Genetic Systems to Investigate Regulation of Oncogenes and Tumour Suppressor Genes in Drosophila

Animal growth requires coordination of cell growth and cell cycle progression with developmental signaling. Loss of cell cycle control is extremely detrimental, with reduced cycles leading to impaired organ growth and excessive proliferation, potentially resulting in tissue overgrowth and driving tumour initiation. Due to the high level of conservation between the cell cycle machinery of Drosophila and humans, the appeal of the fly model continues to be the means with which we can use sophisticated genetics to provide novel insights into mammalian growth and cell cycle control. Over the last decade, there have been major additions to the genetic toolbox to study development in Drosophila. Here we discuss some of the approaches available to investigate the potent growth and cell cycle properties of the Drosophila counterparts of prominent cancer genes, with a focus on the c-Myc oncoprotein and the tumour suppressor protein FIR (Hfp in flies), which behaves as a transcriptional repressor of c-Myc.


Introduction
Tight regulation of developmental signaling is essential for controlling the cell growth and division required for proper formation of tissues and organs. Impaired cell cycle progression can be extremely detrimental; reduced cell cycles can lead to small, impaired organs and, conversely, uncontrolled cell proliferation can lead to tissue overgrowth and cancer formation. In Drosophila, cell growth and cell cycle progression are regulated by a number of key genes, which have been shown to control cell cycle in an analogous manner in all multicellular organisms. The mammalian c-Myc (referred to as Myc from here on) transcription factor and oncogene and its Drosophila orthologue, dMyc, are both key regulators of cell growth and division [1,2]. A collection of genetic experiments, transcriptome analyses and genome binding studies in mammals and Drosophila have revealed that Myc proteins can bind to the promoters and potentially control the transcription of 10% 15% of all genes [2 7] (reviewed in [8]). Thus the regulatory targets of Myc and dMyc include genes from virtually every biochemical and regulatory pathway in the cell, including growth, metabolism, cell cycle progression, differentiation and apoptosis (reviewed in [8 10]).

Cell Cycle Control is Fundamental to Development and is Disrupted in Cancer: The Myc oncogene and FIR Tumour Suppressor are Conserved between Drosophila and Mammals
In addition to regulating a wide range of genes, Myc targets can vary depending on the cell-type and developmental context, a complexity of function which has thwarted attempts to identify a universal transcriptome signature for Myc. In an effort to address these observations, two recent studies ( [11,12] and reviewed in [13]) provide a potential explanation for the cell and context dependent variations between transcriptional signatures associated with Myc. Genome-wide ChIP-Seq analysis suggests that rather than driving its own transcriptional program, Myc behaves as a general amplifier of the cells transcriptional state at the time of Myc activation. By monitoring co-occupancy of the elongating form of RNA polymerase II and the presence of active chromatin marks, both studies observed that after overexpression, Myc protein is loaded quantitatively onto active promoters to enhance transcription. The observation that Myc overexpression does not increase its enrichment within the promoters of silent genes, suggested to the authors that elevation of Myc is not sufficient to activate transcription [11,12].
These new models suggest that Myc behaves as an amplifier of existing cellular states, however, Myc activation has been largely linked with cell proliferation and elevated levels of Myc are sufficient to drive cell growth and cell cycle progression [1,8,14]. Moreover, the capacity of Myc to drive growth is critical to its oncogenic potential [15]. Indeed, many studies in mammals suggest that Myc is an instructive force, rather than a simple reinforcer of cell fate. In mammals, even physiological levels of Myc can drive quiescent cells to proliferate [8,14]. Extensive studies have shown that like mammalian Myc proteins, dMyc is not only essential but is also sufficient for accumulation of cellular mass or cell growth [1,2,16]. In Drosophila, dMyc regulates cell and organismal size; hypomorphic mutants of dMyc are small due to reduced cell growth [1], while null dMyc mutations lead to larval lethality due to a growth arrest [16]. Conversely, overexpression of dMyc produces larger flies [1,16]. A pertinent question then remains: how does Myc achieve this effect on growth?
Transcription of ribosomal DNA (rDNA) is required to produce functional ribosomes, which is one of the most fundamental rate-limiting steps for growth and DNA synthesis. Myc and dMyc are both key regulators of rDNA transcription, ribosome biogenesis and cell growth [1,2]. The Myc oncogene regulates cell growth via three RNA polymerases; RNA polymerase I, II and III [17]. Myc regulates a large number of RNA polymerase II-transcribed genes, many of which encode ribosomal proteins and translation factors [6]. Myc is also very efficient at activating transcription via RNA polymerase I (Pol I) [2,7] and III (Pol III) [18] to drive rDNA transcription, ribosome biogenesis and protein translation. Drosophila microrarrays revealed upregulation of Pol I transcripts in dMyc overexpressing cells, and an associated increase in rDNA transcription, ribosome biogenesis and cell growth [2]. The capacity to drive production of rRNA (ribosomal RNA) is central to mammalian Myc's powerful cell growth effects and oncogenic ability [15]. Small changes, either up or down, in Myc protein levels will modify growth and potentially result in cancer initiation and/or progression, emphasizing the requirement for extremely tight control of Myc expression [19].
In addition to the ability to drive growth in Drosophila and mammals, the Myc protein is also required to couple growth with cell cycle progression. Like Myc, early in the G1 phase of the cell cycle, dMyc activates the genes required for DNA replication and progression through S phase [1,2,4,16,20 22]. Myc and dMyc activate the G1 to S-phase Cyclin/Cyclin-dependent-kinase complexes, CycE/Cdk2 and CycD/Cdk4, in order to trigger DNA replication and S-phase progression [23 25] ( Figure 1A). As the loss of cell cycle control resulting from misregulation of Myc is associated with cancer [26,27], Myc must be tightly regulated during normal development [14,19,28,29]. The importance of maintaining a tight control of Myc levels is reflected by the multiple types of regulation observed for Myc and dMyc, including transcriptional initiation, RNA synthesis, translation and protein stability [8,30]. Therefore, the upstream signaling pathways and transcriptional, translational and proteolytic mechanisms that regulate Myc are of critical importance. In addition, the levels of Myc expression need to be responsive to growth and developmental signals to allow organ and tissue growth in response to nutrients and other external cues [1,20,31 33].
In mammals, one mechanism proposed for tight regulation of Myc levels and a quick response of Myc transcription to growth signals in vitro involves the presence of a paused, but transcriptionally engaged, RNA polymerase II (Pol II) within the Myc promoter [9,29,34,35]. Escape of Pol II from the promoter allows transcript elongation. Biochemical evidence has shown the movement of Pol II and elongation of the Myc transcript is dependent on the Transcription Factor IIH (TFIIH) complex and its DNA helicase subunit, XPB. Interactions with TFIIH/XPB helicase and two DNA structure-sensitive regulatory proteins called FUSE Binding Protein (FBP) and FBP Interacting Repressor (FIR) control the Pol II complex movement within the promoter of the Myc gene [35,36]. In this system, FBP and FIR act as dominant regulators of Myc: FBP is a potent activator of Myc, while FIR is required for repression of Myc transcription. FBP and FIR interact antagonistically and are essential for the transcriptional regulation of Myc transcription, and their actions are mediated by the XPB DNA helicase [9,29,34,35,37]. Consistent with a role for FIR in transcriptional repression of Myc, loss-of-function FIR mutations are associated with increased Myc mRNA levels in primary human colorectal cancer [38]. Mutations of XPB [35,36] and activation of FBP [39] have also been implicated in human cancer. Evidence suggests Drosophila Hfp is the functional homolog of FIR, being essential for repression of dMyc transcription and, furthermore, cell growth in vivo [31,40]. ChIP experiments have revealed enrichment of Hfp proximal to the dMyc transcriptional start site (TSS). In addition, loss of Hfp results in enhanced cell growth, which depends on the presence of dMyc. Furthermore, Hfp physically and genetically interacts with the XPB helicase component of the TFIIH transcription factor complex, Hay, which is required for normal levels of dMyc expression, cell growth and cell cycle progression. Below we review how the Drosophila model has been used to increase our understanding of how Myc expression is correctly patterned and modulated in multicellular organisms in vivo, particularly in response to the Hfp transcriptional repressor.

Larval Imaginal Discs: Models for Connecting Developmental Signals to Organ Growth
Drosophila has been used extensively as a model organism to understand the link between developmental signaling pathways and cell cycle progression [41,42]. A major benefit of this model organism is its short lifecycle. Within 10 days of egg deposition, the Drosophila larvae proceeds through the various stages (1st, 2nd, 3rd instar larvae and pupae) to eclosure as an adult fly. Each Drosophila larvae contains imaginal discs which are the precursors to many adult organs. Imaginal discs develop from invaginations of the embryonic epithelium into head structures (mouth parts, eyes and antenna), appendages (legs and wings) and genitalia. At the 1st instar stage, imaginal discs are bags of around 10 50 undifferentiated cells, which undergo massive growth and proliferation to comprise up to 100,000 cells by the end of the third larval instar. Differentiation starts at the end of the 3rd larval instar and is complete by the end of pupariation, when all adult structures such as the wings, legs and eyes have developed [43]. The 3rd larval instar is therefore a critical stage of Drosophila development, containing the major growth and proliferation of all tissues required to form the adult fly.
The wing imaginal disc gives rise to the adult wing blade, hinge and part of the thorax ( Figure 1B). Cell cycle patterning in the wing imaginal disc pouch has been particularly well characterized [1,41,44 46]. For instance, clear connections have been made between developmental signals and the cell cycle delay that occurs late in the 3rd larval instar along the D/V boundary, which is required for differentiation of the margin of the wing blade ( Figure 1B) [1,30,41,42,44,45]. A major signaling molecule required for wing disc morphogenesis is the Wingless (Wg) protein, the founding member of the Wnt family of secreted morphogens [47,48]. Wg is secreted across the D/V boundary of the wing pouch during the 3rd larval instar and is required for inhibition of cell cycle and formation of the zone of non-proliferative cells (ZNC) ( Figure 1B) [44,46]. This zone gives rise to the adult wing margin and is characterized by reduced S phases as measured by incorporation of bromodeoxyuridine (BrdU) ( Figure 1C). The posterior portion of the ZNC is composed entirely of G1 cells, but the anterior portion of the ZNC can be divided into three subdomains ( Figure 1B). The central domain, where cells express Wg, is comprised of G1 cells and is flanked by the dorsal and ventral subdomains, where proneural transcription factors Achaete (Ac) and Scute (Sc) are expressed to delay these cells in G2 [44]. Wg induces G2 arrest via Ac and Sc, which act to repress the mitotic factor Stg/Cdc25 in the dorsal and ventral domains of the anterior compartment to result in cell cycle exit and differentiation [44]. In the posterior and central domain of the anterior compartment, Wg inactivates dE2F and results in G1 arrest in preparation for differentiation into the vein and intervein components of the adult wing during the pupal stage [44] (Figure 1B).
In the cell division cycles of the wing imaginal disc pouch, DNA synthesis is coupled with cell division; cells grow in G1, initiate DNA replication and enter S phase, which is separated from mitosis by G2. In these cells G1 progression is stimulated by growth factors, which trigger cell growth and activate the G1-S cell cycle machinery. The inhibition of Rbf, a member of the Drosophila Retinoblastoma family [22,49,50], results in release of E2F1 from the inhibitory complex with Rbf, which permits upregulation of E2F1 dependent S phase genes [51]. Imaginal disc cells maintain their size via the tight coupling of cell growth and cell cycle progression. S phase and G2-M can be coupled as CycE and Stg, the rate limiting factors for S-phase and mitosis, respectively, can both be activated by the Drosophila orthologue of human E2F1 protein, dE2F1, which thereby coordinates progression from S-phase into mitosis to maintain organ size [52].

The Drosophila Salivary Gland: Models for Understanding Cell Growth Control
In the mitotically dividing imaginal tissues discussed above, cell growth and cell cycle progression are tightly coordinated to allow cells to maintain a consistent size, which can mask potential effects on cell growth. In contrast, the salivary gland and many other tissues in the Drosophila larvae undergo endoreplication, i.e., cell growth and DNA replication/S phase occur in the absence of cell division to result in large polyploid cells [53]. Indeed, the salivary gland has been used as a tissue to study the effect of dMyc on cell growth. In these endoreplicating cells, dMyc is both necessary and sufficient for rDNA synthesis and growth [2,16]. Salivary gland cells fail to grow and endoreplicate in dMyc mutants, while overexpression of dMyc dramatically increases cell growth (nucleolar size) and nuclear DNA content (DNA replication) [2,16].
In all eukaryotes, the clusters of rRNA genes are organized in the nucleolus, where rRNA transcription, processing and ribosome assembly occur [54]. Thus, nucleolar size is largely proportional to the level of rRNA transcription and provides an indirect measure of ribosomal gene transcription. As predicted, based on previous studies [2] overexpression of dMyc in the salivary gland results in increased nucleolar size, measured using an antibody to the nucleolar protein Fibrillarin in 3rd instar salivary glands ( Figure 1D). In addition, altered rDNA transcription can be measured directly in salivary glands by monitoring abundance of rRNA. As transcription of the rDNA repeat sequence by RNA Pol I begins within the 5' external transcribed spacer (ETS) region, fluorescent in situ hybridisation (FISH) using a DIG-labelled riboprobe for the 5'ETS can be used as a direct read out of rRNA abundance ( Figure 1E). Consistent with the increase in the size of the nucleolus detected with fibrillarin, dMyc overexpression results in more 5'ETS, as expected based on the ability of dMyc to increase ribosomal gene transcription [2].

The Drosophila hemopoietic System; a Model for Mammalian Blood Development and Disease: Key Oncogenic Signaling Pathways Drive Overproliferation of Hemocytes
Disruptions to normal hematopoietic function have been implicated in numerous forms of leukemia, anemia and other blood disorders. The first link between dysregulation of the Myc oncogene and human disease was in Burkitts lymphoma, where the Myc gene is translocated to the immunoglobulin heavy chain locus, which results in amplified Myc expression (reviewed in [55 57]). One of the many problems faced by medical researchers attempting to dissect the underlying cause of hemopoietic disease is the complex genetic control and functional redundancies acquired as a result of gene duplication and evolution in mammals. Morphologically, development of the Drosophila and mammalian blood lineages is considerably diverged. However, molecular genetic studies have revealed that many key hemopoietic signaling pathways and transcription factors are conserved between systems [ 58 63].
In Drosophila, the blood producing lymph gland (LG) is specified prior to the blastoderm stage of embryogenesis [64]. The LG initially consists of 2 lobes, known as the primary lobes, which first appear late in the embryonic stages. Expression of collier, the Drosophila orthologue of the vertebrate transcription factor Early B Cell Factor, is one of the earliest markers of LG precursors [65]. Collier expression becomes increasingly restricted until it is only expressed in the most posterior 2 3 cells of the approximately 20 25 cell LG, the future Posterior Signaling Center (PSC, see Figure 2). The LG persists as the larval hematopoietic organ until the onset of metamorphosis, at which point it disintegrates and releases its contents into the hemolymph. There is no evidence of a hematopoietic organ in the adult Drosophila. The fully mature 3rd instar larval LG (Figure 2) consists of a pair of primary lobes and 2 7 pairs of secondary lobes which develop during successive rounds of proliferation during the early 3rd instar. The primary lobe is divided into three subsections: the medulla (MZ, composed of prohemocytes or hematopoietic precursor cells), the cortex (CZ, comprising differentiating and mature cells) and the PSC, which is responsible for maintaining larval blood cell homeostasis [66 68]. Under normal conditions, two cell types are found in the LG: (1) plasmatocytes and (2) crystal cells [64]. Plasmatocytes make up the vast majority of the hemocyte population (~95%) and function to phagocytose bacterial or fungal infections along with removal of excess cells during development [69].
A number of key signals have been implicated in both the mammalian and Drosophila hematopoietic maintenance as well as lineage specification. Dysregulation of these pathways can lead to loss of hemocyte progenitors and/or differentiation blocks, and can result in LG overgrowth in Drosophila and malignancy in humans. Ras was amongst the first human oncogenes identified and belongs to the family of genes encoding GTP-binding proteins. This family of genes is implicated in regulating cell growth, proliferation, and differentiation and is dysregulated in approximately 30% of human cancers [70,71]. In Drosophila expression of the activated/oncogenic form of Ras (Ras V12 ) specifically in the hemocytes results in a 40-fold increase in hemocyte number [59]. The overproliferation can be suppressed by reducing Cdk2/cyclin E, suggesting the Ras-induced increase in hemocytes is a result of increased E2F activity [59]. These changes in cell proliferation may depend on Myc, as the Ras pathway signals through the MAP kinase pathway to both increase Myc/dMyc protein levels [72 74]. Consistent with this, in the Drosophila wing imaginal disc, Ras activity results in increased dMyc protein accumulation and accelerated cell cycle progression [75]. The ability to generate leukemia-like phenotypes with well-established oncogenes lends weight to the idea that Drosophila may be a useful model for human blood disorders.
Further to this, the pathways implicated in PSC signaling and stem cell behavior also regulate activity of the mammalian HSC niche. The TGF-HSC niche [76] and in Drosophila hematopoiesis the orthologous BMP protein, Dpp, is necessary for maintaining PSC size and, as a result, larval blood homeostasis [63]. Similarly, Wg, and the BMPF1 ortholog, Dlp, are modulated by Dpp signaling and are required for hematopoiesis (Figure 2A) [63]. Thus, many parallels exist between mammalian and Drosophila hematopoietic niches, and given the role of Myc in controlling development of the HSC lineage [77,78] and leukemia [55 57] it will be of interest to investigate potential roles for dMyc in Drosophila hematopoiesis in the future.

Genetic Tools for Manipulating Gene Expression in Drosophila
The UAS-GAL4 system, derived from the yeast S. cereviase, has provided an invaluable genetic tool for studying the manipulation of gene expression during Drosophila development [79]. With a diverse variety of GAL4 drivers, temporal and tissue specific overexpression of a given UAS-transgene can be achieved in a specific tissue of interest. For example, cell cycle biologists combined the UAS-GAL4 system with the FLP/FRT system to generate the Actin<CD2<GAL4 monitor ( Figure 3A [80,81]). Gene expression is controlled by heatshock induced expression of the Flip recombinase (Flp) [82]. Flp will recognise the FRT sites in the Act<CD2<GAL4 cassette results in: (1) CD2 protein expression in the neighboring control clone and (2) the Actin promoter driving GAL4 in the clone, which can be positively marked with co-expression of a fluorescent transgene e.g. UAS-GFP or RFP with the UAS-transgene of interest ( Figure 3A,B) [80,81,83]. This system is ideal for studying effects of manipulating gene expression on cell growth, proliferation, and signaling pathways in comparison to surrounding normal tissue, which is essential for understanding tumorigenesis. For gene knockdown, GAL4 drivers can be used for expression of a UAS-transgene for an inverted repeat targeting the gene of interest [84]. Extensive UAS-RNAi collections have been developed and made publicly available and, as these cover most (97%) of the Drosophila genome (e.g. Vienna Drosophila resource centre [85]), they have provided an invaluable resource for conducting non-biased genetic screens to elucidate novel signaling mechanisms and gene function ( Figure 3C,D). For -generated using a hairpin targeting Hfp, were used to demonstrate that loss of Hfp results is increased growth compared with the surrounding control cells [40]. Moreover, using the dmyc-lacZ enhancer trap described below, we have demonstrated that Hfp is required for repressing dmyc promoter activity (Figure 4, [40]). In mammalian systems, in vitro transcription assays (e.g., luciferase assays) are a standard method for monitoring effects of a particular factor on promoter/enhancer activity. However, if we are to draw connections between the signaling environment and changes to the expression of cell cycle genes, it is essential to investigate transcription from endogenous promoters in vivo. In Drosophila, promoter and enhancer trap activity can be measured using reporter elements inserted into endogenous promoters in order to determine factors capable of modulating transcription of a given gene. Furthermore, random P-element mutagenesis screens generated extensive enhancer trap collections, where each line contains the insertion of a visible reporter (e.g., LacZ or GFP) for monitoring gene activity within a particular region of the genome [86,87]. As these enhancer traps often land in regions of active chromatin, such as gene promoters, they can often be used to measure endogenous promoter or enhancer activity, as an indicator of transcriptional activation of the gene located downstream of the insertion [88].
For example, a random P-lacZ insertional mutagenesis screen for the X-chromosome resulted in many hits in the dMyc promoter [89], providing potential enhancer traps to identify changes in dMyc promoter activity. Analysis of these lacZ insertions revealed that insert [P{lacW}l(1)G0354 [89]], located just prior to the 5'UTR in the dMyc promoter, in a region that is responsive to wing enhancers of dMyc expression ( Figure 4A). By using the -gal antibody, this dMyc-lacZ line can, therefore, be used to monitor changes to dMyc promoter activity in wing imaginal discs [40,42]. In particular, UAS-hfp -dMyc-lacZ enhancer trap background, revealed that Hfp is required for repression of dMyc promoter activity ( Figure 4B and [40]). The enhancer trap appears to reflect changes in dMyc transcription, as mRNA abundance is significantly increased in Hfp loss of function imaginal tissues [40] and can provide an in vivo read out for potential enhancers and suppressors of dMyc transcription.

Conclusion
As detailed above, Myc is a potent mitogen, but despite the large number of transcriptional targets and the associated oncogenic potential of Myc dysregulation, how Myc is regulated at the transcriptional level is largely unknown. In Drosophila using the genetic systems described above, we have demonstrated that repression of dMyc promoter activity and cell growth requires Hfp function ( Figure 4B and [40]). Furthermore, the increased dMyc-lacZ reporter activity, compared to surrounding wild type cells ( Figure 4B) was associated with significantly increased dMyc mRNA levels [40]. Thus like its mammalian counterpart FIR, Hfp behaves as a tumor suppressor to repress dMyc, which suggests that the mechanism proposed for transcriptional repression of c-Myc by FIR is conserved in Drosophila. These data suggest that the loss-of-function FIR mutants described in colorectal cancer may be sufficient to increase Myc expression, which would be predicted to lead to cancer initiation and progression.