Chaperonin TRiC/CCT supports mitotic exit and entry into endocycle in Drosophila

Endocycle is a commonly observed cell cycle variant through which cells undergo repeated rounds of genome DNA replication without mitosis. Endocycling cells arise from mitotic cells through a switch of the cell cycle mode, called the mitotic-to-endocycle switch (MES), to initiate cell growth and terminal differentiation. However, the underlying regulatory mechanisms of MES remain unclear. Here we used the Drosophila steroidogenic organ, called the prothoracic gland (PG), to study regulatory mechanisms of MES, which is critical for the PG to upregulate biosynthesis of the steroid hormone ecdysone. We demonstrate that PG cells undergo MES through downregulation of mitotic cyclins, which is mediated by Fizzy-related (Fzr). Moreover, we performed a RNAi screen to further elucidate the regulatory mechanisms of MES, and identified the evolutionarily conserved chaperonin TCP-1 ring complex (TRiC) as a novel regulator of MES. Knockdown of TRiC subunits in the PG caused a prolonged mitotic period, probably due to impaired nuclear translocation of Fzr, which also caused loss of ecdysteroidogenic activity. These results indicate that TRiC supports proper MES and endocycle progression by regulating Fzr folding. We propose that TRiC-mediated protein quality control is a conserved mechanism supporting MES and endocycling, as well as subsequent terminal differentiation.


Author summary
Endocycle, a cell cycle variant consisting of DNA replication and Gap phases without a mitotic phase, is widespread in nature. Endocycling cells arise from proliferating cells through a switch of the cell cycle mode called mitotic-to-endocycle switching (MES). While the molecular mechanisms regulating progression of the mitotic cell cycle and endocycle have been well studied, the regulatory mechanism of MES and its impact on cell differentiation processes remains poorly understood. Here we used the Drosophila steroidogenic organ to uncover the regulatory factors of MES, and our genetic analyses identified the evolutionarily conserved chaperonin TRiC as a novel regulator of MES and subsequent steroidogenesis. Given that TRiC supports proper folding of numerous proteins including cell cycle regulators, TRiC-mediated protein quality control will be a Introduction A tightly controlled cell cycle is a fundamental system for survival in every organism. The bestknown cell cycle mode is mitotic cell cycle, which is achieved through a sequence of distinct phases including genome DNA synthesis (S), mitotic (M), and intervening Gap (G) phases. Moreover, endocycle, a cell cycle variant without M phase, is commonly observed in protozoa, plants, and animals [1,2]. Endocycling cells undergo repeated rounds of S and G phases without a M phase, which gives rise to polyploidy of genome DNA. Endocycle and polyploidy are closely associated with cell growth; in some cell types, progression of endocycling is required for their terminal differentiation [2][3][4]. Moreover, polyploid genomic DNA has been observed in approximately 37% of all human tumors [5], and several lines of evidence point to the importance of endocycle in tumor development and survival [6][7][8]. Thus, elucidation of the underlying mechanisms regulating initiation and progression of endocycle is a key step to understanding the role of endocycle in normal and pathological cellular processes. The important question in endocycle regulation is how the transition from cell division to endocycle is achieved. Endocycling cells arise from diploid cells through a switch of the cell cycle mode, called the mitotic-to-endocycle switch (MES) [1]. At the molecular level, MES is accomplished by downregulation of mitotic cyclin-dependent kinases (M-Cdks), which leads to the M phase being bypassed. M-Cdk is suppressed through degradation of its binding partners, called mitotic cyclins, including Cyclin A (CycA) and B (CycB). Mitotic cyclins are recognized by Fizzy-related [Fzr, a.k.a. CDH1 (CDC20 Homologue 1)], an activator of a multisubunit ubiquitin ligase anaphase promoting complex/cyclosome (APC/C), to be polyubiquitinated, and then degraded through the ubiquitin-proteasome pathway [1,9]. Fzr triggers MES to support the progression of numerous biological events including morphogenesis, growth, and tissue repair in insects, plants, and mammals [1,[10][11][12][13][14][15][16][17][18]. Fzr-mediated degradation of mitotic cyclins is commonly observed during MES [1,11,16,18], whereas identified upstream regulators of Fzr are diverse among species [1]. It therefore remains unclear whether there are evolutionary conserved regulatory mechanisms of Fzr expression. In addition, it is largely unknown how mitotic exit and endocycle progression are cooperatively regulated.
In this study, we focused on the Drosophila prothoracic gland (PG), an endocrine organ composed of polyploid endocycling cells, for dissecting the molecular mechanisms supporting MES and the progression of endocycle. The PG produces ecdysone, the primary insect steroid hormone that triggers initiation and progression of metamorphosis [19,20]. In Drosophila and other higher Diptera, the PG is part of the ring gland (RG), an endocrine organ complex that comprises the PG, corpus cardiacum (CC), and corpus allatum (CA) [19]. The PG expresses a set of ecdysone biosynthetic genes including Neverland (Nvd), Spookier (Spok), Shroud (Sro), Phantom (Phm), Disembodied (Dib), and Shadow (Sad), whose expression is upregulated before the initiation of metamorphosis to enhance the ecdysteroidogenic activity [21][22][23]. Ecdysone secreted from the PG is converted into its active form, 20-hydroxyecdysone (20E), in peripheral tissues [19]. In Drosophila, PG cells undergo repeated rounds of endocycling during the larval stage, and endocycle progression is essential for activating ecdysone biosynthesis in the PG [22]. After at least three rounds of endocycle (when the C value reaches 32), biosynthesis of ecdysone is initiated to induce the larval-to-pupal metamorphic transition [22]. When endocycle progression is impaired in the PG, ecdysone biosynthesis is not upregulated, and the larva cannot transit into the pupal stage [22]. Because this 'larval arrest' phenotype is readily recognized, PG-selective genetic analysis can be a potentially effective approach to screen novel regulators of MES and endocycle progression. Its relatively large cell size and low cell number (its width is about 200 μm at the late larval stage and the cell number is around 50) also make the PG an attractive model for investigating MES and endocycle progression.
Here we show that PG cells undergo MES through downregulation of mitotic cyclins mediated by Fzr. Fzr protein expression is increased during the MES period in the PG, and knockdown of fzr in the PG causes a block of MES and subsequent ecdysone biosynthesis owing to upregulation of mitotic cyclins. Furthermore, we performed a PG-selective RNAi screen to further elucidate the regulatory mechanism of MES, and identified TCP-1 ring complex (TRiC), a molecular chaperonin complex, as a novel MES regulator. TRiC-deficient PG cells showed a prolonged mitotic period, probably due to impaired nuclear translocation of Fzr, which also caused loss of ecdysteroidogenic activity. Our genetic study provides an important basis for understanding the regulatory mechanisms of endocycle initiation and progression.

PG cells undergo MES
In wild type strain Oregon R, the PG increases its cell number during the 1st instar larval stage [L1, from 0 to 24 hours after hatching (hAH)] [22]. During the 2nd (L2, from 24 to 48 hAH) and the 3rd (L3, from 48 to 96 hAH) instar larval stages, in contrast, the PG cell number does not increase while the DNA content continues to increase and the C value reaches 32-64C by the late L3 stage, 84-96 hAH (Fig 1A) [22]. This suggests that PG cells undergo cell division during L1, execute MES at around 24 hAH, and undergo 3-4 rounds of endocycle during the L2 and L3 stages ( Fig 1A). First we determined that the mean C value in the PG of Oregon R at 84 hAH was 58 (Fig 1B and 1C). In addition, the C values in two transgenic control lines at 96 hAH were 53 and 54, respectively (S1A and S1B Fig). These results indicate that PG cells undergo approximately four rounds of endocycling. Considering that these C values are less than 64, which is achieved by four complete rounds of endocycling, it is suggested that replication of genomic DNA in PG cells is incomplete, which is known as under-replication [1]. We next tested whether DNA replication was suppressed in the heterochromatic region, in which under-replication is commonly observed [1]. 5-Bromodeoxyuridine (BrdU), an analog of thymidine used as a marker of replication activity, was incorporated into the DNA dense-core heterochromatic region of PG cells at 84 hAH, but its frequency was less than the early and middle S-phases (S1C and S1D Fig). This suggests that the genomic DNA of PG cells exhibits under-replication but the replication in the under-replicated region is not severely impaired.
Next, we investigated the cell number, the DNA content, and expression of the mitotic marker pH3 (histone H3 phosphorylated at serine 10) in the PG of Oregon R during larval development. The PG cell number was increased during 12 to 18 hAH (Fig 1D and 1F), suggesting that mitotic cell cycle in the PG is active approximately during 12 to 18 hAH. In contrast, no significant increase in the PG cell number was observed at 24 hAH and thereafter ( Fig  1D and 1F), and the C value started to increase at 18 hAH (Fig 1D and 1G). These results suggest that some, but not all, PG cells start MES from 18 hAH. In contrast, pH3-positive PG cells were detectable at 18 hAH (Fig 1D and 1H), but we could not detect a statistically significant increase in the percentage of pH3-positive PG cells at 18 hAH ( Fig 1H). We also failed to detect pH3 expression in the PG at 12 hAH (Fig 1D and 1H), probably reflecting transient pH3 expression owing to rapid mitotic cycles. To clarify cell cycle phase in PG cells during 0 to 24 hAH, we used Fly Fluorescent Ubiquitin-based Cell Cycle Indicator (Fly-FUCCI) driven by  Fig 1C). Scale bars: 10 μm. (C) Scatter plots with a mean value of relative signal intensity of Hoechst in Oregon R sperm, PG, and CA cells. The mean C value in sperm cells was PG-selective phantom-22-Gal4 (phm-Gal4) [24,25]. The Fly-FUCCI components GFP-fused E2F1 1-230 (GFP.E2F1) and mRFP1-fused CycB 1-266 (mRFP1.CycB) are expressed in patterns consistent with the presence of G1, S, and G2/M cells [24] (S1E and S1F Fig). The percentage of mRFP.CycB-positive/GFP.E2F1-negative cells (= S phase) reached the maximum level at 6 hAH (S1G and S1H Fig), indicating that S-phase of mitotic cell cycle is active at around 6 hAH. In contrast, the percentage of both mRFP.CycB and GFP.E2F1-positive cells (= G2/M phases) was increased at 12 and 18 hAH (S1G and S1H Fig), confirming that mitosis is active between 12 and 18 hAH. Actually, a small number of mRFP.CycB and GFP.E2F1-positive cells showed a metaphase-like nuclear shape at 12 and 18 hAH (arrows in S1G Fig). Furthermore, mRFP.CycB-positive PG cells were dramatically reduced at 24 hAH, and the majority of PG cells was mRFP.CycBnegative/GFP.E2F1-positive at 24 hAH (S1G and S1H Fig), indicating that mitotic cell cycle is downregulated by 24 hAH. Taken together, these observations indicate that mitotic cell cycle is active during a narrow time window, approximately between 12 and 18 hAH, and that PG cells execute MES by 24 hAH. Moreover, Fzr expression, which was visualized by GFP-fused Fzr (Fzr-GFP), was upregulated at 18 hAH and reached the maximum level at 24 hAH (Fig 1E and 1I), supporting the idea that Fzr-mediated MES in the PG starts at around 18 hAH and that PG cells execute MES by 24 hAH. In addition, Fzr-GFP was detectable at lower levels during later stages (Fig 1E and 1I), suggesting that Fzr blocks mitosis continuously during endocycling in the PG.
In contrast to the PG, pH3 expression was not observed in the CA (Fig 1D; CA cells are indicated by the arrowheads), another endocrine organ in the RG complex. This suggests that CA cells do not undergo cell division during the larval stage. However, the mean C value in CA cells reached around 8 at 84 hAH ( Fig 1B and 1C), indicating that CA cells undergo a round of endocycle. Actually the DNA content in CA cells seems to be increased during 24 to 84 hAH (arrowheads in Fig 1D), suggesting that CA cells undergo endocycle during the L2 and L3 stages. To test this possibility, expression of S-phase activator Cyclin E (CycE) and

Fzr-mediated downregulation of mitotic cyclins is required for MES and ecdysone biosynthesis in the PG
To investigate whether Fzr regulates MES in the PG, PG-selective phm-Gal4 was used to overexpress RNAi construct against fzr in the PG. The nuclei of PG cells were labeled by mCherry normalized to 1, and accordingly, the mean C values in PG and CA cells at 84 hAH were set to 58 and 8, respectively. Sample sizes (the number of sperm, PG, and CA cells) are shown above each column. (D) The PG (upper panels) and PG cells in higher magnification (lower panels) of Oregon R at indicated time points. PG cells, mitotic pH3-positive cells, and DNA were detected by anti-Sro antibody (green), anti-pH3 antibody (magenta), and Hoechst (blue), respectively. The arrows and arrowheads indicate pH3-positive PG cells and the nuclei of CA cells, respectively. Scale bars: 50 μm (upper panels) and 10 μm (lower panels). (E) Fzr-GFP expression in the PG at indicated time points. Fzr-GFP (green and white in the upper and lower panels, respectively) was detected using anti-GFP antibody at indicated stages. PG cells are labeled by mCherry.nls driven by PG-selective phm-Gal4 (magenta in the upper panels). The PGs are indicated by dotted lines. The asterisks indicate CC cells expressing DsRed driven by eyeless promoter: 3xP3-dsRed marker in Fzr-GFP transgene. Scale bars: 10 μm. (F-H) Scatter and box plots showing the cell number (F), the C value (G), and the percentage of pH3-positive cells (H) in the PG of Oregon R at indicated time points. Different lowercase letters indicate statistically significant differences (P < 0.05; Steel-Dwass test; see S1 Table). ns, not significant (P > 0.05). Box plots indicate the median (bold line), the 25th and 75th percentiles (box edges), and the range (whiskers). Dot plots show all data points individually. The mean C value at 84 hAH was normalized to 58 (see Fig 1C), and all data points were normalized accordingly. Sample sizes (the number of PGs) are shown above each column. (I) Scatter and box plots showing the relative expression level of Fzr-GFP in the PG at indicated stages. Different lowercase letters indicate statistically significant differences (P < 0.05; Steel-Dwass test; see S1 Table). Box and dot plots as in Fig  1C- carrying the nuclear localization signal (mCherry.nls), and RNAi efficiency was enhanced by overexpression of dicer2. In accordance with the result of ploidy measurement, the mean C value of the controls (phm>dicer2 mCherry.nls) was set to 54 (see S1A and S1B Fig). In the control animals, the PG cell number was slightly increased only during 0 to 12 hAH (Fig 2A and 2B), whereas the C value in the PG started to increase from 24 hAH and continuously increased until 96 hAH (Fig 2A and 2C). In contrast, the PG of fzr RNAi animals (phm>dicer2 mCherry.nls fzr-RNAi) continuously increased its cell number while its C value was around 4C during larval development (Fig 2A-2C). Consistent with these observations, pH3 was continuously detected in the PG of the fzr RNAi animals throughout larval development (Fig 2D and 2E). In addition, we confirmed efficient knockdown of Fzr, visualized by Fzr-GFP, along with upregulation of pH3 in the PG of fzr RNAi at 24, 48, 72, and 96 hAH (S3 Fig). Furthermore, hypomorphic fzr G0418 mutant clone PG cells, introduced by FLP/FRT recombination induced by heat-shock just after hatching (Fig 2F  and 2G), showed a reduced DNA level and, in some mutant clones, metaphase-like DNA distribution ( Fig 2H). Taken together, these results indicate that fzr is required for MES in the PG. Considering that mitotic cell cycle was not arrested in fzr RNAi, we suggest that fzr is unnecessary for progression of mitotic cell cycle, which has been described in a previous study [11].
We also confirmed that, as shown in a previous study [22], knockdown of fzr in the PG caused L3 arrest (S4A-S4C Fig These results indicate that fzr-mediated MES in the PG is required for activation of ecdysone biosynthesis and pupariation. However, fzr RNAi larvae did not show any defects in the L1-L2 and L2-L3 molting (S4A and S4B Fig), which are also triggered by ecdysone. This confirms that Fzr-mediated MES in the PG is required for pupariation but not for the L1-L2 and L2-L3 molting.
Next, we investigated whether Fzr triggers MES through downregulation of mitotic cyclins in the PG. In the controls (phm>dicer2 mCD8::GFP), neither CycA nor B expression was observed in the PG at post-MES stages, including 24, 48, 72, and 96 hAH (Fig 3A-3D, S5 Fig), whereas CycA was upregulated in the PG of fzr RNAi animals (phm>dicer2 mCD8::GFP fzr-RNAi) at the same stages ( Fig 3A-3D, S5 Fig). CycB was also detectable in the PG of fzr RNAi at 48, 72, and 96 hAH (Fig 3A-3D, S5 Fig). These results suggest that fzr-deficient PG cells cannot undergo MES, and ecdysone biosynthesis is inhibited owing to highly expressed mitotic cyclins. To test this possibility, we examined whether knockdown of cycA and B rescues impaired MES and ecdysteroidogenesis in fzr RNAi animals. In contrast to fzr RNAi animals (phm>dicer2 mCherry.nls fzr RNAi) showing a cell number increase and reduced DNA content in the PG (Fig 3E-3G), the C value reached 32-64 and the cell number was restored to around 50 at 96 hAH in the PG of both fzr RNAi + cycA RNAi (phm>dicer2 mCherry.nls fzr-RNAi cycA-RNAi) and fzr RNAi + cycB RNAi animals (phm>dicer2 mCherry.nls fzr-RNAi cycB-RNAi) (Fig 3E-3G). Consistent with this, the pH3-positive PG cell number was reduced in both fzr RNAi + cycA RNAi and fzr RNAi + cycB RNAi animals ( Fig 3H). These observations indicate that fzr induces MES through inactivation of mitotic cyclins in the PG. Moreover, knockdown of cycA or B in the PG of fzr RNAi animals rescued the developmental arrest at L3 (Fig 3I and 3J) and restored the expression level of ecdysone biosynthetic genes and the ecdysteroid concentration (Fig 3K and 3L). Taken together, these results indicate that the Fzr-mediated reduction of the mitotic cyclin protein level induces MES and subsequent ecdysone biosynthesis in the PG.

PG-selective RNAi screen to identify novel MES regulators
To further dissect the regulatory mechanisms of MES, we performed a genetic screen using the Gal4/UAS system and RNAi. A previous study carried out a genome-wide PG-selective RNAi

Fig 2. Fzr is required for MES in the PG. (A)
The PG of the controls (phm>dicer2 mCherry.nls) (upper panels) and fzr RNAi (phm>dicer2 mCherry.nls fzr-RNAi) (lower panels) at indicated time points. pH3 and DNA were detected by anti-pH3 antibody (green) and Hoechst (blue), respectively, and the nuclei of PG cells were labelled by mCherry.nls (magenta). The PGs are indicated by dotted lines. The asterisks indicate nonspecific anti-pH3 antibody signal around CC cells. Scale bars: screen and identified 701 genes whose knockdown cause L3 or L1/L2/L3 arrest (630 and 71 genes, respectively) [26]. We therefore focused on these genes as novel MES regulator candidates and knocked them down in the PG to observe their potential effects on developmental transitions as well as the cell number and the DNA content in the PG (Fig 4A). In our screen, females carrying two copies of phm>mCherry.nls were crossed with UAS-RNAi males to drive a dsRNA or shRNA construct selectively in the PG of their offspring. In addition, RNAi lines not used in a previous genome-wide RNAi screen were used whenever available to exclude potential off-target effects (see S3 Table). Because knockdown of target of rapamycin (tor) and β3-octopamine receptor (Octβ3R) in the PG causes an L3 arrest phenotype [22,27], these two genes were used as positive controls ( Fig 4A). In this screen, we used standard cornmeal/yeast Drosophila culture medium, whereas all other experiments in this study were performed using nutrient-rich German Food (GF).
In the first step analyzing the developmental phenotype (indicated as 'Step 1' in Fig 4A), a larval arrest phenotype was confirmed in 442 genes; more specifically, L1/L2, L1/L2/L3 and L3 arrest were observed in 77 (11%), 18 (3%), and 347 genes (49%), respectively ( Fig 4B). Other phenotypes, including delayed pupariation (7 genes) and embryonic lethality (1 gene), were observed in 8 genes (1%). In contrast, 253 genes (36%) showed no obvious phenotype (NOP; Fig 4B), which were excluded from further analysis. Next, we observed PG cells of 449 RNAi lines that showed either larval arrest or delayed pupariation (indicated as 'Step 2' in Fig 4A). The schematic diagram in Fig 4C shows the developmental change of the cell number and the DNA content in PG cells. Normal PG cells undergo mitotic cell cycle during the early larval stage (i.e. L1), then undergo MES and several rounds of endocycling to increase the DNA content during the L2 and L3 stages ( Fig 4C). By contrast, as in the case of fzr RNAi, animals with MES-deficient PG are expected to be arrested at the larval stage with the PG showing an increased cell number and reduced DNA content ( Fig 4C). In addition, it is expected that a defect in endocycle progression results in a reduced DNA content, whereas blocking the downstream pathway of endocycle does not cause a severe reduction in the DNA content ( Fig  4C). Based on these criteria, the cell number and the DNA content were examined in the PG of each RNAi line using histochemistry at day 6 after crossing ( Fig 4A). Fig 4D shows the mean value of the DNA intensity and the cell number in the PG of 449 RNAi lines, as well as the controls (phm>mCherry.nls) (green plot in Fig 4D, indicated as "Control" in the legend), raised on standard Drosophila medium. We also observed the PG of control animals cultured in nutrient-rich GF at 0, 24, 48, 72, and 96 hAH as a references (black plots in Fig 4D, indicated as "Control cultured on GF" in the legend). Of the 449 genes tested in Step 2, knockdown of 210 genes caused morphological or physiological defects in PG cells, including an abnormal distribution of DNA and nucleolus, apoptotic nuclear condensation, and a nontransparent cytoplasm (S3 Table, Table (Steel-Dwass test). Box and dot plots as in Fig 1C-1E. The mean C value of control at 96 hAH was set to 54 (see S1B Fig). Sample sizes (the number of PGs) are shown above each column. (D) PG cells in higher magnification of the controls (upper panels) and fzr RNAi animals (lower panels) at indicated stages. The arrows indicate pH3-positive PG cells. Scale bars: 10 μm. (E) Scatter and box plots showing the percentage of pH3-positive PG cells of the controls (red) and fzr RNAi animals (blue) at indicated stages. P-values for all pairwise comparisons are shown in S1 Table (Steel-Dwass test). Sample sizes are the same as in B and C. (F) Schematic diagram of FLP/FRT system. Heat-shock-induced FLP recombinase recognizes FRT sites to induce trans-chromosome recombination between chromosome arms carrying fzr G0418 and mCherry.nls, which leads to wild-type and homozygous mutant progenies. (G) Diagram of temperature-shift experiment. Newly-hatched larvae are heat-shocked at 37˚C for 30 min, and then recovered at 18˚C until 120 hAH. (H) fzr G0418 /fzr G0418 clone PG cells. fzr G0418 /fzr G0418 homozygous mutant cells (indicated by dotted lines) do not express mRFP.nls (magenta and white in the left and middle panels, respectively). The PG and nuclear DNA were labelled by anti-Dib antibody (green in the left panels) and Hoechst (blue and white in the left and right panels, respectively), respectively. Scale bars: 10 μm.
https://doi.org/10.1371/journal.pgen.1008121.g002 controls, indicated by green plot in Fig 4D. With this analysis, we identified 31 genes whose knockdown caused a significant cell number increase in the PG (called "MES-related genes" hereafter) (magenta and purple plots in Fig 4D; summarized in S4 Table). To reveal biological processes important for MES, Gene ontology (GO)-term enrichment analysis was performed for MES-related genes, and we found that chaperonin containing tcp1 (cct) genes were significantly enriched in the MES-related gene group. CCT proteins are subunits of the evolutionary conserved molecular chaperon complex, TRiC, which supports proper folding of cytoskeletal proteins and cell cycle regulators [28,29]. Generally, TRiC is a hetero-oligomeric double-ring complex with eight subunits (CCT1-8) per ring [30]. Six cct subunit genes (cct1, 2, 4, 5, 6, and 8) were included in 31 MES-related genes, and knockdown of these genes caused not only an increased cell number but also a severe reduction in the DNA content (magenta plots in Fig  4D). Taken together, our RNAi screen raised the possibility that TRiC is a novel MES regulator in the PG.

TRiC is required for proper MES and endocycle progression in the PG
To investigate the role of TRiC in the PG, each cct subunit gene was knocked down in the PG. In contrast to the controls (phm>mCherry.nls) whose cell number and C value in the PG were around 50 and 53, respectively (Fig 5A-5C, S1A and S1B Fig), the PG cell number of cct1-8 RNAi animals (phm>mCherry.nls cct-RNAi) reached 60-70, and their C value was around 8 at 96 hAH (Fig 5A-5C). These results indicate that cct subunit genes are required for proper MES. In addition, pH3 expression was detected in some, but not all, PGs of cct RNAi larvae (Fig 5D), but we could not observe statistically significant difference in pH3 expression between control and cct RNAi (Fig 5D). This suggests that cct genes are also required for proper progression of mitotic cell cycle. We next confirmed that development was mainly arrested at the L3 stage in cct RNAi animals (phm>cct-RNAi) (Fig 5E and 5F), that CycA and B were detected by specific antibodies against each cyclin (magenta and white in the upper and lower panels, respectively). The PGs are indicated by dotted lines. Scale bars = 50 μm. (C and D) Scatter and box plots showing the percentage of CycA (C) and CycB (D)-positive PG cells of the controls (red) and fzr RNAi animals (blue) at indicated stages. Different lowercase letters indicate statistically significant differences (P < 0.05; Steel-Dwass test; see S1 Table). Box and dot plots as in Fig 1C-1E. Sample sizes (the number of PGs) are shown above each column. (E) The PG (upper panels) and PG cells in higher magnification (lower panels) of fzr RNAi (phm>dicer2 mCherry.nls fzr-RNAi), fzr RNAi + cycA RNAi (phm>dicer2 mCherry.nls fzr-RNAi cycA-RNAi), and fzr RNAi + cycB RNAi (phm>dicer2 mCherry.nls fzr-RNAi cycB-RNAi) at 96 hAH. pH3 and DNA were detected by anti-pH3 antibody (green) and Hoechst (blue), respectively, and the nuclei of PG cells were labelled by mCherry.nls (magenta). The PGs are indicated by dotted lines. The arrows indicate pH3-labelled mitotic PG cells. Scale bars: 50 μm (upper panels) and 10 μm (lower panels). (F-H) Scatter and box plots showing the cell number (F), the C value (G), and the percentage of pH3-positive cells (H) in the PG of fzr RNAi, fzr RNAi + cycA RNAi, and fzr RNAi + cycB RNAi animals at 96 hAH. Different lowercase letters indicate statistically significant differences (P < 0.05; Steel-Dwass test; see S1 Table). The mean C value of fzr RNAi at 96 hAH was set to 4 according to its mean C value at the same time point (Fig 2C). Sample sizes (the number of PGs) are shown above each column.(I) Percentages of pupariated fzr RNAi (phm>dicer2 fzr-RNAi), fzr RNAi + cycA RNAi (phm>dicer2 fzr-RNAi cycA-RNAi), and fzr RNAi + cycB RNAi animals (phm>dicer2 fzr-RNAi cycB-RNAi) are shown. The asterisks indicate statistically significant differences between fzr RNAi and fzr RNAi + cycA RNAi or fzr RNAi + cycB RNAi (P < 0.05; Fisher's test). Sample sizes (the number of animals) are shown above each column. (J) fzr RNAi animals arrested at the L3 stage and pupariated fzr RNAi + cycA RNAi and fzr RNAi + cycB RNAi animals at 120 hAH. (K) Whole-body ecdysteroid levels in fzr RNAi, fzr RNAi + cycA RNAi, and fzr RNAi + cycB RNAi animals at 96 hAH were measured using ELISA. Scatter and box plots show ecdysteroid level of five independent data sets (10 larvae were pooled in each datum). Different lowercase letters indicate statistically significant differences (P < 0.05; Steel-Dwass test; see S1 Table). (L) The expression level of ecdysone biosynthetic genes at 96 hAH was measured using qPCR. Scatter plots with average values of triplicate data sets are shown with SE (ten to fifteen larvae were pooled in each datum). Different lowercase letters indicate statistically significant differences (P < 0.05; Tukey's multiple comparison test; see S1 Table).
https://doi.org/10.1371/journal.pgen.1008121.g003 ecdysteroidogenic gene expression was significantly reduced in cct RNAi (Fig 5G), and that 20E administration restored larval-to-pupal transition in 20%-30% of animals (Fig 5H and  5I). The explanation for why only 20%-30% of cct RNAi was rescued is that the 20E concentration used in this rescue experiment may have been too high to trigger proper pupariation in cct RNAi animals. To test this possibility, we have administrated 20E against cct8 RNAi animals at a concentration of 0.5 mg/g (used mainly in this paper), 0.05 mg/g, 0.005 mg/g, 0.0005 mg/g, and 0 mg/g. As S7 Fig shows, approximately 30% of cct8 RNAi larvae fed on the medium with 0.5 and 0.05 mg/g of 20E undergo pupariation, but there was no significant difference in timing and the percentage of pupariation between these two groups. This suggests that lower efficiency of this rescue experiment is not explained by the concentration of 20E used. Considering that PG produces other humoral factors, including monoamines [27], one potential mechanism is that knockdown of ccts perturbs production of hormones other than ecdysone, which may cause developmental defects. Overall, these results indicate that TRiC is required for ecdysone biosynthesis in the PG to induce the larval-to-pupal transition.
Further observation of PG cells during larval development revealed that individual knockdown of cct4 and 8 caused a delay in both the onset of the DNA content increase and cessation of the cell number increase: In the controls (phm>mCherry.nls), the PG cell number reached around 45 at 24 hAH and its C value was continuously elevated after 24 hAH (Fig 6A-6C). By contrast, in cct4 and 8 RNAi animals (phm>mCherry.nls cct4/8-RNAi), the PG cell number reached around 60 by 48 hAH, and the C value of their PG cells did not increase after 48 hAH (Fig 6A-6C). Consistently, pH3-positive cells were detected in the PGs of cct4 and 8 RNAi animals even after 24 hAH, although not statistically significant (Fig 6D and 6E). These results indicate that mitotic cell cycle is prolonged (i.e., MES is delayed) in the PG of cct RNAi. Furthermore, we found that the rate of the DNA content increase was suppressed in the PGs of cct4 and 8 RNAi animals even after cessation of the cell number increase (Fig 6C). This result indicates that TRiC also regulates endocycle progression in the PG, as well as mitotic cell cycle and MES. To confirm this possibility, the cct4 mutant line named cct4 KG09280 , carrying a P-element insertion on the cct4 coding region that causes loss of cct4 mRNA expression and developmental arrest at the L1/L2 stages (S8 Fig), was used for FLP/FRT-based clonal analysis. FLPout clones carrying a cct4 KG09280 homozygous mutation in the PG showed decreased DNA content (Fig 6F), confirming the importance of TRiC in endocycle progression. Chaperonin-mediated cell cycle control The expression level of ecdysone biosynthetic genes was measured using qPCR. Scatter plots with average values of triplicate data sets are shown with SE (ten to fifteen larvae were pooled in each datum). Different lowercase letters indicate statistically significant differences (P < 0.05; Tukey's multiple comparison test; see S1 Table). (H) Percentages of pupariated cct1-8 RNAi animals, cultured on the medium with 20E (0.5 mg/g) or without 20E from 48 hAH are shown. The asterisks indicate statistically significant differences between 20E-fed and no 20E groups (P < 0.05; Fisher's test). Sample sizes (the number of animals) are shown above each column. (I) cct4 RNAi larva fed on -20E medium and pupariated cct4 RNAi animal fed on +20E medium.   Table (Steel-Dwass test). Sample sizes are the same as in B and C. (F) cct4 KG092808 /cct4 KG092808 FLP-out clone induced by heat shock. Newly-hatched larvae are heatshocked at 37˚C for 30 min, and then recovered at 18˚C until 120 hAH. FRT40A site was used in this clonal analysis. cct4 KG092808 homozygous mutant cell (indicated by

TRiC controls nuclear translocation of Fzr and downregulation of CycA
The above observations raised the question of how TRiC regulates MES and endocycling. To investigate whether TRiC controls MES via Fzr and mitotic cyclin regulation, expression of Fzr-GFP and mitotic cyclins was observed in the PGs of cct4 and 8 RNAi animals. In the controls (phm>mCherry.nls, fzr-GFP), Fzr-GFP was detected in both the cytoplasm and nuclei of PG cells (Fig 7A-7C). In contrast, localization of Fzr-GFP into the nuclei was suppressed in the PGs of cct4 and 8 RNAi larvae (phm>mCherry.nls cct4/8-RNAi, fzr-GFP) at 24 and 48 hAH (Fig 7A-7C), suggesting that TRiC regulates nuclear translocation of Fzr in the PG. Furthermore, probably due to the misregulation of Fzr nuclear translocation, CycA expression in the PG of cct4 RNAi (phm>mCD8::GFP cct4-RNAi) was significantly increased than those in the controls (phm>mCD8::GFP) at 24 hAH (Fig 7D and 7F). CycA was also detectable in the PG of cct8 RNAi (phm>mCD8::GFP cct8-RNAi) although this was not statistically significant (( Fig  7D and 7F). However, we could not observe enhanced CycB expression in cct4 and 8 RNAi (Fig 7E and 7G). These results suggest that TRiC-deficient PG cells cannot undergo proper MES owing to CycA upregulation. To test this possibility, we investigated whether knockdown of cycA restores MES in cct RNAi animals. As Based on our findings, here we propose a working model of TRiC-mediated control of MES and endocycle progression: TRiC downregulates CycA by regulating Fzr nuclear translocation to promote MES and endocycling (Fig 8).

Discussion
MES is an essential cellular process that changes cell states from proliferation to growth and initiates terminal differentiation in multicellular organisms. Here we used the Drosophila steroidogenic organ PG to study regulatory mechanisms of MES and found that PG cells undergo MES in a Fzr-dependent manner to activate ecdysteroid biosynthesis. Furthermore, our RNAi screen identified the evolutionary conserved chaperonin TRiC as a novel regulator of MES and endocycle progression. Further genetic analysis showed that TRiC downregulates CycA at least in part by regulating Fzr nuclear translocation to induce MES and subsequent endocycling. Based on these results, we propose that TRiC-mediated protein quality control is a fundamental mechanism supporting MES and subsequent endocycling that promotes terminal differentiation.

TRiC-mediated regulation of MES and endocycle
We investigated the role of TRiC in regulating Fzr and mitotic cyclin expression in the PG, and found that TRiC is required for nuclear translocation of Fzr (Fig 7). This result suggests that TRiC supports Fzr folding to facilitate its translocation into the nuclei. Furthermore, knockdown of cct subunit gene resulted in increased CycA expression (Fig 7), and knockdown of cycA along with cct subunit genes prevented PG cell number increase (S9 Fig), indicating that TRiC promotes CycA inactivation, which allows PG cells to undergo MES. Because nuclear translocation of Fzr was blocked in cct RNAi, increased CycA expression in the nuclei of cct RNAi is likely due to decreased Fzr translocation into the PG cell nuclei. However, in contrast to CycA, CycB expression was not disturbed in cct subunit RNAi (Fig 7). Thus, we speculate that TRiC is unnecessary for Fzr to recognize CycB. RNAi animals (blue) at indicated stages. Different lowercase letters indicate statistically significant differences (P < 0.05; Steel-Dwass test; see S1 Table). ns, not significant. Sample sizes (the number of PGs) are shown above each column. https://doi.org/10.1371/journal.pgen.1008121.g007 Although the PG cell number was significantly increased in cct RNAi, we could not observe a statistically significant increase in the percentage of pH3-positive PG cells in cct RNAi (Figs 5  and 6). The explanation for why pH3 was not detected frequently is that cct is also required for proper progression of mitotic cell cycle in the PG. Indeed, the PG cell number was not continuously increased in cct RNAi (Fig 6). Actually, it has been reported that TRiC regulates the disassembly of mitotic checkpoint complex [31] and mitotic cell cycle events such as sister chromatid separation [32]. Further elucidation of the regulatory mechanism of TRiC-mediated mitotic cell cycle is an important step to understanding the role of TRiC in cell cycle control.
In addition to the importance of TRiC in MES, we revealed that TRiC also has a critical role in regulating endocycle progression: inhibition of cct subunit genes caused endocycle arrest at around 8C (Figs 5 and 6), and knockdown of cycA together with cct in the PG partially restored endocycle up to 16C (S9 Fig). These results indicate that TRiC-mediated CycA downregulation is also required for progression of endocycling in the PG, perhaps due to reduced nuclear translocation of Fzr. Indeed, it has been reported that CycA as well as Fzr controls endocycle progression in the Drosophila [33][34][35]. However, cycA knockdown was not enough to restore the third round of endocycle in the PG of cct RNAi animals (S9 Fig). This suggests that other downstream factor(s) of TRiC or Fzr promote the third endocycle independently of CycA downregulation. Because Fzr also regulates Geminin degradation, a DNA replication inhibitor, to promote endocycle progression [33], one possible mechanism is that Fzr facilitates entry into the S-phase through suppression of Geminin to execute proper progression of endocycle in the PG. Furthermore, given that TRiC supports numerous proteins' folding, including tubulin and actin [28,29], we propose that TRiC-mediated protein quality control is a fundamental mechanism supporting MES and subsequent endocycling, which leads to the terminal differentiation.

PG is suitable for studying regulatory mechanisms of MES and endocycle
In this study, we used the PG as a model organ to study MES and endocycle regulatory mechanisms because of the organ's simple structure and the correlation between endocycling and ecdysone biosynthesis. PG cells undergo MES at the end of L1 and carry out repeated rounds, at least three times, of endocycle during the L2 and L3, which is essential for activation of ecdysone biosynthesis (Figs 1-3, S1 and S4 Figs). In contrast to PG cells, CA cells do not seem to undergo mitotic cell cycle and perform only one round of endocycle during the larval stage (Fig 1 and S2 Fig). These two types of endocrine cells originate from homologous ectodermal cells, and homeobox (Hox) gene expression controls PG and CA specification during embryogenesis [36]. Because CA and PG cells originate from deformed (Dfd)-and sex comb-reduced (Scr)-positive ectoderms, respectively [36], one possible mechanism is that distinct downstream genetic programs induced by Dfd and Scr determine the timing of MES and the activity of endocycle in these cells.
Our PG-selective RNAi screening identified not only the cct subunit genes required for proper MES but also other MES-related genes (summarized in S3 Table). Considering that enhancement of fzr transcription is a common step to triggering MES in Drosophila and other organisms [1,[10][11][12][13][14][15][16][17][18], our results will provide a solid basis for further investigating regulatory mechanisms of how fzr expression is upregulated to initiate MES. In Drosophila ovarian follicle cells, for example, the Notch signaling pathway promotes fzr transcription during the MES period [10,11]. However, because core components of Notch signaling were not included in our list of MES-related genes (S3 Table), the regulatory mechanisms of MES seem to be distinct between the PG and follicle cells. Thus, other signaling pathways are likely involved in transcriptional regulation of fzr in the PG. Furthermore, we identified a group of genes whose knockdown causes a significant decrease in the DNA content (Fig 4 and S3 Table), suggesting that these genes regulate endocycle progression. Endocycling in the PG is controlled by nutrient signaling, including the insulin/ TOR signaling pathway regulating the third endocycle [22]. However, upstream signaling pathways of the first, second, and fourth endocycle, as well as MES, have not been identified. Thus, detailed and systematic analysis of both MES-related and endocycle-related genes will shed light on the molecular mechanisms of how environmental and genetic cues are integrated into MES and endocycle progression in the PG to determine the onset of ecdysone biosynthesis.
In summary, we have demonstrated the genetic evidence showing the importance of TRiC in the regulation of MES and endocycle. Considering that TRiC suppresses accumulation of mitotic cyclins through the generation of functional Cdh1 protein in yeast [32], we propose that TRiC-mediated regulation of Cdh1/Fzr is an evolutionary conserved mechanism that promotes exit from mitotic cell cycle, including MES. Moreover, several lines of evidence have shown that some cct subunit genes are involved in the survival and proliferation of cancer cells [37]. Considering that cancer tissues possess endocycling cells at a high frequency and that endocycle is considered crucial for tumorigenesis [6][7][8], elucidating the role of TRiC in MES and endocycle progression is a fundamental step to revealing TRiC-mediated control of oncogenesis. This study thus provides a solid basis for revealing genetic programs that control initiation and progression of endocycle.

Drosophila stocks
Genotypes of the flies used in this study are summarized in S5-S7 Tables, and UAS-RNAi lines used for PG-selective RNAi screen and its phenotype were summarized in S3 Table. Fly stocks were maintained on standard Drosophila cornmeal/yeast medium at 18 or 25˚C under a 12-hour light/dark cycle.

Analysis of developmental progression
To obtain larvae just after hatching, parent flies were maintained in the bottle and allowed to lay eggs for 24 hours on grape juice agar plates supplemented with yeast powder. Newly hatched larvae were transferred to vials with nutrient-rich medium named as "German food (GF)" (https://bdsc.indiana.edu/information/recipes/germanfood.html). Larvae were cultured at 25˚C under a 12-hour light/dark cycle, and developmental stages and lethality were scored periodically.

Quantitative RT-PCR (qPCR)
Total RNA was extracted from whole larvae using TRIzol (Thermo). Reverse-transcription was performed using SuperScript III (Invitrogen). cDNA was used as a template for qPCR using Quantifast SYBR Green PCR kit (QIAGEN) and Rotor-Gene Q (QIAGEN). The expression level of target gene was normalized using an endogenous control, ribosomal protein 49 (rp49), and the relative expression level was calculated (relative expression level = expression value of the gene of interest/expression value of rp49). Primer sets used for qPCR are shown in S8 Table. Ecdysteroid measurement Ten larvae were rinsed with distilled water, and collected in a 1.5 ml microcentrifuge tube. The larvae were homogenized in 400 μl of methanol with a plastic pestle at room temperature. The samples were centrifuged at 15,000 g for 5 min at 4˚C, and 60 μl of the supernatant (equivalent to 1.5 larvae) was subjected to vacuum desiccation. Dried extract was re-dissolved in 50 μl of EIA buffer (Cayman Chemical). Ecdysteroid was quantitated by enzyme-linked immunosorbent assay (ELISA) using 20E EIA antiserum, 20E AchE tracer, and Ellman's reagent (Cayman Chemical) according to manufacturer's protocol. Standard 20E was purchased from Sigma.

20E administration
To rescue developmental arrest in RNAi animals, larvae were transferred to GF with 0.5 mg/g 20E at 48 hAH. Larvae transferred to GF without 20E at the same time point were used as control. Developmental stages were scored at 24-hour intervals.

Immunostaining and histochemistry
Larvae were dissected in phosphate buffered saline (PBS) and fixed for 25 min with 4% paraformaldehyde (PFA) in 0.1% PBT (0.1% Triton X-100 in PBS). Tissues were washed with 0.1% PBT three times for 10 min each, permeabilized with 1% PBT (1% Triton X-100 in PBS) for 5 min, blocked with 2% goat serum (Sigma, G9023) in 0.1% PBT for 30 min, and then incubated at 4˚C overnight with primary antibodies diluted in blocking solution. Tissues were washed with 0.1% PBT three times for 10 min each, and incubated at 4˚C overnight with Alexa 488-or Alexa 546-conjugated secondary antibodies (Thermo) in 0.1% PBT. Together with the secondary antibody, Hoechst 33342 (Thermo, H3570) was added at a 1:1500 dilution to detect DNA. After washing with 0.1% PBT three times for 10 min each, tissues were mounted in mounting medium.
Fly-FUCCI probe expressed in the PG was observed as follows: Larvae were dissected in phosphate buffered saline (PBS) and fixed for 25 min with 4% PFA in 0.1% PBT; Tissues were washed with 0.1% PBT three times for 10 min each, and incubated with Hoechst at a 1:1500 dilution at 4˚C overnight; We did use neither anti-GFP nor anti-mRFP antibody; After washing with 0.1% PBT three times for 10 min each, tissues were mounted in mounting medium.
Images were taken with a Zeiss LSM700, and the pictures' properties including the cell number and Fzr-GFP expression level were analyzed using Image J/Fiji [38]. Cell counting was performed using the plugin named Cell Counter. Fzr-GFP signal intensity in the PG was obtained from stacked slices of PG, and normalized by the signal intensity in the CA. Measurement of the DNA content was performed as described below.

Ploidy measurements and DNA quantification
Ploidy measurement was performed as previously described [39] with minor modifications. The C value in the PG and CA was determined using the following methods. Larvae and adult male flies were dissected in 0.7% NaCl. Brain-ring gland complex and testes were treated with 0.5% sodium acetate for 10 min and then fixed for 30 min with 4% paraformaldehyde in 0.1% PBT. After washing twice with 0.1% PBT, tissues were squashed on an APS-coated slide and submerged in liquid nitrogen to remove coverslip. Tissues were dehydrated in ethanol for 15 min and washed with PBS three times for 10 min each. Slides were stained with Hoechst (1:5000) for 15 min, washed with PBS three times for 10 min each, and mounted in mounting medium. Testes within each slides were imaged at the same gain and settings, in order to use the sperm cells as an internal control. Images were taken with a Zeiss LSM700, and analyzed using Image J/Fiji [38]. Regions were drawn around each nucleus using the trace function and the fluorescence intensity was measured within each region. The mean DNA staining intensity of sperm cells (1C) on the same slide was analyzed and set to 1. The average intensities of PGs were calculated on the basis of the C value in sperm cells.
To measure and quantify DNA signal intensity in the PG, a series of images obtained by immunostaining/histochemistry were processed using Image J/Fiji as follows. DNA signal overlapped with binarized and filled Sro signal (Oregon R) or phm>mCherry.nls signal (transgenic lines) was obtained using a function named "Image Calculator". Processed images were stacked, DNA signal in the PG was drawn around their nuclei using the trace function, and the fluorescent intensity was measured within the region. DNA staining intensity in the PG was adjusted using average DNA staining intensity obtained from z-stacked images of the brain lobe. Normalized DNA staining intensity was further divided by the PG cell number to obtain mean DNA intensity per PG cell. In accordance with the result of ploidy measurement, the mean C value in the PG of Oregon R at 84 hAH was set to 58 (see Fig 1B and 1C), and the mean C values in the PG of phm>mCherry.nls and phm>dicer2 mCherry.nls at 96 hAH were set to 53 and 54, respectively, (see S1A and S1B Fig).

BrdU incorporation
Larvae were dissected in Ringer's solution, and the tissues were incubated for 30 min with 100 mM BrdU (Sigma, B5002) diluted in Ringer's solution and then fixed in 4% PFA for 20 min. Fixed tissues were briefly washed twice in 0.01% PBT (0.01% Triton X-100 in PBS), washed in 0.1% PBT twice for 10 min each, and treated with 2N HCl for 30 min. The tissues were briefly washed twice in 0.01% PBT, washed in 0.1% PBT twice for 10 min, and incubated with blocking solution for 30 min. The tissues were incubated with primary antibody against BrdU diluted 1:20 in blocking solution overnight at 4˚C. The tissues were washed in 0.1% PBT for 10 min three times, and incubated 4˚C overnight with Alexa 488 fluor-conjugated secondary antibody (mouse IgG, Thermo Fisher, A-11001) and Hoechst diluted at 1:1000 and 1:1500, respectively, in 0.1% PBT. The tissues were washed in 0.1% PBT for 10 min three times and mounted on a slide glass with mounting medium. Images were taken with a Zeiss LSM700, and their analysis was performed using Image J/Fiji [38].

PG-selective RNAi screening
Ten virgins carrying two copies of PG-selective phm-Gal4 and UAS-mCherry.nls were crossed with five UAS-RNAi males to obtain the offspring in which gene of interest is knocked down in the PG. Parent flies were cultured on standard Drosophila medium in plastic vials for 2 days. Developmental phenotype of their progenies was observed to confirm developmental defects at day 10 after crossing. The PG of RNAi animals showing developmental defect were observed using histochemistry at day 6 after crossing as follows. Larvae at day 6, in which most of control larvae (phm>mCherry.nls) are in wandering stage, were dissected in PBS and fixed for 25 min with 4% PFA in 0.1% PBT. Tissues were washed with 0.1% PBT three times for 10 min each, and stained with Hoechst 33342 (1:1500 in 0.1% PBT) to observe the DNA content and distribution and morphology of PG cells. Images were taken with a Zeiss LSM700, and their analysis was performed using Image J/Fiji [38]. DNA quantification was performed as described above (see Ploidy measurement).

Statistical analysis
Statistical analyses were performed using R (http://www.R-project.org/). Exact P-values of Steel Dwass and Tukey's multiple comparison test in main and supplemental figures are shown in S1 and S2 Tables. All numerical data except for the data obtained in a RNAi screen are shown in S1 Data. Numerical data in the RNAi screen are shown in S3 Table. In our RNAi screen, a significant increase in the PG cell number was determined using Dunnet's multiple comparison test. In all statistical analysis, P < 0.05 was considered to represent a statistically significant difference. GO-term enrichment analysis was performed to evaluate which biological process term is enriched in a group of genes using Reactome resource in PANTHER (http://pantherdb.org/).  Table). Box and dot plots as in Fig 1C-1E. Sample sizes (the number of PGs) are shown above column. (E and F) Schematics of Drosophila FUCCI system. E2F1 1-230 -fused GFP (GFP.E2F1) and CycB 1-266 -fused mRFP1 (mRFP1.CycB) expressed under the control of Gal4/UAS system were degraded through CRL4-and APC/C-dependent manner, respectively (E). Since CRL4 and APC/C-dependent protein degradation are active at S and G1 phase in mitotic cell cycle, G1-, S-, and G2/M-phase cells were labelled by GFP, mRFP1, and both GFP and mRFP1, respectively (F). (G) The expression patterns of GFP.E2F1 (green and white in the upper and middle panels, respectively) and mRFP1.CycB (magenta and white in the upper and lower panels) in the PG of FUCCI reporter-expressing animals (phm>GFP.E2F1 mRFP1.CycB) at indicated time points. DNA was stained by Hoechst (blue in the upper panels). The PGs are indicated by dotted lines. The arrows indicate mitotic cell showing metaphase-like nuclear shape, which were counted as "mitotic cells" (yellow in H). Scale bars: 10 μm. The expression level of ecdysone biosynthetic genes in the controls and fzr RNAi measured using qPCR at indicated time points. Average values of triplicate data sets with SE and scatter plots are shown. Ten to fifteen larvae were pooled in each datum. Different lowercase letters indicate statistically significant differences (P < 0.05; Tukey's multiple comparison test; see S2 Table). (E) Whole-body ecdysteroid levels in the controls and fzr RNAi animals at 96 hAH measured using ELISA. Ecdysteroid levels of five independent data sets are shown by scatter and box plots. Ten larvae were pooled in each datum. The asterisk indicates statistically significant differences (P < 0.05; Welch's two sample t-test). (F) Percentages of pupariated fzr RNAi animals cultured on the medium with 20E (0.5 mg/g) or without 20E from 48 hAH. Sample sizes (the number of animals) are indicated in parentheses. The asterisk indicates statistically significant differences (P < 0.05; Fisher's test).  Fig. Genetic interaction between cct subunits and cycA. (A) The PG (upper panels) and PG cells in higher magnification (lower panels) of cct4 RNAi (phm>mCherry.nls cct4-RNAi), cct4 RNAi + cycA RNAi (phm>mCherry.nls cct4-RNAi cycA-RNAi), cct8 RNAi (phm>mCherry. nls cct8-RNAi), and cct8 RNAi + cycA RNAi (phm>mCherry.nls cct8-RNAi cycA-RNAi) at 96 hAH. pH3 and DNA were detected by anti-pH3 antibody (green) and Hoechst (blue), respectively, and the nuclei of PG cells were labelled by mCherry.nls (magenta). The PGs are indicated by dotted lines. The arrows indicate pH3-positive PG cells. Scale bars: 50 μm (upper panels) and 10 μm (lower panels). (B-D) Scatter and box plots showing the cell number (B), the C value (C), and the percentage of pH3-positive cells (D) in the PG of cct4 RNAi, cct4 RNAi + cycA RNAi, cct8 RNAi, and cct8 RNAi + cycA RNAi animals at 96 hAH. The asterisks indicate statistically significant differences (P < 0.05; Mann-Whitney U test). Box and dot plots as in Fig 1C-1E. The mean C values of cct4 and cct8 RNAi at 96 hAH were set to 9 and 8.8, according to their mean C value at the same time point (Fig 6C). Sample size (the number of PGs) are shown above each column. (E) Percentages of pupariated and L1/L2-and L3-arrested animals in cct4 RNAi, cct4 RNAi + cycA RNAi, cct8 RNAi, and cct8 RNAi + cycA RNAi are shown. Sample sizes (the number of animals) are shown above each column. ns, not significant (P > 0.05; Fisher's test). (TIF) S1 Table. Exact P-value in main figures. � Steel Dwass multiple comparison test �� Tukey's multiple comparison test a All data were obtained in Oregon R b All data were obtained in phm>mCherry.nls, fzr-GFP.  Table. Genotypes of the flies used in this study. � UAS-mCherry.nls or UAS-mCD8::GFP was introduced into the 3rd chromosome when offspring's PGs were observed using histochemistry �� Offsprings carrying GFP-balancer chromosome was removed in all experiment. (XLSX) S8