Tissues in living organisms are shaped via complex processes that are collectively called ‘morphogenesis’. Many researchers have used the fruit fly Drosophila as a model to understand morphogenesis, which occurs both during the development of a Drosophila embryo and during metamorphosis (when the pupa changes to become an adult fly).

Like other insects, adult fruit flies have three main body sections (a head, a thorax and an abdomen), which are further divided into segments. The adult's abdomen forms inside the pupa from precursor cells called histoblasts. These cells are unusual in that they develop in the embryo but remain inactive during the larval stages of life. During pupation, these cells are reactivated by a hormone called ecdysone, and gradually replace the larva's tissues. However, it was not clear how this process was coordinated.

Verma and Cohen have now demonstrated that a small RNA molecule—a microRNA called miR-965—acts in histoblast cells and controls how much these cells divide as well as how they migrate during morphogenesis to replace the larval cells. MicroRNAs regulate other RNAs, called messenger RNAs, typically by targeting them for destruction. This prevents the messenger RNA molecules from being used to make proteins. When flies develop without miR-965, they are mostly normal but have defects in the formation of segments in the abdomen.

Verma and Cohen revealed that miR-965 acts by targeting two important messenger RNAs for destruction. These messenger RNAs encode a protein called String, which regulates histoblast proliferation, and another protein called Wingless. Once the pupa starts to form, the ecdysone hormone reduces the production of miR-965 to increase histoblast proliferation and migration. The miR-965 microRNA in turn reduces the level of ecdysone receptor. The ecdysone hormone acts as an all-or-nothing switch to make an irreversible change from the larval to the pupal stage. The hormone boosts its own activity by increasing expression of its own receptor. This ‘positive feedback loop’ acts like a switch and is very sensitive to small changes in the amount of hormone present.

Verma and Cohen propose that by reducing the levels of the hormone receptor, miR-965 makes the system more stable. This is because the hormone must first overcome the action of miR-965 before it can kick off the positive feedback loop. This takes time, and means that any change in the amount of hormone must be around for a while to have an effect. This mechanism buffers against short-lived, small changes in hormone levels that might throw the switch at the wrong time—a feature known as ‘robustness’. This seems to be a complicated process to go from one state to another (i.e., from a larva to a pupa). But, the existence of the many distinct checks and balances makes sure the switch is thrown only when it is needed.

Tissue morphogenesis is a complex process, through which the organism coordinates cell proliferation and cell death with cell migration and rearrangements to achieve final organ shape and size. Mechanisms controlling these processes play important role in morphogenesis, tissue repair and regeneration, and in cancer (Friedl and Gilmour, 2009; Rorth, 2009). The abdominal epithelium of Drosophila provides a useful model system in which to study the dynamics of tissue morphogenesis in vivo and to explore the genetic and cellular mechanisms that control these complex morphogenetic processes.

During metamorphosis, larval epidermal tissues undergo cellular restructuring and rearrangement to give rise to adult abdominal epithelium. The adult abdominal epithelium is produced from progenitor cells, known as histoblast cells. Histoblasts are small diploid cells, easily distinguishable from the large polyploid larval epidermal cells (LEC) that surround them. Histoblasts are specified in the embryo and lie quiescent throughout larval development (Guerra et al., 1973; Simcox et al., 1991). There are four pairs of histoblast nests in each segment, which merge to assemble the adult abdominal epidermis (Madhavan and Schneiderman, 1977). The anterior and posterior dorsal histoblast nests give rise to the external dorsal cuticle of the abdominal segments (tergites), while the ventral pair give rise to the ventral cuticle (sternites).

During larval stages, histoblasts are arrested in the G2 phase of the cell cycle. At pupariation, an ecdysone pulse triggers the expression of string (cdc25), which activates cyclin/CDK and pushes the histoblasts into rapid proliferation (Edgar and O'Farrell, 1990; Gautier et al., 1991). The proliferative phase is divided into two stages with distinct features (Madhavan and Madhavan, 1980; Ninov et al., 2009). The early ‘division phase’ is characterized by rapid and synchronous cell division without intervening G2 phases to allow for cell growth. Cells double in number and decrease in size with each division, with little change in the size of the nests. This is followed by a phase of slower division, from 15–40 hr of pupal development, in which proliferation is accompanied by longer intervening gap phases to allow cell growth. During this phase, epidermal growth factor receptor (EGFR) and Insulin receptor/PI3K- signaling coordinate the growth of cells with proliferation (Ninov et al., 2009). During the growth phase, the histoblast nests begin to spread to replace the LEC. The expansion of the histoblast nests by cell migration is accompanied by programmed cell death, so that larval cells are replaced by the expanding histoblast population to maintain integrity of the epithelium (Bischoff and Cseresnyes, 2009; Nakajima et al., 2011).

Patterning of the adult abdominal segments makes use of many of the signaling pathways that pattern the embryonic and larval segments, including hedgehog, wingless, decapentaplegic and EGFR. Interactions between morphogen gradients produced by some of these proteins determine anterior-posterior and dorsal-ventral patterning of segments (Sanicola et al., 1995; Shirras and Couso, 1996; Struhl et al., 1997; Kopp et al., 1999; Ninov et al., 2010). Although abdominal segmental patterning has been extensively studied, the molecular mechanisms regulating cell division, migration, cell replacement and their interactions during formation of segments remain less well understood. Here, we provide evidence that the microRNA, miR-965, is required in the histoblast nests, where it acts via regulation of string and wingless to control histoblast proliferation and migration during pupal morphogenesis.

The miR-965 microRNA is located in the first intron of the kismet gene (Figure 1A). kismet and miR-965 are transcribed in the same direction. Quantitative PCR showed that a series of P-element insertional mutants near the first exon of the kismet gene reduced the level of mature miR-965 miRNA, suggesting that kismet and miR-965 arise from a single transcription unit (Figure 1—figure supplement 1).

Figure 1 with 3 supplements see all

Download asset Open asset

To explore the biological functions of miR-965, independently from those of kismet, we produced targeted deletion mutants to remove the miRNA from the kismet intron (Figure 1A; [Chen et al., 2014]). One mutant allele was made by homologous recombination using the pw25 vector (Gong and Golic, 2003). In this allele, the genomic region containing the miRNA hairpin was replaced by a mini-white reporter gene flanked by LoxP sites (Figure 1A, w+KO1). The presence of the intron-containing mini-white marker in the kismet intron was expected to disrupt kismet gene function. Cre-mediated excision of the LoxP-flanked mini-white cassette produced the KO1 allele, in which the miRNA was replaced by a single LoxP site (Figure 1A, w-KO1). A second independent deletion was produced using the pRMCE vector, which was designed to allow subsequent retargeting of the locus using Recombination Mediated Cassette Exchange (RMCE) (Weng et al., 2009). The miRNA hairpin was replaced by a mini-white marker flanked by inverted attP and LoxP sites. The KO2 allele was made by Cre-mediated excision of the LoxP flanked mini-white reporter (Figure 1A, w-KO2). Deletion of the miRNA was verified by quantitative miRNA PCR in animals heteroallelic for the mini-white deleted versions of KO1/KO2 and for the KO2 allele in trans to Df(2L)ED19, a chromosomal deletion that uncovers the kismet gene (Figure 1B). To produce a genetic rescue allele, we used RMCE to replace the mini-white cassette with a fragment containing the miR-965 hairpin (Figure 1A, 965-Rescue). miR-965 expression was restored to normal levels in the 965-Rescue allele (Figure 1B).

In the w- versions of the KO1 and KO2 alleles, DNA in the kismet intron was replaced by a single LoxP site (Figure 1A). To ask whether kismet gene function was compromised in these mini-white excised alleles, we performed molecular and genetic tests. Figure 1C shows RT-PCR to assess splicing of the first kismet intron. A PCR product of normal size was produced in RNA isolated from miR-965^w-KO1/miR-965^w-KO2 and miR-965^w-KO2/Df animals. This qualitative assessment was confirmed by quantitative PCR (Figure 1D). kismet transcript levels were not reduced in the miR-965^w-KO1/miR-965^w-KO2 flies compared to controls, indicating that splicing of intron 1 was not impaired in these mutants. Genetic complementation tests also showed that the miR-965^w-KO1 and miR-965^w-KO2 mutant alleles were viable in trans to kismet¹, a lethal allele, as well as in trans to Df(2L)ED19, which removes kismet. Together, these data provide evidence that the mini-white excised mutant miR-965^w-KO1 and miR-965^w-KO2 alleles do not compromise kismet gene function. For brevity, these mini-white deleted alleles will be referred to as KO1 and KO2 in the text and figures that follow.

miR-965 activity in histoblasts

The abdominal segmentation defects observed in miR-965 mutants suggested that the miRNA might be required in the histoblasts. To visualize miR-965 activity, we made use of a sensor transgene consisting of a ubiquitously expressed GFP reporter with a perfect target site for miR-965 in the 3′ UTR. Sensors of this design allow miRNA activity to be visualized by downregulation of GFP (Brennecke et al., 2003). The control sensor, lacking the miRNA-target site, was expressed at comparable levels in the large polyploid LEC and in the smaller histoblast cells (hb, Figure 2). miR-965 sensor GFP levels were lower in the histoblast nests, compared to the adjacent larval cells, particularly in the histoblast cells near the edge of the nests (Figure 2). This difference was lost when the sensor was introduced into the miR-965 mutant background, providing evidence that the reduced GFP level in the histoblasts is due to miR-965-mediated repression (Figure 2).

Figure 2

Download asset Open asset

Defects in histoblast development

The escargot (esg) gene is expressed in histoblasts from the time of their specification in the embryo until the late pupal stage, and provides a marker to visualize histoblast development (Hayashi et al., 1993). Figure 3A presents still images taken from time-lapse videos of esg-GAL4, UAS-GFP pupae to visualize the first 3 mitotic divisions of the histoblasts. In the controls, cells became smaller and the number of cells doubled after each division (Video 1). This pattern of synchronous division was perturbed in the miR-965 mutant. Asynchronous division of miR-965 mutant histoblasts resulted in the presence of cells of different sizes (Figure 3A, Video 2). In the mutant, a subset of histoblast nuclei became pyknotic, fragmented and disappeared, indicating cell death (Video 3). Cell death was rare during this phase in the control samples. Synchronous division was restored and cell death was suppressed in the miR-965-rescue allele (Figure 3A, Video 4).

Figure 3 with 3 supplements see all

Download asset Open asset

Video 1

Download asset

Video 2

Download asset

Video 3

Download asset

Video 4

Download asset

Subsequently, the rate of histoblast nest expansion was slower in the miR-965 mutant, compared to the controls (Figure 3B). Expansion of the histoblast nests was quantified in segments 3 and 4 by monitoring the speed of migration (Figure 3—figure supplement 1, Videos 5, 6). The average speed of migration of third and fourth histoblast nests in the control samples was 15 µm/hr, compare with ∼6.5 µm/hr in the mutant. The rate of nest expansion was increased to 12–14 µm/hr by restoring miR-965 expression with the rescue allele the miR-965 mutant background (Video 7).

Video 5

Download asset

Video 6

Download asset

Video 7

Download asset

LEC generally undergo programmed cell death as the histoblasts nests start to grow (Nakajima et al., 2011). In miR-965 mutants, we observed that some of these cells were still present in the nests, suggesting a failure to eliminate LECs in the mutant (Figure 3—figure supplement 2). Delayed expansion of the histoblast nests, combined with the persistence of LECs, is likely to be responsible for the gaps observed in adult abdominal segments.

It was noted that the migration defects observed in the videos seem to be more severe than the segmentation defects observed in the adult flies. However, we did not observe significant pre-eclosion lethality (Figure 3—figure supplement 3). Given that there was no evidence for loss of a class of more severely affected animals, we suggest that the apparent difference reflects a delay in tissue replacement in the mutant, so that most animals end up with milder defects by the end of pupariation than were apparent during the early pupal time-window in the videos.

miR-965 regulates string and wingless

miR-965 is predicted to target 69 genes (www.targetscan.org). polycomb, string, wingless, homothorax, Tor, Hsp83 and jumeau were selected for further analysis based on the quality of the predicted target site and on Flybase annotation suggesting roles in segmentation. Of these, only string and wingless (wg) mRNAs were upregulated in RNA isolated from miR-965 mutant pupae, and were also restored to near normal levels in the rescued mutant (Figure 4A,B), as would be expected for functional miRNA targets.

Figure 4 with 1 supplement see all

Download asset Open asset

The wg 3′UTR was used to make a luciferase reporter transgene and tested for regulation by miR-965 in S2 cells. miR-965 expression reduced wg 3′UTR reporter activity, and this regulation was lost in the wg UTR reporter mutated to disrupt pairing with the miRNA seed sequence (Figure 4C, mutated residues shown in red). miR-965 expression also reduced expression of the string 3′UTR luciferase reporter (Figure 4D). Mutation of the predicted target site to disrupt seed pairing partially offset regulation of the string 3′UTR luciferase reporter (Figure 4D, mutated residues in red). We noted the presence of a second potential target site nearby (Figure 4—figure supplement 1). More extensive mutation to disrupt both sites further compromised regulation by miR-965 (Figure 4D). Other non-canonical target sites might be responsible for the remaining regulation by miR-965, however, we do not exclude the possibility that there could also be indirect effects of miR-965 on expression of this reporter. These findings provide evidence that miR-965 can regulate expression of string and wg.

Contributions of string and wingless to the miR-965 mutant phenotype

To ask whether the increases in string and wg expression might be responsible for the miR-965 mutant phenotype, we first asked if overexpressing them in an otherwise normal genetic background could phenocopy the mutant. esg-Gal4 driven expression of a UAS-string transgene produced abdominal segment gaps, segment fusion and polarity reversal phenotypes, similar to those observed in the miR-965 mutant (Figure 5A Figure 5—figure supplement 1). The proportion of flies with defects caused by string overexpression was similar to that in the miR-965 mutant (Figure 5B). string overexpression caused asynchronous histoblast division and apoptosis during the early division phase (Figure 5—figure supplement 2, Video 8). Expansion of the string-overexpressing histoblast nests was slowed, and the histoblasts were unable to fully replace the LECs (Figure 5—figure supplements 2, 3, Video 9), similar to what was observed in miR-965 mutants. Thus, string overexpression was sufficient to reproduce the miR-965 mutant phenotype.

Figure 5 with 5 supplements see all

Download asset Open asset

Video 8

Download asset

Video 9

Download asset

To assess the contribution of wg overexpression, we made use of a UAS transgene directing expression of a temperature sensitive form of Wg protein (Wilder and Perrimon, 1995). Use of Wg^ts was required to allow stage specific activation of Wg in the histoblasts. Continuous expression of wild-type Wg under esg-Gal4 control was lethal. Wg^ts protein is inactive at 25°C. Flies carrying esg-Gal4 and UAS-wg^ts were reared at 25°C or shifted to 18°C in the third larval instar to allow expression of active Wg as the histoblasts began proliferation. This resulted in a phenocopy of the segment gap phenotype in 7% of animals (Figure 5C). Phenocopy was rare in the animals raised continuously at 25°C to maintain low Wg activity (0.4% affected, Figure 5C; p = 0.014 comparing UAS-wg^ts at 18° vs 25°, Fisher's exact test).

Next we asked whether limiting target gene overexpression could suppress the mutant phenotype. Lowering string activity by introducing string mutant alleles reduced the penetrance of the segment gap phenotype in the miR-965 (KO1/KO2) mutant background from 25% to 6–7% (Figure 5D, Video 10, p < 0.001 comparing KO1/KO2 vs KO1/KO2; stg^EY or stg⁴/+, Fisher's exact test). Lowering string activity also suppressed the asynchronous division phenotype and the slow cell migration phenotype of the miR-965 mutant (Figure 5—figure supplements 4, 5, Video 11).

Video 10

Download asset

Video 11

Download asset

Wingless protein (Wg) is expressed in the anterior dorsal histoblast nests, during their growth and migration (Kopp et al., 1999). The level of Wg protein was higher in the miR-965 mutant (Figure 5E). Wg is a secreted protein, and its distribution appears to be broader, reaching some of the LECs in the mutant, perhaps reflecting the higher level of protein produced (Figure 5E). To ask whether this elevated Wg expression contributes to the defects in the mutant, we introduced wg mutant alleles into the miR-965 (KO1/KO2) mutant background. The segment gap phenotype was reduced comparing KO1/KO2 vs KO1 wg^SP-1/KO2 (Figure 5F, p < 0.05, Fisher's exact test; suppression by the wg^I-12 allele was not statistically significant). Next, we tested the effect of removing one copy each of stg and wg. The segment gap phenotype was reduced comparing KO1/KO2 vs KO1 wg^SP-1/KO2; stg⁴/+ (Figure 5F, p < 0.001, Fisher's exact test). Taking out one copy each of stg and wg lowered penetrance of the phenotype to ∼4%, compared with 6–7% for stg/+ alone or 12% for wg/+ alone.

These experiments provide evidence that overexpression of string and wg each contribute to the miR-965 mutant histoblast phenotypes. The effect of limiting Wg expression in the mutant background may appear to be smaller than that of limiting String. However, to make a meaningful comparison of the relative contribution of these two targets, it would be necessary to restore each of them to normal levels. The genetic method used to reduce target activity in the mutant background does not allow precise control over the final target level, so this question cannot be addressed. We also do not exclude the possibility that there could be other functionally significant targets in addition to Wg and String.

Ecdysone mediated regulation of miR-965

Ecdysone pulses at the beginning of pupariation have been shown to induce string expression in order to reactive histoblast proliferation (Ninov et al., 2009). In light of the relationship between miR-965 and string, we asked whether miR-965 expression might be under Ecdysone control at this stage. We made use of a UAS-EcR^RNAi transgene expressed under esg-Gal4 to reduce Ecdysone receptor levels in the histoblasts. RNAi-mediated depletion of EcR mRNA led to an increase in the level of the miR-965 primary transcript and mature miRNA (Figure 6A, Figure 6—figure supplement 1), and to reduced string mRNA levels in RNA samples isolated from early pupae (Figure 6A). EcR binding sites have been identified near the host gene, kismet (Gauhar et al., 2009), consistent with the possibility that ecdysone regulates expression of both kismet and miR-965. When miR-965 was overexpressed in the histoblast cells using esg-GAL4, string transcript levels were reduced (Figure 6B) and histoblast cell divisions were arrested (Figure 6C, Video 12). This phenotype resembles EcR-B mutants, in which histoblast division is compromised (Bender et al., 1997). Pupae overexpressing miR-965 in histoblast cells did not survive beyond ∼12 APF, so it was not possible to monitor the later stages of histoblast migration in this genotype.

Figure 6 with 2 supplements see all

Download asset Open asset

Video 12

Download asset

In light of the finding that EcR limits miR-965 expression, we asked whether Ecdysone signaling might act via miR-965 to regulate string. We introduced a UAS-EcR^RNAi into the miR-965 mutant background and measured the levels of string mRNAs. Depletion of EcR did not reduce string mRNA in animals lacking miR-965 (Figure 6D). This provides evidence that Ecdysone signaling is mediated through regulation of the miR-965 miRNA to regulate histoblast proliferation.

In the course of this analysis, we observed that EcR transcript levels increased in the miR-965 mutant (Figure 6D). Both mature and primary transcript levels increased, suggesting an indirect effect of the miRNA on EcR transcription. To ask if there might also be a post-transcriptional component to the regulation of EcR by miR-965, we used an EcR 3′ UTR reporter transgene linked to GFP (Varghese and Cohen, 2007). GFP expression in the histoblast nests did not increase in the miRNA mutant background, indicating indirect regulation of EcR by miR-965 (Figure 6—figure supplement 2). These experiments provide evidence for a regulatory feedback relationship between miR-965 and the Ecdysone receptor (Figure 6E). EcR activity limits miR-965 expression. miR-965 activity limits EcR primary transcript levels, suggesting an effect on transcription.

Our findings link regulation of the miR-965 microRNA to the onset of histoblast proliferation at the larval to pupal transition. Previous reports have provided evidence that Ecdysone signaling activates string expression to trigger the onset of histoblast proliferation at the beginning of pupal development (Ninov et al., 2009). Our findings provide evidence that Ecdysone signaling works though regulation of miR-965, which in turn regulates string. Interestingly, we also find evidence for negative feedback regulation of miR-965 on EcR. Mutual repression circuitry of this type can contribute a switch-like function: EcR activity lowers miR-965 activity, which allows greater EcR expression/activity by alleviating miR-965 mediated repression. In a circuit of this design, there will be a delay between reduced transcription of the miRNA primary transcript and the decay of the mature miRNA product. Hence sustained EcR activity is needed to throw the switch.

We have previously reported that EcR shows positive transcriptional autoregulation and that this is buffered by miR-14 in a mutual repression circuit (Varghese and Cohen, 2007). Positive feedback allows for a sharp switch-like response, but also makes the system very sensitive to stochastic fluctuation in EcR activity. Coupling EcR positive auto-feedback to miRNA-mediated repression allows a robust switch function upon Ecdysone stimulation, while protecting the system from the effects of biological noise. This study provides evidence that miR-965 plays an analogous role in regulating EcR response and suggests that miR-965 confers robustness to the EcR response in the histoblasts.

Upregulation of string in the miR-965 mutant contributes to the defects in histoblast proliferation. How misregulation of string might contribute to the migration defects is less immediately obvious. Previous work has shown that cell cycle progression in the histoblast population is required to trigger programmed cell death in the surrounding LEC (Nakajima et al., 2011). Those authors provided evidence that cell growth and the expansion of the histoblast nests may be required to elicit LEC apoptosis. Although the mechanism by which expansion of the histoblasts triggers LEC death is not clear, elevated string expression in the miR-965 mutant is likely to be responsible for the cell cycle progression defects during this phase, hindering normal LEC removal and histoblast migration.

Persistence of the LECs might also be a consequence of the increased expression of Wg protein in the mutant histoblast nests. Wg acts in combination with EGFR and Dpp signals to control abdominal segment patterning (Shirras and Couso, 1996; Kopp et al., 1999; Ninov et al., 2009, 2010). These signals are thought to control differential cell adhesion, which may be important for elimination of the LECs as well as for proper segmental fusion of the histoblast nests. Elevated expression of Wg protein may lead to an expanded range of action, perhaps resulting in ectopic Wg activity in the LECs.

Each adult abdominal segment has a well-defined anterior-posterior polarity. Wg is required from 15–20 hr APF for bristle formation and from 18–28 hr APF for tergite differentiation and pigmentation. Overexpression of wg has been shown to cause ectopic bristle formation, and shaggy mutant clones, which constitutively activate wg signaling, can cause polarity reversal in abdominal bristles, while EGFR, FGF, dpp and Notch signaling have no effect on the polarity of bristles in adult epidermis (Lawrence et al., 2002). Wg levels are normally higher in the posterior region of the anterior histoblast nests and lower more anteriorly. Our finding that Wg levels were elevated and that the distribution of Wg was broader than normal suggests ectopic Wg activity throughout the histoblast nest, including cells that normally experience low Wg levels. Ectopic spread of Wg could be responsible for the formation of ectopic bristles and for the occasional instances of polarity reversal observed in the anterior part of tergites in the miR-965 mutants.

Replacement of the larval epidermis during metamorphosis involves regulation of both cell-intrinsic events in the abdominal histoblasts and communication between histoblasts and the larval cells they will replace. miR-965 acts on at least two separate processes required during histoblast morphogenesis. A miRNA with multiple targets can add a layer of regulation, acting across different pathways to integrate their activities (Herranz and Cohen, 2010). In doing so, the miR-965 miRNA appears to contribute to the robustness of this complex morphogenetic system.

1. 1. Bender M
   2. Imam FB
   3. Talbot WS
   4. Ganetzky B
   5. Hogness DS

   (1997) Drosophila ecdysone receptor mutations reveal functional differences among receptor isoforms

   Cell 91:777–788.

   https://doi.org/10.1016/S0092-8674(00)80466-3

   • Google Scholar
2. 1. Bischoff M
   2. Cseresnyes Z

   (2009) Cell rearrangements, cell divisions and cell death in a migrating epithelial sheet in the abdomen of Drosophila

   Development 136:2403–2411.

   https://doi.org/10.1242/dev.035410

   • Google Scholar
3. 1. Brennecke J
   2. Hipfner DR
   3. Stark A
   4. Russell RB
   5. Cohen SM

   (2003) bantam encodes a developmentally regulated microRNA that controls cell proliferation and regulates the pro-apoptotic gene hid in Drosophila

   Cell 113:25–36.

   https://doi.org/10.1016/S0092-8674(03)00231-9

   • Google Scholar
4. 1. Chen YW
   2. Weng R
   3. Cohen SM

   (2011) Protocols for use of homologous recombination gene targeting to produce MicroRNA mutants in Drosophila

   Methods in Molecular Biology 732:99–120.

   https://doi.org/10.1007/978-1-61779-083-6_8

   • Google Scholar
5. 1. Chen YW
   2. Song S
   3. Weng R
   4. Verma P
   5. Kugler JM
   6. Buescher M
   7. Rouam S
   8. Cohen SM

   (2014) Systematic study of Drosophila MicroRNA functions using a collection of targeted knockout mutations

   Developmental Cell 31:784–800.

   https://doi.org/10.1016/j.devcel.2014.11.029

   • Google Scholar
6. 1. Edgar BA
   2. O'Farrell PH

   (1990) The three postblastoderm cell cycles of Drosophila embryogenesis are regulated in G2 by string

   Cell 62:469–480.

   https://doi.org/10.1016/0092-8674(90)90012-4

   • Google Scholar
7. 1. Friedl P
   2. Gilmour D

   (2009) Collective cell migration in morphogenesis, regeneration and cancer

   Nature Reviews. Molecular Cell Biology 10:445–457.

   https://doi.org/10.1038/nrm2720

   • Google Scholar
8. 1. Gauhar Z
   2. Sun LV
   3. Hua S
   4. Mason CE
   5. Fuchs F
   6. Li TR
   7. Boutros M
   8. White KP

   (2009) Genomic mapping of binding regions for the Ecdysone receptor protein complex

   Genome Research 19:1006–1013.

   https://doi.org/10.1101/gr.081349.108

   • Google Scholar
9. 1. Gautier J
   2. Solomon MJ
   3. Booher RN
   4. Bazan JF
   5. Kirschner MW

   (1991) cdc25 is a specific tyrosine phosphatase that directly activates p34cdc2

   Cell 67:197–211.

   https://doi.org/10.1016/0092-8674(91)90583-K

   • Google Scholar
10. 1. Gong WJ
    2. Golic KG

    (2003) Ends-out, or replacement, gene targeting in Drosophila

    Proceedings of the National Academy of Sciences of USA 100:2556–2561.

    https://doi.org/10.1073/pnas.0535280100

    • Google Scholar
11. 1. Guerra M
    2. Postlethwaite JH
    3. Schneiderman HA

    (1973)

    The development of the imaginal abdomen of Drosophila melanogaster

    Developmental Biology 32:361–372.

    • Google Scholar
12. 1. Hayashi S
    2. Hirose S
    3. Metcalfe T
    4. Shirras AD

    (1993)

    Control of imaginal cell development by the escargot gene of Drosophila

    Development 118:105–115.

    • Google Scholar
13. 1. Herranz H
    2. Cohen SM

    (2010) MicroRNAs and gene regulatory networks: managing the impact of noise in biological systems

    Genes & Development 24:1339–1344.

    https://doi.org/10.1101/gad.1937010

    • Google Scholar
14. 1. Kopp A
    2. Blackman RK
    3. Duncan I

    (1999)

    Wingless, decapentaplegic and EGF receptor signaling pathways interact to specify dorso-ventral pattern in the adult abdomen of Drosophila

    Development 126:3495–3507.

    • Google Scholar
15. 1. Lawrence PA
    2. Casal J
    3. Struhl G

    (2002)

    Towards a model of the organisation of planar polarity and pattern in the Drosophila abdomen

    Development 129:2749–2760.

    • Google Scholar
16. 1. Madhavan MM
    2. Madhavan K

    (1980)

    Morphogenesis of the epidermis of adult abdomen of Drosophila

    Journal of Embryology and Experimental Morphology 60:1–31.

    • Google Scholar
17. 1. Madhavan MM
    2. Schneiderman HA

    (1977)

    Histological analysis of the dynamics of growth of imaginal discs and histoblast nests during the larval development of Drosophila melanogaster

    Roux's Archives of Developmental Biology 183:269–305.

    • Google Scholar
18. 1. Nakajima Y
    2. Kuranaga E
    3. Sugimura K
    4. Miyawaki A
    5. Miura M

    (2011) Nonautonomous apoptosis is triggered by local cell cycle progression during epithelial replacement in Drosophila

    Molecular and Cellular Biology 31:2499–2512.

    https://doi.org/10.1128/MCB.01046-10

    • Google Scholar
19. 1. Ninov N
    2. Manjon C
    3. Martin-Blanco E

    (2009) Dynamic control of cell cycle and growth coupling by ecdysone, EGFR, and PI3K signaling in Drosophila histoblasts

    PLOS Biology 7:e1000079.

    https://doi.org/10.1371/journal.pbio.1000079

    • Google Scholar
20. 1. Ninov N
    2. Menezes-Cabral S
    3. Prat-Rojo C
    4. Manjon C
    5. Weiss A
    6. Pyrowolakis G
    7. Affolter M
    8. Martin-Blanco E

    (2010) Dpp signaling directs cell motility and invasiveness during epithelial morphogenesis

    Current Biology 20:513–520.

    https://doi.org/10.1016/j.cub.2010.01.063

    • Google Scholar
21. 1. Rorth P

    (2009) Collective cell migration

    Annual Review of Cell and Developmental Biology 25:407–429.

    https://doi.org/10.1146/annurev.cellbio.042308.113231

    • Google Scholar
22. 1. Sanicola M
    2. Sekelsky J
    3. Elson S
    4. Gelbart WM

    (1995)

    Drawing a stripe in Drosophila imaginal disks: negative regulation of decapentaplegic and patched expression by engrailed

    Genetics 139:745–756.

    • Google Scholar
23. 1. Shirras AD
    2. Couso JP

    (1996) Cell fates in the adult abdomen of Drosophila are determined by wingless during pupal development

    Developmental Biology 175:24–36.

    https://doi.org/10.1006/dbio.1996.0092

    • Google Scholar
24. 1. Simcox AA
    2. Hersperger E
    3. Shearn A
    4. Whittle JRS
    5. Cohen SM

    (1991) Establishment of imaginal discs and histoblast nests in Drosophila

    Mechanisms of Development 34:11–20.

    https://doi.org/10.1016/0925-4773(91)90087-M

    • Google Scholar
25. 1. Struhl G
    2. Barbash DA
    3. Lawrence PA

    (1997)

    Hedgehog organises the pattern and polarity of epidermal cells in the Drosophila abdomen

    Development 124:2143–2154.

    • Google Scholar
26. 1. Varghese J
    2. Cohen SM

    (2007) microRNA miR-14 acts to modulate a positive autoregulatory loop controlling steroid hormone signaling in Drosophila

    Genes & Development 21:2277–2282.

    https://doi.org/10.1101/gad.439807

    • Google Scholar
27. 1. Weng R
    2. Chen YW
    3. Bushati N
    4. Cliffe A
    5. Cohen SM

    (2009) Recombinase-mediated cassette exchange provides a versatile platform for gene targeting: knockout of miR-31b

    Genetics 183:399–402.

    https://doi.org/10.1534/genetics.109.105213

    • Google Scholar
28. 1. Wilder EL
    2. Perrimon N

    (1995)

    Dual functions of wingless in the Drosophila leg imaginal disc

    Development 121:477–488.

    • Google Scholar

Author details

1. Pushpa Verma

   Institute of Molecular and Cell Biology, Singapore, Singapore

   Contribution

   PV, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

2. Stephen M Cohen

   1. Institute of Molecular and Cell Biology, Singapore, Singapore
   2. Department of Cellular and Molecular Medicine, University of Copenhagen, Copenhagen, Denmark

   Contribution

   SMC, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   scohen@sund.ku.dk

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0003-2858-9163

• 1,787

  views

• 389

  downloads

• 24

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.