Molecular basis of cyclin-CDK-CKI regulation by reversible binding of an inositol pyrophosphate

Abstract

When Saccharomyces cerevisiae cells are starved of inorganic phosphate, the Pho80-Pho85 cyclin-cyclin-dependent kinase (CDK) is inactivated by the Pho81 CDK inhibitor (CKI). The regulation of Pho80-Pho85 is distinct from previously characterized mechanisms of CDK regulation: the Pho81 CKI is constitutively associated with Pho80-Pho85, and a small-molecule ligand, inositol heptakisphosphate (IP7), is required for kinase inactivation. We investigated the molecular basis of the IP7- and Pho81-dependent Pho80-Pho85 inactivation using electrophoretic mobility shift assays, enzyme kinetics and fluorescence spectroscopy. We found that IP7 interacts noncovalently with Pho80-Pho85-Pho81 and induces additional interactions between Pho81 and Pho80-Pho85 that prevent substrates from accessing the kinase active site. Using synthetic peptides corresponding to Pho81, we define regions of Pho81 responsible for constitutive Pho80-Pho85 binding and IP7-regulated interaction and inhibition. These findings expand our understanding of the mechanisms of cyclin-CDK regulation and of the biochemical mechanisms of IP7 action.

Figure 1

Reversible regulation of Pho80-Pho85-Pho81-MD by IP₇. (a) Structures of myo-D-inositol derivatives used in this study (P and PP denote phosphate and pyrophosphate, respectively). The absolute configuration of 4/6-IP₇ is not completely defined (either 4-IP₇ or 6-IP₇, which are mirror images of each other). (b) SDS-PAGE of Pho80-Pho85-Pho81 (8 μM each) preincubated with 40 μM 4/6-[β-³²P]IP₇ (5 μCi μmol⁻¹). Coomassie staining of the gel (left); autoradiogram of the gel (right). (c,d) Time-dependent (c) and Pho80-Pho85 concentration-dependent (d) IC₅₀ changes. For c, following 15 (●), 30 (○), 60 (▲) and 120 (△) min of incubation with 4/6-IP₇, the kinase activity of Pho80-Pho85-Pho81-MD (1 nM Pho80-Ph85, 450 nM Pho81-MD) was measured by the addition of buffer containing ATP and Pho4 (1 mM and 1 μM, respectively). Average and s.d. (n = 3) are shown. For d, the IC₅₀ value for the inhibition by 4/6-IP₇ (30 min incubation) was measured at different Pho80-Pho85 concentrations (40 (●), 2 (○) and 0.4 nM (▲); data represent mean values ± s.d.; n = 3).

Figure 2

Characterization of IP₇ binding to Pho80-Pho85-Pho81-MD. (a) Native gel EMSA. Pho80-Pho85 (8 μM), Pho81-MD (8 μM) and divalent metal cations (6 mM each; E, EDTA; Mg, MgCl₂; Ca, CaCl₂; Mn, MnCl₂; Zn, ZnSO₄) were mixed and incubated with 0.2 μM 4/6-[β-³²P]IP₇ (5 μCi mmol⁻¹) for 30 min at 30 °C, followed by electrophoresis at 4 °C. All subsequent binding assays were performed in the presence of 6 mM MgCl₂ unless noted otherwise. (b) Isomer selectivity. Pho80-Pho85-Pho81-MD (8 μM each) was incubated with 0.2 μM 5-[β-³²P]IP₇ or 4/6-[β-³²P]IP₇ (30 °C, 30 min) and analyzed as in a. (c) Specificity of the binding. Binding of 4/6-[β-³²P]IP₇ (0.2 μM) to Pho80-Pho85-Pho81-MD (8 μM each) was monitored as in a in the presence of 400 μM competitors. (d) Binding reversibility. After formation of the complex between Pho80-Pho85-Pho81-MD and 4/6-IP₇ as in a, reversibility of the binding was tested by the addition of 20 mM EDTA (E), or by heating (B; 90 °C, 1 min). N, not treated. (e) Dissociation kinetics. After formation of the complex, samples were diluted in 400 μM unlabeled 4/6-IP₇ at 4 or 30 °C, and the dissociation of the complex was monitored by EMSA as a function of the time following the dilution. (f) Association kinetics. After adding 4/6-[β-³²P]IP₇ to Pho80-Pho85-Pho81-MD and incubating at 4 or 30 °C, a portion of the sample was diluted into ice-cold (4 °C) unlabeled 4/6-IP₇ (400 μM) and analyzed by EMSA. For d and e, data were fit to single exponential equations.

Figure 3

Enzyme kinetic analysis of IP₇-mediated Pho80-Pho85-Pho81-MD inactivation. Effect of 4/6-IP₇ on Pho80-Pho85-Pho81-MD–catalyzed Pho4 phosphorylation. Pho4 phosphorylation was measured at different Pho4 concentrations (a,b) or ATP concentrations (c,d) in the absence (a) or presence (c) of 4/6-IP₇, and the rate constants were plotted against substrate concentrations. Plots of K_m, obtained by fitting data shown in a and b to a Michaelis-Menten equation (see Methods), for Pho4 (b) and ATP (d) against the concentration of 4/6-IP₇. All reactions were carried out with 0.5 nM Pho80-Pho85 and 450 nM Pho81-MD as described in the Methods.

Figure 4

Additional interaction between Pho81-MD and Pho80-Pho85. (a) Binding of Lissamine-labeled Pho81-MD (5 nM) to Pho80-Pho85 in the presence of the indicated concentrations of 4/6-IP₇. Protein-binding-dependent lissamine fluorescence quenching was monitored (λ_ex 532 λ_em 580 nm; data represent mean values ± s.d.; n = 3). (b) Inhibition of Pho80-Pho85 (0.5 nM) by excess Pho81-MD at different concentrations of 4/6-IP₇. (c) Additional Pho80-Pho85-Pho81-MD-4/6-IP₇ complex formation in the presence of excess Pho81-MD. EMSA experiments were carried out with 8 μM Pho80-Pho85, 0.2 μM 4/6-[β-³²P]IP₇ and different concentrations of Pho81-MD.

Figure 5

Dissection of Pho81-MD into binding and inhibitory segments. (a) Design of peptide fragments of Pho81-MD. (b,c) Fluorescence polarization binding of Alexa Fluor 488–labeled Pho81-MD (10 nM) segments to Pho80-Pho85 (data represent mean values ± s.d.; n = 3). (b) Binding of S1 (●), S2 (○) and S3 (▲) to Pho80-Pho85 in the absence of 4/6-IP₇. (c) Binding of S1 to Pho80-Pho85 in the presence of 0 (▲), 5 (△) and 200 (●) μM 4/6-IP₇. (d) Effect of Pho81-MD, S1, S2 and S3 on Pho80-Pho85 inhibition by 4/6-IP₇. Experiments were carried out with 1 nM Pho80-Pho85, 400 nM Pho81-MD or shorter segments, 1 μM Pho4, 500 μM ATP, 0.1 μCi μl⁻¹ [γ-³²P]ATP for 5 min (data represent mean values ± s.d.; n = 3). Symbols note Pho80-Pho85 activity in the presence of IP₇ and Pho81-MD (○), S1 (●), S2 (▲) and S3 (△).

Figure 6

Model of Pho80-Pho85 regulation by 4/6-IP₇ and Pho81. (a) The S3 region of Pho81 binds constitutively to Pho80-Pho85 (left); addition of 4/6-IP₇ triggers a structural change that results in the S1 segment of Pho81 occluding interaction of Pho4 with the kinase active site (right), without affecting ATP binding (white molecule). (b) The effect of binding additional molecules of Pho81-MD. When a high concentration Pho81-MD is present, it can bind and block substrate access to the active site. Our data suggest (Fig. 4c) that 4/6-IP₇ can still bind to this higher order complex.