Regulation of proboscipedia in Drosophila by Homeotic Selector Genes

Corresponding author: Thomas C. Kaufman, HHMI, Department of Biology, Indiana University, Bloomington, IN 47405., kaufman{at}sunflower.bio.indiana.edu (E-mail)

Communicating editor: A. J. LOPEZ

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The gene proboscipedia (pb) is a member of the Antennapedia complex in Drosophila and is required for the proper specification of the adult mouthparts. In the embryo, pb expression serves no known function despite having an accumulation pattern in the mouthpart anlagen that is conserved across several insect orders. We have identified several of the genes necessary to generate this embryonic pattern of expression. These genes can be roughly split into three categories based on their time of action during development. First, prior to the expression of pb, the gap genes are required to specify the domains where pb may be expressed. Second, the initial expression pattern of pb is controlled by the combined action of the genes Deformed (Dfd), Sex combs reduced (Scr), cap'n'collar (cnc), and teashirt (tsh). Lastly, maintenance of this expression pattern later in development is dependent on the action of a subset of the Polycomb group genes. These interactions are mediated in part through a 500-bp regulatory element in the second intron of pb. We further show that Dfd protein binds in vitro to sequences found in this fragment. This is the first clear demonstration of autonomous positive cross-regulation of one Hox gene by another in Drosophila melanogaster and the binding of Dfd to a cis-acting regulatory element indicates that this control might be direct.

THE metameric expression of the homeotic (Hox) genes of the Antennapedia complex (ANT-C) and bithorax complex (BX-C) is crucial to the proper development of Drosophila melanogaster (KAUFMAN et al. 1990 ; MORATA 1993 ). The ANT-C contains the traditional Hox genes labial (lab), proboscipedia (pb), Deformed (Dfd), Sex combs reduced (Scr), and Antennapedia (Antp), each of which encodes a homeodomain containing transcription factor (KAUFMAN et al. 1990 ). Generally speaking, the homeotics confer cellular identity to all of the cells within their domain of expression. The identity that any specific cell adopts is highly dependent on the timing of the Hox gene expression and on the presence of other developmental factors (ROGERS et al. 1997 ; ROGERS and KAUFMAN 1997 ). Consequently, misexpression or loss of expression of the Hox genes can result in homeotic transformations, which may greatly affect the viability of either the adult or larva (LEWIS 1978 ). The homeotics are primarily organized along the anterior/posterior (A/P) axis of the embryos. The specification of the A/P axis is crucial to development and serves to subdivide the embryo into smaller and smaller units wherein the homeotics are expressed and confer identity. The primary and most recognized units of the insect embryo are the segment and parasegment. The specification of the A/P axis, including the expression of the Hox genes, is controlled by a hierarchy or cascade of genes (INGHAM 1988 ). This hierarchy begins with maternally provided factors that are differentially localized within the oocyte. Upon fertilization, these factors then act to establish the primary axes of the embryo (LAWRENCE 1992 ; PANKRATZ and JACKLE 1993 ). In the specification of the A/P axis, gradients of these maternal factors act to establish the expression patterns of the gap and terminal genes that subdivide the embryo into discreet domains (NUSSLEIN-VOLHARD and WEISCHAUS 1980 ; STRUHL et al. 1992 ). In turn, the gap and maternal genes establish the periodic expression patterns of the pair rule genes, which further subdivide the embryo and determine the expression pattern of the segment polarity genes (GOTO et al. 1989 ; PANKRATZ et al. 1990 ; SMALL et al. 1992 ; PANKRATZ and JACKLE 1993 ; GROSSNIKLAUS et al. 1994 ). The initial expression pattern of the Hox genes has been attributed to genes at every level of this cascade (JACK and MCGINNIS 1990 ). In particular, it has been shown that the Dfd expression pattern is dependent on the maternal, gap, and pair rule genes (JACK et al. 1988 ; JACK and MCGINNIS 1990 ). Ubx is regulated in a similar fashion (ZHANG et al. 1991 ; ZHANG and BIENZ 1992 ). Once the expression pattern of the Hox genes has been established, later expression becomes, at least in part, dependent on the action of the trithorax group (trxG) and Polycomb group (PcG) genes. The trxG and PcG genes are thought to function in the maintenance of stable expression and repression of Hox gene expression, respectively (MCKEON and BROCK 1991 ; GINDHART and KAUFMAN 1995 ; SOTO et al. 1995 ; KINGSTON et al. 1996 ; WOLFFE 1996 ). The Hox genes themselves are known to act in conjunction with the region-specific homeotics cap'n'collar (cnc), spalt (salm), and teashirt (tsh), thereby specifying segmental identity by regulating the expression of various combinations of target genes (RODER et al. 1992 ; DE ZULUETA et al. 1994 ; KUHNLEIN et al. 1994 ; MOHLER et al. 1995 ; for a review see ROGERS and KAUFMAN 1997 ).

The Hox gene pb is unusual in that it does not confer identity at the level of the segment, but instead acts to modify structures on segments (i.e., limbs) to become specialized for feeding. Adult Drosophila that are homozygous for pb null alleles have their labial palps transformed into legs (KAUFMAN 1978 ). Consistent with this transformation, pb is expressed in the labial discs and central nervous systems of third instar larvae. However, in the Drosophila embryo, which gives rise to a limbless larva, pb serves no known function. Nevertheless, it is expressed in a well-defined pattern during embryogenesis (RANDAZZO et al. 1991 ). In recent years, homeotic mutations in the beetle Tribolium castanaeum have been identified and characterized. Mutations in the Tribolium Hox gene maxillopedia (mxp), the beetle homolog of pb, result in transformation of the larval mouthparts into legs (BEEMAN et al. 1993 ). This result can be interpreted to mean that pb homologs play a functional role in the development of insect embryos outside the higher Diptera (BEEMAN et al. 1993 ). Indeed, comparison of embryonic expression of pb in insect orders other than the Diptera indicates that the expression pattern of pb has been largely conserved over evolutionary time (DENELL et al. 1996 ; ROGERS and KAUFMAN 1997 ). Taken together, these results suggest that pb expression in Drosophila may reflect the existence of ancient regulatory mechanisms that endure despite the apparent nonfunctional nature of the protein in the Drosophila embryo.

Previous analysis of the pb locus has led to the identification of several important regulatory elements (RANDAZZO et al. 1991 ; KAPOUN and KAUFMAN 1995A ). Some of these show a high degree of sequence conservation when compared with similar regions from D. pseudoobscura (RANDAZZO et al. 1991 ). A 500-bp DNA fragment taken from the second intron has been identified as the minimal regulatory element that, when placed in a reporter construct containing the pb promoter, can partially recapitulate the expression pattern of pb in both the embryo and in the imaginal discs (KAPOUN and KAUFMAN 1995A ). In addition to the enhancer elements in the pb locus numerous pairing-sensitive sites have been identified (KASSIS 1994 ; KAPOUN and KAUFMAN 1995B ). These sites are thought to be bound by large multiprotein complexes made up of the products of the PcG and trxG genes (WOLFFE 1996 ; STRUTT et al. 1997 ). However, unlike the pairing-sensitive sites found in the BX-C, the sites from the pb locus have been found to be insensitive to several mutations in either the PcG or trxG genes (KAPOUN and KAUFMAN 1995B ). Recently it has been shown that the genes Dfd and Scr are required for expression of pb in the Drosophila embryo (D. MILLER, S. HOLTZMAN, A. KALKBRENNER and T. C. KAUFMAN, unpublished results). However, pb is expressed in only a subset of the cells in which Dfd or Scr is expressed. In light of this, we have undertaken a systematic analysis of pb expression in various developmental mutants. Here we report the identification of cnc and tsh as the genes responsible for the restriction of pb expression at its anterior and posterior boundaries, respectively. Additionally, we have identified the PcG genes Posterior sex combs (Psc; MARTIN and ADLER 1993 ) and polyhomeotic (ph; DURA et al. 1985 ; DECAMILLIS et al. 1992 ) as being required to maintain repression of pb expression. Using the expression of pb reporter constructs in these mutant backgrounds, we show that all of these genes function through the 500-bp conserved regulatory element taken from the second intron of pb. Additionally, we show that Dfd can bind in vitro to a Dfd consensus binding sequence (CHAN et al. 1997 ) found in this regulatory element. These results describe a regulatory paradigm for pb that is unlike that of the other Hox genes and that may have been conserved during the evolution of the insects.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks and protein detection:
Embryos from the following stocks were collected and assayed using immunohistochemistry: gap gene mutants: hb¹², kni¹, gt^X11, Kr², btd¹, ems¹, oc^YH; pair rule mutants: ftz^W20, h²⁵, opa¹, odd⁵, slp¹, eve¹, eve^15H6a, prd⁹, run^E9; segment polarity mutants: wg^CX4, en^X31, en⁴, hh^AC, arm¹, ptc^IN, nkd²; homeotic mutants: cnc^PZ, lab^VD1 cnc^PZ, Dfd¹⁶, Dfd¹⁶ cnc^PZ, exd¹, salm¹, tsh⁸, tsh⁸ Scr⁴, Scr⁴, Df(2R)Dll^MP; polycomb group mutants: ash1^B1, pho^1(l(4)29), Pcl¹⁰, E(Z)^S1, E(Z)^S6, E(Pc)*, Asx^D1, esc², kto¹, Pc¹, Scm^D1, sxc¹, vtd⁵ ph⁵⁰³, Psc^IIN48; trithorax group mutants: trx³, dev², osa¹, brm², skd¹, urd², sls¹, ash2², mor², kis². Recombination was used to generate stocks doubly mutant for cnc^PZ and the ANT-C homeotics. The P-element reporter construct P{0.5+pbZR} has been described previously (KAPOUN and KAUFMAN 1995A , KAPOUN and KAUFMAN 1995B ). Recombinants between Psc^IIN48 and P{0.5+pbZR} were isolated based on derepression of the white gene located in the P-element construct.

Embryos were fixed and stained essentially as described by KAPOUN and KAUFMAN 1995A . For examination of Pb accumulation antisera directed against the peptide encoded by the second exon of pb (anti-E2) and against the peptide encoded by the ninth exon of pb (anti-E9) were used (CRIBBS et al. 1992 ). Antibodies raised against Scr (GORMAN and KAUFMAN 1995 ), Dfd (MAHAFFEY et al. 1989 ), and Dll (COHEN et al. 1993 ) were also used. Staining of reporter expression was detected using monoclonal mouse anti-ß-galactosidase from Boehringer Mannheim (Indianapolis).

Microscopy and photography:
Embryos were mounted on slides in methyl salicylate. Subsequent examination was performed on a Zeiss axiophot and photographed using ASA100 print film.

Protein purification and gel mobility shift:
Dfd protein purification and gel mobility shifts were performed essentially as described by DESSAIN et al. 1992 with the following minor modifications. Binding was performed in 10 µl containing 100 mM KCl, 20 mM HEPES, pH 7.4, 0.25 mM EDTA, 1 mM dithiothreitol, 10% glycerol, and 100 µg/ml calf thymus DNA (Boehringer Mannheim). Reactions were then run on a BioRad Mini-PROTEAN II gel running apparatus at room temperature for 1 hr at 120 V.

The DNA fragments used in the gel mobility shift were generated by annealing complementary synthetic oligonucleotides (Oligos Etc.), which generated ends that were complementary to EcoRI and XbaI restriction sites. The annealed fragments were then cloned into Bluescript⁺ (Stratagene, La Jolla, CA) and transformed into Subcloning Efficiency DH5 cells (GIBCO-BRL, Gaithersburg, MD). Transformants were amplified and the DNA was isolated using Plasmid Maxiprep (QIAGEN, Chatsworth, CA). The fragments were excised using EcoRI and XbaI, gel purified, precipitated, and then labeled as described by DESSAIN et al. 1992 .

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Expression pattern of proboscipedia and pb reporter constructs:
The expression pattern of pb has been described previously (RANDAZZO et al. 1991 ). Detectable levels of Pb can first be identified in the mesoderm of the mandibular segment at the onset of germband extension (early stage 10; CAMPUS-ORTEGA and HARTENSTEIN 1985 ). Ectodermal accumulation of Pb begins in the maxillary and labial lobes concurrent with their formation just prior to segmentation. As segmentation of the trunk is completed, a ventral stripe of Pb that overlaps the posterior mandible and anterior maxillary segments becomes apparent. At the same time, the mesodermal expression separates into three discrete patches, one of which becomes associated with the mandible. At later stages, Pb accumulates in the brain and along the length of the ventral nerve cord. Lateral Pb accumulation in a stage 12 embryo is shown in Fig 1A and ventral expression is shown in Fig 1B.

View larger version (57K): In this window In a new window Download PPT slide   Figure 1. The expression pattern of endogenous Pb and pb reporter constructs. In all figures embryos are shown with anterior end to the left. (A and B) Composite images showing endogenous accumulation of Pb. Composite images are created through the superimposition of four different images of the same embryo, thereby showing all the different staining aspects that are not normally in the same plane of focus. (A) Lateral view of wild-type Pb. (B) Ventral view of wild-type Pb. (C and D) Expression of ß-galactosidase (ß-gal) protein in embryos containing pb reporter constructs. (C) Embryo containing pb reporter 7. (D) Embryo containing pb reporter 38. Note that reporter 38 displays the ectopic ß-gal accumulation just anterior to the maxillary lobe in the ectoderm of the mandibular lobe. For all embryos in this figure, large solid arrows indicate staining of the labial lobe, large solid arrowheads indicate staining of maxillary lobe; large open arrows point out ventral ectodermal expression; large open arrowheads indicate mandibular mesoderm; small solid arrows point to expression in the brain; and finally, the small solid arrowhead shows prothoracic-associated mesodermal expression.

Regulatory elements from the second intron of pb have been analyzed previously in P-element reporter constructs and their embryonic expression has been assayed (KAPOUN and KAUFMAN 1995A ). One of these constructs that contains a 500-bp pb enhancer and the lacZ gene fused to the pb promoter is designated P{0.5+pbZR}. This construct recapitulates the embryonic and imaginal disc expression patterns of pb. Two of the P{0.5+pbZR} transgenic lines, referred to as reporter #19.2 and reporter #7, located on the X chromosome and second chromosome, respectively, were used in these experiments (KAPOUN and KAUFMAN 1995A ). To facilitate certain crosses, reporter #19.2 was mobilized off the X chromosome and a line isolated with an insertion on the third chromosome referred to as reporter #38. The lacZ expression pattern is shown for both reporter #7 (Fig 1C) and reporter #38 (Fig 1D). Occasionally these reporter constructs express lacZ outside of the normal domain of pb. Both lines display ectopic expression seen in the dorsal ridge and faintly in a lateral striping pattern along the trunk of the embryo (data not shown). Reporter #38 also consistently generates an additional domain of ectopic expression in the mandibular ectoderm (Fig 1D).

pb is positively regulated by Dfd and Scr through an intronic enhancer element:
pb is positively regulated by both the homeotic genes Dfd and Scr (D. MILLER, S. HOLTZMAN, A. KALKBRENNER and T. C. KAUFMAN, unpublished results). The P{0.5+pbZR} reporters contain the smallest identified regulatory element of pb that is sufficient to recapitulate a pb-like pattern of lacZ expression in the maxillary and labial lobes (Fig 1C and Fig D; KAPOUN and KAUFMAN 1995A , KAPOUN and KAUFMAN 1995B ). Here we used immunohistochemistry to assay and compare the expression patterns of the pb gene and pb reporter #7 in embryos mutant for the genes Dfd and Scr.

In Scr null mutant embryos (Scr⁴; REUTER et al. 1990 ) pb expression is markedly reduced but not eliminated from the labial lobe (Fig 2A). Based on the pattern of expression in embryos doubly stained with antibodies against both Pb and the engrailed (en) gene product, Pb is eliminated almost entirely from the posterior compartment of the labial lobe (data not shown). The staining intensity of the remaining expression in the anterior compartment is markedly reduced when compared to pb expression in the neighboring maxillary lobe. In this mutant background, reporter #7 responds even more strongly than does the endogenous gene. lacZ expression is completely eliminated from the labial lobe (Fig 2B). Similarly, in embryos mutant for a null allele of Dfd (Dfd¹⁶; MERRILL et al. 1987B ), pb expression is significantly lower in the maxillary lobe (Fig 2C). In this case, however, residual expression of pb is found in the posterior compartment of the maxillary lobe. Mutations in Dfd do not affect mesodermal expression of pb in the mandible (Fig 2C and Fig E). lacZ expression is completely eliminated from all but the most posterior edge of the maxillary lobe (Fig 2D). Consistent with these results, lacZ expression is completely eliminated from both the maxillary and labial lobes of Dfd¹⁶ Scr⁴ double-mutant embryos (Fig 2F). Interestingly, the ectopic expression of lacZ in the dorsal ridge is also eliminated in the double mutant embryos. However, ectopic trunk expression is unaffected (data not shown). pb expression in the double mutant is also greatly reduced (Fig 2E), more so than would be expected given the results for either of the single mutants. These results are consistent with previous reports that pb is regulated by Dfd and Scr (D. MILLER, S. HOLTZMAN, A. KALKBRENNER and T. C. KAUFMAN, unpublished results). From these results we also conclude that this regulation takes place in part through regulatory elements found in the P{0.5+pbZR} transgene.

View larger version (56K): In this window In a new window Download PPT slide   Figure 2. Comparison of Pb protein accumulation (A, C, and E) vs. ß-gal accumulation (B, D, and F) in embryos mutant for the ANT-C genes Dfd and Scr. All embryos are homozygous for reporter 7. (A and B) Embryos carrying the Scr4 mutation. In the labial lobe (indicated by arrow), accumulation of Pb (A) is reduced while ß-gal accumulation (B) is completely eliminated. (C and D) Dfd16 mutant embryos show reduced Pb accumulation (C) and elimination of ß-gal (D) in the maxillary lobe (maxillary segment indicated by arrowheads). Note that ß-gal is completely absent from the maxillary segment except in a few cells at the posterior edge of the lobe that are known to express Scr protein (RILEY et al. 1987 ; CARROLL et al. 1988 ). (E and F) Dfd16 Scr4 double mutant embryos show a more severe reduction of Pb accumulation (E) and a complete absence of ß-gal (F) accumulation in both the maxillary and labial lobes.

As both Dfd and Scr are known transcription factors, they could act to directly regulate pb expression. It is known that many of the Antennapedia-class homeotic genes bind a core DNA consensus sequence containing the ATTA motif (EKKER et al. 1994 ). Examination of the sequence of the pb minimal regulatory element reveals that it contains four ATTA motifs (RANDAZZO et al. 1991 ). Since it has been shown that Dfd can bind DNA independently of other cofactors in vitro (REGULSKI et al. 1991 ), we decided to examine whether Dfd protein binds to any of the ATTA-containing sequences from the pb enhancer. Using partially purified Dfd, gel mobility shifts were performed on control DNA fragments and on DNA fragments derived from the pb minimal enhancer. The results of these experiments indicate that only one of the ATTA-containing fragments binds Dfd strongly (Fig 3A and Fig B). The fragment bound by Dfd contains a sequence that exactly matches the Pbx-Hox consensus binding site (CHAN et al. 1997 ). This result, in conjunction with the previously mentioned genetic interactions, argues that Dfd binds pb regulatory elements directly in vivo to regulate pb expression.

View larger version (30K): In this window In a new window Download PPT slide   Figure 3. Gel mobility shift assay using Dfd protein performed on putative Dfd binding sites. (A) Gel mobility shift using a control Dfd binding site (Ctrl; DESSAIN et al. 1992 ) as well as three other sites taken from within the 500-bp regulatory element of pb. Presence of partially purified Dfd extract in the mobility shift is indicated by +. The white arrowhead indicates unbound fragments while the black arrowhead indicates shifted fragments. Extracts prepared from cells without the Dfd-expressing plasmid failed to result in shifts (data not shown). (B) Sequence of the 47-bp DNA fragments used in the gel mobility shift. The ATTA homeotic core binding motifs in each fragment are underlined.

Restriction of pb expression by cap'n'collar and teashirt:
Both Dfd and Scr are capable of driving pb expression in the ectoderm of the embryo (D. MILLER, S. HOLTZMAN, A. KALKBRENNER and T. C. KAUFMAN, unpublished results). However, there is no ectodermal expression of pb in either the mandibular lobe, where Dfd is normally expressed, or in the first thoracic segment where Scr is expressed. This suggests that other genes are involved in the regulation of pb expression and act to prevent activation of pb in the segments anterior and posterior to its normal domain of ectodermal expression. In an attempt to identify additional regulators of pb, a survey of Pb accumulation in various mutant backgrounds was performed. The genes included in this survey were picked for their potential to regulate pb on the basis of their timing and pattern of expression. Table 1 shows a summary of the results of this survey. As a consequence of this survey, the genes cnc and tsh were clearly identified as negative regulators of pb expression in the mandible and first thoracic segments, respectively.

tsh encodes a zinc finger-containing transcription factor that is expressed throughout the trunk of the embryo and is involved in the specification of the thoracic and abdominal segments (DE ZULUETA et al. 1994 ). Embryos homozygous for a strong allele tsh⁸ display weak ectopic expression of pb in the first thoracic segment (Fig 4A and Fig B). To test whether this ectopic expression was due to Scr, pb expression was examined in Scr⁴ tsh⁸ double mutants. In the double mutant there is no ectopic expression of pb (data not shown). When lacZ expression from reporter #38 was examined in a tsh⁸ mutant background, surprisingly strong levels of lacZ were found in the first thoracic segment compared to the relatively weak expression of endogenous pb (Fig 4C). This lacZ expression closely resembles the graded expression pattern of Scr normally seen in the first thoracic segment (GORMAN and KAUFMAN 1995 ).

View larger version (67K): In this window In a new window Download PPT slide   Figure 4. Mutations in the genes tsh and cnc result in ectopic expression of both endogenous pb and pb reporters. (A and B) tsh8 mutant embryos showing patches of ectopic accumulation of Pb in the T1 segment (arrows) from lateral (A) and ventral (B) views. (C) Reporter 38 shows strong ectopic ß-gal expression throughout T1 in a tsh8 mutant embryo. Note that ß-gal accumulation in the mandibular lobe and dorsal ridge is normal for this reporter line. (D) Ectopic Pb appears in the posterior ectoderm of the mandible in cncPZ mutant (arrowhead). (E) A close-up view of the head of a cncPZ mutant embryo reveals Pb accumulation in the posterior mandible (arrowhead) and also faint accumulation in the intercalary lobes (open arrows). The Pb expression visible but out of the plane of focus in the ventral mandible is in the mesoderm. (F) Reporter 7 shows strong ectopic ß-gal expression in the mandible (arrowhead) in a cncPZ mutant background.

cnc is a leucine zipper (bZIP class)-containing transcription factor that is expressed in the clypeolabrum and in the mandibular segment (MOHLER et al. 1991 ). Mutations in this gene cause a transformation of the mandible toward a maxillary identity (MOHLER et al. 1995 ). In cnc^PZ mutant embryos, pb is expressed ectopically in the posterior ectoderm of the mandibular lobe (Fig 4D and Fig E). Analysis of embryos mutant for both Dfd¹⁶ and cnc^PZ indicates that this ectopic expression of pb is dependent on Dfd expression (data not shown). Unexpectedly, faint ectopic pb can also be seen ventrally in the intercalary lobes (Fig 4E). We chose to examine whether lab was responsible for the intercalary expression of pb for several reasons. First, mutations in cnc do not significantly alter either the expression of Dfd or Scr (MOHLER et al. 1995 ). Second, Dfd and Scr are the only genes to date that are known to activate pb expression (D. MILLER, S. HOLTZMAN, A. KALKBRENNER and T. C. KAUFMAN, unpublished results). Perhaps lab, being a fellow homeotic gene, can also activate pb under the proper circumstances. lab is expressed in the intercalary segment and is required for embryonic head morphogenesis (MERRILL et al. 1987A ; DIEDERICH et al. 1989 ). Pb accumulation still occurs in the intercalary lobes in a cnc^PZ lab^VD1 double-mutant background (data not shown). This indicates that lab is not responsible for pb expression in the intercalary lobe. When reporter #7 is placed in a cnc^PZ mutant background, strong ectopic lacZ expression can be detected only in the mandibular lobe (Fig 4F).

Effects of gap, pair rule, segment polarity, and other genes on pb expression:
As shown in Table 1, many of the genes that are members of either the gap, pair rule, or segment polarity genes have some effect on the pattern of pb accumulation. For the most part, mutations in genes of these classes reduce the number of cells expressing pb but do not eliminate Pb entirely from the affected segments (data not shown). In no case do they cause pb to accumulate ectopically. The most striking results were caused by zygotic mutations in the genes buttonhead (btd), giant (gt), and hunchback (hb). btd is a head gap gene required for formation of the mandibular segment (COHEN and JURGENS 1990 ). In btd mutants, no mandibular structures are seen and no pb accumulation occurs anterior of the maxillary segment (data not shown). pb accumulation is normal in the other gnathal segments. Mutations in both gt and hb disrupt the formation of the labial lobe (LEHMANN and NUSSLEIN-VOLHARD 1987 ; PETSCHEK et al. 1987 ) and result in concomitant loss of pb expression therein (data not shown). We have also tested the pb reporter #7 in a hb mutant background and found that there is no lacZ expression in the presumptive labial segment (data not shown). In the case of gt, pb expression is not entirely extinguished. Weak pb accumulation can sometimes be seen in the most dorsal and posterior cells of the presumptive labial segment (data not shown), overlapping with the few remaining cells of the engrailed stripe in the labial segment (PETSCHEK and MAHOWALD 1990 ). For both gt and hb, this reduction or loss of pb expression in the labial lobe cannot be attributed to alterations in the Scr pattern as Scr accumulates in the cells posterior to the maxillary segment (data not shown; PETSCHEK and MAHOWALD 1990 ). It is possible that tsh expression has moved anteriorly to block pb expression. To test this we examined pb expression in tsh⁸ hb¹² double-mutant embryos. Because we find no restoration of pb expression in these embryos, we conclude that tsh is not responsible for loss of pb expression in hb mutants.

Nearly all the pair rule and segment polarity genes affect morphology of the gnathal segments and to varying degrees perturb pb accumulation. In general, mutations in the pair rule genes eliminate either the maxillary or labial lobe, as well as reduce the width of the respective segment. Despite these effects on morphology, pb expression can often be seen in the affected segments. Mutations in the segment polarity genes affect the morphology of both the maxillary and labial lobes. The overall effect is a reduction in the size of these lobes, resulting in a correspondingly reduced number of pb-expressing cells. Of the segment polarity genes tested, wingless (wg) has the strongest effect on pb expression. At early stages in wg^CX4 mutants, no pb expression is apparent in the presumptive labial lobe, though later, as head involution commences, some of these cells do begin to express pb.

A number of other developmentally important genes that we examined in the survey are noteworthy for their apparent lack of effect on pb accumulation. The first of these genes is extradenticle (exd), known as pbx in vertebrates, which encodes a homeobox-containing protein and a known cofactor of several vertebrate and invertebrate Hox genes (VAN DIJK and MURRE 1994 ; RYOO and MANN 1999 ). The Dfd binding site we identified presumably would also bind Exd in conjuction with Dfd (CHAN et al. 1997 ). Surprisingly, exd¹ zygotic mutants showed no effect on pb accumulation. Distal-less (Dll) is another homeobox-containing transcription factor and is required for the formation of limbs (COHEN et al. 1989 ). Dll is expressed in both the maxillary and labial lobes in a fashion analogous to pb, but mutations in Dll have no effect on pb accumulation. Another intriguing gene is salm, which encodes a zinc-finger transcription factor expressed in the gnathal segments and is thought to be required for proper specification of gnathal identity (KUHNLEIN et al. 1994 ). However, salm¹ mutants had no effect on pb expression.

Polycomb group interactions with pb:
After the pattern of Hox gene expression has been established, maintenance of this pattern is dependent on the function of the PcG and trxG genes (MCKEON and BROCK 1991 ; SOTO et al. 1995 ; KINGSTON et al. 1996 ). Here, we assayed accumulation of Pb in embryos mutant for various PcG and trxG genes (see MATERIALS AND METHODS for stocks). The PcG genes Psc and ph were the only genes from either group that showed a detectable interaction with pb. In Psc^IIN48 mutants, ectopic pb accumulates at high levels in the antennal segment and in a complicated pattern throughout the trunk segments (Fig 5A and Fig B). In the abdomen, the ectopic accumulation of pb occurs as a pair of stripes in each segment. Additionally, ectopic pb accumulates in the leg anlagen. Similarly, ph⁵⁰³ mutants result in ectopic accumulation of pb in the antennal segment and in the trunk segments; the pattern is not as clearly defined as in Psc mutants and there is a lower accumulation of pb in the leg anlagen (Fig 5C). While both Dfd and Scr are ectopically expressed in zygotically mutant Psc embryos, neither is expressed in the antennal segment, in the leg anlagen, or in the abdominal segments (data not shown; MCKEON and BROCK 1991 ). Interestingly, neither ph nor Psc causes ectopic accumulation of ß-galactosidase from the P{0.5+pbZR} reporters (data not shown). However, using an eye color assay (KAPOUN and KAUFMAN 1995B ) we see derepression of the P-element white mini-gene in Psc^IIN48 heterozygous adult flies carrying the pb reporter constructs (data not shown). This suggests that regulatory elements in the reporter used here can mediate the activity of the PcG genes but apparently only in the imaginal discs. Moreover, since the pattern of expression of the resident pb locus is changed in Psc and ph mutant backgrounds there must be cis-acting sequences in the native gene that reside outside the tested fragment that are responsible for the observed interaction. Further investigation will be necessary to map these components.

View larger version (58K): In this window In a new window Download PPT slide   Figure 5. PcG mutants result in ectopic accumulation of Pb throughout the embryo. (A) Ventral surface of a PscIIN48 embryo showing ectopic accumulation of Pb in the leg anlagen (arrow). (B) PscIIN48 mutant showing ectopic Pb accumulation in cells of the lateral ectoderm along the entire length of the embryo. Arrowhead indicates ectopic Pb in the fourth abdominal segment. (C) Lateral view of Pb expression in a ph503 mutant embryo. Besides showing highly abnormal morphology, Pb accumulates in the antennal segment (white arrow), laterally along the length of the embryo (arrowhead), and in cells in the center of the leg anlagen (arrow).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have examined pb expression during embryogenesis. While pb plays no known functional role during embryogenesis, its pattern of expression is clearly restricted to the gnathal region where it is known to function during adult development. Our results provide a mechanism by which pb expression is established and that may also serve to determine pb expression subsequently in adult tissues.

Temporal and spatial regulation of pb:
On the basis of these results and previous work, we have developed a working model for the regulation of pb. This model accounts for both temporal and spatial aspects of pb expression (Fig 6). In effect, the regulation of pb can be broken down into early, middle, and late phases. The early phase represents the period prior to the onset of pb expression, during which the gap genes define the domain of pb expression through a presumably indirect mechanism. During the middle phase, the genes cnc, Dfd, Scr, and tsh act to establish the initial expression pattern of pb along the A/P axis. During the late phase we propose that the PcG and trxG genes assume responsibility for maintaining the pattern of pb expression through the later stages of embryogenesis.

View larger version (30K): In this window In a new window Download PPT slide   Figure 6. A model depicting the temporal and spatial regulation of pb during embryogenesis. The regulation of pb has been divided into three phases referred to as early, middle, and late. The early phase corresponds to the time during development when the gap genes are necessary to permit expression of pb during the middle and late phases. The broken lines, depicting regulation of pb by the gap genes, are shown to suggest that these interactions are indirect in nature. During the middle phase, the initial expression pattern of pb is established. The known regulatory interactions, which may be mediated directly or indirectly, are represented by unbroken black lines. The nonautonomous influences, possibly mediated through wg signaling, are represented by dashed gray lines. Finally, during the late phase, the PcG and trxG genes act to prevent further changes in the pb expression pattern despite alterations in the expression patterns of the initial regulators of pb (Scr expression in T1). The overriding activity of the PcG and trxG genes is represented by the gray shading of the squares. IC, intercallary segment; MN, mandibular segment; MX, maxillary segment; LB, labial segment; T1, first thoracic segment; T2, second thoracic segment.

The early phase reflects a requirement for gap gene function for normal expression of pb to occur during later stages. Specifically, btd, gt, and hb have been identified as being required for proper gnathal expression of pb. The function of the head gap gene btd has been shown to be required only during early stages of embryogenesis (WIMMER et al. 1997 ). The expression patterns of gt and hb are such that they are no longer expressed in the labial segment at the time when pb expression begins (BENDER et al. 1988 ; ELDON and PIRROTTA 1991 ). We take this as a strong indication that the gap genes influence pb indirectly. Consistent with this hypothesis, we are unable to identify any gt or hb binding sites (STANOJEVIC et al. 1989 ; ZHANG et al. 1991 ; ZHANG and BIENZ 1992 ) in the regulatory elements of the pb reporter. In the case of hb, we have investigated the role that various trans-acting factors might play in mediating loss of pb expression in the labial segment. We find that expression of Scr, the positive regulator of pb in the labial segment, is not eliminated. Further, we have shown that repression of pb is not attributable to expansion of tsh expression. One possibility is that another negative regulator is being expressed such that Scr can no longer activate pb. Given the negative regulatory interactions that occur between the gap genes (CAPOVILLA et al. 1992 ; STRUHL et al. 1992 ), it is possible that one of the other gap genes might be misexpressed and downregulate pb. On the other hand, it may be misexpression of cnc or some other gene that has yet to be identified. Alternatively, the "hit-and-run" hypothesis, proposed by ZHANG and BIENZ 1992 to explain the long-term repression of Ultrabithorax (Ubx) by hb, may describe how transient expression of the gap genes is required very early in development to permit later expression of pb. In their hypothesis, heritable changes in chromatin structure, mediated by the PcG genes, were invoked to explain how hb regulates Ubx long after hb expression has ceased. In the case of pb regulation, one or more of these gap genes may be required to alter chromatin structure in and around the pb locus, thereby allowing the various trans-acting factors access to the pb cis-acting regulatory elements.

During the middle phase, the initial expression pattern of pb is set by a variety of trans-acting factors. Our focus has been on the identification of those factors that determine the ectodermal pattern of pb expression along the A/P axis of the embryo. We have confirmed that the Hox genes Dfd and Scr act as positive regulators of pb (D. MILLER, S. HOLTZMAN, A. KALKBRENNER and T. C. KAUFMAN, unpublished results) and demonstrated that Dfd can bind to pb regulatory elements in vitro. We think it likely that Scr also regulates pb directly based on the similarity with which mutations in Dfd and Scr affect expression of pb and the pb reporter. In addition to the Hox genes, the region-specific homeotics cnc and tsh have been identified as negative regulators of pb and serve to restrict pb expression to the gnathos. It is not clear whether these genes regulate pb directly, though in the case of tsh we have identified the sequence TGGAAAAGT in the 500-bp regulatory fragment used in the pb reporter; this sequence is very similar to the identified tsh binding site (ALEXANDRE et al. 1996 ). While this regulatory paradigm does not completely describe the regulation of the endogenous gene, based on the presence of pb residual expression, it is sufficient to explain the behavior of the 500-bp pb reporter. This mechanism of regulation places pb downstream of the Hox genes and is the first instance in Drosophila where one Hox gene is positively and directly regulated by another, a distinction previously accorded only to vertebrate Hox genes (GOULD et al. 1997 ). Others (D. MILLER, S. HOLTZMAN, A. KALKBRENNER and T. C. KAUFMAN, unpublished results) have suggested that wg may be mediating the nonautonomous residual expression of pb that is uncovered by mutations in Dfd and/or Scr. With the exception that wg has the strongest effect on pb expression of the segment polarity genes tested, our results shed little light on the mechanism that underlies this phenomenon. However, signalling has been implicated to explain regulation of ectodermal pb function by mesodermal expression of Scr; perhaps the residual expression in the embryo is an example of this pathway (PERCIVAL-SMITH et al. 1997 ). Further experiments, including identification of an enhancer that mediates this residual expression, are needed.

Finally, during the late phase, we have identified two PcG genes that are involved in maintaining repression of pb outside its normal domain of expression. This result supersedes a previous report that the PcG genes do not regulate pb (KAPOUN and KAUFMAN 1995B ). We have yet to identify any trxG genes that are required for the maintenance of pb expression. To function, the PcG genes are thought to assemble on DNA in large multimeric complexes (FRANKE et al. 1992 ). Unlike the genes of the BX-C, which are regulated, to greater and lesser extent, by all the PcG genes that have been tested (MCKEON and BROCK 1991 ; SIMON et al. 1992 ), pb is not regulated by the majority of known PcG genes. Assuming that the PcG genes function similarly at the pb locus, the implication is that not all multimeric complexes can be equal. However, it is not clear how these differences are established. One possibility is that complexes composed of different combinations of PcG genes are formed at different times during development, thereby regulating different loci (WOLFFE 1996 ). Interestingly, a vertebrate homolog of Psc has been shown to bind a specific DNA sequence. This exact sequence is also found in the regulatory elements of the pb reporter construct, indicating that Psc may bind directly, though this remains to be shown (KANNO et al. 1995 ).

In addition to forming quantitatively different complexes, the timing of complex formation may be crucial to the proper expression of pb. Normally, the PcG genes are required after the expression pattern of pb has been established to prevent ectopic expression. Presumably, this ectopic expression would result from newly expressed transcription factors acting inappropriately at the pb locus. This hypothesis may offer an explanation for the differences seen between the expression pattern of endogenous pb and the pb reporter in tsh mutant backgrounds. In this scenario, Scr expression in T1 begins concurrently with the initiation of PcG-mediated repression at the pb locus. In the absence of tsh, competition between activation by Scr and repression by the PcG genes results in the weak and variable pb expression. Because the PcG genes do not regulate ß-galactosidase expression from the pb reporter, it is free to respond strongly to Scr accumulation in T1. Further experiments are required to support this hypothesis.

The evolution of regulation of pb in other insects:
In Drosophila, pb plays a role in specifying limbs to become specialized for feeding. In other insects where it has been examined, pb is expressed in and required for the formation of the larval mouthparts (DENELL et al. 1996 ; for review see ROGERS and KAUFMAN 1997 ). Overall, the expression patterns of Dfd and Scr have also been conserved in other insects (FLEIG et al. 1992 ; KOKUBO et al. 1997 ; ROGERS et al. 1997 ; ROGERS and KAUFMAN 1997 ). Given these results, it is interesting to speculate about whether the regulation of pb may be conserved in other insects. Preliminary results from studies in Tribolium suggest that at least some aspects of pb regulation may be conserved. A deficiency that deletes many of the Tribolium Hox genes except for the lab, pb, and Abd-B homologues has been isolated. Embryos homozygous for this deficiency display a mutant phenotype in which every segment of the Tribolium larva is transformed toward an antennal segment and bears a pair of antennae (STUART et al. 1991 ). These larvae have no mouthparts or walking legs. Further, gain-of-function mutations in Tribolium pb, which result in ectopic expression of Tribolium pb in the antennal segment, are known to transform the antennae into generic feeding palps (DENELL et al. 1996 ). Given these two results, we can infer that pb is not being expressed in embryos containing this deficiency because the antennae are not transformed into feeding palps. If pb is not being expressed, this implies that the positive regulators of pb have been removed by this deficiency. The simplest interpretation that is consistent with these results and the results presented here is that the Dfd and Scr homologs in Tribolium may be required to positively regulate the Tribolium pb much like their counterparts do in Drosophila. Actual proof of this possibility requires in situ data on the expression pattern of the pb homolog in various Tribolium Hox mutants.

	ACKNOWLEDGMENTS

We thank W. McGinnis for the Dfd-expressing bacterial cell line and for his helpful advice concerning the gel mobility shifts; M. Peterson, A. Popadic, and B. Rogers for their helpful discussions and reviews of this manuscript; D. Noecker for his help with the antibody staining of the embryos; D. Verostko for administrative assistance; K. Matthews and the Bloomington Stock Center, and the Kaufman lab as a whole for their support and assistance without which this paper would not be possible. This work was supported by the Howard Hughes Medical Institute (HHMI) and by a National Institutes of Health Predoctoral Fellowship (GM07757) to D.B.R.; T.C.K. is an Investigator of the HHMI.

Manuscript received February 28, 2000; Accepted for publication May 1, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALEXANDRE, E., Y. GRABA, L. FASANO, A. GALLET, and L. PERRIN et al., 1996  The Drosophila Teashirt homeotic protein is a DNA-binding protein and modulo, a HOM-C regulated modifier of variegation, is a likely candidate for being a direct target gene. Mech. Dev. 59:191-204.

BEEMAN, R. W., J. J. STUART, S. J. BROWN, and R. E. DENELL, 1993  Structure and function of the homeotic gene complex (HOM-C) in the beetle, Tribolium castaneum.. Bioessays 15:439-444.

BENDER, M., S. HORIKAMI, D. CRIBBS, and T. C. KAUFMAN, 1988  Identification and expression of the gap segmentation gene hunchback in Drosophila. Dev. Genet. 9:715-732.

CAMPUS-ORTEGA, J., and V. HARTENSTEIN, 1985 The Embryonic Development of Drosophila melanogaster. Springer-Verlag, New York.

CAPOVILLA, M., E. D. ELDON, and V. PIRROTA, 1992  The giant gene of Drosophila encodes a b-ZIP DNA-binding protein that regulates the expression of other segmentation gap genes. Development 114:99-112.

CARROLL, S., S. DINARDO, P. O'FARRELL, R. WHITE, and M. SCOTT, 1988  Temporal and spatial relationships between segmentation and homeotic gene expression in Drosophila embryos: distributions of the fushi tarazu, engrailed, Sex combs reduced, Antennapedia, and Ultrabithorax proteins. Genes Dev. 2:350-360.

CHAN, S.-K., H. D. RYOO, A. GOULD, R. KRUMLAUF, and R. S. MANN, 1997  Switching the in vivo specificity of a minimal Hox-responsive element. Development 124:2007-2014.

COHEN, B., A. A. SIMCOX, and S. M. COHEN, 1993  Allocation of the thoracic imaginal primordia in the Drosophila embryo. Development 117:597-608.

COHEN, S. M. and G. JURGENS, 1990  Mediation of Drosophila head development by gap-like segmentation genes. Nature 246:482-485.

COHEN, S. M., G. BRONNER, F. KUTTNER, G. JURGENS, and H. JÄCKLE, 1989  Distal-less encodes a homeodomain protein required for limb development in Drosophila. Nature 338:432-434.

CRIBBS, D. L., M. A. PULTZ, D. JOHNSON, M. MAZZULA, and T. C. KAUFMAN, 1992  Structural complexity and evolutionary conservation of the Drosophila homeotic gene proboscipedia.. EMBO J. 11:1437-1449.

DECAMILLIS, M., N. S. CHENG, D. PIERRE, and H. W. BROCK, 1992  The polyhomeotic gene of Drosophila encodes a chromatin protein that shares polytene chromosome-binding sites with Polycomb. Genes Dev. 6:223-232.

DENELL, R. E., S. J. BROWN, and R. W. BEEMAN, 1996  Evolution of the organization and function of insect homeotic complexes. Semin. Cell Dev. Biol. 7:527-538.

DESSAIN, S., C. T. GROSS, M. A. KURIORA, and W. MCGINNIS, 1992  Antp-type homeodomains have distinct DNA binding specificities that correlate with their different regulatory functions in embryos. EMBO J. 11:991-1002.

DE ZULUETA, P., E. ALEXANDRE, B. JACQ, and S. KERRIDGE, 1994  Homeotic complex and teashirt genes co-operate to establish trunk segmental identities in Drosophila. Development 120:2278-2296.

DIEDERICH, R. J., V. K. MERRILL, M. A. PULTZ, and T. C. KAUFMAN, 1989  Isolation, structure, and expression of labial, a homeotic gene of the Antennapedia complex involved in Drosophila head development. Genes Dev. 3:399-414.

DURA, J. M., H. W. BROCK, and P. SANTAMARIA, 1985  polyhomeotic: a gene of Drosophila melanogaster required for correct expression of segmental identity. Mol. Gen. Genet. 198:213-220.

EKKER, S. C., D. G. JACKSON, D. P. VON KESSLER, B. I. SUN, and K. E. YOUNG et al., 1994  The degree of variation in DNA sequence recognition among four Drosophila homeotic proteins. EMBO J. 13:3551-3560.

ELDON, E. D. and V. PIRROTTA, 1991  Interactions of the Drosophila gap gene giant with maternal and zygotic pattern-forming genes. Development 111:367-378.

FLEIG, R., U. WALLDORF, W. J. GEHRING, and K. SANDER, 1992  Development of the Deformed protein pattern in the embryo of the honeybee Apis melliferia L. (Hymenoptera). Roux's Arch. Dev. Biol. 201:235-242.

FRANKE, A., M. DECAMILLIS, D. ZINK, N. CHENG, and H. BROCK et al., 1992  Polycomb and polyhomeotic are constituents of a multimeric protein complex in chromatin of Drosophila melanogaster.. EMBO J. 11:2941-2950.

GINDHART, J. G., JR. and T. C. KAUFMAN, 1995  Identification of Polycomb and trithorax group responsive elements in the regulatory region of the Drosophila homeotic gene Sex combs reduced.. Genetics 139:797-814.

GORMAN, M. J. and T. C. KAUFMAN, 1995  Genetic analysis of cis-acting regulatory elements of the Drosophila homeotic gene Sex combs reduced.. Genetics 140:557-572.

GOTO, T., P. MACDONALD, and A. T. MANIATIS, 1989  Early and late periodic patterns of even-skipped expression are controlled by distinct regulatory elements that respond to different spatial cues. Cell 57:413-422.

GOULD, A., A. MORRISON, G. SPROAT, R. A. H. WHITE, and R. KRUMLAUF, 1997  Positive cross-regulation and enhancer sharing: two mechanisms for specifying overlapping Hox expression patterns. Genes Dev. 11:900-913.

GROSSNIKLAUS, U., K. M. CADIGAN, and W. J. GERHING, 1994  Three maternal coordinate systems cooperate in the patterning of the Drosophila head. Development 120:3155-3171.

INGHAM, P. W., 1988  The molecular genetics of embryonic pattern formation in Drosophila. Nature 335:25-34.

JACK, T. and W. MCGINNIS, 1990  Establishment of the Deformed expression stripe requires the combinatorial action of coordinate, gap and pair-rule proteins. EMBO J. 9:1187-1198.

JACK, T., M. REGULSKI, and W. MCGINNIS, 1988  Pair-rule segmentation genes regulate the expression of the homeotic selector gene, Deformed.. Genes Dev. 2:635-651.

KANNO, M., M. HASEGAWA, A. ISHIDA, K. ISONO, and M. TANIGUCHI, 1995  mel-18, a Polycomb group-related mammalian gene, encodes a transcriptional negative regulator with tumor suppressive activity. EMBO J. 14:5672-5678.

KAPOUN, A. M. and T. C. KAUFMAN, 1995a  A functional analysis of 5', intronic and promoter regions of the homeotic gene proboscipedia in Drosophila melanogaster.. Development 121:2127-2141.

KAPOUN, A. M. and T. C. KAUFMAN, 1995b  Regulatory regions of the homeotic gene proboscipedia are sensitive to chromosomal pairing. Genetics 140:643-658.

KASSIS, J. A., 1994  Unusual properties of regulatory DNA from the Drosophila engrailed gene: three "pairing-sensitive" sites within a 1.6-kb region. Genetics 136:1025-1038.

KAUFMAN, T. C., 1978  Cytogenetic analysis of chromosome 3 in Drosophila melanogaster: isolation and characterization of four new alleles of the proboscipedia (pb) locus. Genetics 90:579-596.

KAUFMAN, T. C., M. A. SEEGER, and G. OLSEN, 1990  Molecular and genetic organization of the Antennapedia gene complex of Drosophila melanogaster.. Adv. Genet. 27:309-362.

KINGSTON, R. E., C. A. BUNKER, and A. N. IMBALZANO, 1996  Repression and activation by multiprotein complexes that alter chromatin structure. Genes Dev. 10:905-920.

KOKUBO, H., K. UENO, K. AMANAI, and Y. SUZUKI, 1997  Involvement of the Bombyx Scr gene in development of the embryonic silk gland. Dev. Biol. 186:46-57.

KÜHNLEIN, R. P., G. FROMMER, M. FRIEDRICH, M. GONZALEZ-GAITAN, and A. WEBER et al., 1994  spalt encodes an evolutionarily conserved zinc finger protein of novel structure which provides homeotic gene function in the head and tail region of the Drosophila embryo. EMBO J. 13:168-179.

LAWRENCE, P. A., 1992 The Making of a Fly: The Genetics of Animal Design, pp. 25–77. Alden Press, Oxford.

LEHMANN, R. and C. NÜSSLEIN-VOLHARD, 1987  hunchback, a gene required for segmentation of an anterior and posterior region of the Drosophila embryo. Dev. Biol. 119:402-417.

LEWIS, E. B., 1978  A gene complex controlling segmentation in Drosophila. Nature 276:565-570.

MAHAFFEY, J. W., R. J. DEIDERICH, and T. C. KAUFMAN, 1989  Novel patterns of homeotic protein accumulation in the head of the Drosophila embryo. Development 105:167-174.

MARTIN, E. C. and P. N. ADLER, 1993  The Polycomb group gene Posterior sex combs encodes a chromosomal protein. Development 117:641-655.

MCKEON, J. and H. W. BROCK, 1991  Interactions of the Polycomb group of genes with homeotic loci of Drosophila. Roux's Arch. Dev. Biol. 199:387-396.

MERRILL, V. K., R. J. DIEDERICH, F. R. TURNER, and T. C. KAUFMAN, 1987a  A genetic and developmental analysis of mutations in labial, a gene necessary for proper head formation in Drosophila melanogaster.. Dev. Biol. 135:376-391.

MERRILL, V. K., F. R. TURNER, and T. C. KAUFMAN, 1987b  A genetic and developmental analysis of mutations in the Deformed locus in Drosophila melanogaster.. Dev. Biol. 122:379-395.

MOHLER, J., K. VANI, S. LEUNG, and A. EPSTEIN, 1991  Segmentally restricted, cephalic expression of a leucine zipper gene during Drosophila embryogenesis. Mech. Dev. 34:3-9.

MOHLER, J., J. W. MAHAFFEY, E. DEUTSCH, and K. VANI, 1995  Control of Drosophila head segment identity by the bZIP homeotic gene cnc.. Development 121:237-247.

MORATA, G., 1993  Homeotic gene of Drosophila. Curr. Opin. Genet. Dev. 3:606-614.

NÜSSLEIN-VOLHARD, C. and E. WEISCHAUS, 1980  Mutations affecting segment number and polarity in Drosophila. Nature 287:795-801.

PANKRATZ, M. J., and H. JÄCKLE, 1993 Blastoderm segmentation, pp. 467–516 in The Development of Drosophila melanogaster, edited by M. BATE and A. MARTINEZ ARIAS. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

PANKRATZ, M. J., E. SEIFERT, N. GERWIN, B. BILLI, and U. NAUBER et al., 1990  Gradients of krüpple and knirps gene products direct pair-rule gene stripe patterning in the posterior region of the Drosophila embryo. Cell 61:309-317.

PERCIVAL-SMITH, A., J. WEBER, E. GILFOYLE, and P. WILSON, 1997  Genetic characterization of the role of the two HOX proteins, Proboscipedia and Sex Combs Reduced, in determination of adult antennal, tarsal, maxillary palp and proboscis identities in Drosophila melanogaster. Development 124:5049-5062.

PETSCHEK, J. P. and A. P. MAHOWALD, 1990  Different requirements for l(1)giant in two embryonic domains of Drosophila melanogaster.. Dev. Genet. 11:88-96.

PETSCHEK, J., N. PERRIMON, and A. P. MAHOWALD, 1987  Region specific effects in l(1)giant embryos of Drosophila. Dev. Biol. 119:177-189.

RANDAZZO, F. M., D. L. CRIBBS, and T. C. KAUFMAN, 1991  Rescue and regulation of proboscipedia: a homeotic gene of the Antennapedia complex. Development 113:257-271.

REGULSKI, M., S. DESSAIN, N. MCGINNIS, and W. MCGINNIS, 1991  High-affinity binding sites for Deformed protein are required for the function of an auto-regulatory enhancer of the Deformed gene. Genes Dev. 5:278-286.

REUTER, R., G. E. PANGANIBAN, F. M. HOFFMANN, and M. P. SCOTT, 1990  Homeotic genes regulate the spatial expression of putative growth factors in the visceral mesoderm of Drosophila embryos. Development 110:1031-1040.

RILEY, P., S. CARROLL, and M. SCOTT, 1987  The expression and regulation of Sex combs reduced protein in Drosophila embryos. Development 109:1289-1303.

RÖDER, L., C. VOLA, and S. KERRIDGE, 1992  The role of the teashirt gene in trunk segmental identity in Drosophila.. Development 115:1017-1033.

ROGERS, B. T. and T. C. KAUFMAN, 1997  Structure of the insect head in ontogeny and phylogeny: a view from Drosophila. Int. Rev. Cytol. 174:1-84.

ROGERS, B. T., M. D. PETERSON, and T. C. KAUFMAN, 1997  Evolution of the insect body plan as revealed by the Sex combs reduced expression pattern. Development 124:149-157.

RYOO, H. D. and R. S. MANN, 1999  The control of trunk Hox specificity and activity by Extradenticle. Genes Dev. 13:1704-1716.

SIMON, J., A. CHIANG, and W. BENDER