A Forward Genetic Approach to Mapping a P-Element Second Site Mutation Identifies DCP2 as a Novel Tumor Suppressor in Drosophila melanogaster

The use of transposons to create mutations has been the cornerstone of Drosophila genetics in the past few decades. Second-site mutations caused by transpositions are often devoid of transposons and thereby affect subsequent analyses. In a P-element mutagenesis screen, a second site mutation was identified on chromosome 3, wherein the homozygous mutants exhibit classic hallmarks of tumor suppressor mutants, including brain tumor and lethality; hence the mutant line was initially named as lethal (3) tumorous brain [l(3)tb]. Classical genetic approaches relying on meiotic recombination and subsequent complementation with chromosomal deletions and gene mutations mapped the mutation to CG6169, the mRNA decapping protein 2 (DCP2), on the left arm of the third chromosome (3L). Thus the mutation was renamed as DCP2l(3)tb. Fine mapping of the mutation further identified the presence of a Gypsy-LTR like sequence in the 5′UTR coding region of DCP2, along with the expansion of the adjacent upstream intergenic AT-rich sequence. The mutant phenotypes are rescued by the introduction of a functional copy of DCP2 in the mutant background, thereby establishing the causal role of the mutation and providing a genetic validation of the allelism. With the increasing repertoire of genes being associated with tumor biology, this is the first instance of mRNA decapping protein being implicated in Drosophila tumorigenesis. Our findings, therefore, imply a plausible role for the mRNA degradation pathway in tumorigenesis and identify DCP2 as a potential candidate for future explorations of cell cycle regulatory mechanisms.

Introduction of a functional copy of Rab11 in the homozygous mutant background rescued the ommatidial defect, but failed to rescue the tumorous phenotype and lethality, implying the presence of a second-site mutation, besides the P-element insertion in Rab11. Allowing free recombination, the mutations, viz., the P-element insertion in Rab11 and the second-site mutation devoid of the P-element, were separated into two discrete isogenised lines (Alone and Roy, unpublished). Since the phenotypes shown by the second-site mutation phenocopies mutations in tumor suppressor genes in Drosophila (Gateff and Schneiderman 1969;Gateff 1994;Gateff E, 1978), the mutation was named as l(3)tb [lethal (3) tumorous brain] owing to its location on the third chromosome and the phenotypes manifested. Genetic and molecular analyses mapped the mutation to DCP2 on the left arm of chromosome 3 (cytogenetic position 72A1) and hence the allele was named as DCP2 l(3)tb .
DCP2 codes for the mRNA decapping protein 2, which belongs to the NUDIX family of pyrophosphatases and was identified more than a decade ago through a yeast genetic screen (Dunckley and Parker 1999). Being one of the major components of the decapping complex, DCP2 is conserved in worms, flies, plants, mice, and humans (Wang et al., 2002). DCP2 is activated by DCP1 and they function together as a holoenzyme to cleave the 59 cap structure of mRNA (LaGrandeur and Parker 1998;Coller and Parker 2004;Parker and Song 2004;She et al., 2008).
In humans, DCP2 is upregulated in lung cancer (Watson et al., 2008) and regulates histone mRNA titres (Mullen and Marzluff 2008) and there is some evidence that DCP2 is regulated by the p53-DREAM pathway (Engeland 2018). Recently, Gaviraghi and coworkers showed that the co-activator of mRNA decapping, PNRC1, shuffles the mRNA decapping complex to the nucleolus from the P-bodies, wherein decapping of certain snoRNAs involved in processing of ribosomal RNA is initiated, thereby restricting the oncogenic potential of cells (Gaviraghi et al., 2018;Mugridge and Gross 2018).
In Drosophila, DCP2 plays fundamental roles in the response pathways active following chronic nicotine exposure and its loss mediates locomotor hyperactivity following such exposure (Ren et al., 2012). Also, DCP2 is involved in the decapping of mRNAs involved in cell cycle progression and DNA replication and acts to restrict bunyaviral infections in cell lines and adult flies (Hopkins et al., 2013).
Despite the abundance of information from various model organisms pertaining to the roles of the mRNA decapping machinery, and/or the decay programs in cellular homeostasis, development and stress management, our understanding of the roles of the components of the mRNA decay program(s) and their interaction with the cellular signaling modules vis-à-vis gene expression programs is still in its infancy. Although mRNA decapping plays a significant role in mRNA turnover and translation, widely affecting gene expression (Mitchell and Tollervey 2001;Raghavan and Bohjanen 2004;Song et al., 2010), simultaneous links between mRNA degradation genes, retrotransposons and tumors have not been observed and/or investigated so far. Therefore, the novel allele DCP2 l(3)tb reveals a new perception for functional roles of mutant lesions and the ensuing perturbations in gene regulation in tumor biology.
Analysis of lethal phase in l(3)tb homozygotes For analysis of lethal phase and morphological anomalies associated with the homozygous l(3)tb mutation, embryos were collected at the intervals of 2h on food filled Petri dishes. Embryos from wild type flies were collected as controls. The total number of eggs in each plate was counted and the embryos were allowed to grow at 23°or 18°or 16°(61°). Hatching of embryos and further development of larval stages was monitored to determine any developmental delay. Mutant larvae, at different stages, were dissected and the morphology of larval structures was examined.
F2 virgins were then crossed to ru Pri ca/TM6B males to score the frequency of recombinants in the F3 progeny. Thereafter, all the F3 progeny males obtained, were individually scored for ru, h, th, st, cu, sr, e and ca phenotypes and then they were individually crossed with virgin l(3)tb/TM6B females to identify which of them had the l(3)tb mutation along with other scored markers.
Complementation mapping of the l(3)tb mutation: Complementation analysis of the mutation in l(3)tb allele was carried out in two stages. First, deficiency stocks spanning the entire chromosome 3 (Supplementary Table S1) were used to identify the mutant loci, and second, lethal P-insertion alleles selected from the region narrowed down through recombination and deficiency mapping (Supplementary Table S2) were harnessed to further identify the mutant gene(s) in l(3)tb. In either case, virgin females of y w; +/+; l(3)tb/TM6B,Tb were crossed with the males of the various deficiency stocks and/or the lethal P-insertion alleles and the nontubby F1 males heterozygous for l(3)tb and the deficiency were scored for the phenotype(s).

Fine mapping of l(3)tb mutation
Genomic DNA Isolation, PCR and Southern hybridization: Genomic DNA for polymerase chain reaction (PCR) was isolated by homogenizing 50 male flies from each of the desired genotype or 80-100 third instar larvae from homozygous mutant l(3)tb (Sambrook et al. 1989). Based on the results obtained from genetic mapping, identification of the candidate region in DCP2 was done by overlapping PCR based screening, wherein the entire genomic span of DCP2 was amplified using 28 primer pairs from 3L:15811834..15819523 (Supplementary Tables S4, S5, S6). After identifying the candidate region, it was validated with the primer pair, Dbo_F: 59-ACAACATTCACTCCATGGAACACCT-39 and DCP2_P19_R: 59-TGCTCACCGAACTTTTTCGCGATCT-39. The primer pair DCP2_F: 59-ATAACAAAAAAGTTATGGTACCACCCCCGCG-TTGTATTCT-39 and DCP2_R: 59-AGATTTCGATGTATATGGATC-CGTCCCAACCTTTGCGTCT-39 was designed to amplify the full length gene along with flanking sequences (500 bp on either side). In either case, the thermal cycling parameters included an initial denaturation at 96°(2 min) followed by 30 cycles of 30 s at 94°, 45 s at 72°, and 15 min at 68°. Final extension was carried out at 68°for 20 min. The PCR products were electrophoresed on 0.8% agarose gel with O'GeneRuler 1 kb plus DNA ladder (Thermo Scientific, USA). An 812 bp region (3L: 15825979..15826790) spanning the candidate mutated region in DCP2 was PCR amplified and ligated in pGEM-T vector (Promega) to generate the pGEM-T-812 clone. The $430 bp fragment isolated during primer walking (see below) was purified and ligated in pTopo-TA-XL vector (Invitrogen, USA) to generate the pTopo-TA-XL-430 clone. Digestion, ligation and transformation were performed using standard protocols as described in Sambrook et al. 1989. Southern hybridizations were performed according to Russell and Sambrook 2001. Following electrophoresis and gel pre-treatments, DNA was transferred on to positively charged nylon membranes (Roche, Germany). Hybridizations were performed at 68°with 0.02% SDS, 5X SSC, 0.5% Blocking reagent, and 0.1% laurylsarcosine with probes generated from pGEM-T-812 and pTopo-TA-XL-430 plasmids. DIG labeling and chemiluminescent detection were performed as per the manufacturer's instructions (Roche, Germany). Alternatively, radiolabeling of probes using a-32 P dCTP (BRIT, India) and detection were performed according to Sambrook et al., 1989. Sequencing, primer walking and CNV detection: To confirm the fidelity of amplification automated DNA sequencing was performed (ABI -3130, USA) as per the manufacturer's instructions. Primer walking was initiated with the primers dbo_F and DCP2_P19_R and from the terminal part of the sequence obtained, new primers P19_W2_F 59-GGAGATCTGTTTGAAATATCTCTTCACATT-39 and P19_W2_R 59-GGCGCGTCAGCATTGTTCATACAAAGC-TAC-39 were designed. Long-range PCR with P19_W2 was performed as described previously. Sequence chromatograms were assessed and analyzed with FinchTV 1.4.0, Geospiza Inc. Semi-quantitative assessment of copy number variance (CNV) of the intergenic sequence in DCP2 l(3)tb was determined through PCR analyses. A 156 bp sequence (3L: 15826497..15826652) was chosen to be amplified by CNV_F 59-ACAGTTGGCTCTGTGATAAATGT-39 and CNV_R 59-AGTGCAACGGAAGGGAATCT-39 against an internal control sequence of 153 bp, corresponding to the gene Dsor, amplified by the primer pair Dsor_F 59-CCACCCTGGAGTC-GATATTC-39 and Dsor_R 59-GTCCTTGAACTCGGTGGAGA-39. Thermal cycling parameters included an initial denaturation at 95°( 5 min) followed by 28 cycles of 30 s at 94°, 30 s at 60°, and 30 s at 72°. Final extension was carried out at 72°for 10 min. The PCR products were electrophoresed on 2% agarose gel with a 100-bp DNA ladder (BR Biosciences, India).

RNA isolation and RT-PCR
For semi-quantitative reverse transcription-polymerase chain reaction (RT-PCR), total RNA was isolated from the 50 brains from wandering late third instar larvae from the genotypes mentioned in the respective cases in the results section, using TRI Reagent following the manufacturer's protocol (Sigma-Aldrich, India). RNA pellets were dissolved in nuclease-free water, followed by treatment with 2U of RNase-free DNaseI (Thermo Scientific, USA) for 40 min at 37°. First strand cDNA was synthesized using 200 U of RevertAid reverse transcriptase (Thermo Scientific, USA) and 80 pmol of oligo(dT) 18 primer (Thermo Scientific, USA). PCR was carried out for Cyclin A (amplicon size 77 bp), Cyclin E (amplicon size 163 bp) and Diablo (amplicon size 179 bp) using G3PDH (amplicon size 144 bp) as internal control. In either case, the thermal cycling parameters included an initial denaturation at 95°(5 min) followed by 30 cycles of 30 s at 94°, 45 s at 60°, and 1 min at 72°. Final extension was carried out at 72°for 10 min. The PCR products were electrophoresed on 1.8% agarose gel with a 100 bp DNA ladder (BR Biosciences, India). The forward and reverse primers used are listed below. Immunocytochemistry and Immunoblotting The imaginal discs and/or brain ganglia were collected from wild type Oregon R + wandering 3 rd instar larvae, just before pupation (110 h AEL) and in mutant homozygous l(3)tb from day 6 and day 10/12. The tissues were processed for immunostaining/immunoblotting as described in Banerjee

RESULTS
l(3)tb homozygotes show the classic hallmarks of cancer in Drosophila including developmental delay, abnormal karyotype, larval/pupal lethality along with tumorous brain and wing imaginal disc Developmental analysis of l(3)tb homozygotes showed that while embryos hatched normally and developed alike heterozygous siblings [l(3)tb/TM6B], the third instar larvae reached the wandering stage quite late with the larval stage extending up to 12 or 13 days ( Figure 1B). Although 66.8% of the larvae survived to pupate (Table 1), they died in the pupal stage following bloating, increase in size and cessation of growth ( Figure 1A). Hence, the mutation is lethal, with the lethality being pronounced in the pupal stage.
Lowering the temperature to 16°or 18°reduced the larval mortality, causing 96% of larvae to pupate but did not improve pupal survival (Figure 1 C and D). Analysis of larval brain and imaginal discs in the homozygotes in the early (Day 6) and late (Day 10-12) larval phase showed gross morphological alterations in the size of the larval brain, wing and eye imaginal discs (Figure 2A-G) as compared to the wandering wild type third instar larvae (115 h after larval hatching [ALH]). The brain was smaller in size than the wild type (Oregon R + ), or heterozygous [l(3)tb/TM6B] individuals till 115 h ALH but started showing aberrant growth in the dorsal lobes thereafter, with significant differences in the diameter and area of the lobes. There was no loss of symmetry in the overgrown brain hemispheres in most of the cases, except in some cases where they were deformed and fused with the imaginal discs (Figure 2 J and K). A similar trend in morphological aberration was observed in the wing discs, which remained small initially, but later enlarged sufficiently ( Figure  n■

Figure 2
Morphological and chromosomal alterations in l(3)tb homozygotes. Homozygous l(3)tb mutants show severe morphological alteration in delayed 3 rd instar larval brain, wing and eye-antennal disc. Homozygous l(3)tb mutant 3 rd instar larvae revealed tumorous brain of day 12 (C) as compared to day 6 of homozygous mutant (B) and day 5 of wild type, Oregon R + (A). l(3)tb homozygotes exhibited highly significant differences in the overall circumference of the left and right brain lobes in the delayed stage (day 10) as compared to the respective wild type brain lobes (J). Significant differences were found in the area (mm 2 ) of respective brain lobes of l(3)tb homozygotes and wild type (K). Dimensions of wing and eye-antennal imaginal discs of delayed 3 rd instar larvae from homozygous l(3)tb mutant revealed significant increase in size (D,E,F,G). Length and breadth of wing discs from 3 rd instar larvae of l(3)tb mutant of day 6 were found to be smaller than the wing imaginal discs from wild type, but wing discs from extended larval period (day 10) showed significant increase in size (L). Metaphase chromosome preparation of brain cells (H) from wild type and l(3)tb homozygotes exhibited abnormal karyotypes (I) where l(3)tb homozygotes showed less condensed and extended chromosome morphology as compared to wild type, Oregon R + . ÃÃÃ denotes P , 0.005.
Eye-antennal discs and leg imaginal discs also show morphological and developmental anomalies in l(3)tb homozygous individuals Global analysis of morphological aberrations in the mutant homozygotes showed that besides the tumorous brain and wing imaginal discs, eye-antennal discs and leg imaginal discs were also overgrown with a transparent appearance. Expression of Elav and Dlg in the eye-antennal discs revealed similarities to the developmental perturbations observed in the wing discs and brain ( Figure 4S, V). In the early third instar mutant larvae (Day 6), all photoreceptor cells showed expression of Elav, similar to the wild type tissue ( Figure 4J and N). However, during advanced stages of larval tumorigenesis (Day 10), it dwindled eventually ( Figure 4R). The Elav expressing cells, posterior to the morphogenetic furrow, co-express Dlg and demonstrate the typical ommatidial Figure 3 Tumorous tissues in l(3)tb homozygotes show enhanced mitotic potential. Enhanced mitotic potential observed in the tumourous tissues of homozygous l(3)tb as shown in larval whole brain (A), brain lobes (D, G) and wing imaginal discs (S) immunostained for PH3, a potent mitotic marker. Distribution of PH3 labeled cells counter stained with DAPI cells in wild type (A) and homozygous l(3)tb (Day 6 and Day 10) larval brain (D, G) and also in wild type brain lobes (B, C) and homozygous mutant brain lobes (E, F for day 6; H, I for day 10) indicated high mitotic index as compared to wild type. Similarly, more mitotic positive cells were seen in tumorous wing imaginal discs (day 10) of homozygous mutant l(3)tb (S) as compared to wild type, Oregon R + (P) . DIC images (C, F, I and R, U) illustrate external normal morphology in wild type and more pronounced tumorous phenotypes in homozygous l(3)tb larval brain and wing imaginal discs. Quantitative analysis showed increase in the number of mitotic positive cells in homozygous mutant larval brain lobes and wing imaginal discs as compared to wild type and the difference was highly significant (V). The images are projections of optical sections acquired by confocal microscopy. Staining was done in triplicates with 10 brains and 15 wing imaginal discs in each group. Significant difference is represented as ÃÃÃ P # 0.005 using one-way ANOVA.
arrangement. In the mature mutant larvae, however, the eye discs demonstrate significant deviations from the usual regular arrangement of ommatidia. The leg imaginal discs, which reside close to the brain and wing imaginal discs also show enlargement in size that increases with advancement of larval stage. They show gradual disruption of normal expression of DE-cadherin and Armadillo ( Figure 5), alike tumorous wing discs (see above), implying the mutation and subsequent tumor to affect developmental homeostasis in adjacent tissues as well. 3 rd instar larval brain shows intense staining of Elav (green) in day 6 (D) of homozygous mutant later on show loss of staining in enlarged brain of day 10 (G), while the wild type brain (A) showed normal pattern of Elav staining. Dlg stained the ventral nerve chord and central brain in optic lobes of wild type (B), which is similar to that observed in day 6 of homozygous mutant brain (E) but in delayed larval brain, (day 10), the pattern was altered (H). Scale shown is 100mm. Neuronal tissue from eye imaginal discs also display loss of neurons seen through Elav staining in day 10 (R) as compared to day 6 (N) in homozygous l(3)tb mutant as well as to wild type (J). Pattern of junctional protein, Dlg, in eye imaginal discs is also altered in day 10 (S, V) as compared to day 6 (O) and wild type (K). Counter stain with DAPI shows very weak intensity in day 10 (T) reflecting disintegrating chromatin as compared to day 6 (P) and wild type (L). Scale bar represents 10 mm.
Genetic mapping through meiotic recombination and complementation mapping identify l(3)tb to be allelic to DCP2 The mutation l(3)tb, being recessive and on the third chromosome, was maintained with TM6B balancer. Analysis of meiotic recombination frequencies of an unmapped mutation with known markers is a classical technique that has been routinely employed to identify its cytogenetic position. In order to bring l(3)tb in a chromosome with such markers (ru cu ca), we allowed meiotic recombination to occur between l(3)tb and the eight recessive markers present on the "rucuca" chromosome (Table 2).
113 recombinant males were observed and recombination frequencies were calculated in centiMorgan (cM). Table 3 shows the recombination frequencies of each marker (locus) with the mutation l(3)tb. Preliminary analysis suggested that l(3)tb was close to thread (th) with minimum recombination events between the two loci (2.65%). Further analysis of recombination events be-  Complementation analysis with molecularly defined Drosdel and Exelixis deficiency lines (N = 85), spanning the entire chromosome 3, identified four lines which failed to complement the mutation, viz., Df(3L)BSC774, Df(3L)BSC575, Df(3L)BSC845 and Df(3L)RM95, which was generated in the lab using progenitor RS stocks. Transheterozygotes l(3)tb/Df(3L)BSC575 were pupal lethal and the dying non-tubby larvae showed phenotypes similar to l(3)tb homozygotes, suggesting the mutation to reside between 71F1 n■ and 72A1 on the left arm of chromosome 3. Further analysis using six deletion lines belonging to the above region (71F1-72A2) identified the mutation to reside between 71F4 to 71F5, which, strangely, is a gene desert region. Complementation analyses performed with lethal insertion alleles (N = 26) of genes residing proximal or distal to 71F4-F5 identified two lethal P-element insertion alleles of DCP2 (mRNA decapping protein 2; CG6169), viz., P{GT1}DCP2 BG01766 and PBac {RB}DCP2 e00034 , which failed to complement the l(3)tb mutation ( Figure 7A and C), as well as those deletions which had failed to complement l(3)tb, implying the mutation to be allelic to DCP2 (72A1).  Trans-heterozygotes of DCP2 mutants and l(3)tb show developmental delay, tumorous larval brain with elevated neuroblast numbers, larval/pupal lethality and developmental defects in escapee flies Trans-heterozygotes of l(3)tb with two known alleles of DCP2, viz., P{GT1}DCP2 BG01766 and PBac{RB}DCP2 e00034 , showed developmental delay and third instar larvae persisted in the same stage till Day 10 ALH ( Figure 6B and D), and showed tumorous phenotypes of brain and wing imaginal discs (Supplementary Figure S4), similar to the l(3)tb homozygotes. The expression of Deadpan (Dpn), a marker for neuroblasts, shows an increased number of neuroblasts in the larval brain of the trans-heterozygotes as well as l(3)tb homozygotes ( Figure 7F, K and P). Also, the trans-heterozygous progeny showed a higher mitotic index than the wild-type progeny, similar to the l(3)tb homozygotes ( Figure 7G, L and P). While PBac{RB}DCP2 e00034 /l(3)tb was found to be 100% pupal lethal, P{GT1}DCP2 BG01766 /l(3)tb was only 81.6% lethal ( Figure 6A and B), with the rest 18.4% pupae eclosing as flies. However, the escapee flies showed several developmental abnormalities, viz., defects in the wing (9.5%), thorax closure (3.2%), loss of abdominal para-segments and abdominal bristles (3.2%), and presence of melanotic patches (22.2%), leg defects (41.3%) or eclosion defects (12.7%) (Supplementary Figure  S2). Analysis of compound eyes in these escapees revealed complete loss of regular arrangement of ommatidia and ommatidial bristles   Figure S3). Abnormal external genitalia were also observed in the male escapees (data not shown). Subsequent analysis showed that the trans-heterozygous escapee flies had compromised fertility, with only 40% of the males and 21.7% of the females being fertile (Table 5). The similarity in the pattern of development and the defects associated between the l(3)tb trans-heterozygotes and homozygotes provide strong genetic proof of allelism between l(3)tb and DCP2.

DCP2 l(3)tb is an insertion allele of DCP2
Fine mapping, performed by overlapping PCR, identified the region (Supplementary Figure S5), amplified by the primer pair, Dbo_F, and DCP2_P19_R, to span the candidate region. The region, which is 945 bp (3L: 15826279..15827223) in the wild type and comprises of 59UTR coding region of DCP2, the adjacent intergenic region and the proximal part of the neighboring gene, dbo, showed absence of amplification in the DNA of l(3)tb homozygotes, highlighting it as Figure 9 The DCP2 l(3)tb genome harbors Gypsy-LTR like sequence in 59UTR coding region of DCP2 and expansion of adjacent upstream intergenic AT-rich sequence along with low expression of the neighboring gene, dbo. Gel electrophoretogram showing PCR amplification profile obtained by the primers used (P19_W2_R and P19_W2_F) for the "second step" of walking (B). The alignments of the sequences uncovered in the "first step" are shown as thin lines colored as per homology with the wild type sequence. The region amplified here lies subsequent to the sequence uncovered by the initial primers (A; Dbo_F and Dcp2_P19_R, shown in red arrows). Mentioned alongside the electrophoretogram are the semi-logarithmic estimates of the amplicon size. Shown alongside is the blot of the same hybridized with the probe generated from the $430 bp amplicon. Semiquantitative PCR to detect change in copy number of the intergenic region in the DCP2 l(3)tb genome (C) shows increased amplification of the intergenic sequence in DCP2 l(3)tb genome as compared to the Dsor (control) amplicons in both the genomes. Shown in D is a histogram comparing the fluorescence intensity of PCR amplicons obtained from amplification of the intergenic sequence and the control sequence from the DCP2 l(3)tb genome and the wild type genome. The schematic in E shows the architecture of the mutant allele, DCP2 l(3)tb based on the results obtained from fine mapping. Semi-quantitative RT-PCR analyses of transcription from DCP2 and dbo in the wild type and DCP2 l(3)tb homozygotes (F) shows decreased titer of mRNA from both genes in the tumorous individuals.
the candidate lesion. Long-range PCR using the same pair of primers revealed a large amplicon of $8.5 kb in the mutant against the 945 bp amplicon in the wild type genome, subjected to the same thermal cycling parameters ( Figure 8B). Amplification of the full-length gene DCP2 using primers residing outside the gene revealed a large amplicon of $17 kb from the mutant genome as against the 8.6 kb (3L: 15811576..15820204) wild type amplicon ( Figure 8A). The pGEM-T-812 probe, which corresponds to the candidate region in the wild type, hybridized with all the amplicons implying the fidelity of amplification ( Figure 8A and B). This was further corroborated by the sequencing of the amplicon terminals (data not shown). Southern hybridization of wild type and mutant genomic DNA with HindIII and BamHI enzymes with pGEM-T-812 probe showed completely different banding profiles. While the HindIII digested DNA showed a band at $2.1 kb in the wild type genome, the l(3)tb genome showed a single band at $10 kb, the size difference being almost in agreement with the banding profile exemplified by BamHI digested DNA, wherein, the wild type genome showed a band at $10.2 kb and the mutant at $18 kb ( Figure 8D). These results imply the presence of an insertion in DCP2 gene, suggesting that l(3)tb is essentially an insertion allele of DCP2. Henceforth l(3)tb will be referred to as, DCP2 l(3)tb .
The DCP2 l(3)tb genome harbors Gypsy-LTR Like sequence in 59UTR coding region of DCP2 and expansion of adjacent upstream intergenic AT-rich sequence To identify the functional genomics of mutations, it is essential to deduce the nucleotide sequence of the mutation. Thus a convergent bi-directional primer-walk was initiated with the primer pair, which identified the insertion in DCP2 in the l(3)tb genome. On sequencing, the DCP2-proximal end showed the presence of wild type sequence till 3L: 15826410 after which a 444 bp AT-rich sequence was detected (Supplementary Figure S6 B-1), which did not show any resemblance with the wild type sequence present at the region whereas the dbo-proximal end showed complete wild type sequence profile (3L: 15827143..15826738) (Supplementary Figure S6 B-2). On a homology search to identify the novel sequence obtained, the sequence showed homology with the Gypsy LTR sequence of Drosophila. On searching for DCP2 promoters in the Eukaryotic Promoter Database, SIB and aligning the sequence coordinates of the 444 bp insertion, it was found that the insertion is downstream to the transcription start site (TSS) of DCP2, which is at 3L: 15826420. Subsequently, long-range PCR was first performed with DCP2 l(3)tb and wild type genomic DNA with a new pair of primers from the distal part of the reads obtained above,. Although no amplification was observed with the wild type DNA, the DCP2 l(3)tb DNA showed amplicons of sizes $7.2 kb, $3 kb, $2.8 kb and $430 bp with the 430 bp amplicon showing the highest concentration as observed from the electrophoretogram ( Figure 9B). This amplification profile resembled that of tandem repeat bearing regions. To confirm the repetitive nature of the sequence, Southern hybridization was performed with the same electrophoretogram. The 430 bp amplicon was eluted from the gel, cloned in pGEM-T vector and used as a probe. The probe showed complete hybridization with all of the amplicons indicating the repetitive nature of the sequence present downstream ( Figure 9B). Sequencing of the 430 bp amplicon revealed an AT-rich sequence. Homology search identified the sequence to be homologous to the distal part of the DCP2 UTR and the adjacent intergenic sequence between DCP2 and dbo, the coordinates being 3L: 15826407..15826716. After aligning the Figure 10 The tumor caused by DCP2 is hyperplastic with elevated Cyclin A and E. Immunolocalisation of Cyclin E shows elevated expression in the tumorous larval brains of DCP2 l(3)tb homozygotes (B) as compared to the wild type (A). Semi-quantitative analyses of mRNA expression of Cyclins A and E show similar elevation in the brain of DCP2 l(3)tb homozygotes (C). Expression of Discs-large in the brain (D) and wing discs (E) of the tumorous individuals did not show appreciable loss. At 138h AEL, the wing discs showed increase in cell number concomitant with decrease in cell size (E.C) whereas, at the same stage, the tumorous brain shows increased number of cells at in the optic lobe (D).
present set of reads with the previous set, a sequence duplication was observed for 59-T-A-T-A-39, flanking the Gypsy-LTR insertion (Supplementary Figure S6 B-3 and 4). The present set of sequencing reads also confirmed that the LTR insertion (3L: 15826407..15826407) was indeed prior to the completion of the UTR (3L: 15826423). Copy number variation analyses of the intergenic sequence vs. the internal control through PCR in the wild type and the mutant DCP2 l(3)tb, showed a sharp increase in the amplicon concentration of the intergenic sequence in DCP2 l(3)tb against the internal control as evidenced from the gel electrophoretogram ( Figure 9C). Comparison of the fluorescent intensity of the bands (intra and inter-genotype) showed a relatively high ratio of concentrations of the amplicon to the internal control (Dsor) amplicon, as observed from the graphical analyses ( Figure 9D). Based on the results obtained from rough and fine mapping, the architecture of the mutant allele, DCP2 l(3)tb is depicted in Figure 9E, which shows the bipartite nature of the mutation, viz., amplification of the intergenic sequence between DCP2 and dbo, as well as an insertion of 444 bp Gypsy LTR-like sequence immediately downstream to the TSS of DCP2. DCP2 l(3)tb is a DCP2 hypomorph along with low expression of the neighboring gene, dbo Following the identification of the genomic architecture of the allele, it was imperative to determine the expression potential of the allele. Semi-quantitative RT-PCR analyses confirmed the hypomorphic nature of the allele wherein the homozygous l(3)tb mutant showed extremely low levels of expression of DCP2 ( Figure 9F). Since the intergenic region between DCP2 and dbo is upstream to either gene and bears an expansion, dbo transcript titres were also examined wherein they showed extremely lowered expression ( Figure 9F). At present, it is doubtful whether the lowered dbo level is a cause or an effect of the mutation since Dbo (Smac/Diablo/Henji) is a proapoptogenic molecule which inhibits the inhibitor of apoptotic proteins (IAP), and its lowered levels, therefore, serve as a prognostic marker of tumor progression in human carcinomas (Martinez-Ruiz et al., 2008). Again, dbo expresses strongly in the neuronal tissues and localizes at the synapse (Wang et al., 2016) where its perturbation causes alteration in neuromuscular function. Despite being a multifaceted molecule, ubiquitous knock-down of dbo does not result in developmental abnormalities, viz., developmental delay, lethality or tumorous phenotypes in the brain, and/or wing imaginal discs (Supplementary Figure S7).

The tumor caused by DCP2 is hyperplastic with elevated Cyclin A and E
Since the mutation showed all the hallmarks of classical tumor suppressors (Merz et al., 1990), we sought to characterize the perturbations in cellular physiology caused in the wake of tumourigenesis. RT-PCR analyses depicted elevated levels of Cyclin E (G1/S phase cyclin) and A (G2/M phase cyclin), which are indicative of increased cell proliferation and rapid cell cycles ( Figure 10C). Immunolocalization studies also confirmed the elevated expression of Cyclin E as well ( Figure 10A and B). On observing closely, the regular arrangement of cells in the brain hemisphere and optic lobes in the wild type was severely disrupted in the mutant along with excessive growth and increased number of mitotic nuclei. The enlarged brain lobes, increased number of mitotic nuclei and disruption of the regular arrangement of cells in the mutant concomitant with elevated expression of cyclins A and E validated the tumourous nature of the mutant.
n■ Table 6 Global overexpression of <italic>DCP2</italic> rescues the mutant phenotypes exhibited by <italic>l (3)   When the tumorous brains ( Figure 10D) and wing discs ( Figure 10E) were examined for the expression of the polarity marker Discs large (Dlg), both tissues did not show appreciable loss of polarity. On a closer look, the wing discs at 138 h AEL showed an increase in cell number concomitant with a decrease in cell size ( Figure 10E-C). At the same stage, the tumorous brain shows an increased number of cells in the optic lobe ( Figure 10D). Usually, neoplastic tumors are metastatic and the tumor cells lose their polarity to acquire the mesenchymal-like fate, delaminate from the matrix and migrate. In contrast, hyperplastic tumors do not show appreciable loss of polarity even in later stages of tumourigenesis, since in these tumors the cells do not delaminate but remain adhered to the original tissue matrix, concomitant with cell division (Miles et al. 2011). The expression pattern of Dlg showed retention of polarity at 138 h of development, which is an exceptionally late and delayed 3 rd instar larval stage, implying that tumor is an over-proliferative and hyperplastic. Again, this is in agreement with the Cyclin E staining pattern and taken together they imply excessive cell division, which essentially requires increased and rapid cell cycles.
In both the cases of overexpression, all non-tubby progeny pupated, devoid of any developmental anomalies reminiscent of l(3)tb mutation, and emerged as flies. Thus, the rescue of the mutant phenotypes observed in l(3)tb homozygotes by global overexpression of DCP2 iteratively substantiates the fact the l(3)tb is an allele of DCP2 and that the tumor is caused solely owing to the loss of expression of DCP2.

Discussion and Conclusion
In the present paper we have identified a novel insertion allele of Drosophila DCP2, namely DCP2 l(3)tb, through a mutagenesis screen. DCP2 is conserved in other organisms and is extremely important for a number of growth processes throughout development (Xu et al., 2006;Ma et al., 2013), DNA replication (Mullen and Marzluff 2008;Schmidt et al., 2011), stress response (Hilgers et al., 2006;Xu and Chua 2012), synapse plasticity (Hillebrand et al., 2010), retrotransposition (Dutko et al., 2010) and viral replication (Hopkins et al., 2013). In Arabidopsis, DCP2 loss-of-function alleles show accumulation of capped mRNA intermediates, lethality of seedlings and defects in post-embryonic development, with no leaves, stunted roots with swollen root hairs, chlorotic cotyledons and swollen hypocotyls (Goeres et al., 2007;Iwasaki et al., 2007;Xu et al., 2006). In humans as well, chromosomal deletions of 5q21-22, the region harboring DCP2, is frequently observed in lung cancers (Hosoe et al., 1994;Mendes-da-Silva et al., 2000), colorectal cancer (Delattre et al., 1989) and oral squamous cell carcinoma (Mao et al., 1998). Hence, DCP2 has an unexplored role in the development and/or cell cycle progression across phyla, which needs to be investigated. Since the physiology of an organism is tightly regulated by the optimized titres of gene expression programs, a global loss of DCP2 may lead to perturbation in mRNA titres, which may alter the cellular response to such dismal conditions and eventually lead to drastic physiological disorders such as tumourigenesis. The exact mechanism(s) by which a loss of DCP2 leads to tumourigenesis needs further analysis and identification of the novel allele, DCP2 l(3)tb , unravels that DCP2 could be a probable candidate for future explorations of tumourigenesis.

ACKNOWLEDGMENTS
The authors acknowledge Prof. L. S. Shashidhara, IISER Pune, India, Prof. Volker Hartenstein, University of California, USA for sharing the Anti-Armadillo and Anti-Deadpan antibodies, respectively and to Prof. Aike Guo, Institute of Biophysics, China for sharing UAS-DCP2 fly line. We sincerely acknowledge Prof. Subhash C. Lakhotia, Prof. Figure 11 Global overexpression of DCP2 rescues mutant phenotypes associated with l(3)tb. Global over-expression of DCP2 using ubiquitous GAL4 drivers [Act5C-GAL4 (A) or Tub-GAL4 (B)] in the mutant homozygous l(3)tb individuals rescued the larval and pupal lethality. The graphs depict the fate of the non-tubby progeny, all of which pupated, devoid of any developmental anomalies and emerged as healthy flies.
Rajiva Raman and Dr. Rachana Nagar of our lab for assistance in fine mapping of the mutation and in analyses and interpretation of the results obtained therein. We duly acknowledge the National Facility for Laser Scanning Confocal Microscopy, UGC-CAS, DST-FIST of Department of Zoology, Banaras Hindu University and DST-PURSE, UGC-UPE to the Institute. We sincerely thank Department of Science and Technology (DST) for financial support to JKR.