Functional characterization of the human Cdk10/Cyclin Q complex

Abstract

Cyclin-dependent kinases (CDKs) are key players in cell cycle regulation and transcription. The CDK-family member Cdk10 is important for neural development and can act as a tumour suppressor, but the underlying molecular mechanisms are largely unknown. Here, we provide an in-depth analysis of Cdk10 substrate specificity and function. Using recombinant Cdk10/CycQ protein complexes, we characterize RNA pol II CTD, c-MYC and RB1 as in vitro protein substrates. Using an analogue-sensitive mutant kinase, we identify 89 different Cdk10 phosphosites in HEK cells originating from 66 different proteins. Among these, proteins involved in cell cycle, translation, stress response, growth signalling, as well as rRNA, and mRNA transcriptional regulation, are found. Of a set of pan-selective CDK- and Cdk9-specific inhibitors tested, all inhibited Cdk10/CycQ at least five times weaker than their proposed target kinases. We also identify Cdk10 as an in vitro substrate of Cdk1 and Cdk5 at multiple sites, allowing for a potential cross-talk between these CDKs. With this functional characterization, Cdk10 adopts a hybrid position in both cell cycle and transcriptional regulation.

Keywords: CDK10; Cyclin M; Cyclin Q; RNA polymerase II; cell cycle; transcription.

Figure 1.

Cdk10 shares substrates with cell cycle and transcriptional CDKs. (a) Domain architectures of human Cdk10 (UniProt accession number Q15131) and human Cyclin Q (Q8N1B3). M117—gatekeeper methionine, T133—phosphorylation site, results in increased Cdk10 degradation, D163—catalytic centre, T196—conserved T-loop threonine phosphorylation is required for kinase activation. (b) Coomassie-stained SDS–PAGE analysis of 1 µg recombinant human Cdk10/CycQ. (c) Mass spectrometry-based detection of phosphorylation within recombinant human Cdk10 expressed in Sf9 insect cells. (d) Radiometric kinase assay; 0.2 µM Cdk10/CycQ, 25 µM GST-RB1 (761–928), 10 µM GST-CTD_[52], 25 µM His-c-MYC (17–167) or 25 µM GST-SRSF7 (1–248) were incubated with 1 mM ATP containing 0.35 µCi γ[³²P]-ATP for 30 min at 30°C. Data represent mean ± s.d. of duplicate measurements; cpm, counts per minute. (e) SDS–PAGE analysis (Coomassie) of 2 µg recombinant human MBP-Cdk10/His-CycQ, MBP-Cdk10T196A/CycQ and MBP-Cdk10T196E/CycQ. (f) Radiometric kinase assay of Cdk10 T-loop mutants. The experiments were performed similarly as in (d).

Figure 2.

RNA pol II CTD phosphorylation by Cdk10. (a) Time course of a radiometric kinase assay. Incubation of 10 µM GST-CTD_[52], 1 mM ATP containing 0.35 µCi γ[³²P]-ATP with or without 0.2 µM Cdk10/CycQ (mean ± s.d. of duplicate measurements; cpm, counts per minute). (b) SDS–PAGE analysis (Coomassie) of 2 µg (top panel) and immunoblot analysis (lower panels) of 100 ng GST-CTD_[52] after phosphorylation by Cdk10/CycQ. (c) Radiometric kinase assay (top panel) and immunoblot analysis (lower panels) of GST-CTD_[52] after incubation with the respective kinase for 30 min (radiometric assay: mean ± s.d. of duplicate measurements; immunoblot: representative blots of two independent experiments). (d) Radiometric kinase assay. Three hepta-repeats with either no or continuous phosphorylation marks or serine to lysine substitution were provided as substrate (mean ± s.d. of triplicate measurements). (e) Radiometric kinase assay. Time-dependent phosphorylation of non-modified CTD_[3] and K7-CTD_[3] peptides (mean ± s.d. of triplicate measurements).

Figure 3.

Inhibition of Cdk10 by small molecules. (a) Chemical structures of the inhibitors tested against Cdk10/CycQ. (b) Radiometric dose–response measurements at 0.2 µM kinase, 50 µM c-MYC (17-167), 1 mM ATP containing 0.35 µCi γ[³²P]-ATP and indicated concentration of inhibitor (mean ± s.d. of duplicate measurements).

Figure 4.

A chemical genetic screen identifies novel protein substrates of Cdk10/CycQ. (a) Schematic of the identification of Cdk10/CycQ substrates. Samples were thio-phosphorylated by the addition of Cdk10^as/CycQ and PhEt-ATPγS. As control, lysates were incubated with Cdk10^wt/CycQ. Proteins were digested by trypsin and thio-phosphorylated peptides were captured with iodo-acetyl beads. After elution, peptides were analysed by mass spectrometry. In parallel, an aliquot of the sample was alkylated with para-nitrobenzylmesylate (PNBM) and analysed by immunoblot to monitor labelling efficiency. (b) SDS–PAGE analysis (Coomassie) of 2 µg wild-type Cdk10/CycQ and analogue-sensitive Cdk10M117G/CycQ complexes. (c) Immunoblot analysis of thio-phosphorylated proteins in HeLa cell lysate. (d) Consensus sequence of the Cdk10 phosphosites deduced from the dataset. Upper panel: all phosphosites; lower panel: restricted to phosphosites followed by a proline. (e) Comparison of (S/T)P site protein substrates of the kinases Cdk10/CycQ (this paper), Cdk1/CycB1 [42] and Cdk9/CycT1 [43].

Figure 5.

In vitro validation of Cdk10 substrates. (a) SDS–PAGE analysis (Coomassie) of UBAP2L (355–652), NPM1 (isoform2), MBP-HSF1 (271–384), LIMK2 (149–242), SRRT (409–602), GST-TRIM28 (1–835), GST-U2AF2 (2-475, isoform2), GST-HDGF (1–240), 2 µg each (arrowheads); kDa, kilodalton; M, marker. (b) Radiometric kinase assay. Substrates at a concentration of 25 µM (UBAP2L, NPM1, SRRT, LIMK2, U2, HDGF) 15 µM (HSF1), 10 µM (TRIM28) were incubated with 0.2 µM Cdk10/CycQ for 1 h at 30°C. U2 and HDGF were measured separately from the other samples, indicated by the dashed line (mean ± s.d. of duplicate measurement). (c) Radiometric kinase assay. The same as in (b), except that the background of the w/o kinase control was subtracted for reasons of clarity (mean ± s.d. of duplicate measurement).

Figure 6.

CDK cross-talk to regulate Cdk10 activity. (a) Immunoblot analysis of thio-phosphorylation (upper panel) or MBP (lower panel) of 1 µg MBP-Cdk10 after incubation with 0.2 µM of the indicated kinase for 60 min. Asterisks indicate GST-CycB1/GST-Cdk1 and GST-CycA2/GST-Cdk2 autophosphorylation. (b) Immunoblot analysis of thio-phosphorylation of 1 µg MBP-Cdk10 and MBP-Cdk10T196A after incubation with 0.05 µM Cdk1/CycB1 or Cdk5/p35NCK. Below, Ponceau S staining of the membrane. (c) Results of the mass spectrometric determination of phosphorylation sites of MBP-Cdk10 after 20 min incubation without any additional kinase or after incubation with Cdk1/CycB1, Cdk5/p35 or Cdk10/CycQ. The asterisk indicates phosphosites that were identified only after longer incubation times of 40 or 120 min.