Digitor/dASCIZ Has Multiple Roles in Drosophila Development

In this study we provide evidence that the spindle matrix protein Skeletor in Drosophila interacts with the human ASCIZ (also known as ATMIN and ZNF822) ortholog, Digitor/dASCIZ. This interaction was first detected in a yeast two-hybrid screen and subsequently confirmed by pull-down assays. We also confirm a previously documented function of Digitor/dASCIZ as a regulator of Dynein light chain/Cut up expression. Using transgenic expression of a mCitrine-labeled Digitor construct, we show that Digitor/dASCIZ is a nuclear protein that is localized to interband and developmental puff chromosomal regions during interphase but redistributes to the spindle region during mitosis. Its mitotic localization and physical interaction with Skeletor suggest the possibility that Digitor/dASCIZ plays a direct role in mitotic progression as a member of the spindle matrix complex. Furthermore, we have characterized a P-element insertion that is likely to be a true null Digitor/dASCIZ allele resulting in complete pupal lethality when homozygous, indicating that Digitor/dASCIZ is an essential gene. Phenotypic analysis of the mutant provided evidence that Digitor/dASCIZ plays critical roles in regulation of metamorphosis and organogenesis as well as in the DNA damage response. In the Digitor/dASCIZ null mutant larvae there was greatly elevated levels of γH2Av, indicating accumulation of DNA double-strand breaks. Furthermore, reduced levels of Digitor/dASCIZ decreased the resistance to paraquat-induced oxidative stress resulting in increased mortality in a stress test paradigm. We show that an early developmental consequence of the absence of Digitor/dASCIZ is reduced third instar larval brain size although overall larval development appeared otherwise normal at this stage. While Digitor/dASCIZ mutant larvae initiate pupation, all mutant pupae failed to eclose and exhibited various defects in metamorphosis such as impaired differentiation, incomplete disc eversion, and faulty apoptosis. Altogether we provide evidence that Digitor/dASCIZ is a nuclear protein that performs multiple roles in Drosophila larval and pupal development.


Introduction
Skeletor is the founding member of the Drosophila spindle matrix complex (reviewed in [1,2]) and localizes to interband regions of polytene chromosomes during interphase [3]. In order to determine other members of the complex interacting with Skeletor we have performed yeast two-hybrid interaction assays, and previously, we identified the chromodomain protein, Chromator, as such an interaction partner [3]. Another protein identified in this screen was a novel zinc-finger protein (CG14962) that we have named Digitor [4,5]. Recently, this protein was independently identified by sequence alignment in a search for orthologs of human ASCIZ by Zaytseva et al. [6] and referred to as dASCIZ by these workers. Digitor/dASCIZ has four zincfinger domains at the amino-terminal end and a SQ/TQ cluster domain (SCD) with six TQT motifs in the carboxy-terminal domain [4,6]. TQT motifs have previously been identified in Dynein light chain binding proteins [7,8] and Digitor/dASCIZ has been shown to interact directly with Cut up, the DYNLL1 homolog in Drosophila [6].
By RNAi knock-down of Digitor/dASCIZ in developing wing discs Zaytseva et al. [6] demonstrated a role for Digitor/dASCIZ in regulating Dynein light chain expression as well as in mitotic progression. However, in order to determine global consequences of the absence of Digitor/dASCIZ during development in this study we have identified and characterized a Pelement insertion that by qRT-PCR analysis appears to be a true null allele. Furthermore, we characterize the localization of Digitor/dASCIZ using a HA-and mCitrine-tagged Digitor/ dASCIZ transgene. We provide evidence that Digitor/dASCIZ is a nuclear protein that performs multiple roles in Drosophila larval and pupal development.
Skeletor bait was used to screen the Clontech Matchmaker 0-21 hr embryonic Canton-S yeast two-hybrid cDNA library as well as a Drosophila 0-2 hr embryonic yeast two-hybrid library (generous gift of Dr. L. Ambrosio, Iowa State University) as previously described [3,14]. Positive cDNA clones were isolated from both libraries, retransformed into yeast cells containing the Skeletor bait to verify the interaction, and sequenced. Homology searches identified the interacting clones as comprised of partial coding sequences from the CG14962 locus.
Pull-down experiments. For in vitro pull down assays of Digitor/Skeletor interactions, a Skeletor cDNA fragment encoding residues 215-474 was subcloned in frame into the pET28a vector (Novagen) containing 6x His-tags. A Digitor cDNA fragment corresponding to amino acid residues 144 to 376 without the NH 2 -terminal zinc finger domains was subcloned in-frame into the pGEX-4T3 vector (Amersham Biosciences) with a GST-tag. As a control construct the amino-terminal fragment of the JIL-1 kinase [19,20] corresponding to the first 260 amino acid residues was subcloned in frame into the pET28a vector. The tagged protein constructs were expressed in BL21 cells (Novagen). For His pull-down assays, approximately 10 μg of His-Skeletor or His-JIL-1 were coupled with Ni-NTA beads (Qiagen) and incubated with 10 μg of GST-Digitor in 1X PBS (pH 7.4) for 4 hr at 4˚C. The protein complex-coupled beads were given a total of 5 washes of 20 min each (2 washes with 0.04% PBST, 2 washes with 0.02% PBST, and a final wash in 1X PBS) and analyzed by SDS-PAGE and immunoblotting using anti-GST mAb 8C7 [3]. Similarly for GST pulldown assays, GST-Digitor or GST alone was bound to glutathione agarose beads (Sigma) and incubated with His-tagged Digitor in 1X PBS (pH 7.4) for 4 hr at 4˚C. The protein complex coupled beads were given a total of 5 washes of 20 min each (2 washes with 0.04% PBST, 2 washes with 0.02% PBST, and a final wash in 1X PBS) and analyzed by SDS-PAGE and immunoblotting using anti-His antibody (Boehringer Mannheim).
For in vitro pull-down assays of Digitor/Cut up interactions, a Digitor cDNA fragment encoding amino acids 256 to 388 was subcloned in-frame into the pET28a vector and a fulllength Cut-up cDNA fragment encoding amino acids 1-89 was subcloned into the pMAL-C2X vector (New England BioLabs) with a MBP tag. The tagged protein constructs were expressed in BL21 cells. For His pull-down assays, approximately 10 μg of His-Digitor or His-Skeletor were coupled with Ni-NTA beads (Qiagen) and incubated with 10 μg of MBP-Cut-up in 1X PBS (pH 7.4) for 4 hr at 4˚C. The protein complex coupled beads were given a total of 5 washes of 20 min each (2 washes with 0.04% PBST, 2 washes with 0.02% PBST, and a final wash in 1X PBS) and analyzed by SDS-PAGE and immunoblotting using anti-MBP antibody (New England BioLabs). Similarly for the MBP pulldown assays, MBP-Cut up or MBP alone was bound to amylose resin (New England BioLabs) and incubated with His-tagged Digitor 1X PBS (pH 7.4) for 4 hours at 4˚C. The protein complex-coupled beads were given a total of 5 washes of 20 min each (2 washes with 0.04% PBST, 2 washes with 0.02% PBST, and a final wash in 1X PBS) and analyzed by SDS-PAGE and immunoblotting using anti-His antibody (Boehringer Mannheim).

Immunohistochemistry
Standard polytene chromosome squash preparations were performed as in Cai et al. [21] using either 1 or 5 minute fixation protocols and labeled with antibody as described in Jin et al. [19]. S2 cell immunocytochemistry using 4% paraformaldehyde fixation protocols were performed as described in Qi et al. [22]. Immunocytochemistry for H3S10ph in dissected third instar larval brains using 4% paraformaldehyde fixation was performed according to the protocol of Wu and Luo [23] and Acridine Orange labeling of third instar larval brains was performed as in Sullivan et al. [24]. BrdU labeling of third instar larval brains was performed essentially as in Truman et al. [25] and in Bolkan et al. [26]. In brief, dissected brains were incubated in Shields and Sang M3 insect medium (Sigma) containing 1 mg/ml 5-bromo 2-deoxyuridine (BrdU; Sigma) for 1 hr. The brains were then fixed in 4% paraformaldehyde, permeabilized in 0.04% PBST, and treated with 2N HCl in PBST for 30 min at room temperature. After a series of washes BrdU incorporation was detected by performing immunocytochemistry. DNA was visualized by staining with Hoechst 33258 (Molecular Probes) in PBS. The appropriate speciesand isotype-specific TRITC-, and FITC-conjugated secondary antibodies (Cappel/ICN, Southern Biotech, Jackson Immuno Research) were used (1:200 dilution) to visualize primary antibody labeling. All antibody-labeling washes were repeated 3 times for 10 min each with PBST (PBS + 0.2% Triton X-100). As previously described [22] the final preparations were mounted in 90% glycerol containing 0.5% n-propyl gallate and examined using epifluorescence optics on a Zeiss Axioskop microscope and images were captured and digitized using a cooled Spot CCD camera. Images were imported into Photoshop where they were pseudocolored, image processed, and merged. In some images non-linear adjustments were made to the channel with Hoechst labeling for optimal visualization of chromosomes.

Expression of tagged constructs in S2 cells
A full-length Digitor cDNA was cloned into the pMT/V5-HisA vector (Invitrogen) with an inframe GFP tag at the COOH-terminus using standard methods [10]. Additionally, a full-length Cut up construct with an mCherry tag and a full-length H2Av construct with a GFP tag were generated in the pAFHW vector (Invitrogen). The fidelity of all constructs was verified by sequencing at the Iowa State University DNA Facility. As previously described [22] Drosophila Schneider 2 (S2) cells were cultured in Shields and Sang M3 insect medium (Sigma) supplemented with 10% fetal bovine serum, antibiotics and L-Glutamine at 25˚C. The S2 cells were transfected with the different tagged constructs or in combination using a calcium phosphate transfection kit (Invitrogen). Expression was either constitutive (Cut up-mCherry and H2Av-GFP) or induced by 0.5 mM CuSO 4 (Digitor-GFP). Cells expressing the constructs were harvested 18-24 h after induction and affixed onto Concavalin-A coated coverslips for immunostaining and Hoechst labeling or images of the fluorescently tagged constructs were captured live and digitized using a cooled Spot CCD camera.

Timelapse confocal microscopy
Timelapse imaging of the fluorescently-tagged constructs in live syncytial embryos was performed using a Leica TCS SP5 tandem scanning microscope as previously described in Ding et al. [27] and in Yao et al. [28]. In short, 0-1.5 h embryos were collected from apple juice plates, and aged 1 h. The embryos were manually dechorionated, transferred onto a cover slip coated with a thin layer of heptane glue, and covered with a drop of Halocarbon oil 700. Timelapse image sequences of a single z-plane or of z-stacks covering the depth of the mitotic apparatus were obtained using a Plan-apochromat 63X 1.4 NA objective.

Phenotypic analysis
Wild-type and Digitor mutant pupae were synchronized at the white prepupal stage (T = 0) and incubated at 25˚C for either 24 hr, 90 hr, or 168 hr in a humidified incubator to document development during metamorphosis. Images of the pupae were obtained using a cooled Spot CCD camera. Analysis of pupal developmental stages was done in accordance with Bainbridge and Bownes [29]. For analysis of the relative brain size of wild-type compared to the Digitor mutant, third instar larvae brains were dissected in PBS and imaged using a cooled Spot CCD camera. The area of the two optic lobes from wild-type and mutant larvae was determined using Image J software (NIH) and results compared using a Student's two-tailed t-test. To compare wild-type and Digitor mutant third instar larval body size, 22 larvae of each genotype were weighed on an analytical balance. Four replicates for each genotype were compared using a Student's two-tailed t-test. For analysis of the relative BrdU incorporation and H3S10ph antibody-labeling in wild-type and Digitor mutant third instar larval brains we obtained images of the fluorescently labeled preparations using a cooled Spot CCD camera. The area of the two optic lobes and thoracic segments from wild-type and mutant larvae was outlined and average pixel intensity of the fluorescence determined using the Image J software (NIH). Background fluorescence was determined in the abdominal segments and subtracted. The results were compared using a Student's two-tailed t-test.

RNA isolation and gene expression analysis
RNA was isolated from wandering third instar larvae or staged pupae using UltraClean1 tissue and cells RNA isolation kit (MO BIO Laboratories, Inc.) according to the manufacturer's instructions. 1 μg of RNA was DNAse (Invitrogen) treated and converted into the first strand of cDNA using the RT-PCR SuperScript III kit (Invitrogen) according to the manufacturer's instructions. For determining changes in gene expression quantitative PCR was carried out with the cDNA template using Brilliant1 II SYBR Green QPCR Master Mix (Stratagene) in an Mx4000 PCR machine (Stratagene). Cycling parameters used were 10 min at 95˚C, followed by 40 cycles of 30 s at 95˚C, 30 s at 55˚C, and 30 s at 72˚C. Fluorescence intensities were plotted against the number of cycles using an algorithm provided by Stratagene. Relative abundance in transcripts was calculated according to the 2-ΔΔCt method [30].
For analysis of changes in ecdysone signaling in staged pupae, RT-PCR was conducted as follows: total RNA was isolated from whole wild-type and EP(3)3709/EP(3)3709staged prepupae and pupae at designated time-points. 1 μg of RNA was converted into cDNA and analyzed by RT-PCR. Digital images of the RT-PCR products after agarose gel separation containing ethidium bromide [10] were obtained with a SONY CCD camera and densitometric analysis performed using Image J software (NIH). Relative level of expression was determined by normalizing average pixel intensity of the experimental bands to the average pixel intensity of RP49 bands. Thermocycler conditions used were as follows: initial denaturation at 94˚C for 5 min followed by 94˚C for 1 min, 60˚C for 1min, 72˚C for 1 min, and a final 10 min 72˚C elongation period. The cycle numbers used varied from 22-35 depending on the individual genes and primer pairs analyzed. The primer pairs used in this study are listed in S1 Table. For RT-PCR analysis, results were obtained from three independent biological replicates.

Paraquat induced stress test
A paraquat resistance test was performed essentially as described in Vermeulen et al. [31] and in Bonilla et al. [32]. In brief, for each genotype, a total of 50 adult flies were assayed per sex. Virgin flies from both wild-type and Digitor heterozygous mutant backgrounds from a single egg batch grown under non-crowding conditions were collected within the same day, sex-segregated, and aged in standard food-containing vials for 5 d at 25˚C. The aged flies were then starved for 6 hr by transferring them into vials containing only water soaked filter papers to avoid desiccation effects and to also ensure that no food remained in their digestive tract at the end of the starvation period. Afterwards, the flies were transferred to vials containing filter paper soaked with 20 mM paraquat in a 5% sucrose solution. Resistance was scored as percent survival at 12 hr intervals for a total of 48 hr after the start of paraquat exposure. Five independent replicates were conducted for each experimental condition.

The spindle matrix protein Skeletor interacts with Digitor/dASCIZ
In order to identify candidates for proteins comprising the spindle matrix macromolecular complex (reviewed in [1,2]) we conducted yeast two-hybrid interaction assays using a Skeletor bait construct containing amino acids 215 through 474 that alone was unable to activate transcription of the reporter genes [3]. Two different embryonic yeast two-hybrid libraries were screened (0-2 hr and 0-21 hr) and interacting clones from both libraries comprising CG14962 coding sequences were identified. The CG14962 locus is located on the third chromosome and has two alternative transcripts due to variant use of different 5' exons; however, the coding sequence is identical and without introns for both transcripts (Flybase). The resulting protein has four zinc-finger domains at the amino-terminal end and a SCD with six TQT motifs in the carboxy-terminal domain and has homology to human ASCIZ (ATM substrate Chk2-interacting Zn 2+ finger) that was originally identified as a stress-response protein involved in signaling induced by DNA damage [31]. Due to the zinc-finger domains CG14962 was named Digitor by Sengupta et al. [4,5] and dASCIZ by Zaytseva et al. [6] based on its homology. Analysis of the isolated Digitor/dASCIZ yeast two-hybrid library clones suggested that the interaction region with Skeletor was COOH-terminally located and independent of the zinc-finger region.
To confirm the physical interaction of Digitor/dASCIZ with Skeletor, we performed in vitro pull down experiments using a His-tagged Skeletor fusion protein containing the yeast two-hybrid bait region and a Digitor-GST fusion protein without the four zinc-finger motifs ( Fig 1A). Whereas GST protein alone or beads only were not able to pull down Skeletor, the Digitor-GST fusion protein pulled down a band corresponding to the size of Skeletor-His that was detected by His-antibody (Fig 1B). In the reverse experiment, Skeletor-His fusion protein was able to pull down Digitor-GST as detected by GST-antibody ( Fig 1C). However, beads only and a His-tagged JIL-1 kinase control fusion protein did not (Fig 1C). These results support the existence of a direct physical interaction between Skeletor and Digitor/dASCIZ that does not require the zinc-finger region.
The presence of the TQT motifs in Digitor/dASCIZ indicates an interaction with the Dynein light chain as has been previously reported for Cut up in Drosophila by Zaytseva et al. [6]. We provide further evidence for this interaction by in vitro pulldown assays with a Histagged COOH-terminal fragment of Digitor containing the six TQT motifs ( Fig 1A) and a full length Cut up MBP fusion protein. As illustrated in Fig 1D whereas MBP protein alone or beads only were not able to pull down Digitor, the Cut up-MBP fusion protein pulled down a band corresponding to the size of Digitor-His that was also detected by His-antibody. In the reverse experiment, Digitor-His fusion protein was able to pull down Cut up-MBP as detected by MBP-antibody ( Fig 1E). However, beads only and a His-tagged Skeletor control fusion protein did not (Fig 1E). . To further document the interaction with Cut up we co-transfected S2 cells with Digitor-GFP and Cut up-mCherry (Fig 3). Cut up-mCherry in singly transfected cells was localized both to the nucleus and in the cytoplasm (241 out of 241, 100%). However, in 73 out of 235 (31%) co-transfected cells, Cut up-mCherry was only found in the nucleus together with Digitor-GFP (Fig 3A). In cells co-transfected with Cut up-mCherry and an H2Av-GFP control construct, Cut up-mCherry retained its dual localization to both the nucleus and cytoplasm (210 out of 210 cells examined) (Fig 3B). Taken together these data suggest that overexpressed Digitor-GFP binds and recruits Cut up-mCherry to the nucleus providing evidence for an in vivo interaction between Digitor/dASCIZ and Cut up.
To further explore the nature of the nuclear localization of Digitor/dASCIZ we expressed a 3xHA-Digitor-mCitrine full length construct ( Fig 4A) transgenically in third instar larvae under UAS-Gal4 control. Under non-stressed conditions when driven by the sgs3-Gal4 driver, 3xHA-Digitor-mCitrine binds to polytene chromosomes at interband regions in a complementary pattern to that of Hoechst ( Fig 4B) and co-localizes with Skeletor ( Fig 4C). Interestingly, enhanced levels of 3xHA-Digitor-mCitrine but not of Skeletor at developmental puff regions [33] was especially prominent (Fig 4B and 4C). Furthermore, many but not all of these puffs were also labeled by antibody to the paused form of RNA Polymerase II (Pol IIo ser5 ) (Fig 5). This pattern of labeling is compatible with a general role for Digitor/dASCIZ as a transcriptional regulator of development. Since human ASCIZ is known to be a stress response protein [34,35] we performed heat shock experiments on third instar salivary glands (Fig 5) in order to examine whether this pattern changes under stressed conditions. Polytene chromosome squashes were made from salivary glands at control conditions (before heat shock) as well as 5 min, 1.5 hr, and 5 hr after heat shock. The preparations were double labeled with Pol IIo ser5 a marker for heat shock puff regions [36]. In response to heat shock treatment there is a redistribution of Pol IIo-ser5 labeling which is reduced at most sites, while being dramatically upregulated at heat shock puffs where transcription of heat shock-activated genes is occurring [36] (Fig 5). In contrast, 3xHA-Digitor-mCitrine was released only from developmental puff regions and was not recruited to heat-shock induced puffs (Fig 5). While the non heat-shock pattern of Pol IIo ser5 labeling was restored after 1.5 hr the binding of 3xHA-Digitor-mCitrine to developmental puff regions recovered on a slower time scale and was only restored 5 hr after heat-shock (Fig 5). This dynamic relocation of Digitor/dASCIZ from developmental puff regions during heatshock conditions is compatible with the hypothesis that Digitor/dASCIZ may modulate gene expression at these sites and that this modulation changes during the stress response.
The interaction of Digitor/dASCIZ with Skeletor suggest a possible interaction with the spindle matrix during cell division. To explore this possibility we performed time-lapse imaging of 3xHA-Digitor-mCitrine together with tubulin-mCherry during mitosis in syncytial Drosophila embryos. The results showed (Fig 6 and S1 Movie) that a considerable fraction of 3xHA-Digitor-mCitrine remains in the nuclear space after nuclear envelope breakdown (NEB) embedding the forming microtubule spindle apparatus, a defining feature of spindle matrix proteins [28]. After NEB there are no diffusion barriers even for greater than 2000 kDa molecules [28] suggesting that Digitor/dASCIZ's confinement to the spindle region during mitosis is mediated by molecular interactions with the spindle matrix complex. A likely mechanism for such interactions would be Digitor/dASCIZ's direct physical association with Skeletor.

Digitor/dASCIZ is an essential protein
In order to determine the phenotypic consequences of the absence of Digitor/dASCIZ during development we have identified a P element, EP(3)3709, inserted 27 bp downstream of the Digitor/dASCIZ ATG start codon that is likely to be a true null allele. To verify this we performed qRT-PCR of transcripts from Digitor/dASCIZ, Cut up, Skeletor, and the neighboring gene CG32280 extracted from Wild-type and EP(3)3709 homozygous third instar larvae, respectively. The Dynein light chain gene Dynll1 (Cut up in Drosophila) has been established to be an evolutionarily conserved transcriptional target of ASCIZ proteins down regulating its To verify that the observed lethality was caused by impaired Digitor/dASCIZ function we used 3xHA-Digitor-mCitrine as a rescue construct using a Act5C-Gal4 driver line. Thus, in the following cross: 3xHA-Digitor-mCitrine, w+/CyO, y+; EP(3)3709/TM6b, Sb,Tb, e s X Act5C-Gal4, w+/CyO,y+; EP(3)3709/TM6b, Sb, Tb, e s rescue would be indicated by the presence of adult progeny that lack the Sb phenotype. Out of 249 adult flies examined from such a cross we found 26 flies with the non-Sb phenotype (S2 Table). This indicates that the 3xHA-Digitor-mCitrine construct can provide partial rescue function (31%) supporting that the EP(3)3709 is a null or strong hypomorphic Digitor/dASCIZ allele and that Digitor/dASCIZ is an essential gene.
The enrichment and localization of Digitor/dASCIZ to developmental puff regions suggest a role in organ maturation and metamorphosis. An early phenotype we observed in homozygous EP(3)3709 (Digitor mutant) crawling third instar larvae was a severely reduced brain size (Fig 8A and 8B). We quantified this aspect by comparing the area of the two optic lobes of brains from Wild-type and Digitor mutant larvae, respectively. As illustrated in Fig 8B there was a statistically significant (p<0.001) reduction of this parameter in Digitor mutant larvae to about half that of Wild-type larvae. Interestingly, this was not a consequence of a general decline in larval development as the size and weight of Digitor mutant larvae were indistinguishable from that of Wild-type larvae (Fig 8C and 8D). A small brain phenotype could be the result of decreased cell proliferation, increased cell death, or a combination of both. To distinguish between these scenarios we first assessed the levels of cell death and apoptosis in brains from Digitor mutant and Wild-type larvae by Acridine Orange (Fig 8E) and TUNEL (Fig 8F) labeling. The results showed that compared to Wild-type there was only a slight increase in cell death in the Digitor mutant that was not likely to be able to account for the markedly decreased brain size. However, both BrdU labeling, a marker for actively replicating cells [26] and H3S10ph labeling, a marker for cells in mitosis [38] showed significantly fewer cells in the actively dividing proliferation brain zones (Fig 8G-8J) in the Digitor mutant. Taken together these results suggest that impaired cell proliferation was the main cause of the mutant small brain phenotype.

Digitor/dASCIZ is required for metamorphosis
Digitor mutants are pupal lethals, however, the exact time at which lethality occurs is not precise and varies from very early metamorphosis with the formation of a brown puparium with no detectable differentiation of tissue, to late metamorphosis with dead pharate adults. To further analyze the contribution of Digitor/dASCIZ in metamorphosis, Digitor mutants were selected as white prepupae and monitored through prepupal and pupal development and analyzed as in Bainbridge and Bownes [29]. At 24 hours after puparium formation (APF), the wild-type population showed posteriorly displaced imagos with proper imaginal disc eversion, and legs and wings were fully extended along the abdomen (Fig 9A). Gas bubbles could be clearly detected in the anterior end whereas salivary glands were completely degenerated and green malphigian tubules were visible in the abdomen (Fig 9A). At this time-point, most of the Digitor mutant population was either decayed or undifferentiated without proper head or imaginal disc eversion (Fig 9C and 9D). In some cases within the pupal case the gas bubble was still present posteriorly displacing the imago anteriorly. In the remaining 45% of the mutant population that was not desiccated and had the correct gas bubble displacement, formation of adult structures was impaired (Fig 9B). In these pupae the imaginal discs everted normally and the clear pupal cuticle was formed, but within this cuticle the integrity of the epithelial sheets in the imaginal tissues was severely compromised (Fig 9B). An ecdysone pulse approximately at 10-12 hr post puparium formation normally triggers programmed cell death of the larval salivary glands [39,40]. To determine if this response is defective in the Digitor mutant, presence or absence of salivary glands was analyzed in mutant pupae at 24 hr APF. The results showed that salivary glands still could be detected in 12 out of 13 mutant animals analyzed within the 45% mutant pupae population that was not desiccated and had the correct gas bubble displacement suggesting a defective ecdysone response. At 90 hours APF, the wild-type population showed bright red pigmentation in the eyes with the full complement of bristles; the abdomen was segmented and developed; wings were folded and almost black in color (Fig 9E). In contrast, at this time point, 52% of the mutant population appeared shriveled and decayed (Fig 9G). Another 30% of the population had impaired development comparable to the 24 hr APF mutants (Fig 9F). In the remaining 18% of the population some rudimentary development of adult organs and appendages could be detected (Fig 9H). After 96 hours APF the wild-type population eclosed as adult flies ( Fig 9I); however, there were no eclosion events observed within the Digitor mutant population, which was analyzed for another 72 hr. At 168 hr APF, 53% of the mutant population were shriveled, decayed and degenerated (Fig 9K), and another 33% of the population were without distinct adult organs inside the pupal cuticle ( Fig 9J). Interestingly, in a small population (13%) further signs of development compared to the 96 hr APF-mutants could be detected with pigmentation of the eyes along with bristles in head, thorax, legs and abdomen (Fig 9L). However, the different populations of mutant pupae described all failed to eclose with no escapers and eventually degenerated. Taken together, these results suggest a critical role for Digitor/dASCIZ during metamorphosis.
Defective ecdysone signaling in the Digitor mutants during early metamorphosis A major regulator of metamorphosis and subsequent eclosion is the ecdysone-signaling cascade that is tightly controlled by the expression of ecdysone-regulated genes [41,42]. Defective ecdysone signaling caused by an aberrant expression of one or more genes of the signaling cascade has been implicated in impaired metamorphosis in numerous studies [39,[43][44][45][46][47][48][49][50]. Analysis of the pupal phenotypes in the Digitor mutant raised the possibility that these phenotypes were triggered by changes in expression of key ecdysone-regulated genes. To test this hypothesis, we analyzed the expression of a set of ecdysone-regulated genes, namely, ECR, E74, BR-C, βFTZ-F1, and E93 at 3 time-points: at white puparium formation (T 0 ), at the mid-prepupal period (T 8 measured as 8 hr after T 0 ), and at the time just after head eversion (T 14 measured as 14 hr after T 0 ).
In the experiments RNA was isolated from white prepupae (T 0 ), mid prepupae (T 8 ) and early pupae (T 14 ) from both wild-type and Digitor mutant backgrounds, converted into cDNA and analyzed for the expression of the aforementioned genes by RT-PCR (Fig 10). All experiments were performed in triplicate. Digitor/dASCIZ expression was significantly reduced in the mutants at all time points examined (Fig 10F). Many of the ecdysone-regulated transcripts surveyed registered a slight decrease in abundance relative to wild-type at both T 8 and T 14 ( Fig  10); however, only the expression of BR-C and E93 at T 8 and BR-C and Mdh2 at T 14 were significantly reduced (Fig 10B, 10E, and 10G) with likely consequences for development. BR-C is a key ecdysone inducible gene required for metamorphosis and has been shown to perform a plethora of functions including morphogenesis of imaginal discs to form continuous and compact adult structures [46,51,52]. Hence downregulation of BR-C may account for many of the observed defects and general lack of development in Digitor mutant pupae. The expression of E93 was significantly decreased at T 8 but not at T 14 (Fig 10E). Previous studies have shown that E93 is induced in the larval midgut at the mid prepupae stage catalyzing the destruction of midgut gastric caeca and midgut shortening [39]. Thus reduced expression of E93 in the Digitor mid-prepupal mutant is likely to interfere with larval midgut destruction. Moreover, studies have indicated that induction of E93 is also involved in salivary gland histolysis in response to the ecdysone pulse at 12 hr APF [39,53]. However, at this time point we observed no statistically significant reduction in the levels of E93 in the Digitor mutant ( Fig 10E). Nonetheless, an instance of E93 independent block in salivary gland cell death was reported in Med24 mutants where a dramatic reduction in Mdh2 expression was suggested to interfere with salivary gland histolysis [40]. As illustrated in Fig 10G expression of Mdh2 was also significantly reduced in the Digitor mutant pupae at T 14 providing a possible mechanism for the observed block in salivary gland histolysis. Taken together these results suggest that select subsets of ecdysone-regulated genes as well as Mdh2 are down regulated during early metamorphosis in the Digitor mutant background contributing to defective and incomplete metamorphosis. Impaired DNA damage response in the Digitor mutant DNA double-strand break (DSB) repair is an important mechanism for maintaining genomic stability (reviewed in [54]). DSBs trigger activation of the DNA damage response pathway which activates a number of protein kinases including the ATM kinase that rapidly phosphorylate the carboxy-terminus of the histone 2A variant (H2Av) (reviewed in [55]). This phosphorylation is evolutionarily conserved [56] and in Drosophila the phosphorylated form is denoted γH2Av with the S137 residue being phosphorylated [17]. Thus, a standard assay for DNA damage detection and recognizing DSBs is to probe for elevated levels of γH2Av using anti-γH2Av antibody [17]. Since the ASCIZ family of proteins may have a general role in modulating ATM kinase signaling [57] we compared γH2Av levels in protein extracts from wildtype and Digitor mutant third instar larvae. As illustrated by the immunoblots in Fig 11A there are greatly elevated levels of γH2Av in the mutant larva. We quantified this difference by determining the average pixel density of the anti-γH2Av antibody labeled bands normalized to the pixel density of anti-Lamin Dm0 antibody labeled loading controls. The data from three independent determinations indicate about a 3-fold increase in the levels of γH2Av and concomitant DSBs in the Digitor mutant as compared to wild-type ( Fig 11B). These data strongly suggest that Digitor/dASCIZ is involved in the DNA damage response in Drosophila.

Increased susceptibility to paraquat induced stress in Digitor heterozygous mutants
Since accumulation of DSBs as indicated by elevated levels of γH2Av was 3-fold higher in the Digitor homozygous mutants compared to the wild-type controls (Fig 11), we hypothesized that reduced levels of Digitor/dASCIZ may also decrease the resistance to paraquat-induced oxidative stress. Paraquat is known to induce DNA damage by generation of free radicals [32,58]. To test this we generated w 1118 ; EP(3)3709/TM6B, Tb 1 heterozygous mutant flies and compared them to wild-type controls. In the experiments we administered 20 mM paraquat in  Fig 12A) and female (Fig 12B) heterozygous mutant flies with males being more affected with no surviving flies after only 12 hr (Fig 12A). These data suggest that Digitor/dASCIZ plays a critical role in conferring resistance to paraquat induced oxidative stress in Drosophila.

Discussion
In this study we present evidence that in Drosophila the spindle matrix protein Skeletor interacts with Digitor/dASCIZ, the human ASCIZ ortholog. We detected this interaction in a yeast two-hybrid screen and confirmed it by pull-down assays. The transgenic expression of a mCitrine-labeled Digitor construct shows that Digitor/dASCIZ protein is localized to interband and developmental puffed chromosomal regions during interphase but that it redistributes during mitosis to the spindle region. We have identified and characterized by RT-PCR analysis a P element insertion allele of Digitor/dASCIZ that appears to be a true null allele. When homozygous this allele results in a complete pupal lethal phenotype, indicating that Digitor/ dASCIZ is an essential gene. This pupal lethality was partially rescued (31% viability) by the transgenic expression of 3xHA-Digitor-mCitrine in a homozygous Digitor/dASCIZ mutant background. Rescue was not complete, likely due to differences in the levels of expression of the transgene compared with wild type gene expression. Our analysis of the phenotypic consequences of the absence of Digitor/dASCIZ during development combined with its dynamic nuclear localization suggest that Digitor/dASCIZ has multiple roles in Drosophila development.
Digitor/dASCIZ regulation of mitosis. Previously, Zaytseva et al. [6] using RNAi knockdown of Digitor/dASCIZ in the posterior compartment of imaginal wing discs provided evidence that decreased Digitor/dASCIZ leads to impaired mitosis with severe spindle and chromosome alignment defects as well as to reduced wing size. These effects were mainly attributed to a decrease in Dynein light chain (Cut up) expression when levels of Digitor/dASCIZ are downregulated [6]. However, we show that Digitor/dASCIZ, in addition to binding to Cut up itself ( [6]; this study), has direct physical binding interactions with the spindle matrix protein Skeletor. It has been demonstrated that the Drosophila spindle matrix protein Megator homolog Tpr in mammalian cells associates with the Dynein complex during mitosis [59]. Furthermore, live imaging analysis indicated that Digitor/dASCIZ is confined to the spindle region during mitosis at a time after NEB when where there are no diffusion barriers. Thus, these findings suggest the possibility that Digitor/dASCIZ may be playing a direct role in mitotic progression as a member of the spindle matrix and/or Dynein complex in addition to serving as a transcriptional regulator of Dynein light chain expression.
Digitor/dASCIZ is required for tissue maturation and metamorphosis. The enrichment and localization of Digitor/dASCIZ to developmental puff regions suggest a role in organ maturation and metamorphosis. Furthermore, the dynamic relocalization of Digitor/dASCIZ from developmental puff regions during heat-shock conditions is compatible with the hypothesis that Digitor/dASCIZ may modulate gene expression at these sites and that this modulation changes during the stress response. An early phenotype we observed in Digitor mutant crawling third instar larvae was a severely reduced brain size although overall larval development at this stage appeared normal with the size and weight of the mutant larvae indistinguishable from that of wild-type larvae. While we detected only a small increase in cell death and apoptosis in Digitor mutant brains there was a marked decrease in the number of actively dividing cells in the proliferating brain zones suggesting impaired cell proliferation was the main cause of the mutant small brain phenotype. This is in contrast to the findings of Zaytseva et al. [6] indicating that the underlying reason for reduced wing size after Digitor/dASCIZ RNAi knock down was anaphase arrest of mitotic cells caused by Dynein complex-dependent spindle defects leading to enhanced apoptosis. However, considering the pleiotropic effects of Digitor/ dASCIZ the consequences of its absence is likely to be tissue and context dependent. For example, we provide evidence that the actual programmed cell death of larval salivary gland cells during pupation does not occur in the absence of Digitor/dASCIZ. In the Digitor null mutant pupation is initiated; however, no eclosion occurs and pupal development is impaired and arrested at various stages. RT-PCR analysis of transcript levels of ecdysone-regulated genes controlling metamorphosis and subsequent eclosion in the ecdysone-signaling cascade showed a significant decrease in some but not all of the genes examined. We therefore propose that the pupal phenotypes in the Digitor mutant is a result of changes in expression of a subset of ecdysone-regulated genes as well as Mdh2 and that Digitor/dASCIZ plays a role in metamorphosis by acting as a transcriptional regulator of these genes. However, it should be noted that the experiments do not rule out that these effects are an indirect consequence of timing defects and/or Cut up dysregulation. In mammals ASCIZ also has several ATM-kinase independent developmental functions and is required for B cell maturation as well as for lung, kidney, and brain organogenesis [60][61][62]. Some of these defects are a consequence of dynein light chain down regulation [8,35]; however, in other cases the effects have been suggested to be caused by modulation of developmental signaling cascades such as the Wnt pathway [62].
Digitor/dASCIZ is involved in the DNA damage response. Mammalian ASCIZ has a role in orchestrating ATM-kinase but not NBS1 dependent DNA damage repair in response to agents that perturb chromatin structure such as chloroquine or by osmotic and oxidative stress [35,63]. Here we provide direct evidence for an evolutionarily conserved role for Digitor/dAS-CIZ in the DNA damage response in Drosophila. In the Digitor null mutant larvae levels of γH2Av were greatly elevated indicating accumulation of DSBs in the absence of Digitor/ dASCIZ. Furthermore, reduced levels of Digitor/dASCIZ decreased the resistance to paraquatinduced oxidative stress resulting in increased mortality in a stress test paradigm. Moreover, Zaytseva et al. [6] have shown that GFP-tagged Digitor/dASCIZ expressed in HeLa cells accumulate within 5 min along tracks of laser-induced oxidative DNA damage reminiscent of the association of human ASCIZ with oxidative DNA damage-induced foci [34].
Thus, in summary we show that Digitor/dASCIZ has multiple developmental functions in Drosophila and plays critical roles in regulation of metamorphosis and organogenesis as well as in the DNA damage response. In addition, we provide evidence that Digitor/dASCIZ may be contributing to mitotic progression not just as a regulator of Dynein light chain transcription but also through direct physical interactions with the spindle matrix (e.g. Skeletor). It will be of interest in future studies to further elucidate the many functional roles of Digitor/dAS-CIZ and their interplay in the highly amenable Drosophila model system.
Supporting Information S1 Movie. Confocal timelapse imaging of transgenically expressed Digitor-mCitrine and tubulin-mCherry in a syncytial Drosophila embryo. Confocal timelapse imaging of transgenically expressed Digitor-mCitrine and tubulin-mCherry in a syncytial Drosophila embryo (Leica TCS SP5 tandem scanning microscope). The image sequence illustrates the dynamic relationship of Digitor/dASICZ (in green) relative to microtubule spindle formation (in red) through two mitotic cycles. The results suggest that Digitor-mCitrine may associate with the spindle matrix after invasion of microtubules into the nuclear space. The timelapse covers a period of 20 min. (MOV) S1 Table. Primers for the genes analyzed by RT-PCR.