RT-PCR showed that Abs transcripts are found throughout all developmental stages, but no ABS RNA could be detected by in situ hybridizations or on Northern blots, nor is Abstrakt protein detectable in normal embryos stained with antibodies against Abstrakt. Thus, ABS RNA and protein does not appear to be very abundant. When Abstrakt is expressed at high levels using the GAL4/UAS ectopic expression system, a clear signal is seen by immunofluorescence in the areas in which the UAS-abs transgene is expressed. In older embryos (> stage 12), the signal is restricted to the nuclei. In younger embryos, the subcellular localization of Abstrakt varies between predominantly nuclear localization, homogeneous distribution throughout the cell, and enrichment in the cytoplasm. No correlation between Abstrakt localization and specific stages of the cell cycle has been observed (Irion, 1999).
Embryos homozygous for the non-conditional abstrakt alleles or the deficiency Df(3R)231-5 develop without apparent gross defects, but fail to hatch. To assay functions at other stages, the temperature-sensitive allele abs14B was used. abs14B animals were transferred from room temperature to 32°C at various stages of development. Shifts to restrictive temperature lead to lethality at all stages of the life cycle. Although embryos homozygous for the non-conditional alleles develop until hatching, abs14B embryos cease to develop and show gross morphological defects at restrictive temperature. Thus, abs is essential for embryogenesis. The embryos homozygous for the non-conditional alleles must therefore use the wild-type gene product provided by their heterozygous mothers to complete embryogenesis in the absence of functional zygotic Abstrakt. A wild-type paternal copy of the gene is sufficient to rescue abs14B embryos if they are shifted to restrictive temperature after gastrulation. The temperature-sensitive phenotype is identical in progeny from mothers homozygous for abs14B allele and mothers heterozygous for abs14B and a deletion of the gene. No dominant effects of the abs14B allele were seen in animals heterozygous for the abs14B and one or more wild-type copies of the gene (Irion, 1999).
It seemed possible that Abstrakt, as a DEAD-box protein, might be involved in general RNA processing or metabolism, and that cells in the mutant animals eventually run out of essential components and commit apoptosis. When TUNEL assays were performed on abs14B mutants, however, no increased rate of apoptosis is seen. Instead, apoptosis is almost completely suppressed, and even the wild-type pattern of cells undergoing programmed cell death is not seen in mutant embryos (Irion, 1999).
To test at which level the apoptotic pathway is interrupted in abs mutants, the expressions of reaper (rpr), a nuclear regulator of apoptosis in Drosophila, and dredd, a target gene of reaper encoding a caspase were examined. Both reaper and dredd are expressed properly in abs14B mutant embryos. This shows that the apoptotic regulators are activated in the mutant and components of the apoptotic pathway are expressed, but the program is not completed (Irion, 1999).
Ectopic cell death induced by overexpression of reaper is also greatly suppressed in abs mutants. The amnioserosa is the only tissue in abs mutants in which extensive apoptosis was found after overexpressing reaper. Ectopic apoptosis induced by overexpression of a second regulator, hid, is not completely suppressed in abs mutants, consistent with previous findings that hid is a more potent inducer. Rather than being blocked, apoptosis in abs14B mutant ovaries occurs prematurely in follicle cells and nurse cells. This points to different modes of regulation of apoptosis during embryogenesis and oogenesis, in line with recent findings that reaper, hid and grim are not essential for programmed cell death of the nurse cells (Irion, 1999).
In Drosophila, both neural and muscle progenitors divide asymmetrically. In these cells the Inscuteable (Insc) protein complex coordinates cell polarity and spindle orientation. Abstrakt (Abs) is a DEAD-box protein that regulates aspects of cell polarity in oocytes and embryos. A conditional allele of abs was used to investigate its role in neural and muscle progenitor cell polarity. In neuroblasts loss of apical Insc crescents, failure in basal protein targeting, and defects in spindle orientation were observed. In the GMC4-2a cell loss of apical Insc crescents, defects in basal protein targeting, and equalization of sibling neuron fates are observed; muscle precursors show a similar equalization of sibling cell fates. These phenotypes resemble those of insc mutants; indeed, abs mutants show a striking loss of Insc protein levels but no change of insc RNA levels. Furthermore, the Abs protein physically interacts with insc RNA. These results demonstrate a novel role for Abs in the posttranscriptional regulation of insc expression, which is essential for proper cell polarity, spindle orientation, and the establishment of distinct sibling cell fates within embryonic neural and muscle progenitors (Irion, 2003).
Mitotic neuroblasts form an apical cortical protein complex containing Bazooka (the Drosophila homolog of nematode and mammalian Par-3), Par-6, atypical Protein Kinase C, Inscuteable (Insc), Partner-of-Inscuteable, and Gαi proteins . These apical proteins have three functions: to promote basal cell fate determinant localization, to orient the mitotic spindle along the apical/basal axis, and to promote the formation of an asymmetric spindle leading to the generation of daughters of unequal size. The basally localized determinants include Miranda (Mir) and Numb (Nb), which were used as markers in this study. Their basal localization ensures their preferential segregation into the basal daughter cell, called ganglion mother cell (GMC), during neuroblast division and ensures proper GMC fate specification (Irion, 2003).
To assay abs function, a temperature-sensitive allele was used in combination with a small deficiency uncovering the abs locus (abs14B/Df(3R)231-5, hereafter referred to as abs14B embryos) in which the maternally contributed Abs protein can be inactivated by a shift to the restrictive temperature. Wild-type embryos at the restrictive temperature and abs14B embryos at the permissive temperature show normal apical (Insc) and basal (Mir) cortical protein crescents in mitotic neuroblasts, as well as normal apicobasal orientation of the mitotic spindle. In contrast, abs14B embryos that are shifted to the restrictive temperature display severe defects in neuroblast polarity: Mir frequently shows uniform cortical distribution or occasionally accumulates as mispositioned lateral crescents. Furthermore, mitotic spindles occasionally fail to orient along the apical-basal axis. The similarity of these phenotypes and those that were previously reported for mutations affecting components of the Insc complex prompted the assaying of Insc protein localization in abs mutants. Interestingly, Insc protein is not detectable above background levels at the restrictive temperature in abs14B mutant neuroblasts, although apical Insc localization is not affected in abs14B embryos at the permissive temperature or in wild-type embryos at the restrictive temperature. It is concluded that loss of abs function leads to the loss of detectable Insc protein in neuroblasts and generates the phenotype previously seen in insc mutants. The simplest interpretation is that Insc expression and/or Insc protein stability is impaired in abs mutants, which leads to the observed defects in neuroblast asymmetric cell division (Irion, 2003).
To determine if abs has a more general role in regulating Insc levels and asymmetric cell division, asymmetric division of the ganglion mother cell GMC4-2a, which produces a pair of identified sibling neurons, RP2 and RP2sib, was examined. During wild-type GMC4-2a divisions, the mitotic spindle is apicobasally oriented; Insc is localized to the apical cortex, whereas Numb is localized as a basal cortical crescent and segregates preferentially into the more basal daughter cell, where it acts to downregulate Notch (N) signaling and induce the RP2 cell fate. The RP2 sibling cell does not inherit Numb, cannot downregulate N signaling, and adopts the secondary RP2sib fate. GMC4-2a and RP2 express the Even-skipped (Eve) transcription factor, but RP2sib does not; thus, there is only one Eve+ cell at the RP2 position in wild-type embryos. However, in abs14B embryos shifted to the restrictive temperature prior to GMC4-2a division, approximately 32% of the hemisegments had a duplicated Eve+ cell at the RP2 position. This phenotype was rarely seen either in control embryos (from a stock homozygous for abs14B along with two copies of a functional abs+ transgene, henceforth referred to as abs24:14B, that rescues the abs lethality subjected to the same temperature-shift regime or in abs14B embryos at the permissive temperature. The duplicated Eve+ cells are likely to be duplicated RP2 neurons because they express two additional markers (22C10 and Zfh1) for mature RP2 neurons (Irion, 2003).
To elucidate the origin of the duplicated RP2 neurons, anti-Eve staining was used to follow the development of the GMC4-2a lineage in wild-type and abs14B embryos. The results indicate that the extra RP2 neuron arises as the result of a transformation of the RP2sib to the RP2 cell fate (Irion, 2003).
Pon directly binds Numb protein and reflects the localization of Numb in all cells analyzed so far. In control abs24:14B and in wild-type embryos shifted to the restrictive temperature (33°C), Pon localizes as a basal crescent in mitotic GMC4-2a. In abs14B embryos subjected to the same temperature shift regime, approximately 50% (18/34) of metaphase GMC4-2a cells show cortical distribution, misplaced crescents, or weak basal crescents of Pon, and approximately 25% of the cells show no obvious Pon crescents. Hence, the symmetric segregation of Numb to both daughter cells in a proportion of the dividing GMC4-2a cells could account for the RP2 duplication phenotype seen in the abs14B embryos (Irion, 2003).
Because the abs phenotype is similar to the insc phenotype in both neuroblasts and GMC4-2a, Insc localization during the GMC4-2a cell division was investigated. In control embryos, Insc always forms an apical crescent in metaphase GMC4-2a cells. In contrast, at the restrictive temperature, the majority of the abs14B mutant GMC4-2a cells showed no clear apical crescents of Insc. Consistent with the finding that Insc localization is affected in abs14B, the duplicated RP2 cells seen at the restrictive temperature exhibit equal nuclear size, as is also seen in insc embryos but not in mutants that disrupt sibling cell fate choice at the postmitotic level (Irion, 2003).
The role of abs during embryonic muscle progenitor divisions was investigated. The muscle progenitor P15 divides asymmetrically to produce two daughter cells with distinct identities. Numb is asymmetrically localized in the dividing P15 and preferentially segregates to the daughter cell that will become the founder for the single Eve-positive muscle DA1; the sibling cell is Eve-negative. abs14B embryos subjected to a 45 min pulse at the restrictive temperature showed duplications of the Eve-positive DA1 in 34% (23/68) of the hemisegments. In the control abs24:14B embryos, 135/136 of the hemisegments showed a single Eve-positive DA1. Thus, abs is also required for the asymmetric division of some muscle progenitors (Irion, 2003).
The abs and insc mutant phenotypes in asymmetrically dividing cells are very similar, and abs mutants show a loss of Insc protein crescents in neuroblasts, in GMCs, and throughout the embryo. Thus, the abs phenotype can be most simply modeled as a defect in establishing or maintaining normal levels of apical Insc protein in all of these cell types. The loss of Insc crescents could be caused either by an overall decrease in the levels of Insc or by a failure to localize Insc correctly in these cells. In situ hybridization experiments revealed no reduction in insc RNA expression, so abs does not appear to regulate insc at the transcriptional level. Western blots were used to test whether the total amount of Insc protein was affected in abs mutant embryos. The Insc protein migrates as an approximately 100 kDa band. Wild-type and abs14B embryos were shifted to the restrictive temperature and analyzed after 0, 30, and 60 min. The levels of Insc protein decreased progressively in abs14B embryos until they were nearly undetectable at 60 min, whereas they remained constant or even increased (depending on the age distribution of embryos at the beginning of the experiment) in wild-type embryos. Other proteins remain constant, and several proteins can be translated de novo at the restrictive temperature, indicating that abs function is not generally required for protein synthesis. Together, these data indicate that the most upstream defect associated with a reduction in abs function is a reduction in the levels of the Insc protein (Irion, 2003).
If Abs indeed acts on asymmetric cell divisions by maintaining high levels of Insc, it should be possible to circumvent the requirement for Abs at least in part by raising Insc levels experimentally. To test this, the GAL4-UAS system was used to express high levels of insc within neuroblasts in embryos lacking functional Abs protein. This led to a marked rescue of the RP2 phenotype (Irion, 2003).
Because Abs is a DEAD-box protein, it seemed conceivable that it might exert its effect on Insc protein levels by a direct interaction with insc RNA. A yeast-three hybrid assay was used to test this. The assay is based on the interaction of the HIV-1 RNA binding protein Rev with RNA molecules containing a Rev responsive element (RRE). Rev is fused to the GAL4 DNA binding domain, whereas the putative RNA binding protein, in this case Abs, is fused to the activation domain. The two fusion proteins are then bridged by a hybrid RNA consisting of an RRE-containing sequence fused to the RNA to be tested, in this case insc RNA. insc RNA is clearly able to interact with Abs in this system. Both the full-length RNA and a construct lacking the 5' third of the RNA show an interaction. However, no single fragment of the 3' part of the RNA was found to be able to interact with Abs (Irion, 2003).
Thus Abs directly binds Insc mRNA in vitro; loss of Abs leads to lowered Insc protein levels but not lowered mRNA levels, and loss of Abs leads to a failure to properly localized cell fate determinants in at least three asymmetrically dividing cell types: neuroblasts, GMCs and muscle progenitors. It is concluded that Abs has a role in controlling cell polarity and asymmetric cell division in multiple cell types, in part through the posttranscriptional regulation of Insc levels (Irion, 2003).
Benz, J., Trachsel, H. and Baumann, U. (1999). Crystal structure of the ATPase domain of translation initiation factor 4A from Saccharomyces cerevisiae - the prototype of the DEAD box protein family. Structure 7: 671-679.
Irion, U. and Leptin, M. (1999). Developmental and cell biological functions of the Drosophila DEAD-box protein Abstrakt. Curr. Biol. (23): 1373-81.
Irion, U., et al. (2004). Abstrakt, a DEAD box protein, regulates Insc levels and asymmetric division of neural and mesodermal progenitors. Curr. Biol. 14: 138-144. 14738736
Schmucker, D., Jackle, H. and Gaul, U. (1997). Genetic analysis of the larval optic nerve projection in Drosophila. Development 124: 937-48.
Schmucker, D., Vorbruggen, G., Yeghiayan, P., Fan, H. Q., Jackle, H. and Gaul, U. (2000). The Drosophila gene abstrakt, required for visual system development, encodes a putative RNA helicase of the DEAD box protein family. Mech Dev. 91: 189-96.
Subramanya, H. S., Bird, L. E., Brannigan, J. A. and Wigley, D. B. (1996). Crystal structure of a DExx box DNA helicase. Nature 384: 379-383.
date revised: 6 June 2000
Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.