectopically it has no affect on tsh transcription (this work) and therefore it coexists with tsh products forming prothoracic patterns in each thoracic segment (Gibson et al., 1990).

Fig. 13.

Transcript accumulation of tsh in HSU (A) and HSA (B) embryos, following three heat shock pulses during embryogenesis. Embryos are at the extended germ band stage despite being aged for 10 hours. As well as the normal distribution of transcripts in the thoracic and abdominal segments (compare to Fig. 4), tsh is expressed ectopically in the head region. In A, tsh expression is detected in parts of the procephalon (pc) including the clypeolabrum (cl), mandibular (Md), maxillary (Mx) and labial (Lb) segments. Similar ectopic tsh mRNA accumulations are observed in HSA embryos; in B expression is absent in the labial and maxillary segments. Similar patterns of tsh expression are also observed in HSU embryos. In the populations of HSA or HSU embryos the numbers of cells expressing tsh in the head is variable. For 1 hour heat shock treatments, tsh is expressed almost exclusively in the procephalon of both HSA and HSU embryos. Abbreviations as in Figs 3, 4 and 5 except for the eighth abdominal segment (A8).

Fig. 13.

Transcript accumulation of tsh in HSU (A) and HSA (B) embryos, following three heat shock pulses during embryogenesis. Embryos are at the extended germ band stage despite being aged for 10 hours. As well as the normal distribution of transcripts in the thoracic and abdominal segments (compare to Fig. 4), tsh is expressed ectopically in the head region. In A, tsh expression is detected in parts of the procephalon (pc) including the clypeolabrum (cl), mandibular (Md), maxillary (Mx) and labial (Lb) segments. Similar ectopic tsh mRNA accumulations are observed in HSA embryos; in B expression is absent in the labial and maxillary segments. Similar patterns of tsh expression are also observed in HSU embryos. In the populations of HSA or HSU embryos the numbers of cells expressing tsh in the head is variable. For 1 hour heat shock treatments, tsh is expressed almost exclusively in the procephalon of both HSA and HSU embryos. Abbreviations as in Figs 3, 4 and 5 except for the eighth abdominal segment (A8).

Fig. 14.

The effect of ubiquitous Ubx and Antp expression in the presence and absence of tsh function. Cuticular phenotypes of wild type (A and C) and a tsh null mutation (B and D) carrying a construct with the structural genes for the Antp (HSA) or Ubx (HSU) proteins fused to the control region of a heat shock promoter, following three heat shocks of 25 minutes each during embryogenesis. As described previously, ectopic Ubx protein (A) may replace the thoracic and three head segments with A1 patterns (Gonzalez-Reyes et al., 1990). In addition, prothoracic patterns, as seen here by the ectopic beard (b), may differentiate in the labial segment. Note the presence of sclerotic cuticle in the posterior part of the labium (arrows). Ectopic Antp protein (C) induces mesothoracic denticle belts in the prothorax and in the head positions (see Gibson and Gehring, 1988). In the absence of tsh activity and in the presence of ubiquitous Ubx or Antp protein (B and D), ectopic denticle belts differentiate in the same ectopic positions as wild type and in the prothorax; no denticle belt differentiates normally in tsh mutations in this position (compare to Fig. 2B). In B, note that the prothorax resembles the labial segment in that head cuticle forms in both segments posterior to the denticle belts (arrows). Also note that sclerotic cuticle in the trunk region has been eliminated in the tsh mutations (B and D) and that segments are larger and better defined, compared to tsh mutations (Fig. 2B).

Fig. 14.

The effect of ubiquitous Ubx and Antp expression in the presence and absence of tsh function. Cuticular phenotypes of wild type (A and C) and a tsh null mutation (B and D) carrying a construct with the structural genes for the Antp (HSA) or Ubx (HSU) proteins fused to the control region of a heat shock promoter, following three heat shocks of 25 minutes each during embryogenesis. As described previously, ectopic Ubx protein (A) may replace the thoracic and three head segments with A1 patterns (Gonzalez-Reyes et al., 1990). In addition, prothoracic patterns, as seen here by the ectopic beard (b), may differentiate in the labial segment. Note the presence of sclerotic cuticle in the posterior part of the labium (arrows). Ectopic Antp protein (C) induces mesothoracic denticle belts in the prothorax and in the head positions (see Gibson and Gehring, 1988). In the absence of tsh activity and in the presence of ubiquitous Ubx or Antp protein (B and D), ectopic denticle belts differentiate in the same ectopic positions as wild type and in the prothorax; no denticle belt differentiates normally in tsh mutations in this position (compare to Fig. 2B). In B, note that the prothorax resembles the labial segment in that head cuticle forms in both segments posterior to the denticle belts (arrows). Also note that sclerotic cuticle in the trunk region has been eliminated in the tsh mutations (B and D) and that segments are larger and better defined, compared to tsh mutations (Fig. 2B).

In conclusion, the tsh gene seems to play a key role in determining the identity of the anterior prothorax indepen-dently of the trunk homeotic activities. The specificity of action of the Scr gene seems to be determined by the presence or absence of tsh gene products; when tsh activity is missing, a labial segment is made and when tsh is present a prothoracic segment forms (see Fig. 1).

Independent and similar roles for tsh, Antp and BX-C genes for trunk identity

Our results show that the tsh, Antp and BX-C genes have common and independent functions for trunk identity, which can account for the lack of a clear homeotic trans-formation in tsh mutations. First, homeotic genes together with tsh are required for trunk identity since in their absence (Fig. 10A) ventral trunk identity disappears (i.e. no denti-cle belts). Second, if any one of these genes is active (Figs 2 and 13; Struhl, 1983), denticle belts differentiate showing that this global aspect of trunk identity is under the control of each of these genes independently. Third, high levels of Ubx+ or Antp+ products can partially rescue the tsh phenotype in the trunk (Fig. 14B and D). Finally, any one of these genes repress the lab gene in the trunk (Fig. 6) again indicating a common and overlapping function for these genes.

Normal Antp and BX-C gene function modulates and requires tsh+ activity

It is obvious that, although tsh and homeotic gene products have common functions, they also have unique ones. For example, tsh+ function is required throughout the trunk domain whereas the trunk homeotic genes have restricted functions within it. This difference is linked to our observations that tsh activity is required for normal homeotic gene function throughout the trunk.

The simultaneous presence of tsh+ and homeotic gene activities is always correlated with a specific trunk segmental identity (Fig. 1). We show that the Antp and Ubx proteins, when expressed ectopically and at high levels, lead to activation of tsh transcription in the head, in regions corresponding to the segments thought to be transformed to trunk (Gonzalez-Reyes and Morata, 1991). In the absence of tsh+ activity the ventral denticle belts, even in ectopic positions, are abnormal (Figs 2B, 14B and D), showing that tsh+ activity is required for normal trunk development. In specific domains of the trunk, tsh and particular homeotic genes define unique segmental identities; as described above, tsh and Scr combine for the identity of the protho-rax, tsh and Antp for the mesothorax and so on throughout the posterior trunk (see Fig. 1).

We show that the Antp and Ubx gene products could act as direct or indirect activators of the tsh gene. Very few bona fide targets of these trunk homeotic genes are known (reviewed by Andrew and Scott, 1992). For example, Gould et al. (1990) have described two targets of the Ubx gene and Wagner-Bernholz et al. (1991), have analysed several putative target genes of Antp, one of which may be the spalt (sal) gene. Furthermore, we note that Antp and Ubx proteins are unable to induce ectopic tsh expression in all parts of the head (Fig. 13) suggesting that other factors determine whether or not the tsh gene will be regulated by the homeotic genes. A similar type of differential behaviour has been noted for repression of the sal gene, following ectopic expression of the Antp gene (Wagner-Bernholz et al., 1991); repression of sal expression occurs in the antennal discs but not in the brain or wing discs.

It seems likely that the abdA+ and AbdB+ gene activities are also required for the maintenance of tsh transcription; in the absence of the Antp and BX-C genes at the begin-ning of germ band retraction, tsh transcription is not detected in the posterior compartments of segments in the entire trunk domain (Fig. 5G and H); since these cells dif-ferentiate head identities (Struhl, 1981; Sato et al., 1985), this indicates that the regulatory effects observed between the Antp, the BX-C and tsh genes are functionally relevant.

In conclusion, Antp and BX-C genes regulate and require tsh products for determining the specific identities of the different trunk segments from the posterior prothorax to the eighth abdominal segment.

Co-operative repression of head genes by the trunk homeotic genes

We argue above that tsh, Antp and BX-C genes have independent roles for trunk identity. Clearly when all these gene functions are missing ventral trunk patterns are replaced with structures deriving normally from the procephalon and/or acron (Fig. 10), indicating that tsh, in combination with these homeotic genes, is required for repressing head formation in the trunk tagmata.

Mutations in the tsh gene disrupt the trunk of the embryo and our results suggest that this in part is due to the failure of repression of at least one procephalic gene in this domain, the lab gene (Fig. 6C). As shown in Figs 2B and 12 the ectopic expression of lab in tsh embryos is functionally significant and therefore, a mixture of trunk (Antp and BX-C) and procephalic (lab) homeotic gene products coexist in the trunk, accounting in part for the disrupted thorax and abdomen of tsh mutations. The expression of lab is also repressed by the trunk homeotic genes in homologous dorsal-lateral groups of cells in the trunk (Fig. 6D and E). Sato et al. (1985) showed that Antp BX-C mutants develop head cuticle in dorsal positions. Taken together these observations suggest that the ectopic expression of lab directs these cells into this head developmental path-way in these mutant embryos.

Although the combination of tsh, Antp and BX-C mutations results in a clear transformation of ventral trunk to head (Fig. 10A), the precise segmental origin of this cuticle is in doubt, but it derives from the anterior head. The dere-pression of the lab gene (Fig. 10D and E) is limited to epidermal cells located laterally in the extended germ band embryo whereas the head cuticle is located ventrally in the larva. In our opinion, the combination of trunk homeotic genes is therefore critical for the repression of other head genes in the trunk region. Consistent with this idea is that lab mutations cannot totally rescue the tsh phenotype in the trunk (Figs 2B and 12). Not all of the trunk region is replaced with head patterns (Fig. 10A) in the absence of tsh, Antp and BX-C genes since in the dorsal position trunk patterns develop. It seems probable that in the dorsal position other genes are responsable for trunk development.

Surprisingly, we find that the effect of the absence of tsh and the trunk homeotic genes on the expression of the lab, as well as that of the Dfd and Scr, genes is not simply additive but synergistic (Figs 9-11). These results suggest that tsh and the homeotic genes have the same function in a common set of cells to repress these head genes. The tsh gene codes for a zinc finger protein which therefore is a potential DNA binding factor (Fasano et al., 1991). Thus, it is tempting to speculate that tsh regulates directly, together with the homeotic genes, an overlapping set of genes. Further experiments are required to confirm this idea at the molecular level.

The co-operative effects of tsh with some of the classical trunk homeotic genes is also described for restricting the expression of the Scr and Dfd genes to the gnathal segments. However, the ectopic expression of these genes is not found in all trunk segments (Figs 9B and 11), as is the case for the procephalic gene lab. From a morphological standpoint, the gnathal segments appear to be modified trunk segments. Furthermore, on the basis of the distinct genetic mechanisms involved in establishing the head and trunk domains, several authors have speculated that the gnathocephalon is part of the trunk (Finkelstein and Perrimon, 1991; Cohen and Jürgens, 1991; Gonzalez-Reyes and Morata, 1991). We show that the gnathal genes seem to be distinct, compared to the procephalic genes, in terms of their regulation by the trunk genes (see also Jack et al., 1988). Therefore, the gnathocephalic region can be considered as a distinct morphological unit, with particular genetic properties, compared to the procephalic and the trunk domains.

What is the evolutionary significance of tsh?

Several authors have speculated that a primitive homeobox gene may be implicated in the distinction of trunk from the ends of the embryo and that via duplication and divergence of this primitive function, segmental specialisation occurred (Stuart et al., 1991; Akam et al., 1988). If this hypothesis is true, what might be function of tsh? In Diptera, Snodgrass (1935) has suggested that there was considerable selective pressure on the development of larval features and therefore tsh could have evolved for the development of dipteran larval trunk. In this respect it is noteworthy that deletion of homeotic genes in the more primitive insect Tri-bolium, cause trunk to antennal transformations (Stuart et al., 1991). However, we would like to raise the point that thoracic to antennal transformations in Tribolium are observed in the leg appendages and possibly not in the proximal parts to which the legs are attached. In Drosophila, null mutations of the Antp gene cause a similar transformation of parts of the leg discs but not apparently in the most proximal (or distal) regions (Struhl, 1981). Since tsh is active only in the proximal parts of the leg discs (data not shown), perhaps in evolutionary terms tsh has been con-served for identifying this particular proximal trunk identity in Drosophila as well as in Tribolium.

An alternative hypothesis is that a set of genes distinct from the homeotic ones are defining a ground state (Lewis, 1963), i.e. the identity of segments in the absence of all homeotic gene activities. In this hypothesis, the homeotic genes act to modify the segmental ground state and spe-cialize segmental identities. For the trunk region in Drosophila, this basal identity is prothoracic-like (Struhl, 1983; Sato et al., 1985). We show that tsh+ function is critical for this identity and is modulated by, and required for the normal function of, the Antp and BX-C genes. Our results support the hypothesis that tsh+ (and presumably other gene activities) are contributing to the ground state identity of segments of the Drosophila embryo.

Elements of both these hypotheses could be true. A prim-itive homeobox gene similar to Antp and a zinc finger protein similar to tsh may have evolved in conjunction in the primitive ancestor. In Drosophila, we show that tsh and specific homeotic genes act in concert to determine trunk identity. Evolutionary studies are necessary to distinguish between these possibilities.

We thank T. Kaufman, W. McGinnis, G. Struhl, M. Scott, B. Wakimoto and G. Morata for mutant stocks; T. Kaufman, P. Ingham and A. Martinez-Arias for probes; M. Wilcox for anti-bodies; L. Fasano, N. Coré, B. Jacq and E. Alexandre for dis-cussions; Uwe Waldorf for advice on in situ hybridization; G. Morata for advice on heat shocking; P. Weber for his help with the photography and especially J. Pradel, B. Jacq and B. Oliver for their suggestions on the manuscript. This work was supported by the CNRS and by Ministère de la Recherche et de la Tech-nologie (M.R.T.) and the Association pour la Recherche contre le Cancer (A.R.C.) fellowships to L. R.

Akam
,
M.
and
Martinez-Arias
,
A.
(
1985
).
The distribution of Ultrabithorax transcripts in Drosophila embryos
.
EMBO J
.
4
,
1689
1700
.
Akam
,
M.
(
1987
).
The molecular basis for metameric pattern in the Drosophila embryo
.
Development
101
,
1
22
.
Akam
,
M.
,
Dawson
,
I.
and
Tear
,
G.
(
1988
).
Homeotic genes and the control of segment diversity
.
Development Supplement
104
,
123
133
.
Andrew
,
D.
and
Scott
,
M.
(
1992
).
Downstream of the homeotic genes
.
The New Biologist
4
,
5
15
.
Bermingham
,
J.
,
Martinez-Arias
,
A.
,
Petitt
,
M.
and
Scott
,
M.
(
1990
).
Different patterns of transcription from the two Antennapedia promoters during Drosophila embryogenesis
.
Development
109
,
553
566
.
Boncinelli
,
E.
,
Somma
,
R.
,
Acampora
,
D.
,
Pannese
,
M.
,
D’Esposito
,
M.
and
Simeone
,
A.
(
1988
).
Organization of the human homeobox genes
.
Hum. Reprod
.
3
,
880
886
Campos-Ortega
,
J.
and
Hartenstein
,
V.
(
1985
).
In: The Embryonic Development of Drosophila melanogaster
.
Berlin
:
Springer-Verlag
.
Carroll
,
S.
,
Laymon
,
R.
,
McCuthcheon
,
M.
,
Riley
,
P.
and
Scott
,
M.
(
1986
).
The localization and regulation of Antennapedia protein expression in Drosophila embryos
.
Cell
47
,
113
122
.
Casanova
,
J.
(
1989
).
Mutations in the spalt gene of Drosophila cause ectopic expression of Ultrabithorax and Sex combs reduced
.
Wilhelm Roux’s Arch. Dev. Biol
.
198
,
137
140
.
Chadwick
,
R.
and
McGinnis
,
W.
(
1987
).
Temporal and spatial distribution of transcripts from the Deformed gene of Drosophila
.
EMBO J
6
,
779
789
.
Cohen
,
S.
(
1990
).
Specification of limb development in the Drosophila embryo by positional cues from segmentation genes
.
Nature
343
,
173
177
.
Cohen
,
S.
and
Jürgens
,
G.
(
1991
).
Drosophila headlines
.
Trends Genet
.
7
,
267
272
.
Coleman
,
K
,
Poole
,
S.
,
Wier
,
M.
,
Soeller
,
W.
and
Kornberg
,
T.
(
1987
).
The invected gene of Drosophila: sequence analysis and expression studies reveal a close kinship to the engrailed gene
.
Genes Dev
.
1
,
19
28
.
Diederich
,
R.
,
Merrill
,
V.
,
Pultz
,
M.
and
Kaufman
,
T.
(
1989
).
Isolation, structure and expression of labial, a homeotic gene of the Antennapedia complex involved in Drosophila head development
.
Genes Dev
3
,
399
414
.
Duboule
,
D.
and
Dolle
,
P.
(
1989
).
The structural and functional organization of the murine Hox gene family resembles that of Drosophila homeotic genes
.
EMBO J
.
8
,
1497
1505
Duncan
,
I.
(
1987
).
The Bithorax complex
.
A. Rev. Genet
.
21
,
285
319
.
Fasano
,
L.
,
Röder
,
L.
,
Coré
,
N.
,
Alexandre
,
E.
,
Vola
,
C.
,
Jacq
B.
and
Kerridge
,
S.
(
1991
).
The teashirt gene is required for the development of Drosophila embryonic trunk segments and encodes a protein with widely spaced zinc finger motifs
.
Cell
64
,
63
79
.
Finkelstein
,
R.
and
Perrimon
,
N.
(
1991
).
The molecular genetics of head development in Drosophila melanogaster
.
Development
112
,
899
912
.
Garcia-Bellido
,
A.
(
1975
).
Genetic control of wing disc development in Drosophila
.
CIBA Foundation Symposium
29
,
161
182
.
Ghysen
,
A.
and
O’Kane
,
C.
(
1989
).
Neural enhancer-like elements as specific cell markers in Drosophila
.
Development
105
,
35
52
.
Ghysen
,
A.
,
Dambly-Chaudiere
,
C.
,
Aceves
,
E.
,
Jan
,
L.
and
Jan
,
Y.
(
1986
).
Sensory neurons and peripheral pathways in Drosophila embryos
.
Wilhelm Roux’s Arch. Dev. Biol
.
195
,
281
289
.
Gibson
,
G.
and
Gehring
,
W.
(
1988
).
Head and thoracic transformations caused by ectopic expression of Antennapedia during Drosophila development
.
Development
102
,
657
675
.
Gibson
.,
G.
,
Schier
,
A.
,
Lemotte
,
P.
and
Gehring
,
W.
(
1990
).
The specifities of Sex combs reduced and Antennapedia are defined by a distinct portion of each protein that includes the homeodomain
.
Cell
62
,
1087
1103
.
Glaser
,
L.
and
Shilo
,
B-Z.
(
1991
).
The Drosophila FGF-R homolog is expressed in the embryonic tracheal system and appears to be required for directed tracheal cell extension
.
Genes Dev
.
5
,
697
705
.
Gonzalez-Reyes
,
A.
and
Morata
,
G.
(
1990
).
The developmental effect of overexpressing a Ubx product in Drosophila embryos is dependant on its interactions with other homeotic products
.
Cell
61
,
515
522
.
Gonzalez-Reyes
,
A.
,
Urquia
,
R.
,
Gehring
,
W.
,
Struhl
,
G.
and
Morata
,
G.
(
1990
).
Are cross-regulatory interactions between homeotic genes functionally significant?
Nature
334
,
78
80
.
Gonzalez-Reyes
,
A.
and
Morata
,
G.
(
1991
).
Organization of the Drosophila head as revealed by ectopic expression of the Ultrabithorax product
.
Development
113
,
1459
1471
.
Gould
,
A.
,
Brookman
,
J.
,
Strutt
,
D.
and
White
,
R.
(
1990
).
Targets of homeotic gene control in Drosophila
.
Nature
348
,
308
312
.
Graham
,
A.
,
Papalopulu
,
N.
and
Krumlauf
,
R.
(
1989
).
The murine and Drosophila homeobox gene complexes have common features of organization and expression
.
Cell
57
,
367
378
.
Hafen
,
E.
,
Levine
,
M.
and
Gehring
,
W.
(
1984
).
Regulation of Antennapedia transcript distribution by the bithorax complex in Drosophila
.
Nature
307
,
287
289
Harding
,
K.
,
Wedeen
,
C.
,
McGinnis
,
W.
and
Levine
,
M.
(
1985
).
Spatially regulated expression of homeotic genes in Drosophila
.
Science
233
,
1236
1242
.
Hazelrigg
,
T.
and
Kaufman
,
T.
(
1983
).
Revertants of dominant mutations associated with the Antennapedia gene complex in Drosophila melanogaster: cytology and genetics
.
Genetics
105
,
581
600
.
Ingham
, (
1988
).
The molecular genetics of embryonic pattern formation in Drosophila
.
Nature
335
,
25
33
.
Irish
,
V.
,
Martinez Arias
,
A.
and
Akam
,
M.
(
1989
).
Spatial regulation of the Antennapedia and Ultrabithorax homeotic genes during Drosophila early development
.
EMBO J
.
8
,
1527
1537
.
Jack
,
T.
,
Regulski
,
M.
and
McGinnis
,
W.
(
1988
).
Pair-rule segmentation genes regulate the expression of the homeotic selector gene, Deformed
.
Genes Dev
.
2
,
635
645
.
Jürgens
,
G.
,
Lehman
,
R.
,
Schardin
,
M.
and
Nüsslein-Volhard
,
C.
(
1986
).
Segmental organization of the head in the embryo of Drosophila melanogaster
.
Wilhelm Roux’s Arch. Dev. Biol
.
195
,
359
377
.
Jürgens
,
G.
(
1988
).
Head and tail development of the Drosophila embryo involves spalt, a novel homeotic gene
.
EMBO J
.
7
,
189
196
.
Jürgens
,
G.
and
Weigel
,
D.
(
1988
).
Terminal versus segmental development in the Drosophila embryo: the role of the homeotic gene fork head
.
Roux Arch. Dev. Biol
.
197
,
345
354
.
Kaufman
,
T.
,
Seeger
,
M.
and
Olsen
,
G.
(
1990
).
Molecular organization of the Antennapedia gene complex of Drosophila melanogaster
.
Adv. in Genetics
pp
309
362
.
Kerridge
,
S.
and
Morata
,
G.
(
1982
).
Developmental effects of some newly induced Ultrabithorax alleles of Drosophila
.
J. Embryol. Exp. Morphol
.
68
,
211
234
.
Kuroiwa
,
A.
,
Kloter
,
U.
,
Baumgertner
,
P.
and
Gehring
,
W.
(
1985
).
Cloning of the Sex combs reduced gene in Drosophila and in situ localization of its transcripts
.
EMBO J
.
4
,
3757
3764
.
Kuziora
,
M. A.
and
McGinnis
,
W.
(
1988
).
Different transcripts of the Drosophila Abdominal-B gene correlate with distinct genetic subfunctions
.
EMBO J
.
7
,
3233
3244
Lewis
,
E.
(
1963
).
Genes and developmental pathways
.
Am. Zool
3
,
33
56
.
Lewis
,
E.
(
1978
).
A gene complex controlling segmentation in Drosophila
.
Nature
276
,
565
570
Lohs-Schardin
,
M.
,
Cremer
,
C.
and
Nüsslein-Volhard
,
C.
(
1979
).
A fate map of the larval epidermis of Drosophila melanogaster: localized cuticle defects following irradiation of the blastoderm with an ultraviolet laser microbeam
.
Devl. Biol
.
73
,
239
255
.
Mahaffey
,
J.
and
Kaufman
,
T.
(
1987
).
Distribution of the Sex combs reduced gene products in Drosophila melanogaster
.
Genetics
117
,
51
60
.
Martinez-Arias
,
A.
and
Lawrence
,
P.
(
1985
).
Parasegments and compartments in the Drosophila embryo
.
Nature
313
,
639
642
.
Martinez-Arias
,
A.
(
1986
).
The Antennapedia gene is required and expressed in parasegments 4 and 5 of the Drosophila embryo
.
EMBO J
.
5
,
135
141
Martinez-Arias
,
A.
,
Ingham
,
P.
,
Scott
,
M.
and
Akam
,
M.
(
1987
).
The spatial and temporal deployment of Dfd and Scr transcripts throughout the development of Drosophila
.
Development
100
,
673
683
.
McGinnis
,
N.
,
Kuziora
,
M.
and
McGinnis
,
W.
(
1990
).
Human Hox-4.2 and Drosophila Deformed encode similar regulatory specificities in Drosophila embryos and larvae
.
Cell
63
,
969
976
.
McGinnis
,
W.
,
Levine
,
M.
,
Hafen
,
E.
,
Kuroiwa
,
A.
and
Gehring
,
W.
(
1984
).
A conserved DNA sequence in homeotic genes of the Drosophila Antennapedia and Bithorax complexes
.
Nature
308
,
428
433
.
Merrill
,
V.
,
Diederich
,
R.
,
Turner
,
F.
and
Kaufman
,
T.
(
1989
).
A genetic and developmental analysis of mutations in labial, a gene necessary for proper head formation in Drosophila melanogaster
.
Devl. Biol
.
135
,
376
391
.
Pattatucci
,
A.
,
Otteson
,
D.
and
Kaufman
,
T.
(
1991
).
A functional and structural analysis of the Sex combs reduced locus of Drosophila melanogaster
.
Genetics
129
,
423
441
.
Riley
,
P.
,
Carroll
,
S.
and
Scott
,
M.
(
1987
).
The expression of the Sex combs reduced protein in Drosophila embryos
.
Genes Dev
.
1
,
716
730
.
Sanchez-Herrero
,
E.
,
Vernos
,
I.
,
Marco
,
R.
and
Morata
,
G.
(
1985
).
Genetic organization of the Drosophila bithorax complex
.
Nature
.
313
,
108
113
Sanchez-Herrero
,
E.
and
Crosby
,
M.
(
1988
).
The Abdominal-B gene of Drosophila melanogaster : overlapping transcripts exhibit two different spatial distributions
.
EMBO J
.
7
,
2163
2173
.
Schneuwly
,
S.
and
Gehring
,
W.
(
1985
).
Homeotic transformation of thorax into head: developmental analysis of a new Antennapedia allele in Drosophila melanogaster
.
Devl. Biol
.
108
,
377
386
.
Sato
,
T.
,
Hayes
,
P.
and
Denell
,
R.
(
1985
).
Homeosis in Drosophila: Roles and spatial patterns of expression of the Antennapedia and Sex combs reduced loci in embryogenesis
.
Devl. Biol
.
111
,
171
192
Scott
,
M.
and
Weiner
,
A.
(
1984
).
Structural relationships among genes that control development: sequence homology between the Antennapedia, Ultrabithorax, and fushi tarazu loci of Drosophila
.
Proc. Natl. Acad. Sci. U.S.A
.
81
,
4115
4119
.
Snodgrass
,
R.
(
1935
).
Principles of Insect Morphology
.
McGraw-Hill
,
New York
.
Struhl
,
G.
(
1981
).
A homeotic mutation transforming leg to antenna in Drosophila
.
Nature
292
,
335
338
.
Struhl
,
G.
(
1983
).
Role of the esc+ gene product in ensuring the selective expression of segment specific homeotic genes in Drosophila
.
J. Embryol. exp. Morph
.
76
,
297
331
.
Stuart
,
J.
,
Brown
,
S.
,
Beeman
,
R.
and
Denell
,
R.
(
1991
).
A deficiency of the homeotic complex of the beetle Tribolium
.
Nature
350
,
72
74
.
Wagner-Bernholz
,
J.
,
Gibson
,
G.
,
Schuh
,
R.
and
Gehring
,
W.
(
1991
).
Identification of target genes of the homeotic gene Antennapedia by enhancer detection
.
Genes Dev
.
5
,
2467
2480
.
Wakimoto
,
B.
and
Kaufman
,
T.
(
1981
).
Analysis of larval segmentation phenotypes associated with the Antennapedia gene complex of Drosophila melanogaster
.
Devl. Biol
81
,
51
64
.
Wirz
,
J.
,
Fessler
,
L.
and
Gehring
,
W.
(
1986
).
Localization of the Antennapedia protein in Drosophila embryos and imaginal discs
.
EMBO J
.
5
,
3327
3334
.
Wright
,
T.
,
Hodgetts
,
R.
and
Sherald
,
A.
(
1976
).
The genetics of dopa-decarboxylase in Drosophila melanogaster. I. Isolation and characterization of deficiencies that delete the dopa-decarboxylase dosage-sensitive region and α-methyl-dopa-hypersensitive locus
.
Genetics
84
,
267
285
.