SEARCH

SEARCH BY CITATION

Keywords:

  • Drosophila;
  • pRb;
  • PP1;
  • cell cycle

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

G1 Cyclin/Cdk complexes phosphorylate and inactivate the pRb tumor suppressor by preventing its ability to bind and repress E2F transcription factors. Current molecular and biochemical evidence suggests that type 1 protein phosphatases (PP1) dephosphorylate and thereby activate pRb, but the functional significance of this has not been addressed in the context of animal development. Here, we use genetic analyses to determine the role of PP1 in the regulation of Rbf1 activity during Drosophila development. While Rbf1 is required for E2f1 inhibition and G1 arrest in the embryonic epidermis and for the periodic expression of E2f1 target genes during endocycle S phase in the embryonic midgut and larval salivary gland, PP1 is not. PP1 regulates periodic cyclin E protein accumulation in ovarian nurse cells independently of Rbf1, which is dispensable for endocycle regulation in this tissue. We conclude that PP1 is not a major regulator of the Rbf1/E2F1 pathway in Drosophila. Developmental Dynamics 236:2567–2577, 2007. © 2007 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

As one of the earliest tumor suppressors described, pRb has a well-established role in cell-cycle control during development. pRb inhibits the activity of the E2F transcription factor, which regulates many genes that contribute to cell-cycle progression (DeGregori,2002; Trimarchi and Lees,2002; Dimova and Dyson,2005). pRb activity with respect to E2F is controlled by phosphorylation. Hypo-phosphorylated pRb binds E2F and represses gene expression. Upon phosphorylation by G1 Cyclin/Cyclin-dependent kinase (Cyc/Cdk) complexes, including Cyclin D/Cdk4 (CycD/Cdk4) and Cyclin E/Cdk2 (CycE/Cdk2), pRb dissociates from E2F, which then activates the transcription of target genes that contribute to S phase entry.

While large efforts have been made to characterize the kinases responsible for pRb phosphorylation, less is known about the phosphatases that contribute to pRb dephosphorylation. In the current view, various Cyc/Cdk complexes maintain pRb in a hyper-phosphorylated state throughout much of the cell cycle, and hypo-phosphorylated pRb accumulates during mitosis beginning in anaphase (Tamrakar et al.,2000). There is some evidence that type 1 protein phosphatases (PP1) directly mediate pRb dephosphorylation (Alberts et al.,1993; Ludlow et al.,1993; Nelson et al.,1997; Nelson and Ludlow,1997; Yan and Mumby,1999). pRb associates with PP1 both in vitro and in vivo (Durfee et al.,1993; Nelson et al.,1997; Puntoni and Villa-Moruzzi,1997), and this association occurs through the C-terminal region of pRb (Tamrakar and Ludlow,2000). Interestingly, PP1 interacts with both hyper-phosphorylated and hypo-phosphorylated pRb (Nelson and Ludlow,1997; Tamrakar et al.,1999). CycD/Cdk4 and CycE/Cdk2 phosphorylate pRb at different residues, and some experiments suggest that the dephosphorylation of these residues is site-specific and temporally-regulated (Rubin et al.,2001). Although there are multiple PP1 isoforms in mammalian cells (Ceulemans and Bollen,2004), PP1δ appears to be the most active toward pRb (Nelson et al.,1997; Rubin et al.,2001).

Although the relationship between PP1 and pRb has been extensively characterized at the molecular level in mammalian cell culture, it has not been analyzed in the context of animal development where genetic means can assess the contribution of PP1 activity to pRb function and E2F target gene expression in vivo. To address this, we used phenotypic analysis of Drosophila PP1 mutations to explore the role of PP1 in pRb regulation during development. As in mammals, the pRb homolog in Drosophila (Rbf1) that regulates cell cycle–coupled gene expression is controlled by phosphorylation (Du et al.,1996). Rbf1 negatively regulates the E2f1 transcription factor, which is required for replication factor gene expression (Duronio et al.,1995; Royzman et al.,1997; Du and Dyson,1999). The Rbf1/E2f1 network functions throughout development, including during G1 cell-cycle exit (Du and Dyson,1999; Shibutani et al.,2007), the proliferation of diploid imaginal cells (Xin et al.,2002), and the regulation of the non-canonical endocycle, in which cells become polyploid through repeated rounds of S phase that are not interrupted by mitosis (Duronio et al.,1998; Royzman et al.,2002; Weng et al.,2003). Drosophila thus provides a unique opportunity to analyze the contribution of PP1 to Rbf1/E2f1 control in various situations.

PP1 is an evolutionarily well-conserved protein that has been implicated in a number of processes including glycogen metabolism, mitosis, muscle contraction, and RNA processing (Cohen,2002). It exists as a holoenzyme composed of a regulatory subunit and a catalytic subunit (Ceulemans and Bollen,2004). In Drosophila, there are two subtypes of PP1 catalytic subunits: PP1α, which is homologous to mammalian PP1α and PP1γ, and PP1β, which is homologous to mammalian PP1β/δ (Dombradi et al.,1993). While there are three genes that encode PP1α-isozymes (PP1α13C, PP1α87B, and PP1α96A), there is only one gene that encodes for PP1β (PP1β9C or flw). Although the protein sequences of PP1α and PP1β are very similar, each subtype appears to have a unique function in the fly (Dombradi et al.,1993). PP1β9C specifically regulates non-muscle myosin (Vereshchagina et al.,2004), although PP1α96A may participate in this process (Kirchner et al.,2007), while PP1α87B has been implicated in mitotic progression and acts as a suppressor of position-effect variegation (Axton et al.,1990; Dombradi et al.,1990; Baksa et al.,1993).

Rbf1 activity is highly regulated during Drosophila development. In the early embryo, Rbf1 is expressed but kept inactive presumably by hyper-phosphorylation via constitutively active CycE/Cdk2, and this results in the ubiquitous transcription of E2f1-target genes (Shibutani et al.,2007). Later in embryogenesis, when epidermal cells exit the cell cycle and arrest in G1, Rbf1 is required to repress E2f1-target gene expression (Du and Dyson,1999; Shibutani et al.,2007). The mechanisms by which Rbf1 is converted to an active inhibitor of E2f1 as cells initially exit the cell cycle remain incompletely understood, but presumably involve a conversion of Rbf1 from a hyper- to a hypo-phosphorylated form (Shibutani et al.,2007). Rbf1 also functions to regulate the endocycle: chronic E2f1 activity in the absence of Rbf1 activity reduces the ploidy of many larval tissues (Weng et al.,2003), likely because periodic expression of E2f1 targets like Cyclin E is necessary for re-activation of replication origins during each endocycle S phase (Follette et al.,1998; Weiss et al.,1998; Lilly and Duronio,2005).

By analyzing PP1 catalytic subunit mutants, we asked whether PP1 activity modulates Rbf1 activity during cell-cycle exit and endocycle progression in vivo. Our data indicate that the expression of Rbf1/E2f1-regulated genes is unchanged in PP1-mutant embryonic cells that have exited the cell cycle in G1 and in endocycling cells of the embryonic midgut and the larval salivary gland. However, we have identified a role for PP1 in the regulation of the endocycle in ovarian nurse cells.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

PP1 Catalytic Subunits Are Expressed in the Same Embryonic Cells as Rbf1

We began our analysis by examining PP1 catalytic subunit expression during embryogenesis. If PP1 regulates Rbf1, then it should be expressed in the same cells as Rbf1. We examined PP1α13C, PP1α87B, and PP1α96A expression during embryogenesis by in situ hybridization. mRNA for each PP1 catalytic subunit is provided maternally (Fig. 1A,E,I), and is ubiquitously expressed in germ band–extended embryos (Fig. 1B,F,J). Following germ band retraction, all three PP1 catalytic subunits are expressed in the G1-arrested epidermis (Fig. 1C,G,K, arrows), the proliferating CNS (Fig. 1D,H,L, double arrows), and the endocycling midgut (Fig. 1D,H,L, arrowheads), all tissues where Rbf1 is also expressed (Stevaux et al.,2002; Keller et al.,2005). PP1α87B is the most highly expressed of these three catalytic subunits, and this is consistent with it comprising approximately 80% of total PP1 activity in Drosophila (Dombradi et al.,1990). PP1α96A expression is less robust while PP1α13C expression is barely detectable, similar to previous observations (Dombradi et al.,1993). Unlike the PP1α isoforms, PP1β9C has a restricted pattern of expression. PP1β9C mRNA is provided maternally, but is not expressed in germ band–extended embryos and zygotic expression is limited to the musculature in germ band–retracted embryos (BDGP, Gene Expression Patterns). These expression data are consistent with the possibility that PP1α13C, PP1α87B, and PP1α96A could regulate Rbf1 in the embryo.

thumbnail image

Figure 1. Embryonic expression patterns of Drosophila PP1 catalytic subunits. Whole mount in situ hybridization of w1118 embryos with PP1α13C (A–D), PP1α87B (E–H), and PP1α96A (I–L) probes. Note expression in the epidermis (arrows), central nervous system (double arrows), and the midgut (arrowheads).

Download figure to PowerPoint

PP1 Is Not Required for Rbf1-Dependent Repression of E2f1 Target Genes in G1-Arrested Epidermal Cells

Drosophila embryos execute a stereotyped, developmentally controlled cell-cycle program that is well characterized (Lee and Orr-Weaver,2003). The earliest cell cycles lack G1 phase, with S phase beginning immediately after mitosis. The first G1 appears at the beginning of germ band retraction, after which many cells remain arrested in G1 (e.g., cells of the epidermis) while others (e.g., midgut) re-enter S phase and begin endocycles. Rbf1 has an established role in E2f1 repression and the maintenance of G1 arrest in the embryonic epidermis (Du and Dyson,1999). This can be visualized in whole mount preparations of embryos that are both labeled with BrdU to detect S phase nuclei and hybridized with a probe for RnrS, a well-characterized E2f1-target gene encoding the small subunit of ribonucleotide reductase (Duronio et al.,1995; Shibutani et al.,2007). In wild type germ band–retracted embryos, G1-arrested epidermal cells do not incorporate BrdU and do not express RnrS (Fig. 2A, asterisk). In contrast, Rbf1-mutant embryos (lacking maternal and zygotic gene function) inappropriately express RnrS throughout the epidermis, indicative of ectopic E2f1 activity, and contain cells that have re-entered the cell cycle as measured by ectopic BrdU incorporation (Fig. 2B, asterisk).

thumbnail image

Figure 2. PP1 does not regulate Rbf1 activity in the embryonic epidermis. A–D: Stage 14 embryos pulse-labeled for 15 min with BrdU (cyan) were hybridized with an RnrS probe (red). Asterisks mark the epidermis. Scale bars = 50 μm. A:w1118. The arrow marks wild type RnrS expression in the malphigian tubules. B:Rbf114 maternal and zygotic null embryo. Note the ectopic RnrS expression and BrdU incorporation in the epidermis. C:PP1α87B1/PP1α87B87Bg-3-PCNA EmGFP maternal and zygotic null embryo. D:PP1α87B87Bg-3-PP1α96A2/PP1α87B87Bg-3-PP1α96A2. The arrow and arrowhead in C and D denote wild type RnrS expression in the malpighian tubules and trachea, respectively. E: Stage 14 UAS HA-NIPP1Dm/da-Gal4 embryo pulse-labeled for 15 min with BrdU (cyan) and hybridized with an RnrS probe (red) and stained for HA-NIPP1 (green). Note the HA-NIPP1 expression throughout the epidermis. The arrow and arrowhead denote wild type RnrS expression in the malphigian tubules and trachea, respectively. Scale bar = 50 μm. The embryos in A–E are oriented dorsal-laterally with the plane of focus on the epidermis. Consequently, the replicating cells of the central nervous system are not visible (but see Fig. 3). F: Schematic of interactions among G1-S regulators in Drosophila.

Download figure to PowerPoint

thumbnail image

Figure 3. Rbf1 and Dap regulate the embryonic midgut endocycle independently of PP1. A,B: Stage-14 embryos pulse-labeled for 15 min with BrdU (cyan) and hybridized with an RnrS probe (red). A:w1118. B:Rbf114 maternal and zygotic null embryo. C–E: In situ hybridization of stage 13 embryos with an RnrS probe. Arrows denote the central midgut. C:w1118. D: UAS Rbf1/midgut-Gal4. E: UAS Rbf-280/midgut-Gal4. F–I: Stage 14 embryos pulse-labeled for 15 min with BrdU (cyan), hybridized with an RnrS probe (red), and stained with either anti-GFP (not shown), anti-β-galactosidase (not shown), or anti-HA antibodies (green). F:PP1α87B1/PP1α87B87Bg-3-PCNA EmGFP maternal and zygotic null embryo. GFP staining confirmed the PP1α87B87Bg-3-PCNA EmGFP chromosome. G:PP1α87B87Bg-3-PP1α96A2/PP1α87B87Bg-3-PP1α96A2. The absence of GFP staining from the balancer chromosome indicated PP1α87B-PP1α96A double mutant embryos. H: UAS HA-NIPP1Dm/midgut-Gal4. HA staining confirmed HA-NIPP1 expression in the midgut. The double arrow denotes the anterior and posterior midgut while the asterisk (RnrS panel) marks yolk autofluorescence. I:dap4454/dap4454. The absence of β-galactosidase staining from the balancer chromosome indicated dap mutant embryos. In A, B, F, G, and I, the arrow indicates the central midgut and the arrowheads indicate the anterior and posterior midgut. Note that in these embryos, expression of RnrS and incorporation of BrdU are unaffected in the central nervous system (small arrow). Scale bar = 50 μm.

Download figure to PowerPoint

Although Rbf1 transcript and protein are present throughout early embryogenesis (Stevaux et al.,2002; Keller et al.,2005), Rbf1 does not act to repress E2f1-target gene expression in the epidermis until cells begin to enter G1 arrest during germ band retraction. We recently provided genetic evidence that the appearance of Rbf1 repressor activity involves a conversion of Rbf1 from a hyper- to a hypo-phosphorylated form (Shibutani et al.,2007). To determine if PP1 is involved in this conversion, we analyzed E2f1-target gene expression and BrdU incorporation in PP1-mutant embryos. We hypothesized that if PP1 was required for Rbf1 dephosphorylation (and thus activation as an E2f1 repressor), then PP1-mutant embryos would phenocopy Rbf1-mutant embryos and display ectopic RnrS expression and BrdU incorporation in the epidermis. Initially, we analyzed embryos that were both maternally and zygotically-mutant for PP1α87B, the major PP1 catalytic subunit in Drosophila (Dombradi et al.,1990). These embryos were generated with the hypomorphic PP1α87B1 allele because we were unable to obtain eggs from mothers with a PP1α87B null mutant germ line. PP1α87B1 is a G-to-A point mutation that results in the conversion of Gly220 to Ser (Dombradi and Cohen,1992). Although the reduction in PP1α87B activity in PP1α87B1 is similar to the reduction seen in null mutants, this allele was designated a hypomorph because a transcript is made and there are fewer aberrant mitotic figures seen in larval neuroblasts compared to null mutants (Axton et al.,1990; Baksa et al.,1993). Unlike Rbf1 mutants, there was no inappropriate expression of RnrS or ectopic BrdU incorporation in the epidermis of PP1α87B1-mutant embryos (Fig. 2C, asterisk). The same result was obtained by using null mutant alleles to remove the zygotic contribution of PP1α87B, PP1α13C, or PP1α96A (data not shown).

The high degree of sequence identity among these three catalytic subunits (Dombradi et al.,1993) and their similar embryonic expression patterns (Fig. 1) suggest that they could function redundantly in Rbf1 control. To address this, we analyzed RnrS expression and BrdU incorporation in embryos multiply mutant for zygotic gene function. There was no inappropriate expression of RnrS and no ectopic BrdU incorporation in the epidermis of either PP1α87B-PP1α96A double mutant embryos (Fig. 2D, asterisk) or PP1α13C; PP1α87B-PP1α96A triple mutant embryos (see Experimental Procedures section). These data suggest that the removal of the zygotic contribution of PP1α13C, PP1α87B, and PP1α96A does not affect Rbf1 activity in the embryonic epidermis. To confirm this result, we over-expressed NIPP1 in the epidermis. NIPP1 is an inhibitor of PP1 that when over-expressed results in phenotypes similar to PP1 mutants (Parker et al.,2002; Bennett et al.,2003). UAS HA-NIPP1Dm was expressed throughout the epidermis with the strong, ubiquitous da-Gal4 driver. In spite of the strong expression of HA-NIPP1 (Fig. 2E, asterisk), we could not detect ectopic RnrS expression or BrdU incorporation. Although we cannot be certain that we have completely eliminated PP1 function in the embryo, these results suggest that Rbf1 is capable of maintaining G1 arrest, through the inhibition of E2f1-target gene expression, independently of type 1 protein phosphatase activity.

Another potential mechanism by which Rbf1 activation could occur is through the inhibition of the Rbf1 kinase CycE/Cdk2, which is constitutively active early in embryogenesis (Sauer et al.,1995; Xin et al.,2002). We previously showed that in some cells entering G1 arrest Rbf1 becomes activated by the developmentally controlled expression of p27Dap, a specific inhibitor of CycE/Cdk2 (de Nooij et al.,1996; Lane et al.,1996; Shibutani et al.,2007). To test the possibility that p27Dap and PP1 cooperate in the appearance of active Rbf1, we analyzed RnrS expression and BrdU incorporation in germ band–retracted, p27Dap; PP1α87B double zygotic mutant embryos. Similar to germ band–retracted p27Dap single mutants, there was no inappropriate expression of RnrS in the epidermis and no ectopic BrdU incorporation (data not shown and de Nooij et al.,1996; Lane et al.,1996) suggesting that these pathways do not act redundantly.

Rbf1 Regulates Embryonic Midgut Endocycles

Although we could not detect a role for PP1 during G1 cell cycle exit in the embryonic epidermis, PP1 could regulate Rbf1 in cycling cells. We therefore chose to examine an endocycling cell population, because periodic S phase–coupled expression of E2f1-target genes contributes to normal endocycle progression (Duronio et al.,1998). Following G1 arrest, cells of the midgut re-enter S phase and begin endocycles, where gap phase and S phase alternate (Lee and Orr-Weaver,2003). Separate regions of the embryonic midgut differentially enter and exit S phase, generating a characteristic pattern of coincident BrdU incorporation and RnrS expression. For example, during stage 14 the central midgut cells are in S phase and, therefore, incorporate BrdU and express RnrS (Fig. 3A, arrow), while the anterior and posterior midgut cells are in a gap phase and do not incorporate BrdU nor express RnrS (Fig. 3A, arrowheads).

Previous experiments suggested that Rbf1 is necessary for normal midgut endocycle regulation (Du and Dyson,1999). To examine this, we analyzed RnrS expression and BrdU incorporation in stage 14 embryos lacking maternal and zygotic Rbf1. In a Rbf1-mutant embryo, both BrdU incorporation and RnrS expression occur throughout the midgut rather than just in the central midgut as seen in a wild type embryo (compare Fig. 3B to 3A). The increase in BrdU-positive cells in the anterior and posterior midgut (arrowheads in Fig. 3B) at this stage suggests that Rbf1 mutant cells spend less time in the endocycle gap phase (i.e., the gap phases are shorter). It is difficult to precisely determine whether the cells of the anterior and posterior midgut re-enter S phase precociously or whether they fail to exit the previous S phase on schedule. Regardless, these results suggest that the control of E2f1 activity by Rbf1 is required for the proper regulation of the endocycle in the midgut.

To test whether Rbf1 activity was controlled by phosphorylation in the midgut, we utilized a mutant version of Rbf1 (Rbf-280) containing mutations in four Cdk consensus sites that cannot be inhibited by the activity of G1 Cyc/Cdk complexes (e.g., CycE/Cdk2) (Xin et al.,2002). Driving UAS Rbf-280 expression with a midgut-specific driver resulted in the termination of RnrS expression in the central midgut (compare Fig. 3E to 3C, arrows). Little change was seen after the over-expression of wild type Rbf1 (Fig. 3D, arrow), suggesting that the termination of RnrS was specific to UAS Rbf-280. These results suggest that Rbf-280 can bypass the normal mechanism of Rbf1 control in the embryo, and are consistent with Rbf1 activity being regulated by phosphorylation in the endocycling midgut.

Type 1 Protein Phosphatases Do Not Regulate Rbf1 Activity in the Endocycling Midgut

To test if PP1 modulates Rbf1 activity in the midgut, we analyzed RnrS expression and BrdU incorporation in PP1 mutants. We hypothesized that if PP1 was involved in the dephosphorylation of Rbf1 and thus activation of Rbf1 repressor activity in the midgut, then PP1-mutant embryos would have the same phenotype as Rbf1-mutant embryos. However, unlike Rbf1 mutants and similar to stage 14 wild type embryos, RnrS expression and BrdU incorporation were seen only in the central midgut of maternal and zygotic PP1α87B1 mutants (Fig. 3F, arrow), zygotic PP1α87B-PP1α96A double mutant embryos (Fig. 3G, arrow), and zygotic PP1α13C; PP1α87B-PP1α96A triple mutant embryos (see Experimental Procedures). In an attempt to remove all PP1 function in the midgut, we drove the expression of UASHA-NIPP1Dm with a midgut-specific driver and found that there was no inappropriate expression of RnrS or ectopic BrdU incorporation in the anterior and posterior midgut (Fig. 3H, double arrow). Therefore, similar to what we observed during G1 cell cycle exit in the epidermis, these data suggest that Rbf1 activity is regulated properly in endocycling embryonic midgut cells in the absence of PP1 function.

p27Dap Regulates Embryonic Midgut Endocycles

The oscillation of CycE/Cdk2 activity is required for repeated rounds of endocycle S phase in Drosophila (Lilly and Duronio,2005). The induction of CycE/Cdk2 activity is required to trigger DNA synthesis during S phase (Knoblich et al.,1994) while the termination of CycE/Cdk2 activity is needed for the assembly of new replication origins during the endocycle gap phase (Sauer et al.,1995; Weiss et al.,1998). CycE/Cdk2 activity oscillation is accomplished in part by the periodic expression of p27Dap, which is necessary for normal endocycle regulation (de Nooij et al.,2000; Hong et al.,2003,2007).

To test if p27Dap modulates Rbf1 activity in a cycling population of cells, we analyzed RnrS expression and BrdU incorporation in the endocycling midgut of p27Dap-mutant embryos. Similar to what was observed in a Rbf1-mutant embryo, there was both ectopic RnrS expression and BrdU incorporation throughout the midgut rather than just in the central midgut as in wild type embryos (compare Fig. 3I to 3A). This result is consistent with a model whereby p27Dap contributes to the control of E2f1 activity in endocycling cells via regulating the state of Rbf1 phosphorylation. p27Dap also plays a more direct role in controlling replication origin licensing in the endocycle, especially in the ovary (Hong et al.,2003,2007).

To test if p27Dap and PP1 cooperate in Rbf1 regulation in the endocycling midgut, we analyzed RnrS expression and BrdU incorporation in the midgut of germ band–retracted, p27Dap; PP1α87B double zygotic mutant embryos. There was ectopic RnrS expression and BrdU incorporation throughout the midgut rather than just in the central midgut as seen in wild type embryos (data not shown and Fig. 3A). The phenotype of these double mutant embryos was similar to, but not detectably worse than, p27Dap single mutants (Fig. 3I), suggesting that p27Dap and PP1 do not cooperate in controlling Rbf1 activity in this tissue.

Type 1 Protein Phosphatases Do Not Regulate Rbf1 Activity in Larval Salivary Glands

One interpretation of our failure to detect a replication or RnrS expression phenotype in the embryo is that our genetic manipulations did not sufficiently reduce PP1 activity because of an abundant maternal load of PP1 catalytic subunits. We therefore analyzed larval stages of development when aberrant phenotypes arise in PP1 mutants (Axton et al.,1990; Baksa et al.,1993). We chose to examine E2f1-target gene expression in endocycles of the larval salivary gland, because Rbf1 activity is required in this tissue and because PP1 has an essential role during mitosis that precludes a careful analysis of E2f1 activity in diploid cells (Axton et al.,1990). Without Rbf1, the entire salivary gland is smaller than wild type and the oscillation of Cyclin E protein (and therefore CycE/Cdk2 activity) is lost (Weng et al.,2003).

To test if PP1 regulates Rbf1 activity during the larval endocycle, we analyzed Cyclin E protein expression in PP1α87B-mutant salivary glands. Cyclin E is a known target of E2f1 in endocycling salivary gland cells (Weng et al.,2003), and Cyclin E protein accumulation oscillates, being high during S phase and low during gap phase (Follette et al.,1998; Weiss et al.,1998). We also used a PCNA-EmGFP reporter construct to monitor E2f1 activity. This transgenic reporter accurately depicts E2f1-target gene expression in embryonic and larval cell cycles (Fig. 4). The asynchrony of the endocycles in the salivary gland results in patterned Cyclin E protein and PCNA-EmGFP expression. In a wild type salivary gland, cells in gap phase contain no Cyclin E protein or PCNA-EmGFP expression (Fig. 5A, large arrow). Cells with high levels of both Cyclin E and PCNA-EmGFP expression (Fig. 5A, small arrow) are likely in early S phase, while cells with high Cyclin E but low PCNA-EmGFP expression likely represent late S phase (Fig. 5A, arrowhead). These patterns of Cyclin E protein and PCNA-EmGFP expression were observed in PP1α87B-mutant salivary glands (Fig. 5B and C, large and small arrows). There was no significant difference in the numbers of Cyclin E (P = 0.8) or PCNA-EmGFP (P = 0.8) -positive nuclei between wild type and PP1-mutant salivary glands. This suggests that E2f1 was properly regulated during the salivary gland endocycle in this mutant. We also analyzed Cyclin E protein expression after over-expressing UAS HA-NIPP1Dm using a salivary gland–specific driver. This treatment results in small salivary glands with disrupted morphology, suggesting that PP1 activity is necessary for normal salivary gland development. Nonetheless, Cyclin E protein accumulation still oscillated, suggesting that E2f1 activity was regulated (Fig. 5D). Taken together, these results suggest that Rbf1 regulation of E2f1 in the larval salivary gland is not dependent on PP1.

thumbnail image

Figure 4. PCNA-EmGFP transgene expression in the embryo and larva reports E2f1 activity. A–I: PCNA-EmGFP embryos stained with anti-GFP antibodies (green). A: Syncytial embryo showing the maternal contribution of PCNA-EmGFP. B: Stage 5, blastoderm. C: Stage 6. D: Stage 9. E: Stage 10. Asterisk marks epidermal expression. F: Stage 11. Expression declines in the epidermis (asterisk) as these cells arrest in G1. G: Stage 13. PCNA-EmGFP is expressed in the mitotically dividing brain lobes (BL) and ventral nerve cord (VNC), and in the endocycling hindgut (HG) and anterior and posterior midguts (AMG and PMG, respectively). H: Stage 14. PCNA-EmGFP is expressed in the VNC, as well as the endocycling central midgut (CMG), malphigian tubules (MT), and anal pad (AP). I: Stage 14. PCNA-EmGFP is expressed in the VNC but not the G1-arrested epidermis (asterisk). J:PCNA-EmGFP eye and antennal discs from third instar larva stained with anti-GFP antibodies (green). PCNA-EmGFP is expressed in the asynchronously dividing cells in the anterior region of the eye disc (arrow) and in the second mitotic wave (arrowhead). The lack of expression between these two regions denotes the G1-arrested cells in the morphogenetic furrow. K:PCNA-EmGFP early third instar larval salivary gland stained with anti-GFP antibodies (green) and DAPI (blue). PCNA-EmGFP is expressed in a subset of cells (arrow) because the endocycles are asynchronous. Scale bars = 50 μm.

Download figure to PowerPoint

thumbnail image

Figure 5. PP1 does not regulate Rbf1 activity in the larval salivary gland. A–D: Salivary glands from early third instar larvae. A–C: Salivary glands were stained with anti-Cyclin E (red), anti-GFP (green), and DAPI (blue). Scale bars = 50 μm. A:PCNA EmGFP. B:PP1α87B1-PCNA EmGFP/PP1α87B87Bg-3. C:PP1α87B87Bg-6-PCNA EmGFP/Df (3R) Exel 6164. The large arrows (A–C) mark cells with low levels of both Cyclin E and PCNA-EmGFP. The small arrows (A–C) denote cells with high levels of both Cyclin E and PCNA-EmGFP while the arrowhead (A) identifies a cell with high levels of Cyclin E but low levels of PCNA-EmGFP. D: UAS HA-NIPP1Dm/AB1-Gal4 salivary glands were stained with anti-Cyclin E (green) and DAPI (blue). Scale bar = 20 μm.

Download figure to PowerPoint

Type 1 Protein Phosphatases Regulate Cyclin E Protein Oscillations in Ovarian Nurse Cell Nuclei

The Drosophila ovary serves as another tissue in which to study the possible contribution of PP1 to gene expression during the endocycle. Each ovary consists of approximately 16 ovarioles, egg chamber assembly lines that contain all developmental stages of oogenesis. At the most anterior region of each ovariole is the germarium, a specialized structure that contains both germ line (GSCs) and somatic stem cells (SSCs). GSCs divide asymmetrically to produce a daughter stem cell and a cystoblast (Lin and Spradling,1997). The cystoblast undergoes four synchronous mitotic divisions with incomplete cytokinesis to produce a 16-cell cyst. Before exiting the germarium, the germ line cyst is encapsulated by somatic follicle cells to create an egg chamber. One of the 16 cyst cells will adopt the oocyte fate while the remaining 15 will adopt a nurse cell fate. The nurse cells become highly polyploid through multiple endocycles to support the growth of the oocyte (Swanhart et al.,2005).

Like the endocyling cells of the salivary gland, Cyclin E protein oscillations are critical for the regulation of the endocycle in the nurse cells (Lilly and Duronio,2005). To assess the role of both Rbf1 and PP1 in nurse cell endocycle control, we analyzed Cyclin E protein expression in Rbf1 and PP1α87B-mutant egg chambers. Representative pictures of wild type and PP1α87B-mutant egg chambers are shown in Figure 6A and B, respectively. Because the nurse cell endocycles are asynchronous, Cyclin E accumulates in only a subset of the nurse cells in any one egg chamber (arrows). We hypothesized that if E2f1-regulated transcription was a major contributor to the oscillation of Cyclin E protein accumulation, then in the absence of Rbf1-mediated repression we would observe more Cyclin E–positive nurse cells than in wild type. In addition, we would observe a similar phenomenon in PP1-mutant nurse cells if PP1 activity was necessary to generate hypo-phosphorylated Rbf1. We quantified the number of Cyclin E–positive nurse cell nuclei in stage 8 wild type, Rbf1-mutant, and PP1α87B-mutant egg chambers and the results are depicted in Figure 6C. The average number of Cyclin E–positive nurse cell nuclei was similar in wild type and Rbf1-mutant egg chambers (P < 0.101). This indicates that Cyclin E expression is not dependent on the Rbf1/E2f1 pathway in this tissue. However, there was significantly more Cyclin E–positive nurse cell nuclei in PP1α87B mutants compared to both wild type (P < 0.00005) and Rbf1-mutant (P < 0.006) egg chambers, suggesting that Cyclin E protein oscillations are disrupted in PP1α87B-mutant egg chambers. These data are consistent with a Rbf1-independent role for PP1 in the regulation of the endocycle in the ovarian nurse cells.

thumbnail image

Figure 6. PP1α87B regulates Cyclin E protein oscillations in nurse cell endocycles. A,B: Stage-8 egg chambers were stained with anti-Cyclin E (red), phalloidin (green), and DAPI (blue). Phalloidin, which stains f-actin, was used to outline the nurse cells and DAPI was used to mark nurse cell nuclei. A:w1118. The arrow denotes Cyclin E staining in a nurse cell. Cyclin E also accumulates to high levels in the oocyte nucleus (arrowhead). B:PP1α87B1 germ line clone. Scale bars = 50 μm. C: Quantification of the number of Cyclin E–positive nurse cell nuclei in stage 8 egg chambers of w1118, Rbf114 germ line clones and PP1α87B1 germ line clones.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

pRb phosphorylation has an established role in controlling the activity of E2F transcription factors during the cell cycle, particularly at the G1-S transition, and this mechanism of cell-cycle regulation is well conserved (Cobrinik,2005). In this work, we used genetic analyses and the well-described cell-cycle programs of Drosophila development to ask whether type 1 protein phosphatases contribute to Rbf1/E2f1 regulation in several tissues. We hypothesized that PP1 may be necessary for Rbf1 repressor activity from mammalian data suggesting that PP1 is required for the dephosphorylation of pRb during mitosis (Ludlow et al.,1993; Nelson et al.,1997; Nelson and Ludlow,1997; Yan and Mumby,1999; Rubin et al.,2001). Using the expression of E2f1 targets as a readout, we found no evidence that PP1 is necessary for Rbf1 activity during G1 arrest in the embryonic epidermis and during the embryonic midgut, larval salivary gland, and ovarian nurse cell endocycles. Since we cannot be absolutely certain that PP1 function was completely eliminated by our genetic manipulations, it is possible that a small amount of PP1 activity could dephosphorylate and thus activate Rbf1. We were also unable to detect a physical interaction between PP1α87B and Rbf1 by co-immunoprecipiation experiments with embryonic extracts (data not shown). We conclude that either another phosphatase is involved in the dephosphorylation of Rbf1, or that it acts redundantly with PP1. Along with PP1, there is some evidence in mammalian cell lines that PP2A regulates the phosphorylation status of pRb (Garriga et al.,2004).

Drosophila contain one other pRb and one other E2F family member, called Rbf2 and E2f2, that function as part of a complex (MMB/dREAM) that controls the expression (primarily via repression) of a large number of genes important for developmental processes independent of the cell cycle (Stevaux et al.,2002; Dimova et al.,2003; Korenjak et al.,2004; Lewis et al.,2004; Stevaux et al.,2005). Rbf1 appears to act redundantly with Rbf2 in this complex. Although some data suggest that phosphorylation by G1 cyclin-dependent kinases is insufficient to reverse E2f2-mediated repression (Frolov et al.,2003), it is possible that PP1 could regulate the activity of Rbf2 or Rbf1 when bound to E2f2, and thus modulate the expression of MMB/dREAM target genes. This would provide a cell cycle–independent link between PP1 and pRb family members, which was not addressed by our experiments.

Our data show that Rbf1 is necessary for proper regulation of E2f1-target gene expression and S phase in the endocycle of the embryonic midgut. In a Rbf1 mutant, cells of the anterior and posterior midgut inappropriately replicate DNA and express RnrS at a time when these cells should be in a gap phase (Du and Dyson,1999). These results suggest that Rbf1 is either necessary for exit from the endocycle S phase or required to maintain the endocycle gap phase. We observed a similar phenotype in p27Dap mutants. Because p27Dap protein oscillates during the endocycle (de Nooij et al.,2000), we suggest that any cycling between hyper- and hypo-phosphorylated forms of Rbf1 (i.e., between S phase and gap phase) may be controlled primarily by the presence or absence of this CycE/Cdk2 inhibitor, rather than regulation of a phosphatase. Perturbing the oscillation of p27Dap protein disrupts the normal oscillation of CycE/Cdk2 activity in the nurse cells of the ovary (Hong et al.,2007).

Interestingly, PP1 may be required for the proper regulation of the endocycle in nurse cells. We found more Cyclin E–positive nurse cell nuclei in stage 8 PP1α87B-mutant egg chambers compared to both wild type and Rbf1-mutant egg chambers, suggesting that the oscillation of Cyclin E protein may be disrupted in this mutant. There is yeast two hybrid data from Drosophila suggesting that PP1α87B physically interacts with Cdk2 (Stanyon et al.,2004), and this interaction has also been described for one of the PP1 isoforms in mammalian cell culture (Flores-Delgado et al.,2007). CycE/Cdk2 complexes may play an important role in the turnover of Cyclin E protein during the endocycle (Lilly and Spradling,1996). Through its interaction with Cdk2, PP1 could be directly involved in regulating Cyclin E protein levels during the nurse cell endocycle. Similar to E2f1 and Dp mutants (Royzman et al.,2002), Cyclin E protein oscillations were not disrupted in Rbf1-mutant egg chambers, suggesting that the contribution of E2f1-target gene expression to the regulation of the nurse cell endocycle is minimal. This differs from endocycle control in the embryo where E2f1-target gene expression is required for the proper timing of the endocycle (Duronio et al.,1998), suggesting that endocycles may be modified by different regulatory pathways in a tissue-specific manner.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Drosophila Strains

w1, w1118, AB1-Gal4, actin-Gal4/TM6B, da-Gal4, midgut-Gal4, w ovoD FRT 14A-B/C(1)DX, y f/Y; hsFLP, hsFLP, y w; DrMio/TM3, w; FRT 82B, piM 87E, w; FRT 82B ovoD/ st β Tub85D ss e/TM3, In(1)wm4h; PP1α87B1 /TM3, w; Df(3R)Exel 6164/TM6B, w1118; Df(3R)Exel 7357/TM6B, and Df(1)ED7294, w1118/FM7j, B were obtained from the Bloomington Stock Center. dap4454/CyO has been described previously (Lane et al.,1996). UAS HA-PP1α87B/TM3, UAS HA-NIPP1Dm/TM6B, PP1α96A2e/TM6B, and PP1α87B87Bg-3-PP1α96A2e /TM6B were generously provided by Luke Alphey (Bennett et al.,2003; Vereshchagina et al.,2004; Kirchner et al.,2007). UAS Rbf-280/TM3, UAS Rbf1, and Rbf114 FRT14A-B/FM7 as well as w; PP1α87B87Bg-3/TM6B and w; PP1α87B87Bg-6e/TM6B were kindly provided by Wei Du and Adelaide Carpenter, respectively (Axton et al.,1990; Du and Dyson,1999; Xin et al.,2002). w1118; PCNA-EmGFP, dap4454/CyO wg-lacZ, w; PP1α87B87Bg-6-PCNA EmGFP e /TM3 kr-GFP, w; PP1α87B87Bg-3-PCNA-EmGFP/TM3 kr-GFP, PP1α87B1-PP1α96A2/TM3 twi-GFP, PP1α87B87 Bg-6-Df(3R)Exel 7357/TM6B, and hsFLP y w; PP1α87B1 FRT82B/TM3 were constructed for this study. Rbf114 and PP1α87B1 germ line clones were generated as described (Du and Dyson,1999). PP1α13C, PP1α87B, and PP1α96A triple mutants were constructed by crossing Df(1)ED7294/+; PP1α87B87 Bg-3-PP1α96A2/+ females with either PP1α87B87 Bg-3-PP1α96A2/TM3 twi-GFP or PP1α87B87 Bg-6-Df(3R)Exel 7357/TM6B males. In total, over 1,000 embryos from these parents were analyzed for RnrS expression and over 500 for BrdU incorporation.

PCNA EmGFP Transgene Construction

A modified version of the PCNA-GFP fusion construct reported in Thacker et al. (2003) was constructed by replacing one copy of GFP with three copies of Emerald GFP (pUAST:3X Emerald GFP plasmid kindly provided by Chris Doe). Transgenic animals were created by P-element-mediated germline transformation of a w1 strain.

RNA In Situ Hybridization and BrdU Labeling

Embryos were dechorionated, fixed in a 1:1 mixture of 4% formaldehyde in PBS/heptane for 25 min, and devitellinized with methanol. For BrdU labeling, dechorionated embryos were permeabilized with octane, pulse-labeled with 1 mg/ml BrdU for 15 min in Schneider's Drosophila medium prior to fixation. Embryos were stored in methanol at −20°C.

In situ hybridization with digoxigenin-labeled antisense RNA probes was performed as described (Kearney et al.,2004). Fluorescent detection of hybrids (FISH) was achieved with the TSA Fluorescence System (Perkin Elmer) using a 30-min incubation in TSA-Cy3 or TSA-Fluorescein. For all triple fluorescent staining (i.e., FISH, anti-protein, anti-BrdU), embryos were first processed for FISH, then for immuno-detection of proteins, and finally for BrdU detection by acid denaturation of chromosomes (Schubiger and Palka,1987).

Immunostaining

Ovaries were dissected in Schneider's Drosophila medium and fixed in a 1:4 mixture of 2% paraformaldehyde in PBS/heptane for 20 min. Larval salivary glands were dissected in PBS and fixed in 4% formaldehyde in PBS for 20 min. Both tissues were incubated with primary antibody overnight at 4°C. Primary antibodies were: mouse anti-BrdU monoclonal antibody (1:100, Becton Dickinson), rat anti-BrdU monoclonal antibody (1:100, Abcam), rabbit anti-β galactosidase (1:200, Chemicon), rabbit anti-GFP (1:2,000, Abcam), mouse anti-HA (1:50, Covance), and mouse anti-CycE (1:5, gift of Helena Richardson) (Richardson et al.,1995). Secondary antibodies were: goat anti-mouse Oregon Green (1:1,000, Molecular Probes), goat anti-mouse-Cy5 (1:500, Jackson), goat anti-mouse-Cy3 (1:500, Jackson), donkey anti-rat-Cy5 (1:500, Jackson), and goat anti-rabbit-Cy2 (1:500, Jackson). Filamentous actin was detected in ovaries by incubation with Oregon green-conjugated phalloidin (1:1,000 in PBS, Invitrogen) for 20 min, and DNA in both salivary glands and ovaries was labeled with DAPI (1:1,000, Pierce). A Student's t-test assuming unequal variances was used for statistical calculations of Cyclin E–positive nurse cell nuclei. An unpaired t-test with Welch correction was used to analyze Cyclin E– and PCNA-EmGFP-positive nuclei in salivary glands. Stained embryos, salivary glands, and ovaries were mounted with Fluoromount-G (Southern Biotech) and visualized with either a Nikon Eclipse E800 microscope (for DIC imaging in Fig. 1) or a Zeiss LSM 510 scanning confocal microscope.

Co-Immunoprecipitations and Western Blotting

Immunoprecipitations from 0–8 hr and 8–16 hr actin-Gal4 UAS HA-PP1α87B and w1118 embryos were performed as described (Peifer et al.,1993), and analyzed by SDS-PAGE (4–15% pre-cast gradient gel, Bio-Rad) and Western blotting with mouse anti-Rbf1 (DX-3, 1:4; Du et al.,1996) and mouse anti-HA (1:1,000, Covance). The secondary antibody was ECL™-Sheep anti-mouse HRP (1:5,000 for Rbf1 and 1:2,000 for HA) from Amersham Biosciences.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank Luke Alphey, Adelaide Carpenter, Chris Doe, Wei Du, and Helena Richardson for reagents, Corbin Jones for help with statistics, Shusaku Shibutani, Jan LaRocque, and Joe Kearney for a critical reading of the manuscript, Duronio lab members for helpful discussion, and Tony Perdue for assistance with confocal microscopy.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES