Compensatory growth in novel Drosophila Akt1 mutants

Organisms, tissues and cells are genetically programmed to grow to a specific largely pre-set size and shape within the appropriate developmental timing. In the event of mutation, cell death, or tissue damage, the remaining cells may increase their rate of growth to compensate and generate an intact, potentially smaller, tissue or organism in order to achieve the desired size. A delay in the developmental timing could aid in this process. The insulin receptor signalling pathway with its central component, the Akt1 kinase, and endpoint regulator, the transcription factor foxo, plays a significant role in the control of growth. Drosophila melanogaster is an excellent model organism with a well-studied life cycle and a consistently developing compound eye that can undergo analysis to compare changes in the properties of adult ommatidia as an indicator of growth. Imprecise excision of a PZ P-element inserted in the upstream region of Akt1 generated several novel hypomorphic alleles with internally deleted regions of the Pelement. These mutations lead to small, viable Drosophila that present with delays in development. Suppression of this phenotype by the directed expression of Akt1+ indicates that the phenotypes observed are Akt1 dependent. Somatic clones of the eyes, consisting of homozygous tissue in otherwise heterozygous organisms that develop within a standard timeframe, signify that more severe phenotypes are masked by an extension in the time of development of homozygous mutants. Generation of Drosophila having the hypomorphic Akt1 alleles and a null allele of the downstream target foxo result in a phenotype very similar to that of the foxo mutant and do not resemble the Akt1 mutants. The developmental delay of these novel Akt1 hypomorphs results in a latent phenotype uncovered by generation of somatic clones. The compensatory growth occurring during the extended time of development appears to be implemented through alteration of foxo activity. Production of clones is an effective and informative way to observe the effects of mutations that result in small, viable, developmentally delayed flies.

Findings: Imprecise excision of a PZ P-element inserted in the upstream region of Akt1 generated several novel hypomorphic alleles with internally deleted regions of the Pelement. These mutations lead to small, viable Drosophila that present with delays in development. Suppression of this phenotype by the directed expression of Akt1 + indicates that the phenotypes observed are Akt1 dependent. Somatic clones of the eyes, consisting of homozygous tissue in otherwise heterozygous organisms that develop within a standard timeframe, signify that more severe phenotypes are masked by an extension in the time of development of homozygous mutants. Generation of Drosophila having the hypomorphic Akt1 alleles and a null allele of the downstream target foxo result in a phenotype very similar to that of the foxo mutant and do not resemble the Akt1 mutants. Conclusion: The developmental delay of these novel Akt1 hypomorphs results in a latent phenotype uncovered by generation of somatic clones. The compensatory growth occurring during the extended time of development appears to be implemented through alteration of foxo activity. Production of clones is an effective and informative way to observe the effects of mutations that result in small, viable, developmentally delayed flies.

Background
The cell is the basic structural unit of all living organisms. The overall size of a cell can either augment or limit its ability to perform essential functions. Consequently size homeostasis is pertinent for the fitness and function of cells. Even slight disruption of this homeostasis can lead to disease, thus it is critical to understand the complex mechanisms that control cell growth. Drosophila melanogaster develops quickly through a sequence of three feeding and growing larval stages followed by pupation and eclosion [1] and is an ideal model system in which to study cell growth.
A crucial point in the control of growth in Drosophila is the achievement of the critical mass, the minimum weight required for transition from larvae to pupae, upon which any further feeding, or lack of feeding, will not prevent this change [2,3]. Drosophila larvae, when fed generously, can grow to, or past, the critical weight within four days. Restriction of dietary proteins slows this process, while total absence can halt growth completely [4]. Once larvae have reached the critical weight required for pupation, they may continue to feed for a period of time before undergoing the transition [5]. Several factors can influence the rate of growth during the larval stages including nutrition, temperature, density of organisms present in the environment, and underlying genetic mechanisms [6][7][8][9][10]. Slowed growth, due to genetic mechanisms or nutrient conditions, characteristically results in larvae that develop into smaller adults. While many mutations can influence growth; some alter the growth of individual organs, some retard overall growth without changing the final adult size, the mutations which slow growth and lead to a reduction in the overall organ and body size may be the most intriguing.
The conserved insulin receptor (InR) signalling pathway is implicated in the management of final adult size. In Drosophila, this highly conserved pathway has been shown to control cell size and growth, and to regulate body size and nutrient usage [11,12]. When any of the seven Drosophila insulin-like peptide (Ilp) genes are overexpressed, growth rates in larvae and adults are greatly increased, and ablation of the medial neurosecretory cells in the brain (the main producer of Ilps) leads to a decrease in the growth rate and final size [13]. Overexpression of the upstream components of the pathway, including the ligand (Ilps), the insulin receptor (Inr) and the insulin receptor substrate (chico), in Drosophila results in larger than normal flies, while mutation or loss of function of these components results in size reduction and developmental delay [14]. This reinforces the pivotal role of insulin receptor signalling in the control of growth.
The Akt1 kinase is a central component of the insulin receptor signalling pathway. When Drosophila Akt1 is overexpressed, it is shown to increase cell size but not proliferation, or number of cells, by overriding the control mechanisms that are responsible in determining the final size of cells [15]. Loss of Akt1 can result in lethality [16] while hypomorphic activity can result in the production of smaller adults [15]. A key downstream target of Akt1, the transcription factor foxo, mediates the transcriptional regulation of the insulin pathway and controls several important cellular functions including metabolism, cell cycle regulation, DNA repair, apoptosis and protection of the cell against oxidative stress [17][18][19][20]. Through these diverse functions, the transcription factor foxo can facilitate the end result of Akt11 activity upon the regulation of cell growth and survival.
In order to explore the influence of Akt1 activity upon cell growth, a series of novel Akt1 hypomorphs were generated through imprecise excision of a P-element situated in the control region upstream of the gene's coding region. A subset of these hypomorphs were selected, based upon reduction in adult size, and were further characterized with replacement analysis to confirm the reduction in size was due to altered Akt1 activity. Due to the extended time required to reach eclosion by the homozygotes, somatic clones of the eye were generated to produce a mutant phenotype less influenced by a developmental delay. Finally, to further investigate the dynamic interaction between Akt1 and foxo, Drosophila lines with novel hypomorphic alleles and a null version of the downstream target foxo gene were created and the potential for epistasis was evaluated. Our intent through these experiments is to better understand the effect of extended development time upon the overall phenotype of the novel Akt1 hypomorphs.

Drosophila stocks, media and culture
The initial P-element insertion line ry 506 P{PZ}Akt11 04226 / TM3, ry RK , Sb 1 , Ser 1 (Akt1 04226 ) was obtained from the Bloomington Drosophila Stock Center. This line contains a P-element inserted within the 5′ untranslated region of the Akt1 gene on the third chromosome. Initial reports of this allele state that it is semi-lethal [21], but we, in addition to other researchers [22,23] have found the homozygotes to be viable. The control line w 1118 ; P[FRT; w + ] 2A P[ry + neo R FRT] 82B Akt1 + was derived from lines obtained from Norbert Perrimon of Harvard University. The PΔ2-3, ry + line was utilized to generate the novel Akt1 mutants [24]. The foxo null mutant line w; foxo W124X was obtained from Drs. E. Hafen and M. Junger [25] of the University of Zurich. Wild-type Oregon R (OrR) stock was obtained from the Bloomington Drosophila Stock Center and w 1118 was obtained from Dr. Howard Lipshitz from the University of Toronto. Stocks and crosses were maintained on a standard medium containing cornmeal, molasses, yeast, agar and water. Routinely, stocks were kept at room temperature (22 ± 2°C) while crosses and experiments were carried out at 25°C.

Generation of Drosophila lines
Hypomorphic alleles of Akt1 04226 were generated via P-element excision by crosses to a line containing a stable source of transposase, PΔ2-3. The critical class offspring of the dysgenic males and Ly/TM3, Sb ry females were selected based upon loss of the PZ P-element by the presentation of the rosy eye colour phenotype. These novel alleles were expected to differ from the Akt1 + line and Akt1 04226 by the resultant alterations of the PZ P-element and/or the adjacent Akt1 sequences. To allow for clonal analysis, recombinants of w; P[FRT;w + ] 2A P[ry + neo R FRT] 82B and the novel derivatives of Akt1 04226 were generated and balanced over TM6B, Hu Tb e. Of the derivatives generated, a subset of these recombinants were selected for analysis based on the appearance of non-Tubby homozygotes.
Replacement studies were carried out by generating independent lines of w 1118 ; UAS-Akt1 + /CyO; Akt1 m / TM6B and w 1118 ; arm-GAL4/CyO; Akt1 m /TM6B where m represents each of the novel Akt1 mutant alleles. Crosses between these lines generated the critical class of w 1118 ; UAS-Akt1 + /arm-GAL4; Akt1 m /Akt1 m to be analyzed.
Generation of flies bearing both the novel mutant Akt1 alleles and a null foxo allele was performed via standard recombinant methods. As Akt1 52 and Akt1 57 exhibited the greater developmental delay, these alleles were selected for recombination with the null allele of foxo. From these combinations, a series of lines were selected based upon adult phenotypes and confirmed through PCR and sequencing.

Molecular characterization of the novel hypomorphs
Homozygous wild type, novel hypomorphic and double mutant Drosophila samples were collected from crosses of adult heterozygous female virgins to heterozygous males of each genotype. Ten homozygous male flies were collected upon eclosion and aged three to five days before being flash-frozen at −70°C. DNA was extracted from each sample via a standard phenolchloroform protocol. The Flybase database (http://flybase.org) includes the complete sequence for the PZ P-element positioned within the Akt1 04226 line, and the National Center for Biotechnology Information (NCBI) database (http://www.ncbi.nlm.nih.gov/) includes the complete gene sequence for Akt1 + . To design oligonucleotides indicating the breakpoint region of each novel mutant, a series of oligonucleotides both flanking and spanning the P-element insertion site was carried out via Primer3. PCR analysis via primers spanning the PZ Pelement revealed forward primer sites present and positioned near the breakpoints. A reverse primer positioned near the end of the PZ P-element allowed for generation of breakpoint fragments using HotStart Taq Polymerase (Qiagen Inc.) in an Eppendorf Mastercycler gradient thermal cycler through standard methods. Gels were photographed with a ChemiImageTM Ready 4400 v5.5 photodocumentation system. Purification and sequencing of the PCR products was completed at the Genomics and Proteomics (GaP) facility, Memorial University of Newfoundland.

Analysis of developmental timing
Heterozygotes are identified based upon the presence of Humeral (Hu), an allele of Antennapedia carried by the TM6B, Hu Tb e balancer chromosome that results in extra bristles on the outer edges of the prothorax, while homozygous mutant flies lack this marker. Heterozygous females and males of the novel Akt1 hypomorphs were transferred to fresh media, incubated at 25°C for six hours to allow for egg-laying then removed. Vials were returned to the incubator immediately after removal of adult flies and examined each morning for fifteen days. Observations included days until pupation and eclosion. Pupae and adult flies were scored as Tubby heterozygotes or non-Tubby mutants and used to generate developmental delay line graphs in Graph-Pad Prism Version 5.03.

Biometric analysis of Drosophila eyes
Critical class males of the homozygous mutant Akt1 alleles, the transgenic rescues, the somatic clones and the double mutant lines were collected and aged for three days. Flies were then flash-frozen at −70°C before preparation for scanning electron microscopy. Preparation included mounting upon aluminum SEM studs, desiccated and sputter coating in gold. Images were taken with either a Hitachi S-170 or S-570 Scanning Electron Microscope as per standard methods and analyzed using NIH Image J software [27].

Results
Three novel Akt1 mutants retain portions of inserted PZ P-element Molecular characterization of the three small viable Akt1 mutants revealed internally deleted versions of the PZ P-element at the original point of insertion ( Figure 1). Analysis indicates that the retained sections are from both ends of the PZ P-element. Akt1 87 possesses the largest deleted region (11029 base pairs) between nucleic acids 2184 (within lacZ) and 13213 (within ry + ) of the PZ P-element sequence. The next largest deletion (9532 base pairs) is in Akt1 52 between nucleic acids 2754 (within lacZ) and 12307 (within ry + ). Lastly Akt1 57 has the smallest deletion of the three mutants of 3259 base pairs between nucleic acid 4315 (within the HSP70 polyadenylation control region) and 7574 (within ry + ). Each deletion includes a part of the ry + gene, responsible for the phenotype (ry − ) upon which these mutants were selected. No alteration to the Akt1 coding region sequence was detected in the three mutants.

Three novel Akt1 mutants are developmentally delayed
In the development from embryo to adult, Akt1 mutant heterozygotes are similar to controls ( Figure 2). The formation of heterozygous (Tubby) pupae occurs in a similar time-frame to the control lines (data not shown). The time required to eclose by adult Akt1 heterozygotes (Humeral) and control flies is nine to ten days ( Figure 2). Emergence of the homozygous adult flies is delayed by two to four days. One allele, Akt1 87 , is delayed until day 12 while the other alleles are delayed until day 14 ( Figure 2). The extended time of development of homozygotes may be required for the production of the adult mutants.
The eyes of novel Akt1 mutants are reduced in ommatidia size and number Biometric analysis of homozygous Akt1 mutant eyes indicates there is an overall decrease in both ommatidia number and size when compared to controls (Table 1, Figure 3). The control eyes had a number of 676.4 ± 13.2 ommatidia per eye (OPE) and an ommatidia area of 222.6 (±4.3) um 2 , which was the largest overall. The original Pelement insertion mutant Akt1 04226 has the smallest   The largest of the mutants is Akt1 57 with an ommatidia area of 197.6 ± 3.5 um 2 and an ommatidia number of 579.4 ± 11 OPE.
Transgenic replacement partially rescue the phenotype of mutant homozygotes Ubiquitous expression of wild-type Akt1 in the background of the homozygous mutants results in a partial rescue of both ommatidia size and number (Table 1, , ON = Ommatidia Number, P1 = P-value when compared to Akt1 + control, P2 = P-value when compared to transgenic rescue counterpart, P3 = P-value when compared to somatic clone counterpart, S = significant, NS = not significant. Figure 3). The transgenic control Akt1 + is the largest eye overall having an ommatidia area of 218.9 ± 2.7 um 2 and a total number of 705.8 ± 17.6 OPE. The transgenic expression of Akt1 + in the background of the original P-element insertion mutant results in eyes that are only slightly smaller than the control having an ommatidia area of 204.7 ± 1.8 um 2 and a total of 683.2 ± 3.9 OPE. In all cases the size of the ommatidia and the total count of ommatidia for the mutants with transgenic replacement of wild-type Akt1 + does not differ significantly from the    Table 1. Green bars represent the homozygotes; purple bars represent the clones and blue bars represent the transgenic rescues. The original P-element insertion mutant Akt1 04226 is smaller than the control but larger than the novel mutants in both size and number. All three novel Akt1 mutants are reduced in both the area and number of ommatidia as both homozygotes or clones but reach almost comparable measurements compared to the control when rescued by a transgene. Error bars represent standard error of the mean (p = <0.05). Further statistical analysis can be found in Table 1.
control ( Figure 3). The average ommatidia area is very similar for Akt1 52 and Akt1 57 being 206.3 ± 3.6 um 2 and 206.2 ± 3.2 um 2 respectively; while the area for Akt1 87 is larger at 212.9 ± 2.5 um 2 . The ommatidia number for the partially rescued mutants is similar for Akt1 57 and Akt1 87 being 644.6 ± 13.5 OPE and 643.6 ± 7.1 OPE respectively, with Akt1 52 having a few more ommatidia at 701 ± 12.6 OPE.
Somatic clones of the eye have a more severe phenotype than the homozygotes Given the developmental delay of the Akt1 homozygotes, analysis of somatic clones of the eye was carried out. The FLP recombinase was driven by the eyeless promoter to direct expression in the developing eye tissue.
In the presence of FLP, homologous chromosomes undergo mitotic recombination between the FRT sites located on chromosome pairs. Heterozygous parent cells can produce both homozygous Akt1 mutant cells containing two copies of the mutant allele, and cells containing two copies of Akt1 + . In this system, the Akt1 + daughter cells are lost due to the presence of a linked recessive cell lethal mutation located on the same arm of the chromosome bearing the Akt1 + allele. Thus, in the eyes of clone bearing flies, the surviving cells bear two copies of the Akt1 mutant allele under investigation. The clone of the control is the largest and most consistent in size when compared to its homozygous and transgenic rescued counterpart having an ommatidia area of 212.1 ± 7.1 um 2 and number of 704.3 ± 5.5 OPE. The cloned original P-element insertion mutant Akt1 04226 is reduced in both size and number compared to its original homozygous version with an average area of 179.3 ± 8.3 um 2 and ommatidia number of 508.3 ± 35.4 OPE, yet is comparable in size and number to two of the cloned novel mutants, Akt1 57 and Akt1 87 , which have an ommatidia area of 175.7 ± 9.6 um 2 and 173.8 ± 7.9 um 2 and a count of 483 ± 24.5 OPE and 501 ± 33.2 OPE respectively. The measurement and count of ommatidia for both of these mutants is significantly smaller than that of their homozygous versions. Of all the mutants, Akt1 52 exhibits the most severe phenotype with the greatest decrease in ommatidia area (148.7 ± 3.9 um 2 ) and number (261 ± 14.3 OPE) when compared to both its homozygous counterpart as well as with the other cloned mutants.

Akt1-foxo double mutant lines reveal an epistatic relationship
Drosophila lines having both the novel Akt1 mutant alleles in combination with a null foxo mutant allele resemble the original foxo mutant more closely than the Akt1 mutants ( Figure 4;

Discussion
Viable novel Akt1 hypomorphs were generated via imprecise P-element excision and were found to retain internally deleted versions of the original PZ P-element upstream of the Akt1 gene's protein coding region ( Figure 1). Three selected Akt1 hypomorphs were characterized phenotypically as small in size and delayed in terms of developmental time. In Drosophila, the development from egg to adult involves three larval stages plus pupation before the non-growing sexually mature adult fly arises. The timing of transition between these stages is dependent upon the rate of growth. The insulin receptor signalling pathway is a major contributor in the control of growth and has been implicated in the control of the onset of metamorphosis in Drosophila [28]. Ablation of insulin producing cells within the larval brain decreases the growth rate and delays metamorphosis in Drosophila, as does a loss-of-function mutation of the insulin receptor [29,30]. As Akt1 is a central component of the insulin receptor signalling pathway, it is not surprising that these novel hypomorphic alleles result in a delay of development and overall smaller adult organisms. Due to the extension in the time for the novel mutants to undergo eclosion, a comparison of the phenotypes of the eye for both homozygous mutants and somatic clones was undertaken. The clone eyes are comprised of homozygous mutant tissue in a heterozygous organism that develops within a relatively normal timeframe. Biometric analysis of the eyes of these mutant clones revealed enhanced severity of the decreased growth phenotype. Adult organisms, as well as their organs and tissues, have a tendency to develop within a range of normal overall size, such that the cellular composition may vary from a large number of small cells, to a small  Table 2. Green bars represent the controls, purple bars represent the novel Akt1 mutants and the null foxo mutant, blue bars represent the double mutant homozygotes. In analysis of both the ommatidia size and number, the double mutants have a larger number of smaller ommatidia in comparison to the original Akt1 mutants, but are comparable in both size and number to the null foxo mutant. Error bars represent standard error of the mean (p = <0.05). Further statistical analysis can be found in Table 2. number of large cells. Cell growth includes an increase in cell number and cell size, and while not mutually exclusive, both can be regulated by distinct extracellular processes [11,[31][32][33], including the insulin receptor signalling pathway, which is highly conserved between invertebrates and mammals [34]. Reduced expression or loss of Akt1, the central component of insulin receptor signalling, can result in the production of smaller animals or, if severe, lethality [35,16,15]. The smaller eyes observed in the homozygotes is expected with the lower expression of Akt1 in these novel mutants. Compensatory growth is widespread and occurs in the surviving cells of damaged tissues to generate final structures of near normal overall size [36,37]. This growth consists of remodeling the existing tissue to regenerate the full body plan in response to tissue damage leading to the development of a smaller but still complete and intact organism. In order to maintain tissue homeostasis, cells that survive the tissue damage can compensate for those that are lost by increasing their rate of proliferation and cell divisions. Cells in Drosophila that have experienced an increase in cell death via radiation showed an increase in proliferation by the surviving cells [38]. Compensatory proliferation has been shown to lead to the development of normal-sized adult wings even when 40-60% of cells in the wing disc of Drosophila are either killed or rendered incapable of further proliferation [39]. These novel Akt1 hypomorphs have been shown to be developmentally delayed and result in the formation of small adult flies. The mutant clone eyes show a more severe phenotype due to the reduced replacement of missing tissue without the extended time during development. Clearly, the extended period of time required for these mutants to develop allows compensatory proliferation to generate smaller but intact adults.
In order to begin to understand the mechanisms responsible for the observed compensatory growth, double mutant lines of the novel Akt1 hypomorphs and an amorphic allele of foxo, a gene encoding a key downstream target of Akt1, were generated. The transcription factor foxo is known as a major effector of insulin receptor signalling and has been implicated in the control of cell growth. Overexpression of the mammalian homologues of foxo, as well as Drosophila foxo, leads to growth arrest [7,40] which can be suppressed with increased insulin receptor signalling. This suppression is ineffective when the foxo transcription factor has been made incapable of phosphorylation, and thus nuclear exclusion, by Akt1. In addition to this, foxo governs the expression of target genes that encode factors that regulate cell growth such as the eukaryotic initiation factor 4Ebinding protein (4E-BP) gene and cell cycle regulators including p27 kip1 . The 4E-BP product is a negative regulator of protein synthesis and has been shown to strongly influence the regulation of cell growth [41]. When foxo is upregulated, so is 4E-BP, which binds to the messenger RNA 5′ cap-binding protein eIF4E to inhibit protein synthesis and cell growth. In humans, p27 kip1 inhibits cyclindependent kinases (cdks) [42], which aid in promoting the transitions between cell-cycle phases. Overexpression of p27 kip1 in human cells leads to cell-cycle arrest in the G1 phase, and when foxo and, subsequently, p27 kip1 , is upregulated [41,43]. Co-expression of foxo and constitutively active Ras2, which can induce G1/S progression and cell proliferation, is able to partially rescue the phenotype in the eye that is observed with an overexpression of foxo alone [7]. An increase in foxo activity appears to result in a decrease in cell proliferation. The double mutants, having both hypomorphic alleles of Akt1 and null alleles of foxo, more closely resemble the foxo mutants. Analysis of ommatidia number shows an epistatic effect whereas an argument could be made in the comparison of ommatidia area for a slight synergistic enhancement of the phenotype. Regardless this suggests that without the presence of the foxo gene product, the hypomorphic alleles of Akt1 do not cause the same reduction of growth and strongly suggests that foxo is necessary for the processes that lead to compensatory growth.

Conclusions
Through the generation of clones, we were able to uncover a more severe effect of these Akt1 hypomorphs upon the control of growth. Originally, the generation of somatic clones was utilized to study homozygous tissue in a heterozygous organism when the homozygotes themselves were not viable. However, when used to study homozygous tissues of viable, yet small and developmentally delayed organisms, this system can expose a subtle phenotype previously obscured by compensatory proliferation. Developmental delay is a common phenomenon associated with many genetic mutations and could potentially play a significant role in the final phenotype. Generation of somatic clones would eliminate this developmental timing factor, thereby clarifying the impact a genetic mutation has on cellular processes including growth.

Animal ethics
This study was conducted under the approval of the Animal Care Committee of Memorial University of Newfoundland as a Category of Invasiveness Level A protocol under the project title of "Genetic, biochemical and molecular analysis of cell survival and cell death in Drosophila melanogaster" (protocol number: 14-09-BS).