A conserved chronobiological complex times C. elegans development

Abstract

The mammalian PAS-domain protein PERIOD (PER) and its C. elegans orthologue LIN-42 have been proposed to constitute an evolutionary link between two distinct, circadian and developmental, timing systems. However, while the function of PER in animal circadian rhythms is well understood molecularly and mechanistically, this is not true for LIN-42's function in timing rhythmic development. Here, using targeted deletions, we find that the LIN-42 PAS domains are dispensable for the protein's function in timing molts. Instead, we observe arrhythmic molts upon deletion of a distinct sequence element, conserved with PER. We show that this element, designated CK1δ-binding domain (CK1BD), mediates stable binding to KIN-20, the C. elegans CK1δ/ε orthologue. We demonstrate that CK1δ phosphorylates LIN-42 and define two conserved helical motifs in the CK1BD, CK1BD-A and CK1BD-B, that have distinct roles in controlling CK1δ-binding and kinase activity in vitro. KIN-20 and the LIN-42 CK1BD are required for proper molting timing in vivo, and loss of LIN-42 binding changes KIN-20 subcellular localization. The interactions mirror the central role of a stable circadian PER-CK1 complex in setting a robust ~24-hour period. Hence, our results establish LIN-42/PER - KIN-20/CK1δ/ε as a functionally conserved signaling module of two distinct chronobiological systems.

Figure 1.. lin-42ΔPAS mutations cause mild heterochronic phenotypes while lin-42ΔCK1BD mutations cause asynchronous molting

(A) Schematic depiction of the lin-42 genomic locus. The gene structure of the lin-42 a, b, and c isoforms are shown. The location of sequences encoding putative PAS-A, PAS-B, SYQ, and LT sequence motifs are indicated. Published lin-42 deletion alleles n1089 and ok2385 are indicated in dark and light grey, respectively, newly generated targeted ΔPAS and ΔCK1BD deletions in navy and teal, respectively. (B) Heatmaps showing trend-corrected luminescence traces from the indicated genotype. Each horizontal line represents one animal. Traces are sorted by entry into the first molt. Darker color indicates low luminescence signal and corresponds to the molts. (C) Bar plot quantifying the percentage of animals of the indicated genotype with precocious complete or partial alae at the L3-L4 molt. (D) Bar plot quantifying the percentage of animals of the indicated genotype that exhibited egg-laying defects as determined by the presence of hatched larvae in the animal. (E) Bar plot depicting the average number of live progeny from hermaphrodites of the indicated genotype. (F) Bar plot quantifying the percentage of animals of the indicated genotype that arrested or died as larvae. (C-F) Statistical significance was determined using an ordinary one-way ANOVA. p<0.05 was considered statistically significant. **** indicates < 0.0001. Error bars in C-F represent standard deviation.

Figure 2. LIN-42 and KIN-20 interact in vivo

(A) Volcano blot comparing protein enrichments in 3xFLAG::LIN-42b+c and 3xFLAG::SART-3 (control) immunoprecipitations, determined by mass spectrometry. (B) mRNA expression profile for lin-42 (dark blue) and kin-20 (light blue) mRNAs throughout larval development. Data Source: Meeuse et al., 2020. (C) Western blot with extracts from 3xflag::lin-42b+c; 3xflag-ha::kin-20 (HW3479) larvae collected hourly for 11 hours, starting at 13 h after plating synchronized L1 stage animals on food at 25°C. Top part probed with anti-FLAG-HRP (1:1,000), lower part with anti-actin-1 (1:7,500). Arrows indicate bands for LIN-42b, KIN-20 and ACT-1 (see Fig. S5A). (D) Anti-HA pulldown from 3xflag::lin-42b+c; 3xflag-ha::kin-20 animals. Blot probed with anti-FLAG (1:1,000).

Figure 3. The LIN-42 SYQ and LT domains constitute a functional CK1-binding domain

(A) Schematic representing PER2 and LIN-42 protein domains. Protein constructs used in this study, CK1BD + Tail and CK1BD, are indicated i. CK1BD, Casein Kinase 1-Binding Domain; CBD, CRY-binding domain. (B) Representative pulldown assay of human CK1 and biotinylated LIN-42 CK1BD + Tail proteins using the indicated protein variants. (C) Values for K_D from kinetic analysis of BLI data (D-H) based on a 2:1 heterogeneous ligand binding model and global analysis (Octet). Mean ± SEM, ordinary one-way ANOVA. p<0.05 was considered statistically significant. **** indicates p< 0.0001. (D-H) Bio-layer interferometry (BLI) data for indicated LIN-42 protein binding to immobilized, biotinylated CK1. Inset values represent the concentrations of LIN-42 for individual binding reactions. Model fit to association and dissociation over time is represented by black lines. Data shown from one representative experiment of n ≥ 3 assays.

Figure 4. The extended C-terminus of LIN-42 is phosphorylated by CK1

(A) Representative ³²P-radiolabeled ATP enzymatic assay of CK1 activity on LIN-42 CK1BD + Tail and CK1BD wild-type and indicated domain deletion mutants. (B) Quantification of band intensity depicted in (A) and replicates. Model fit to Michaelis- Menten, n = 3. (C) Fold-change quantification of band intensity from panels A & B compared to CK1 autophosphorylation at 120 minutes. Ordinary one-way ANOVA. p<0.05 was considered statistically significant. ****p < 0.0001, n ≥ 3, mean ± SEM. (D) Representative ADP-Glo enzymatic assay comparing activity of CK1 on LIN-42 CK1BD + Tail wild-type (maroon), CK1BD-ΔA (orange), CK1BD-ΔB (yellow), and CK1BD-ΔA/B (green). (E) Values of k_cat/K_m calculated from (D). n=3, mean ± SEM, Ordinary one-way ANOVA. p<0.05 was considered statistically significant. * indicates p<0.03 and ** indicates p<0.002. (F) LIN-42 phosphorylation sites identified via mass spectrometry from in vivo samples from whole worm lysate collected hourly from synchronized L1 animals from 14 hrs (early L2) to 25 hrs (late L3) after plating at 25°C (top arrows) and in vitro kinase reactions of CK1 activity on LIN-42 CK1BD + Tail (bottom arrows). Serine, threonine, and tyrosine residues are indicated via vertical blue, purple, and pink ticks, respectively, along the CK1BD + Tail construct schematic.

Figure 5. kin-20 null and catalytically dead alleles exhibit larval arrest and asynchronous molting similar to lin-42(ΔCK1BD) mutants

(A) Heatmaps showing trend-corrected luminescence traces from the indicated genotype. Each horizontal line represents one animal. Traces are sorted by entry into the first molt. Darker color indicates low luminescence signal and corresponds to the molts. (B) Bar plot depicting the average number of live progeny from hermaphrodites of the indicated genotype. (C) Bar plot quantifying the percentage of animals of the indicated genotype that exhibited egg-laying defects as determined by the presence of hatched larvae in the animal. (D) Bar plot quantifying the percentage of animals of the indicated genotype with precocious complete or partial alae at the L3-L4 molt. (E) Bar plot quantifying the percentage of animals of the indicated genotype that arrested or died as larvae. (B-E) Statistical significance was determined using ordinary one-way ANOVA. p<0.05 was considered statistically significant. ** indicates p<0.0001****, indicates p<0.0001. Error bars in B-E represent standard deviation.

Figure 6.. The LIN-42 CKBD is required for KIN-20 nuclear localization.

(A) Confocal images of splitgfp::kin-20 animals staged according to vulva morphology (Mok et al., 2015) from late L3 to adult. Arrows indicate seam cell nuclear (white) and cytoplasmic (blue) localization; white arrowheads indicate nuclear hyp7 localization. Scale bar=20 μm. (B) Confocal images of gfp::lin-42b+c; wrmscarlet::kin-20 staged according to vulva morphology. An arrow indicates seam cell, an arrowhead hyp7 nuclear localization. Scale bar=20 μm. (C) Confocal images of splitgfp::kin-20 in wild type and lin-42ΔCK1BD mutant animals during early - mid L4 stage based on the vulvae morphology. Arrows indicate seam cell nuclear (white) and cytoplasmic (blue) localization, white arrowheads indicate hyp7 nuclear localization. Scale bar=20 μm. (D) Quantification of GFP::KIN-20 signal in hyp7 (nucleus), seam cells (nucleus) and hyp7 (cytoplasm) of wild-type (blue) and lin-42ΔCK1BD (orange) animals. Each condition with n=9 mid-L4 stage worms.