BIOLOGICAL OVERVIEW

Legs and antennae are considered to be homologous appendages. The fundamental patterning mechanisms that organize spatial pattern are conserved, yet appendages with very different morphology develop. The distal antenna (dan) and distal antenna-related (danr) genes encode novel 'pipsqueak' motif nuclear proteins that probably function as DNA binding proteins serving as sequence-specific transcription factors but may serve instead as more general chromatin modification factors. dan and danr are expressed in the presumptive distal antenna, but not in the leg imaginal disc. Ectopic expression of dan or danr causes partial transformation of distal leg structure toward antennal identity. Mutants that remove dan and danr activity cause partial transformation of antenna toward leg identity. Therefore it is suggested that dan and danr contribute to differentiation of antenna-specific characteristics. Antenna-specific expression of dan and danr depends on a regulatory hierarchy involving homothorax and Distal-less, as well as cut and spineless. It is proposed that dan and danr are effector genes that act downstream of these genes to control differentiation of distal antennal structures (Emerald, 2003).

dan and danr were identified in a large-scale modular misexpression screen of ~8500 EPg elements, as insertions that caused abnormalities in wing development when expressed under sd^Gal4 control and small rough eyes when expressed under ey^Gal4 control. Expression of dan using EPg J3-220, under control of Dll^Gal4 causes transformation of distal leg structures toward distal antennal identity. The claws, found on the tip of the fifth tarsal segment, are transformed into the distal-most part of the antenna, the arista. In addition, there is some overall morphological abnormality of the tarsal region. Expression of danr from EPg35635 using Dll^Gal4 causes loss of the claws but does not produce an overt transformation to arista (Emerald, 2003).

EPg J3-220 and EPg 35635 lie approximately 45 kb apart on chromosome 3R, 263 bp and 39 bp upstream of the predicted genes, CG11849 and CG13651. To verify that the predicted genes tagged by the EP-element insertions were responsible for the observed overexpression phenotypes, UAS-dan and UAS-danr transgenic strains were generated. Six independent UAS-dan transformants and five independent UAS-danr transformants were tested and found to be lethal when expressed with the Dll^Gal4 driver. However when expressed using dpp^Gal4, UAS-dan and UAS-danr showed distal leg to antenna transformation. Molecular markers of antennal identity were also examined in the imaginal discs. The zinc-finger protein Spalt is expressed in antenna, but not in leg discs. Ectopic expression of dan can induce limited expression of Spalt in the leg disc, consistent with the observed transformation toward antennal identity. These observations suggested a role for dan and danr in specification of the identity of distal antennal structures (Emerald, 2003).

Dan plays an important role in specification of antennal identity downstream of spineless (ss), but rescue of the spineless mutation by Dan suggests that there may be additional genes acting downstream from ss to specify antennal identity. spineless mutants lead to ectopic expression of Antennapedia and concomitant loss of Dan/Danr expression and cause a strong phenotypic transformation of distal A3 and arista to tarsus. To determine whether morphological transformation depends on loss of Dan/Danr, use was made of Gal4 to direct Dan expression in the ss mutant discs. ptc^Gal4 directed expression of Dan causes strong suppression of the arista-to-tarsus transformation in the ss mutant antenna. ptc^Gal4 is expressed in a stripe of cells adjacent to the AP boundary in the antenna region of the disc. Dan expression does not repress ectopic expression of Antp in the ptc^Gal4 stripe of the mutant discs. This suggests that Dan can direct antennal differentiation in the presence of Antp, and overcome the ability of Antp to cause transformation to tarsus. Remarkably, this transformation can affect the entire distal arista, even though ectopic Dan is expressed in only a subset of Antp-expressing cells. These observations suggest that Dan plays an important role in specification of antennal identity (Emerald, 2003).

Both Dan and Danr contribute to specification of antennal identity. danr single mutants produce a partial transformation of arista to tarsus. A similar, but slightly stronger phenotype results when both dan and danr are deleted. Reduced dan activity in the dan^ems3 mutants or reduced Dan expression caused by RNA interference produces a milder version of the same phenotype (Emerald, 2003).

An additional line of evidence to indicate that both genes contribute to distal antenna identity comes from examining genetic interactions with spineless^aristapedia. ss^a mutants lose Dan/Danr expression and express Antennapedia ectopically in the antenna disc. Restoring Dan expression is able to partially suppress the transformation to antenna, implicating Dan as an effector of ss^a function. The consequences of removing one copy each of Dan and Danr was examined in a ss mutant background. The spineless^114.4 allele shows a mild transformation of the basal capsule of the arista when heterozygous, suggesting that the reduced level of ss activity in this allele is not sufficient to support normal development. Removing one copy of danr using the danr^ex35 deletion in this background causes a modest increase in the size of the basal capsule and in the number of ectopic bristles. The dan danr^ex56 deletion causes a stronger phenotype, with the basal capsule adopting a two-segment structure with multiple bracted bristles and obvious tarsal morphology. Flies heterozygous for the dan danr^ex56 deletion are morphologically normal. Thus, reduction of both Dan and Danr gene dose leads to a more severe phenotype under conditions where ss activity is limiting. Even more extreme arista transformation phenotypes are observed when one copy of ss is removed in animals homozygous for the dan danr^ex56 deletion (Emerald, 2003).

Genetic interactions were also observed with Dll. Double heterozygotes for danr and Dll or dan danr and Dll show ectopic tarsal bristles in the basal capsule of the arista (in this case the phenotypes are similar in strength). The ss/dan danr double heterozygotes produces a more complete transformation phenotype than the dan danr homozygous mutants. This raises the possibility that there may be additional genes acting downstream from ss to specify antennal identity (Emerald, 2003).

Insect antennae develop in the absence of input from HOM-C genes. In the anterior head of Drosophila where Antennapedia-complex and Bithorax complex genes are not expressed, expression of hth and Dll overlaps and promotes antenna development. One consequence of overlapping expression of Dll and Hth is sustained expression of ss in the distal antenna. ss is expressed in the leg and antenna discs in second instar, but its expression is not maintained in the leg (Emerald, 2003).

Loss of Hth activity has been shown to cause transformation of arista to tarsus, presumably because of loss of ss. It has been suggested that uniform expression of Hth in second and early third instar antennae might be responsible for its role in specification of distal antenna identity. However, the results of this study indicate that Hth can have a non-autonomous effect on the expression of Dan in the antenna. Hth-expressing cells sort out from the distal part of the leg. Nonetheless they are able to induce Dan expression in cells that remain integrated in the distal leg. This observation is best explained by a non-autonomous induction of Dan in response to a signal from Hth-expressing cells. Responsiveness to this signal apparently requires Dll, which limits it to the distal region. These effects are presumably mediated by regulation of ss, which is required for Dan and Danr expression. These observations provide an explanation for the apparently non-autonomous role of Hth together with Dll in the distal antenna (Emerald, 2003).

ss is also required to induce Dan and Danr and to repress Antp expression. Repression of Antennapedia may be mediated in part by repression of Cut. The findings described above implicate Dan and Danr as downstream effectors of ss that promote development of distal antennal structures. Remarkably, expression of Dan or Danr under Gal4 control can restore antenna development and prevent transformation of antenna to leg in the ss mutant, even when Antp is present. A striking feature of these results is that there appears to be non-autonomous activity. Transformation is blocked in cells expressing Dan and Danr, as well as in nearby cells that do not express these proteins. The identity of the genes responsible for these non-autonomous effects in antenna specification remains to be determined. In view of recent reports of non-autonomous effects of vein/EGFR signaling in development of distal leg pattern, it will be of interest to learn if there is a link to this pathway in the non-autonomous effects of Dan and Danr (Emerald, 2003).

What are the molecular functions of Dan and Danr? The 'pipsqueak motifs' found near the N terminus of Dan and Danr are most closely related to those in the 'transposase group' of pipsqueak motif proteins, which includes the Pogo transposase and human centromere protein B (CENP-B). The pipsqueak motifs of both Pogo and CENP-B are DNA-binding domains. NMR-spectroscopy of the CENP-B pipsqueak motif demonstrates that it has a helix-turn-helix structure. Interestingly, the glutamate residue that has been transformed to a lysine in the pipsqueak motif of the dan^ems3 allele lies at the same position as the arginine that is required for DNA binding by the Pogo pipsqueak motif. Thus, the dan^ems3 mutation may interfere with Dan function by abrogating the ability of Dan to bind DNA, or changing its specificity. Although these data suggest that Dan/Danr bind to DNA, it remains unclear whether they act as sequence-specific transcription factors or more general chromatin modification factors. Biochemical analysis of Dan/Danr and identification of interacting proteins will be required to address this issue in detail (Emerald, 2003).

PROTEIN STRUCTURE

Amino Acids - 678 & 419

Structural Domains

The predicted proteins encoded by dan and danr are similar, showing 25% identity overall. This similarity extends through the entire sequence. In addition, Dan has a C-terminal extension of more than 200 amino acids. The most conserved region is a 64 amino acid sequence beginning with the N terminus, where Dan and Danr share 92% identity. Within this region, both Dan and Danr contain the newly identified 'pipsqueak' motif (Siegmund, 2002), a helix-turn-helix structure that is likely to be involved in DNA binding. Outside the pipsqueak motif, the Dan and Danr proteins contain no regions of significant sequence similarity to other known proteins but show short blocks of strong similarity to one another (Emerald, 2003).

date revised: 10 March 2003

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.