A New Genetic Locus Controlling Growth and Proliferation in Drosophila melanogaster

Corresponding author: Pierre Léopold, Developmental Biology and Cancer Research, UMR6543 CNRS, Parc Valrose, 06108 Nice Cedex 2, France., leopold{at}unice.fr (E-mail)

Communicating editor: A. J. LOPEZ

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Multicellular organisms grow through both proliferation and growth of their individual cells. We have conducted a P-element-based misexpression screen for genes whose upregulation alters wing disc growth during development. One particular group of four P elements, all inserted at cytological location 61C7-8, exhibited specific overgrowth upon misexpression in proliferating imaginal tissues. Clonal analysis revealed that upon misexpression, cell number was increased but cell size was not affected, indicating that cell growth and proliferation were induced in a coordinate manner. Loss of function at the locus produced small flies with reduced cell number, consistent with the presence of a gene encoding a positive growth regulator. We characterized a new transcription unit initiating in a region adjacent to the P insertions, which generated a complex series of polyadenylated transcripts. Although these RNAs were induced in response to misexpression, none was sufficient by itself to recapitulate overgrowth when overexpressed. This suggested either that a particular combination of these transcripts was necessary or that other sequences are involved.

MOST animals are developmentally programmed to grow to a characteristic adult size. The size of an organ is also intrinsically determined, as is each individual cell size. Elucidation of the different mechanisms that control cell size, cell proliferation, cell death, and their coupling with developmental cues has become a particularly challenging issue for both developmental and cellular biology (CONLON and RAFF 1999 ).

Drosophila larvae provide a useful system to study tissue growth. During the 4 days of larval life, extensive tissue growth takes place via two distinct mechanisms. In most larval tissues, endoreduplication allows rapid amplification of the genome and a considerable increase in cell volume (EDGAR and ORR-WEAVER 2001 ). This is exemplified in the salivary glands where cell volume is increased by a factor of 200 during larval growth. In imaginal discs, which give rise to the different adult appendages, concomitant cell growth and proliferation allow tissues to form by increasing cell number while maintaining a roughly constant cell size during development (NEUFELD and EDGAR 1998 ). These differences illustrate the complex relationships between cell growth (increase in cell volume) and cell proliferation (increase in cell number) that exist in vivo (TAPON et al. 2001 ).

Recent experiments have investigated these relationships in proliferating Drosophila imaginal disc. In this system, clonal overexpression of the cell cycle activator dE2F accelerates the pace of cell divisions without affecting cell growth, thus leading to a decrease in cell size. Conversely, overexpressing the dE2F-inhibitor RBF slows the cell cycle without affecting cell growth, thus increasing cell size (NEUFELD and EDGAR 1998 ). This suggests that the cell cycle machinery can be activated independently of the cell growth machinery and that growth does not depend upon cell division. Whether and how cell growth can control cell proliferation are more complex questions. Recent work has exemplified the role of several signaling pathways in the control of imaginal disc cell growth. The insulin receptor signaling pathway exists in flies and was recently shown to control tissue growth in vivo (STOCKER and HAFEN 2000 ). Loss-of-function mutations in dInr (the unique Drosophila insulin/IGF receptor), chico (IRS), PI3K, dAKT/PKB, or dS6K lead to sizeable growth defects in adults and larvae, but conserve proportions between body parts. Clonal analysis shows that downregulation of this signaling pathway leads to a reduction of cell size accompanied with a slowdown of the cell cycle. Interestingly, overexpression of PI3K and other downstream components causes an increase in cell size, accompanied in some cases with a reduction of G₁ phase, but in fine without acceleration of the division rate (for review, see WEINKOVE et al. 1999 ). This establishes that, in imaginal discs, PI3K-mediated cell growth is necessary for cell division to occur at the normal rate but not sufficient to push cells to proliferate faster. Other growth regulators like dRas and dMyc have similar effects: in contrast to their described mitogenic function in tissue culture cells, their overexpression in disc cells promotes growth and shortens G₁ but is unable to promote increased cell proliferation (JOHNSTON et al. 1999 ; PROBER and EDGAR 2000 ; for review, see PROBER and EDGAR 2001 ).

In contrast, clonal overexpression in imaginal discs has revealed that CyclinD/Cdk4 (D/K4) increases growth and proliferation coordinately (i.e., with no alteration of cell cycle phasing or average cell size; DATAR et al. 2000 ; MEYER et al. 2000 ). So far, this is the only well-characterized example of such a coupling with cell division in growing Drosophila tissues. Genetic data suggest that, in this case, growth control and cell cycle control might be elicited through separate targets. For example, the pocket protein Rbf counteracts the proliferative action of D/K4 in discs but does not suppress the overgrowth phenotypes elicited by ectopic D/K4 expression in nonmitotically active cells (DATAR et al. 2000 ; MEYER et al. 2000 ; XIN et al. 2002 ). This suggests that D/K4 induces proliferation through an inhibitory phosphorylation of Rbf, which in turn causes activation of dE2F1 as described in mammalian models. In accordance with this hypothesis, ectopic dE2F1 is capable of activating both Cyclin E and Cdc25^string in wing disc, leading to accelerated proliferation (NEUFELD and EDGAR 1998 ). In contrast, neither dE2F nor dRBF proteins have been shown to control cell growth in imaginal disc and the mechanism by which D/K4 activates disc cell growth is currently unknown.

To better understand the links between cell growth and proliferation, we have conducted a misexpression screen (RORTH 1996 ) for genes whose upregulation alters growth or proliferation of imaginal disc cells. One group of mutant lines defines a new genetic locus involved in both growth and proliferation control.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks:
yw,P{y+}UAS;+;+, yw;CyO/Sp;23,Sb/TM6b, yw;CyO/Sp;+ and yw;+;TM3,Sb/Dr [Berkeley Drosophila Genome Project (BDGP) and UY Consortium] were used for the generation of UY collection. All strains were isogenized before mutagenesis was performed.

• Gain-of-function screen: MS1096;+;+, yw;enGAL4,+ (Bloomington Stock Center), w,MS1096,UAS-PI3K^CAAX;+;+, w,MS1096, UAS-PI3K^D954A;+;+ (gifts from S. Leevers), w;UAS-dInR and w;UAS-dPTEN/CyO (gift from E. Hafen).

• Analysis of induced transcripts: w; +;daGAL4.

• Proliferation and growth rate measurements: w;+;P[w⁺,Act>CD2>GAL4] and w,hsFLP; Tub>GAL4,UAS-GFP/CyO (gift from B. Edgar).

• Weight measurement: sn³;+;Df(3L)emc/TM2Ubx; ban¹/TM3 (HIPFNER et al. 2002 ); EP3622 (Szeged Stock Center).

Generation of the UY collection:
The P{y+}UAS transposon, derived from the Casper4 vector, contains a yellow marker, five copies of Gal4-binding sites and the hsp70 promoter (J. MERRIAM, personal communication). The P element is designed so that the GAL4-dependent transcription starts from the basal hsp70 promoter within the P element and extends into the flanking DNA throughout the 5' end. To generate the UY line collection, UY element was mobilized from the X chromosome by standard methods in a yw genetic background. Mobilization was carried out only in the males so that insertions on the X chromosome were counterselected. New insertions on chromosomes 2 or 3 were mapped and balanced for further analysis. To minimize genetic background effects, all the lines used in this process (transposon or transposase-bearing line and lines used for balancing the insertions) were previously crossed for at least six generations against a Canton-S reference line. Nine research teams of the UY Consortium collaborated to generate 2500 independent lines.

Gain-of-function screen:
To identify genes affecting imaginal tissue growth, UY lines were first crossed with the MS1096 wing driver. To test the effects of different levels of expression, parents from initial crosses were serially transferred and progeny from individual crosses were raised at 18°, 25°, and 29° during larval and pupal stages. Phenotypes at 29° were always stronger. The selected lines were then crossed with the enGAL4 driver for further analysis (second round of the screen). In this line, GAL4 is expressed in the posterior compartment of imaginal wing disc so that the anterior compartment constitutes an internal control under the same growth conditions. For the selected lines, microscopic images of dissected wings were captured using a CCD camera. Wing surface of the posterior region of the adult wing was measured with the histogram function of Adobe Photoshop and compared with the corresponding anterior region values. Trichome density was calculated by counting the number of hairs in a defined square repositioned at three different places in both compartments. Lines showing significant difference of growth under these new conditions were kept for the third round of the screen. To select lines allowing misexpression of genes specifically implicated in growth control, we tested for interactions between misexpression and the PI3K pathway. Males from individual UY lines were crossed with w/MS1096,UAS-Dp110^CAAX or UAS-Dp110^D954A females (LEEVERS et al. 1996 ) and MS1096; UY/CyO or TM3 balanced males were crossed with w;UAS-dINR or UAS-dPTEN/CyO females (BROGIOLO et al. 2001 ). Phenotypes of lines carrying UY and one of the precited transgenes were compared to phenotypes of lines carrying UY alone. This interaction scheme allowed us to detect and select for suppression of the UY misexpression phenotype. Postmitotic induced cellular growth was tested using the GMR-GAL4 driver.

Molecular characterization of the selected UY lines:
We performed molecular characterization for 43 of 131 of the lines picked after the second round on the basis of phenotypic analysis with the enGAL4 driver. To characterize the regions of P insertion, flanking genomic DNA was amplified by PCR rescue. Total DNA from individual UY lines was digested with MspI or HinpII and ligated under diluted conditions. Genomic DNA immediately downstream of the P element (5' end) was then amplified by inverse PCR using OUY31 (5' ATTGATTCACTTTAACTTGCAC 3') and OUY52 (5' ACACAACCTTTCCTCTCAACAA 3') P-specific primers. PCR products were sequenced using OUY53 primer (5' ATACTTCGGTAAGCTTCGGCTATCGACG 3'). This allowed precise localization of P insertions and identification of known or predicted genes located immediately downstream from the 5' end of the UY element. All sequences were performed on a Perkin-Elmer (Norwalk, CT) Genetic Analyzer 310. Sequence similarity searches were performed using the BLASTN or BLASTX program (ALTSCHUL et al. 1997 ) with National Center for Biotechnology Information, BDGP, and Celera nucleotide and protein databases. For the selected lines, molecular characterization allows identification of the following predicted genes: UY15, CG3166; UY48 and 669, CG5261; UY158, CG8895; UY510 and 2050, CG4827; UY562, CG5393; UY582, 3006, 3122, 1206, and 1779, CG8205; UY636, CG8531; UY641, CG6889; UY681, CG11128; UY752, CG11901; UY822, CG2525; UY1290, CG7123; UY788, 3167, 1240, and 1342, CG6494; UY2825, CG4531; UY1505, CG12214; and UY1783, CG3909.

Analysis of wild type and GAL4-induced transcripts:
GAL4-induced transcripts were analyzed by reverse transcriptase (RT)-PCR using F₁ larvae from UY1678;daGAL4 crosses grown at 29°. Total RNA was isolated using standard protocol (Trizol; Sigma, St. Louis) and reverse-transcribed using long template reverse transcriptase (Roche) with oligo(dT) or specific downstream primers. PCR was performed using the Expand long template PCR kit (Roche) using either OUY52' (5' ACACAACCTTTCCTCTCAACAAGCAAACGTGC 3') or specific 5' genomic primers and the 3' primers used for RT. PCR products were gel-purified, blunt-ended with Klenow fragment of DNA polymerase (Promega, Madison, WI), phosphorylated with T4 nucleotide kinase (Promega), and subcloned into SmaI-digested pUC18 vector. Independent clones were sequenced with vector primers. We identified numerous transcripts but only 1–4 are described here. Transcripts 1 and 2 were identified by RT-PCR on daGAL4;UY1678 larvae using OUY52' and oligo(dT), 3 was identified by RT-PCR on wild-type larvae using MB40 (5' CTGTCGTCGTCGGCCATTCGGGTTC 3') and MB05 (5' CGTAGTCAAATCATTGGAGTTAC 3') genomic primers, and 4 was identified by screening a 0- to 4-hr wt embryo cDNA library with a 2-kb genomic probe [nucleotides (nt) 12980–14824 on AE003469].

Northern blot analyses were performed either on total RNAs or on doubly affinity-purified poly(A)⁺ RNA (as mentioned). The 5' P-element probe was amplified by PCR using OUY50 (5' GCAAAGTGAA CACGTCGCT 3') and OUY36 (5' CATGATGAAATAACATAAGG 3') primers.

Proliferation and growth rate measurements:
Fifty larvae from 0- to 3-hr collections were transferred to yeasted vials 24 hr after egg deposition (AED). Misexpression flip-out clones were induced at 37°, either 40 or 72 hr AED and were evaluated 115 hr AED for cell-doubling time, clonal growth, or FACS analysis as described in NEUFELD et al. 1998 . Clones were imaged on a Leica DMR confocal microscope. Cell doubling times were calculated using the formula (log 2/log n)h, where n is the median number of cells per clone and h the age of the clone (in hours). Clone areas were measured using the histogram function of Adobe Photoshop. FACS analysis was performed on a Becton Dickinson FACS Vantage; cell size (FSC) and cell cycle phasing were determined with the Cell Quest software (Becton Dickinson). Controls were done in parallel, and each experiment was scored blind and performed at least twice.

Weight measurements:
For the weight measurements, we analyzed the progeny of the following crosses: w with w; w with sn³;+;Df(3L)emc/TM2Ubx; UY3207 with w; UY3207 with sn³ ;+;Df(3L)emc/TM2Ubx, armGAL4/CyO; UY3207/TM2Ubx with w and armGAL4/CyO; and UY3207/TM2Ubx with sn³;+;Df(3L)emc/TM2Ubx. Fifty L1 larvae from 0- to 12-hr collections were transferred to yeasted vials, 50 larvae per vial, and raised at 25°. Male progeny of various crosses were collected for 24 hr, sorted according to genotype as indicated by visible markers and aged for 3 days after eclosion. The weight of the females being considerably influenced by the weight of the ovaries, we used only the male progeny to minimize eventual sterility effects (MEYER et al. 2000 ). Collections were repeated on three consecutive days. Pools of 50 flies were weighed on a precision balance (Mettler) and the average fly weight was calculated using at least three different measurements (n on Fig 6). We used the weight of male flies averaged from the three consecutive collections to calculate the weight ratio (average weight of test genotype population/average weight of control population).

View larger version (54K): In this window In a new window Download PPT slide   Figure 1. Misexpression at the clochette locus enhances tissue growth. UY1678 and UY3207 insertion lines were crossed with either MS1096 (A) or enGAL4 (B) wing drivers. MS1096 expresses Gal4 preferentially in the dorsal compartment of the wing imaginal disc and UY3207 misexpression causes a downward curvature of the wing. When crossed with enGAL4, only the posterior compartment presents overgrowth. (C) Area and hair density were measured on 10 separate wings for both test and control conditions. Hair density is not changed by misexpression. Test area is enlarged and control area is reduced by a compensation mechanism.

View larger version (119K): In this window In a new window Download PPT slide   Figure 2. clochette misexpression alters postmitotic cell growth in the eye disc. GMR-GAL4 driver expresses Gal4 in postmitotic cells of the eye disc. Misexpression using UY3207 enhances overall eye and individual ommatidial size. Lateral view SEMs x200, enlarged x500 for ommatidial view.

View larger version (72K): In this window In a new window Download PPT slide   Figure 3. UY1678-induced overgrowth is compensated by reducing activation of the PI3K pathway. MS1096 driver is used to induce misexpression in the wing disc. Dorsal overgrowth induces a downward curvature of the wing blade, which is suppressed by coexpressing Dp110D954A, a dominant-negative form of the Dp110 subunit of PI3-kinase.

View larger version (44K): In this window In a new window Download PPT slide   Figure 4. UY1678-induced cellular overgrowth is coupled with an increase of proliferation. (A) FACS analysis of flip-out GAL4 overexpression clones induced at 72 hr AED and analyzed 115 hr AED. Light traces (control flipped-out GFP+ cells) and dark traces (UY3207 flipped-out GFP+ cells) are superimposed for both relative cell sizes (FSC) and cell cycle profiles (DNA). Internal comparison with GFP- profiles is not shown, as GFP- and GFP+ cell cycle profiles showed a small but reproducible difference, in both control and test experiments. (B) For cell-doubling time measurement (DT), flip-out GAL4 clones were induced at 72 hr and analyzed at 115 hr AED. The distribution of number of cells per clone is presented. Number of clones scored (n) and DT (hours) are indicated. Clone surfaces were also measured to evaluate growth rate; in this case, clones were induced at 40 hr and fixed at 115 hr AED for analysis. The distribution of clone surface is presented. Number of scored clones (n) is indicated. All controls were done in parallel.

View larger version (48K): In this window In a new window Download PPT slide   Figure 5. The clochette locus is essential for organismal growth. Reducing cot function limits organismal growth. Adult male flies from the indicated genotypes, grown under controlled nutritional conditions, are shown. Male progeny were selected and weight was measured for pools of 50 flies (n, number of pools analyzed). Weight ratio to wild-type flies was calculated: Df(3L)emc/+, 0.80; arm-GAL4;P(UY)3207, 1.06; and Df(3L)emc;P(UY)3207; arm-GAL4, 0.93.

View larger version (26K): In this window In a new window Download PPT slide   Figure 6. Map of the clochette locus. The four UY and the EP insertions are located between nt 11,404 and 12,980 on AE003469. Arrows represent the sense of transcription from the P elements. Annotated genes (solid arrows) and ESTs (striped arrows) in the region are shown. The four major identified cDNAs are represented. Open reading frames are solid, except for cDNA 4 considered as a nonspliced transcript. Nos. 1 and 2 correspond to GAL4-induced transcripts starting from the UY1678 promoter region. An ATG present in the transposon sequence is used for translation of the longest open reading frame. No. 3 corresponds to an endogenous polyadenylated transcript present in wt embryos and larvae. No. 4 is an abundant 2.8-kb cDNA colinear with genomic sequences, which is present in various Drosophila cDNA libraries and in wt as well as under deregulated conditions. All four cDNAs (1–4) as well as cDNAs corresponding to CG12030 (glucose epimerase, 5) and CG3200 (6) have been reintroduced in transgenic flies for UAS-directed misexpression, with addition of a Cavener consensus for facilitating translation. Limits of the ban1 deletion are indicated. RE64518 corresponds to a truncated form of the 2.8-kb unspliced RNA (4).

Transgenic flies:
cDNAs were cloned into the pUAST vector (BRAND and PERRIMON 1993 ) and constructs were introduced into the germline by injection in the presence of transposase as previously described (RUBIN and SPRADLING 1982 ). For CG12030, full-length cDNA sequence for the LD27852 expressed sequence tag (EST) was obtained from GenBank.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

A misexpression screen for wing growth defects:
As part of a consortium of laboratories (see MATERIALS AND METHODS), we established a collection of 2400 UY independent lines, each containing a single UY insertion on chromosomes 2 or 3. The UY transposon contains five copies of the UAS_Gal4 sequences, which, in the presence of the Gal4 activator, allow directional transcription from the downstream basal hsp70 promoter through the 5' end of the transposon and into the adjacent genomic region. We first screened the UY collection for wing phenotypes using the MS1096 line, which predominantly expresses Gal4 in the dorsal part of the wing pouch during disc development. Our previous studies had shown that expressing growth regulators in the wing disc using this driver introduces curvatures of the wing blade due to differential alteration of tissue growth between dorsal and ventral wing surfaces. We selected 295 lines, most of which (95%) presented an upward curvature of the wing, indicating a growth reduction in the overexpressing dorsal tissue (Table 1). In cases of severe growth inhibition, the wings appeared crumpled with strong structural deformation. Ten lines presented a downward curvature of the wing indicating increased growth in the dorsal compartment (Fig 1A). We then screened the collection using the enGAL4 driver, which drives expression in the posterior part of the wing disc. This second screen confirmed overgrowth effects observed for the 10 lines already identified and gave additional information about final cell size and cell number (see below). Finally, all lines presenting a growth phenotype were tested for genetic interaction with the growth-promoting PI3-kinase pathway. Growth reductions were challenged by coexpression of the Dp110 catalytic subunit of PI3K and overgrowth by coexpression of a dominant-negative form of Dp110 (Dp110^D954A). In all cases, overgrowth was suppressed by coexpression of Dp110^D954A. In contrast, only 1 line showed suppression of growth inhibition by expression of Dp110. This suggested that overgrowth phenotypes were more reliable than growth suppression, which could be due in part to nonspecific cell death. Among the group of 10 lines that showed growth increase upon misexpression, 4 lines (UY3207, UY1079, UY1678, and UY789) presented P insertions localized within a 3-kb interval at cytological location 61C8 and directed transcription of the same downstream genomic sequences (nt 11,404–12,947 on AE003469, see Fig 5 and molecular analysis below). We named this genetically defined locus clochette (cot).

Misexpression at the cot locus increases organ growth:
UY1678 corresponds to the most proximal insertion site in cot. When it was crossed with the en-GAL4 driver, we observed a strong enlargement of the posterior compartment of the adult wing (Fig 1B). Quantification of trichome on the surface of the wing blade indicated that trichome density in wings of en-GAL4>UY1678 flies was identical to that of controls (Fig 1C). Since each trichome corresponds to one cell of the wing blade, this indicated that cot misexpression did not alter cell size. Measurements indicated a 31% surface increase in posterior wing compartment of en-GAL4>UY1678 flies as compared to controls (Fig 1C). Growth increase was restricted to the posterior part of the wing where misexpression was induced, indicative of a cell-autonomous response. Interestingly, overgrowth in the posterior compartment was accompanied by a reduction of the anterior compartment, suggesting that a nonautonomous cross-compartment compensation was taking place. Nevertheless, overall wing size was still increased under these conditions.

To test whether cot misexpression could induce growth in other tissues during development, we transactivated UY1678 in the eye imaginal disc using the GMR-GAL4 line, which drives Gal4 expression in postmitotic cells posterior to the morphogenetic furrow. As in the adult wing, cot misexpression provoked a strong enlargement of the adult eye (Fig 2). In this case, individual ommatidia were strongly enlarged, suggesting that the size of individual cells composing each ommatidium was increased. Remarkably, the appearance of the eye surface was well preserved, and optical sections of adult retina did not reveal any severe abnormality (not shown). Thus, while cell growth was strongly induced, differentiation of ommatidial cells occurred normally.

As part of the initial screen, these growth effects were challenged by coexpression with a dominant-negative form of PI3K (Dp110^D954A) or dPTEN, a negative regulator of the PI3-kinase pathway. In both cases, overgrowth was suppressed (Fig 3). Remarkably, Dp110^D954A expression provoked an almost complete reversion to normal wing size and morphology. This suggested that misexpression of the cot locus was acting exclusively on organ growth without interfering with other functions during disc development.

Cell growth and proliferation rates are affected by misexpression of the cot locus:
We studied the cycling parameters of proliferating wing disc cells under misexpression conditions. For this purpose, misexpression was induced in random clones using the flip-out GAL4 technique. This technique allows the distinction between deregulated (GFP-positive) and control (GFP-negative) cells. Dissociated disc cells were analyzed by FACS for cell size and DNA content 44 hr after induction. cot misexpression did not modify cell size (FSC values) relative to control, in accordance with our previous evaluation of hair densities in adult wings. Cell cycle phasing was also unaltered compared to control (Fig 4). We measured cell-doubling times by counting the average number of cells in misexpression and control clones induced at identical times. Under these conditions, cell-doubling time was reduced by 15% in misexpression clones as compared to control clones (DT values, Fig 4). Thus, misexpression of the cot locus induced an acceleration of the division rate with no variation in the average cell size. This suggested that, during this period of rapid cell proliferation, cell growth was also increased. Indeed, measurement of clone surfaces indicated a growth acceleration in deregulated clones compared to control clones (Fig 4).

Therefore, in mitotically active cells, cot misexpression accelerated growth and proliferation coordinately with no alteration in cell size or cell cycle phasing. In contrast, enlargement of final ommatidium size in the adult retina suggested that cot misexpression could promote cell growth in postmitotic eye cells, leading to cellular hypertrophy.

The cot locus is required for normal growth:
Since transactivation at the cot locus elicited strong tissue overgrowth, we predicted that reducing cot copy number would affect final adult size. Df(3L)emc presents a deletion of 150 kb across the cot locus. For animals carefully grown under controlled nutritional conditions, average weight of Df(3L)emc/+ heterozygous flies was decreased by 20% as compared to wt (Table 2). This weight decrease was due to a reduction of overall body size. When Df(3L)emc was present in trans to UY3207, the weight reduction was enhanced, indicating that a growth phenotype due to the P insertion was revealed in the sensitized genetic background of the deficiency (Fig 5; Table 2). Accordingly, this aggravation was reverted when UY3207 was precisely excised from the locus (not shown). Inversely, in a Df(3L)emc /UY3207 background, ubiquitous UY3207 misexpression using the armGAL4 driver rescued the growth defect almost to the level of wt flies (Fig 5; Table 2). Thus partial loss of function obtained by combining Df(3L)emc and UY3207 could be rescued by enhanced transcription of the cot locus. Closer examination of the growth phenotypes revealed that wings from Df(3L)emc/+ animals were 12% smaller than those from wt (Table 2). In animals with Df(3L)emc in trans to UY3207, wing size was further reduced. Comparable data were obtained in a different genetic background using EP3622, a P(UAS) insertion from the EP collection (RORTH 1996 ) localized 36 bp upstream of UY3207 (Table 2; Fig 6), which induced similar overgrowth by misexpression (data not shown). By contrast, insertions UY1079, UY1678, and UY789, positioned on either side of the EP3622;UY3207 group, did not enhance the Df(3L)emc wing size reduction, suggesting that these insertion points do not alter cot function (Table 2; Fig 6).

While this work was under review, HIPFNER et al. 2002 reported the identification of a growth-controlling locus named bantam (ban) in the vicinity of cot. To narrow down the localization of the cot gene, we made use of a smaller deletion, ban¹, which removes 22 kb in the region common to both loci (HIPFNER et al. 2002 ). Heterozygous ban¹/+ animals were slightly smaller than w¹¹¹⁸ controls and presented a reduction of wing surfaces due to a reduction of cell number, without affecting cell size (Table 2). We then observed animal wings in different heteroallelic combinations of ban¹ and P insertions at the cot locus. ban¹/UY3207 and ban¹/EP3622 wings were further reduced, unlike wings from ban¹/UY1079, ban¹/UY789, or ban¹/UY1678 animals, confirming the genetic interactions observed with Df(3L)emc (Table 2). Additionally, this aggravation was due strictly to a further decrease in cell number since we observed no effect on cell size (Table 2). These results establish that the cot locus positively regulates growth through coupled cell growth and cell proliferation and fall in complete agreement with our gain-of-function analysis.

We further tested whether ban and cot could correspond to the same locus. EP3622 was described as a hypomorphic ban allele (HIPFNER et al. 2002 ). Although neither EP3622 nor UY3207 presented a phenotype as heterozygous, EP3622/UY3207 flies presented a reduction of body size comparable to the one observed in ban¹/UY3207, ban¹/EP3622, or EP3622/EP3622 animals (Table 2). This indicates that UY3207 and EP3622 do not complement each other and thus establishes that ban and cot correspond to the same locus (hereafter called ban/cot).

Molecular characterization of the ban/cot locus:
For all five lines, P elements are inserted so that misexpression induces transcription of genomic regions on the same DNA strand, toward ascending numbers on AE003469 (Fig 6). The first downstream gene, CG12030, which encodes a glucose epimerase, is located 7 kb downstream of UY1678, the 3'-most P insertion. Northern blot analysis indicated that misexpression induced a twofold increase in transcription of this gene. This could be either a direct effect of misexpression or an indirect effect due to extra cell growth. We generated transgenic flies expressing CG12030 under the control of the Gal4 UAS. Ectopic CG12030 expression was not sufficient to recapitulate the overgrowth phenotype observed in the UY lines. Other predicted genes located up to 50 kb away on both sides of the P insertion sites were also tested on Northern blots but none was induced by misexpression in the UY lines (Table 3). These data suggested that transactivation of an unannotated sequence might be responsible for the overgrowth phenotypes.

We searched for other transcription units in the region using either genomic probes in Northern blotting and cDNA library screens or RT-PCR with primers corresponding to genomic sequences (Fig 6 and Fig 7). This analysis revealed a complex family of alternatively spliced transcripts present at very low levels in wt larvae and embryos (detected by RT-PCR only, data not shown), together with an abundant 2.8-kb unspliced species, which could correspond to a nonmatured precursor (detected in library screens, RT-PCR, and Northern blot experiments; Fig 7, lane 6; also represented as 4 on Fig 6). The spliced RNAs contained short open reading frames potentially encoding a series of short peptides (<120 amino acids) presenting no homology to known proteins. These RNAs were overexpressed under misexpression conditions, from either UY3207 or UY1678. In this case a major band was detected on Northern blots using either a P-derived probe (Fig 7, lanes 2 and 3) or a cDNA probe (Fig 7, lane 5), which corresponds to misexpression of the major unspliced RNA. A smear of shorter overexpressed transcripts was also visualized on Northern blot using poly(A)⁺ RNAs (Fig 7, bracket on lane 5). These likely correspond to the spliced transcripts detected by RT-PCR in wt tissues. Indeed, RT-PCR experiments also detected spliced transcripts after misexpression (represented as 1, 2, and 3 on Fig 6), with the splicing diversity found in wild-type cells.

View larger version (29K): In this window In a new window Download PPT slide   Figure 7. Northern analysis of misexpression-induced transcripts. GAL4-induced transcripts at the clochette locus were analyzed by Northern blot on total (lanes 1, 2, and 3) or poly(A)+ RNAs (lanes 4, 5, and 6) from daGAL4, daGAL4;UY1079, and daGAL4;UY1678 larvae as well as wt embryos. For both UY1678 and UY3207 lines, GAL4 induces expression of a major transcript starting within the UY element (lanes 2 and 3). This transcript (2*) is also detected using cDNA probe 1 (5*). Its size suggests that it corresponds to a truncated form of cDNA 4. In control larvae, no transcript could be detected using cDNA 1 probe. In wt embryos, this probe allowed visualization of a major 2.7-kb band (lane 6) corresponding in size to cDNA 4, also detected in RT-PCR experiments and in several cDNA libraries.

We then tested the ability of the three major RT-PCR species (1–3 on Fig 6), as well as the 2.8-kb cDNA (4 on Fig 6), to recapitulate overgrowth when induced under UAS control in transgenic animals. None of these constructs reproduced the overgrowth phenotype of the UY flies, suggesting either that a precise combination of multiple transcripts is necessary or that other sequences are responsible for the growth effect.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

By using a gain-of-function screen, we have identified a genetic locus that we initially called clochette (cot), which is involved in growth control. All misexpression phenotypes obtained with the five different UY insertions at this locus were similar and varied only in intensity. In each case, sizes of adult appendages were larger than those of control. This effect could be induced at the level of the entire animal by using a ubiquitous driver (see Fig 5). At the cellular level, cell size was not affected and cell number was increased, indicating a coordinate induction of cell growth and cell division.

Genetic characterization of loss-of-function phenotypes defined a locus functioning as a positive growth regulator. We found allelic combinations that reduced adult body size by reducing cell number without affecting cell size. This is in complete accordance with the cellular phenotype obtained in our gain-of-function experiments, which provoke a coupled increase in growth and proliferation. Further genetic experiments establish that the cot locus is allelic to the ban locus, recently published while our work was under review (HIPFNER et al. 2002 ).

The identical orientation of the four different UY insertions at the locus strongly suggests that the misexpression phenotypes are caused by transcription of the genomic region downstream of the 5' ends of UY insertions. Candidate annotated genes either were not transactivated upon misexpression or did not recapitulate the growth phenotype when overexpressed under UAS control. Our molecular analysis of the locus reveals a new transcription unit responsible for the generation of a complex family of transcripts initiating in the region of the P insertions and transcribing in the expected direction. Transcription of these different RNAs is activated upon misexpression, but separate ectopic expression of the three most abundant mature transcripts did not induce overgrowth. The major transcript we found is an unspliced RNA that spans >2.8 kb and could correspond to a precursor for different RNA species. This suggests that a precise control on splicing might regulate relative abundance of the different transcripts in vivo. In this respect, we cannot exclude that a specific combination of several induced transcripts might be necessary to promote overgrowth. Interestingly, UY3207 and EP3622, the two P-element insertions that significantly enhance the growth phenotype over the deletion, are positioned in the 5' region of the most complete RNA we identified, whereas UY1678, which does not genetically interact with the deletion, is inserted in a predicted intronic region. This correlates with the fact that a transcription unit corresponding to the cot gene might indeed be located in the vicinity. Attempts to identify EMS-induced revertants of the GAL4-dependent overgrowth phenotype have been unsuccessful, leaving open the possibility that a less canonical, non-protein-coding mechanism might be involved (F. ROUYER and E. BLANCHARDON, unpublished data; our unpublished data).

Although we showed that overgrowth could be efficiently compensated by coexpression of a dominant-negative version of PI3-kinase, we do not think that the ban/cot growth regulator is part of the PI3-kinase pathway. At the cellular level, when PI3K activation increases cell size (WEINKOVE et al. 1999 ), ban/cot misexpression accelerated cell divisions without change in individual cell size. In accordance with this gain-of-function analysis, heteroallelic combinations at the ban/cot locus established that cell number is decreased when cell size is not affected. This precise coupling between cell growth and cell proliferation is not frequently observed in gain-of-function experiments and is similar to what is obtained when the Cdk4/cyclin D growth inducer is overexpressed (PROBER and EDGAR 2000 ). Accordingly, Cdk4 loss of function leads to a decrease in adult body size (MEYER et al. 2000 ), which is comparable to the most effective ban/cot loss-of-function combination. However, genetic analysis suggests that these two regulators might function independently (HIPFNER et al. 2002 ). Further experiments will be necessary to test what the respective contributions of ban/cot and Cdk4/cyclin D activities are in the coupling of cell growth and cell division.

	FOOTNOTES

¹ These authors contributed equally to this work.

	ACKNOWLEDGMENTS

We thank the laboratories of the consortium for UY mutagenesis and especially colleagues from P. Thérond and F. Rouyer laboratories for sharing unpublished results. We thank Sally Leevers, Jacques Montagne, and Stephen Cohen for providing flies and Nicolas Tapon for careful reading of the manuscript. This work was supported by the Centre National de la Recherche Scientifique, the Institut National de la Santé et de la Recherche Médicale, the Association pour la Recherche contre le Cancer, and the Fondation pour la Recherche Médicale.

Manuscript received July 4, 2002; Accepted for publication March 19, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALTSCHUL, S. F., T. L. MADDEN, A. A. SCHAFFER, J. ZHANG, and Z. ZHANG et al., 1997  Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25:3389-3402.[Abstract/Free Full Text]

BRAND, A. H. and N. PERRIMON, 1993  Targeted gene expression as a means of altering cell fates and generating dominant phenotypes. Development 118:401-415.[Abstract]

BROGIOLO, W., H. STOCKER, T. IKEYA, F. RINTELEN, and R. FERNANDEZ et al., 2001  An evolutionarily conserved function of the Drosophila insulin receptor and insulin-like peptides in growth control. Curr. Biol. 11:213-221.[Medline]

CONLON, I. and M. RAFF, 1999  Size control in animal development. Cell 96:235-244.[Medline]

DATAR, S. A., H. W. JACOBS, A. F. DE LA CRUZ, C. F. LEHNER, and B. A. EDGAR, 2000  The Drosophila cyclin D-Cdk4 complex promotes cellular growth. EMBO J. 19:4543-4554.[Medline]

EDGAR, B. A. and T. L. ORR-WEAVER, 2001  Endoreplication cell cycles: more for less. Cell 105:297-306.[Medline]

HIPFNER, D. R., K. WEIGMANN, and S. M. COHEN, 2002  The bantam gene regulates Drosophila growth. Genetics 161:1527-1537.[Abstract/Free Full Text]

JOHNSTON, L. A., D. A. PROBER, B. A. EDGAR, R. N. EISENMAN, and P. GALLANT, 1999  Drosophila myc regulates cellular growth during development. Cell 98:779-790.[Medline]

LEEVERS, S. J., D. WEINKOVE, L. K. MACDOUGALL, E. HAFEN, and M. D. WATERFIELD, 1996  The Drosophila phosphoinositide 3-kinase Dp110 promotes cell growth. EMBO J. 15:6584-6594.[Medline]

MEYER, C. A., H. W. JACOBS, S. A. DATAR, W. DU, and B. A. EDGAR et al., 2000  Drosophila Cdk4 is required for normal growth and is dispensable for cell cycle progression. EMBO J. 19:4533-4542.[Medline]

NEUFELD, T. P. and B. A. EDGAR, 1998  Connections between growth and the cell cycle. Curr. Opin. Cell Biol. 10:784-790.[Medline]

NEUFELD, T. P., A. F. DE LA CRUZ, L. A. JOHNSTON, and B. A. EDGAR, 1998  Coordination of growth and cell division in the Drosophila wing. Cell 93:1183-1193.[Medline]

PROBER, D. A. and B. A. EDGAR, 2000  Ras1 promotes cellular growth in the Drosophila wing. Cell 100:435-446.[Medline]

PROBER, D. A. and B. A. EDGAR, 2001  Growth regulation by oncogenes—new insights from model organisms. Curr. Opin. Genet. Dev. 11:19-26.[Medline]

RORTH, P., 1996  A modular misexpression screen in Drosophila detecting tissue-specific phenotypes. Proc. Natl. Acad. Sci. USA 93:12418-12422.[Abstract/Free Full Text]

RUBIN, G. M. and A. C. SPRADLING, 1982  Genetic transformation of Drosophila with transposable element vectors. Science 218:348-353.[Abstract/Free Full Text]

STOCKER, H. and E. HAFEN, 2000  Genetic control of cell size. Curr. Opin. Genet. Dev. 10:529-535.[Medline]

TAPON, N., K. H. MOBERG, and I. K. HARIHARAN, 2001  The coupling of cell growth to the cell cycle. Curr. Opin. Cell Biol. 13:731-737.[Medline]

WEINKOVE, D., T. P. NEUFELD, T. TWARDZIK, M. D. WATERFIELD, and S. J. LEEVERS, 1999  Regulation of imaginal disc cell size, cell number and organ size by Drosophila class I(A) phosphoinositide 3-kinase and its adaptor. Curr. Biol. 9:1019-1029.[Medline]

XIN, S., L. WENG, J. XU, and W. DU, 2002  The role of RBF in developmentally regulated cell proliferation in the eye disc and in Cyclin D/Cdk4 induced cellular growth. Development 129:1345-1356.

This article has been cited by other articles:

				A. Khokhar, N. Chen, J.-P. Yuan, Y. Li, G. N. Landis, G. Beaulieu, H. Kaur, and J. Tower Conditional Switches for Extracellular Matrix Patterning in Drosophila melanogaster Genetics, March 1, 2008; 178(3): 1283 - 1293. [Abstract] [Full Text] [PDF]

				C. M. A. Coelho, B. Kolevski, C. D. Walker, I. Lavagi, T. Shaw, A. Ebert, S. J. Leevers, and S. J. Marygold A Genetic Screen for Dominant Modifiers of a Small-Wing Phenotype in Drosophila melanogaster Identifies Proteins Involved in Splicing and Translation Genetics, October 1, 2005; 171(2): 597 - 614. [Abstract] [Full Text] [PDF]

				J. J. Krupp, L. E. Yaich, R. J. Wessells, and R. Bodmer Identification of Genetic Loci That Interact With cut During Drosophila Wing-Margin Development Genetics, August 1, 2005; 170(4): 1775 - 1795. [Abstract] [Full Text] [PDF]