SEARCH

SEARCH BY CITATION

Keywords:

  • Drosophila;
  • deflated;
  • Ints7;
  • developmental signalling;
  • snRNA processing

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The Drosophila gene deflated (CG18176; renamed after the pupal lethal abdominal phenotype of mutant individuals) is a member of a conserved gene family found in all multicellular organisms. The human orthologue of deflated (Ints7) encodes a subunit of the Integrator complex that associates with RNA polymerase II and has been implicated in snRNA processing. Since loss-of-function analyses of deflated have not yet been reported, we undertook to investigate deflated expression patterns and mutant phenotypes. deflated mRNA was detected at low levels in proliferating cells in postblastoderm embryos and GFP tagged protein is predominately nuclear. Generation and analysis of four mutant alleles revealed deflated is essential for normal development, as mutant individuals displayed pleiotropic defects affecting many stages of development, consistent with perturbation of cell signalling or cell proliferation. Our data demonstrate multiple roles in development for an Ints7 homologue and to demonstrate its requirement for normal cell signalling and proliferation. Developmental Dynamics 238:1131–1139, 2009. © 2009 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

RNApolII transcribes a number of types of RNAs, including mRNAs and small nuclear RNAs (snRNAs). Integrator is a newly identified 12-subunit complex that associates with and forms part of the transcriptional regulatory network of human RNA polymerase II (Baillat et al.,2005). Of the 12 subunits initially identified biochemically, seven (Ints 1, Ints 3, Ints 6, Ints7, Ints 8, Ints 11, and Ints 12) copurify with the largest subunit of human RNApolII C-terminal domain (CTD). The CTD domain is required for both the transcription and 3′ processing of the U1 and U2 snRNAs (Uguen and Murphy,2003), with the 3′ processing of these transcripts likely to be due to the activity of the Ints11/Ints9 subcomplex acting within the larger Integrator complex. RNAi depletion of either Ints11 (the presumptive catalytic subunit) and Ints1 (the largest Integrator subunit) results in unprocessed U1 and U2 primary transcripts (Baillat et al.,2005). However, aside from Ints11 and Ints9, the biochemical or cellular roles of the remaining Integrator subunits are presently unclear, despite their strong evolutionary amino acid sequence conservation.

Here we describe the initial genetic characterisation of the Drosophila orthologue of the Ints7 Integrator subunit, which is located on the left arm of chromosome 3 within polytene band 67C5. In line with the gene-naming conventions for Drosophila, we have renamed this gene “deflated” (defl) after its late pupal-lethal abdominal phenotype (described below). Preexisting functional data of orthologues of Ints7/deflated include cursory reports from large-scale mutagenesis experiments in zebrafish (Golling et al.,2002) and RNAi knockdown in C. elegans (Kamath et al.,2003). Both of these reports demonstrate that the orthologue of Ints7/deflated is required for normal development in these species. However, in neither was a detailed analysis performed. Since in-depth developmental and genetic analyses have not yet been reported, we undertook to characterise deflated function in Drosophila.

To study the function of deflated in detail, we generated four independent germline mutations in Drosophila, which enabled us to determine the requirements for deflated during development. Our analyses of homozygous and transheterozygous alleles, mutant individuals partially rescued by wildtype deflated transgenes, and DEFL overexpression studies reveal mutant phenotypes consistent with defects in cell proliferation and/or cell signalling.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

deflated Encodes an Evolutionarily Conserved Protein

The protein encoded by the Drosophila deflated gene (DEFL) is predicted to be 1,001 amino acids in length and to have a molecular weight of 112 kDa (Fig. 1A). While deflated encodes a well-conserved protein, with a single orthologue present in all metazoan genomes examined to date, homology searches fail to assign it to any previously studied protein superfamily. Homology alignments of representative DEFL orthologues show that while DEFL proteins have a high degree of conservation, only the N-terminus shows an identifiable conserved protein domain (Fig. 1B). This domain contains tandem copies of a degenerate protein motif related to the HEAT repeat family as well as a putative cyclin binding site (Fig. 1B). As proteins with high levels of similarity to DEFL are predicted from genome project data obtained from both plant and animal species (Fig. 1C), and no significant similarity could be detected in any coding sequences identified in yeast (S. cerevisiae and S. pombe), fungal (Aspergillus nidulans), or bacterial genomes, DEFL is likely to play a cellular role unique to multicellular organisms.

thumbnail image

Figure 1. DEFL protein conservation and mutant gene structures. A: A schematic of the Deflated protein (AAK92871) showing conserved putative motifs including N-terminal HEAT repeats and conserved cyclin binding site. B: Homology alignment of the N-termini of DEFL orthologues from human, zebrafish, and Drosophila showing the conserved putative cyclin binding site (CYC) within the conserved HEAT domain. C: Dendrogram of CLUSTALW aligned DEFL/Ints7 protein sequences from Human (NP 056249), Mouse (NP 848747), Chicken (NP 001006399), Zebrafish (NP 775374), Sea Urchin (XP 789961), Nematode (NP 496477), Arabidopsis (NP 193739), Mosquito (XP 317927), and Drosophila (NP 648352). D: Diagram of the genomic region of deflated showing the exon/intron structure and the direction of transcription of deflated and the flanking genes nbs and CG18177. The dashed line above deflated exon 1 indicates the region corresponding to the riboprobe used for the mRNA in situ hybridisation in Figure 3. Triangle indicates location of the EP-element (EP(3)3301) inserted 276 bp from the end of the deflated coding region used in generating the deflated alleles. All alleles comprise small deletions (indicated by parentheses) into the end of the deflated coding region. E: The truncated DEFL proteins predicted by the molecular analyses of the three deflated alleles. Numbers correspond to amino acids encoded by the wild type deflated sequence.

Download figure to PowerPoint

thumbnail image

Figure 3. deflated mRNA expression in wildtype embryos. A–H: An antisense deflated mRNA probe shows that deflated mRNA is not expressed in (A) syncytial or (B) cellularised embryos. Expression is first detected during (C) the beginning of gastrulation in the epidermis and in (D) the beginning of the gut formation. As the embryo gets older, expression is strongest in the gut but is also observed in the (E–G) central nervous system (CNS) and (H) possibly the peripheral nervous system (PNS). I, J: A sense deflated mRNA probe shows no specific staining under identical conditions. All embryos are shown anterior to the left, posterior to the right, dorsal up, and ventral down.

Download figure to PowerPoint

Generation of deflated Alleles

Using the EP(3)3301 strain, which harbours an EP transposable element inserted 276 bp downstream of the deflated coding region (Fig. 1D), yet displays no discernable phenotypes, we generated deflated mutant alleles by P-element mobilization. These alleles, denoted P{w+mC=EP} deflP, P{w+mC=EP} deflL, P{w+mC=EP} deflZ, and P{w+mC=EP}deflΔ (hereafter referred to as deflP, deflL, deflZ, and deflΔ, respectively), comprise deletion of sequences flanking the original EP insertion site. deflP is a 2,131-bp deletion that includes 1,857 bp of the deflated coding region. In deflL and deflZ, 1,560 and 702 bp have been deleted, removing 910 and 52 bp of the defl coding region, respectively (Fig. 1D). In deflΔ, the entire deflated coding region and half of the neighbouring nbs coding region is deleted (data not shown). Consequently, deflΔ is expected to yield no protein at all, and the three other alleles are predicted to generate C-terminal truncations of the DEFL protein, with deflP predicted to encode amino acids 1–401 (relative to the 1,001 amino acid wild type protein), deflL amino acids 1–698, and deflZ amino acids 1–986 (Fig. 1E). DNA sequence analysis of RT-PCR amplified defl transcripts produced from each of the mutant alleles is consistent with these predictions (data not shown).

deflated Is Required at Various Developmental Stages

Homozygous loss of deflated function was found to be lethal, with no homozygous adult flies of any of the four alleles recovered from sib crossing of balanced heterozygotes. Examination of sib cross-cultures established the lethality period for all alleles to be towards the end of the second-instar larval stage. Homozygous third-instar larvae were absent from all cultures and many homozygous second-instar larvae were observed to wander out of the food before becoming sluggish and dying. Intercrossing of deflated alleles revealed deflP, deflL, deflZ, and deflΔ failed to complement one another, with the lethality period of deflP/deflL, deflP/deflΔ, and deflL/deflΔ being indistinguishable from the homozygous deflated individuals. In contrast, the lethal period for most deflL/deflZ, deflP/deflZ, and deflΔ/deflZ individuals occurred during the pupal stage. Additional crosses with alleles from another complementation group recovered in the mutagenesis screen revealed a second recessive mutation on the chromosome carrying deflZ (data not shown). As this second site mutation is closely linked to the deflZ allele, analyses of deflZ were restricted to experiments involving animals transheterozygous at the defl locus, ensuring the second site mutation did not contribute to the observed phenotype(s).

Although many deflP/deflZ, deflL/deflZ, and deflΔ/deflZ individuals die during the pupal stage, 60% of expected deflP/deflZ individuals (38 transheterozygous adults /164 total adult progeny) and 18% of expected deflL/deflZ and deflΔ/deflZ individuals (21/252 and 7/84, respectively) reached adulthood. These data indicate that deflP is a hypomorphic allele, yielding more DEFL function than either deflΔ or deflL, which are likely to be null alleles. On the other hand, deflZ behaves as a weak hypomorphic allele, supplying sufficient DEFL function to allow some transheterozygous individuals to reach adulthood. The phenotypic strength of the alleles, with the exception of deflP, correlates with the degree of coding region loss at the molecular level. The deflPphenotypic data suggest the presence of an inhibitory domain in the 698–986aa interval, or the existence of more complicated regulatory effects that will require further experimentation to resolve.

The majority of transheterozygous adult escapers (80% (32/38) of deflP/deflZ and 100% (21/21) of deflL/deflZ) were male, although deflL/deflZadult females occasionally arose in mass cultures. These data indicate that male development requires less DEFL function, on average, than female development. The observation that mutants carrying presumptive null alleles can continue development until the second instar stage may be due to perdurance of maternal DEFL protein, consistent with the observation of defl mRNA being upregulated in ovaries (Chintapalli et al.,2007).

Whereas wildtype flies at the pharate adult stage have formed all of the adult structures, but have not yet eclosed from the pupal case (Fig. 2A), most deflated transheterozygotes that survive to a similar stage of development display abdominal defects of varying severity (Fig. 2B). Normal cuticle formation was found anteriorly, with head, thorax, and between one and four anterior abdominal tergites forming. However, posterior abdominal tergites were frequently absent, and external genitalia were not observed in such situations. In many deflated transheterozyous individuals, the abdomen exhibited a “deflated balloon” appearance, which led us to name this gene (previously designated CG18176) deflated. deflated transheterozygous adult escapers did not show any obvious abdominal defects. However, they often lacked vigour and frequently died within a few days of eclosing. Other morphological defects observed in these adult escapers included minor defects in wing vein morphology, where the L5 vein fails to extend to the wing margin (Fig. 2C), and bent humeral bristle morphology (Fig. 2D).

thumbnail image

Figure 2. deflated mutant individuals display pleitrophic phenotypes. A, B: Dissected pharate adults. A: Wild type pharate adult showing normal abdomen development. B:deflL/deflZ pharate adult shows abnormal abdominal development with absence of posterior tergites and external genitalia (arrows). C: Adult wings showing the abnormal development of the L5 vein in deflL/deflZ adult escapers (arrow). D: Humeral bristles found frequently bent in deflL/deflZ escapers (arrow). E: Egg with normal dorsal appendages and chorion. F: Eggs laid by P{UAS-defl}AV3; deflL/L females frequently display short misshapened dorsal appendages and/or (G) thin translucent chorions. H: Eye-antennal and (I) wing imaginal discs are small and malformed in P{UAS-defl}AV3; deflL/L third-instar larvae.

Download figure to PowerPoint

Although male adult deflP/deflZ, deflL/deflZ, and deflΔ/deflZ transheterozygous escapers are fertile, female escapers demonstrated reduced fertility. Examination of eggs laid by these females showed a high frequency of eggs with thin chorion and/or abnormal dorsal appendage formation. Forty percent of eggs laid by deflP/deflZ females (224/557) and 64% of eggs laid by deflL/deflZ females (35/55) displayed defective chorion and dorsal appendage formation, compared to 7% from wild type females (40/609).

deflΔ/deflZ transheterozygotes were not examined in detail due to these females laying very few eggs. Although the dorsal appendage and chorion defects are the most probable cause of the decreased fecundity observed, additional embryonic defects could not be ruled out, as the frequency of hatching from morphologically normal eggs was also reduced (3% for deflP/deflZ (4/121), compared to 87% (296/340) for wildtype).

To determine whether the mutant phenotypes described above were indeed associated with the deflated locus, transgenic lines carrying a single copy of the deflated cDNA (clone GH11567), under the control of the Gal4-inducable promoter of the UAST vector (Brand and Perrimon,1993), were generated. Even in the absence of Gal4-induced transgene expression, second chromosome insertion lines (P{w+mC=UAS-defl}AV3, P{w+mC=UAS-defl}BA2 and P{w+mC=UAS-defl}BB1) rescued deflL/deflL, deflP/deflP, and deflΔ/deflΔ second instar lethality and allowed development to continue until the larval/pupal boundary or into adulthood without any obvious sex bias (Table 1). The P{w+mC=UAS-defl}BB1 line fully rescued deflL/deflLindividuals and partially rescued deflP/deflP and deflΔ/deflΔ individuals. In the case of deflP/deflP, adult escapers were observed, but at a frequency less than expected from anticipated mendelian ratios. On the other hand, partially rescued deflΔ/deflΔ individuals developed until the white pupae stage only, consistent with the reported homozygous mutant phenotype of the flanking nbs gene (Ciapponi et al.,2006), which is also mutated in deflΔ. The ability of wildtype deflated transgenes to fully or partially ameliorate the phenotypes associated with molecular lesions mapping to the deflated coding region confirms that these phenotypes are indeed associated with perturbation of DEFL function.

Table 1. Rescue of deflated Homozygous Individuals by Low Level Expression of a deflated cDNA Transgene
P{UAS-defl} transgenic lineaPhenotype of homozygous individuals
deflL/deflLdeflP/deflP
  • a

    All tested transgenes were on the second chromosome.

  • b

    nd, not done.

NoneDie at end of 2nd instar stageDie at end of 2nd instar stage
AV3Die as late 3rd instar larvaeNdb
 Contain psuedotumours and small imaginal discs 
BA2Die as white pupaeDie as white pupae
  Contain psuedotumours
BB1100% rescue63% rescue
 Females infertileFemales fertile

Phenotypic defects observed in P{UAS-defl} rescued individuals were observed to correlate with the strength of each allele inferred from the transheterozygous experiments. Homozygous adult deflP females rescued by the P{UAS-defl}BB1 transgene were of normal fertility, whereas P{UAS-defl}BB1 rescued deflL/deflL females were infertile due to chorion and dorsal appendage defects (Fig. 2E–G), which were similar to those seen in eggs laid by transheterozygous (deflP/deflZ and deflL/deflZ) female escapers. In contrast, P{UAS-defl}BB1 rescued adult males of both deflL/deflLand deflP/deflP were fertile. This is consistent with our prior observation that male development requires less DEFL function, on average, than female development. In P{UAS-defl}AV3 partially-rescued deflL/deflL individuals, which only survive to the third instar larval/pupal boundary, melanotic pseudotumours (data not shown) and, upon dissection, amorphic eye-antennal (Fig. 2H) and wing (Fig. 2I) imaginal discs were observed. Differences observed in the efficiency of rescue achieved by P{w+mC=UAS-defl} transgenic lines in the various mutant backgrounds are most likely a reflection of variations in local chromatin environment of the P{w+mC=UAS-defl} transgene insertions affecting the level and timing of defl expression. Increasing DEFL transgene expression via the introduction of GAL4-expressing transgenes also ameliorates these phenotypes and produces adult escapers with varying efficiencies (data not shown).

deflated mRNA Is Expressed in Post-Blastoderm Embryos

To examine the pattern of deflated mRNA expression, we performed in situ hybridisation on wild type embryos and third-instar larval imaginal discs. In imaginal disc material we were unable to detect a deflated mRNA signal above background (data not shown), and in embryos could only detect low levels of staining with an antisense probe (Fig. 3A–H) relative to sense probe controls (Fig. 3I and J). deflated mRNA expression was not observed during the maternally driven syncytial divisions or in early cellularised embryos (Fig. 3A,B). deflated mRNA was detected in early epidermal tissues of cellularised embryos undergoing gastrulation (Fig. 3C,D). In later stages, deflated expression was strongest in the gut, with some expression in the central nervous system (CNS) (Fig. 3E,F) and possibly in the peripheral nervous system (PNS) (Fig. 3G). Expression of deflated was observed to continue late into embryogenesis (Fig. 3H). Our observations that deflated mRNA is expressed at low levels in embryogenesis is supported by previously reported microarray studies showing deflated mRNA does not vary much in its expression, with peaks of expression in late embryogenesis and during pupariation being only twofold greater than the lowest levels of expression observed during development (Arbeitman et al.,2002).

There are two striking features of the deflated expression pattern. The first is that deflated expression is switched on post-cellularisation, not long after zygotic expression occurs on a large scale (Hooper et al.,2007). The second feature is that the majority of deflated expression patterns reflects the general pattern of post-blastoderm cell proliferation during embryogenesis. After cell cycle 16, cells of the epidermis arrest in G1 phase of the cell cycle and do not resume cycling until the first-instar larval stage (Du and Dyson,1999). In contrast, cells of the CNS and PNS continue to undergo mitotic cycles after cell cycle 16. This pattern is similar to what we observe for deflated expression; it appears to drop in the epidermis but continues in the CNS and possibly the PNS. Furthermore, gut cells in the embryo undergo repeated rounds of endoreplication during the second half of embryogenesis (Smith and Orr-Weaver,1991) and we find that deflated expression also persists in these cells. However, we do see lower levels of deflated expression in cells that have stopped proliferating, including the epidermis after cycle 16 (Fig. 3G), and in the gut and epidermis in late stage embryos (Fig. 3H), indicating that deflated expression is not solely restricted to proliferating cells.

GFP-Tagged DEFL Localises to Nuclei in Embryos

In the absence of an antibody to DEFLATED, we generated transgenic lines of two constructs in which GFP was fused in frame to the C-terminus of DEFL, one cloned in pUAST (denoted P{w+mC=UAS-defl::gfp}) and the other in pUASP (denoted P{w+mC=UASp- defl::gfp}), allow ng optimal somatic and germline expression, respectively. Transgenic lines expressing these constructs were tested for their ability to rescue the second-instar lethality of deflated homozygote mutants. All transgenes tested were capable of rescue. In particular, two independent P{w+mC= UAS-defl::gfp} lines could fully rescue deflL/deflL individuals and one P{w+mC= UASp-defl::gfp) line could rescue deflΔ/deflΔ individuals to the pupal stage. These data are very similar to the rescue capabilities of untagged DEFL, indicating that the presence of GFP at the C-terminus does not significantly interfere with DEFL function. To examine the cellular localisation of GFP-tagged DEFL, we expressed UASP P{w+mC=UASp- defl::gfp} in embryos by nanos-Gal4 induced expression. The gross development of these flies was normal, indicating that increased levels of DEFL protein during oogenesis and embryogenesis do not adversely affect development. We examined the DEFL::GFP-expressing embryos by live confocal microscopy and observed DEFL::GFP localisation to be similar at all embryonic stages examined, with strong nuclear localisation and weaker localisation to the cytoplasm, a result seen more clearly in syncytial embryos where there are fewer nuclei (Fig. 4). This localisation is consistent with the observed association of human DEFL homologue (Ints7) with RNA polymerase II and proteins implicated in snRNA processing, both of which occur in the nucleus.

thumbnail image

Figure 4. DEFLATED::GFP localises mainly to nuclei in embryos. A: A syncytial embryo showing strong DEFL::GFP localisation to the nucleus and weaker localisation to the cytoplasm. B: The predominantly nuclear localisation of DEFL::GFP is retained in older embryos as indicated in the (B') higher magnification insert. Embryos were overexpressing DEFL::GFP in the germline induced by nanos-GAL4 expression in a P{w+mC=UASp-defl::gfp} background.

Download figure to PowerPoint

Overexpression of deflated Results in Bristle and Wing Defects

Since deflated mRNA is expressed at low levels (Fig. 3; Arbeitman et al.,2002), we reasoned that overexpression would be likely to result in an observable phenotype. We expressed both the untagged and GFP-tagged DEFL proteins from Gal4-inductable transgenes using a number of lines that express Gal4 in various tissues and developmental stages. Unexpectedly, we only observed discernable phenotypes when two copies of the deflated cDNA were overexpressed at 29°C (elevated temperatures increase GAL4-induced expression). Even under these conditions, the majority of Gal4 lines used to induce DEFL expression produced no obvious phenotype. However, notable effects were observed on wing development, where strong overexpression driven by spalt-Gal4, MS1096-Gal4, and scalloped-Gal4 resulted in wings that had an extra posterior cross-vein or were small and crumpled (Fig. 5B–D). Low-level ubiquitous drivers such as armidillo-Gal4 and escargot-Gal4 resulted in a mild phenotype where machrochaetae on the notum, particularly the scutellum, were frequently duplicated (Fig. 5F and G). In addition, an unusual phenotype was observed with the driver T1096-Gal4, where the notum displayed two laterally positioned fissures (Fig. 5H). These data indicate that excess DEFL can be tolerated by most cells, with the wing and notum being the most sensitive structures.

thumbnail image

Figure 5. Overexpression of deflated results in wing and bristle defects. A: A wildtype wing with the five longitudinal veins (L1–L5) and the two cross veins (anterior, ACV; posterior, PCV) labelled. B–D, F–H: Overexpression of P{UAS-defl}AV3, P{UAS-defl}Z1 at 29°C resulted in wing and bristle defects. B: Overexpression by spaltmajor-Gal4 resulted in an ectopic anterior cross vein (asterisk). C, D: Overexpression by MS1096-Gal4 and scalloped-Gal4 resulted in a reduced and crumpled wing. E: A wild type thorax showing the characteristic pattern of macrochatae. The curled macrochatae are a result of the preparation for scanning electron micrography and do not represent pre-fixation bristle morphology. F, G: Overexpression by escargot-Gal4 and armadillo-Gal4 resulted in the duplication of macrochatae (blue arrows). H: Overexpression by T1096-Gal4 resulted in the formation of “scar”-like fissures on either side of the thorax (red arrows).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

In this study, we describe an investigation of the developmental roles of deflated, the Drosophila homologue of the Integrator complex subunit Ints7. Like previous reports of loss of Ints7 homologue function in zebrafish and C. elegans, we find that deflated is essential for normal development. Mutation of the zebrafish defl homologue (Dkfzp434b168) and knockdown of the C. elegans defl homologue (2L877) by RNAi leads to craniofacial and locomotor defects (respectively) and lethality (Golling et al.,2002; Kamath et al.,2003). As these reports were of large-scale screens, phenotypes and underlying cellular defect(s) were not examined in detail, Our data demonstrate that deflated is also required for early Drosophila development and, as a result of investigating the phenotypes of an allelic series, that DEFL function is required at multiple stages throughout development. That either the whole Integrator complex, or some subcomplexes, play essential roles in development is consistent with the observation that targeted disruption of the Ints1 homologue in mice causes growth arrest and apoptosis in early blastocyst embryos (Hata and Nakayama,2007). Whether these essential roles are due to activity of the whole Integrator complex, or certain subcomplexes, still needs to be determined.

The phenotypes observed in Drosophila as a result of reduction or absence of DEFL indicate the principal defects are in cell proliferation and cell signalling. deflated mutants show abdominal defects, melanotic tumours, small imaginal discs, and eggs with abnormal chorion and dorsal appendages. These phenotypes are similar to those observed for the S-phase regulators E2f and Dp (Royzman et al.,1997,1999,2002; Myster et al.,2000; Bosco et al.,2001; Cayirlioglu et al.,2001; Frolov et al.,2001; Royzman et al.,2002). Similar abdominal defects are also observed in mutants of M-phase regulators myc and cdc2, where the abdominal histoblasts do not maintain a proper G2 arrest and instead re-enter the cell cycle (Stern et al.,1993; Hayashi and Yamaguchi,1999; Fung et al.,2002). Similar phenotypes affecting embryonic dorsal appendages, wing vein morphology, and bristle duplications have also been observed following disruption in cell signalling pathways (Tomoyasu et al.,1998; Wharton et al.,1999; Culi et al.,2001; de Celis,2003). Our finding that deflated may be required for cell signalling is further supported by reports that implicate deflated in the negative regulation of both the Imd innate immunity and the RTK/ERK signalling pathways (Foley and O'Farrell,2004; Friedman and Perrimon,2006). The finding that loss of deflated results in a similar phenotypic spectrum suggests that the processes of cell proliferation and cell signalling may be the most sensitive to proper snRNA processing.

Alternatively, DEFL may perform a separate biochemical function and not be directly involved in snRNA biosynthesis despite being intimately associated with RNA pol II CTD and other proteins involved in snRNA production. The observation that DEFL/Int7 is not found in unicellular eukaryotic organisms, which nevertheless produce snRNAs, supports this contention. The Drosophila zygotic genome is not transcribed on a large scale until just after cellularisation (Hooper et al.,2007), which is not long before deflated is first expressed. Our observations that deflated mRNA expression correlates with known patterns of proliferation in postblastoderm embryos, and that loss of deflated function results in phenotypes that are similar to those seen in other proliferation mutants, suggests the possibility that deflated plays a role in coordinating transcription with cell-cycle progression in response to certain cell–cell signaling stimuli. The finding that deflated homologues share an evolutionary conserved correlation of expression with proliferation regulators Dp and cdt4 (Stuart et al.,2003) supports this assertion. This is consistent with evidence suggesting that pre-mRNA splicing is linked to cell-cycle progression, as components of the pre-mRNA splicing machinery are targets of the S-phase cyclin Cyclin E in mammals (Seghezzi et al.,1998) and in Drosophila mutation of the SR protein splicing factor B52 can restore S-phase in de2f1 mutant cells by preventing the correct splicing of de2f2 pre-mRNA (Rasheva et al.,2006). A role for DEFL/Ints7 in cell-cycle regulation is also supported by a recent study of the human Ints7 promoter, which revealed negative regulation in response to increased B-Myb, c-Myb, and p53 expression (Nakagawa et al.,2008).

This study examined in detail the developmental requirements of a DEFL/Ints7 homologue. We have demonstrated that deflated is essential for development and functions at multiple stages. deflated mRNA expression is low in embryos and as such deflated mutants can be rescued by low-level leaky expression from cDNA transgenes. GFP-tagged protein localises mainly to nuclei, which is expected based on the functional association of the Integrator complex with RNA polII and snRNA biogenesis. Loss of deflated results in phenotypes that can be explained by perturbations in cell proliferation and cell signalling, implying that these cellular process may be either directly controlled by DEFL/Ints7 or indirectly via snRNA availability.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Fly Stocks and Generation of deflated Alleles

All flies were grown at 25°C on standard cornmeal-treacle media. The stocks w1118; P{w+mC=EP}EP3301 (BDGP Project Members, 2000-), w;; TM3 Sb[1]/TM6B p[Xp] Tb[1] and w[*];; Dr[1]/ TMS P[ry[+t7.2]=Delta2-3} 99B were obtained from the Bloomington Stock Centre. For overexpression we used the following Gal4 expressing lines; nanos-Gal4p{w[+mC] Scer\GAL4[nos.UTRT:Hsim\VP16] =GAL4::VP16-nosUTR}, MS1096-Gal4 P[GawB]Bx[MS1096], armadillo-Gal4 P[GAL4-arm.S], spalt-Gal4, scalloped-Gal4; If/CyO P{GawB}sd[SG29.1], T1096-Gal4, and escargot-Gal4 (Tweedie et al.,2009). deflated alleles were generated by P-element local hopping and detected by a multiplex PCR screen with primers to the P-element inverted repeat sequence (PF2 5′-CGACGGGACCACCTTATGTTAT-3′) and to the deflated genomic region (gscreena 5′-TTAACTATGGCTTGGAATACGCCTACA-3′; gscreenb 5′-ATTGTGTACCAATTCCAAGGAGCAGTT-3′; gscreenc 5′-ATCTGCGATAAATGTGAGTGAG- GAAAC-3′; gscreend 5′-GTAGGGGAATGTCACAATCGGCTTAT-3′; gscreene 5′-TTTGGTCTTCCGTGTTATGTGCT- TACT-3′; gscreenf 5′-GCTTGTTCACAGAATCCTCAAGAGAAG-3′; and gscreeng 5′-ATTTGGCCAGGGATCTGAAAAA- CTT-3′). Precise mapping of the alleles was performed either through inverse PCR (Huang et al.,2000) using the restriction enzyme MspI and the primers 752 (5′-GCATGTCCGTGGGGTTTGAATTAAC-3′) and 829 (5′CTGTCTCACTCAGACTGAATACGACACTCAG-3′) or the generation of PCR products using a primer that binds the genomic region at the very start of the CG18177 coding region (5′-TACCGGTACTAGTACTCACAGTGTGAATTGTGCC-3′) and a primer specific to either the 3′ (5′-CTGTCTCACTCAGACTCAATACGACACTCAG-3′) or 5′ (5′-CGTCCGCAGACAACCTTTCCTC-3′) end of the P-element depending on the orientation of the P-element. The 3′ or 5′ primer was then used in the subsequent sequencing reaction of the PCR product.

Generation of deflated cDNA Transgenic Flies

The deflated cDNA clone (GH11567) was obtained from Berkeley Drosophila Genome Project. It was subcloned into pUAST via EcoR1 and Xho1 restriction enzyme sites. To generate the deflated::gfp-tagged constructs, the deflated cDNA was amplified by PCR to generate a product that introduced a Kpn1 site 5′ and a HindIII site 3′ and also abolished the stop codon (primers used were 5′-ATCTGCGGTACCTGTGAGTGAGGAAAC- CTG-3′ and 5′-CCCCCAAGCTTATAAACCTCCTCGTCTGTCCC-3′). This fragment was cloned into pALX190 to fuse the gfp coding region to the 3′ end of the deflated cDNA. The fused deflated cDNA and GFP-coding regions were then excised using the restriction enzymes Kpn1 and Xba1 and subcloned into both pUAST and pUASP via these same sites. Transgenic flies were created using standard procedures (Spradling,1986).

RNA In Situ Hybridisation

mRNA in situ hybridisation to embryos was performed according to Lehner and O'Farrell (1990), except hybridisations were performed at 55°C and tRNA was omitted from the hybridisation. Template for the riboprobe synthesis was generated by PCR using genomic DNA as template and the primers deflfT7 (5′-TAATACGACTCACTATAGGCTG- CTCCTGATTCCCGTGT-3′) and deflr (5′-TGCGGTGGAAAAGGCGTACT-3′) for the anti-sense probe and deflrT7 (5′-TAATACGACTCACTATAGGTGC- GGTGGAAAAGGCGTACT-3′) and deflf (5′-GCTGCTCCTGATTCCCGTGT-3′) for the sense probe. The riboprobe was synthesized and detected by use of the DIG labelling and detection kits (Roche) according to the manufacturer's directions.

Microscopy

Light micrographs were taken on an Olympus BX51 microscope or an Olympus SZ60 dissecting microscope equipped with an Olympus DP50 digital camera. Fluorescent images were taken on a BioRad 2000 laser scanning confocal microscope with Lasersharp 2000 (ver. 5) software. Wings were prepared for microscopy by soaking whole flies in dissection solution (50% ethanol, 50% lactic acid) at least overnight. Wings were dissected and mounted in dissection solution and imaged immediately. Flies were prepared for scanning electron microscopy as follows: dehydrated through a 25, 50, 75, and 100% aqueous acetone gradient, air-dried, mounted onto metal stubs, and coated with platinum. They were imaged using a Joel JSM 5410LV scanning electron microscope with “SemAfore” software. All images were processed by ImagePro and Graphics Converter.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

This work was generously supported by a grant to W.D.W. from the Queensland Cancer Fund. We thank the Bloomington Stock Centre and Gary Hime for fly lines, the Berkeley Genome project for the deflated cDNA clone GH11567, Metaxia Vlassi for assistance in bioinformatics, and Kevin Blake for assistance with the scanning electron and confocal microscopes.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES