Overexpression Beadex Mutations and Loss-of-Function heldup-a Mutations in Drosophila Affect the 3' Regulatory and Coding Components, Respectively, of the Dlmo Gene

Corresponding author: Daniel Segal, Department of Molecular Microbiology and Biotechnology, Tel-Aviv University, Tel-Aviv 69978, Israel., dsegal{at}ccsg.tau.ac.il (E-mail).

Communicating editor: T. C. KAUFMAN

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

LIM domains function as bridging modules between different members of multiprotein complexes. We report the cloning of a LIM-containing gene from Drosophila, termed Dlmo, which is highly homologous to the vertebrate LIM-only (LMO) genes. The 3' untranslated (UTR) of Dlmo contains multiple motifs implicated in negative post-transcriptional regulation, including AT-rich elements and Brd-like boxes. Dlmo resides in polytene band 17C1-2, where Beadex (Bx) and heldup-a (hdp-a) mutations map. We demonstrate that Bx mutations disrupt the 3'UTR of Dlmo, and thereby abrogate the putative negative control elements. This results in overexpression of Dlmo, which causes the wing scalloping that is typical of Bx mutants. We show that the erect wing phenotype of hdp-a results from disruption of the coding region of Dlmo. This provides molecular grounds for the suppression of the Bx phenotype by hdp-a mutations. Finally, we demonstrate phenotypic interaction between the LMO gene Dlmo, the LIM homeodomain gene apterous, and the Chip gene, which encodes a homolog of the vertebrate LIM-interacting protein NLI/Ldb1. We propose that in analogy to their vertebrate counterparts, these proteins form a DNA-binding complex that regulates wing development.

LIM domains constitute a novel subclass of cysteine-rich motifs and are found in various proteins involved in key processes during development and differentiation (reviewed in CURTISS and HEILIG 1998 ; DAWID et al. 1998 ; JURATA and GILL 1998 ). They are composed of ~55 residues with the consensus sequence CX₂CX_16-23HX₂CX₂CX_16-23CX₂C (where X is any amino acid), and they bind two atoms of Zn²⁺. The "LIM" acronym is derived from the first three homeodomain proteins in which these domains were recognized: lin-11, which functions in asymmetric division of Caenorhabditis elegans secondary vulval blast cells (FREYD et al. 1990 ); Isl1, which binds to the rat insulin I gene enhancer (KARLSSON et al. 1990 ) and has a major function in motor neuron development (PFAFF et al. 1996 ); and mec-3, which is essential for differentiation of touch receptor neurons in C. elegans (WAY and CHALFIE 1988 ). LIM proteins often contain multiple (up to five) LIM domains in tandem. While originally identified in LIM homeodomain proteins, LIM proteins are also found to contain other known domains, such as kinase or GAP domains, as well as transcriptional activation domains (reviewed in SANCHEZ-GARCIA and RABBITTS 1994 ; TAIRA et al. 1995 ). Still, other LIM proteins contain no additional domain of known function [LIM-only proteins (LMO)]. Despite the structural resemblance of the LIM domain to the GATA-1-type zinc finger, there is no evidence that LIM domains bind to DNA. Rather, growing evidence indicates that they mediate interaction with other proteins (CURTISS and HEILIG 1998 ; JURATA and GILL 1998 ).

The nuclear localization of many LIM proteins and the fact that they often contain known transcriptional activation domains prompted the suggestion, and subsequently the demonstration, that the LIM domains play a role in modulation of the activity of the transcriptional activation domains associated with them. For example, the transcriptional activation capability of the amphibian LIM homeodomain protein Xlim-1 is relieved when its LIM domains are either deleted (TAIRA et al. 1994 ) or bound to the interacting protein NLI/Ldb1 (AGULNICK et al. 1996 ; BREEN et al. 1998 ). These observations suggest that the LIM domains of Xlim-1 negatively regulate its transcriptional activation activity. In other cases, LIM domains have been shown to positively regulate transcription activation activity. For example, the murine LIM homeodomain protein Lhx3 acts synergistically with the pituitary-specific POU homeodomain protein Pit-1 in activation of transcription from several pituitary-specific promoters (BACH et al. 1995 ). The association of these proteins is mediated by the LIM domains of Lhx3 and POU domains of Pit-1 (BACH et al. 1995 ).

LIM domains can also mediate binding of two LIM proteins resulting in homodimers, as in the case of the cysteine-rich protein (CRP; FEUERSTEIN et al. 1994 ; SANCHEZ-GARCIA et al. 1995 ) or in heterodimers, e.g., CRP and Zyxin (SADLER et al. 1992 ; SCHMEICHEL and BECKERLE 1994 ).

LMO proteins serve to link different transcription factors that either contain or lack LIM domains. For example, two transcription factors, the zinc finger GATA-1 protein and the bHLH Tal1 protein, synergize in activating transcription from a target gene when they are bridged by the LIM domains of LMO2 (WADMAN et al. 1997 ). A Nuclear LIM-Interacting protein (NLI, also termed Ldb1) has recently been shown to mediate binding of LIM proteins to their partners in various transcription complexes (VISVADER et al. 1997 ; BREEN et al. 1998 ; JURATA et al. 1998 ).

Certain LIM proteins are cytoplasmic (e.g., Zyxin, CRP, CRIP, Paxillin, MLP; CURTISS and HEILIG 1998 ; JURATA and GILL 1998 ). There they also serve as adaptor molecules between various cytoskeletal proteins. For example, Paxillin, which is found at focal adhesion sites, contains binding sites for Vinculin and for the focal adhesion tyrosine kinase FAK (BROWN et al. 1996 ). Zyxin contains a proline-rich, -actinin-binding domain (CRAWFORD and BECKERLE 1992 ), and MLP binds actin filaments (ARBER and CARONI 1996 ). These interactions are mediated by the respective LIM domains.

Taken together, these in vitro studies demonstrate a role for LIM domains in the assembly of different proteins into functional transcription complexes or into higher-order components of the cytoskeleton.

Drosophila offers a unique opportunity to study the role of LIM proteins in the context of the whole organism and to identify by genetic means the proteins they interact with. Here we report the isolation of a LMO gene from Drosophila. The gene has been independently isolated by ZHU et al. 1995 and was termed Dlmo. We show that the 3' untranslated region (UTR) of Dlmo contains various motifs [AT-rich elements (ARE) and Brd-like boxes] that have been implicated in negative post-transcriptional regulation of various genes. We demonstrate that hypermorphic mutations at the Beadex (Bx) locus disrupt the 3'UTR of Dlmo, leading to overexpression of the gene. Furthermore, we show that loss-of-function mutations at the adjacent heldup-a (hdp-a) locus represent lesions in the coding region of Dlmo, thus providing a molecular basis for the genetic interaction between Bx and hdp-a mutations. Finally, we demonstrate phenotypic interactions among mutations in Dlmo, in the apterous LIM homeodomain gene, and in the Chip gene, which is homologous to the vertebrate NLI protein. These observations suggest that in analogy to their vertebrate counterparts, these three proteins form a DNA-binding complex that regulates wing-specific genes.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Drosophila stocks:
Strains were maintained and crosses were conducted on cornmeal-molasses medium at 25°. For egg collection, flies were transferred to bottles attached to egg-laying plates (3% Bacto-agar, 2% sugar, 1.5 g/liter methylparaben) supplemented with live yeast paste at 25°. Description of balancer chromosomes and markers can be found in LINDSLEY and ZIMM 1992 . The Canton-S strain was used as a wild-type stock. The following strains were used: ₂ and C(1)DX, yf; ₂ (kindly provided by W. R. ENGELS), Bx¹, Bx³, Inscy w Bx^M, Inscy w and Df(1)N19/FM6 (provided by the Bloomington Stock Center), Bx², and Dp(1:1)Bx^r, B Bx^r,car/C(1)DX, yf and C(1)DX, ywf (provided by the Mid America Stock Center). Deficiency Df(1)N19 deletes 17C1-2 to 18A1 and the cytology of Dp(1;1)Bx^r is 17A1-17A4; 17E1-17F3 (LINDSLEY and ZIMM 1992 ). Strains used for interaction studies were as follows: yw; Chip^e5.5/CyO, Df(2R)Kr⁴, Kr^B80, Dp(1;2)y⁺ (kindly provided by D. DORSETT), ap^56f, Dp(2;2)41A, al² cy cn² L⁴ sp²/In(2L)CyIn(2R)Cy, and pr cn Dp(2;Y)C (obtained from the above-mentioned stock centers).

P-element-induced mutagenesis:
Males from the ₂ strain (ENGELS and PRESTON 1984 ) were crossed to females from the M strain C(1)DX, y w f. The resulting dysgenic male progeny were crossed in groups of 15–20 to Df(1)N19/FM6 or Bx³/Bx³ tester females. Female offspring from these test crosses that exhibited the relevant phenotype (erect or scalloped wings in the first cross or normal wings in the second cross) were allowed to mate individually with their sibling males, and the resulting male offspring were crossed to compound-X P cytotype females, C(1)DX,y f; ₂, to establish a stable stock. To avoid clusters, only one line from each bottle of the cross was subsequently used for further analysis.

Classification of the scalloped wing phenotype:
Wing margin abnormalities were routinely classified according to the number of notches at the anterior and posterior margins of the wing: rank 1, normal wings; rank 2, one to two notches at the posterior margin; rank 3, more than two notches at the posterior margin; rank 4, more than two notches at the posterior margin plus one to two notches at the anterior margin; rank 5, more than two notches at the posterior and anterior margin; rank 6, notches as in rank 5 plus blisters on the wing blade; rank 7, strap wings; rank 8, club wings (see Figure 5).

View larger version (64K): In this window In a new window Download PPT slide   Figure 1. Nucleotide sequence of the 3'UTR of Dlmo. Numbering is according to the sequence in ZHU et al. 1995 . The stop codon is in lowercase letters. The two putative polyadenylation signals are in bold letters. The two BamHI sites and the position of primers 14 and 15 used in Figure 3 are indicated. The ARE motifs and the T stretch are boxed. Brd-like boxes are underlined. The proximal breakpoint of the deletion in three of the Bx mutants is indicated by dotted lines. Primer 13, which delimits the coding region, is indicated.

View larger version (30K): In this window In a new window Download PPT slide   Figure 2. Map of the Dlmo region at 17C. (A) The position of the Dlmo transcript (solid box) and the P-element insertion site are indicated on the restriction map of the region. Restriction sites are as follows: BamHI, B; SacI, C; EcoRI, R; SmaI, M; HindIII, H; SalI, L. The dotted line indicates the SacI fragment used as a probe for Southern analysis. (B) Enlargement of the map around the Dlmo gene. The exon (boxes)-intron (solid lines) organization of the Dlmo transcript is outlined. Exons Ia and II are not included in the map depicted in A. The coding region is indicated by the dotted boxes. The two LIM domains are enclosed in bold boxes, the ARE region is indicated in wavy lines, and the Brd-like boxes are shown as diamonds. Arrowheads depict position and direction of primers used for PCR analysis of Bx and hdp mutants. (C) Characterization of newly generated Bx alleles and hdp-a allele. Hatched boxes represent deleted sequences. Open boxes indicate potentially deleted sequences (exact breakpoint ends not determined). Positions of P-element insertions are shown for Bx120-4-1 and Bx122-3-1.

View larger version (124K): In this window In a new window Download PPT slide   Figure 3. PCR analysis of existing Bx alleles (Bx1, Bx2, Bx3, and BxM) and two wild-type strains, Canton S (CS) and Inscy w (Ins), used to generate BxM. Primers 14 and 15 flanking the two BamHI sites present in the 3'UTR of Dlmo were used (see Figure 1). BstEII digest was used as size markers.

View larger version (22K): In this window In a new window Download PPT slide   Figure 4. Scheme of crosses used to produce mutants with erect wings and scalloped wings (Cross A) and suppressors of Bx (Cross B).

View larger version (53K): In this window In a new window Download PPT slide   Figure 5. Abnormal wing phenotype of Bx mutants. Wing margin severity defect was ranked according to the number of notches at the anterior and posterior margins of the wing (see MATERIALS AND METHODS).

Standard DNA techniques:
Restriction site mapping, Southern blotting, subcloning, library screening with ³²P-labeled probes, and isolation of genomic DNA were carried out essentially as described by SAMBROOK et al. 1989 . The genomic library used was in FIX II (Stratagene, La Jolla, CA).

PCR screen for LIM-containing genes:
Genomic DNA, a cDNA library constructed from 0–4-hr-old embryos (courtesy of N. BROWN), as well as cDNA generated from Canton-S embryos, larvae, pupae, or adults, were used as templates.

Partially degenerate primers were designed according to conserved amino acid sequences from the LIM domains of the Drosophila apterous gene and other LIM domain-containing proteins from vertebrates and plants (BALTZ et al. 1992 ), taking into consideration the codon usage of Drosophila. The following primers were used:

P1: 5'AGACACTGCAGCAGAAGCAGTTGACGTGAAAAAC 3'

P2: 5'AACAAGAATTCTGGCGCGCATGAC 3'

P3: 5'TGCGGCGAGCTCATC/ACAGGAC/TCGCTA/TT/CC/TTCCTC 3'

P4: 5'GCTAACCTCGAGGTAA/GTCCCGT/CTTGCA 3'

P5: 5'CGAATTCTGCG/TCCGGCTGCGGC 3'

The designed primers had the potential to amplify a single LIM domain (primer pairs P3 + P4 or P5 + P4) or tandem LIM domains (primer pairs P3 + P1, P5 + P1, P2 + P3, or P5 + P2). A variety of temperature ranges were used for annealing (37, 42, or 55°). Amplified fragments were subcloned in pBluescript and sequenced at the sequencing unit of Tel-Aviv University.

Northern analysis:
PolyA⁺ RNA was prepared from 0–4-hr-old embryos of different genotypes using the mRNA purification kit of Pharmacia (Piscataway, NJ). ³²P-labeled riboprobes were synthesized using the 1.8-kb cDNA of Dlmo and rp49 as templates. Northern blots were quantified using ImageMaster DTS and ImageMaster 1D software (Pharmacia).

In situ hybridization:
In situ hybridization to larval imaginal discs was performed according to the method of TAUTZ and PFEIFLE 1989 . The 1.8-kb cDNA clone of Dlmo was labeled with digoxigenine (Boehringer Mannheim, Indianapolis) and used as a probe. We took special care to perform the procedure on the discs of all strains simultaneously and under the same conditions. Discs were mounted in glycerol and photographed.

PCR analysis of mutants:
Genomic DNA was used for long and short PCR reactions. Long PCR was performed using the Expand Long Template PCR System (Boehringer Mannheim), and short PCR was performed using Taq polymerase (Appligene) according to the manufacturer's instructions. The following set of 17 primers was designed according to the sequence of the 1.8-kb cDNA of Dlmo and of adjacent genomic sequences: primer 1, nt 4–26; primer 2, nt 173–186; primer 3, nt 360–383; primer 4, nt 4011–4033; primer 5, nt 79–101; primer 6, nt 275–295; primer 7, nt 518–540; primer 8, nt 649–661; primer 9, nt 841–860; primer 10, nt 938–960; primer 11, nt 1027–1046; primer 12, nt 1041–1061; primer 13, nt 1328–1350; primer 14, nt 1549–1568; primer 15, nt 2098–2117; primer 16, nt 3230–3253; primer 17, nt 6401–6425.

For each primer, the position of the corresponding sequence is given according to the numbering in Figure 2 of ZHU et al. 1995 . Primers corresponding to downstream and upstream genomic sequences also follow this numbering. Note that primers 5 and 6 correspond to exon Ib in ZHU et al. 1995 . The orientation of the primers is indicated in Figure 2B.

Two primers were designed according to 5' and 3' ends of the P element:

5' P element: 5'ATACTTCGGTAAGCTTGCGCTATC3'

3' P element: 5'CATACGTTAAGTGGATGTCTCTTG3'

PCR fragments were cloned into the pGEM-T-vector (Promega, Madison, WI), when deemed necessary, and were sequenced.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Isolation of Dlmo and the structure of its 3'UTR:
Various combinations of partially degenerate PCR primers, designed according to conserved LIM sequences, were used to amplify single or tandem LIM domains from templates of either genomic DNA, an embryonic cDNA library, or cDNA of different developmental stages of Drosophila melanogaster. One PCR-amplified fragment, produced from an embryonic cDNA template using primers P3 and P4 (see MATERIALS AND METHODS), was found upon sequencing to contain a novel LIM consensus motif. This PCR fragment was used as a probe to isolate cDNA clones from a 3–12-hr-old embryonic cDNA library. A 1.8-kb embryonic cDNA clone was isolated and sequenced. It was found to comprise 1798 nucleotides capable of encoding a 266-amino acids-long protein with two tandem LIM motifs and no homeodomain or any other known domain. Thus, it is a LMO protein. While this work was in progress, the same sequence was reported by ZHU et al. 1995 , who termed it Drosophila LMO (Dlmo). Our cDNA contains exons 2–5, identified by ZHU et al. 1995 . The first LIM domain of the conceptual Dlmo protein is 79 and 69% similar to the first LIM domain of the human LMO1 and LMO2, respectively, whereas the second LIM domain of Dlmo is 94 and 60% similar to the second LIM domain of the human counterparts, respectively.

The structure of the 776-nucleotide-long 3'UTR of Dlmo has particular bearing on the analysis of mutants in this gene (described below). It contains four AT_3-5A boxes, seven AT₃ motifs, and several stretches of Ts, the longest of which is T₁₂, collectively referred to as ARE (see Figure 1), which in various eukaryotic genes increase destabilization of the transcript (CHEN and SHYU 1995 ). In addition, the 3'UTR of Dlmo contains five AGT_3-4A sequences (Figure 1) that are closely related to the Drosophila Brd box (AGCTTTA). Brd boxes also mediate negative post-transcriptional regulation, probably by conferring instability upon the transcript (LAI and POSAKONY 1997 ). Thus, the 3'UTR of Dlmo may be involved in negative regulation of Dlmo.

Bx mutations affect the 3'UTR of Dlmo:
The 1.8-kb Dlmo cDNA was used as a probe to isolate genomic clones from a FIX Drosophila genomic library. Three partially overlapping clones were isolated. In situ hybridization to polytene chromosomes using the 1.8-kb Dlmo cDNA as a probe indicated that the gene maps to band 17C on the X chromosome (data not shown). This is in accord with the mapping by ZHU et al. 1995 . The 17C region has been previously the subject of chromosomal walks, and a total of nearly 250 kb of genomic DNA from this region have been isolated and physically mapped (MATTOX and DAVIDSON 1984 ; MARIOL et al. 1987 ). Comparison of the published restriction maps from these walks with the map derived from the three genomic phage clones we have isolated indicated potential overlap. Southern analysis of genomic clones from the walk (clones 13, 24, and 19; Figure 4 in MATTOX and DAVIDSON 1984 ; kindly provided by W. MATTOX) probed with the 1.8-kb Dlmo cDNA verified that Dlmo is contained within the genomic sequences included in these clones (Figure 2A).

Several genes are known to reside in the 17BC region (FlyBase; LINDSLEY and ZIMM 1992 ). Of these, recessive mutations in hdp-a (causing erect wings) and dominant mutations in Bx (causing scalloping of the wing margin) have been previously mapped to a 0.4-kb BamHI fragment within that part of the 17C chromosomal walk, which we have shown to overlap Dlmo. Because much of the 3'UTR of Dlmo resides within a 0.4-kb BamHI fragment (Figure 1), we examined whether hdp-a or Bx mutations localize to it. Five Bx mutants are available (Bx¹, Bx², Bx³, Bx^J, and Bx^M), all of which are insertional mutants of different transposable elements (MATTOX and DAVIDSON 1984 ), but all previously described hdp-a mutants have been lost.¹ We have performed PCR reactions using primers flanking the 0.4-kb BamHI fragment in the 3'UTR of Dlmo (Figure 1), as well as genomic DNA from these five Bx mutants as templates. These reactions amplified a fragment longer than the expected 0.4-kb fragment from each of the Bx mutants (Figure 3). These results indicate that Bx¹, Bx², and Bx^J contain inserts ~8 kb long, and Bx³ and Bx^M contain inserts 9 and 14 kb long, respectively, within the 0.4-kb BamHI fragment of the 3'UTR of Dlmo. The sizes of the inserts in these mutants are in agreement with the results of MATTOX and DAVIDSON 1984 , which were derived from restriction mapping and Southern analysis. These results demonstrate that Bx mutations localize to Dlmo.

The inserts in the Bx mutations analyzed disrupt the 3'UTR of Dlmo. We hypothesize that these mutations may exert their phenotypic effect by interfering with the function of the putative negative regulatory motifs present in the 3'UTR, resulting in overexpression of the Dlmo transcript. A critical prediction of this hypothesis is that removal of these motifs from the 3'UTR of Dlmo would result in a mutant phenotype characteristic of insertional Bx mutations.

To produce deletions in Dlmo that might result in Bx-like wing scalloping, and in the hope of generating new hdp-a mutations, we mobilized P elements in the wild-type ₂ strain by hybrid dysgenesis. This strain contains multiple copies of P element, one of which maps to 17C2-3 (O'HARE and RUBIN 1983 ). Hybrid dysgenesis in ₂ has been shown to be an effective means for producing Bx and heldup mutants (ENGELS and PRESTON 1984 ), most of which were attributable to the P element in 17C2-3 (SIMMONS et al. 1984 ). Comparison of the sequence of the genomic 1.8-kb BamHI fragment flanking the P element in 17C2-3 (O'HARE and RUBIN 1983 ; courtesy of K. O'HARE) with the sequence of Dlmo indicates that this P element in ₂ resides 707 bp downstream of the 3' end of our cDNA, suggesting that deletions extending into Dlmo could be generated by imprecise excision of this P element.

Generation and characterization of new Bx mutants:
Males from the ₂ strain were crossed to M cytotype C(1)DX, y w f females. The resulting dysgenic sons were crossed to Df(1)N19/FM6 females (Figure 4, Cross A), and the ~60,000 FM6-bearing female progeny were screened for wing scalloping. To identify which of these may be Bx mutants, we used the following two criteria: (1) Wing scalloping in Bx mutants is suppressed when combined with a deletion of the 17BC chromosomal region. Flies heterozygous for such a deletion, e.g., Df(1)N19, have normal wings. (2) Bx wing scalloping is augmented when combined with a chromosome carrying a duplication of the normal 17BC region. Flies heterozygous for such a duplication, e.g., Dp(1;1)Bx^r, have normal wings and rarely display very mild scalloping (LIFSCHYTZ and GREEN 1979 ). We crossed FM6-bearing female progeny displaying scalloped wings to males carrying Df(1)N19 and scored the female offspring for amelioration of the wing scalloping. Putative Bx mutants thus identified were subsequently crossed to Dp(1;1)Bx^r, and female offspring were scored for augmentation of wing scalloping. Based on these two criteria, 17 out of the 34 dominant X-linked wing scalloping mutants generated are Bx mutants (Table 1).

All 17 new Bx mutants are homozygous viable. The degree of wing scalloping varied between the different mutants, and they are listed in Table 1 in their order of severity. In each mutant, scalloping was more pronounced in the homozygotes than in the heterozygotes. We have classified scalloping according to its severity, where rank 1 is normal wing and rank 8 is a nearly strap-like wing (Figure 5; Table 1). We find that a population of flies carrying any Bx allele displays a characteristic distribution of severity ranks of wing scalloping. Suppression of scalloping in Bx/Df(1)N19 was therefore recognizable as a shift in the distribution of the wing phenotypes toward the less severe ranks, as compared to Bx/+ (Figure 6). The suppressed phenotype appeared to be directly correlated to the severity of scalloping caused by the original Bx mutation. The milder the original scalloping was, the closer the suppressed phenotype was to wild-type wings (Table 1; Figure 6). Suppression was therefore barely noticeable for the mildest mutants, such as Bx^120-4-1, but was very conspicuous in more severe mutants, such as Bx³, Bx^10-5-2, and Bx^4-5 (Figure 6). Likewise, augmentation of any Bx allele by Dp(1;1)Bx^r was evident as a shift of the wing phenotypes toward the more severe ranks (Table 1; Figure 6). Here, too, the resulting phenotype appeared to be directly correlated to the severity of scalloping caused by the original mutation, and augmentation of scalloping was readily observed, even in the mildest mutants. For example, marked augmentation of wing scalloping is seen for Bx³ and the newly induced Bx alleles Bx^120-4-1 and Bx^4-5, but it is less evident for Bx^10-5-2.

View larger version (47K): In this window In a new window Download PPT slide   Figure 6. Distribution of wing defects according to severity (see Figure 5) in the existing Bx3 allele and in three newly generated Bx alleles, Bx10-5-2, Bx4-5, and Bx120-4-1. Each histogram represents the distribution of wing scalloping in hetrozygous females and shows the suppression of this phenotype by Df(1)N19 and augmentation by Dp(1;1)Bxr. Approximately 100 flies were scored for each genotype.

These observations suggest that Bx mutations are hypermorphs and that the degree of severity of the Bx-engendered wing scalloping depends on the quantity of a certain gene product (see also LIFSCHYTZ and GREEN 1979 ). Taken together, these phenotypic results and our demonstration that Bx mutations interfere with the 3'UTR of Dlmo, which may negatively regulate the stability of the Dlmo transcript, suggest that this gene product is likely to be DLMO. We propose that the hypermorphic nature of these mutations results from overexpression of Dlmo, resulting from abrogation of negative control motifs in its 3'UTR.

Molecular analysis of Bx mutants:
PCR and Southern analyses were performed on genomic DNA of 11 out of the 17 newly induced Bx mutants. Genomic DNA from the ₂ strain, used for generating these mutants, served as a control. The primers used for these reactions were designed according to sequences from the cDNA of Dlmo and from genomic sequences flanking the site of insertion of the 17C2-3 P element in ₂ (see Figure 2B and MATERIALS AND METHODS for exact location of the primers).

Interestingly, this analysis provides evidence that 8 of the 11 newly induced Bx mutants examined have lesions confined to the 3'UTR of Dlmo and the downstream flanking genomic sequences. For example, the mutants Bx^4-5 and Bx^110-4-1 amplified fragments that are 1 and 0.6 kb long, respectively, when using primer pair 14 + 16, while the ₂ control gave a 4.5-kb fragment, suggesting they have deletions in the region delimited by these primers (Figure 7). Likewise, the mutant Bx^10-5-2 amplified a fragment 0.8 kb long when using primer pair 11 + 16, as compared to 5.2 kb in the control, suggesting it has a deletion in the region between their corresponding genomic sequences (Figure 7). In the mutants Bx^17-8-1 and Bx^113-4-1, the primer pair 14 + 17 amplified fragments 2.5 and 1.8 kb long, respectively, while the corresponding control fragment is 7.7 kb long. Given that the ₂ strain has a 2.9-kb P element inserted 0.7 kb downstream of Dlmo (Figure 2A; O'HARE and RUBIN 1983 ), the sizes of the deleted genomic fragments in these mutants, excluding the P element, are as follows: Bx^4-5, 0.6 kb; Bx^110-4-1, 1 kb; Bx^10-5-2, 1.5 kb; Bx^17-8-1, 2.3 kb; and Bx^113-4-1, 3 kb. In the mutant Bx^17-7-1, no fragments were amplified using any primer pair downstream of primer 13, which delimits the 3' end of the coding region, even when using the most downstream primer available, primer 17 (Figure 7). This indicates that the deletion in Bx^17-7-1 extends downstream beyond sequences corresponding to primer 17 (Figure 2B). Taken together, in each of these six Bx mutants, the deletion removes part of the 3'UTR of Dlmo, but leaves the coding region of the gene intact. We have confirmed these results by Southern analysis (data not shown).

View larger version (64K): In this window In a new window Download PPT slide   Figure 7. PCR-amplified products from new Bx mutants.

To get a more accurate estimate of the extent of the deletions, we cloned and sequenced the PCR fragments of the mutants Bx^4-5 and Bx^110-4-1 amplified using primer pair 14 + 16, and the fragment of Bx^10-5-2 amplified using primer pair 11 + 16 (Figure 7). Sequence analysis has confirmed that the deletion in each of these mutants extends from the site of the P-element insertion in the ₂ strain, removing the P element entirely, into the 3'UTR of Dlmo. The portion of the 3'UTR sequences deleted is different in the three mutants. In Bx^4-5, the last 226 bp of the 3'UTR of Dlmo are missing, including one ATTTA motif, the T stretch, and one Brd-like box (Figure 1). In Bx^110-4-1, an additional 126 bp have been deleted from the 3'UTR, removing all ARE motifs and four Brd-like boxes (Figure 1). In Bx^10-5-2, the deletion is larger, leaving only the first 107 bp of the 3'UTR and removing all the putative negative regulatory elements present in the 3'UTR of Dlmo (Figure 1). Taken together, the PCR and sequence analyses of these Bx mutants support the hypothesis that the Bx mutant phenotype results from abrogation of the 3'UTR of Dlmo. Interestingly, the extent of the deleted segment from the 3'UTR of Dlmo in these mutants is directly correlated with the severity of their wing scalloping, where Bx^10-5-2 has the most severe phenotype, Bx^110-4-1 is less severe, and Bx^4-5 is the mildest of the three (Table 2). In addition, the 3'UTR DNA lesions in Bx^110-4-1 and Bx^10-5-2 differ only in that the former lacks an additional Brd-like box. This may provide functional evidence that Brd-like boxes in Dlmo have a negative regulatory role.

The PCR products amplified from the mutants Bx^122-3-1 and Bx^120-4-1 using primer pair 11 + 15 (8 and 4 kb, respectively) were longer than the corresponding fragment in ₂ (1 kb, Figure 7), indicating that sequences ~7 and 3 kb long, respectively, have inserted into the fifth exon of the Dlmo gene in each of these mutants. Because the PCR product of primer pairs 14 + 15 and 11 + 13 in these mutants has the same size as in the control (Figure 7), we deduce that the inserts in these two mutants are localized to the 179-bp region between the sequences corresponding to primers 13 and 14, in the 5'-most portion of the 3'UTR of Dlmo (Figure 1 and Figure 2C).

To further characterize the insertions in these two mutants, we used primers designed according to the 5' and 3' ends of the P-element sequences, in combination with primer 11 (Figure 2B; see MATERIALS AND METHODS). The combination using the 5' primer of the P element yielded fragments of 2 kb in the mutant Bx^122-3-1 and of 0.5 kb in the mutant Bx^120-4-1. These findings support the conclusion that in both mutants, the insertion was in the interval mentioned above. In the mutant Bx^122-3-1, the insertion consists of an excised P element with 1.7 kb of flanking genomic sequences at its 5' end and 2.5 kb of flanking genomic sequences at its 3' end (Figure 2C). In mutant Bx^120-4-1, the insertion consists of the P-element sequences only (Figure 2C).

In addition, no fragment was amplified using primer pair 14 + 16 in these two mutants. On the other hand, normal size fragments were amplified from their DNA using either primer pair 14 + 15 or the primer pairs located upstream to them. This suggests that in addition to the insertion, they have lesions removing sequences downstream of Dlmo, leaving its coding region intact (Figure 2C).

The mutants Bx^3-5-4, Bx^16-7, and Bx^111-1-8 had no visible difference from the control ₂ strain in all primer pairs used. Because of the limited resolution of the PCR technique, we cannot exclude the possibility that they have small lesions in the Dlmo gene that are responsible for their wing scalloping. This should be resolved by sequencing.

Dlmo expression is affected in Bx mutants:
Northern analysis of poly(A)⁺ RNA extracted from 0–4-hr-old embryos of several Bx mutants (Bx¹, Bx², Bx³, and Bx^M) and a control strain (Canton-S) was performed using the 1.8-kb cDNA of Dlmo as a probe. A 2.0-kb transcript exists in the control strain, while all Bx mutants examined have a truncated transcript that is ~0.5 kb shorter (Figure 8A). The longer transcript in Bx² may be the result of transcription termination within the gypsy element (see DISCUSSION).

View larger version (86K): In this window In a new window Download PPT slide   Figure 8. Expression of Dlmo in Bx mutants. (A) Northern analysis of wild-type, Canton S (CS), and four insertional Bx mutants. Poly(A)+ RNA was extracted from 0- to 4-hr-old embryos and probed with Dlmo cDNA. The bottom panel shows the reprobing with rp49 as a loading control. (B) Ratio of Dlmo transcript (Bx/CS) normalized according to rp49. (C) Expression pattern of Dlmo in wild-type (CS) and Bx3 mutant wing imaginal discs.

To correct for the amounts of poly(A)⁺ RNA loaded in each lane, the membrane was rehybridized with a probe of the ribosomal protein RP49. Scanning for quantification of the Dlmo RNA and the rp49 RNA and normalizing for RNA loading indicated that the Dlmo transcript is overexpressed (two- to fourfold) in the Bx mutants as compared to the wild-type strain (Figure 8B).

Because the Bx phenotype is manifested in the wing, we compared the expression of Dlmo in the wing imaginal discs of Bx³/Bx³ and the wild-type Canton-S. The overexpression of Dlmo observed in the mutant embryos (Figure 8A and Figure B) is also evident in their discs (Figure 8C), which display stronger staining.

Generation and characterization of hdp-a mutants:
All hdp-a mutants that have been generated previously were lost. The fact that hdp-a mutants suppress the wing scalloping phenotype of Bx mutants (LIFSCHYTZ and GREEN 1979 ) prompted us to try and generate new hdp-a alleles, by hybrid dysgenesis, to study the molecular basis underlying this interaction.

Two phenotypic criteria were used to identify recessive hdp-a mutants: (1) uncovering of the hdp-a mutation by Df(1)N19 and (2) the ability of hdp-a mutants to suppress the dominant wing scalloping of Bx. These two strategies were used to screen for hdp-a mutants.

Strategy 1: Approximately 30,000 Df(1)N19-carrying female offspring from Cross A were screened for the erect wing phenotype (Figure 4). Six such independent mutants were recovered, and they were subsequently confirmed to be X-linked, recessive, and homozygous viable. We crossed the six erect wing mutants to Bx³ and observed amelioration of wing scalloping in the female offspring of four (hdp-a^7-5-4, hdp-a^115-5-1, hdp-a^19-4-1, and hdp-a^13-7-13, Table 3). These four mutants do not complement each other, and all transheterozygous combinations of them display erect wings. These results suggest that these four mutants represent lesions in the hdp-a gene.

Strategy 2: Dysgenic males were crossed to Bx³/Bx³ females, and their female offspring were scored for suppression of the dominant wing scalloping (Figure 4, Cross B). Approximately 10,500 chromosomes were screened, and 20 such independent, X-linked Bx suppressors were isolated. These could correspond to lesions in hdp-a or in other second-site Bx-suppressor genes. Four of the Bx suppressors (hdp-a^32-4-14, hdp-a^26-3-7, hdp-a^128-1-5, and hdp-a^28-3-3) also have a recessive erect wing phenotype that is uncovered by Df(1)N19. These results suggest that they have lesions in hdp-a. All four mutants do not complement each other for the erect wing phenotype, nor do they complement the four hdp-a mutants generated in strategy 1. We conclude that these four Bx suppressors are hdp-a alleles, comparable to the hdp-a mutants generated in strategy 1.

Ranking the severity of the erect wing phenotype in these eight hdp-a mutants was difficult. The severity of the hdp-a mutants was easier to measure by determining the extent of their suppression of the wing scalloping of Bx mutants. The different degrees of severity of the eight hdp-a mutants, using this criterion, are depicted in Table 3. These hdp-a mutants also suppress the wing scalloping of the 17 newly generated Bx mutants (see Table 1 for suppression of the new Bx alleles by hdp-a^32-4-14), supporting the conclusion that they are indeed new Bx alleles.

Based on the degree of suppression of Bx³, the hdp-a^32-4-14 allele is the most severe hdp-a allele in our collection.

hdp-a mutants represent loss of function of Dlmo:
Given the close proximity of Bx and hdp-a mutations on the genetic map (0.0045 map units, LIFSCHYTZ and GREEN 1979 ) and their phenotypic interaction, we were interested to examine what gene in the vicinity of Dlmo is affected by hdp-a mutations.

PCR analysis was performed on the hdp-a^32-4-14 allele, which was shown to suppress completely the wing scalloping phenotype of Bx³ using primer pairs covering the entire Dlmo transcript (Figure 2B). All primer pairs corresponding to exons Ia, II, Ib, III, and IV amplified fragments identical in size to those amplified in the parental strain ₂ (data not shown). When primer pairs designed according to sequences corresponding to exon V were used, however, no amplified fragments were obtained. These results indicate that exon V was entirely absent in this hdp-a mutant, resulting in the loss of approximately half of the coding sequence of Dlmo, including part of the second LIM domain (Figure 2C). Another PCR reaction using primer 16, which is located downstream of the insertion site of the P element, in combination with primer 9, located in exon IV, indicated that the deletion in this mutant extends beyond the insertion site of the P element (Figure 2C).

Additional evidence supporting this conclusion was obtained from Southern analysis of hdp-a^32-4-14 using either the 1.8-kb cDNA of Dlmo or a genomic 5.9-kb SacI fragment as a probe (Figure 2A).

Thus, the loss-of-function nature of the hdp-a^32-4-14 mutation results from disruption of the coding region of Dlmo. This result, combined with the overexpression of Dlmo in Bx mutants, provides molecular grounds for explaining the phenotypic interaction between Bx and hdp-a mutations.

Dlmo interacts genetically with ap and Chip:
The LIM motifs in LIM proteins function in protein-protein binding and, in some cases, mediate LIM-LIM interaction between different LIM proteins (CURTISS and HEILIG 1998 ; JURATA and GILL 1998 ). In Drosophila, recessive mutations in the LIM homeodomain gene apterous (ap) cause truncated wings. The ap gene has been shown to be a key regulator of wing development (COHEN et al. 1992 ). Because both Dlmo and ap contain LIM domains and both affect wing development, we examined whether they interact. We generated various Bx-ap double heterozygotes and examined the morphology of their wings. The results are summarized in Table 4.

Double heterozygotes for overexpression mutations of Dlmo and for loss-of-function mutations of ap exhibit abnormal wing morphology markedly different from the phenotype of either of the two mutants alone. For example, slight overexpression of Dlmo combined with an ap mutation results in conspicuous augmentation of wing scalloping (Dp(Dlmo)/+; ap56f/+, rank 3, vs. Dp(Dlmo)/+, rank 1–2). Further increase in overexpression of Dlmo in ap heterozygotes leads to dramatic enhancement of wing scalloping (Bx³/+; ap^56f/+, rank 5–6). The synergistic effect of Bx and ap mutations suggests that Dlmo and ap interact during wing development.

We further examined how manipulation of the levels of these gene products affects the wings. Heterozygotes for loss of function of both Dlmo and ap have normal wings (Df(Dlmo/+; ap^56f/+, rank 1). Likewise, elevated levels of ap⁺ [e.g., three doses in Dp(ap⁺)/+ or four doses in Dp(2;Y)C; Dp(2;2)41A/+] do not affect wing morphology (rank 1, Table 4; M. SHORESH and D. SEGAL, unpublished observations, respectively). However, when combined with slight or marked overexpression of Dlmo, the elevated levels of ap⁺ augment the wing scalloping of Bx [e.g., Dp(Dlmo)/+; Dp(ap⁺)/+, rank 6, and Bx³/+; Dp(ap⁺)/+, rank 6–7]. These results corroborate the conclusion that Dlmo and ap interact during wing development, and they imply that this interaction is sensitive to the dosage of their gene products.

In vertebrates, the NLI protein (also called Ldb1) has been shown to mediate the binding of LIM proteins to various transcription factors (AGULNICK et al. 1996 ; VISVADER et al. 1997 ; BREEN et al. 1998 ; JURATA and GILL 1998 ). Recently, the Drosophila homolog of NLI, called Chip, has been isolated (MORCILLO et al. 1997 ). Interestingly, loss-of-function mutants of Chip cause, in single doses, very mild nicks in the posterior wing margin (MORCILLO et al. 1997 ). This phenotype is distinct from the Bx or ap wing scalloping. The CHIP protein binds in vitro the LIM domains of AP (MORCILLO et al. 1997 ), and the ap-Chip interaction (e.g., ap^56f +/+ Chip^e5.5) results in dramatic truncation of the wing blade (Table 4; MORCILLO et al. 1997 ). Given these observations and the fact that Dlmo is a LIM-containing gene that interacts with ap, we wanted to examine whether Chip mutants and Dlmo mutants interact. Reduction in the level of Dlmo does not affect the loss-of-function Chip phenotype (Df(Dlmo)/+; Chip^e5.5/+; Table 4); however, elevation of the level of Dlmo in the Chip mutants (e.g., in Bx³/+; Chip^e5.5/+) results in a synergistic effect on wing development (rank 6). These results suggest that Chip and Dlmo interact, and this interaction is sensitive to the relative dosage of their gene products.

These phenotypic interactions indicate that ap, Dlmo, and Chip share a role in the regulation of wing margin development in Drosophila. The in vitro binding of AP and CHIP and the analogy to their vertebrate counterparts collectively suggest that these three proteins form a DNA-binding complex regulating wing-specific genes.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

LIM-containing genes in Drosophila:
The Dlmo gene belongs to a growing family of animal and plant genes encoding LIM proteins. LIM proteins have key roles in diverse processes during development and differentiation (CURTISS and HEILIG 1998 ; DAWID et al. 1998 ; JURATA and GILL 1998 ). The LIM-containing genes identified to date in Drosophila exemplify the diverse and pivotal roles of LIM proteins. The apterous (ap), islet (isl), and Arrowhead (Awh) genes all encode, in addition to the LIM domains, a homeodomain, and are thus likely to be transcription factors. The ap gene is a key regulator of dorso-ventral patterning in the wing (DIAZ-BENJUMEA and COHEN 1993 ; BLAIR et al. 1994 ) and is required for specification of embryonic muscle precursors (BOURGOUIN et al. 1992 ). In addition, ap is expressed in the embryonic central nervous system (CNS) and is required for projection of axons along their appropriate pathways (LUNDGREN et al. 1995 ). ap has also been implicated in neuroendocrine regulation of adult reproduction (ALTARATZ et al. 1991 ; RINGO et al. 1991 ). The vertebrate homolog of ap, Lhx2, is expressed in the embryonic nervous system, and mice homozygous for a null Lhx2 mutation display massive brain defects (PORTER et al. 1997 ). The Drosophila isl gene, like its vertebrate homologs, Islet-1 and Islet-2, is expressed in a subset of embryonic motor neurons and interneurons, and loss of its function causes defects in axon pathfinding and targeting (PFAFF et al. 1996 ; THOR and THOMAS 1997 ). The Awh gene is required for the establishment of a subset of imaginal tissues, the abdominal histoblasts, and the salivary imaginal rings (CURTISS and HEILIG 1997 ). In addition, two homologs of the vertebrate cytoplasmic muscle LIM proteins (MLP) have been cloned from Drosophila, Mlp60A, and Mlp84B. No mutants have been described so far in either of these two genes; however, accumulating data suggest that like their vertebrate homologs, the Drosophila Mlp genes have a role in myogenesis (STRONACH et al. 1996 ). Thus, key roles of LIM proteins appear to have been conserved from insects to mammals (CURTISS and HEILIG 1998 ; DAWID et al. 1998 ; JURATA and GILL 1998 ).

Dlmo is the sole LMO gene identified so far in Drosophila (ZHU et al. 1995 ). On the other hand, the family of vertebrate LMO genes includes three genes, LMO1, LMO2, and LMO3 (FORONI et al. 1992 ). They function in mammalian hematopoiesis and, like the LIM homeodomain proteins, appear to play a role in transcription as they localize to the nucleus and associate with other known transcription factors (CURTISS and HEILIG 1998 ; DAWID et al. 1998 ; JURATA and GILL 1998 ).

Negative regulatory elements in the 3'UTR of Dlmo:
The 3'UTR of Dlmo contains multiple AREs and five Brd-like boxes. AREs are found in the 3'UTR of many mRNAs that code for proto-oncogenes, nuclear transcription factors, and cytokines (for reviews see CHEN and SHYU 1995 ; ROSS 1995 ; ABLER and GREEN 1996 ). They represent the most common determinant for RNA stability in eukaryotic cells. Numerous studies in various in vivo and in vitro systems have shown that deletion or disruption of AREs results in more stable mRNAs, whereas the addition of AREs to the 3' of reporter genes causes destabilization of the transcript. The mechanisms by which AREs direct mRNA degradation and the cis- or trans-acting factors involved are largely unknown. Thus, the AREs in the 3'UTR of the Dlmo gene are likely to be involved in negative post-transcriptional regulation. Interestingly, like their Drosophila homolog, the mammalian LMO1 and LMO2 genes contain AREs in their 3'UTR (MCGUIRE et al. 1989 ; ROYER-POKORA et al. 1991 ).

In addition to the AREs, the 3'UTR of Dlmo contains five heptanucleotide AGTTTTA sequence motifs that are closely related to the AGCTTTA motif, termed Brd box, found in the 3'UTR of the Bearded (Brd) gene and in many genes involved in Notch signaling during cell fate specification in the adult peripheral nervous system of Drosophila (LAI and POSAKONY 1997 ; LEVITEN et al. 1997 ). One of the five Brd-like boxes in the 3'UTR of Dlmo is located within the interval containing the AREs, whereas the remaining four are located upstream. The Brd boxes have been shown to be negative regulatory elements (LAI and POSAKONY 1997 ).

Bx lesions abrogate negative regulatory elements in the 3'UTR of Dlmo:
The results presented in this article demonstrate that the genetically defined Bx locus corresponds to the 3'UTR of Dlmo. Insertion of a P element or a retrotransposon in the 3'UTR of Dlmo can result in a dominant wing scalloping phenotype similar to that caused by removal of most or all the AREs and Brd-like boxes in the 3'UTR. We therefore surmise that the insertions into the 3'UTR of Dlmo—by retrotransposons or P elements or by deletion of parts of the 3'UTR of Dlmo—similarly abrogate the negative regulatory effect of the ARE and Brd-like motifs. Consequently, the level of the Dlmo transcript in the Bx alleles examined is two- to fourfold higher than that of the wild type, as expected if the ARE and Brd-like boxes had an RNA-destabilizing effect (CHEN and SHYU 1995 ; LAI and POSAKONY 1997 ). Transcriptional up-regulation caused by regulatory elements in the transposon can be ruled out because a similar hypermorphic phenotype is exhibited in Bx mutants lacking the 3'UTR region. In addition, we find correlation between the extent of the 3'UTR sequences missing in the three deletion-associated alleles and the severity of their wing scalloping.

Similar effects of transposon insertions on 3'UTR negative regulatory motifs have been reported for two other mutants in Drosophila. The dominant gain-of-function mutation Ser^D in the Serrate (Ser) gene results from insertion of the Tirant retrotransposon in the 3'UTR of Ser, causing termination of the transcript in Ser^D within the transposon's long terminal repeat, at a AAUAAA hexanucleotide that probably serves as a polyadenylation signal (THOMAS et al. 1995 ). As a consequence of this premature termination, the Ser transcript is shorter by 600 nucleotides, which contain eight AREs. The Ser transcript and protein were found to be more abundant in the Ser^D mutant than in the wild type. The higher level of Ser transcript was shown to result from greater stability of the transcript in the mutant rather than from higher rate of transcription (THOMAS et al. 1995 ).

A second example is the insertion of the blood retrotransposon in the 3'UTR of the Brd gene, which causes a dominant gain-of-function phenotype (LEVITEN et al. 1997 ). LAI and POSAKONY 1997 have demonstrated that the 3'UTR of the Brd gene confers negative regulatory activity on heterologous reporter genes in vitro and in transgenic flies, and this activity is strongly dependent on the integrity of the Brd boxes. This indicates that Brd is normally regulated negatively by these boxes. The nullifying effect of the blood insert on the RNA-destabilizing activity of the Brd boxes is caused by premature termination of Brd transcription, resulting in a transcript lacking two of the three Brd boxes. This affects both Brd RNA and protein levels.

The Dlmo transcript in Bx¹, Bx², Bx³, and Bx^M is ~0.5 kb shorter than in the wild type. Insertion of various retrotransposons, including copia and gypsy, in different genes in Drosophila causes premature termination of transcription of the host gene (for a review see SMITH and CORCES 1991 ). Transcription often terminates in polyadenylation signals present in the retrotransposon, as has been proposed for the Ser^D mutation (THOMAS et al. 1995 ). In other cases, the retrotransposon insert was shown to potentiate the utilization of an upstream cryptic polyadenylation signal (DORSETT 1990 ). A cryptic polyadenylation signal is located immediately after the translation stop signal in the 3'UTR of Dlmo (Figure 1). Therefore, retrotransposons inserted in the 3'UTR of Dlmo in the Bx¹, Bx², Bx³, and Bx^M mutants may cause truncation of the transcript by either of these means. In Bx², we observed, in addition to the truncated Dlmo transcript, a transcript larger than the wild type. In this mutant, the shorter transcript may represent termination at the cryptic polyadenylation signal, and the larger one may represent termination within the transposable element. Given the resolution of the Northern blot, it is not possible to determine which of these two alternative mechanisms operates in each of these Bx alleles. At any rate, the truncated Dlmo transcript is devoid of most if not all of the 3'UTR negative regulatory motifs. A polyadenylation signal is located in the P element 150 nucleotides downstream of its 5' end. Therefore, the P-element insertions in Bx^120-4-1 and Bx^122-3-1 may have an effect similar to that of the retrotransposons in abrogating the negative regulatory elements in the 3'UTR of Dlmo. Likewise, the Dlmo transcript in the deletion-associated alleles Bx^4-5, Bx^110-4-1, and Bx^10-5-2 lacks most if not all of its 3'UTR. In these mutants, the Dlmo transcript may terminate at the cryptic polyadenylation site in the beginning of its 3'UTR. Alternatively, because the centromere-proximal breakpoint of these deletions is at the site of the P-element insertion in the ₂ strain, it is possible that Dlmo transcription terminates in sequences that are centromere proximal to this site. Indeed, we find that a polyadenylation signal resides 380 nucleotides downstream of the P-insertion site in ₂.

Overexpression of Dlmo causes wing scalloping:
The Bx wing scalloping can be brought about by supernumerary copies of the normal 17BC chromosomal region, each likely having the normal Dlmo gene along with its control regions. This suggests that the abnormal wing morphology results from overexpression of the gene in those cells in which it is normally expressed, albeit at lower levels, rather than from spatial or temporal misexpression. A similar wing scalloping is brought about by Bx mutations that cause overexpression of the Dlmo gene. Therefore, we assume that Dlmo is expressed under its normal spatial-temporal control in these mutants also. This assumption is corroborated by the similar pattern of distribution of the Dlmo transcript in wild-type and Bx mutant imaginal discs, except that in the latter, the level of the transcript appears elevated. Thus, the scalloped wing phenotype is exclusively the result of disruption or deletion of the 3'UTR of the gene. Because wing scalloping is the only overt mutant phenotype in Bx mutants, whether heterozygous or homozygous, the overexpression of Dlmo apparently does not interfere with functions in which the Dlmo product may participate in cells, other than those at the wing margin. It will be interesting to examine the consequences of directed misexpression of Dlmo in cells or stages where it is not normally expressed because LMO proteins serve as bridges between different proteins (CURTISS and HEILIG 1998 ; DAWID et al. 1998 ; JURATA and GILL 1998 ).

heldup-a mutations correspond to loss-of-function of Dlmo and interact with Bx:
Recessive hdp-a mutations have been genetically mapped to close proximity (0.0045 map units) centromere distal of Bx mutations. Furthermore, hdp-a mutations have been reported to suppress in one dose the dominant wing scalloping of Bx mutations either in cis or in trans (LIFSCHYTZ and GREEN 1979 ). Based on these observations, the hypermorphic nature of Bx mutations has been proposed to result from overexpression of a nearby structural gene, possibly hdp-a (LIFSCHYTZ and GREEN 1979 ; MATTOX and DAVIDSON 1984 ). Although all previously existing hdp-a alleles have been lost, we were able to regenerate hdp-a mutants in two ways, and both groups recapitulate the two characteristics of the previous alleles, namely erect wings and suppression of the dominant wing scalloping of Bx mutants. Molecular analysis has been carried out on one of them, hdp-a^32-4-14, demonstrating that it has a deletion of a major part of the coding region of Dlmo, including part of the second LIM domain, suggesting that hdp-a corresponds to loss-of-function of Dlmo. This conclusion is supported by the results obtained by MATTOX and DAVIDSON 1984 from restriction mapping and Southern analysis of one of the previously existing hdp-a alleles. They found that hdp-a^D30r harbors a small deletion extending from the insertion site of the P element in ₂, removing the 0.4-kb fragment to which Bx mutations have been mapped and extending upstream of it. Comparison to the map of Dlmo suggests that the deletion in hdp-a^D30r has removed part of the coding region of Dlmo. Loss of function of Dlmo could be also caused by mutations disrupting the promoter region of the gene.

Assuming that hdp-a mutations cause loss of the functional Dlmo product, we can explain in molecular terms the suppression of the Bx dominant wing scalloping by recessive hdp-a mutations. We propose that in Bx mutants, the Dlmo product is overexpressed because of abrogation of the negative control elements in its 3'UTR. Likewise, duplications of the normal 17BC region result in excess of the Dlmo protein, causing in turn a scalloping phenotype comparable to that of Bx mutants. When either of these duplications or Bx mutations are combined with a deletion of the chromosomal 17BC region or with hdp-a mutations, which likely cause loss of function of Dlmo protein, the net amount of the Dlmo product is reduced to approximately the wild-type level, resulting in normal wing morphology.

The anatomical cause for the erect wings in hdp-a mutants is unknown at this time. Mutations in many genes in Drosophila affect wing posture. Most of them affect either components of the wing muscles or their innervation (reviewed in BERNSTEIN et al. 1993 ). Dlmo may be required for either function because LIM proteins are often expressed in the nervous system and muscles (CURTISS and HEILIG 1998 ; DAWID et al. 1998 ; JURATA and GILL 1998 ). Indeed, we find that in the embryo, Dlmo expression is restricted primarily to the CNS (M. SHORESH and D. SEGAL, unpublished observations).

The mutant phenotypes of lesions in the Dlmo gene involving wing margin defects and abnormal wing posture, as well as the limited information we have about the spatial distribution of its transcript in the wing imaginal discs and embryonic CNS, suggest that this LMO protein participates in diverse processes during development and differentiation of Drosophila. In this respect, it resembles its vertebrate homologs, which are expressed in the embryonic CNS and in the hematopoietic system and are involved in a multitude of processes during animal development (CURTISS and HEILIG 1998 ; DAWID et al. 1998 ; JURATA and GILL 1998 ).

Dlmo may participate in a DNA-binding complex regulating wing development:
Overexpression of Dlmo in certain imaginal wing disc cells causes scalloping. The interactions we observe between ap and Chip mutations and their analogy to the interactions between their mammalian counterparts enable us to propose a model to explain the role of Dlmo in wing development. Mutations in ap and Chip cause varying degrees of wing scalloping. Studies by different groups have demonstrated that the NLI protein, of which Chip is a homolog, is capable of specifically binding to various LIM homeodomain proteins, and as a dimer facilitates the formation of heteromeric complexes between LIM-containing transcription factors (for reviews see CURTISS and HEILIG 1998 ; DAWID et al. 1998 ; JURATA and GILL 1998 ). Likewise, in Drosophila, the Chip-encoded protein has been shown to bind in vitro the apterous LIM homeodomain transcription factor (MORCILLO et al. 1997 ). This binding is biologically significant because flies transheterozygous for ap and Chip mutations display marked augmentation of wing scalloping (MORCILLO et al. 1997 ; results presented in this article).

LMO2, the vertebrate homolog of Dlmo, has been recently found to serve as a bridging molecule, assembling a DNA-binding complex that includes various transcription factors (CURTISS and HEILIG 1998 ; DAWID et al. 1998 ; JURATA and GILL 1998 ). For example, LMO2 is an obligatory component in the formation of an oligomeric complex that includes the zinc finger protein GATA-1 and the Tal1 and E47 basic helix-loop-helix proteins. LMO2 facilitates the binding of this complex to DNA that contains both recognition motifs for binding of bHLH proteins and GATA-1–binding sites (VISVADER et al. 1997 ; WADMAN et al. 1997 ). Moreover, NLI is a partner in these DNA-binding complexes and mediates their assembly. Thus, LMO2 and NLI may be general bridging modules between various transcription factors that often contain LIM domains. This could also be the case in Drosophila, where DLMO and CHIP may be responsible for the assembly of various transcription complexes. Indeed, we demonstrated that mutations in Bx and either ap or Chip interact to augment wing scalloping. In view of both the binding of the AP (LIM homeodomain transcription factor) and CHIP (NLI homolog) proteins to each other (MORCILLO et al. 1997 ), as well as the role of LMO2 (the vertebrate homolog of Dlmo) in the assembly of transcription factor complexes, we propose that in Drosophila, the AP, CHIP, and DLMO proteins form a similar DNA-binding complex. The identity of the genes regulated by this complex remains to be elucidated. However, some of them may be genes identified downstream of ap in the regulatory hierarchy of wing margin formation (DIAZ-BENJUMEA and COHEN 1993 ; BLAIR et al. 1994 ).

In mammals, GATA-1, Tal-1, LMO2, and NLI have been shown to coexpress in erythroid cells (reviewed in CURTISS and HEILIG 1998 ; DAWID et al. 1998 ; JURATA and GILL 1998 ). Loss-of-function mutations in the former three have a similar phenotype, namely failure of hematopoietic development that leads to lethality (OSADA et al. 1995 ). Recent studies on oligomeric complexes involving the mammalian LMO2 and NLI have demonstrated that depending on the conditions, they may function either to facilitate or inhibit formation of the corresponding DNA-binding complexes. The balance of the different constituent molecules in these complexes appears to be important for the inhibitory or synergistic effect, as well as the presence or absence of other interacting proteins (BREEN et al. 1998 ; JURATA et al. 1998 ). Likewise, we find that mutations in ap, Chip, and Bx cause a similar phenotype, abnormal wing morphology. Our observations on the wing phenotypes of flies carrying different doses of these three genes suggest that their interaction is sensitive to the balance between the corresponding proteins. Overexpression or underexpression of ap augments the wing scalloping caused by overexpression of Dlmo. This underscores the importance of the fine tuning of the relative levels of these proteins. Overexpression of Lhx2, NLI, or LMO2 in erythroid cells maintains their proliferative undifferentiated state (WU et al. 1994 ; VISVADER et al. 1997 ). We suggest that in analogy, overexpression of Dlmo in certain cells of the wing imaginal discs disrupts the presumptive DNA-binding complex, leading to abnormal wing development.

Overexpression of LMO genes is oncogenic:
Overexpression of two human LMO genes, LMO1 and LMO2, causes neoplasia. LMO1 is disrupted by a t(11;14)(p15; q11) T cell translocation involving the TCR locus in T cell acute lymphoblastic leukemia, (T-ALL; BOEHM et al. 1988 ). LMO2 is associated with another breakpoint (11p13) that involves a distinct TCR locus in T-ALL (ROYER-POKORA et al. 1991 ). This breakpoint disrupts the chromosome 5' to the coding region of LMO2 and was proposed to cause deregulation of the gene leading to leukemia (BOEHM et al. 1991A , BOEHM et al. 1991B ). In both cases, the result of the translocation is overexpression of the LMO product. Transgenic mice targeting expression of LMO1 and LMO2 to T cells develop overt malignancies, as in T-ALL. The neoplasia likely results from disruption of the stoichiometry between the LMO proteins and the transcription factors they recruit to functional DNA-binding complexes. Interestingly, the frequency of tumor incidence was proportional to the amount of transgene expression in T cells of the transgenic mice (MCGUIRE et al. 1992 ; LARSON et al. 1994 ). Overexpression of their Drosophila homolog, Dlmo, which leads to abnormal wing development, may serve as an in vivo model for T-ALL that is amenable to genetic manipulation.

	FOOTNOTES

¹ The mutant designated hdp-a¹⁰² in FlyBase is not uncovered by the chromosomal deletion Df(1)N19; hence, is not an allele of hdp-a and should be renamed (data not shown).

	ACKNOWLEDGMENTS

We are grateful to BILL ENGELS, DALE DORSETT, and the Drosophila Stock Centers at Bloomington and Bowling Green for fly strains. We thank WILLIAM MATTOX and BRIGITTE ROYER-POKORA for providing clones and for sharing unpublished results. KEVIN O'HARE kindly provided the genomic sequence around the P element at 17C2-3 in ₂, and NICK BROWN provided the cDNA library. We are indebted to DALE DORSETT, PATRICK MORCILLO, and members of our lab for stimulating discussions. This work was supported in part by a grant from The Israel Science Foundation to D.S.

Manuscript received April 13, 1998; Accepted for publication June 12, 1998.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AGULNICK, A. D., M. TAIRA, J. J. BREEN, T. TANAKA, and I. B. DAWID, 1996  Interactions of the LIM domain binding factor Ldb1 with LIM homeodomain proteins. Nature 384:270-272[Medline].

ABLER, M. L. and P. J. GREEN, 1996  Control of mRNA stability in higher plants. Plant Mol. Biol. 32:63-78[Medline].

ALTARATZ, M., S. W. APPLEBAUM, D. S. RICHARD, L. I. GILBERT, and D. SEGAL, 1991  Regulation of juvenile hormone synthesis in wild-type and apterous mutant Drosophila. Mol. Cell. Endocrinol. 81:205-216[Medline].

ARBER, S. and P. CARONI, 1996  Specificity of single LIM motifs in targeting and LIM/LIM interaction in situ. Genes Dev. 10:289-300[Abstract/Free Full Text].

BACH, I., S. J. RHODES, R. V. PEARSE, T. HEINZEL, and B. GLOSS, 1995  P-LIM, a LIM homeodomain factor, is expressed during pituitary organ and cell commitment and synergizes with Pit-1. Proc. Natl. Acad. Sci. USA 92:720-772.

BALTZ, R., J.-L. EVRARD, C. DOMON, and A. STEINMETZ, 1992  A LIM motif is present in a pollen-specific protein. Plant Cell 4:1465-1466[Free Full Text].

BERNSTEIN, S. I., P. T. O'DONNELL, and R. M. CRIPPS, 1993  Molecular genetic analysis of muscle development, structure and function in Drosophila. Int. Rev. Cytol. 143:63-153[Medline].

BLAIR, S. S., D. L. BROWER, J. B. THOMAS, and M. ZAVORTNIK, 1994  The role of apterous in the control of dorsoventral compartmentalization and PS integrin gene expression in the developing wing of Drosophila. Development 120:1805-1815[Abstract].

BOEHM, T., R. BAER, I. LAVENIR, A. FORSTER, and J. J. WATERS et al., 1988  The mechanism of chromosomal translocation t(11:14) involving the T-cell acceptor C delta locus on human chromosome 14q11 and a transcribed region of chromosome 11p15. EMBO J. 7:385-394[Medline].

BOEHM, T., L. FORONI, Y. KANEKO, M. F. PERUTZ, and T. H. RABBITTS, 1991a  The rhombotin family of cysteine rich LIM domain oncogene: distinct members are involved in T-cell translocations to human chromosome 11p15 and 11p13. Proc. Natl. Acad. Sci USA 88:4367-4371[Abstract/Free Full Text].

BOEHM, T., M. G. SPILLANTINI, M. V. SOFRONIEW, and T. H. RABBITTS, 1991b  Developmentally regulated and tissue specific expression of mRNAs encoding the two alternative forms of the LIM domain oncogene rhombotin: evidence for thymus expression. Oncogene 6:695-703[Medline].

BOURGOUIN, C., S. E. LUNDGREN, and J. B. THOMAS, 1992  apterous is a Drosophila LIM domain gene required for the development of a subset of embryonic muscles. Neuron 9:549-561[Medline].

BREEN, J. J., A. D. AGULNICK, H. WESTPHAL, and I. B. DAWID, 1998  Interactions between LIM domains and the LIM domain-binding protein Ldb1. J. Biol. Chem. 273:4712-4717[Abstract/Free Full Text].

BROWN, M. C., J. A. PEROTA, and C. E. TURNER, 1996  Identification of LIM3 as the principal determinant of Paxillin focal adhesion localization and characterization of a novel motif on Paxillin directing vinculin and focal adhesion kinase binding. J. Cell Biol. 135:1109-1123[Abstract/Free Full Text].

CHEN, C. Y. A. and A. B. SHYU, 1995  AU-rich elements: characterization and importance in mRNA degradation. Trends Biochem. Sci. 20:465-470[Medline].

COHEN, B., M. E. MCGUFFIN, C. PFEIFLE, D. SEGAL, and M. COHEN, 1992  apterous, a gene required for imaginal disc development in Drosophila, encodes a member of the LIM family of developmental regulatory proteins. Genes Dev. 6:715-729[Abstract/Free Full Text].

CRAWFORD, A. W. and M. C. BECKERLE, 1992  Purification and characterization of Zyxin, an 82000-dalton component of adherens junctions. J. Biol. Chem. 266:5847-5853[Abstract/Free Full Text].

CURTISS, J. and J. S. HEILIG, 1997  Arrowhead encodes a LIM homeodomain protein that distinguishes subsets of Drosophila imaginal cells. Dev. Biol. 190:129-141[Medline].

CURTISS, J. and J. S. HEILIG, 1998  DeLIMiting development. BioEssays 20:58-69[Medline].

DAWID, I. B., J. J. BREEN, and R. TOYAMA, 1998  LIM domains: multiple roles as adaptors and functional modifiers in protein interactions. Trends Genet. 14:156-162[Medline].

DIAZ-BENJUMEA, F. J. and S. M. COHEN, 1993  Interaction between dorsal and ventral cells in the imaginal disc directs wing development in Drosophila. Cell 75:741-752[Medline].

DORSETT, D., 1990  Potentiation of a polyadenylation site by a downstream protein-DNA interaction. Proc. Natl. Acad. Sci. USA 87:4373-4377[Abstract/Free Full Text].

ENGELS, W. R. and C. R. PRESTON, 1984  Formation of chromosome rearrangements by P factors in Drosophila. Genetics 107:657-678[Abstract/Free Full Text].

FEUERSTEIN, R., X. WANG, D. SONG, N. E. COOK, and S. A. LEIBHABER, 1994  The LIM/double zinc finger motif functions as a protein dimerization domain. Proc. Natl. Acad. Sci. USA 91:10655-10659[Abstract/Free Full Text].

FORONI, L., T. BOEHM, L. WHITE, A. FORSTER, and P. SHERRINGTON et al., 1992  The Rhombotin gene family encodes related LIM-domain proteins whose differing expression suggests multiple roles in mouse development. J. Mol. Biol. 226:747-761[Medline].

FREYD, G., S. K. KIM, and H. R. HORVITZ, 1990  Novel cysteine-rich motif and homeodomain in the product of Caenorhabditis elegans cell lineage gene lin-11. Nature 344:876-879[Medline].

JURATA, L. W. and G. M. GILL, 1998  Structure and function of LIM domains. Curr. Top. Microbiol. Immunol. 228:75-113[Medline].

JURATA, L. W., S. L. PFFAF, and G. M. GILL, 1998  The nuclear LIM domain interactor NLI mediates homo- and heterodimerization of LIM domain transcription factors. J. Biol. Chem. 273:3152-3157[Abstract/Free Full Text].

KARLSSON, O., S. THOR, Y. NORBERG, H. OHLSSON, and E. EDLUND, 1990  Insulin gene enhancer binding protein Isl-1 is a member of a novel class of proteins containing both a homeodomain and a Cys-His domain. Nature 344:879-882[Medline].

LAI, E. C. and J. W. POSAKONY, 1997  The Bearded box, a novel 3'UTR sequence motif, mediates negative post-transcriptional regulation of Bearded and Enhancer of Split complex gene expression. Development 124:4847-4856[Abstract].

LARSON, R. C., P. FISCH, T. A. LARSON, I. LAVENIR, and T. LANGFORD et al., 1994  T cell tumors of disparate phenotype in mice transgenic for RBTN-2. Oncogene 9:3675-3681[Medline].

LEVITEN, M. W., E. C. LAI, and J. W. POSAKONY, 1997  The Drosophila gene Bearded encodes a novel small protein and shares 3'UTR sequence motifs with multiple Enhancer of Split Complex genes. Development 124:4039-4051[Abstract].

LIFSCHYTZ, L. and M. M. GREEN, 1979  Genetic identification of dominant overproducing mutations: the Beadex gene. Mol. Gen. Genet. 171:153-159[Medline].

LINDSLEY, D. L., and G. G. ZIMM, 1992 The Genome of Drosophila melanogaster. Academic Press, San Diego.

LUNDGREN, S. E., C. A. CALLAHAN, S. THOR, and J. B. THOMAS, 1995  Control of neuronal pathway selection by the Drosophila LIM homeodomain gene apterous. Development 121:1769-1773[Abstract].

MARIOL, M. C., T. PREAT, and B. LIMBOURRG-BOUCHON, 1987  Molecular cloning of fused, a gene required for normal segmentation in the Drosophila melanogaster embryo. Mol. Cell. Biol. 7:3244-3251[Abstract/Free Full Text].

MATTOX, W. W. and N. DAVIDSON, 1984  Isolation and characterization of the Beadex locus of Drosophila melanogaster: a putative cis-acting negative regulatory element for the heldup-a gene. Mol. Cell. Biol. 4:1343-1353[Abstract/Free Full Text].

MCGUIRE, E. A., R. D. HOCKETT, K. M. POLLOCK, M. F. BARTHOLDI, and S. J. O'BREIN et al., 1989  The t(11:14)(p15:q11) in a T-cell acute lymphoblastic leukemia cell line activates multiple transcripts, including Ttg-1, a gene encoding a potential zinc finger protein. Mol. Cell. Biol. 9:2124-2132[Abstract/Free Full Text].

MCGUIRE, E. A., C. E. RINTOUL, G. M. SELAR, and S. J. KORSMEYER, 1992  Thymic expression of Ttg-1 transgenic mice result in T-cell acute lymphoblastic leukemia/lymphoma. Mol. Cell. Biol. 12:4186-4196[Abstract/Free Full Text].

MORCILLO, P., C. ROSEN, M. K. BAYLIES, and D. DORSETT, 1997  Chip, a widely expressed chromosomal protein required for segmentation and activity of a remote wing margin enhancer in Drosophila. Genes Dev. 11:2729-2740[Abstract/Free Full Text].

O'HARE, K. and G. M. RUBIN, 1983  Structures of P transposable elements and their sites of insertion and excision in the Drosophila melanogaster genome. Cell 17:415-427.

OSADA, H., G. GRUTZ, H. ALEXON, A. FORSTER, and T. H. RABBITTS, 1995  Association of erythroid transcription factors: complexes involving the LIM protein RBTN2 and the zinc-finger protein GATA-1. Proc. Natl. Acad. Sci. USA 92:9585-9589[Abstract/Free Full Text].

PFAFF, S. L., M. MENDELSON, C. L. STEWART, T. EDLUND, and T. M. JESSELL, 1996  Requirement for LIM homeobox gene Isl1 in motor neuron generation reveals a motor neuron-dependent step in interneuron differentiation. Cell 84:309-320[Medline].

PORTER, F. D., J. DRAGO, Y. XU, S. S. CHEEMA, and C. WASSIF et al., 1997  Lhx2, a LIM homeodomain gene, is required for eye, forebrain, and definitive erythrocyte development. Development 124:2935-2944[Abstract].

RINGO, J., R. WERCZBERGER, M. ALTARATZ, and D. SEGAL, 1991  Female sexual receptivity and juvenile hormone synthesis are defective in mutants of the apterous gene of Drosophila melanogaster. Behav. Genet. 21:453-469[Medline].

ROSS, J., 1995  mRNA stability in mammalian cells. Microbiol. Rev. 59:423-450[Abstract/Free Full Text].

ROYER-POKORA, B., U. LOOS, and W. D. LUDWIG, 1991  TTG-2, a new gene encoding a cysteine-rich protein with the LIM motif, is overexpressed in acute T-cell leukemia with the t(11:14)(p13:q11). Oncogene 6:1887-1893[Medline].

SADLER, I., A. W. CRAUFORD, J. W. MICHELSEN, and M. C. BECKERLE, 1992  Zyxin and cCRP: two interactive LIM domain proteins associated with the cytoskeleton. J. Cell Biol. 119:1573-1587[Abstract/Free Full Text].

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

SANCHEZ-GARCIA, I. and T. H. RABBITTS, 1994  The LIM domain: a new structural motif found in zinc-finger-like proteins. Trends Genet. 10:315-320[Medline].

SANCHEZ-GARCIA, I., H. ALEXON, and T. H. RABBITTS, 1995  Functional diversity of LIM proteins: amino-terminal activation domains in the oncogenic proteins RBTN1 and RBTN2. Oncogene 10:1301-1306[Medline].

SCHMEICHEL, K. L. and M. C. BECKERLE, 1994  The LIM domain is a modular protein-binding interface. Cell 79:211-219[Medline].

SIMMONS, M. J., J. D. RAYMOND, N. A. JOHNSON, and T. M. FAHEY, 1984  A comparison of mutation rates for specific loci and chromosome regions in dysgenic hybrid males of Drosophila melanogaster. Genetics 106:85-94[Abstract/Free Full Text].

SMITH, P. A. and V. G. CORCES, 1991  Drosophila transposable elements: mechanisms of mutagenesis and interactions with the host genome. Adv. Genet. 29:229-300[Medline].

STRONACH, B. E., S. E. SIEGRIST, and M. C. BECKERLE, 1996  Two muscle-specific LIM proteins in Drosophila. J. Cell Biol. 134:1179-1195[Abstract/Free Full Text].

TAIRA, M., H. OTANI, J. P. SAINT-JEANNET, and I. B. DAWID, 1994  Role of the LIM class homeodomain protein Xlim-1 in neural and muscle induction by the Spemann organizer in Xenopus. Nature 372:677-679[Medline].

TAIRA, M., J.-L. EVRARD, A. STEINMETZ, and I. B. DAWID, 1995  Classification of LIM proteins. Trends Genet. 11:431-432[Medline].

TAUTZ, D. and C. PFEIFLE, 1989  A non-radioactive in situ hybridization method for the localization of specific RNAs in Drosophila embryos reveals a translational control of the segmentation gene hunchback. Chromosoma 98:81-85[Medline].

THOMAS, U., F. JONSSON, S. A. SPEICHER, and E. KNUST, 1995  Phenotypic and molecular characterization of Ser^D, a mutant allele of the Drosophila gene Serrate. Genetics 139:203-213[Abstract].

THOR, S. and J. B. THOMAS, 1997  The Drosophila islet gene governs axon pathfinding and neurotransmitter identity. Neuron 18:397-409[Medline].

VISVADER, J. E., X. MAO, Y. FUJIWARA, K. HAHM, and S. H. ORKIN, 1997  The LIM domain binding protein Ldb1 and its partner LMO2 act as negative regulators of erythroid differentiation. Proc. Natl. Acad. Sci. USA 94:13707-13712[Abstract/Free Full Text].

WADMAN, I. S., H. OSADA, G. G. GRUTZ, A. D. AGULNICK, and H. WESTPHAL et al., 1997  The LIN-only protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding complex which includes the Tal1, E47, GATA-1 and Ldb1/NLI proteins. EMBO J. 16:3145-3157[Medline].

WAY, J. C. and M. CHALFIE, 1988  mec-3, a homeobox-containing gene that specifies differentiation of the touch receptor neurons in C. elegans. Cell 54:5-16[Medline].

WU, H.-K., H. H. Q. HENG, D. P. SIDEROVSKI, W. F. DONG, and Y. OKUNO et al., 1994  Consistent high level expression of a human LIM/HOX gene, hLH-2 in chronic myelogenous leukemia and chromosomal localization to 9q33-341. Blood 84(Suppl.):295a.

ZHU, T. H., J. BODEM, E. KEPPEL, R. PARO, and B. ROYER-POKORA, 1995  A single ancestral gene of the human LIM domain oncogene family LMO in Drosophila: characterization of the Drosophila Dlmo gene. Oncogene 11:1283-1290[Medline].

This article has been cited by other articles:

				S. Zenvirt, Y. Nevo-Caspi, S. Rencus-Lazar, and D. Segal Drosophila LIM-Only Is a Positive Regulator of Transcription During Thoracic Bristle Development Genetics, August 1, 2008; 179(4): 1989 - 1999. [Abstract] [Full Text] [PDF]

				F. Bejarano, C. M. Luque, H. Herranz, G. Sorrosal, N. Rafel, T. T. Pham, and M. Milan A Gain-of-Function Suppressor Screen for Genes Involved in Dorsal-Ventral Boundary Formation in the Drosophila Wing Genetics, January 1, 2008; 178(1): 307 - 323. [Abstract] [Full Text] [PDF]

				J. I. Pueyo and J. P. Couso Chip-mediated partnerships of the homeodomain proteins Bar and Aristaless with the LIM-HOM proteins Apterous and Lim1 regulate distal leg development Development, July 1, 2004; 131(13): 3107 - 3120. [Abstract] [Full Text] [PDF]

				D. J. van Meyel, J. B. Thomas, and A. D. Agulnick Ssdp proteins bind to LIM-interacting co-factors and regulate the activity of LIM-homeodomain protein complexes in vivo Development, May 1, 2003; 130(9): 1915 - 1925. [Abstract] [Full Text] [PDF]

				E. C. Lai Drosophila Tufted Is a Gain-of-Function Allele of the Proneural Gene amos Genetics, April 1, 2003; 163(4): 1413 - 1425. [Abstract] [Full Text] [PDF]

				L. Chen, D. Segal, N. A. Hukriede, A. V. Podtelejnikov, D. Bayarsaihan, J. A. Kennison, V. V. Ogryzko, I. B. Dawid, and H. Westphal Ssdp proteins interact with the LIM-domain-binding protein Ldb1 to regulate development PNAS, October 29, 2002; 99(22): 14320 - 14325. [Abstract] [Full Text] [PDF]

				U. Weihe, M. Milan, and S. M. Cohen Regulation of Apterous activity in Drosophila wing development Development, November 15, 2001; 128(22): 4615 - 4622. [Abstract] [Full Text] [PDF]

				J. Parsch, J. A. Russell, I. Beerman, D. L. Hartl, and W. Stephan Deletion of a Conserved Regulatory Element in the Drosophila Adh Gene Leads to Increased Alcohol Dehydrogenase Activity but Also Delays Development Genetics, September 1, 2000; 156(1): 219 - 227. [Abstract] [Full Text]

				S. Abdelilah-Seyfried, Y.-M. Chan, C. Zeng, N. J. Justice, S. Younger-Shepherd, L. E. Sharp, S. Barbel, S. A. Meadows, L. Y. Jan, and Y. N. Jan A Gain-of-Function Screen for Genes That Affect the Development of the Drosophila Adult External Sensory Organ Genetics, June 1, 2000; 155(2): 733 - 752. [Abstract] [Full Text]

				M Milan and S. Cohen Temporal regulation of apterous activity during development of the Drosophila wing Development, January 7, 2000; 127(14): 3069 - 3078. [Abstract] [PDF]

				D. van Meyel, D. O'Keefe, S Thor, L. Jurata, G. Gill, and J. Thomas Chip is an essential cofactor for apterous in the regulation of axon guidance in Drosophila Development, January 5, 2000; 127(9): 1823 - 1831. [Abstract] [PDF]

				R. A. Rollins, P. Morcillo, and D. Dorsett Nipped-B, a Drosophila Homologue of Chromosomal Adherins, Participates in Activation by Remote Enhancers in the cut and Ultrabithorax Genes Genetics, June 1, 1999; 152(2): 577 - 593. [Abstract] [Full Text]

				E. Torigoi, I. M. Bennani-Baiti, C. Rosen, K. Gonzalez, P. Morcillo, M. Ptashne, and D. Dorsett Chip interacts with diverse homeodomain proteins and potentiates Bicoid activity in vivo PNAS, March 14, 2000; 97(6): 2686 - 2691. [Abstract] [Full Text] [PDF]