Supporting Information

Files in this Data Supplement:

Supporting XXXXX ?

Fig. 6. Plk1 kinase activation with mitotic extracts. Several controls were performed to ensure Plk1 activation is specific of the mitotic extract. Lane 1, nickel beads treated with mitotic extracts. Lane 2, activated Plk1 nickel beads assayed in the absence of histone H1. Lane 3, activated Plk1 beads assayed in the presence of histone H1. (Upper) The ³²P incorporation autoradiography. (Lower) The histone H1 loaded in a Coomassie-stained SDS/PAGE. Therefore, H1 phosphorylation is specific of Plk1 action.

Fig. 7. Plk1 kinase activity is modified upon peptide binding. (a) Recombinant Plk1 was subjected to an in vitro kinase assay. Purified Plk1 is barely active by itself (Plk1) (Fig. 3a, lane 1), and only when it is preincubated with a mitotic extract from HeLa cells is fully activated (Plk1*) (Fig. 3a, lane 2). Recombinant Plk1 was assayed in the presence of increasing amounts (see Results) of unprimed (b Left) and primed (c Left), Cdc25C peptide. The presence of any of both peptides is not able to elicit Plk1 activity. In contrast, when Plk1 is preactivated with a mitotic extract, Plk1* (b and c Right), the subsequent Cdc25C peptides binding increases Plk1 activity about two times. The same experiment was performed with nonnatural peptides, (d and e), showing a decrease in Plk1 activity after preactivation.

Fig. 8. Ca ribbon representation of the human PBD structure. The protein coordinates for the representation have been taken from the PBD/Cdc25C-P complex, the only one that shows the connecting loop (CL). The secondary structural elements are colored according to the scheme at the bottom: The linker in yellow, Polo Box I (PBI) in green, Polo Box II (PBII) in blue, and the CL in magenta. The PDB is oriented with the target peptide binding cleft on top showing a clear pseudo-two-fold axis between the b-sheets perpendicular to the plane. This pseudo two-fold axis relates PBI and PBII. A first rotation of 180° in the perpendicular axis and a following one in the horizontal axis shows the back side of the structure and a closer view of the CL.

Fig. 9. Stereo diagram of the PBD-binding pocket in complex with the Cdc25C target peptides. The PBD is colored in green and the CdC25-P and Cdc25C peptides are colored in purple and cyan, respectively. Shown is a detailed view of the residues that constitute the PBD-binding pocket and the interactions with the primed and unprimed Cdc25C target peptides

Fig. 10. (a) Superposition of the Ca of the PBD apo structure (green), the PBD/Cdc25C complex (pink), and the PBD/Cdc25C-P complex (blue). The CL is visible only in the case of the PBD/Cdc25C-P complex but not in the apo form or when the unprimed target peptide is bound. (b) Detailed view of the unphosphorylated (pink) and the phosphorylated (yellow) Cdc25C target peptide structures. Clear conformational changes both in Pro-6 and Leu-1 and 2 can be observed. The position of Thr-5 is rather well conserved in both cases despite the phosphorylation state. (c) Comparison of the PBD-Cdc25C-P and the PBD nonnatural peptide structures. Superposition of the structures of the PBD-Cdc25C-P (green) and the PBD nonnatural phosphorylated peptide (pink) complexes.

Fig. 11. Quantitation of EGFP-Plk1 D400-603 FRAP experiment at the centrosomal region in PC3 cells. The half time of recovery (t_1/2 = 6.9 ± 0.2 seconds) and the mobility fraction (M_f = 32%) are very similar to the WT Plk1 and Plk1 H538A/K540M (Fig. 1 b and c).

Table 1. PBD peptide affinities

Table 2. Data collection and refinement statistics

*Values in the highest-resolution shell are given in parentheses.

SI Materials and Methods

Transfected HeLa cells with GFP-Plk1 cDNA (WT and H538A/K540M mutant) were grown on coverslips for 20 h after transfection, fixed in 4% formaldehyde in PBS for 10 min, and permeabilized in cold (-20°C) methanol. They were then incubated with 20 mM glycine/PBS for 10 min, followed by an additional incubation with 3% BSA in PBS for 20 min. Microtubules and centrosomes were stained with antibodies against a-tubulin (YL1/2) and g-tubulin (Sigma; GTU88), respectively, and the corresponding Alexa dye-tagged secondary antibodies (Molecular Probes). DNA was counterstained with 4-6-diamidine-2-phenylindole (2 mg/ml). Finally, cells were mounted with Mowiol and analyzed with a Leica SP2 Confocal Scanning Microscope.

Fluorescence recovery after photobleaching (FRAP) of WT and mutant EGFP-Plk1-expressing HeLa cells was performed 18-20 h after transfection with EGFP-Plk1 and EGFP-H538A/K540M Plk1 plasmids. The 35-mm dishes with coverslip bottoms were directly mounted onto a Leica SP2 Confocal Scanning Microscope equipped with an argon laser. The growing medium was kept at 37°C and 5% CO₂ in a sealed chamber using The Cube and The Brick devices from Life Imaging Services. The 488-nm laser and a 63x plan Apo lens with a 1.4 NA were used in bleaching and imaging experiments. A laser power of 5-7% of 5mW was used in image acquisitions, and 100% of 5 mW was used in photobleaching. The time for each image acquisition ranges from 2 to 2.5 seconds, which did not significantly influence the fluorescent intensity through multiple acquisitions. An area of »10 mm² where the centrosome is embedded was bleached with an iteration of 5. Several images were collected before the bleach, and immediately after the photobleach at 3- to 5-seconds time lapses during at least 4 min. At least 10 cells were analyzed for each experiment. The final recovery models were generated and half-times calculated using the Axelrod method (1).

The Leica LAS AF quantification software was used for this purpose. The EGFP signal associated with different Plk1 constructs and the Alexa-594 signal associated with gamma-tubulin staining were captured in a Leica SP5 confocal. Equal laser intensity, optical thickness, and photomultiplier power acquisition settings were kept for all of the analyzed cells, to maintain the same conditions for each experiment. Only low EGFP-expressing cells were quantified to avoid overexpression artifacts. Fluorescence centrosomal mean intensity (intensity/area) was related to the total cell fluorescence to compensate for differences in expression. Finally, EGFP-Plk1 centrosomal fluorescence values were related to the g-tubulin intensity value to compensate possible changes in centrosomal size for a more accurate final analysis. More than 50 cells were analyzed for each construct.

The cDNA sequence corresponding to residues 1-603 of full-length human Plk1 was amplified by PCR and cloned into vector pFastBac HTa (Invitrogen) using BspHI (NcoI) and XhoI restriction sites for baculovirus-mediated protein expression in Sf9 insect cells. The recombinant pFastBac HTa vector containing the His-6 tagged fusion protein with a TEV cleavage site was used to generate recombinant baculovirus using the Bac-to-Bac method following the manufacturer's protocols. After Sf9 infection, cells were cultured for 3 days and then collected by centrifugation. Cells where disrupted by sonication and the protein was purified using a Ni-NTA column and gel filtration chromatography. Fractions corresponding to the pure protein were collected and concentrated to 2 mg/ml, flash frozen in liquid nitrogen, and stored at -80°C.

The cDNA sequence corresponding to residues 367-603 of human polo like kinase 1 was amplified by PCR and cloned into vector pGEX-6P-2 (Pharmacia). Protein expression and purification were performed as described (2).

Four micrograms of the recombinant His-tagged full-length Plk1 was mixed with Ni²⁺ beads and incubated for 60 min to allow binding. Subsequently, the His-tagged full-length Plk1 protein was incubated with either an interphasic or mitotic extract (100 mg) prepared from asyncronic HeLa cell cultures serum starved or arrested at prometaphase with nocodazole, respectively. After incubation, Plk1-Ni²⁺ beads were washed in lysis buffer three times and an extra wash in lysis buffer plus 0.5 M NaCl to remove possible non-specific bound proteins. Finally, Plk1-Ni²⁺ beads were washed in kinase buffer (10 mM Hepes-NaOH, pH 7.4/150 mM NaCl/10 mM MgCl₂/1 mM EGTA) supplemented with 0.5 mM DTT/5 mM sodium fluoride/1 mM sodium orthovanadate. Pure Plk1 was then eluted by addition of 300 mM imidazole. The kinase reaction was initiated by the addition of 0.2 mM ATP/2 mCi [g-³²P]ATP (Amersham Pharmacia) to the eluted Plk1 preparation, and 0.5 mg of Histone H1 (Roche) as substrate. After incubation for 30 min at 37°C, the reaction was stopped with SDS/PAGE sample buffer and analyzed by electrophoresis and autoradiography. The phosphorylated H1 was quantified by densitometric scanning. Kinase assays were carried out in the presence of both Cdc25C target peptide and the nonphysiological peptide in their primed and unprimed states. The eluted Plk1 after incubation with the HeLa extracts was mixed with peptides in a protein/target peptide molar ratio of 1:0.5, 1:1, and 1:1.5 for 60 min at room temperature. The kinase reaction was begun by the addition of 0.2 mM ATP/2 mCi [g-³²P]ATP (Amersham Pharmacia)/0.5 mg of Histone H1 (Roche) substrate. Again, after incubation for 30 min at 37°C, the reaction was stopped by the addition of 10 ml of SDS/PAGE sample buffer and analyzed by electrophoresis and autoradiography. The phosphorylated H1 was quantified by phosphoimager scanning (Amersham Pharmacia).

Cdc25C, nonphysiological, and control target peptides (Cdc25C-P sequence LLCS[pT]PNGL, Cdc25C sequence LLCSTPNGL, Nonphysiological-P sequence MAGPMQS[pT]PLNGAKK, nonphysiological sequence MAGPMQSTPLNGAKK, control-P sequence CGERKK[pT]LSGTPNY, and control sequence CGERKKTLSGTPNYI) were synthesized by GENOSPHERE Biotechnologies and checked by mass spectrometry. Fluorescence experiments were performed using a PTI fluorimeter. All of the measurements were carried out in 10 mM Hepes, pH 7.0/500 mM NaCl/0.5 mM EDTA/1 mM DTT at 20°C. For emission spectra, fluorescence emission was monitored with a bandpass of 6 nm between 310 and 340 nm. The fluorescence experiments were done by increasing the concentration of the target peptides from 0 to 30 mM (PBD) and 0 to 50 mM (Plk1). Binding curves were analyzed with Origin 6.0 scientific graphing and analysis software. Calorimetry measurements were performed using VP-ITC microcalorimeter (MicroCal Inc.). Experiments involved 8-ml injections of peptide solutions (100-150 mM) into a sample cell containing15 mM PBD in 20 mM Tris, pH 8.0/500 mM NaCl/1 mM EDTA/3 mM DTT. Thirty injections were performed with a spacing of 240 s using a reference power of 25 mCal/s. Binding isotherms were analyzed and plotted using Origin Software 7.0 (MicroCal Inc.). The concentrations of the peptide samples were measured by one dimensional proton NMR comparing the intensity of the peptide methyl signals with the intensity of the internal chemical shift standard DSS (4,4-dimethyl-4-silapentane-1-sulfonate). The concentration of the DSS was previously measured by the same method with a solution of N-acetyl-tryptophanamide whose concentration was determined spectrophotometrically.

Crystals of the PBD, PBD-Cdc25C-P, and PBD-Cdc25C were obtained as described in ref. 2. All data were collected using synchrotron radiation at the ESRF and SLS. Diffraction images were processed using HKL2000 (3) and Mosflm (4). Programs of the CCP4 package (Collaborative Computational Project 1994, ref. 5) were used for subsequent crystallographic calculations. Statistics for the crystallographic data are summarized in SI Table 1.

The structure was solved using the molecular replacement method as implemented in the program MOLREP (6). In both crystals, PBD-Cdc25C-P and PBD-Cdc25C, the search model was based on a polyalanine backbone derived from the PDB 1UMW found in the Protein Data Bank (7). The coordinates from the nonphysiological peptide were deleted in the search model. Both for PBD-Cdc25C-P and PBD-Cdc25C, the correct solution was the highest peak in rotation and translation searches including data between 15.0- and 4.0-Å resolution. The final correlation coefficient and R factor were 0.55/0.45 and 0.53/0.42, respectively. A 2F_o-F_c map showed clear and contiguous electron density for the peptide backbone and for many of the side chains of the protein. In both crystals, extra density was clearly seen in the peptide binding pocket after molecular replacement

PBD, PBD-Cdc25C-P, and PBD-Cdc25C were refined using the following strategy. Five percent of the reflections of the data set were set aside for free R factor calculations during refinement (8). The electron density map was calculated using only the working set of reflections and the model was rebuilt where the electron density supported changes and the side chains were fitted into the density. The resulting models were then refined against 1.95-, 2.10-, and 2.8-Å data set using CNS (9) for the first round of the refinement, including a rigid body minimization followed by simulated annealing (Cartesian starting at 5,000 K). Further rounds of model rebuilding were performed using the program O (10). Refinement proceeded with the program REFMAC5 (11) including a rigid-body refinement as the first step. We made use of Babinet bulk solvent correction (12) combined with overall anisotropic scaling and individual anisotropic temperature factor refinement using maximum likelihood as implemented in REFMAC5 in the case of the PBD and the PBD-Cdc25CP structures. Isotropic B factor refinement was used in the case of the PBD-Cdc25C structure. Several rounds of rebuilding using the program O and the placement of the peptides and the water molecules into the electron density, resulted in the final model. The statistics after crystallographic refinement of this model are summarized in supplementary material. All of the structure superpositions were performed with the use of the program O lsq routine (10).

The crystal structures of the PBD and its complexes with Cdc25C-P and Cdc25C target peptides have been solved by molecular replacement and refined respectively to 1.95-, 2.10-, and 2.80-Å resolution. Overall, all three structures show a common scaffold that was previously described. In each polo box, the six b-strands form an antiparallel b-sheet building a shallow cavity where the target peptide binds both in the case of Cdc25C and Cdc25C-P complexes (Fig. 3; SI Figs. 9 and 10). After a careful comparison of the three models, we found that the main difference in the protein structure among the three different crystal structures, the apo and the PBD-Cdc25C, PBD-Cdc25C-P complexes, arises from the absence of the 20-residue loop in the apo structure and the PBD-Cdc25C (residues Ala-493-Arg-507 for the apo and Glu-488-Arg-507 for the PBD-Cdc25C complex). This loop (connecting loop, CL) joins both polo boxes and flanks the binding site of the target peptide (SI Figs. 7 and 8). However, when the primed peptide was bound to the PBD the connecting loop was well defined and therefore built into the structure.

In both crystal structures where the peptide is bound, only 7 of its 10 residues could be modeled into the electron density map despite its phosphorylation status. Even though the target peptide binding occurs in the same pocket independently of its phosphorylation state, a detailed comparison of the binding mode shows differences depending on the threonine phosphorylation status. These differences affect the conformation of the PBD binding cleft amino acids as well as the target peptide conformation in the binding pocket (SI Figs. 10 and 11b). A close view reveals the binding differences and common features between the primed and unprimed target peptide (SI Fig. 10 b and c). While the conformation of the central core (Cys-3, Ser-4, and Thr-5) of the target peptide is similar in both cases (SI Figs. 10 and 11b), there are conformational changes in the main and side chains of several amino acids, as well as in their protein-peptide hydrogen bonding networks depending on the peptide phosphorylation state. The main differences can be observed both in Pro-6, which dramatically changes its conformation after phosphorylation in Thr-5, and Leu-1 and Leu-2 whose side chains, in the case of the unprimed target peptide, are oriented toward one side of the cavity, where they associate in hydrophobic contacts with Phe-535 and a hydrogen bond of Leu-1 carbonyl with the side chain of Arg-516 NH2 (2.66 Å). However, their conformation is different when the primed peptide is bound. In this case, both Leu-1 and Leu-2 are involved in an extensive network of hydrophobic interactions with Val-415, Tyr-417, Tyr-485, and Arg-516, which are located on the other side of the cavity, and the main chain hydrogen bond with the Arg-516 side chain is disrupted. The conservation of the Thr-Ser-Cys conformation in the central core of the target peptide allows the unprimed Thr-5 interaction with the side chain of His-538 (Thr5OG1 His538ND1, 3.5 Å). This is one of the residues involved in phosphate binding together with Lys-540, as it has previously been described in the PBD complex with a nonphysiological peptide. The presence of the phosphate group in Thr-5 side chain promotes the interaction of O1P with His538ND1 (2.65 Å) and O2P with Lys-540 NZ (2.69 Å). Moreover, the phosphate moiety is associated with seven water molecules which form an extensive hydrogen bond network. These water molecules are absent when the unprimed peptide is bound.

Several PBD Cdc25C peptide main chain interactions are similar despite the target peptide phosphorylation status. They are mainly due to the carbonyl main chain of Trp-414 with the amide group from Ser-4 (2.78 Å) (SI Fig. 10). It has been observed that this residue seems to play an important role in Plk1 subcellular localization. Indeed, its mutation (W414F) disrupts proper Plk1 subcellular localization without disturbing its kinase activity. Remarkably, the presence of Trp-414 displaying its indole ring on the bottom of the binding cleft favors contacts with Ser-4, Cys-3, and Leu-2 main chains that are conserved in both PBD-Cdc25C and PDB-Cdc25C-P structures. These data indicate that Trp-414 could be involved not only in localization but also participate in the molecular recognition of the target substrate independently of its phosphorylation state. Both processes are essential for proper Plk1 function.

As mentioned before, there is a clear change in Pro-6 conformation between the unprimed and the primed target peptide (SI Fig. 11b). In the PDB-Cdc25C complex structure, Pro-6 conformation probably promotes a conformational change in the connecting loop with respect to the structure observed in the PDB-Cdc25C-P complex. This is mainly due to the location of the unprimed target peptide in an area of the binding pocket, where the main chain of this loop lies in the PBD-Cdc25-P complex. On the other hand, whereas in the PBD-Cdc25-P complex the Asn-7 side chain is associated in a hydrogen bond network with three water molecules and contacts the carbonyl of His-489 through its amide group (2.89 Å), none of these interactions could be observed in the unprimed peptide complex due to the absence of density to build the connecting loop between polo box I and polo box II .

A comparison between the structures of the PBD-Cdc25C-P and the PBD in complex with the nonnatural peptide reveals differences in the peptide binding mode (SI Fig. 11c). Even though the p-Thr and the following Ser maintain similar conformations and contact similar residues in the PBD in both cases, the positions and bond lengths as well as the number of water molecules that can interact with O1P and O2P in Cdc25C-P are different from those that have been described for the non natural peptide. It is noteworthy that a close view of the Cdc25C-P phosphate moiety reveals its interaction (O1P and O2P) with the side chain from two basic residues, Arg-518 and Lys-556, which belong to other crystallographically related molecules. Each contact is mediated by a water molecule (Fig. 3b). This reveals that the phosphate binding cleft in the PBD is not buried inside the protein core but relatively exposed in the shallow binding pocket. Therefore, it is accessible to other residues located outside the PBD. Additional differences are observed in the peptide conformation both in the C- and N-terminal regions. The presence of polar amino acids in the C-terminal region of the nonnatural peptide (Gln-His-Met) contrasts with the hydrophobic ones in these positions in the case of Cdc25C-P peptide (Cys-Leu-Leu). This distinct chemical character promotes an array of different interactions between the ligand and the protein, which are reflected in the different accommodations that the side chains adopt (Fig. 5a). The differences are minor in the residues in the N-terminal side before the p-Thr (Asn-Pro).

1. Axelrod D, Koppel DE, Schlessinger J, Elson E, Webb WW (1976) Biophys J 16:1055-1069.

2. García-Álvarez B, Ibanez S, Montoya G (2006) Acta Crystallogr F62:372-375.

3. Otwinowski Z, Minor W (1997) Methods in Enzymology, Vol 276 (Academic, New York).

4. Leslie AGW (1992) Joint CCP4+ ESF-EAMCB Newsletter on Protein Crystallography, no. 26.

5. Collaborative Computational Project (1994) Acta Crystallogr D50:760-763.

6. Vagin A, Teplyakov A (2000) Acta Crystallogr D56:1622-1624.

7. Benson DA, Karsch-Mizrachi I, Lipman DJ, Ostell J, Rapp BA, Wheeler DL (2000) Nucleic Acids Res 28:15-18.

8. Brünger AT (1992) Nature 355:472-475.

9. Brünger AT, Adams PD, Clore GM DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, et al. (1998) Acta Crystallogr D54:905-921.

10. Jones TA, Zou J-Y, Cowan SW, Kjeldgaard M (1991) Acta Crystallogr A47:110-115.

11. Murshudov GN, Vagin AA, Lebedev A, Wilson KS, Dodson EJ (1999) Acta Crystallogr D55:247-255.

12. Moews PC, Kretsinger RH (1975) J Mol Biol 91:201-228.