Fig 1.
Loss of CDK8 or CycC leads to defective pupal morphology and delayed larval–pupal transition.
(A–D) Compared to the control (A; w1118), cdk8 (B; w1118; +; cdk8K185) and cycC (C; w1118; +; cycCY5) single mutants, as well as the cdk8-cycC double mutants (D; w1118; +; cdk8K185, cycCY5) fail to evert their anterior spiracles and the prepupae are partially separated from the pupal case (arrows). Scale bar in (A): 0.2mm. (E and F) These defects were rescued in transgenic lines carrying genomic fragments of wild-type cdk8+ (E; w1118; cdk8+-EGFP; cdk8K185) or cycC+ (F; w1118; cycC+-EGFP; cycCY5), respectively. (G) The larval-to-pupal transition was analyzed by observing the percentage of pupariated animals after egg laying (AEL) once every 12 hr. (H) The time from egg deposition to pupariation in cdk8, cycC mutants, and the rescued animals. * p < 0.05; ** p < 0.01 based on t-tests. Underlying numerical data and statistical analysis for Fig 1G and 1H can be found in S1 Data.
Fig 2.
Expression of EcR-target genes is defective in cdk8 or cycC mutant larvae.
(A) Microarray analyses of genes whose products are related to functions of the ecdysone and JH. The RpL32 (Rp49, green arrow head) serves as the negative control, and cdk8 and cycC are positive controls (red arrow heads). (B and C) The microarray data were validated by qRT-PCR in the L3 wandering larvae. * p < 0.05; ** p < 0.01 based on t-tests. (D–F) Compared to the control (salivary glands in D and D’ are from the same larvae; genotype: w1118; EcRE-lacZ;+), the expression of EcRE-lacZ was significantly reduced in cdk8 (E versus E’, from the same larvae; genotype: w1118; EcRE-lacZ; cdk8K185) and cycC (F versus F’, from the same larvae; genotype: w1118; EcRE-lacZ; cycCY5) mutant genetic backgrounds. Scale bar in (F): 0.1mm. Underlying numerical data and statistical analysis for Fig 2B and 2C can be found in S1 Data.
Fig 3.
The effects of cdk8 or cycC mutation on biosynthesis of ecdysteroids.
(A) Expression of the Halloween genes analyzed by qRT-PCR in the L3 wandering larvae of the following genotypes: w1118 (black bars), cdk8K185 (red), and cycCY5 (blue). (B) Expression of genes encoding factors that are involved in regulating degradation of 20E (Cyp18a1), expression of the Halloween genes (mld, kni, and vvl), the neuropeptide prothoracicotropic hormone PTTH (ptth), and metabolism (tor and InR) in cdk8 and cycC mutant L3 wandering larvae. (C) Ecdysteroid titers determined by ELISA in cdk8 and cycC mutant animals from early L3 larvae to white prepupal stage. (D) Supplying the EcR ligand 20-hydroxyecdysone (20E, 200 μM) in food did not rescue the aberrant morphology of the cdk8 and cycC mutants. (E and F) Supplying 20E in food efficiently rescued the developmental arrest caused by either ablating PG cells using PG-specific (phm-Gal4) expression of reaper (E) or PG-specific knockdown of spok using RNAi (F). (G) Quantification of the effect of 20E (200 μM) on the time from egg deposition to pupariation in cdk8 and cycC mutants. * p < 0.05; ** p < 0.01 based on t-tests. Underlying numerical data and statistical analysis for Fig 3A, 3B, 3C, and 3G can be found in S1 Data.
Fig 4.
The levels and subcellular distribution of EcR and USP in cdk8 and cycC mutants in the third instar larvae and pupae.
(A) Western blot of EcR and USP in wild-type animals in early L3 (84 hr AEL), L3 wandering (112 hr AEL), white prepupal (120 hr AEL), and pupal stages (72 hr APF). (B and C) Western blot analyses of the protein levels of EcR, USP, CDK8, and CycC in cdk8 and cycC mutants at the L3 wandering stage (B) and the white prepupal stage (C). The arrows in USP blots mark the 54 kDa full-length USP, and the arrows in EcR blots indicate the EcR-B1 isoform. (D) Western blot of EcR and USP in the nuclear and cytoplasmic fractions from early third instar larvae (L3) to white prepupal stages in cdk8 or cycC mutants.
Fig 5.
The levels of EcR and USP on polytene chromosome correlate with the expression of EcR target genes in salivary glands.
(A–C) Immunostaining of EcR (A, B, and C) and USP (A’, B’, and C’) in w1118 (control), cdk8K185, and cycCY5 polytene chromosomes, which were also stained with DAPI (A”, B”, and C”). Scale bar in (C”): 20μm. (D) Quantification of EcR target gene expression in salivary glands by qRT-PCR. (E) ChIP assay of USP binding to EcR target gene promoters in the w1118, cdk8K185, and cycCY5 mutant larvae. Ctrl: control; GPNS: guinea pig normal serum; IP: immunoprecipitation. * p < 0.05; ** p < 0.01 based on t-tests. Underlying numerical data and statistical analysis for Fig 5D and 5E can be found in S1 Data.
Fig 6.
Biochemical interactions between CDK8-CycC and EcR-USP.
(A) EcR-B1 co-immunoprecipitates with CDK8 in white prepupae; USP serves as the positive control. (B) USP co-immunoprecipitates with CDK8 in white prepupae; EcR serves as the positive control. (C) Mediator subunits that can co-immunoprecipitate with EcR or USP are identified by LC-MS/MS analysis. Subunits of the CDK8 submodule are shown in color. The results were combined from two biological replicates. (D) Immunoprecipitation of EcR by using anti-USP (guinea pig) antibody in white prepupae of w1118, cdk8K185, and cycCY5 mutants, and the input is shown in (D’). (E) CDK8 has a LXXLL motif that is highly conserved from yeast to human. The LXXLL motif is highlighted in red. (F) The LXXLL motif in Med14 is conserved from Drosophila to humans, but it is not present in Caenorhabditis elegans. (G) Schematic diagram of the EcR-B1 and USP protein depicting the two activating domains (AF1 and AF2), DNA-binding domain (DBD) and the ligand-binding domains (LBD). (H) Yeast two-hybrid analyses show that EcR-AF1, but not EcR-AF2 or USP-AF1/2, can directly bind to CDK8 and Med14. Underlying numerical data and statistical analysis for Fig 6H can be found in S1 Data.
Fig 7.
CDK8-CycC may couple nutrient intake, lipid biosynthesis, and developmental timing.
(A) Model for the CDK8-SREBP/EcR regulatory network: In response to nutrient intake, CDK8-CycC may coordinately regulate lipogenesis by directly inhibiting SREBP-activated gene expression and developmental timing by activating EcR-activated gene expression during the larval–pupal transition. Arrows represent activation, and blunt arrows represent inhibition. (B) The protein levels of CDK8, CycC, SREBP, EcR-B1 and USP (upper band is the 54 kDa full-length USP) in wild-type larvae from L3 (92 hr AEL) to the WPP stage (120 hr AEL). For SREBP, the lower band (approximately 49 kDa, arrow) is the mature nuclear form, while the upper band (53–54 kDa) is the N-terminal fragment of SREBP after cleavage by the S1P (see S5B and S5C Fig for detailed analyses of these SREBP isoforms). (C–K) The mRNA levels of cdk8, cycC, EcR, usp, E75, E78, SREBP, dFAS, and dACS from L3 (92 hr AEL) to the WPP stage (120 hr AEL). The black bars represent the wandering stage, while the white bars represent the WPP stage. The x-axis represents the number of hours AEL. * p < 0.05; ** p < 0.01 based on t-tests. Underlying numerical data and statistical analysis for Fig 7C, 7D, 7E, 7F, 7G, 7H, 7I, 7J, and 7K can be found in S1 Data.
Fig 8.
The effects of starvation on the CDK8-SREBP/EcR regulatory network and timing for the larval–pupal transition.
(A) The protein levels of CDK8, CycC, mature SREBP (arrow), EcR-B1 (arrow), and USP (arrow, full-length USP) in feeding versus starved larvae from 84 hr to 100 hr AEL. (B–G) The mRNA levels of cdk8, EcR, E74, SREBP, dFAS, and dACS in feeding versus starved larvae from 84 hr to 100 hr AEL. The x-axis represents the number of hours AEL. * p < 0.05; ** p < 0.01 based on t-tests. (H) The effect of starvation and refeeding on the timing of the larval–pupal transition in wild-type (w1118) larvae. The larvae enter into L3 at 72 hr AEL and reach critical weight between 80 and 82 hr AEL. The larvae were starved starting 84 hr AEL and the timing of pupariation was analyzed once every two hours. For the refeeding experiments, the larvae were put back on normal food after 10 hr of starvation (94 hr AEL). Underlying numerical data and statistical analysis for Fig 8B, 8C, 8D, 8E, 8F, 8G, and 8H can be found in S1 Data.
Fig 9.
The effects of refeeding of starved larvae on the CDK8-SREBP/EcR regulatory network.
(A) The experimental scheme for the refeeding treatment. Briefly, the wild-type larvae were starved for 10 hr (from 84 hr AEL to 94 hr AEL), and they were then transferred back onto normal food. Samples were collected after 1, 2, 3, 6, 9 hr after refeeding for further analyses. (B) The protein levels of CDK8, CycC, SREBP, EcR-B1 and USP in starved versus refed larvae after 1, 2, 3 hr of refeeding. (C) The protein levels of CDK8, EcR-B1, and USP in starved versus refed larvae after 3, 6, 9 hr of refeeding. The control (anti-actin) is the same as the S12A Fig (D–H) The mRNA levels of cdk8, EcR, SREBP, dFAS, and dACS in starved versus refed larvae after 1, 2, 3 hr of refeeding. The x-axis represents the number of hours for refeeding. * p < 0.05; ** p < 0.01 based on t-tests. Underlying numerical data and statistical analysis for Fig 9D, 9E, 9F, 9G, and 9H can be found in S1 Data.