Autophosphorylation of conserved yeast and human casein kinase 1 isozymes regulates Elongator-dependent tRNA modifications

Abstract

Casein kinase 1 (CK1) family members are crucial for ER-Golgi trafficking, calcium signalling, DNA repair, transfer RNA (tRNA) modifications, and circadian rhythmicity. Whether and how substrate interactions and kinase autophosphorylation contribute to CK1 plasticity remains largely unknown. Here, we undertake a comprehensive phylogenetic, cellular, and molecular characterization of budding yeast CK1 Hrr25 and identify human CK1 epsilon (CK1ϵ) as its ortholog. We analyse the effect of Hrr25 depletion and catalytically inactive mutants in vivo and show that perturbations in CK1 activity lead to stress-induced growth defects, morphological abnormalities, and loss of Elongator-dependent tRNA modification. We use purified Hrr25 protein to identify distinct autophosphorylation patterns and phospho-sites on several physiological substrates in vitro and find only human isozyme CK1ϵ can replace yeast Hrr25 functions essential for tRNA modification and cell proliferation in vivo. Furthermore, we demonstrate that human and yeast CK1 orthologs share conserved autophosphorylation sites within the kinase domains, which regulate their activities and mutually exclusive interactions with Elongator subunit Elp1 and Sit4, a phosphatase antagonist of Hrr25. Thus, autophosphorylation controls CK1 activity and regulates the tRNA modification pathway. Our data offer mechanistic insights into regulatory roles of CK1 that are conserved between yeast and human cells and reveal a complex phosphorylation network behind CK1 plasticity.

Graphical Abstract

Figure 1.

Hrr25 is crucial for yeast cell morphogenesis and viability. (A) Scheme of Hrr25 domain architecture with emphasis on the ADP bound catalytical triad (PDB 4xhg). (B) Phase contrast microscopy images and morphology of exponentially growing yeast strains with indicated genetic backgrounds (bar: 20 μm). (C) The Elongator complex, which catalyses U₃₄ modification of tRNA (xcm⁵U), is regulated by Hrr25 kinase and Sit4 phosphatase. (D) LC-MS/MS profiles of U₃₄ modification states from the indicated strain backgrounds. Modified nucleoside signal was normalized using the total uridine (U) content. ncm⁵U, mcm⁵U, and mcm⁵s²U signals were normalized against WT, the s²U signal against elp3Δ. The statistical significance was tested with a two-tailed t-test (***P <.001, **P <.01, *P <.05, ns > 0.05). EV, empty vector; ns, not significant. (E) Gene shuffle scheme based on 5-FOA plasmid chase-out assay. A lethal hrr25Δ allele is rescued by HRR25 on a URA3 plasmid that can be selected against by 5-FOA if another plasmid (HRR25::LEU2) is present as substitute. (F) Gene shuffles show Hrr25 catalytic activity is essential for hrr25Δ rescue and cell viability. Ten-fold serial cell dilutions of indicated strains were replica spotted on control (Ctrl) medium without or with 5-FOA and cultivated for 3 days at 30°C.

Figure 2.

Mapping of p-sites on Atg19, Atg34, and Elp123. (A) Representative size exclusion chromatography profile and SDS–PAG E gel of FL Hrr25 purified from Hi5 cells. Ultraviolet signal is scaled to 1. (B) Hrr25 FL is capable of in vitro phosphorylation. All samples were incubated for 1 h at 30°C. Conditions without ATP were used as control. Gel bands were excised from the gel and subjected to MS for identification of p-sites. (C) Domain representation of Atg19 and Atg34 with p-sites highlighted as red lines. (D) Subunit domain representation of the Elp123 subcomplex. Hrr25-dependent phosphorylations are plotted as red lines. (E) Structural overview of p-sites on the Elp123 subcomplex identified in this study. Colouring as in panel (D). (F) In vitro GST pull-down assay. The N-terminus of Elp1 is the main docking site of Hrr25 on the Elongator complex. GST-Elp1 constructs were used as a bait, free GST was used as a control.

Figure 3.

CK1ϵ is the human homolog of Hrr25. (A) Multiple sequence alignment of human and yeast CK1. All isozymes share a highly conserved kinase domain (violet) of comparable size and location. (B) A phylogenetic tree of human and yeast CK1 modified after [87] and [88]. Branch length is proportional to evolutionary distance. (C, D) 5-FOA chase-out assays. Solely, human CK1ϵ complements lack of Hrr25 function and CK1ϵ catalytic inactivation (K38R) abolishes hrr25 mutant complementation in vivo. (E) LC-MS/MS profiles to demonstrate rescue of U₃₄ modification defects in hrr25 mutant (E52D) by CK1ϵ. Signals for modified nucleoside were standardized with total uridine (U) content. ncm⁵U, mcm⁵U, and mcm⁵s²U signals were normalized against WT, s²U signals against elp3Δ. Statistical significance is based on two-tailed t-test with P-values as in Fig. 1. EV, empty vector; ns, not significant.

Figure 4.

Hrr25 autokinase activity. (A) Hrr25 autophosphorylation in vivo. WB analysis of yeast lysates obtained from indicated strains and subjected to Phos-tag™ (left panel) or conventional SDS–PAGE (right panel). (B) Hrr25 autophosphorylation in vitro. Purified FL Hrr25 and Hrr25_1-394 and their respective K38R mutants were incubated with or without ATP for 1 h at 30°C and subjected to SDS–PAGE followed by Coomassie staining. (C) Identification of in vitro autophosphorylation sites of Hrr25 by MS. Structural overview of detected p-sites (red spheres) plotted on the Hrr25_1-394 structure (PDB 5cyz). (D) p-site mutagenesis shows Hrr25 autophosphorylation mediates zymocin sensitivity and SUP4 read-through. Serial cell dilutions of the indicated strains were replica spotted on medium lacking (control, Ctrl) zymocin or containing 100% (v/v) (+ Zymocin). For the SUP4 assay, that monitors ade2-1^ochre read-through, strains were replica spotted on medium containing (Ctrl) or lacking adenine (−Ade). Cultivation was for 3–5 days at 30°C, respectively. (E) LC-MS/MS reveals hrr25 phosphoablative mutations trigger progressive U₃₄ tRNA modification defects. For modification measurements and statistical significance analysis, see Figs 1 and 3. EV, empty vector.

Figure 5.

Conservation of Hrr25 p-sites in CK1ϵ. (A) Structural alignment of Hrr25 and CK1ϵ identifies conserved autophosphorylation sites. Structures of Hrr25 (PDB 5cyz, 1–394; purple, kinase domain; green, central domain) and CK1ϵ (PDB 4hni, 1–294; kinase domain, blue) were aligned using PyMOL. Positions of conserved p-sites in Hrr25 (purple) and CK1ϵ (blue) are indicated by red spheres. (B) Cross-species gene shuffle assay. The capacity of indicated single and multiple CK1ϵ p-site mutations to rescue the inviability of hrr25Δ mutant was tested using the 5-FOA chase-out assay as detailed in Figs 1 and 3. (C) LC-MS/MS shows lack of U₃₄ modification caused by hrr25 mutant (E52D) can be partially substituted by CK1ϵ p-site mutants. For modification measurements, ncm⁵U, mcm⁵U, mcm⁵s²U, and s²U normalization against WT or elp3Δ and statistical significance analysis, see Figs 1 and 3. EV, empty vector; ns, not significant. (D) Progressive loss of phenotypic rescue of hrr25 mutant (E52D) by CK1ϵ p-site mutants. Serial dilutions of the indicated strains were grown under conditions to assay rescue of thermosensitivity at 37°C, zymocin resistance and loss of SUP4 read-through (for details, see Fig. 1 and Supplementary Figs S1 and S3, respectively).

Figure 6.

Hrr25 autophosphorylation defines kinase interactions with Elp1 and Sit4. (A, B) GST pull-down assay addressing Hrr25–Sit4 (A) and Hrr25–Elp1 (B) interactions. GST-Sit4 and GST-Elp1 were used as a bait, free GST served as a negative control. Interaction was visualized with SDS–PAGE (top) and WB with Hrr25-specific antibodies (bottom). (C) GST pull-down assay showing interaction between Sit4 and Elp1 is not mutually exclusive with Hrr25 binding. (D) trans-autophosphorylation of Hrr25 does not exert strong influence on Hrr25–Sit4 interaction, but significantly impairs Elp1 interaction. WB analysis of GST pull-down (top membrane) and input (bottom membrane). The 5% of catalytically active Hrr25 WT provided trans-autophosphorylation to 95% of Hrr25 K38R incapable of autophosphorylation. Hence, the majority of readouts corresponds to the effect of autophosphorylation in trans. (E) AlphaFold model of Hrr25–Sit4 complex coloured according to confidence. (F) Autophosphorylation positively regulates kinase interaction with Elp1. Proteins were isolated from indicated strains and subjected to anti-c-Myc IP followed by WB analysis of either input control or coprecipitated proteins. To assure nonspecific binding to the beads, a control without protein was included Immunodetection of Myc-tagged Elp1 and Hrr25 employed anti-c-Myc or anti-Hrr25 antibodies. Cdc19 detection served as protein loading control.

Figure 7.

Differential roles of CK1 Hrr25 cis- and trans-autophosphorylation for the Elongator pathway in yeast.cis-autophosphorylation releases the CK1 from a Hrr25–Sit4 complex characterized in this study. Autophosphorylation of Hrr25 in trans decreases Hrr25 capacity to bind Elp1, the main scaffolding subunit of the Elongator complex. Upon binding, Hrr25 provides multiple phosphorylation of the Elp123 subcomplex, enabling a faithful tRNA modification process. In yeast, lack of Hrr25 kinase activity (K38R) results in pleiotropic phenotypes including major perturbations in cell morphology.