Drosophila casein kinase I alpha regulates homolog pairing and genome organization by modulating condensin II subunit Cap-H2 levels

Abstract

The spatial organization of chromosomes within interphase nuclei is important for gene expression and epigenetic inheritance. Although the extent of physical interaction between chromosomes and their degree of compaction varies during development and between different cell-types, it is unclear how regulation of chromosome interactions and compaction relate to spatial organization of genomes. Drosophila is an excellent model system for studying chromosomal interactions including homolog pairing. Recent work has shown that condensin II governs both interphase chromosome compaction and homolog pairing and condensin II activity is controlled by the turnover of its regulatory subunit Cap-H2. Specifically, Cap-H2 is a target of the SCFSlimb E3 ubiquitin-ligase which down-regulates Cap-H2 in order to maintain homologous chromosome pairing, chromosome length and proper nuclear organization. Here, we identify Casein Kinase I alpha (CK1α) as an additional negative-regulator of Cap-H2. CK1α-depletion stabilizes Cap-H2 protein and results in an accumulation of Cap-H2 on chromosomes. Similar to Slimb mutation, CK1α depletion in cultured cells, larval salivary gland, and nurse cells results in several condensin II-dependent phenotypes including dispersal of centromeres, interphase chromosome compaction, and chromosome unpairing. Moreover, CK1α loss-of-function mutations dominantly suppress condensin II mutant phenotypes in vivo. Thus, CK1α facilitates Cap-H2 destruction and modulates nuclear organization by attenuating chromatin localized Cap-H2 protein.

Fig 1. CK1α inactivation alters interphase chromosome morphology.

(A) Representative micrographs of Drosophila cultured S2 cells displaying different chromatin-gumball phenotype classifications. Scale, 2.5μm. (B-E) Micrographs of 6 day RNAi-treated Kc cells stained with DAPI to visualize DNA. Depletion of Slimb (C) or CK1α (D) but not control (B) promotes the chromatin-gumball phenotype, while double RNAi with CK1α and Cap-H2 (E) suppresses this phenotype. Scale, 5μm. (F) Frequency histogram of the nuclear phenotypes in S2 cells after 6-day depletion of the indicated proteins via RNAi; (n = 2200–4200 cells per treatment). (G-H) Micrographs of S2R+ cells stained with DAPI to visualize DNA. Treatment of cells with the CK1 inhibitor D4476 (80μM) (H) for 8 hours promotes chromatin-gumball phenotype not observed with control DMSO treatment (G). Scale, 5μm. (I) Frequency histogram of the nuclear phenotypes in S2R+ cells after treatment with control (DMSO) or CK1 inhibition (D4476); (n = 110–170 cells per treatment). p-value = * < 0.05, ** < 0.01, *** = < 0.005, **** = < 0.001, ***** = < 0.0001 (calculated by using students’ t-test in Microscoft excel). Horizontal lines on histograms represent comparisons between percentage of normal cells of two treatments at either end of the line. Error bars in all figures indicate SEM. (A-E,G-H) Maximum projection image of multiple z-slices.

Fig 2. RNAi of CK1α leads to dispersal of centromeres in Kc cells.

(A-G) RNAi treated Kc cells immunostained for centromeric protein (CID, green), Lamin Dm0 (red), and counterstained for DNA (DAPI, blue). CK1α RNAi (E) induces abnormal CID dispersal similar to Slimb RNAi (D), which is suppressed by double RNAi of CK1α + Cap-H2 (F) or CK1α + SMC2 (G). Scale, 2.5μm. (H) Histogram showing average number of CID spots per nucleus after RNAi depletion of the indicated protein. Slimb and CK1α depletion results in a significant increase in average number of CID spots per nucleus; (n = 115–180 cells per treatment). (I) Histograms showing the distribution of CID spots per nucleus. CK1α depletion results in an increased proportion of cells with higher than normal CID spots; (n = 115–150 cells per treatment). N.S. = No significance. p-value = * < 0.05, ** < 0.005, *** < 0.0001 (calculated by using students’ t-test in Microscoft excel). Statistical comparisons are between RNAi treatments and control, unless denoted by horizontal line between bars. Error bars in all figures indicate SEM. (A-G) Maximum projection image of multiple z-slices.

Fig 3. CK1α depletion increases chromosome compaction and unpairing activity in Kc cells.

(A-F) Micrographs of RNAi treated Kc cells stained for DAPI (blue) and three different euchromatic FISH probes specific for the X chromosome (green, red, and white). Single FISH spot for each probe signifies that the homologs are paired and multiple FISH spots indicate the unpairing of chromosomes. Scale, 5μm. (G) Histogram showing the average 3D distance between pairwise FISH spots in microns in RNAi treated Kc cells (n = 45–100 cells per treatment). CK1α RNAi significantly reduces the distance between FISH spots, and this reduction is suppressed by double RNAi of CK1α + Cap-H2 or CK1α + SMC2. (H) Histogram showing the average number of FISH spots per nucleus in RNAi depleted Kc cells (n = 50–110 cells per treatment). CK1α RNAi significantly increases the number of FISH spots, and this increase is suppressed by double RNAi of CK1α + Cap-H2 or CK1α + SMC2. (I-L) Micrographs of RNAi treated Kc cells stained with FISH probes specific to heterochromatin on Chromosome 2R (green) and 3R (red) and counterstained for DNA (DAPI, blue). Cap-H2 overexpression (J) and CK1α depletion (K) induces unpairing of heterochromatin, which is suppressed by double RNAi of CK1α + Cap-H2 (L). Scale, 2.5μm. (M) Histogram showing average number of heterochromatin FISH spots per nucleus after RNAi depletion of the indicated protein. CK1α depletion results in a significant increase in number of FISH spots; (n = 40–78 cells per treatment). (N) Diagram of X chromosomes. Blue represents the DNA, yellow diagonal stripes represents the centromere, and green (X1), red (X2), and white (X3) rectangles represent FISH probes used for compaction and unpairing experiments. Depletion of CK1α via RNAi results in compaction and unpairing of chromosomes. N.S. = No significance. p-value = * < 0.05, ** < 0.01, *** < 0.005, **** < 0.001, ***** < 0.0001 (calculated by using students’ t-test in Microscoft excel). Error bars in all figures indicate SEM. (A-F,I-L) Maximum projection image of multiple z-slices.

Fig 4. CK1α is required for polytene pairing in Drosophila salivary glands.

(A-B) Salivary gland nuclei from control larvae (43B>Gal4; Gal4 under a salivary gland specific driver) (A) and larvae expressing hairpin RNAi to CK1α (CK1α ^RNAi) driven by 43B>Gal4 (B) were hybridized with FISH probes specific to a region of Chromosome 2L (green) and Chromosome X (red) and counterstained with DAPI (blue). Chromosomes are highly paired in control (fewer FISH foci) (A) nuclei whereas expression of a TRiP hairpin RNAi targeting CK1α (B) induces unpairing of the chromosomes (multiple FISH foci). (C) Histogram showing average number of FISH spots per nucleus after CK1α RNAi is expressed in the salivary glands. CK1α depletion results in an increase in number of FISH spots for both 2L and X probes; (n = 24–38 nuclei per genotype). *p-value < 7.2x10⁻⁷ (calculated by using students’ t-test in excel). Error bars indicate SEM. (A-B) DAPI channel image is a single z-slice from the nucleus and FISH channel images are from maximum projection image of multiple z-slices. Scale, 10μm.

Fig 5. CK1α mutations suppress condensin II loss of function unpairing phenotype in polyploid nurse cells.

(A-F) Micrographs of stage 10 nurse cell nuclei from control (triple balancer) (A), CK1α^8B12/+ heterozygote mutant (B), CK1α^EP1555/+ heterozygote mutant (C), condensin II double heterozygous mutant SMC4^k00819/+; Cap-H2^z3–0019/+ (D), CK1α^8B12/+ heterozygote in condensin II double heterozygous background (CK1α^8B12/+; SMC4 ^k00819/+; Cap-H2^z3–0019/+) (E), and CK1α^EP1555/+ heterozygote in condensin II double heterozygous background (CK1α^EP1555/+; SMC4^k00819/+; Cap-H2^z3–0019/+) (F) were stained with FISH probes specific to Chromosome 2L (green) and Chromosome X (red) and counterstained with DAPI (DNA, blue). Condensin II loss of function mutants (SMC4^k00819/+; Cap-H2^z3–0019/+) (D) show a defect in chromosome unpairing (fewer FISH spots), which is strongly suppressed when a CK1α ^8B12/+ heterozygous mutation (E) is introduced into this background (dispersal of FISH spots). CK1α ^EP1555/+ shows similar but weaker suppression when introduced into condensin II double heterozygous background (F). (G) Histogram showing the average number of FISH spots for each probe in stage 10 nurse cells (n = 15–26 nurse cells per genotype). Error bars indicate SEM. *p-value < 2.2x10⁻¹⁰ (calculated by using students’ t-test in MS Excel). (A-F) Maximum projection image of multiple z-slices. Scale, 10μm.

Fig 6. CK1α limits chromatin bound Cap-H2 levels and activity in Drosophila salivary glands.

(A-B) Micrographs of salivary glands from larvae of wild-type (Oregon-R) and CK1α heterozygous mutants (CK1α^8B12 /+ and CK1α ^EP1555 /+) squashed, immunostained for Cap-H2 (green) and counterstained for DNA (DAPI, magenta) on the same glass microscope slide. (C) Micrographs of salivary glands from Drosophila larvae of control (43B>Gal4) and larvae expressing RNAi to CK1α (CK1α^RNAi) squashed, immunostained for Cap-H2 (green) and counterstained for DNA (DAPI, magenta) on the same glass microscope slide. Short and long camera exposures were used in order to show that chromatin bound Cap-H2 levels are increased in both CK1α mutants and CK1α RNAi depleted cells. (C) CK1α RNAi resulted in loss of polytene banding and suggests that chromosomes are unpaired (see Fig. 4). (D) Histogram showing quantitation of fluorescence intensity of Cap-H2 bound to squashed salivary gland chromosomes. CK1α reduction via mutation or RNAi results in significant enrichment of Cap-H2 over same slide controls, assessed by comparing normalized Cap-H2 fluorescence (gray value) intensities (see methods for details); (n = 5–6 fields of squashed chromosomes per genotype). Images shown were not used for analysis. Images used for fluorescence intensity quantitation were acquired such that pixel saturation was minimized. Error bars indicate SEM. p-value = * < 0.05, ** < 0.0001 (calculated by using students’ t-test in Microscoft excel). Statistical comparisons are between mutant or RNAi depleted cells and their control on the same slide (mutant control = OregonR; RNAi control = 43B>Gal4). Maximum projection images of multiple z-slices are shown for all panels. Scale, 20μm in all panels.

Fig 7. Cap-H2 degradation is CK1α dependent.

(A) Whole cell lysates from RNAi treated S2 stable line expressing an inducible Cap-H2-EGFP were immunoblotted for the indicated proteins. Both Slimb and CK1α depletion stabilizes Cap-H2-EGFP. Armadillo is targeted for destruction by Slimb and CK1α and is used to confirm knockdown of CK1α and Slimb. (B) Kc cells were transiently transfected with a plasmid containing an inducible Cap-H2-EGFP, and then treated with either DMSO (control) or 80μM of CK1 inhibitor (D4476) for 8 hours, and lysates were immunoblotted for the indicated proteins. Anti-alpha-Tubulin was used as protein loading control in panels A and B. (C) Amino acid sequence of C-terminal 23 amino acids of Drosophila Cap-H2. Consensus Slimb binding domain DSGISS highlighted in grey. This contains three S-X-X-S motifs, where the underlined serines denote potential priming sites and bold serines indicate potential CK1α phosphorylation sites and X can be any amino acid. (D) Cellular fractions of RNAi treated Kc stable line expressing an inducible Cap-H2-EGFP immunoblotted for the indicated proteins. CK1α depletion stabilizes Cap-H2-EGFP in both the whole cell lysate (WCL, top immunoblot) and the chromatin bound (P3, bottom immunoblot) cellular fractions. Anti-alpha-Tubulin was used as loading control for WCL, Anti-Histone-H3 was used as loading control for P3, and Anti-Lamin-Dm0 was used as loading control for both WCL and P3. (E) Fold enrichment of Cap-H2-EGFP protein levels from (D). CK1α depletion stabilizes whole cell (WCL) and chromatin bound (P3) Cap-H2-EGFP protein levels when normalized to Lamin. Calculated using densitometry in ImageJ. Error bars indicate SEM. p-value = * < 0.05, (calculated by using students’ t-test in Microscoft excel). (n = 4 biological replicates). (F) Immunoprecipitations and immunoblots from RNAi treated Kc cells, transiently transfected with inducible EGFP as a negative control and Kc cells stably expressing an inducible Cap-H2-EGFP. Anti-Cap-H2-EGFP immunoprecipitates Slimb in control, CK1α, and PKA/GSK3β/CK1α depleted cells expressing Cap-H2-EGFP. GFP tag only transfected cells did not immunoprecipitate Slimb. Anti-armadillo was used to verify CK1α and PKA/GSK3β/CK1α depletion and anti-Lamin-Dm0 and anti-alpha-tubulin was used as protein loading control. (G) Immunoprecipitations and immunoblots from RNAi trated S2 cells, transiently co-transfected with inducible Cap-H2-EGFP and inducible 3X-Flag-Ubiquitin. Anti-Cap-H2-EGFP immunoprecipitates 3X-Flag-Ub in both control and CK1α depleted cells expressing Cap-H2-EGFP.

Fig 8. Two-step model of Cap-H2 eviction from chromatin.

(A) SMC dimers (red and orange) are associated with DNA (blue line) in the “open” conformation, resulting in “relaxed” chromatin. There may be regions where Cap-H2 (green oval) is bound with or without Slimb (magenta square) nearby. (B) ATP binding to SMC dimers promotes the “closing” of the SMC subunits, resulting in the “compaction” of chromatin. Cap-H2 binding to SMC dimers traps a closed SMC2/4 dimer and inhibits ATP hydrolysis. Inhibition of ATP hydrolysis by Cap-H2 is thought to maintain SMC dimers in the “closed” conformation and promotes axial chromosome compaction. (C) CK1α negatively regulates Cap-H2. This may occur by directly phosphorylating Cap-H2 or indirectly by CK1α targeting an unknown regulator of Cap-H2. (D) CK1α promotes Cap-H2 removal from chromatin and degradation. Note that Slimb interaction with Cap-H2 is not likely to be dependent on CK1α activity. In chromatin regions where Slimb and CK1α are absent (condensin II complex on the right), Cap-H2 could remain on the chromatin, as it avoids Slimb and CK1α mediated degradation. (E) Cap-H2 removal from chromatin results in the opening of the SMC dimer, resulting in the “relaxation” of chromatin (left side), whereas on chromatin where Slimb is absent, Cap-H2 remains chromatin bound and the chromatin remains axially compact (right side).