Both Proboscipedia (PB) and Sex Combs Reduced (SCR) activities are required for determination of proboscis identity. Here we show that simultaneous removal of PB and SCR activity results in a proboscis-to-antenna transformation. Dominant negative PB molecules inhibit the activity of SCR indicating that PB and SCR interact in a multimeric protein complex in determination of proboscis identity. These data suggest that the expression pattern of PB and SCR and the ability of PB and SCR to interact in a multimeric complex control the determination of four adult structures. The absence of PB and SCR expression leads to antennal identity; expression of only PB leads to maxillary palp identity; expression of only SCR leads to tarsus identity; and expression of both PB and SCR, which results in the formation of a PB-SCR-containing complex, leads to proboscis identity.

However, the PB-SCR interaction is not detectable in vitro and is not detectable genetically in the head region during embryogenesis, indicating the PB-SCR interaction may be regulated and indirect. This regulation may also explain why ectopic expression of SCRQ50K and SCR do not result in the expected transformation of the maxillary palp to an antennae and proboscis, respectively.

Previous analysis of the requirements of SCR activity for adult pattern formation has shown that ectopic expression of SCR results in an antenna-to-tarsus transformation, but removal of SCR activity in a clone of cells does not result in a tarsus-to-arista transformation. Here we show in five independent assays the reason for this apparent contradictory requirement of SCR activity in tarsus determination. SCR activity is required cell nonautonomously for tarsus determination. Specifically, we propose that SCR activity is required in the mesodermal adepithelial cells of all leg imaginal discs at late second/early third instar larval stage for the synthesis of a mesoderm-specific, tarsus-inducing, signaling factor, which after secretion from the adepithelial cells acts on the overlaying ectodermal cells determining tarsus identity.

This study characterizes a combinatorial interaction between two HOX proteins; a mechanism that may have a major role in patterning the anterior-posterior axis of other animals.

The conserved Hox genes function in laying out the body plan along the anterior-posterior axis of many, and maybe most, animal phyla (Carroll, 1995; Slack et al., 1993). All Hox genes encode a protein that contains the DNA-binding protein domain, the homeodomain (HD) (McGinnis and Krumlauf, 1992). The activity of the Hox genes were initially identified in Drosophila by the phenotypes produced by loss-of-function or gain-of-function alleles. A mutation in a Drosophila Hox gene results in the transformation of one body part into another (Lewis, 1978; Kaufman et al., 1990). Ironically, most of the information on how the HOX proteins work has come from the analysis of the role of Ultrabithorax (UBX) in gut development and not their role in the determination of segmental identity (Capovilla et al., 1994). The reason for this is the lack of knowledge about what genes are specifically regulated by HOX proteins during determination of segmental identity, and the lack of knowledge about what activities are required for determination of segmental identity (Andrews and Scott, 1992).

In gut development, UBX activity is required in the visceral mesoderm for the synthesis of Decapentaplegic (DPP) (Capovilla et al., 1994). Using a change of specificity mutation, UBX has been shown to act directly as a transcriptional activator of dpp expression (Capovilla et al., 1994). UBX binds the dpp regulatory region via the HD. However, UBX alone is unable to recognize with high affinity the cis-regulatory dpp sequences. UBX requires the cofactor Extradenticle (EXD) (Chan et al., 1994). Both EXD and UBX activities are required for determination of the correct segmental identity of parasegments 5 and 6 (Peifer and Wieschaus, 1990). EXD interacts with UBX and other HOX proteins through a hexapeptide sequence of amino acids found in most HOX products (Johnson et al., 1995; Chan et al., 1996). This UBX-EXD interaction is considered generally important because of the properties of HOX activity. Individual HOX proteins perform very distinct and specific functions during development. Therefore, it is very surprising the HDs of the various HOX proteins recognize basically the same DNA-binding sites (Gehring et al., 1994), but conversely that the HDs of the various HOX proteins determine much of the specificity of their function (Chan and Mann, 1993; Furukubo-Tokunaga et al., 1993; Gibson et al., 1990; Mann and Hogness, 1990; Zeng et al., 1993). The interaction with EXD and maybe other cofactors allows an explanation of these two properties of HOX activity.

An observation that can also be explained by HOX cofactor interactions is the presence of more segments with unique identity than Hox genes to determine their identity: 14 segments and 8 Hox genes. It has been suggested that HOX proteins interact with one another combinatorially in determination of positional value along the anterior-posterior axis, particularly in vertebrate systems, where there is extensive overlap of the expression domains of the HOX proteins (Hunt and Krumlauf, 1991). However, there is no evidence in vertebrates that this actually occurs. The embryonic expression pattern of the HOX proteins that pattern the thoracic and abdominal segments of Drosophila generally do not result in coexpression of two HOX proteins in the same cell (Peifer et al., 1987). There is extensive coexpression of Proboscipedia (PB) with Deformed (DFD) and Sex Combs Reduced (SCR) in the head (Pultz et al., 1988; Chadwick and McGinnis, 1987; Kuroiwa et al., 1985; Mahaffey and Kaufman, 1987). During embryogenesis, PB and DFD are coexpressed in the maxillary segment, and PB and SCR are coexpressed in the labial segment. In the imaginal discs, PB and SCR are coexpressed in the labial imaginal disc (Randazzo et al., 1991; Pattatucci and Kaufman, 1991; Glickman and Brower, 1988).

A second interesting point about HOX function is the function of the genes that they are known to regulate. Although HOX proteins act as transcription factors regulating gene expression cell autonomously, there are good examples of HOX-regulated genes being components of cell nonautonomous developmental pathways. Ultrabithorax (UBX) activity in the visceral mesoderm binds to the dpp promoter (Capovilla et al., 1994). The product of the dpp gene is a secreted factor with significant amino acid similarity to tumor growth factor β (Padgett et al., 1987). In the endoderm, DPP is required for expression of LAB and the formation of the second gut constriction (Immergluck et al., 1990).

Here we show that PB and SCR, which are coexpressed in the labial segment, interact in the determination of the labial structure, the proboscis. We also show that SCR activity is required cell nonautonomously for tarsus determination.

Stock construction

For a description of the genetic markers and balancer chromosomes used here, see Lindsley and Zimm, 1992. The flies were maintained on standard Drosophila media supplemented with Baker’s yeast. The various chromosomes described in this study were constructed in standard Drosophila crossing schemes (Table 1).

Table 1.

Table of stocks

Table of stocks
Table of stocks

FLP-mediated mitotic recombination

The crosses used for generating the FLP-mediated recombination events are as follows: Scr7 clones in the proboscis, stocks DJ102 and APS112 (Table 1) were crossed; pb27Scr2 clones in the proboscis, stocks GS3 and AP201 were crossed; pb27 clones in the proboscis, stocks GS3 and APS202 were crossed; pb27Scr2 clones in apb27/pb20 mutant background, stocks APS121 and DJ402 were crossed; pb27 clones in a pb27/pb20 mutant background, stocks APS121 and DJ403 were crossed. The first instar larval progeny of these crosses were heat shocked for 30–45 minutes at 36.5°C. Progeny with the appropriate genotype were screened for clones.

Construction of the heat-shock promoter/ Scr and pb fusion genes

The hsp/Scr and hsp/Scpq50K fusion genes were constructed as follows. The unique XbaI site of pNMT4 was deleted by filling in the XbaI site of pNMT4 with deoxynucleotides and ligating, yielding pNMT4AX (Schneuwly et al., 1987). pPL20 was digested with ClaI and XbaI, and this DNA was used in subsequent PCRs (LeMotte et al., 1989). All PCR reactions used the high-fidelity, heat-stable DNA polymerase, Pfu. To construct the two hsp/Scra/b fusion genes, the following PCRs were performed. Primers APS-5 (ACGCGCGGC- CGCACGACAACGACCCCTGGCT) and APS-15 (ACGCGCGGC- CGCTCTAGAGATCGGGCCACACTCTCTAC) were used in a PCR reaction using pPL20 as the template (LeMotte et al., 1989). The PCR product was digested with NotI and gel isolated. This NotI fragment, which encodes the N-terminal portion of SCR, was inserted into the unique NotI site of pNMT4AX, yielding pNMT4AXNScr. Primers APS-16 (TGTGGCCCGATCTCTCTAGACAAGGGCGTCC) and APS-17 (TCGCTCTAGACAGTACCCGAAAAGTGCCAC) were used in a PCR and the resulting fragment was digested with XbaI and gel isolated. This fragment was inserted into the unique XbaI site of pNMT4AXNScr. Insertion of this XbaI fragment creates a full-length Scr gene, which is the same as wild-type Scr except for the change of amino acid 226 and 227 from Asn Lys to Arg Ser. Two identical pP{hsp/Scr, ry+} plasmids were constructed from completely independent PCR reactions. Scra has the expected DNA sequence.

To construct the two pP{hsp/Scp250K, ry+} genes, the Q50K change was introduced at the last step described for the construction of pP{hsp/Scr, ry+}. A PCR was performed using primers APS-10 (TGCGCCGGTTCTTGAACCAGATCTT) and APS-16, and a PCR was performed with APS-9 (AAGATCTGGTTCAAGAACCG- GCGCA) and APS-17. Gel-isolated fragments from both of these PCR reactions were used in a PCR with primers APS-16 and APS- 17. The resulting fragment, which now contains the Q50K mutation, was digested with XbaI and inserted into the XbaI site of pNMT4AXNScr, yielding pP{hsp/ScrQ50K, ry+}. Two completely independent PCRs were used to construct hsp/ScrQ50Ka/b. The Q50K change was confirmed by DNA sequencing.

At the beginning of this project, no full-length pb cDNA existed (Cribbs et al., 1992). Using PCR, we amplified a pb-coding region fragment from a pupal total cDNA population. cDNA made from pupal poly(A)+ RNA with AMV reverse transcriptase was used in a PCR with primers APS-1 (ACGCGCGGCCGCAATTGAAATA- GAAAGAATC) and APS-2 (ACGCGCGGCCGCGACACTGAATA- GAAATACA). The product was reamplified with APS-1 and APS-2, cut with NotI and the fragment gel isolated. This fragment was inserted into the unique NotI site of pNMT4. This yielded pP{hsp/pba, ry+} and pP{hsp/pbb, ry+}. Both pba and pbb cDNAs are derived from the 2-μ-4b spliced form of PB mRNA (Cribbs et al., 1992). To construct the pbQ50K genes, the initial PCR product made with APS- 1 and APS-2 was used in two PCRs. One PCR used primers APS-1 and APS-4 (GCGGCGGTTTTTGAACCAAACTT), and the other used the primers APS-2 and APS-3 (AAGTTTGGTTCAAAAAC- CGCCGC). The products of these two reactions were purified and used in a PCR with the primers APS-1 and APS-2. The full-length product isolated from this PCR contains the Q50K change. The product was digested with Not! and inserted into pNMT4 yielding two constructs pP{hsp/pbQ50Ka, ry+} and pP {hsp/pbQ50Kb, ry+}. The Q50K change was confirmed by DNA sequencing. The pbQ50Ka and pbQ50Kb cDNAs are derived from the 2-μ-4a and 2-μ-4b spliced forms of PB mRNA, respectively (Cribbs et al., 1992). The hsp/Scr and hsp/pb fusion genes were introduced into the Drosophila genome by P-element transformation (Rubin and Spradling, 1982).

FLP-mediated ectopic expression (Flip-out) constructs

We constructed a FLP-mediated ectopic expression vector (flip-out) that contained a unique Not! site into which we inserted our various Not! fragments containing Scr- and pb-coding regions. The components that we used to assemble this vector were kindly supplied by G. Struhl (Struhl and Basler, 1993). The 2.1 kb BaraHI-Sal! fragment from J35, which carries a single FRT site, a transcriptional termination region and the unique NotI site, was isolated and inserted between the Bgl!I and Xho! sites of pW6 (Klemenz et al., 1987), yielding pW6Not!. The 2.6 kb EcoRI-Kpnl fragment carrying the Tub αl promoter was derived from KB700 by digesting KB700 with Not!, filling in the ends using DNA polymerase and attaching EcoR! linkers. This fragment was isolated and digested with EcoRI and Kpnl, and the resulting fragment was inserted between the unique Kpnl and EcoRI sites of pW6NotI, yielding pP{w+, Notl >Tub α1}. Before the last step, the unique Not! site of J35 was deleted by filling it in and religating, yielding J35ANotI. The Xbα1 fragment carrying the y gene, flanked by two partial FRT sites, was inserted into the unique Xbα1 of pP{w+, NotI> Tubal} resulting in pP{w+, NotT>y+>Tubα1}. !nto the unique Not! site of this vector, were inserted NotI fragments carrying pba, pbQ50Ka, pbQ50Kb, Scra, Scrb and ScrQ50Ka. The P-element carrying these constructs was inserted into the Drosophila genome by P-element-mediated transformation (Rubin and Spradling, 1982).

The y+ gene between the FRT sites was excised by crossing any one of the established lines with males of stocks GS1 or DJ400. The progeny of these crosses were heat shocked for 20 minutes to 1 hour at 36.5°C to induce expression of FLP protein from the heat-shock promotor/FLP fusion gene. For the assay of rescue of the pb phenotype by PBa and PBQ50Kb, stock APS303 was crossed with APS301 and APS303 was crossed with APS302, respectively (Table 1). The progeny was heat shocked for 30 minutes at 36.5°C during embryogenesis 0 and 24 hours AEL.

Cuticle preparations

First instar larvae and dissected adult heads were mounted in 50% Hoyer’s 50% lactic acid (Wieschaus and Nüsslein-Volhard, 1986).

In situ hybridization

We found that a good salivary gland marker can be made from the Sp6 promoter of the pSPT19-Neo plasmid supplied as a control in many kits. The in situ hybridization was performed essentially as described in Tautz and Pfeifle (1989), except that the embryos were not digested with proteinase K, but were incubated in acetone for 5 minutes on ice.

Determination of the temperature-sensitive period of the pb1 allele

Flies from the stocks 2097 and 2178 were crossed. The time that the hatched first instar larvae were collected was defined as 24 hours AEL. The vials were placed at 18°C or 28.5°C and, at specific times, vials were either shifted up or down. The time of pupation at 18°C and 28.5°C was used to normalize the results to 25°C.

Clones of pb Scr null mutant proboscis cells adopt antennal identity

Previous genetic observations suggest that PB and SCR activities may interact (Fig. 1). (i) The proboscis of a null pb mutant is transformed into a pair of tarsi (Kaufman, 1978), and (ii) these alleles also result in reduced maxillary palps, which some investigators have interpreted as a transformation of the maxillary palps into antennae (Fig. 1A) (Kaufman, 1978). (iii) Ectopic expression of PB from a heat-shock promoter/pb fusion gene (Cribbs et al., 1995), or in a small clone of cells from a Tubulin al (Tub al) promoter/pb fusion gene (Fig. 1B) result in the transformation of the antennae into maxillary palps. (iv) Ectopic expression of SCR from a heat-shock proraoter/Scr fusion gene results in the transformation of the aristae into tarsi (Fig. 1C) (Gibson et al., 1990). (v) The proboscis of semilethal loss-of-function Scr alleles, and clones of Scr null mutant cells in the proboscis adopt maxillary palp identity (Fig. 1D) (Pattatucci et al., 1991; Struhl, 1982).

Fig. 1.

The phenotypes of the individual loss- or gain-of-function of PB or SCR activity. (A) Loss of PB activity in a pb27/pb20 hemizygous individual. (B) Gain of PBa activity in the antenna. This is an example of a FLP-mediated ectopic expression experiment using the P{w+, pba>y+> Tubal}B line. (C) Gain-of-SCR activity in the antenna. (D) Loss of SCR activity in the proboscis. (A,C) The arrows indicate a claw, the arrowheads indicate a pulvilli, and beside the diamond are the sex combs. (B,D) The filled circles indicate maxillary palp-like bristles, and the filled square marks the thickened portion of the arista.

Fig. 1.

The phenotypes of the individual loss- or gain-of-function of PB or SCR activity. (A) Loss of PB activity in a pb27/pb20 hemizygous individual. (B) Gain of PBa activity in the antenna. This is an example of a FLP-mediated ectopic expression experiment using the P{w+, pba>y+> Tubal}B line. (C) Gain-of-SCR activity in the antenna. (D) Loss of SCR activity in the proboscis. (A,C) The arrows indicate a claw, the arrowheads indicate a pulvilli, and beside the diamond are the sex combs. (B,D) The filled circles indicate maxillary palp-like bristles, and the filled square marks the thickened portion of the arista.

That both PB and SCR activities are required for determination of proboscis identity, and that individual expression of PB and SCR activities determine maxillary palp and tarsus identities, respectively, suggests a simple model for determination of four developmental identities. We propose that the expression patterns of PB and SCR determine antenna, maxillary palp, tarsus and proboscis identities. Specifically, the absence of PB and SCR expression, the default state, leads to antennal identity, expression of only PB activity leads to maxillary palp identity, expression of only SCR activity leads to tarsus identity, and expression of both PB and SCR activities leads to proboscis identity. A prediction of this simple model is that a proboscis primordial cell that is unable to express either PB or SCR will adopt antennal identity.

We have tested this prediction by constructing a chromosome with a Flip recombinase target site (FRT), and the null pb27 allele and the null Scr2 allele. This chromosome was used in a combined Flip-mediated mitotic recombination (Golic, 1991; Xu and Rubin, 1993) and minute technique experiment (Basler and Struhl, 1994) to generate clones of proboscis cells that were homozygous for the pb27Scr2 alleles. These mutant proboscis cells develop with antennal fate (Fig. 2A), confirming the prediction of the simple model. The transformation is not just the transformation of the proboscis into an arista (Fig. 2A), but also a third antennal segment with the appropriate sensilla (Fig. 2B). Also observed, were thick thorny bristles (zahnborsten) indicative of a second antennal segment transformation (Fig. 2B). Although we do not observe defined first and second antennal segments in the pb27Scr2 clones of cells, we nevertheless think that pb27Scr2 mutant proboscis cells represent a transformation to a true complete antenna.

Fig. 2.

(A,B) A pb27Scr2 mosaic analysis and (C,D) ectopic expression of PBQ50Ka transforms the tarsal mouthparts of a pb null mutant to arista. (A,B) Clones ofpb27Scr2 cells in the proboscis. (A) Transformation to an arista and (B) transformation to a third antennal segment with second antennal segment bristle. (C) The aristal mouthparts of a pb27, P{hsp/pbQ50Ka, ry+}/pb20 individual from a cross of APS110 and 2172 that had been heat shocked at 36.5°C for 15-30 minutes between 67 and 72 hours AEL. (D) The same head shown in C but viewed from below to demonstrate the complete lack of pseudotrachea. Asterisk (*), aristal transformations; filled squares, the sensilla trichodea; arrow, a thorny second antennal segment bristle.

Fig. 2.

(A,B) A pb27Scr2 mosaic analysis and (C,D) ectopic expression of PBQ50Ka transforms the tarsal mouthparts of a pb null mutant to arista. (A,B) Clones ofpb27Scr2 cells in the proboscis. (A) Transformation to an arista and (B) transformation to a third antennal segment with second antennal segment bristle. (C) The aristal mouthparts of a pb27, P{hsp/pbQ50Ka, ry+}/pb20 individual from a cross of APS110 and 2172 that had been heat shocked at 36.5°C for 15-30 minutes between 67 and 72 hours AEL. (D) The same head shown in C but viewed from below to demonstrate the complete lack of pseudotrachea. Asterisk (*), aristal transformations; filled squares, the sensilla trichodea; arrow, a thorny second antennal segment bristle.

Genetic evidence with PBQ50K molecules that PB and SCR interact in a multimeric complex

PB and SCR are nuclear-localized, homeodomain-containing proteins suggesting that they both function as transcriptional regulators (Pultz et al., 1988; Glickman and Brower, 1988). If this is the case, two mechanisms for the role of PB and SCR in proboscis determination may be proposed. In both models, PB regulates a set of PB-regulated genes that, when expressed in isolation, determine maxillary palp identity. Similarly, SCR regulates a set of SCR-regulated genes that, when expressed in isolation, determine tarsal identity. In one model, expression of both sets of PB-regulated genes and SCR-regulated genes in the same cell determines proboscis identity. In a second model, expression of PB and SCR proteins in the same cell leads to formation of a PB-SCR-containing, heteromeric, protein complex that regulates a novel set of genes that determines proboscis identity, the PB-SCR-regulated genes. If the second model is correct, it should be possible to design dominant negative PB and SCR molecules that will inhibit the activity of the other.

In choosing the mutations used for the designed dominant negative PB and SCR molecules, we thought the properties of previously described change of DNA-binding specificity mutants made them ideal candidates (Percival-Smith et al., 1990; Hanes and Brent, 1991; Schier and Gehring, 1992). Both PB and SCR have a glutamine at position 50 of the homeo- domain (HD), and we have created pb and Scr genes where this glutamine has been substituted for a lysine. This change is expected to change the DNA-binding specificity of PB and SCR from Antennapedia class DNA-binding sites to Bicoid class DNA-binding sites, as has been extensively documented for other HDs (Percival-Smith et al., 1990; Capovilla et al., 1994). The result of this change would be that the PBQ50K and SCRQ50K molecules, as well the PBQ50K SCR and PB- SCRQ50K-containing complexes, would not only have diminished affinity for their normal interaction site, but would also have an increased affinity for another set of sites dragging the PBQ50K and SCRQ50K molecules, as well as the PBQ50K SCR and PB-SCRQ50K-containing complexes, away from their normal site of interaction.

We fused the pbQ50Ka gene behind a heat-shock promoter, and introduced this hsp/pbQ50Ka fusion gene into the Drosophila genome by P-element-mediated transformation. Induction of PBQ50Ka expression does not affect wild-type proboscis formation. This inability to produce a phenotype may be due to PBQ50Ka not being vastly overexpressed, as is the case with the dominant negative approach in yeast (Hall and Johnson, 1987; Herskowitz, 1987). To avoid this problem, we assayed a situation where SCR activity was alone to complex with PBQ50Ka, such that the PBQ50Ka molecule did not have to compete with the wild-type PB molecule for complex formation with SCR. In the simple model proposed, SCR activity in a pb mutant is left ON by itself in the proboscis precursor cells to determine tarsus identity. If PBQ50Ka has the properties of a dominant negative molecule, we would expect that the tarsal mouthparts of a pb mutant would be transformed to aristae, as aristal identity represents a regulatory state where both PB and SCR activity are OFF. We found that induction of PBQ50Ka expression in a pb27/pb20 null mutant combination results in the transformation of the tarsal mouthparts into aristae (Fig. 2C,D). This indicates that PBQ50Ka is a dominant negative molecule. This proboscis-to-arista transformation is not associated with pseudotrachea formation (Fig. 2D) as are pb hypomorphic alleles (Kaufman, 1978).

Although ectopic expression of PBQ50Ka does result in the expected transformation of a dominant negative PB molecule, an expected control result would be that ectopic expression of PB from a hsp/pb fusion gene should rescue the pb null mutant phenotype (Cribbs et al., 1995; Randazzo, 1991). The toxicity of ectopic expression of PBa/b did not allow this control. It is still important to demonstrate that PBQ50K molecules have reduced activity relative to PB molecules if the claim that PBQ50Ka has the genetic properties of a dominant negative molecule is correct. Another concern is that PBa and PBQ50Ka are derived from different spliced forms of PB mRNA 2-μ-4b and 2-μ-4a, respectively. To address these concerns, we investigated whether PBa expressed in a clone of cells from the Tub al promoter would rescue the pb null phenotype and PBQ50Kb would not. The pba and pbQ50Kb genes were inserted into pP{w+, NotI>y+>Tubα1}, and introduced into the Drosophila genome by μ-element-mediated transformation. Clones of cells expressing PBa in a pb27/pb20 individual exhibit rescue of the pb null phenotype: the antenna is transformed to a maxillary palp (Fig. 3B), the reduced maxillary palp is rescued (Fig. 3E) and the tarsal mouthparts are rescued to a proboscis (Fig. 3H). These rescue phenotypes were observed in two independent transformed lines. Clones of cells expressing PBQ50Kb in a pb27/pb20 individual exhibit the phenotype expected of a dominant negative PB molecule: the antennae are not transformed to maxillary palps (Fig. 3C), the reduced maxillary palps are not rescued (Fig. 3F) and the tarsal mouthparts are transformed towards antennae (Fig. 3I,K,L). We observed the tarsus-to-arista transformation (Fig. 3I), the presence of the antenna sense organ, the sacculus (Fig. 3K), and the presence of thick thorny bristles, the zahnborsten (Fig. 3L). These phenotypes were observed in two independent transformed lines.

Fig. 3.

PBQ50Kb has reduced PB activity. All panels are pb27/pb20 individuals. (A–C) Antennae; (D–F) maxillary palps; (G–I) proboscises. (A,D,G) No expression of PB protein; (B,E,H) expression of PBa; (C,F,I) expression of PBQ50Kb. (J) An antennal sacculus; (K) a sacculus in the proboscis of apb27/pb20 individual expressing PBQ50Kb and (L) a zahnborsten in a pb27/pb20 individual expressing PBQ50Kb. Filled circles, maxillary palp-like bristles; asterisks, arista-like structures; arrowhead, sacculi; arrow in H, the pseudotrachae; arrow in L, the zahnborsten.

Fig. 3.

PBQ50Kb has reduced PB activity. All panels are pb27/pb20 individuals. (A–C) Antennae; (D–F) maxillary palps; (G–I) proboscises. (A,D,G) No expression of PB protein; (B,E,H) expression of PBa; (C,F,I) expression of PBQ50Kb. (J) An antennal sacculus; (K) a sacculus in the proboscis of apb27/pb20 individual expressing PBQ50Kb and (L) a zahnborsten in a pb27/pb20 individual expressing PBQ50Kb. Filled circles, maxillary palp-like bristles; asterisks, arista-like structures; arrowhead, sacculi; arrow in H, the pseudotrachae; arrow in L, the zahnborsten.

Genetic evidence that the interaction between PB and SCR is regulated

Although the analysis with PBQ50K molecules strongly suggests an interaction between PB and SCR molecules, two expected transformations were not observed. Ectopic expression of SCR, from a heat-shock promoter or a Tubulin a1 promoter Scr fusion gene, does not transform the maxillary palp into a proboscis and ectopic expression of SCRQ50K does not reduce the maxillary palps (data not shown). We have also performed assays to detect a direct interaction between in vitro synthesized PB and SCR, but all our assays failed to detect an interaction. These assays were: coimmunoprecipitation, cofractionation, crosslinking and cooperative binding to DNA. We have a suggestion for this inability to detect a direct interaction in vitro that is based on some peculiarities of the embryonic phenotype of loss of PB function and ectopic expression of PB protein. A direct PB–SCR interaction may not be detectable in vitro because the PB-SCR interaction is regulated in vivo.

One of the original observations that initiated this work was that PB and SCR are coexpressed in the labial segment during embryogenesis and in the labial imaginal disc of third instar larvae. However, one of the most interesting and useful genetic properties ofpb null alleles is their adult viable phenotype; PB has no detectable activity during embryogenesis (Pultz et al., 1988). The cuticle and salivary glands of pb null mutant first instar larvae are wild type. Indeed, it seems that embryonic labial segmental identity is determined by SCR activity alone (Struhl, 1983; Pattatucci et al., 1991). The embryonic phenotype of Scr null mutant alleles is reduction of the T1 beard, duplication of the maxillary sense organs, disruption of head involution and no salivary gland formation (Pattatucci et al., 1991; Zeng et al., 1993; Panzer et al., 1992). Loss of both PB and SCR activities has the same cuticular phenotype as loss of SCR activity alone, indicating that PB has no activity even when SCR activity is absent (data not shown). Ectopic expression of SCR protein results in ectopic T1 beards in T2 and T3, and ectopic salivary glands in the head (Gibson et al., 1990; Panzer et al., 1992; Zeng et al., 1993). All these previously described phenotypes lead to the question, ‘Why does PB activity expressed in the embryonic labial segment not interact with SCR activity and inhibit SCR activity in determination of embryonic labial identity?’

Since PB activity is dispensable during embryogenesis, it was surprising that ectopic expression of PBa/b protein during embryogenesis had such a strong phenotype (Fig. 4A). One 1020 minute heat shock at 5 hours AEL results in inhibition of germband retraction, inhibition of head involution and reduction of the T1 beard (Fig. 4A). Reduction of the T1 beard is one of the phenotypes of a Scr null allele indicating that PB inhibits SCR activity during T1 beard formation. To characterize this further, we ectopically coexpressed PB with SCR. Ectopic expression of SCR alone results in the formation of ectopic T1 beards in T2 and T3 (Fig. 4C) (Gibson et al., 1990); however, with ectopic coexpression of PB with SCR, T1 beard formation in all thoracic segments is suppressed (Fig. 4B; Table 2). PBQ50Ka also inhibits SCR activity in induction of ectopic T1 beards (data not shown). Also, we observe the same suppression of ectopic beard formation by PB using SCRa. These data indicate that PB inhibits SCR activity during embryogenesis in T1 beard formation.

Table 2.

Ectopic expression of PB inhibits the phenotype caused by ectopic expression of SCR in the thorax but not the head

Ectopic expression of PB inhibits the phenotype caused by ectopic expression of SCR in the thorax but not the head
Ectopic expression of PB inhibits the phenotype caused by ectopic expression of SCR in the thorax but not the head
Fig. 4.

PB activity inhibits SCR activity in the thorax but not the head. (A–C) First instar larval cuticle preparations; (D–F) in situ hybridizations with a salivary gland-specific probe to 10 hours AEL embryos. (A,D) The result of ectopic expression of PB; (B,E) the result of ectopic expression of both PB and SCR and (C,F) the result of ectopic expression of SCR. The arrows in A and B indicate the reduced T1 beards, and the asterisks in C indicate the ectopic T1 beards in T2 and T3; note that the ectopic T1 beards are lacking in B. The squares in D-F are beside the normal salivary glands. The arrowheads point to the ectopic salivary glands induced by ectopic expression of SCR. Note the presence of ectopic salivary glands when both PB and SCR are ectopically coexpressed.

Fig. 4.

PB activity inhibits SCR activity in the thorax but not the head. (A–C) First instar larval cuticle preparations; (D–F) in situ hybridizations with a salivary gland-specific probe to 10 hours AEL embryos. (A,D) The result of ectopic expression of PB; (B,E) the result of ectopic expression of both PB and SCR and (C,F) the result of ectopic expression of SCR. The arrows in A and B indicate the reduced T1 beards, and the asterisks in C indicate the ectopic T1 beards in T2 and T3; note that the ectopic T1 beards are lacking in B. The squares in D-F are beside the normal salivary glands. The arrowheads point to the ectopic salivary glands induced by ectopic expression of SCR. Note the presence of ectopic salivary glands when both PB and SCR are ectopically coexpressed.

Although ectopic expression of PB protein reduced the T1 beard, the other cuticular phenotype of an Scr null allele, an ectopic maxillary sense organ, was not found. We performed an extensive series of heat shocks at different times during embryogenesis, but did not observe ectopic maxillary sense organs. However, recent analysis of this Scr homeotic transformation has questioned whether this Scr phenotype is a duplication of the maxillary sense organ (Pederson et al., 1996). Another phenotype of a Scr null allele not observed with ectopic expression of PB is inhibition of salivary gland formation (Fig. 4D) (Panzer et al., 1992). To characterize this further, we ectopically coexpressed PB and SCR at 3 hours AEL to determine if PB could inhibit the ability of ectopically expressed SCR to induce ectopic salivary glands. The frequency of ectopic salivary gland formation was similar when SCR was expressed alone, as when PB and SCR were coexpressed (Table 2; Fig. 4E). As observed with loss of PB activity, salivary gland specification is not affected by ectopic expression of PB.

These experiments suggest that the PB–SCR interaction is regionally specific during embryogenesis: occurring in the thorax, but not the head. This result indicates that the PB–SCR interaction is regulated. If the PB–SCR interaction is regulated during embryogenesis, it may be regulated during adult determination, potentially explaining why ectopic expression of SCRQ50K did not inhibit PB activity in maxillary palp formation, and also why ectopic expression of SCR does not result in a maxillary palp-to-proboscis transformation. The regulation of the PB–SCR interaction also suggests that there is a factor(s) mediating PB and SCR complex formation.

SCR activity is required cell nonautonomously for tarsus determination

In the course of this study of the PB-SCR interaction, we found that our various stocks allowed us to address a fundamental problem with our model for the determination of the various adult structures. Although ectopic expression of SCR is able to induce an arista-to-tarsus transformation (Fig. 1C), loss of SCR activity in a mosaic analysis does not result in a tarsus- to-arista transformation (Struhl, 1982).

Ectopic expression of PBQ50Ka and PB proteins transform tarsi toward aristae

The goal of ectopic expression of the dominant negative PBQ50K molecule was to inhibit SCR activity. Indeed PBQ50Ka/b do inhibit SCR activity in determination of the tarsal mouthparts of a pb27/pb20 null mutant (Figs 2C,D; 3). This effect of PBQ50Ka/b suggests that PB and SCR determine proboscis identity by forming a PB–SCR-containing, heteromeric complex. Another goal of ectopic expression of PBQ50K and PB was to inhibit SCR activity via complex formation in other regions of the body plan besides the labial segment and labial imaginal disc. During embryogenesis, SCR is expressed in, and is required for, determination of T1 segmental identity (Pat- tatucci et al., 1991). We have shown that ectopic expression of PB and PBQ50K inhibits SCR activity required for determination of embryonic T1 segmental identity (Fig. 4; Table 2).

The larval expression pattern of SCR is complex. SCR is expressed in the first leg imaginal disc of third instar larvae and highly expressed in a patch of ectoderm cells that are the primordia of the sex combs (Glickman and Brower, 1988; Pat- tatucci and Kaufman, 1991). SCR is also expressed in all leg discs in the mesoderm primordia, the adepithelial cells and throughout the labial imaginal disc. The requirement of SCR activity defined by genetic analysis is also complex (Struhl, 1982; Gibson et al., 1990). Our initial expectation for the phenotype of ectopic expression of PBQ50Ka during the larval stage had been a transformation of the first leg identity to second leg identity and the reduction in the number of sex combs.

Ectopic expression of PBQ50Ka from a hsp/pbQ50Ka fusion gene at late second and early third instar larval stage by administration of a short 10 or 20 minute heat shock is very toxic. The survivors exhibit a number of leg phenotypes: an increased number of sex combs on male first legs (Fig. 5D), ectopic sex combs on lower tarsal segments of the male first leg, a twisted and shortened leg phenotype (Fig. 5E) and a tarsus-to-arista transformation of the ends of all tarsi (Fig. 5B,C). Ectopic expression of PBa/b during third instar larval stage by administration of a short 10 minute heat shock is very toxic. The survivors of ectopic expression of PBa/b have similar phenotypes as those produced by ectopic expression of PBQ50Ka. Although we did not get the expected phenotype with ectopic expression of PBQ50Ka in the leg, anything but, the aristal transformation of all legs was interesting because of the expression pattern of SCR common to all the leg imaginal discs. SCR is expressed in the mesodermal adepi- thelial cells of all leg imaginal discs (Glickman and Brower, 1988). If PBQ50Ka and PBa/b were inhibiting SCR activity to result in the tarsus-to-arista transformation of all six legs, SCR must be determining tarsal identity via a cell nonau- tonomous mechanism, that is, SCR activity expressed in the adepithelial cells must be dictating tarsal identity in the ectoderm.

Fig. 5.

The results of ectopic expression of PBQ50Ka on adult leg development. (A) A wild-type tarsus. In all panels: arrowhead, pulvillus; arrows claw; asterisk, arista-like structure. (B) A first leg and (C) a third leg that developed after ectopic expression of PBQ50Ka. In both panels the posterior claw is absent, and in C it looks as if the claw is replaced by an arista-like structure indicated beside the asterisk. The example in B has arista-like structures developing further up the tarsus. (D,E) Also the result of ectopic expression of PBQ50Ka. (D) Ectopic rows of sex combs developing on a male first leg, and (E) the shortening and malformation of a first leg. (E) The tarsus is also transformed toward aristal identity.

Fig. 5.

The results of ectopic expression of PBQ50Ka on adult leg development. (A) A wild-type tarsus. In all panels: arrowhead, pulvillus; arrows claw; asterisk, arista-like structure. (B) A first leg and (C) a third leg that developed after ectopic expression of PBQ50Ka. In both panels the posterior claw is absent, and in C it looks as if the claw is replaced by an arista-like structure indicated beside the asterisk. The example in B has arista-like structures developing further up the tarsus. (D,E) Also the result of ectopic expression of PBQ50Ka. (D) Ectopic rows of sex combs developing on a male first leg, and (E) the shortening and malformation of a first leg. (E) The tarsus is also transformed toward aristal identity.

Mosaic analysis with null Scr and pb alleles

We have repeated the mosaic analysis performed by Struhl (1982) with the null Scr1 and Scr2 alleles and have also found that the tarsi are not transformed to aristae. These results are explained by the proposed nonautonomous mechanism of SCR function. In a mosaic analysis, a clone of Scr mutant cells is surrounded by Scr+ cells. Also, the markers used in our studies, and others, can only be scored in the ectoderm. The non- autonomous mechanism proposes that tarsus determination of the ectoderm does not require SCR activity in the ectodermal cells, but requires that a SCR-dependent, tarsus-inducing, signal factor is synthesized in the mesoderm. Hence, in a mosaic organism with a Scr1 clone in the ectoderm of the tarsus, SCR activity in the mesoderm is still directing synthesis of the tarsus inducer. Also important to note is that a clone of Scr1 adepithelial cells will not secrete the tarsus-inducing factor, but the surrounding Scr+ adepithelial cells will still be secreting the tarsus-inducing factor; thus, Scr1 clones in the mesoderm produce no phenotype.

This cell nonautonomous proposal also has independent support from a mosaic analysis with the pb27 null allele in the proboscis. This experiment was originally conceived as a control, because a previous mosaic analysis with a pb1ssa chromosome had shown a reproducible proboscis-to-tarsus transformation (Struhl, 1981b). The null pb phenotype of a completely mutant organism is a tarsal transformation of the proboscis (Kaufman, 1978). Hence, it was surprising that 50 % (Fig. 6A) of pb27 clones in the proboscis adopt aristal identity, 15% a mixed aristal and tarsal identity (Fig. 6B), and 35% tarsal claws and pulvilli (Fig. 6C). The proposed cell nonautonomous requirement for tarsal determination on a SCR-dependent, tarsus-inducing, mesodermally synthesized, signaling factor and the interaction between PB and SCR for proboscis determination explains this result nicely. In the mosaic analysis, a clone of pb27 ectodermal cells is produced. Remember that the markers that we used, Sb M, can only be scored in ectodermal derivatives. In these clones of cells, SCR activity is no longer associated with PB activity, but SCR activity expressed in ectodermal cells is not sufficient for synthesis of the tarsus inducer. PB activity present in the non pb27 mutant cells (that is wild type for PB activity) of the mesoderm is interacting with SCR activity such that SCR is not free by itself for the synthesis of the mesoderm-specific, tarsus inducer. This situation results in the clone of pb27 ectodermal cells adopting an aristal identity because no tarsus inducer is synthesized.

Fig. 6.

The phenotypes of clonal loss of PB activity in the proboscis. (A) An aristal transformation, (B) a mixed aristal/tarsal transformation and (C) a tarsal transformation. Asterisks, arista-like structures; arrow, a claw; arrowhead, a pulvillus.

Fig. 6.

The phenotypes of clonal loss of PB activity in the proboscis. (A) An aristal transformation, (B) a mixed aristal/tarsal transformation and (C) a tarsal transformation. Asterisks, arista-like structures; arrow, a claw; arrowhead, a pulvillus.

Clones of pb Scr cells in the labial region of a pb mutant adopt tarsus identity

The nonautonomous mechanism is a robust explanation of our results as it also explains why infrequent tarsal transformations are observed in a pb27 mosaic analysis, but never observed in a pb27Scr2 mosaic analysis. FLP-mediated mitotic recombination is very efficient; a mosaic organism has many independent clones (Xu and Rubin, 1993). The tarsal transformation of the proboscis observed in the pb27 mosaic analysis could be a result of two clones of pb27 cells: one clone marked with the ectodermal markers and a second unmarked clone of cells in the mesoderm. This results in a clone of cells in the mesoderm that can synthesize the SCR-dependent, tarsus-inducing, signaling factor and a clone of ectoderm cells that can respond to the synthesis of this factor (pb27). However, when the same double clone situation occurs in a pb27Scr2 mosaic analysis, the clone produced in the mesoderm lacks SCR activity and can not produce the tarsus-inducing factor, which explains why pb27Scr2 clones invariantly adopt aristal identity (Fig. 2A,B). If these explanations are correct, then an ectodermal clone of pb27Scr2 mutant cells in a pb null mutant background will adopt tarsal identity and not aristal. This result is expected because SCR activity in the mesoderm of a pb null mutant is released from interaction with PB such that the tarsus inducer will be synthesized, and removal of SCR activity in the ectoderm will have no effect.

We have tested this prediction using FLP-mediated mitotic recombination with a fly of the genotype y w; P{hspFLP}/+; P{ry+, neor, FRT}82Bpb27Scr2P{w+, ry+}90E/P{ry+, neor, FRT}82Bpb20Sb63bM(3)95A2P{y+, ry+}96E. Clones of pb27Scr2 cells in the ectoderm of apb27/pb20 mutant do adopt tarsal identity, and not aristal identity as in a pb+ organism (Fig. 7A). This experiment has some technical problems. The flies with this genotype are very sick and only a few eclose so that a majority of adults had to be dissected from the pupal cases. The transformed proboscis everts with a low frequency (Fig. 7A). Also, the pb null phenotype is mixed; in this genetic background, we often observe mixed tarsal and aristal transformations (Fig. 7B). However, even with this problem with the pb phenotype, it is important to point out that the pb27Scr2 clones in a wild-type background never adopt tarsal identity, but in a pb background they do adopt tarsal identity (Fig. 7A). Thus, the prediction is correct, pb27Scr2 clones of proboscis precursor cells will adopt tarsal identity in a pb mutant.

Fig. 7.

The phenotypes of clones ofpb27Scr2 cells andpb27 cells in a pb27/pb20 mutant background. (A) A clone ofpb27Scr2 cells in a pb27/pb20 mutant background. The arrow points to two tarsal claws, one is y+ and the other is y. (Insert) An enlargement of this area where the y+ claw is indicated by a black arrow and the y claw is indicated by the white arrow. (B) The phenotype of the pb27/pb20 combination in this experiment. Asterisk, an arista; arrow, two tarsal claws. (C) A clone of pb27 cells in the proximal portion of the transformed proboscis. The arrowheads indicate Sb+ y bristles. (D) A clone of pb27Scr2 cells in the proximal portion of the transformed proboscis. The arrowhead indicates the clone that is transformed to third antenna segment identity.

Fig. 7.

The phenotypes of clones ofpb27Scr2 cells andpb27 cells in a pb27/pb20 mutant background. (A) A clone ofpb27Scr2 cells in a pb27/pb20 mutant background. The arrow points to two tarsal claws, one is y+ and the other is y. (Insert) An enlargement of this area where the y+ claw is indicated by a black arrow and the y claw is indicated by the white arrow. (B) The phenotype of the pb27/pb20 combination in this experiment. Asterisk, an arista; arrow, two tarsal claws. (C) A clone of pb27 cells in the proximal portion of the transformed proboscis. The arrowheads indicate Sb+ y bristles. (D) A clone of pb27Scr2 cells in the proximal portion of the transformed proboscis. The arrowhead indicates the clone that is transformed to third antenna segment identity.

An interesting observation made in this experiment was that pb27 homozygous clones in the proximal proboscis in a pb27/pb20 background have a pb phenotype (Fig. 7C). But pb27Scr2 homozygous clones in the proximal proboscis of a pb27/pb20 background show a transformation of the proximal leg-like structure to third antennal segment identity (Fig. 7D). We find this interesting because Scr Antp Ubx homozygous clones of cells in the proximal leg adopt third antennal segment identity Struhl, 1982). Indeed, pb27Scr2 clones in a pb background in the proboscis are a phenocopy of Scr Antp Ubx clones in the leg, and ectopic expression of SCR from a hsp/Scr fusion gene in the antenna (Gibson et al., 1990).

Cell nonautonomous requirement of SCR activity for tarsus determination in ectopic expression experiments of SCR

The cell nonautonomous requirement of SCR activity for tarsus determination proposes that the arista-to-tarsus transformation induced by ectopic expression of SCR from a hsp/Scr fusion gene is due to expression of SCR in the adepithelial cells (Gibson et al., 1990). Hence, ectopic expression of SCR, specifically in the ectoderm of the arista, should not induce an arista-to-tarsus transformation. We have tested this using flip- out Scr ectopic expression constructs, P{w+, Scra/b>y+>Tub a1}. A clone of ectoderm cells that expresses SCR activity is not transformed to tarsal identity (Fig. 8A). Also, tarsal structures are observed in arista that are not expressing SCR protein; in Fig. 8B a claw is observed at the end of a y+ arista. We have observed arista-to-tarsus transformation in the three lines that express SCRa and the one line that expresses SCRb. This transformation of the ectoderm without apparent expression of SCR also supports a cell nonautonomous requirement for SCR activity in determination of tarsus identity. As has been observed previously (Zecca et al., 1995), we found that ectopic expression of SCR from the Tub a1 promoter was very low, which may explain why ectopic sex combs on the second and third legs were not observed.

Fig. 8.

The phenotype of clonal ectopic expression of SCR activity. (A) The result of ectopic expression of SCR activity in the ectoderm cells of the arista. The arista in A is y, but wild type in structure. (B) A tarsal transformation in a FLP-mediated ectopic expression experiment with P{w+, Scra>y+>Tuba1}C. The antenna is y+, but the tip of the arista has a claw indicated by the arrow. The end of the transformed arista is shown in the insert.

Fig. 8.

The phenotype of clonal ectopic expression of SCR activity. (A) The result of ectopic expression of SCR activity in the ectoderm cells of the arista. The arista in A is y, but wild type in structure. (B) A tarsal transformation in a FLP-mediated ectopic expression experiment with P{w+, Scra>y+>Tuba1}C. The antenna is y+, but the tip of the arista has a claw indicated by the arrow. The end of the transformed arista is shown in the insert.

When is SCR activity required for tarsus determination?

From our analysis of the PB-SCR interaction, we propose that the class ofpb alleles that result in the proboscis-to-arista transformation produce a PB protein inactive for most PB activities except the ability to interact with SCR. Expression of these PB proteins with SCR results in SCR being sequestered into an inactive PB-SCR complex such that only arista identity can be determined. The first pb allele identified was the temperaturesensitive pb1 allele. We propose that the temperature-sensitive component of PB1 protein activity is its ability to interact with and inhibit SCR activity: at 18°C an inactive PB1-SCR complex forms, and at 28.5°C PB1 is completely inactive leaving SCR active to determine tarsus identity. The temperature-sensitive period of the pb1 allele was reported to be late third instar larval/ early pupal stage (Villee, 1944).

We have re-examined the temperature-sensitive period of the pb1 allele, as the temperature-sensitive period reflects when SCR activity is required for tarsus determination more than when PB activity is required. In a temperature-shift-up protocol, SCR activity is being turned ON at a specific time after hatching. In a temperature-shift-down protocol, SCR activity is being turned OFF at a specific time after hatching. The temperature-sensitive period for the pb1 allele, and hence when SCR activity is required for tarsus determination, is between 65 hours and 100 hours after egg laying at 25°C, which corresponds to late second/early third instar larval stage (Fig. 9). This period corresponds with the period when ectopic expression of PBQ50Ka results in the tarsus-to-arista transformation, and corresponds with the period when ectopic expression of SCR from a heat-shock promoter results in the arista-to-tarsus transformation.

Fig. 9.

The temperature-sensitive period of the pb1 allele. The times after egg laying that the larvae of a cross between stocks 2178 and 2097 were either shifted up from 18°C to 28.5°C (▄), or shifted down from 28.5°C, to 18°C (•), are normalized to development at 25°C. The time intervals of the major stages after hatching are indicated below the time axis. The percentage of pb1/pb27 individuals that exhibited an aristal transformation were plotted; the percentage of tarsal transformations was essentially the reciprocal. The solid bar indicates the temperature-sensitive period of the pb1 allele, and the dashed thick bar indicates the temperature-sensitive period of the Scr8 allele determined in an independent study (Pattatucci et al., 1991).

Fig. 9.

The temperature-sensitive period of the pb1 allele. The times after egg laying that the larvae of a cross between stocks 2178 and 2097 were either shifted up from 18°C to 28.5°C (▄), or shifted down from 28.5°C, to 18°C (•), are normalized to development at 25°C. The time intervals of the major stages after hatching are indicated below the time axis. The percentage of pb1/pb27 individuals that exhibited an aristal transformation were plotted; the percentage of tarsal transformations was essentially the reciprocal. The solid bar indicates the temperature-sensitive period of the pb1 allele, and the dashed thick bar indicates the temperature-sensitive period of the Scr8 allele determined in an independent study (Pattatucci et al., 1991).

A model for the determination of the proboscis.

The data presented in the first part of this paper support a model where both the expression patterns of PB and SCR and the ability of PB and SCR to interact in a multimeric complex control the determination of four adult structures. We propose that the choice of which of the four developmental pathways that result in antennae, maxillary palps, tarsi and the proboscis are entered is controlled by the expression patterns of PB and SCR (Fig. 10A). Specifically, the absence of PB and SCR expression, which we have defined for this discussion as the default state, leads to antennal identity; expression of only PB leads to maxillary palp identity; expression of only SCR leads to tarsus identity and expression of both PB and SCR leads to proboscis identity (Fig. 10A). Also proposed in this model is that PB regulates a set of PB-regulated genes required for maxillary palp determination and SCR regulates a set of SCR- regulated genes required for tarsus determination. But when PB and SCR are expressed together, PB and SCR form a PB- SCR-containing, heteromeric complex that regulates a novel set of genes, the PB-SCR-regulated genes. We also suggest that PB-SCR complex formation is regulated, that is, it may be an indirect interaction involving other cofactors (Fig. 10A).

Fig. 10.

The model for the PB-SCR interaction and the determination of four developmental fates (A), and the model for the cell nonautonomous requirement of SCR activity in a wild-type leg (B). (A) The four squares represent four cells each expressing one of four possible expression patterns of PB and SCR. The expression of pb and Scr genes is indicated by the wavy arrow and the presence of an ellipsoid representing SCR and a sphere representing PB. In the lower part of each cell are represented three sets of genes, the PB- regulated, SCR-regulated and PB-SCR-regulated genes. Although the figure is drawn showing activation of the expression of these sets of genes, our intention is only to indicate some form of regulation: positive, negative or both. The protein factor marked with a question mark is meant to indicate only that it regulates the PB-SCR interaction. In model B, the two squares represent an adepithelial cell on the left and a distally located ectodermal cell on the right. In the lower part of each cell are represented three classes of genes, the SCR-regulated, tarsus inducer gene, the tarsus determination genes and the PB-SCR-regulated genes. The arrow between the cells indicates the action of the secreted tarsus inducer.

Fig. 10.

The model for the PB-SCR interaction and the determination of four developmental fates (A), and the model for the cell nonautonomous requirement of SCR activity in a wild-type leg (B). (A) The four squares represent four cells each expressing one of four possible expression patterns of PB and SCR. The expression of pb and Scr genes is indicated by the wavy arrow and the presence of an ellipsoid representing SCR and a sphere representing PB. In the lower part of each cell are represented three sets of genes, the PB- regulated, SCR-regulated and PB-SCR-regulated genes. Although the figure is drawn showing activation of the expression of these sets of genes, our intention is only to indicate some form of regulation: positive, negative or both. The protein factor marked with a question mark is meant to indicate only that it regulates the PB-SCR interaction. In model B, the two squares represent an adepithelial cell on the left and a distally located ectodermal cell on the right. In the lower part of each cell are represented three classes of genes, the SCR-regulated, tarsus inducer gene, the tarsus determination genes and the PB-SCR-regulated genes. The arrow between the cells indicates the action of the secreted tarsus inducer.

The results presented in this paper that support this model are as follows. (i) Proboscis progenitor cells unable to express PB and SCR adopt antennal fate (Fig. 2A,B). The transformation of the proboscis includes an arista, the third antennal segment, and elements of the first or second antennal segments. (ii) Using dominant negative molecules PBQ50Ka/b, we have produced Scr loss of function phenotypes. The results with dominant negative PB molecules strongly suggest that PB and SCR interact in a multimeric protein complex. (iii) The interaction between PB and SCR seems to be indirect. We have been unable to detect a direct interaction between PB and SCR in vitro, and we also have genetic evidence that the PB-SCR interaction is regulated during Drosophila embryogenesis. We detect PB inhibiting SCR activity in T1 beard formation in the thorax; however, PB does not inhibit SCR activity in formation of salivary glands in the head. This suggests that the PB-SCR interaction is regionally specific during embryogenesis, occurring in the thorax and not the head. This may also explain why we were unable to observe a reduction of the maxillary palps with ectopic expression of SCRQ50K, and a maxillary palp to labial palp transformation with ectopic expression of SCR. There are many models that can explain this regional specificity, but all involve the presence of a factor(s) in addition to PB and SCR. This additional factor may be involved in the post-translational modification of either PB or SCR which would be required for the interaction to occur, or the factor may be a bifunctional protein, or protein complex, with one domain that binds PB and another domain that binds SCR.

The model is drawn to resemble the α1–α2 hypothesis intentionally (Herskowitz, 1989). In yeast, the two homeodomain- containing proteins encoded by MATα2 and MATα1 genes interact (Goutte and Johnson, 1988; Dranginis, 1990; Li et al., 1995). In haploid α cells, the α2 protein forms a homodimer that binds to specific α2-binding sites in α-specific genes repressing their expression. In α/α diploid cells where both α2 and α1 are expressed, α-specific genes are still repressed by the α2 protein. In addition, α2 and α1 proteins form a heterodimer that binds to specific α1/α2-binding sites in haploid-specific genes repressing their expression. Overexpression of a dominant negative a2 molecule in an α/ α cell results in the derepression of both α-specific genes and haploid-specific genes (Hall and Johnson, 1987). This yeast interaction is similar to the PB–SCR interaction in both form and the methods used to demonstrate it. However, the PB-SCR interaction may be indirect, where the α2 α1 interaction is direct (Li et al., 1995). The yeast example is also raised because the amino acid sequence of the EXD protein in the homeodomain shows similarity with MATα1 (Rauskolb et al., 1993) and the UBX-EXD interaction is one of the few well-characterized interactions between two homeodomain-containing proteins in vivo (Chan et al., 1994; Johnson et al., 1995; Peifer and Weischaus, 1990). However, the PB-SCR interaction is the only reported interaction so far between two HOX proteins in vivo.

SCR activity is required cell nonautonomously for tarsus determination

In the second part of the paper, we explain why, when ectopi- cally expressed from a heat-shock promoter, SCR induces an arista-to-tarsus transformation (Gibson et al., 1990). We propose a model for this transformation, and indeed all six wild-type tarsi, where SCR activity is required cell nonau- tonomously for tarsus determination. Specifically, we propose that SCR activity, expressed in all leg imaginal discs in the mesodermal adepithelial cells, is required for the synthesis of a tarsus-inducing signaling factor. This secreted tarsusinducing, signaling factor induces the overlaying ectoderm cells to adopt a tarsal fate as opposed to an aristal fate (Fig. 10B). This proposal is supported by five independent results. (i) Ectopic expression of PBQ50K or PB result in a tarsus-to- arista transformation of all six tarsi because much of SCR activity expressed in the adepithelial cells is complexed with PBQ50K or PB reducing the amount of tarsus-inducing, signaling factor made. (ii) The nonautonomous requirement of SCR activity explains why a tarsus-to-arista transformation was not observed in a Scr1 mosaic analysis. Because, although the ectoderm cells lack a functional Scr gene, the underlying mesoderm still expresses active SCR and hence secretes the tarsus-inducing, signaling factor. (iii) A null pb mutant phenotype is a proboscis-to-tarsus transformation; however, clones of pb mutant cells usually adopt aristal identity. This is because in a completely pb mutant organism, SCR is free in the adepithelial cells of the labial imaginal discs to direct synthesis of the mesoderm-specific, tarsus-inducing, signaling factor, but in a mosaic analysis, even though SCR activity is free in the ectoderm, no signaling factor is synthesized in the mesoderm because SCR is still complexed with PB, directing proboscis identity. (iv) When a mosaic analysis with a pb27Scr2 chromosome is performed in a wild-type background, an invariant transformation of the proboscis to an antenna is observed; however, if the same mosaic analysis is performed in a pb genetic background, the invariant aristal transformation is not observed. This is because, in a pb background, the ade- pithelial cells are synthesizing the SCR-dependent, tarsusinducing, signaling factor which induces the pb27Scr2 clone of cells to adopt tarsus identity. In a wild-type background, the SCR-dependent, tarsus-inducing, signaling factor is not synthesized, because SCR is bound up in a PB-SCR complex directing proboscis identity in wild-type cells. (v) Ectopic expression of SCR in the ectoderm cells of the arista does not result in a transformation of the arista to a tarsus; however, tarsal transformations are observed when there is no apparent expression of SCR in the arista. This is because SCR expression in the ectoderm can not result in the synthesis of a tarsus-inducing, signaling factor. However, ectopic expression of SCR in the mesoderm, which can not be detected with the yellow marker, does result in the synthesis of the tarsus inducing signaling factor. This interlocking set of results strongly supports the proposed model.

The genetic data demonstrate a cell nonautonomous requirement for SCR activity, and we propose that SCR activity expressed in the adepithelial cells is required for the synthesis of a secreted tarsus-inducing, signaling factor. We have three reasons for proposing this. (i) SCR protein is expressed in the adepithelial cells of all three leg imaginal discs. The only other site of expression of SCR in the leg discs is specific to the first leg imaginal disc (Glickman and Brower, 1988; Pattatucci and Kaufman, 1991). (ii) Leg imaginal discs are composed of only two germ layer derivatives: the mesoderm and the ectoderm. The adepithelial cells are the primordial cells of the adult mesoderm (Currie and Bate, 1991). (iii) In the Scr1 mosaic analysis, we observed whole legs that are Scr’ in the ectoderm and, in clonal ectopic expression of SCR, we have observed antennae that are completely y+ with tarsal transformations. These two observations indicate that there is not a small patch of the ectoderm that transiently expresses SCR and secretes the tarsus inducer which patterns the whole ectoderm.

The proposed nonautonomous model for SCR function is an example of induction. SCR is required for synthesis of a signaling factor in the mesoderm of the leg imaginal discs that acts on the overlaying ectodermal cells directing tarsal determination. This is similar to the role of UBX activity in gut development (Immergluck et al., 1990). Another well-characterized example of induction is the requirement of Lin-3 secreted from the anchor cell of the C. elegans gonad for epidermal vulva development (Hill and Sternberg, 1992). A result of our work is the description of the characteristics of an SCR-regulated gene. The SCR-regulated gene, the tarsusinducer, should be expressed in the adepithelial cells of all six leg discs but not the adepithelial cells of the antenna disc. The gene product of this SCR-regulated gene should be a secreted factor that is either the ligand or involved in the activation of a ligand required for tarsus determination. This SCR-regulated gene may also be expressed during embryogenesis under the control of SCR, and required for gastric caeca formation (Reuter and Scott, 1990).

Arista versus antenna and tarsus versus leg

The arista and tarsus are the distal portions of the antenna and leg. The phenotype of the transformations of the proboscis shown in Fig. 7A,D indicates that the determination of the proximal and distal portions of the leg occurs by distinct mechanisms. There are two further relevant points: first, SCR activity is required cell nonautonomously only for tarsus determination, and is required cell autonomously for determination of first leg identity (Struhl, 1982); second, in a completely pb null mutant organism, the proboscis is transformed into a structure that is mainly a tarsus, which at the base has reduced proximal leg parts (tibia and femur) (Kaufman, 1978). These transformed mouthparts have first leg identity. SCR activity is required cell autonomously for sex comb determination. The interesting observation made in the pb27Scr2 mosaic analysis in a pb background was that pb27Scr2 clones in the proximal regions of the leg-like mouthparts adopted third antennal segment identity (Fig. 7D). There are several points that are interesting about this transformation when compared to previously reported transformations (Struhl, 1981a, 1982). The first leg transformation of the proboscis requires SCR activity alone; remove both PB and SCR activity and an antenna forms (Fig. 2A,B). But to get a clone of cells adopting third antennal segment identity in the normal thoracic first leg, both SCR and ANTP activity need to be removed (Struhl, 1982). Removing ANTP activity alone has no effect and removing SCR activity results in the transformation of the first leg into a second leg.

We suggest that the Drosophila leg is made up of two developmental fields: the tarsus and the proximal leg. These two developmental fields may correlate with the nuclear (proximal) versus cytoplasmic (distal) intracellular localization of Extradenticle, and the distal expression of Distalless (Gonzalez-Crespo and Morata, 1996; Aspland and White, 1997). Also, that there are four genetic pathways working in leg determination. The first pathway is the cell nonautonomous SCR-dependent, tarsus-inducing, signal pathway, and this lays down the plan for the basic unmodified tarsus. The second pathway is the relatively cell autonomous proximal leg pathway, which can be activated by the expression of SCR, ANTP or UBX and which lays out the basic plan for the proximal leg (Struhl, 1982). The third and fourth pathways are cell autonomous pathways that SCR and UBX control. A basic leg plan results in second leg identity, but expression of SCR or UBX in both the proximal and distal portions of this basic plan brings about modifications resulting in first or third leg identity, respectively.

Although this model explains much of our, and others, results, it does not explain why ectopic expression of ANTP transforms the antenna into a complete pair of second legs, including the tarsus (Schneuwly et al., 1987). How does ectopic expression of ANTP activity result in tarsus determination when it is a SCR-dependent process? It is possible that ANTP may activate expression of SCR in the antennal adepi- thelial cells.

Temporal requirements of SCR activity

This model for leg determination, using four genetic pathways, also explains why SCR has two temperature-sensitive periods. The temperature-sensitive period of pb1 is the late second/early third instar larval stage. This period is when SCR activity is required for determination of tarsal identity. In an independent study using the temperature-sensitive allele Scr8, the temperature-sensitive period was determined to be late third instar larval/early pupal stage (Pattatucci et al., 1991). This study scored a different phenotype, the requirement of SCR activity for the formation of sex combs. This suggests that SCR activity is required cell nonautonomously during late second/early third instar larval stage for tarsus determination, and later cell autonomously during late third instar larval/early pupal stage to modify the basic leg plan. These observations and explanations are similar to those characterizing the requirement of mab-5 activity during C. elegans development (Salser and Kenyon, 1996), and the requirement of UBX activity during Drosophila embryogenesis (Castelli-Gair and Akam, 1995). Collectively, all these observations demonstrate that a HOX product is required in a variety of developmental pathways at distinct temporal stages, and that the complex spatial and temporal expression pattern of HOX proteins is an important component of their function.

We thank the Bloomington stock center for fly stocks. We thank Gary Struhl for fly stocks, DNA constructs and helpful advise. We thank Danielle J. Hayden for assistance in constructing some of the fly stocks and H. Leung for help with the scanning electron microscope. We thank Marc Perry for his comments on the manuscript. This work was supported by a grant to A. P.-S. from the Medical Research Council of Canada.

Andrews
,
D. J.
and
Scott
,
M. P.
(
1992
).
Downstream of the homeotic genes
.
The New Biologist
4
,
5
15
.
Aspland
,
S. E.
and
White
,
R. A. H.
(
1997
).
Nucleocytoplasmic localisation of extradenticle protein is spatially regulated throughout development in Drosophila
.
Development
124
,
741
747
Basler
,
K.
and
Struhl
,
G.
(
1994
).
Compartment boundaries and the control of Drosophila limb pattern by hedgehog protein
.
Nature
368
,
208
214
.
Capovilla
,
M.
,
Brandt
,
M.
and
Botas
,
J.
(
1994
).
Direct regulation of decapentaplegic by Ultrabithorax and its role in Drosophila midgut morphogenesis
.
Cell
76
,
461
475
Carroll
,
S. B.
(
1995
).
Homeotic genes and the evolution of arthopods and chordates
.
Nature
376
,
479
485
.
Castelli-Gair
,
J.
and
Akam
,
M.
(
1995
).
How the Hox gene Utrabithorax specifies two different segments: the significance of spatial and temporal regulation within metameres
.
Development
121
,
2973
2982
.
Chadwick
,
R.
and
McGinnis
,
W.
(
1987
).
Temporal and spacial distribution of transcripts from the Deformed gene of Drosophila
.
EMBO J
.
6
,
779
789
.
Chan
,
S-K.
and
Mann
R. S.
(
1993
).
The segment identity functions of Ultrabithorax are contained within its homeodomain and carboxy-terminal sequences
.
Genes and Dev
7
,
796
811
.
Chan
,
S-K.
,
Jaffe
,
L.
,
Capovilla
,
M.
,
Botas
,
J.
and
Mann
,
R. S.
(
1994
).
The DNA binding specificity of Ultrabithorax is modulated by cooperative interactions with Extradenticle, another Homeoprotein
.
Cell
78
,
803
813
.
Chan
,
S-K.
,
Popperl
,
H.
,
Krumlauf
,
R.
and
Mann
R. S.
(
1996
).
An extradenticle-induced conformational change in a HOX protein overcomes an inhibitory function of the conserved hexapeptide motif
.
EMBO J
.
15
,
2476
2487
.
Cribbs
,
D. L.
,
Pultz
,
M. A.
,
Johnson
,
D.
,
Mazzulla
,
M.
and
Kaufman
,
T. C.
(
1992
).
Structural complexity and evolutionary conservation of the Drosophila homeotic gene proboscipedia
.
EMBO J
.
11
,
1437
1449
.
Cribbs
,
D. L.
,
Benassayag
,
C.
,
Randazzo
,
F. M.
and
Kaufman
,
T. C.
(
1995
).
Levels of homeotic protein function can determine developmental identity: evidence from low-level expression of the Drosophila homeotic gene proboscipedia under Hsp70 control
.
EMBO J
.
14
,
767
778
.
Currie
,
D.
and
Bate
,
M.
(
1991
).
Development of adult abdominal muscles in Drosophila: Adult myoblasts express twist and are associated with nerves
.
Development
113
,
91
102
Dranginis
,
A. M.
(
1990
).
Binding of yeast a1 and oi2 as a heterodimer to the operator DNA of a haploid-specific gene
.
Nature
347
,
682
685
.
Furukubo-Tokunaga
,
K.
Flister
,
S.
and
Gehring
,
W. J.
(
1993
).
Functional specificity of the Antennapedia homeodomain
.
Proc. Natl. Acad. Sci. USA
90
,
6360
6464
.
Gehring
,
W. J.
,
Qian
,
Y. Q.
,
Billeter
,
M.
,
Furukubo-Tokunaga
,
K.
,
Schier
,
A. F.
,
Resendez-Perez
,
D.
,
Affolter
,
M.
and
Wüthrich
,
K.
(
1994
).
Homeodomain-DNA recognition
.
Cell
78
,
211
223
.
Gibson
,
G.
,
Schier
,
A. F.
,
LeMotte
,
P.
and
Gehring
,
W. J.
(
1990
).
The specificities of Sex combs reduced and Antennapedia are defined by a distinct portion of each protein that includes the homeodomain
.
Cell
62
,
1087
1103
.
Glickman
,
M. A.
and
Brower
,
D. L.
(
1988
).
Expression of Sex combs reduced protein in Drosophila larvae
.
Dev. Biol
.
127
,
113
118
.
Golic
,
K. G.
(
1991
).
Site-specific recombination between homologous chromosomes in Drosophila
.
Science
252
,
958
961
.
Gonzalez-Crespo
,
S.
and
Morata
,
G.
(
1996
).
Genetic evidence for the subdivision of the arthropod limb into coxopodite and telopodite
.
Development 1
22
,
3921
3928
Goutte
,
C.
and
Johnson
,
A. D.
(
1988
).
a1 protein alters the DNA binding specificity of 0i2 repressor
.
Cell
52
,
875
882
Hall
,
M. N.
and
Johnson
,
A. D.
(
1987
).
Homeodomain of the yeast repressor 0 2 is a sequence-specific DNA-binding domain but is not sufficient for repression
.
Science
237
,
1007
1012
.
Hanes
,
S. D.
and
Brent
,
R.
(
1991
).
A genetic model for interaction of the Homeodomain recognition helix with DNA
.
Science
251
,
426
430
.
Herskowitz
,
I.
(
1987
)
Functional inactivation of genes by dominant negative mutations
Nature
329
,
219
222
.
Herskowitz
,
I.
(
1989
).
A regulatory hierarchy for cell specialization in yeast
.
Nature
342
,
749
757
Hill
,
R. J.
and
Sternberg
,
P. W.
(
1992
).
The gene lin-3 encodes an inductive signal for vulval development in C. elegans
.
Nature
358
,
470
476
.
Hunt
,
P.
and
Krumlauf
,
R.
(
1991
).
Deciphering the Hox code: Clues to patterning brachial regions of the head
.
Cell
66
,
1075
1078
.
Immergluck
,
K.
,
Lawrence
,
P. A.
and
Bienz
,
M.
(
1990
).
Induction across germ layers in Drosophila mediated by a genetic cascade
.
Cell
62
,
261
268
.
Johnson
,
F. B.
,
Parker
,
E.
and
Krasnow
,
M. A.
(
1995
).
Extradenticle protein is a selective cofactor for the Drosophila homeotics: Role of the homeodomain and YPWM amino acid motif in the interaction
.
Proc. Natl. Acad. Sci. USA
92
,
739
743
.
Kaufman
,
T. C.
(
1978
).
Cytogenetic analysis of chromosome 3 in Drosophila melanogaster: isolation and characterization of four new alleles of proboscipedia (pb) locus
.
Genetics
90
,
579
596
.
Kaufman
,
T. C.
,
Seeger
,
M. A.
and
Olsen
,
G.
(
1990
).
Molecular and genetic organization fo the Antennapedia gene complex of Drosophila melanogaster
.
In Advances in Genetics: Genetic Regulatory Hierarchies in Development
,
P. R. F.
Wright
, ed., Vol.
27
. pp.
309
362
.
Klemenz
,
R.
,
Weber
,
U.
and
Gehring
,
W. J.
(
1987
).
The white gene as a marker in a new P-elemnt vector for gene transfer in Drosophila
.
NAR
10
,
3947
3959
Kuroiwa
,
A.
,
Kloter
,
U.
,
Baumgartner
,
P.
and
Gehring
,
W. J.
(
1985
).
Cloning of the homeotic Sex combs reduced gene in Drosophila and in situ localization of its transcripts
.
EMBO J
.
4
,
3757
3764
.
LeMotte
,
P. K.
,
Kuroiwa
,
A.
,
Fessler
,
L. I.
and
Gehring
,
W. J.
(
1989
).
The homeotic gene Sex combs reduced of Drosophila: gene structure and embryonic expression
.
EMBO J
.
8
,
219
227
Lewis
,
E. B.
(
1978
).
A gene complex controlling segmentation in Drosophila
.
Nature
276
,
565
570
.
Li
,
T.
,
Stark
,
M. R.
,
Johnson
,
A. D.
and
Wolberger
,
C.
(
1995
).
Crystal structure of the MATa1/MAToi2 homeodomain heterodimer bound to DNA
.
Science
270
,
262
269
.
Lindsley
,
D.
and
Zimm
,
G.
(
1992
).
The Genome of Drosophila melanogaster
.
Academic Press, San Diego, CA
.
Mahaffey
,
J.
and
Kaufman
,
T. C.
(
1987
).
Distribution of the Sex combs reduced gene products in Drosophila melanogaster
.
Genetics
117
,
51
60
.
Mann
,
R. S.
and
Hogness
,
D. S.
(
1990
).
Functional dissection of Ultrabithorax proteins in D. melanogaster
.
Cell
60
,
597
610
.
McGinnis
,
W.
and
Krumlauf
,
R.
(
1992
).
Homeobox genes and axial patterning
.
Cell
68
,
283
302
.
Padgett
,
R. W.
,
St Johnston
,
D.
and
Gelbart
,
W. M.
(
1987
).
A transcript from a Drosophila pattern gene predicts a protein homologous to the transforming growth factor-beta family
.
Nature
325
,
81
84
.
Panzer
,
S.
,
Weigel
,
D.
and
Beckendorf
,
S. K.
(
1992
).
Organogenesis in Drosophila melanogaster: embryonic salivary gland determination is controlled by homeotic and dorsoventral patterning genes
.
Development
114
,
49
57
.
Pattatucci
,
A.
and
Kaufman
,
T. C.
(
1991
).
The homeotic gene Sex combs reduced of Drosophila melanogaster is differentially regulated in the embryonic and imaginal stages of development
.
Genetics
129
,
443
461
.
Pattatucci
,
A. M.
,
Otteson
,
D. C.
and
Kaufman
,
T. C.
(
1991
).
A functional and structural analysis of the Sex combs reduced locus of Drosophila melanogaster
.
Genetics
129
,
423
441
.
Pederson
,
J. D.
,
Kiehart
,
D. P.
and
Mahaffey
,
J. W.
(
1996
).
The role of HOM-C genes in segmental transformations: Reexamination of the Drosophila Sex combs reduced embryonic phenotype
.
Dev. Biol
.
180
,
131
142
.
Peifer
,
M.
,
Karch
,
F.
and
Bender
,
W.
(
1987
).
The bithorax complex: control of segmental identity
.
Genes and Dev
.
1
,
891
898
.
Peifer
,
M.
and
Wieschaus
,
E.
(
1990
).
Mutations in the Drosophila gene extradenticle affect the way specific homeodomain proteins regulate segmental identity
.
Genes and Dev
.
4
,
1209
1223
.
Percival-Smith
,
A.
,
Müller
,
M.
,
Affolter
,
M.
and
Gehring
,
W. J.
(
1990
).
The interaction with DNA of wild-type and mutant fushi tarazu homeodomains
.
EMBO J
.
9
,
3967
3974
.
Pultz
,
M. A.
,
Diederich
,
R. J.
,
Cribbs
,
D. L.
and
Kaufman
,
T. C.
(
1988
).
The proboscipedia locus of the Antennapedia Complex: a molecular and genetic analysis
.
Genes and Dev
.
2
,
901
920
.
Randazzo
,
F. M.
,
Cribbs
,
D. L.
and
Kaufman
,
T. C.
(
1991
).
Rescue and regulation of proboscipedia: a homeotic gene of the Antennapedia complex
.
Development
113
,
257
271
.
Rauskolb
,
C.
,
Peifer
,
M.
and
Wieschaus
,
E.
(
1993
).
extradenticle, a regulator of homeotic gene activity, is a homolog of the homeobox-containing, Human proto-oncogene pbx1
.
Cell
74
,
1101
1112
.
Reuter
,
R.
and
Scott
,
M. P.
(
1990
).
Expression and function of the homeotic genes Antennapedia and Sex combs reduced in embryonic midgut of Drosophila
.
Development
109
,
289
303
.
Rubin
,
G. M.
and
Spradling
,
A. C.
(
1982
).
Genetic transformation of Drosophila with transposable element vectors
.
Science
218
,
348
353
.
Salser
,
S. J.
and
Kenyon
,
C.
(
1996
).
A C. elegans Hox gene switches on, off, on, and off again to regulate proliferation, differentiation, and morphogenesis
.
D evelopment
122
,
1651
1661
.
Schier
,
A. F.
and
Gehring
,
W. J.
(
1992
).
Direct homeodomain-DNA interaction in the autoregulation of the fushi tarazu gene
.
Nature
356
,
804807
.
Schneuwly
,
S.
,
Klemenz
,
R.
and
Gehring
,
W. J.
(
1987
).
Redesigning the body plan of Drosophila by ectopic expression of the homeotic gene Antennapedia
.
Nature
235
,
816
818
.
Slack
,
J. M. W.
,
Holland
,
P. W. H.
and
Graham
,
C. F.
(
1993
).
The Zootype and the Phylotypic Stage
.
Nature
361
,
490
492
.
Struhl
,
G.
(
1981a
).
A homeotic mutation transforming leg to antenna in Drosophila
.
Nature
292
,
635
638
.
Struhl
,
G.
(
1981b
).
Anterior and posterior compartments in the proboscis of Drosophila
.
Dev Biol
.
84
,
372
385
Struhl
,
G.
(
1982
).
Genes controlling segmental specification in the Drosophila thorax
.
Proc. Natl. Acad. Sci. USA
79
,
7380
7384
.
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
.
Struhl
,
G.
and
Basler
,
K.
(
1993
).
Organizing activity of wingless protein in Drosophila
.
Cell
72
,
527
540
.
Tautz
,
D.
and
Pfeifle
,
C.
(
1989
).
A non-radioactive in situ hybridization method for the localization of specific RNAs in Drosophila embryos reveals translational control of the segmentation gene hunchback
.
Chromosoma
98
,
81
85
.
Villee
,
C. A.
(
1944
).
Phenogenetic studies of the homoeotic mutants of Drosophila melanogaster II. The effects of temperature on expression of proboscipedia
.
J. Exp. Biol
.
96
,
85
102
.
Wieschaus
,
E.
and
Nüsslein-Volhard
(
1986
).
Looking at embryos
,
In Drosophila: A practical Approach
,
D. B.
Roberts
ed. (
IRL Oxford
) pp
199228
.
Xu
,
T.
and
Rubin
,
G. M.
(
1993
).
Analysis of genetic mosaics in developing and adult Drosophila tissues
.
Development
117
,
1223
1237
.
Zecca
,
M.
,
Basler
,
K.
, and
Struhl
,
G.
(
1995
).
Sequential organizing activities of engrailed, hedgehog and decapentaplegic in the Drosophila wing
.
Development
121
,
2265
2278
.
Zeng
,
W.
,
Andrews
,
D. J.
,
Mathies
,
L. D.
,
Horner
,
M. A.
and
Scott
,
M.P.
(
1993
).
Ectopic expression and function of the Antp and Scr homeotic genes; the N terminus of the homeodomain is critical to functional specificity
.
Development
118
,
339
352
.