Local Transposition of a hobo Element Within the decapentaplegic Locus of Drosophila

Corresponding author: Stuart J. Newfeld, Department of Biology, Mail Code 1501, Arizona State University, Tempe, AZ 85287-1501., newfeld{at}asu.edu (E-mail)

Communicating editor: V. G. FINNERTY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have efficiently mobilized a phenotypically silent hobo transgene inserted within the cis-regulatory heldout region of the decapentaplegic (dpp) locus in Drosophila melanogaster. The goal of our experiment was to identify germline transmission of a local transposition event within the dpp locus that meets two specific criteria. First, excision of the hobo construct does not generate an adult mutant phenotype, suggesting minimal alteration to the original site of insertion. Second, we required a new insertion of the hobo transgene into the Haploinsufficient region of the locus ~25 kb away. Genetic and molecular criteria are used to evaluate candidate germlines. In a pilot study, this local transposition event occurred independently in two individuals. Both of the transposition events appear to be new insertions into the dpp transcription unit. One insertion is between the two protein-coding exons, and the other is in the 3'-untranslated region of exon three. Strains carrying these insertions are valuable new reagents for the analysis of dpp function and molecular evolution. These results further support the use of the hobo system as an important tool in Drosophila genetics.

ENDOGENOUS, transposable elements have been developed into powerful genetic tools in a wide variety of species. In Drosophila melanogaster, experimental systems based on the P and hobo elements are used widely. These transposons belong to the same superfamily of mobile elements, those displaying inverted terminal repeats and transposing via DNA intermediates (HARTL and LOZOVSKAYA 1995 ). As a result, their respective genetic systems share many characteristics. For example, both P and hobo transposon systems are capable of efficient germline transformation (BLACKMAN et al. 1987 ) and enhancer trapping mutagenesis (SMITH et al. 1993 ). One feature of the P-element system not yet fully explored for hobo is local transposition (TOWER et al. 1993 ; ZHANG and SPRADLING 1993 ). Here, we describe experiments designed to recover new insertions in a specific genomic region that result from the local jump of a hobo transgene. Our strategy is based on the well-characterized molecular genetics of the decapentaplegic (dpp) locus.

The dpp gene encodes a secreted signaling protein belonging to the transforming growth factor-ß (TGF-ß) family (PADGETT et al. 1987 ) that is required for a variety of developmental decisions. Members of the TGF-ß family share the following features: precursor polypeptides are proteolytically cleaved to generate an N-terminal fragment (the proregion) thought to be involved in dimerization and a C-terminal fragment that forms the biologically active ligand (reviewed in MASSAGUE 1990 ). Among the developmental events influenced by dpp are the determination of dorsal ectoderm in the early embryo (IRISH and GELBART 1987 ), gut morphogenesis (IMMERGLUCK et al. 1990 ; PANGANIBAN et al. 1990 ), and proper differentiation of adult wings (POSAKONY et al. 1991 ). By several criteria, the developmental functions of dpp were separated into three genetic domains [shortvein (shv), Haploinsufficient (Hin), and imaginal disk specific (disk)] that span nearly 60 kb (ST. JOHNSTON et al. 1990 ).

The Hin region is a roughly 8-kb block of DNA that contains one of the five classes of dpp transcript (including the common protein-coding exons 2 and 3) and all regulatory sequences necessary for normal dorsal-ventral patterning of the embryo (HOFFMANN and GOODMAN 1987 ). Embryos that contain only a single functional copy of this region are inviable because of severe defects along the dorsal-ventral axis (WHARTON et al. 1993 ). On either side of the Hin region are large arrays of cis-regulatory sequences. The shv region lies distal to the Hin region. The shv region contains four alternatively spliced first exons and regulatory sequences that govern expression in the embryonic epidermis and midgut (ST. JOHNSTON et al. 1990 ; HURSH et al. 1993 ). The disk region lies proximal to the Hin region and is not transcribed. This region is composed of 25 kb of regulatory sequence beginning ~2 kb beyond the polyadenylation site in dpp transcripts. Sequences in the disk region control expression of dpp along the anterior/posterior compartment boundary of imaginal disks. Seven distinct enhancers have been identified using reporter genes, and each directs expression within a subset of the overall dpp pattern (BLACKMAN et al. 1991 ). One of the most distant enhancers is phenotypically uncovered by a small deletion of 2 kb (dpp^d-ho; ST. JOHNSTON et al. 1990 ). When homozygous, this deletion fixes the wings in a heldout position because of the absence of specific sensory structures (SPENCER et al. 1982 ). The specific sequences responsible for this phenotype are unknown.

A strain was created in which a hobo transgene was inserted within the heldout enhancer region (SMITH et al. 1993 ). Interestingly, the strain H{Lw2}dpp^151h is homozygous viable and fertile, and there is no obvious mutant phenotype. However, the allele H{Lw2}dpp^151a created by imprecise excision of the hobo construct displayed the heldout phenotype when in trans to dpp^d-ho (SMITH et al. 1993 ). To critically examine the usefulness of hobo elements for local jumping mutagenesis, we designed an experiment to identify a specific local transposition event within the dpp locus. We desired a precise excision of H{Lw2}dpp^151h from the heldout region followed by a new insertion into the Hin region ~25 kb away. In a pilot screen, we identified two such events. Both transposition events appear to result in insertions within the dpp transcription unit. Strains containing these new insertions represent valuable reagents for the analysis of dpp and for further studies of the hobo element system.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly strains and genetic characterization of hobo mobilization:
All stocks were devoid of endogenous hobo elements (E strains) unless otherwise indicated. The original line y¹ w^67c23; H{Lw2}dpp^151h containing a hobo transgene inserted in the heldout region of dpp is described in SMITH et al. 1993 . The Sp Bl Dp(2;2)DTD48 dpp^d-ho strain carrying a duplication of the dpp locus is described in SPENCER et al. 1982 . The composite parental strain y¹ w^67c23; H{Lw2}dpp^151h Dp(2;2)DTD48 dpp^d-ho was created by recombination; individual w⁺ (from the mini-white marked hobo construct) recombinant males were tested for the ability to suppress lethality resulting from dpp haplo-insufficiency by crossing to dpp^H61-bearing females. The dpp^H61 line contains endogenous hobo elements (H strain). The net dpp^d-ho dp Sp cn sca bw transvection tester strain was not examined for the presence of hobo elements. Heat-shock-induced mobilization of H{Lw2}dpp^151h by P{ry⁺: HSH2}CyO was conducted as described in CALVI and GELBART 1994 with the following modification: cultures were brooded every 2 days and heat shocked every 2 days for a total of three heat shocks for each of the three broods. All dpp mutant strains and visible phenotypic markers are described in FLYBASE 1998 .

Molecular characterization of hobo transposition within the dpp locus:
Standard methods of DNA isolation, digestion, and Southern blot hybridization were used (SAMBROOK et al. 1989 ). To demonstrate hobo mobilization in each candidate line, genomic DNA was digested with SspI. A Southern blot containing this DNA was hybridized with pSV-ß-gal (Promega, Madison, WI). pSV-ß-gal is a probe for ß-galactosidase sequences located in the H{Lw2} transgene. For the identification of new hobo insertions in the 8-kb Hin region, genomic DNA from each candidate line was digested with EcoRI. A Southern blot containing this DNA was hybridized with the dpp cDNA H1. H1 is a full-length cDNA reverse transcribed from a "class B" dpp transcript (NEWFELD et al. 1997 ). The same Southern blot was then stripped of probe and rehybridized with pH{Lw2}. For the identification of new hobo insertions in a 5-kb region bounded by the two dpp protein-coding exons, genomic DNA from each candidate line was digested with NheI (bp 12732) and ScaI (bp 17630). A Southern blot containing this DNA was hybridized with H1 and then stripped and rehybridized with pH{Lw2}. For the identification of new hobo insertions in dpp protein-coding exons 2 and 3, genomic DNA from each candidate line was digested with XbaI. There are four XbaI sites in the Hin region (bp 12772, 13754, 14853, and 17073). The first pair of sites nearly brackets exon 2 (bp 12607–13474), and the second pair of sites brackets the exon 3 open reading frame (bp 15192–16090). Numbers in parentheses indicate the location of the feature in the dpp shv/Hin sequence (GenBank accession number U63857; NEWFELD et al. 1997 ).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Mobilization of H{Lw2}dpp^151h:
The starting point for our examination of hobo local transposition was the transgenic line H{Lw2}dpp^151h (SMITH et al. 1993 ). We planned to identify a specific hobo-mediated event, precise excision of H{Lw2}dpp^151h from the heldout region followed by a new insertion into the Hin region. A composite physical and genetic map of the dpp locus (polytene subdivision 22F1-2 on chromosome arm 2L) in this transgenic line is shown in Figure 1A (adapted from NEWFELD et al. 1997 ). Note the relative position of the Hin and heldout regions, at least 20 kb apart. The 2-kb heldout region contains the phenotypically silent hobo transgene, and the 8-kb Hin region is our local jump target.

View larger version (20K): In this window In a new window Download PPT slide   Figure 1. Genetic and molecular map of the dpp locus. (A) Schematic of the D. melanogaster dpp locus with molecular coordinates from the 22F1-2 chromosome walk. The shaded rectangles above the coordinate line delineate the approximate locations of three genetically defined regions of the locus. The locations of exons and their splicing patterns are shown below the coordinate line; filled rectangles represent protein-coding sequences, and open rectangles indicate untranslated regions. The position and direction of transcription for two tRNATyr genes within the dpp locus are indicated by arrows. The coordinates of the portion of the disk region that, when deleted, lead to the heldout wing posture phenotype are indicated (ST. JOHNSTON et al. 1990 ). The hobo transgene H{Lw2}dpp151h inserted within this 2-kb segment is also shown. (B) Schematic of simple molecular events that may result from the mobilization of H{Lw2}dpp151h from the dpp disk region. Five possible scenarios are shown, and the experimental tests used to identify unwanted events (scenarios i–iv) are listed.

Alterations in the Hin region are impossible to recover in a dpp diploid genome as a result of dominant lethality caused by dpp haplo-insufficiency. Therefore, we recombined a chromosomal duplication of the dpp locus [Dp(2;2)DTD48 dpp^d-ho] onto chromosome arm 2R of this strain. A crucial part of our strategy required the duplicated copy of dpp be deleted for the heldout region.

H{Lw2}dpp^151h was mobilized using a stable source of hobo transposase under the control of the heat-shock promoter inserted on the CyO balancer chromosome (Figure 2; parental cross). Mass matings and multiple heat shocks were used to generate candidate individuals of the appropriate genotype. Under these conditions, somatic mobilization of the hobo construct was seen in nearly 100% of the adults containing the transgene and transposase (Figure 2; F₁ generation). This was determined by mosaicism for white expression (w⁺) in the eye from the mini-white-marked hobo construct. This rate of somatic mobilization is similar to that seen by CALVI and GELBART 1994 when examining individuals that contain two hobo transgenes undergoing a single heat-shock induction of the transposase.

View larger version (42K): In this window In a new window Download PPT slide   Figure 2. Genetic scheme used in the identification of local transposition of H{Lw2}dpp151h. The mobilization of H{Lw2}dpp151h takes place in the F1 male germline. For a description of genetic symbols, see FLYBASE 1998 . Individuals derived from test crosses were not maintained. Details are described in the text.

Mobilization of H{Lw2}dpp^151h may result in a variety of genomic events at the dpp locus. Five examples of possible outcomes are shown in Figure 1B. Those listed are simple events, such as deletions or deletion/transposition combinations. The existence of a duplication of the dpp locus on chromosome arm 2R allows the recovery of unwanted Hin region deletions as well as Hin region hobo insertions. Four of the events (Figure 1B, i–iv) are unwanted outcomes, and the experimental tests used to identify these events are listed. Complex chromosomal events such as deletion/inversion combinations are also possible, but these will fail the transvection test.

Genetic characterization of [H{Lw2}dpp^151h]^jump candidate strains:
In a pilot study, candidate males containing the transposition-capable genotype (Figure 2; F₁ generation) were individually mated to groups of dpp^hr27-bearing tester females. The dpp^hr27 allele contains a nonconservative amino acid substitution in the proregion (WHARTON et al. 1996 ) and acts as a strong hypomorph (WHARTON et al. 1993 ). In the progeny of this cross, transpositions that affect the Hin region of H{Lw2}dpp^151h will generate phenotypically heldout adults. This is because the two remaining copies of dpp (sufficient to overcome haplo-insufficiency) are dpp^hr27, which encodes a defective protein, and the copy on 2R, which is a heldout allele (dpp^d-ho). The dpp^hr27 and the dpp^d-ho alleles are unable to complement each other and generate a wild-type wing phenotype through transvection because of their location on opposite chromosome arms. See below for a discussion of transvection at the dpp locus. However, a heldout phenotype is also generated in the progeny of this cross by two unwanted classes of transposition events. These are events that affect the heldout region only or both the heldout and Hin regions of H{Lw2}dpp^151h. Transposition resulting in Hin-affected alleles will be distinguished from both heldout-affected and heldout plus Hin-affected alleles in subsequent genetic tests.

The dpp^hr27 tester females were also homozygous for z¹. We included z¹ because it acts as an enhancer of the heldout phenotype of dpp^d-ho/dpp^hr4 individuals (GELBART and WU 1982 ). We hoped it would also enhance the heldout phenotype of local jump heterozygotes in which the Hin region of H{Lw2}dpp^151h was affected. We focused our attention on male progeny that were hemizygous for z¹ and displayed heldout wings and w⁺ eyes (Figure 2; F₂ generation). The presence of w⁺ eyes indicates that the hobo transgene remains in the F₂ genotype after segregation of the homologs in the F₁ male germline, suggesting that a local jump is a possibility.

From ~500 fertile F₁ males, we recovered a total of 41 w⁺ heldout males representing 30 different germline events. In several cases, multiple w⁺ heldout males arose among the F₂ progeny of a single F₁ cross. For these clustered males, every individual was carried separately through the remainder of the scheme. In some clusters, distinctions between individuals were seen in the haplo-insufficiency test. This indicates that under extremely efficient conditions of transposase induction, multiple independent events can occur in a single male germline. We also recovered w⁺ heldout females (heterozygous for z¹) at a slightly lower rate. However, because of their location on distinct chromosome arms, 50% of all recombination events in these unbalanced females will separate [H{Lw2}dpp^151h]^jump and Dp(2;2)DTD48 dpp^d-ho in the next generation. This prevents the efficient recovery of Hin-affected alleles of [H{Lw2}dpp^151h]^jump in the F₃ generation. We chose not to characterize these females further.

Each of the F₂ w⁺ heldout males was individually mated to groups of dpp^H61-bearing tester females (Figure 2; F₂ cross A). The dpp^H61 allele is a 2-kb deletion that removes nearly all of the third exon, including the ligand domain (ST. JOHNSTON et al. 1990 ). Lines that carry this allele are haplo-insufficient for dpp. The line is maintained over a dpp embryonic rescue construct containing Hin region sequences inserted in the CyO balancer (CyO23; WHARTON et al. 1993 ). In the test cross, the dpp rescue construct will segregate away from dpp^H61, allowing us to determine if dpp haplo-insufficiency has been restored through involvement of the Hin region in alleles of [H{Lw2}dpp^151h]^jump. Flies of the genotype [H{Lw2}dpp^151h]^jump Dp(2;2)DTD48 dpp^d-ho/dpp^H61 will have only one functional copy of dpp (the one on 2R) if the Hin region of H{Lw2}dpp^151h has been affected by the hobo mobilization. These flies will not survive. As a result, the progeny class they would have contributed to (Cy⁺) will be reduced. In addition, dpp^H61/dpp^hr27 flies (also Cy⁺) will die because of dpp haplo-insufficiency. Thus, the creation of a dpp haplo-insufficient allele at the [H{Lw2}dpp^151h]^jump locus in F₂ w⁺ heldout males is easily detected by the absence of any Cy⁺ progeny. Restoration of dpp haplo-insufficiency eliminates the possibility that the hobo mobilization affected only the heldout region of the dpp locus in [H{Lw2}dpp^151h]^jump strains (Figure 1B, ii).

Each of the F₂ w⁺ heldout males was subsequently individually mated to groups of double-balancer females homozygous for z¹ (Figure 2; F₂ cross B) to generate F₃ individuals suitable for creating balanced stocks of each allele of [H{Lw2}dpp^151h]^jump. We were unable to use balanced progeny from the dpp^H61 mating (Figure 2; F₂ cross A) because dpp^H61 is an H strain. The presence of endogenous hobo elements would result in the remobilization of [H{Lw2}dpp^151h]^jump at some future point in the stock. Sibmating of balanced F₃ individuals of the genotype [H{Lw2}dpp^151h]^jump Dp(2;2)DTD48 dpp^d-ho/CyO was performed only for lines derived from F₂ w⁺ heldout males that demonstrated restoration of dpp haplo-insufficiency (Figure 2; F₂ test). As a second criterion for choosing F₃ progeny for stock construction, only F₂ cross B (Figure 2) progeny containing exclusively w⁺ males and females of the appropriate genotype were chosen. The presence of w⁺ eyes in every individual of the desired genotype indicates that the hobo transgene remains in the F₃ genotype after two generations of segregation from the original mobilization, again suggesting that a local jump remains a possibility. These F₃ individuals and stocks derived from them are again homozygous for z¹. We continued to maintain a z¹ background, hoping that the heldout phenotype would be enhanced in suitable genotypes during future experiments.

Nearly 50% of the F₂ w⁺ heldout males were eliminated by these two criteria. Balanced F₃ stocks were created representing 24 F₂ w⁺ heldout males from 11 different F₁ clusters. However, there was wide variation in the extent of dpp haplo-insufficiency in the F₃ lines. Six of the F₃ lines were essentially haplo-insufficient. These lines gave a small number of escapers (Cy⁺) in the progeny of the cross to dpp^H61. In these lines, escapers appear at roughly the same rate as in the progeny of a cross between dpp^H61 and wild type (Canton-S). Fourteen F₃ lines gave a moderate number of escapers compared to dpp^H61 flies crossed to wild type, yet Cy⁺ flies among the progeny were far fewer than expected by Mendelian ratios (33%). In these lines, we observed 10–30% of the expected number of Cy⁺ progeny. Four F₃ lines gave a large number of escapers. In these lines, we observed 35–65% of the expected number of Cy⁺ progeny.

We chose to pursue the F₃ lines with moderate and high escaper rates for the following reasons. First, the only reported P-element allele of dpp (dpp¹⁰⁶³⁸; TWOMBLY et al. 1996 ) is an insertion at the boundary between the shv and Hin regions (coordinate 83 on the dpp chromosome walk; ST. JOHNSTON et al. 1990 ). The insertion results in a recessive, embryonic-lethal dpp allele, not a haplo-insufficient allele. Second, the dpp^e87 allele is a 0.5-kb deletion that spans the boundary of the shv and Hin regions. The dpp^e87 allele is homozygous viable and fertile, but it fails to complement other recessive, embryonic-lethal dpp alleles (ST. JOHNSTON et al. 1990 ). We did not want to eliminate the possibility of recovering hobo insertions in this portion of the Hin region.

To further test the cosegregation of the w⁺ eye phenotype with the [H{Lw2}dpp^151h]^jump chromosome, the F₃ stocks were examined for the presence of homozygous individuals (Cy⁺). In each stock, these were uniformly heldout and w⁺. The cosegregation of w⁺ eyes and the otherwise unmarked [H{Lw2}dpp^151h]^jump chromosome through three generations implies an 87.5% probability (for each stock) that the hobo transgene resides on this chromosome. Proof that the hobo transgene is inserted in the dpp locus of the [H{Lw2}dpp^151h]^jump chromosome in the F₃ lines requires molecular data.

The final genetic test of the experiment exploits the allelic interaction known as transvection, which is defined as synapsis-dependent, intragenic complementation (LEWIS 1954 ). Transvection at the dpp locus has been studied by GELBART 1982 . The best-characterized dpp genotypes that display transvection are heterozygous for dpp^d-ho and any one of a number of recessive, lethal mutations in the Hin region (e.g., dpp^hr4). Genotypes that contain this combination of mutant alleles are phenotypically wild type in the absence of any chromosomal rearrangement that interferes with the proper pairing of homologs. According to one model (PIRROTA 1990 ), dpp^d-ho complements dpp^hr4 because the wild-type copy of the heldout enhancer on the dpp^hr4 chromosome acts in trans to promote transcription of the wild-type DPP protein encoded by the dpp^d-ho allele on the homolog. We used transvection to identify H{Lw2}dpp^151h mobilization events that affected both the original site of insertion in the heldout region and the Hin region (Figure 1B, Figure I and iv), as well as mobilizations that involve complex chromosomal rearrangements.

Several males from each F₃ stock were crossed to groups of females carrying dpp^d-ho. Trans-heterozygous (Cy⁺) progeny were examined for wing posture (Figure 2; F₃ test). These flies carry three copies of dpp—two copies of dpp^d-ho and [H{Lw2}dpp^151h]^jump. The only possible location for a normal heldout enhancer is at [H{Lw2}dpp^151h]^jump. If the heldout enhancer is intact and if there have been no chromosomal rearrangements, the heldout enhancer should promote the transcription in trans of the wild-type DPP protein encoded by the dpp^d-ho allele on the homolog. This results in wild-type wing posture. However, if the heldout enhancer is affected at the [H{Lw2}dpp^151h]^jump locus, or if there has been a chromosomal rearrangement, then the Cy⁺ progeny will have a heldout phenotype. Two of the F₃ lines from a single F₁ cluster failed to show normal wing posture in this test.

Molecular characterization of [H{Lw2}dpp^151h]^jump candidate strains:
Having identified and eliminated a number of lines that carry several distinct classes of unwanted chromosomal events (Figure 1B, Figure I, ii, and iv), we conducted a molecular characterization of the remaining candidate lines. In the following experiments, candidate lines are organized into groups according to the number of escapers in the haplo-insufficiency test (see the legend to Figure 3 for details). Our first experiment was designed to provide molecular evidence of hobo mobilization in each candidate line. A Southern blot containing genomic DNA digested with SspI from balanced lines containing the original hobo transgene H{Lw2}dpp^151h, the recombinant parental line for our experiment H{Lw2}dpp^151h Dp(2;2)DTD48 dpp^d-ho, and all candidate [H{Lw2}dpp^151h]^jump lines was hybridized with pSV-ß-gal. This is a probe for the ß-galactosidase sequences located in the H{Lw2} transgene. Since SspI has two restriction sites in the ß-galactosidase gene, the probe will hybridize to two restriction fragments. One is an internal fragment of 2 kb that is unaffected by mobilization of the transgene. The other fragment contains the 3' end of H{Lw2} and the genomic sequences flanking the transgene. If the hobo transgene in the [H{Lw2}dpp^151h]^jump chromosome no longer resides in the same location as in the original H{Lw2}dpp^151h or the parental H{Lw2}dpp^151h Dp(2;2)DTD48 dpp^d-ho chromosome, the chimeric restriction fragment will change in size in the [H{Lw2}dpp^151h]^jump strains. Altered fragment size in a [H{Lw2}dpp^151h]^jump line reflects the incorporation of new flanking sequences.

View larger version (54K): In this window In a new window Download PPT slide   Figure 3. Molecular evidence for the mobilization of H{Lw2}dpp151h. Autoradiograph of a Southern blot containing SspI-digested genomic DNA from F3 candidate [H{Lw2}dpp151h]jump balanced lines probed with pSV-ß-gal. Restriction fragment size markers are indicated in kilobases. Lane O, original balanced line containing H{Lw2}dpp151h; lane P, recombinant parental balanced line containing H{Lw2}dpp151h Dp(2;2)DTD48 dppd-ho; lanes 1–24, F3 balanced lines containing [H{Lw2}dpp151h]jump Dp(2;2)DTD48 dppd-ho. F3 candidate lane groupings are as follows: 1–4, lines that showed large numbers of escapers in the haplo-insufficiency test; 5 and 6, lines that failed the transvection test; 7–18, lines that had a moderate number of haplo-insufficiency escapers; 19–24, haplo-insufficient lines with a small number of escapers. Individual lanes represent the following strains: 1, H{Lw2}dpp14; 2, H{Lw2}dppF18; 3, H{Lw2}dppF22; 4, H{Lw2}dppT2; 5, H{Lw2}dpp15D; 6, H{Lw2}dpp15B; 7, H{Lw2}dppF3; 8, H{Lw2}dppF4; 9, H{Lw2}dppF5; 10, H{Lw2}dppF6; 11, H{Lw2}dppF7; 12, H{Lw2}dppF11; 13, H{Lw2}dppF13; 14, H{Lw2}dppF14; 15, H{Lw2}dppF16; 16, H{Lw2}dppF17; 17, H{Lw2}dppF19; 18, H{Lw2}dppF21; 19, H{Lw2}dppF2; 20, H{Lw2}dppF8; 21, H{Lw2}dppF9; 22, H{Lw2}dppF10; 23, H{Lw2}dppF12; 24, H{Lw2}dppF20.

As shown in Figure 3, nearly all candidate lines show chimeric fragments of a different size than in the original and parental lines. Only lane 18 appears similar to the parental lane. This could result from a new hobo insertion that is the same distance from a genomic SspI restriction site as the original insertion was from its flanking SspI site. Thus, the data suggest that we successfully mobilized the hobo transgene in all candidate lines. Intriguingly, several lanes in Figure 3 (lanes 8, 9, 12, and 24) display two chimeric fragments. These lanes also show an increase in the intensity of the signal from the internal fragment in comparison to the original and parental lines. This suggests that there are two hobo transgenes in each of these lines, and that they are inserted at different genomic locations. In two of the lanes in Figure 3 (lanes 12 and 24), one of the chimeric fragments appears similar in size to the parental fragment, suggesting that one copy of the transgene remains in the original location. We did not pursue this finding any further.

We focused our efforts on identifying new hobo insertions in the dpp Hin region. A Southern blot of genomic DNA digested with EcoRI from balanced lines containing the original hobo transgene, the recombinant parental line, and all candidate lines was probed with the dpp cDNA H1. In EcoRI-digested genomic DNA, the H1 probe will hybridize to an 8-kb fragment that contains the Hin region with its two protein-coding exons and to a 15-kb fragment that contains the distant 5' noncoding exon (see Figure 1). EcoRI has three restriction sites within the hobo transgene and therefore a change in size of the 8-kb Hin region restriction fragment indicates a hobo insertion in the Hin region. The altered size of any Hin region fragment indicates that it is now defined by genomic and hobo transgene EcoRI sites.

As shown in Figure 4A, new restriction fragments are detected in lanes 2 and 12 (lines H{Lw2}dpp^F18 and H{Lw2}dpp^F11). In lane 2, a single new fragment >8 kb is detected. In Figure 4 (lane 12), two fragments <8 kb are detected. This result suggests that new hobo insertions now flank the two protein-coding exons in H{Lw2}dpp^F18 and separate the two protein-coding exons in H{Lw2}dpp^F11. In all lanes, the H1-hybridizing restriction fragments derived from Dp(2;2)DTD48 dpp^d-ho and from the CyO balancer chromosome (8 and 15 kb) are unaffected. The H{Lw2}dpp^F11 and H{Lw2}dpp^F18 lines did not derive from the same F₂ w⁺ heldout male germline, indicating that the new insertions are independent events.

View larger version (89K): In this window In a new window Download PPT slide   Figure 4. Molecular identification of new hobo insertions in the dpp Hin region. Autoradiographs of a single Southern blot containing EcoRI-digested genomic DNA from F3 [H{Lw2}dpp151h]jump balanced lines probed with (A) the dpp cDNA H1 or (B) pH{Lw2}. (A) New restriction fragments are detected in lanes 2 and 12 (H{Lw2}dppF18 and H{Lw2}dppF11). The wild-type restriction fragments derived from Dp(2;2)DTD48 dppd-ho and the CyO balancer chromosome remain unchanged. (B) Hybridizing restriction fragments highlighted by arrows in lanes 2 and 12 are the same fragments that show altered size in A. Note that A is an overnight exposure, and B is a 4-day exposure that leads to the difference in the sharpness of the autoradiographic signal from fragments that hybridize to both probes. The background of weakly hybridizing fragments seen in B is common on genomic Southern blots probed with hobo sequences (BLACKMAN et al. 1987 ). Lane designations follow those in Figure 3.

We wanted to confirm that hobo transgene sequences are contained within the new restriction fragment detected with H1. This is important because the Hin region restriction fragment can also change in size if the mobilized hobo transgene deleted chromosomal material adjacent to its original site of insertion. If a hobo-induced deletion removes some but not all of the Hin region, the H1 probe will detect an altered Hin region fragment. However, if the altered fragment is the result of a mobilization-induced deletion, hobo sequences may not be associated with the new Hin region fragment. We took the Southern blot shown in Figure 4A, removed the H1 probe, and reprobed it with pH{Lw2}, as shown in Figure 4B. We chose to use pH{Lw2} as a probe instead of pSV-ß-gal so that we could detect chimeric restriction fragments containing the 5' or the 3' ends of the hobo transgene. This allows us to detect hybridization of hobo sequences to fragments that previously hybridized with H1 regardless of the orientation of the new transgene insertion.

Since EcoRI has three restriction sites within the hobo transgene, we predict four hybridizing fragments in each lane in this experiment. Two of the fragments are internal, and two are chimeric fragments that contain hobo sequences and either 5' or 3' flanking genomic DNA. The internal fragments are nearly equal in size. They appear in Figure 4B as a strongly hybridizing 4-kb doublet in all lanes. The chimeric fragments that contain flanking genomic DNA are of unpredictable size. The large (>20 kb) fragment or fragments hybridizing in all lanes, as well as other strongly hybridizing fragments, e.g., Figure 4B, lanes 2, 8, 9, 12, and 24, likely represent these chimeric fragments. By overlaying the autoradiographs, it is clear that the new fragment in lane 2 and one of the new fragments in lane 12 from Figure 4A also hybridize to pH{Lw2}. These fragments are highlighted by arrows in Figure 4B. Taken together, the data in Figure 4 strongly support the existence of a new hobo transgene insertion in the Hin region in lines H{Lw2}dpp^F18 and H{Lw2}dpp^F11.

To further specify the location of the new hobo insertions in the H{Lw2}dpp^F18 and H{Lw2}dpp^F11 strains, we conducted additional molecular analyses on these lines. We wanted to determine if the insertions were in a 5-kb region roughly bounded by the two dpp protein-coding exons. A Southern blot of genomic DNA digested with NheI and ScaI from balanced lines containing the original transgene, the recombinant parental line, and the H{Lw2}dpp^F18 and H{Lw2}dpp^F11 lines was probed with H1. These enzymes each have a single restriction site within the Hin region. NheI does not cut in the hobo transgene. ScaI cuts very close to one end of the hobo transgene.

As shown in Figure 5A (left), new restriction fragments are detected in lanes 1 and 2 (H{Lw2}dpp^F18 and H{Lw2}dpp^F11, respectively). In lane 1, a single new fragment <5 kb is detected. In lane 2, two new fragments, one <5 kb and one >5 kb, are detected. This result is consistent with our EcoRI experiment, confirming that a new hobo insertion flanks the protein-coding exons in the H{Lw2}dpp^F18 strain and separates the two protein-coding exons in the H{Lw2}dpp^F11 line. In all lanes, the H1-hybridizing fragments derived from Dp(2;2)DTD48 dpp^d-ho and from the CyO balancer chromosome (5 and 9 kb) are unaffected.

View larger version (42K): In this window In a new window Download PPT slide   Figure 5. Molecular specification of the location of new hobo insertions in the dpp Hin region. Autoradiographs of a single Southern blot containing restriction enzyme-digested genomic DNA from two F3 [H{Lw2}dpp151h]jump balanced lines containing new hobo insertions in the Hin region (H{Lw2}dppF18 and H{Lw2}dppF11). The blot was probed with (A) the dpp cDNA H1 or (B) pH{Lw2}. In the four lanes on the left, the genomic DNA was cut with NheI and ScaI. In the four lanes on the right, the genomic DNA was cut with XbaI. (A) New restriction fragments are detected on the left in lanes 1 and 2 (NheI and ScaI). No new fragments are seen on the right (XbaI). The wild-type restriction fragments derived from Dp(2;2)DTD48 dppd-ho and the CyO balancer chromosome remain unchanged. (B) Hybridizing restriction fragments highlighted by arrows in lanes 1 and 2 (NheI and ScaI) are the same fragments showing altered size in A. Lane O, original balanced line containing H{Lw2}dpp151h; lane P, recombinant parental balanced line containing H{Lw2}dpp151h Dp(2;2)DTD48 dppd-ho; lane 1, H{Lw2}dppF18; lane 2, H{Lw2}dppF11.

Our final diagnostic genomic digest was designed to determine if the insertions landed in the dpp open reading frame contained in exons 2 and 3. A Southern blot of genomic DNA digested with XbaI from balanced lines containing the original transgene, the recombinant parental line, and the H{Lw2}dpp^F18 and H{Lw2}dpp^F11 lines was probed with H1. There are four XbaI sites in the Hin region. The first pair of sites nearly brackets exon 2 and delineates a 1-kb fragment. The second pair of sites brackets the coding region of exon 3 and defines a 2-kb fragment. XbaI does not cut in the hobo transgene. As shown in Figure 5A (right), no new fragments are detected.

We then took the Southern blot shown in Figure 5A, removed the H1 probe, and rehybridized it with pH{Lw2}. As described above, we were looking to confirm that hobo transgene sequences are coincident with the altered restriction fragments seen in Figure 5A. Since NheI does not cut in the transgene and ScaI cuts once in the transgene (1300 bp from the 5' end), we expect two hybridizing chimeric fragments of unpredictable size. Both fragments will contain hobo sequences and flanking genomic DNA. Note the hybridizing restriction fragments highlighted by arrows in lanes 1 and 2 (Figure 5B). By overlaying the autoradiographs, it is clear that the new fragment in lane 1 and both new fragments in lane 2 from Figure 5A (left) also hybridize to pH{Lw2}. From the relative intensity of hybridization of the fragments in lane 1, it appears that the small fragment that also hybridizes to H1 contains just the 5' end of the pH{Lw2} transgene. When hybridizing pH{Lw2} to the XbaI digest, we found no fragments hybridizing in common with those in Figure 5A (data not shown). Taken together, the data in Figure 5 strongly support the presence of new hobo insertions in both lines in the Hin region between NheI and ScaI, but not in the protein-coding regions of the two exons defined by XbaI.

A composite restriction map of the Hin region from lines H{Lw2}dpp^F18 and H{Lw2}dpp^F11 that summarizes our molecular data is shown in Figure 6. We present the most likely location for the new hobo insertions in these lines based upon the relative position of the restriction enzymes and the hybridization patterns of our Southern blots. The detection of only one new fragment in the H{Lw2}dpp^F18 strain suggests that the most likely location for the new hobo insertion in this line is in the 3'-untranslated region of dpp exon 3. The detection of two new fragments in the H{Lw2}dpp^F11 strain implies that the most likely location for the new hobo insertion in this line is the intron between the protein-coding exons.

View larger version (12K): In this window In a new window Download PPT slide   Figure 6. Molecular map of new hobo insertions in the dpp Hin region. Schematic of the endogenous dpp Hin region from the hobo-bearing chromosome in lines H{Lw2}dppF11 and H{Lw2}dppF18. The molecular coordinates and depiction of dpp exons follow that in Figure 1. The positions of restriction enzymes used for the molecular characterization of new hobo insertions in these strains are shown above the coordinate line. E, EcoRI; N, NheI; S, ScaI; X, XbaI. The NheI site is 145 bp from the splice acceptor site of exon 2. The distance between NheI and the closest XbaI site is 50 bp. The distance between the second and third XbaI sites (these sites define a restriction fragment not recognized by the dpp cDNA H1) is 1099 bp. The distance between ScaI and the closest XbaI site is 552 bp. The most likely location for the new hobo insertions in each of these lines is shown. For H{Lw2}dppF18, which flanks the protein-coding exons, this is the 552-bp region between the 3' XbaI and the ScaI sites rather than the 50-bp region between the NheI and the 5' XbaI sites.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The development of endogenous, transposable elements into powerful genetic tools has had an enormous impact on our understanding of organismal biology. This study further characterizes the hobo element system in D. melanogaster and compares its utility to the well-established P-element system. Relying upon the exquisite molecular genetics of the dpp locus, we have demonstrated that hobo, like P, is capable of local transposition. We recovered two chromosomes that experienced a precise excision and new insertion of a hobo transgene roughly 25 kb apart. The independent hobo local transposition events resulted in new insertions in the dpp transcription unit. One insertion is between the two protein-coding exons (line H{Lw2}dpp^F11), and the other is in the 3' untranslated region of exon 3 (line H{Lw2}dpp^F18). A summary of the experiment describing the results of each genetic and molecular test is shown in Table 1.

Interestingly, line H{Lw2}dpp^F11 had a moderate number of escapers in the dpp haplo-insufficiency test, and line H{Lw2}dpp^F18 had a large number of escapers. In these lines, it seems unlikely that the new transgene insertions have a major affect on the DPP protein. In line H{Lw2}dpp^F11, the insertion could be removed from the dpp transcript by RNA splicing. In line H{Lw2}dpp^F18, the insertion could be removed by proper translation termination. The effect of the new insertions on dpp function may be caused by problems associated with the dpp transcript. The large transgene (13 kb) may interfere with splicing or translation termination to a limited extent in these lines.

None of the haplo-insufficient lines (those with a small number of escapers) contained a hobo insertion in the Hin region. One possible interpretation of our molecular data is that the haplo-insufficient lines resulted from large, hobo mobilization-induced deletions that removed the entire disk and Hin regions (Figure 1B, iii). The lines that showed escapers from dpp haplo-insufficiency that did not have hobo insertions in the Hin region may have smaller hobo mobilization-induced deletions. These deletions may remove varying amounts of chromosomal material between the heldout region and the Hin region. For example, a number of chromosomal inversions have breakpoints near the tRNA^Tyr genes that remove all disk region sequences. These inversions are lethal as trans-heterozygotes (ST. JOHNSTON et al. 1990 ). Perhaps in the background of dpp^H61 and dpp^d-ho, as in our haplo-insufficiency test, a percentage of dominant lethality is achieved.

A more intriguing possibility is that the various levels of dpp haplo-insufficiency in candidate lines without Hin region insertions are the result of new insertions in second-site enhancers of dpp. In the haplo-insufficiency test, the [H{Lw2}dpp^151h]^jump chromosome was paternal in origin. Thus, our experiment could be viewed as a mutational screen for zygotic enhancers of dpp. As described above, there is an 87.5% probability (for each stock) that the hobo transgene still resides on the second chromosome. Complementation tests of candidate lines without Hin region insertions using alleles from known dpp enhancers on chromosome 2, such as Mothers against dpp (NEWFELD et al. 1996 ), saxophone, and thickveins (BRUMMEL et al. 1994 ), would be the first step in evaluating this exciting possibility.

No P-element insertions have been reported in the portions of the Hin region where the new hobo insertions are suggesting that these elements have different insertional preferences. To critically test this proposal, we conducted an analogous excision/new insertion experiment using a P-element transgene (PZ; MLODZIK and HIROMI 1992 ). In the E32 strain, the PZ transgene is inserted in the 5' untranslated region of the out-at-first gene, ~6 kb further from the dpp Hin region than H{Lw2}dpp^151h (BERGSTROM et al. 1995 ; MERLI et al. 1996 ). The P-element version of our mobilization/new insertion scheme contained several minor technical differences from the hobo experiment. In the F₁ cross, the tester allele was dpp^hr4 carried on a chromosome marked with Sp and the DTD48 duplication of dpp on 2R. The genetic scheme was started by jumping the P element in females, which was shown by ZHANG and SPRADLING 1993 to be more efficient than jumping in males. In the P-element scheme, z¹ was not introduced.

We conducted a screen of 1200 fertile, transposition-capable females. Each female carried the E32 chromosome and a copy of the engineered P-element transposase 2-3 (reviewed in RIO 1990 ) inserted on the CyO balancer chromosome via a hobo transgene (CyOHOP1; B. CALVI and W. GELBART, unpublished data). This screen generated 182 F₂ heldout progeny of both sexes (155 clusters). In this scheme, the continued presence of the transgene could not be followed by eye color, as in our hobo experiment. All F₂ heldout flies were tested for dpp haplo-insufficiency by crossing to flies carrying In(2LR)Gla over CyO23. If haplo-insufficiency for dpp is now present, no offspring of the genotype E32^jump over In(2LR)Gla will survive, and the E32^jump chromosome can be recovered over CyO23. Progeny carrying the E32^jump chromosome are distinguishable from those inheriting the dpp^hr4 chromosome using Sp. In this much larger screen, none of the 182 F₂ heldout flies demonstrated any level of dpp haplo-insufficiency.

Our interpretation of this result is that all the F₂ heldout flies in the P-element scheme had affected heldout enhancers only, reflecting mobilization-induced deletions. This suggests that the dpp Hin region is refractory to P insertion. A limited exception would be at the shv/Hin boundary, where dpp¹⁰⁶³⁸ is inserted. The results of our P and hobo experiments for insertion preference at the dpp locus support the findings of SMITH et al. 1993 , who detected distinct insertion preferences for P and hobo elements in a genome-wide survey. The expanded use of hobo transgenes will facilitate our understanding of aspects of the D. melanogaster genome that are not accessible with P elements.

The new hobo insertions adjacent to the dpp protein-coding exons may be suitable substrates for exploring another aspect of the P-element system that has not yet been demonstrated for hobo. These Hin region insertions are excellent candidates for experiments in gene replacement via transposable element-induced gap repair (reviewed in GLOOR and LANKENAU 1998 ). In this technique, the gap in the chromosome created by P-element excision is repaired from the DNA sequence of a homologous (but nonidentical) template at an ectopic chromosomal site. This process results in the replacement of sequences at the original site of P-element insertion and the loss of the P element. Using this method, tracts of ~1.4 kb of new sequence have been introduced onto otherwise wild-type chromosomes.

The first step in exploring the feasibility of this method for hobo is to determine the exact site of the H{Lw2}dpp^F18 and H{Lw2}dpp^F11 insertions in the dpp Hin region. This is done using restriction sites for plasmid rescue contained in the H{Lw2} construct (SMITH et al. 1993 ). The exact location of the insertion is obtained by sequencing the genomic region of the rescued plasmid. Once the site of insertion is known, constructs that contain the desired gap repair template can be injected, and the gene replacement method can be tested in lines that contain both transgenes. One template that would provide valuable new information about dpp function, if successfully transferred, would be an epitope-tagged ligand for studying DPP protein activity. For improving our understanding of dpp molecular evolution, the possibility of transferring dpp homologs from other species (e.g., grasshopper; NEWFELD and GELBART 1995 ) is very exciting. Larger questions about the developmental role and molecular evolution of the Drosophila TGF-ß family could be addressed by replacing dpp with 60A (WHARTON et al. 1991 ) or screw (ARORA et al. 1994 ).

In summary, our studies have revealed similarities and differences between the hobo and P genetic systems. Both hobo and P are amenable to local jumping strategies, but the elements vary in their insertion site preference. hobo transgenes will insert into genomic regions refractory to P elements. These results suggest that the continued exploitation of both genetic systems is the best approach to understanding the genome of D. melanogaster. In addition, strains generated in our experiments are valuable new reagents for exploring the hobo genetic system and for understanding dpp function and molecular evolution.

	ACKNOWLEDGMENTS

Jeff Sekelsky, Brian Calvi, Des Smith, and Vern Twombly provided valuable advice in the design of the genetic schemes. Wayne Rindone, Joe Chillemi, Susan Russo, and Tessa Cigler provided much needed assistance during the fly pushing stages of the project. Brian Calvi provided the CyOHOP1 strain. Dan Eberl provided a compiled DNA sequence of the H{Lw2} transgene. The genetic screen was conducted in Bill Gelbart's laboratory, and we thank him for valuable comments on the manuscript. This work was supported by a Faculty Grant in Aid from Arizona State University to S.J.N.

Manuscript received June 5, 1998; Accepted for publication September 18, 1998.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ARORA, K., M. LEVINE, and M. O'CONNOR, 1994  The screw gene encodes a ubiquitously expressed member of the TGF-ß family required for specification of dorsal cell fates in the Drosophila embryo. Genes Dev. 21:2588-2601.

BERGSTROM, D. E., C. MERLI, J. CYGAN, R. SHELBY, and R. K. BLACKMAN, 1995  Regulatory autonomy and molecular characterization of the Drosophila out at first gene. Genetics 139:1331-1346[Abstract].

BLACKMAN, R., R. GRIMAILA, M. KOEHLER, and W. M. GELBART, 1987  Mobilization of hobo elements residing within the decapentaplegic gene complex: suggestion of a new hybrid dysgenesis system in Drosophila melanogaster.. Cell 49:497-505[Medline].

BLACKMAN, R., M. SANICOLA, L. A. RAFTERY, T. GILLEVET, and W. M. GELBART, 1991  An extensive 3' cis-regulatory region directs the imaginal disk expression of decapentaplegic, a member of the TGF-ß family in Drosophila. Development 111:657-666[Abstract].

BRUMMEL, T. J., V. TWOMBLY, G. MARQUES, J. WRANA, and S. J. NEWFELD et al., 1994  Characterization and relationship of DPP receptors encoded by the saxophone and thickveins genes in Drosophila. Cell 78:251-262[Medline].

CALVI, B. R. and W. M. GELBART, 1994  The basis for germline specificity of the hobo transposable element in Drosophila melanogaster.. EMBO J. 13:1636-1644[Medline].

FLYBASE,, 1998  FlyBase—a Drosophila database. Nucleic Acids Res. 26:85-88. (http://www.flybase.bio.indiana.edu/).[Abstract/Free Full Text].

GELBART, W. M., 1982  Synapsis-dependent allelic complementation at the decapentaplegic gene complex in Drosophila melanogaster.. Proc. Natl. Acad. Sci. USA 79:2636-2640[Abstract/Free Full Text].

GELBART, W. M. and C.-T. WU, 1982  Interactions of zeste mutations with loci exhibiting transvection effects in Drosophila melanogaster.. Genetics 102:179-189[Abstract/Free Full Text].

GLOOR, G. and D. LANKENAU, 1998  Gene conversion in mitotically dividing cells: a view from Drosophila. Trends Genet. 14:43-46[Medline].

HARTL, D. L., and E. R. LOZOVSKAYA, 1995 The Drosophila Genome Map: A Practical Guide. R. G. Landes, Austin, TX.

HOFFMANN, F. M. and W. GOODMAN, 1987  Identification in transgenic animals of the Drosophila decapentaplegic sequences required for embryonic dorsal pattern formation. Genes Dev. 1:615-625[Abstract/Free Full Text].

HURSH, D., R. PADGETT, and W. GELBART, 1993  Cross regulation of decapentaplegic and Ultrabithorax transcription in the embryonic visceral mesoderm of Drosophila. Development 117:1211-1222[Abstract].

IMMERGLÜCK, K., P. LAWRENCE, and M. BIENZ, 1990  Induction across germ layers in Drosophila mediated by a genetic cascade. Cell 62:261-268[Medline].

IRISH, V. F. and W. M. GELBART, 1987  The decapentaplegic gene is required for dorsal-ventral patterning of the Drosophila embryo. Genes Dev. 1:868-879[Abstract/Free Full Text].

LEWIS, E. B., 1954  The theory and application of a new method of detecting chromosomal rearrangements in Drosophila melanogaster.. Am. Nat. 88:225-239.

MASSAGUÉ, J., 1990  The TGF-ß family. Annu. Rev. Cell Biol. 6:597-641.

MERLI, C., D. E. BERGSTROM, J. CYGAN, and R. K. BLACKMAN, 1996  Promoter specificity mediates the independent regulation of neighboring genes. Genes Dev. 10:1260-1270[Abstract/Free Full Text].

MLODZIK, M. and Y. HIROMI, 1992  Enhancer trap method in Drosophila: its application to neurobiology. Methods Neurosci. 9:397-414.

NEWFELD, S. J. and W. M. GELBART, 1995  Identification of two Drosophila TGF-ß family members in the grasshopper Schistocerca americana.. J. Mol. Evol. 41:155-160[Medline].

NEWFELD, S. J., E. H. CHARTOFF, J. M. GRAFF, D. A. MELTON, and W. M. GELBART, 1996  Mothers against dpp encodes a conserved cytoplasmic protein required in DPP/TGF-ß responsive cells. Development 122:2099-2108[Abstract].

NEWFELD, S. J., R. W. PADGETT, S. D. FINDLEY, B. G. RICHTER, and M. SANICOLA et al., 1997  Molecular evolution at the decapentaplegic locus in Drosophila. Genetics 145:297-309[Abstract].

PADGETT, R., R. D. ST. JOHNSTON, and W. M. GELBART, 1987  A transcript from a Drosophila pattern gene predicts a protein homologous to the TGF-ß family. Nature 325:81-84[Medline].

PANGANIBAN, G., R. REUTER, M. SCOTT, and F. M. HOFFMANN, 1990  A Drosophila growth factor homolog, decapentaplegic, regulates homeotic gene expression within and between germ layers during midgut morphogenesis. Development 110:1041-1050[Abstract/Free Full Text].

PIRROTA, V., 1990  Transvection and long-distance gene regulation. Bioessays 12:409-414[Medline].

POSAKONY, L. G., L. A. RAFTERY, and W. M. GELBART, 1991  Wing formation in Drosophila melanogaster requires decapentaplegic function along the anterior-posterior compartment boundary. Mech. Dev. 33:69-82.

RIO, D., 1990  Molecular mechanisms regulating Drosophila P element transposition. Annu. Rev. Genet. 24:543-578[Medline].

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SMITH, D., J. WOHLGEMUTH, B. R. CALVI, I. FRANKLIN, and W. M. GELBART, 1993  hobo enhancer trapping mutagenesis in Drosophila reveals an insertion specificity different from P elements. Genetics 135:1063-1076[Abstract].

SPENCER, F., F. M. HOFFMANN, and W. M. GELBART, 1982  decapentaplegic: a gene complex affecting morphogenesis in Drosophila melanogaster.. Cell 28:451-461[Medline].

ST. JOHNSTON, R. D., F. M. HOFFMANN, R. K. BLACKMAN, D. SEGAL, and R. GRIMAILA et al., 1990  Molecular organization of the decapentaplegic gene in Drosophila melanogaster.. Genes Dev. 4:1114-1127[Abstract/Free Full Text].

TOWER, J., G. KARPEN, N. CRAIG, and A. SPRADLING, 1993  Preferential transposition of Drosophila P elements to nearby chromosomal sites. Genetics 133:347-359[Abstract].

TWOMBLY, V., R. BLACKMAN, H. JIN, J. GRAFF, and R. PADGETT et al., 1996  The TGF-ß signaling pathway is essential for Drosophila oogenesis. Development 122:1555-1565[Abstract].

WHARTON, K., G. THOMSEN, and W. M. GELBART, 1991  Drosophila 60A gene, another TGF-ß family member, is closely related to human bone morphogenetic proteins. Proc. Natl. Acad. Sci. USA 88:9214-9218[Abstract/Free Full Text].

WHARTON, K., R. P. RAY, and W. M. GELBART, 1993  An activity gradient of decapentaplegic is necessary for the specification of dorsal pattern elements in the Drosophila embryo. Development 117:807-822[Abstract].

WHARTON, K., R. P. RAY, S. FINDLEY, H. DUNCAN, and W. M. GELBART, 1996  Molecular lesions associated with alleles of decapentaplegic identify residues necessary for TGF-ß/BMP cell signaling in Drosophila melanogaster.. Genetics 142:493-505[Abstract].

ZHANG, P. and A. SPRADLING, 1993  Efficient and dispersed local P-element transposition from Drosophila females. Genetics 133:361-373[Abstract].

This article has been cited by other articles:

				S. J. Newfeld and N. T. Takaesu An Analysis Using the hobo Genetic System Reveals That Combinatorial Signaling by the Dpp and Wg Pathways Regulates dpp Expression in Leading Edge Cells of the Dorsal Ectoderm in Drosophila melanogaster Genetics, June 1, 2002; 161(2): 685 - 692. [Abstract] [Full Text] [PDF]