BIOLOGICAL OVERVIEW

abstrakt, a gene originally identified genetically by its effect on axon outgrowth and fasciculation of Bolwig's nerve (Schmucker, 1997), encodes a DEAD-box helicase that is essential for survival at all stages throughout the life cycle of the fly. Mutants show specific defects in many developmental processes, including cell-shape changes, localization of RNA, apoptosis and axon guidance. Abstrakt is not globally required for RNA splicing, transport, subcellular localization or translation. Nevertheless, there is a widespread requirement for Abstrakt during post-transcriptional gene expression. Abstrakt must affect processing of specific subsets of RNAs, suggesting that differential post-translational control is important to the development of the fly (Irion, 2000).

In Drosophila, 12 genes are known that code for DEAD-box proteins, including the Drosophila homologs of P68 (Dm62) and eIF-4a. One of the best studied, vasa, was originally identified as a member of the posterior class of maternal effect genes in Drosophila. It is required for the translation of Oskar and Nanos mRNAs during the assembly of the pole plasm and for the localization and translation of Gurken mRNA during oogenesis. In addition to vasa, mutants exist for belle (bel), hel25E (hel), pitchoune (pit) and eIF-4a. A common feature of the phenotypes of these mutants are larval growth defects, but most mutants also show specific characteristics. For example, strong loss-of-function alleles in eIF-4a affect segmentation of the embryo, a process that is particularly sensitive to changes in protein levels; mutations in hel lead to lethality when homozygous, but when heterozygous act as enhancers of position-effect variegation (Irion, 2000 and references therein).

The larval visual system consists of a pair of photoreceptive organs: Bolwig's organs (BO) and their optic nerves, Bolwig's nerves (BN). Each BO consists of only 12 photoreceptors, unipolar neurons, whose axons fasciculate to form the BN. The BN projects ipsilaterally along the surface of the optic lobe anlagen (OLA), the primordia of the adult optic lobes, from where it leaves the surface and grows toward a small target area in the central brain. The growth of the BN proceeds in three phases, during which the nerve changes direction at two intermediate targets, P1 and P2, which are located on the surface of the OLA (Schmucker, 1997).

In a genetic screen for recessive mutations on the third chromosome that disrupt the Bolwig's nerve (BN) projection, mutations in 13 genes were isolated that affect the projection in distinct phases of its development, suggesting that different genes and mechanisms are involved at different steps of the projection (Schmucker, 1997). One of the mutations isolated in this screen was abstrakt (abs). In abs mutant embryos, defects in BN development become evident shortly after the onset of axon outgrowth. Whereas in wild-type embryos the outgrowing BN axons fasciculate rapidly and project as a tightly fasciculated nerve, abs mutant embryos show 2±3 axon bundles, which explore a broad area, show abrupt directional changes and often fail to reach the second intermediate target. Furthermore, the distal-most portions of the BN branches are significantly thicker than in wild-type, containing fine filopodia-like axonal extensions that extend over several cell diameters (Schmucker, 1997). In abstrakt mutants, the formation of the BO is normal with respect to both cell number and differentiation, and the OLA and the second intermediate target P2 also appear morphologically intact, thus excluding the possibility that the axonal projection defects are merely a secondary consequence of abnormalities in cell number, cell fate, or tissue organization (Schmucker, 1997). abs appears to be essential for the directed and fasciculated early outgrowth of the BN, as well as for its navigation at later stages (Schmucker, 1997).

Given its role in the establishment of the larval visual system in the embryo, attempts were made to determine whether abs is also required for the development of the adult visual system. This analysis of the postembryonic requirement for abs function was made possible by the isolation of the temperature sensitive allele abs^E2. abs^E2/abs^LR7 transheterozygotes, which survive to the third instar larval stage at 25°C and thus represent a partial loss-of-function situation, were examined for axonal projection defects in their developing adult visual system using 24B10 antibodies, which specifically label photoreceptor axons. In the wild-type, the photoreceptor axons leave the optic stalk and fan out according to their retinotopic position. The axons of the outer photoreceptors R1-R6 terminate in the first optic ganglion, the lamina, while the axons of the inner photoreceptors R7 and R8 grow deeper into the brain and terminate in the second optic ganglion, the medulla. The overall organization of the retinal projections is disrupted in abs^E2/abs^LR7 animals. The fan of retinal projections into the optic lobe is irregular in shape, axons terminate prematurely and often cluster aberrantly. As in the larval visual system, there are no apparent defects in photoreceptor differentiation in the developing adult system under these weak hypomorphic abs conditions, since both Elav and Chaoptin are expressed, and the number of ommatidia appears to be normal as well. However, some irregularities in the photoreceptor cell layer are observed, including misrotation of ommatidia, additional photo- receptors, and sometimes fewer photoreceptors, suggesting that abs function may be pleiotropic. This notion is confirmed by the fact that abs^E2/abs^LR7 larvae, apart from their visual system-specific defects, often show melanotic tumors at various positions in their bodies, reflecting a broader requirement for abs function. abs mutant embryos also show abnormal axonal projections in the CNS. These defects may be responsible for the embryonic lethality of abs mutants (Schmucker, 2000).

Thus, during postembryonic stages, abs function is not only required in the developing visual system of the adult, but, given the occurrence of melanotic tumors in the mutant, may be required also for the maintenance of normal cellular behavior in a broad range of tissues. For the larval visual system, there is clear evidence that the effects of (zygotic) loss of abs function are likely to be due to a direct involvement of the abs gene product in axon guidance and not a secondary consequence of abnormalities in cell number, cell fate, or tissue organization (Schmucker, 1997). In the developing adult visual system, it is not clear whether abs has multiple parallel functions or whether some of the effects observed are secondary; the molecular role of Abs as a putative RNA helicase does not help to differentiate between these two possibilities. In any case, the broader requirement for abs function is consistent with the ubiquitous expression of the abs transcript throughout development. Importantly, the abs transcript is already present in early embryos prior to the onset of zygotic expression, which is likely to be due to maternally provided RNA. This maternal contribution could potentially mask a requirement for abs function both during early and later stages of development. It is therefore possible that embryos lacking both maternal and zygotic abs function may show more severe developmental abnormalities. Another question in need of further investigation is the tissue specificity of abs requirement (Schmucker, 2000).

As noted above, Abstrakt is essential for survival at all stages throughout the life cycle of the fly: mutational defects in many embryonic developmental processes include cell-shape changes, localization of RNA and apoptosis. To understand which cellular and developmental processes are affected by loss of abs function, defects were analysed in more detail at various stages of embryogenesis and during oogenesis. When embryos mutant for temperature sensitive Abstrakt are shifted to the restrictive temperature at any point from egg laying up to approximately 3 hours, they continued to develop normally up to gastrulation, producing a fully cellularized blastoderm without any demonstrable defects. Abstrakt, therefore, is not essential for nuclear cleavages, migration of the nuclei to the periphery of the egg, or cellularization of the pole cells. Significantly, it is also not needed for the first process directed by the zygotic genome, the cellularization of the syncytial blastoderm. This shows that the genes required for cellularization must be transcribed, and their RNAs exported and translated normally in the absence of functional Abs. Moreover, normal expression patterns of the early zygotic pattern formation genes (for example, twist, snail, hairy, eve) are established by the cellular blastoderm stage, demonstrating that abs is not required for any of the steps in the production of functional proteins from these genes or their upstream regulators. Because the mRNAs of several of these genes need to be spliced, this result excludes a general role for Abstrakt in splicing. Nevertheless, although a normal cellular blastoderm is produced, development at restrictive temperature is significantly slower than in wild-type embryos (Irion, 1999).

At later stages of embryogenesis, embryos mutant for temperature sensitive abs^14B show defects within 30 minutes of a shift to 32°C. Serious morphological defects are seen during gastrulation. The ventral cells of the blastoderm, which normally undergo a series of cell-shape changes, are unable to do this properly. Even where weak apical cell constrictions result in the formation of an indentation, one of the earliest processes during ventral cell shape changes, the movement of nuclei from the apical towards the basal side of the cell, does not occur in the mutants. Interestingly, this phenotype does not resemble that of mutants in either twist or snail, genes encoding the two essential transcription factors required for this process. This shows that Abstrakt either affects only a subset of the gene products controlled by Twist and Snail, or the genes that act in parallel to Twist and Snail (Irion, 1999).

The behaviour of the neighboring ectodermal cells demonstrates that a loss of Abstrakt function does not simply result in a cessation of all cellular activity, or constitute a block to cell-shape changes in general. In wild-type embryos, neuroblasts delaminate from the neuro-epithelial layer, while the epithelial cells become cuboidal. In the mutant, no delamination of neuroblasts occurs, and the epithelial cells become extremely tall and thin. Gene expression patterns, again, appear normal as far as they have been tested (Irion, 1999).

When embryos are shifted after the neuroblasts have segregated, the neuroblasts continue to develop, to activate expression of the neural differentiation marker 22C10 and to extend neurites. The first cells to express 22C10, the ventral midline precursors vMP2 and dMP2 were examined in greater detail. These cells normally extend their neurites toward the posterior and the interior of the embryo. In mutants, neurite outgrowth appears to be randomized, with a preference for growth towards the exterior of the embryo (Irion, 1999).

The pole cells in wild-type embryos are internalized during gastrulation, then pass through the posterior midgut epithelium, contact the mesoderm and migrate anteriorly toward the somatic gonadal precursor cells. They are guided to their correct targets by a combination of repellants and attractants, controlled by the genes wunen and columbus (see HMG Coenzyme A reductase). In abs^14B mutants that were shifted to restrictive temperature after gastrulation, the pole cells that have passed through the midgut epithelium disperse widely within the embryo. In mutants shifted before gastrulation, the pole cells are not internalized and they distribute over the surface of the embryo. This phenotype is stronger than those observed for any of the genes known to be required for pole cell migration. Abstrakt might be involved in the production of one of the guiding signals by interfering with the function of Wunen or Columbus or other, similarly acting proteins (Irion, 1999).

Both of the above cases show that lack of abs function does not compromise the ability of cells to undergo their differentiation program or cellular behaviour typical of their differentiated state. Thus, neurite outgrowth still occurs, and pole cells are able to migrate over large distances, but both types of cells appear to fail to recognize or react to the cues that direct their morphogenetic behavior in the spatially appropriate manner (Irion, 1999).

One reason for this defect might be a loss of subcellular order or polarity. Various aspects of cell polarity were indeed affected in abs mutants. The first abnormality detected was in the localization of mRNAs in the blastoderm. In the wild-type blastoderm, the mRNAs of several genes are distributed unevenly in the cell. For example, the transcripts of short gastrulation (sog) and crumbs (crb) are tightly apposed to the apical cell surface. In abs^14B mutants, CRB mRNA is not seen apically but predominantly at the level of the nuclei, while SOG mRNA forms a gradient from the apical towards the basal side of the cell. In the case of crb, the mislocalization of RNA is also seen at later stages and in other tissues. For example, the tight apical localization in the hindgut epithelium is disturbed in abs mutants. Remarkably, the Crumbs protein is synthesized properly and targeted to the apical cell surface, showing that apical-basal cell polarity in the epithelial cells is correctly established and maintained in abs mutants (Irion, 1999).

Other unevenly distributed mRNAs are those of the pair-rule genes fushi tarazu, even-skipped and hairy, which are localized in the area between the nucleus and the apical cell surface. The localization of these transcripts is not affected in abs mutants. This shows that lack of Abstrakt does not lead to a breakdown of subcellular RNA sorting in general, but that only a subset of transcripts require Abstrakt for their correct targeting (Irion, 1999).

A further manifestation of cell polarity is the regular orientation of mitotic spindles in the post-blastoderm mitotic domains. These are groups of synchronously dividing cells, and in most of them the spindles are oriented parallel to the surface of the embryo. In abs mutants, both the synchrony of division and the regularity of spindle orientation are disrupted. Although the patterning of the mitotic domains, controlled through the transcriptional regulation of String (Cdc25), occurs properly, not all cells in a domain divide at the same time and the mitotic spindles are positioned randomly. In the mispositioned spindles, the centrosomes are mispositioned, together with the spindles (Irion, 1999).

To obtain embryos without either maternal Abstrakt protein or RNA, X-rays were used to induce homozygous mutant germline clones. No mutant clones were obtained, suggesting that abs function is necessary for oogenesis. To test this, ovaries of females homozygous for abs^14B that were shifted to restrictive temperature immediately after eclosion were analysed (pupae kept at restrictive temperature do not eclose). Even after 2 days, the ovaries of these females contained no mature eggs. In most cases, the egg chambers had not developed beyond stage 9. Frequently, the nucleus in abs^14B oocytes does not reach its anterior dorsal position within the oocyte and oskar (osk) mRNA is mislocalized to the center of the oocyte. Less often, a mislocalization of bicoid (bcd) mRNA to the anterior and posterior pole of the oocyte is found. The oldest egg chambers appear to be degenerating, and on analysis by TUNEL staining, their follicle cells are found to be undergoing apoptosis (Irion, 1999).

As a member of the DEAD-box protein family, Abstrakt probably participates in some aspect of RNA processing. The gene was initially identified by a very specific phenotype, the failure of Bolwig's nerve to fasciculate and project normally. By contrast, the variety of phenotypes and tissues affected in the temperature-sensitive mutants suggest, at a first glance, a widespread function for abs. These results do not, however, support a general, ubiquitous function in cellular RNA metabolism for Abstrakt. A general function for Abstrakt in the production of protein from all mRNAs is excluded by several findings, the most important being that morphogenesis and pattern formation up to gastrulation are largely unaffected in abs^14B embryos. This means that the proteins required for these processes are translated properly and, in the case of genes transcribed from the zygotic genome, their mRNAs are synthesized, processed and exported from the nucleus properly. The proteins made in the absence of Abstrakt include at least the maternal determinant Bicoid, the cyclins CycA and CycB, Twist, Snail, Decapentaplegic (Dpp), the segmentation gene products, and the cellularization proteins Nullo, Serendipity-alpha and Bottleneck. At later stages of development, inactivation of Abstrakt also leads to serious defects, but does not interfere with the production of functional mRNA and protein of a number of genes, for example, 22C10, Engrailed, Even-skipped and Twist. Finally, in cases in which experiments have found a more direct effect of abs on RNA, namely, on the subcellular localization of specific mRNAs, only a subset of RNAs responded to the loss of abs function. In summary, these findings indicate that the role of abs is likely to be restricted to a subset of RNAs. The pool of mRNAs expressed in each cell may be less homogeneous, and different transcripts might have more options for processing than might have been assumed. The variety of phenotypes described here should provide assays for the further investigation of the cell biological function of Abstrakt. Conversely, the biochemical and cell biological investigation of Abstrakt, including the identification of its protein and RNA partners, should reveal new levels of control of gene activity during development (Irion, 1999).

PROTEIN STRUCTURE

Amino Acids - 619

Structural Domains

The Abs protein is a member of the DEAD box family of ATP-dependent RNA helicases, which includes eIF-4A, p68, Vasa and Xenopus An3. Their function as RNA helicases is required to control RNA structure in many cellular processes, in particular mRNA splicing, ribosome assembly and translation initia- tion. The hallmark of this family of proteins is eight highly conserved sequence motifs, whose spacing is well conserved. These include the ATPase A motif (AxxGxGKT) involved in ATP binding; the ATPase B (DEAD) involved in ATP hydrolysis, and the SAT and HRIGR motifs, which are important for helicase and protein-RNA interactions, respectively. In Abs all eight motifs are present in their appropriate spacing. Six of the eight motifs are absolutely conserved: the four mentioned above (ATPase A and B, SAT and HRIGR), whose function has been determined, and the GG and TPGR motifs, whose function is not yet known. However, two of the motifs that have been implicated in oligonucleotide binding, PTRELA and RG-D, show conservative substitutions in their amino acid sequence, serine for threonine in the PTRELA motif and lysine for arginine in the RG-D motif. Interestingly, within the large family of DEAD box proteins only two have been found that have the same substitutions, both of which are Arabidopsis thalia proteins. Both Arabidopsis proteins show a very high degree of overall sequence similarity with Abs, suggesting that these proteins are Abs orthologs. However, nothing is known about the biochemical or biological function of these proteins (Schmucker, 2000).

Abstrakt shows the highest homologies with Vasa and Dm62 (P68), among the known Drosophila DEAD-box proteins. The closest relatives in yeast are DBP1/DED1, DBP2 and DBP3 (around 50% similarity and 40% identity in each case). The only sequence found in the databases that encodes a protein with higher similarity is a human expressed sequence tag (EST) cluster (GenBank accession numbers T30201 and T30202). The corresponding human cDNA was cloned by RT-PCR and it was found that the encoded protein is indeed the closest known Abstrakt homolog, showing 88% similarity and 76% identity to Drosophila Abstrakt in the central region. Significantly, it has the same amino-terminal and carboxy-terminal extensions as Drosophila Abstrakt (78% similarity and 64% identity over the whole protein), which is not the case for any other DEAD-box proteins. When the sequence databases were screened with the human sequence, no other protein with an equal or higher degree of similarity was found. This gene has therefore been called hABS (GenBank accession number AF195417) (Irion, 2000).

Two abs mutant alleles that have been isolated in chemical mutagenesis screens were sequenced. In the strong loss-of-function allele abs^IIF1, the absolutely conserved glycine in the HRIGR motif is replaced by a serine. This motif has been shown to be involved in RNA binding and ATP hydrolysis, and mutants in this region have no helicase activity. However, the mutational analysis has so far largely focused on the role of the arginine residues, and the effects of glycine replacements have not been studied. The finding that the replacement of serine for glycine in the HRIGR motif abrogates Abs function underscores the importance of the strict conservation of this motif and demonstrates that the glycine residue does not tolerate even a conservative change. In the temperature-sensitive partial loss-of-function allele abs^E2, the proline at position 254 is replaced by leucine. This proline residue is strongly conserved among DEAD box proteins, but does not belong to any of the absolutely conserved sequence motifs. The finding that a nonconservative exchange of this residue leads to a temperature-sensitive Abs protein product suggests that the proline residue plays an important role for the structure of the protein, either for its activity or for the initial folding of the polypeptide chain during protein synthesis. According to a molecular model of eIF-4A, which is based on the crystal structure of the NS3 helicase of the hepatitis C virus, these two mutated residues appear at close range to one another from opposite sides of an interdomain cleft in which the residues of the PTRELA and the RG-D motifs are involved in oligonucleotide binding (Subramanya, 1996; Benz, 1999). Further functional and crystallographic studies will help elucidate the mechanistic implications of the amino acid exchanges found in wild-type and mutant Abs proteins (Schmucker, 2000).

date revised: 6 June 2000

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.