Expression of an S phase-stabilized version of the CDK inhibitor Dacapo can alter endoreplication

Abstract

In developing organisms, divergence from the canonical cell division cycle is often necessary to ensure the proper growth, differentiation, and physiological function of a variety of tissues. An important example is endoreplication, in which endocycling cells alternate between G and S phase without intervening mitosis or cytokinesis, resulting in polyploidy. Although significantly different from the canonical cell cycle, endocycles use regulatory pathways that also function in diploid cells, particularly those involved in S phase entry and progression. A key S phase regulator is the Cyclin E-Cdk2 kinase, which must alternate between periods of high (S phase) and low (G phase) activity in order for endocycling cells to achieve repeated rounds of S phase and polyploidy. The mechanisms that drive these oscillations of Cyclin E-Cdk2 activity are not fully understood. Here, we show that the Drosophila Cyclin E-Cdk2 inhibitor Dacapo (Dap) is targeted for destruction during S phase via a PIP degron, contributing to oscillations of Dap protein accumulation during both mitotic cycles and endocycles. Expression of a PIP degron mutant Dap attenuates endocycle progression but does not obviously affect proliferating diploid cells. A mathematical model of the endocycle predicts that the rate of destruction of Dap during S phase modulates the endocycle by regulating the length of G phase. We propose from this model and our in vivo data that endo S phase-coupled destruction of Dap reduces the threshold of Cyclin E-Cdk2 activity necessary to trigger the subsequent G-S transition, thereby influencing endocycle oscillation frequency and the extent of polyploidy.

Keywords: CDK inhibitor; CRL4Cdt2; Cell cycle; Drosophila; Endocycle; Modeling; Polyploidy.

Fig. 1.

Dap contains a PIP degron that mediates S phase-coupled destruction. (A) Alignment of human (Hs) p21, Xenopus (Xl) Xic1, C. elegans (Ce) CKI-1, and Dap PIP degrons. PCNA- and Cdt2-binding residues highlighted in red; consensus PIP degron at top. The alanine substitution mutations in the Dap PIP degron are italicized. (B,C) Flow cytometry of hsp70-GFP-Dap cells analyzed 2 h after 30-min heat shock. (B) The percentage of GFP-positive cells in S phase (green bars) is plotted with the percentage of cells in the total population in S phase (white bars; includes GFP-positive and GFP-negative cells). Error bars indicate standard deviation (s.d.), *P<0.05. (C) Representative cell cycle profiles with solid traces indicating the total cell population and green traces indicating GFP-positive cells. (D) 7-h-old wild-type embryo stained with anti-Dap antibodies (green) and BrdU (red). White arrows indicate a BrdU-negative cell with high Dap expression, red arrows indicate a BrdU-positive cell with low Dap expression. White box outlines area shown in higher magnification. (E-G) 5-h-old embryos expressing GFP, GFP-Dap, or GFP-Dap^mDeg using tub-Gal4 and stained with anti-GFP antibodies (green) and BrdU (red). Yellow arrows in E,G indicate GFP-positive cells that are also BrdU positive, white and red arrows in F as in D. Note that GFP-Dap^mDeg accumulates in cells undergoing S phase (G). (H) Quantification of GFP fluorescence in BrdU-negative (left) and BrdU-postive (right) cells from 5-h-old tub>GFP-Dap and tub>GFP-Dap^mDeg embryos. Each point represents a single cell; three embryos per genotype were used for quantification. Middle bars represent mean; error bars indicate s.d.; ns, not significant; ****P<0.00005.

Fig. 2.

S phase-stabilized Dap expression does not disrupt the embryonic epidermal cell cycle program. (A-C) 6-h-old embryos expressing GFP (A) or GFP-Dap (B,C) transgenes with tub-Gal4 and stained for GFP (green) and BrdU (red). Epidermal cells in all three genotypes enter and progress normally through S phase of cell cycle 16 (S16). (D,E) Germ band-retracted embryos expressing GFP (D) or Dap (E) transgenes with prd-Gal4 and stained with anti-GFP (green) and anti-Dlg (red) antibodies to mark the cell cortex to indicate cell size. White boxes outline area shown in higher magnification.

Fig. 3.

S phase-stabilized Dap expression disrupts endocycle progression in the embryonic gut. (A-C) Stage 13 embryos expressing GFP (A) or GFP-Dap (B,C) transgenes with tub-Gal4 and stained for GFP (green) and BrdU (red). The first endocycle S phase in the anterior midgut (amg) and posterior midgut (pmg) normally occurs concurrently. (D) Quantification of percentage of anterior midgut cells undergoing S phase in Stage 13 embryos. (E) Relative cell density in the anterior midguts of Stage 13 embryos. Each dot represents one embryo. Middle bars represent mean; error bars indicate s.d.; ns, not significant; *P<0.05; **P<0.005.

Fig. 4.

S phase-stabilized Dap expression disrupts endocycle progression in ovarian follicle cells. (A-C) Stage 9 egg chambers expressing GFP (A) or GFP-Dap (B,C) transgenes with c323-GAL4 and stained for GFP (green), EdU (red), and DAPI (blue). (D,E) Higher magnification views (white boxes in B and C, respectively) of EdU-negative nuclei with high GFP-Dap or GFP-Dap^mDeg (white arrows), an EdU-positive nucleus with low GFP-Dap expression (red arrows), and EdU-positive nuclei with different amounts of GFP-Dap^mDeg (yellow arrows/arrowheads). Nuclei with extensive EdU incorporation (yellow arrow) typically had less GFP-Dap^mDeg than nuclei with more modest EdU incorporation (yellow arrowhead). (F) Quantification of GFP fluorescence in EdU-negative (left) and EdU-postive (right) cells from Stage 10 ovarian follicle cells. Each point represents a single cell; three egg chambers per genotype were used for quantification. (G) Quantification of the percentage of follicle cells undergoing S phase in Stage 10 egg chambers. Each dot represents one egg chamber. Middle bars represent mean; error bars indicate s.d.; *P<0.05; ****P<0.00005.

Fig. 5.

S phase-stabilized Dap expression disrupts endocycle progression in salivary glands. (A) Wild-type third instar larval salivary gland stained with anti-Dap antibodies (green), DAPI (blue) and EdU (red). White box outlines the higher magnification region showing an EdU-negative, G phase nucleus with high levels of Dap (white arrow), an extensively labeled S phase nucleus with no Dap (red arrow), and an S phase nucleus with a low level of Dap protein that does not overlap with the EdU signal (yellow arrow). (B-D) Early third instar larval salivary glands expressing GFP (B) or GFP-Dap transgenes (C,D) with ptc-GAL4 and stained for GFP (green), DAPI (blue) and EdU (red). fb, fat body. (E,F) Higher magnification images of ptc>GFP-Dap and ptc>Dap^mDeg salivary glands (white boxes in right panels of C,D) showing EdU-negative nuclei with high GFP-Dap or GFP-Dap^mDeg (white arrows), an EdU-positive nucleus with very low or no GFP-Dap (red arrow), and an EdU-positive nucleus with high GFP-Dap^mDeg (yellow arrow). (G) Immunoblot of equal numbers of salivary glands of the indicated genotypes using anti-GFP antibodies to detect GFP-Dap. Anti-lamin is a loading control. (H) Quantification of GFP fluorescence in EdU-negative (left) and EdU-postive (right) cells from salivary glands. Each point represents a single cell; three glands per genotype were used for quantification. (I-K) Quantification of the S phase index (I), nuclear size (J), and DAPI intensity (K) in salivary glands expressing GFP or GFP-Dap transgenes. In J and K ptc>GFP-Dap and ptc>GFP-Dap^mDeg measurements were normalized to the average nuclear size and DAPI intensity in ptc>GFP glands, respectively. Middle bars indicate the mean; error bars indicate s.d.; ns, not significant; ***P<0.0005; ****P<0.00005.

Fig. 6.

Mathematical modeling of the Drosophila endocycle. (A) Network diagram of the molecular interactions within an existing endocycle model (Zielke et al., 2011). Our additions to the previous model are in bold. (B) The Zielke et al., (2011) endocycle model, which did not include Dap, showing oscillation of the levels or activity of key cell cycle regulators. The x and y axes display arbitrary units. (C) The revised endocycle model including Dap (orange curves). (D) Removal of CRL4^Cdt2 regulation of endogenous Dap eliminates endocycle oscillations. (E-H) Modeling exogenous Dap. Addition of wild-type, exogenous Dap at either 1× (E) or 10× (G) relative levels reduces endocycle frequency. Removal of CRL4^Cdt2 regulation of exogenous Dap at either level of expression (F and H, respectively) eliminates endocycle oscillations. (I) Reduction of S phase degradation of endogenous Dap alters endocycle periodicity.