CDK-dependent Hsp70 Phosphorylation controls G1 cyclin abundance and cell-cycle progression

Abstract

In budding yeast, the essential functions of Hsp70 chaperones Ssa1-4 are regulated through expression level, isoform specificity, and cochaperone activity. Suggesting a novel regulatory paradigm, we find that phosphorylation of Ssa1 T36 within a cyclin-dependent kinase (CDK) consensus site conserved among Hsp70 proteins alters cochaperone and client interactions. T36 phosphorylation triggers displacement of Ydj1, allowing Ssa1 to bind the G1 cyclin Cln3 and promote its degradation. The stress CDK Pho85 phosphorylates T36 upon nitrogen starvation or pheromone stimulation, destabilizing Cln3 to delay onset of S phase. In turn, the mitotic CDK Cdk1 phosphorylates T36 to block Cln3 accumulation in G2/M. Suggesting broad conservation from yeast to human, CDK-dependent phosphorylation of Hsc70 T38 similarly regulates Cyclin D1 binding and stability. These results establish an active role for Hsp70 chaperones as signal transducers mediating growth control of G1 cyclin abundance and activity.

Graphical abstract

Figure 1

Mutation of a Conserved Threonine in the N Terminal of Ssa1 Affects Chaperone Function (A) Alignment of Hsp70 proteins from yeast and human cells. A cyclin-dependent kinase consensus (S/TP) site at T36 in Ssa1 is conserved across the yeast Hsp70s and among mammalian Hsp70 isoforms (including Hsp70 and Hsc70 as shown here). (B) T36 resides adjacent to the ATP-binding pocket on Hsc70. T36 is marked in red on the Hsc70 structure (PDB: 1YUW). NBD is nucleotide-binding domain, SBD is substrate binding domain. (C) Temperature sensitivity of the Ssa1 T36 phosphomutants. Cells expressing wild-type Ssa1 (SSA1) or Ssa1 phosphosite mutants T36A or T36E as their sole Ssa protein were spotted in serial 10-fold dilutions onto YPD media and were incubated for 48 hr at 30°, 37° or 39°C. (D) Mutation of T36 affects nucleotide binding. The intrinsic ATPase activity of WT, T36A and T36E Ssa1 was assayed and the k_cat values (min^-1) are shown as mean ± SD (n = 3). Nucleotide binding of WT, T36A and T36E was calculated by isothermal calorimetry. The K_d (μM) for the Ssa1 proteins are shown as mean ± SD (n = 3). (E) Ssa1 T36E mutation confers a large G1 cell phenotype. Cells grown at 30° C to mid-log phase were analyzed by flow cytometry. Cells in G1 stage of cell cycle were gated based on 1N DNA content, and the size distribution based on Forward Scatter (FSC) of each strain is plotted. (F) T36E cells display delayed G1/S progression. Cells were synchronized in G1 by nitrogen starvation then released into rich media and incubated at 30° C. At 20 min intervals, cells were fixed and assessed for DNA content by flow cytometry. (G) Halo assay of response of SSA1, T36A and T36E to αF pheromone reveals an arrest defect in T36A, expressing the nonphosphorylatable Ssa1 mutant. (H) Ssa1 T36 is phosphorylated in response to αF. His₆-Ssa1 was purified from untreated or αF-treated SSA1 or T36A cells. Ssa1 phosphorylation was assessed with CDK phosphosite-specific anti-Phos-T-P antibody. See also Figure S1.

Figure 2

Phosphorylation Alters the Ssa1 Interactome (A) Scheme for proteomic analysis of Ssa1 interactome. Cells expressing wild-type (SSA1) or T36A mutant His₆-Ssa1 were exposed to αF. Ssa1 complexes were affinity purified and digested. Peptides from Ssa1-T36A interactors were isotopically labeled with ¹⁸O, mixed 1:1 with wild-type Ssa1 interactor peptides, and analyzed by quantitative LC-MS/MS. (B) Functional classification of the Ssa1 interactome. Ssa1 interactors were categorized by cellular function using GO Slim analysis, and plotted by relative enrichment compared to occurrence in the nonessential genome. (C) Chaperones and cochaperones in the Ssa1 interactome. The 18 chaperone/cochaperone proteins identified were analyzed for interactions using STRING and visualized by Cytoscape, grouped by homology and function. Pink nodes represent proteins that decreased interaction with Ssa1 upon T36 phosphorylation. (D) Differential effect of T36 phosphorylation on Sis1 and Ydj1 binding. His₆-Ssa1 was pulled down from WT and T36A cells and probed for Ydj1 or Sis1 coprecipitation. See also Table S1 and Figure S2.

Figure 3

Interaction of T36 Mutants with Cln3 and the G1/S Transcription Pathway (A) Western blotting of cell extracts reveals decreased Cln3 protein levels in T36E cells. A long exposure reveals residual expression in T36E. (B) Mutation of T36 does not alter CLN3 transcription. SSA1, T36A and T36E cells transformed with a CLN3-lacZ reporter were assessed for CLN3 transcription. Each value represents mean ± SD (n = 3). (C) Cln3 lacking PEST domains displays similar abundance in SSA1, T36A and T36E cells, indicating the effect of T36E is mediated by ubiquitin-proteasome degradation. (D) Ssa1 phosphorylation regulates Cln3 binding. HIS₆-Ssa1-WT, T36A and T36E were purified from cells expressing HA-Cln3ΔPEST2. Immunoblotting reveals Cln3 binding induced by nitrogen starvation in SSA1 is absent in T36A and constitutive in T36E. (E) Dissociation of Ydj1 from Ssa1 upon T36 phosphorylation allows Cln3 to bind Ssa1. Ydj1 wild-type or deletion mutant cells expressing either His₆-Ssa1 WT or T36A, along with stabilized, PEST mutant Cln3 were treated with 15 μg/ml nocodazole for 3 hr. Ssa1 was pulled down to determine T36 phosphorylation by anti-Phos-T-P blotting and assess Ssa1-Cln3 interaction by coprecipitation of HA-Cln3ΔPEST2. (F) Schematic of regulation of G1/S transcription by Cdk1-Cln3 and Bck2. (G) Overexpression of Whi5 is lethal in T36E cells. SSA1, T36A or T36E cells transformed with either GAL1-WHI5-FLAG or empty vector pRS316 were spotted in serial 10-fold dilutions on either glucose or galactose medium and incubated for 48 hr at 30°C. (H) Bck2 is essential for cell viability in T36E cells. Both ssa1-4Δ and ssa1-4Δ bck2Δ cells were transformed with LEU2-marked plasmids carrying SSA1, ssa1-T36E or ssa1-T36E or empty vector control (V) and then spotted in serial 10-fold dilutions on either YPD or 5-FOA media to evict a URA3-marked wild-type SSA1 plasmid, then incubated for 48 hr at 30°C. (I) T36E cells are defective in CLN2 expression. SSA1, T36A and T36E cells, bearing deletions of CLN3 or WHI5 as indicated, were transformed with a CLN2-lacZ plasmid and CLN2 expression measured by β-galactosidase assay. Each value represents the mean ± SD (n = 3). See also Figure S3.

Figure 4

Ssa1 Is Phosphorylated by CDKs Cdk1 and Pho85 (A) Ssa1 is phosphorylated in G2 or in response to environmental conditions. Cells expressing HIS₆-Ssa1 were either sorted into G1 or G2/M populations via flow cytometry or subjected to nitrogen starvation (-N) or αF stimulation (α). HIS₆-Ssa1 was purified and T36 phosphorylation was assessed with anti-Phos-T-P antibody. (B) Cdk1 and Pho85 phosphorylate Ssa1 in vitro. Recombinant His₆-Ssa1 was used as a substrate in an in vitro kinase assay with HA-Cdk1 or HA-Pho85 purified from cells treated with nocodazole (Noc), nitrogen starvation (-N) or αF (α). Ssa1 phosphorylation was assessed with anti-Phos-T-P antibody. (C) Cdk1 phosphorylates Ssa1 in vivo. Cdk-as1 cells expressing His₆-Ssa1 (WT or T36A) were treated with 15 μg/ml nocodazole ± 10 μM 1-NM-PP1 for 3 hr. His₆-Ssa1 was purified and T36 phosphorylation was assessed. (D) Phosphorylation of Cln3 by Cdk1 is independent of Ssa1 T36. Recombinant His₆-Cln3 was used as a substrate in an in vitro kinase assay with HA-Cdk1 purified from SSA1, T36A or T36E cells treated with nocodazole. Cln3 phosphorylation was assessed with anti-Phos-T-P antibody. (E) Pho85 mediates Ssa1 T36 phosphorylation in nitrogen starvation. SSA1 cells, SSA1 cells bearing a pho85Δ mutation and T36A cells expressing His₆-Ssa1 proteins were grown in media ± nitrogen source for 5 hr. Ssa1 was pulled down to determine T36 phosphorylation. (F) Ssa1 phosphorylation functions downstream of Pho85 with regard to Cln3 degradation. SSA1 cells and pho85Δ mutant cells expressing SSA1, T36A or T36E were assayed for the effect of nitrogen starvation on Cln3 levels, revealing that Pho85 acts via phosphorylation of T36. (G) Pho85 regulates Ssa1-Cln3 interaction during nitrogen starvation. SSA1 cells and pho85Δ mutant cells expressing His₆-SSA1, T36A or T36E and stabilized, PEST mutant Cln3 were subjected to nitrogen starvation. Coprecipitation of HA-Cln3ΔPEST2 with His₆-Ssa1 was assessed. (H) Cln3 is a direct Pho85 kinase substrate. Recombinantly expressed His₆-Cln3 was used as a substrate in an in vitro kinase assay with HA-Pho85 purified from cells treated with nitrogen starvation. Cln3 phosphorylation was assessed with anti-Phos-S/T antibody. See also Figure S4.

Figure 5

Ssa Proteins 1, 2, 3 and 4 Interact with Cln3 and a Specific Complement of Pho85 Cyclins (A) Ssa1-4 bind Cln3 upon nitrogen starvation. Cells expressing His₆-Ssa1, 2, 3, and 4 and stabilized, PEST mutant Cln3 were grown in media ± nitrogen source for 5 hr. Coprecipitation of HA-Cln3ΔPEST2 with His₆-Ssa1 was assessed. (B) Ssa1-4 interact with Pcl proteins in yeast two-hybrid assays. β-galactosidase activity was measured in protein extracts obtained from PJ694a/α cells transformed with the appropriate AD-Pcl and BD-Ssa1 fusions. Data in white bars represent AD-control fusion versus BD-Ssa. The data in grey represent cells transformed with BD-control versus AD-Pcl and data in black represent the BD-Ssa1 versus AD-Pcl interaction. Each value represents the mean ± SD (n = 3). Bars with ^∗ represent significant Ssa-Pcl interactions as compared to control as indicated by brackets (p < 0.01). (C) Ssa1 interacts with Clg1 and Pcl2. Cells expressing His₆-Ssa1 and either HA-Clg1 or HA-Pcl2 were grown in media ± nitrogen source for 5 hr. Ssa1 was pulled down and coprecipitation of Clg1 or Pcl2 with Ssa1 was determined. (D) Interaction of Pho85 with Ssa1 is dependent on Clg1 and Pcl2. WT and clg1Δ pcl2Δ mutant cells expressing His₆-Ssa1 and HA-Pho85 were grown in media ± nitrogen source for 5 hr. Ssa1 was pulled down and coprecipitation of Pho85 was assessed. See also Figure S5.

Figure 6

Phospho-Regulated Chaperone-Mediated Cyclin Destruction Is Conserved in Mammalian Cells (A) Mammalian Hsc70 is phosphorylated on T38, altering Cyclin D1 binding. Hsc70 was pulled down from HEK293 cells previously transfected with HaloTag-Hsc70 WT, T38A, T38E or empty control vector along with stabilized HA-Cyclin D1^T286A. Hsc70 phosphorylation and interaction with Cyclin D1 were determined by western blot. (B) Overexpression of T38E Hsc70 alters Cyclin D1 accumulation and Rb protein phosphorylation. HEK293 cells were transfected with HaloTag-Hsc70 WT, T38A, T38E or empty control vector. After 72 hr, cells were harvested and levels of Cyclin D1 and Ser780 phosphorylated Rb were determined by western blot. (C) Overexpression of T38E Hsc70 alters Rb-mediated transcription. HEK293 cells were cotransfected in triplicate with HaloTag-Hsc70 WT, T38A, T38E or empty control vector and promoter-luciferase constructs to report expression of Cyclin D1 or Cyclin A1 or a negative control. After 72 hr, luciferase secreted into the media was assayed. Normalized luciferase activity is calculated by dividing luciferase-based luminescence by constitutively expressed alkaline phosphatase produced from the same construct. Each value represents the mean ± SD (n = 3). See also Figure S6.

Figure 7

Model for Regulation of the Cell Cycle through Ssa1 Phosphorylation (A) In response to nitrogen starvation or exposure to prolonged αF mating pheromone, Ssa1 is phosphorylated on T36 by Pho85 activated by Pcl cyclins Clg1 and Pcl2. T36 phosphorylation triggers Ydj1 dissociation and promotes binding of Cln3, which may be phosphorylated by Pho85 on the PEST domains. Binding to Ssa1 and phosphorylation promote degradation of Cln3, decreasing G1/S cell cycle transcription and prolonging G1. (B) In G2/M phase, Cdk1 activated by Clb cyclins phosphorylates Ssa1, driving exchange of Cln3 for Ydj1, and targets Cln3 PEST domains. Both processes drive Cln3 degradation, preventing accumulation of Cln3 and resetting the cell for the next G1.

Figure S1

T36A and T36E Cells Lose Clonogenicity at 42°C Compared to SSA1, Related to Figure 1 SSA1, T36A and T36E cells grown at 30°C in YPD media to an OD₆₀₀ of 0.5. Aliquots of cells were incubated at 42°C for the indicated times, were plated onto YPD plates and incubated for 2 days at 30°C. Colonies were counted and percentage cell viability was determined. Data shown are the mean and SD (n = 3).

Figure S2

Interactome Map for Ssa1, Related to Figure 2 Known protein-protein interactions were calculated by processing the data obtained from Ssa1 Mass spectrometry studies through STRING (string-db.org) and are displayed using Cytoscape. Ssa1 is shown in yellow.

Figure S3

Ssa1 T36 Is Phosphorylated in Asynchronous Cells, Related to Figure 4 Cells expressing His₆-tagged Ssa1 were treated with nocodazole or subjected to nitrogen starvation (5 hr). His₆-tagged Ssa1 was purified and T36 phosphorylation was assessed with anti-Phos-T-P antibody.

Figure S4

The CDK Pho85 Mediates Cln3 Degradation in Response to Pheromone, Related to Figure 4 (A) Pho85 mediates Ssa1 T36 phosphorylation in αF-treated cells. WT and pho85Δ mutant cells expressing His₆-tagged Ssa1 were treated with αF. Total Cln3 levels were assessed. Ssa1 was purified and phosphorylation of Ssa1 determined. (B) Pho85 regulates Ssa1-Cln3 interaction in response to mating pheromone. WT and pho85Δ mutant cells expressing His₆-tagged Ssa1 and stabilized, PEST mutant Cln3 were treated with αF and examined for Ssa1-Cln3 interaction.

Figure S5

T36 Phosphorylation Alters Interaction between Ssa1 and a Subset of Pcl Proteins in Yeast Two-Hybrid Assays, Related to Figure 5 β-galactosidase activity was measured in protein extracts obtained from PJ694a/α cells transformed with the appropriate AD-Pcl and BD-Ssa1 fusions. Data shown are the average and SD (n = 3).

Figure S6

Chemical Inhibition of CDKs Results in Decreased Hsc70 T38 Phosphorylation, Related to Figure 6 HEK293T cells transfected with HaloTag-Hsc70 were grown in media lacking or supplemented with 5μM SU9516 for 24 hr. HaloTag-Hsc70 was purified and Hsc70 T38 phosphorylation was assessed with anti-Phos-T-P antibody.