PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
ACS Chem Biol. Author manuscript; available in PMC 2013 May 18.
Published in final edited form as:
Published online 2012 February 14. doi:  10.1021/cb200469a
PMCID: PMC3355227
NIHMSID: NIHMS354787

Identification of High Affinity Polo-like Kinase 1 (Plk1) Polo-box Domain Binding Peptides Using Oxime-based Diversification

Abstract

In an effort to develop improved binding antagonists of the polo-like kinase 1 (Plk1) polo-box domain (PBD), we optimized interactions of the known high affinity 5-mer peptide, PLHSpT using oxime-based post-solid-phase peptide diversification of the N-terminal Pro residue. This allowed us to achieve up to two orders-of-magnitude potency enhancement. An X-ray crystal structure of the highest affinity analogue in complex with Plk1 PBD revealed new binding interactions in a hydrophobic channel that had been occluded in X-ray structures of the unliganded protein. This study represents an important example where amino acid modification by post solid-phase oxime ligation can facilitate the development of protein-protein interaction inhibitors by identifying new binding pockets that would not otherwise be accessible to coded amino acid residues.

Members of the polo subfamily of protein kinases (collectively, Plks) play pivotal roles in cell proliferation. Among them, Plk1 has been studied most extensively because of its ability to promote tumorigenesis in human cells.1 Plk1 offers two distinct drug targets within one molecule — an N-terminal catalytic domain and a C-terminal polo-box domain (PBD). Over the years, a great deal of effort to generate anti-Plk1 inhibitors has been focused on targeting the catalytic domain. However, large body of evidence suggests that the PBD serves as a cis-acting phospho-binding module that is essentially required for bringing the catalytic activity of Plk1 to its binding targets at specific subcellular locations.24 Thus, inhibition of PBD-dependent Plk1 function could serve as an attractive alternative that is worthwhile to explore.

We have shown that post solid-phase peptide diversification by introducing aminooxy functionality at either the 3- or 4-position of a Pro pyrrolidine ring and subjecting the resulting peptides to oxime-based ligation reactions can have significant utility in the study of protein-protein interaction (PPI) inhibitors.57 The chemical stability of oximes8 and the ease with which they can be prepared in high purity, makes oxime ligation an attractive method for the post-solid-phase construction of peptide libraries that can be biologically evaluated directly without purification.57 Oxime ligation can be a highly effective way to explore peptide-protein interactions in ways that are not possible using coded amino acids.

The polo-box interacting protein 1 (PBIP1) is a known Plk1 substrate that undergoes phosphorylation at T78 to form a Plk1 PBD-binding ligand.9 The PBIP1-derived 5-mer peptide, 74-PLHSpT-78 (1) represents a minimal sequence that specifically interacts with the Plk1 PBD with a high affinity (Kd = 0.45 µM), but not with the two closely-related Plk2 and Plk3 PBDs.10 Co-crystal structures of the Plk-1 PBD in complex with short phosphopeptides3, 4, 10, 11 show that while interaction of the SpT motif is similar in all of these structures, for PLHSpT (PDB: 3HIK) and PPHSpT (PDB: 3C5L) the N-terminal Pro residues bind in greatly different orientations.10 The combined protein binding surface encompassing these Pro residues covers a broad region that seemed to afford an ideal opportunity for oxime-based post solid-phase ligation originating from the N-terminal Pro residue (Suporting Information Figure S1). Herein we report an application of this approach to explore structural variations originating from the Pro residue, which resulted in the identification of analogues that exhibit greater than two orders-of-magnitude enhanced Plk1 PBD affinities.

Oxime libraries were prepared to diversify the Pro residue of peptide 1. The pyrrolidine ring of the original residue was maintained, since biochemical data indicate that this Pro residue is important both for Plk1 PBD-binding affinity and specificity relative to the closely related kinases Plk2 and Plk3.10 Fmoc-based solid-phase peptide synthesis employing orthogonally-protected trans (4R) and cis (4S) aminooxyproline derivatives5 provided the corresponding peptides 2 and 3, respectively, bearing free aminooxy-groups at the 4-position of the Pro residues (Figure 1). Oxime–containing peptide libraries (4 and 5) were then prepared by conjugating peptides 2 and 3 with a set of 8 selected aldehdyes (ah). The choice of these aldehydes was based on their diversity (charge, size, aromatic, aliphatic, heteroaryl, etc.). The resulting peptides were directly evaluated for their ability to compete with an immobilized PBIP1-derived p-T78 peptide in ELISA-based Plk1 inhibition assays using HEK 293A cell lysates expressing GFP-HA-fused Plk1 (Supporting Information Fig. S6).10 The peptides containing oximes formed from 3-phenylpropanal (4b and 5b) showed the greatest affinity enhancement relative to the parent 1, with 4b being slightly more potent than 5b.

Figure 1
Structures of peptides and reagents used in the generation of oxime libraries.

More focused libraries were then prepared to explore structural features of 4b and 5b. Decreasing (4i and 5i) or increasing (4j and 5j) the alkyl chain length by one methylene group decreased affinity with the trans (4R) isomers (4i and 4j) being slightly more potent than the corresponding cis (4S) isomers (5i and 5j) (Supporting Information Figure S7). A methyl scan (4k – 4m) conducted on the phenyl ring of 4b, had little effect on binding affinity, although substitution at the meta-position (4l) slightly increased affinity (Supporting Information Figure S8). Based on the higher affinity of 4l, additional meta-substituted analogues (4n – 4p) were prepared, and all of these exhibited binding potencies similar to 4b, although the 3-methoxy substituent (4n) slightly increased the binding affinity, while the 3-phenyl substituent (4p) slightly decreased the affinity. Replacing the oxime functionality of 5b with an amidooxy group (6, Figure 2a) was also deleterious.

Figure 2
(a) Structures of peptides discussed in the text. (b) Peptide pull-down assay showing ability of the indicated peptides (top) to pull down Plk1 – 3 from mitotic 293T cell lysates expressing kinase-inactive Flag-Plk1 (K82M), Flag-Plk2 (K108M) or ...

The analogues 4b and 5b were converted to the corresponding 4-phenylbutyl ethers (7 and 8, respectively, Figure 2a) using pre-formed proline derivatives. The ethers removed the imine portion of the original oximes, while maintaining a similar chain extension. Both 7 and 8 had significantly higher binding affinities than their corresponding oximes, with the trans (4R) isomer (7) being more potent than the cis (4S) epimer (8) (Supporting Information Figure S8). Binding was phospho-dependent, since a near total loss of affinity was observed for the variant in which the phosphothreonine residue had been replaced by a threonine residue 7(pT5T, Supporting Information Figure S9). Additionally, in order to exclude possible inhibition by non-specific “promiscuous” mechanisms arising from hydrophobicity that are unrelated to specific interactions with PBD12 we made use of the fact that the “SpT” dipeptide motif is critical for high affinity Plk1 PBD-binding and that replacement of the serine residue by an alanine (S/A replacement) typically abrogates binding.4 We observed that S/A variants, 7(S4A) and 8(S4A), showed a significant loss of affinity relative to the corresponding parent peptides (Supporting Information Figure S9). This argued strongly that binding of 7 and 8 was specific in nature.

The ELISA-based Plk1 inhibition data (Supporting Information Figure S6 – S9) provided relative binding affinities that served to guide structural modifications. In order to quantitate the binding affinities of selected analogues, the assays were repeated using an expanded range of concentrations (Supporting Information Figure S10). This allowed an estimation of IC50 values: 1 (20 µM); 4b (0.43 µM); 7 (0.04 µM); 7* (0.20 µM) and 7(S4A) (43 µM) [where 7* indicates replacement of the pT residue with (2S,3R)-2-amino-3-methyl-4-phosphonobutyric acid (Pmab) as a phosphatase-stable pT mimetic13].

Binding affinities were also determined independently using fluorescence polarization techniques, which measured the ability of peptides to compete with a 5-carboxyfluorescein-labeled variant of the peptide GPMQSpTPLNG-OH (9) (5-CF-9) for binding to purified Plk1 PBD protein (Table 1).14 In this latter assay, the WT 5-mer parent peptide 1 (40 ± 2% inhibition at 2.56 µM concentration) was slightly less potent than the control 10-mer peptide (9, IC50 = 1.12 ± 0.26 µM). The isomeric oximes 4b and 5b were approximately an order-of-magnitude more potent than 1 (IC50 = 0.122 ± 0.024 µM and 0.433 ± 0.083 µM, respectively). Consistent with the ELISA-based inhibition assay, the trans-isomer bound with higher affinity than the cis-isomer. Conversion of the oximes 4b and 5b to their corresponding ether analogues was accompanied by another order-of-magnitude increase in affinity (7, IC50 = 0.014 ± 0.001 µM and 8, IC50 = 0.038 ± 0.009 µM, respectively), which represents an approximate two orders-of-magnitude enhancement relative to the WT parent peptide 1. The Pmab-containing versions of 7 and 8 bound with less affinity than their pThr-containing parents. This was observed for both 7 (7*, IC50 = 0.086 ± 0.017 µM; 6-fold less potent) and 8 (IC50 = 0.038 ± 0.009 µM as compared to 8*, IC50 = 0.114 ± 0.003 µM; 3-fold less potent).

Table 1
Plk1 PBD-binding IC50-values.a

We introduced onto 7 and selected variants, N-terminal Cys residues tethered by n-hexanoylamide chains and covalently conjugated the resulting peptides to SulfoLink Coupling Gel. We then measured the relative abilities of these preparations to bind to Plk1, Plk2 or Plk3, when exposed to lysates of mitotic 293T cells containing Flag-fused kinase dead forms of Plk1 (K82M), Flag-Plk2 (K108M) or Flag-Plk3 (K52R) (Figure 2b). While confirming our previous findings that 1 is highly specific for Plk1,1015 a faint band corresponding to binding of peptide 7 to Plk2 was observed in addition to a very intense band associated with its binding to Plk1. A Plk2 band was not seen for the Pmab-containing analogue 7*, although more than 200-fold and approximately 6-fold reduced Plk1 PBD binding affinities of 1 and 7* relative to 7 could render binding of these peptides to Plk2 too faint for detection.

In order to determine the molecular basis for the enhanced binding affinity of 7, we solved the X-ray co-crystal structure of 7 in complex with Plk1 PBD (Supporting Information Table S3 and Figure S12). The “HSpT” residues of 7 were nearly super-imposable with those of the PBD-bound 1 in the 3HIK structure (Figure 3a). However, significant structural differences were observed with the Leu residue, where the psi angle (ψ = −1.9° and 157.9° for 1 and 7, respectively) placed the adjacent N-terminal Pro residues in nearly opposing directions (Figure 3a). This is reminiscent of what is observed in PBD-bound “PPHSpT” (PDB: 3C5L), where the latter’s P2 psi angle (ψ = 145.4°) places its P1 residue in a similar position as 7 (compare with Supporting Information Figure S1). In this new orientation, the trans-(4R) phenylbutyloxy substituent on the P1 pyrrolidine ring of 7 is directed across the β2 and β3 sheets of PBD, where it terminates with its phenyl ring nestled against the αB helix. The net effect is to reveal a channel that is occluded in the binding of parent peptide 1 (compare Figure 3b and 3c).

Figure 3
X-ray co-crystal structures of Plk1 PBD complexed with peptides 1 and 7 and 10. (a) PBD in complex with 1 (PBD 3HIK; protein backbone and peptide shown in red) superimposed on the complex with 7 (protein backbone in grey with peptide 7 colored by atom). ...

The interactions of the −(CH2)4-Phe moiety of 7 are superimposable with those of a recently reported −(CH2)8-Phe adduct originating from the histidine imidazole ring of 1 (peptide 10, PDB: 3RQ7) (Figure 3d).15 The binding of the phenyl ring of 7 was also similar to what was recently reported for the phenyl ring in the F1 residue of “FDPPLHSpTA” (peptide 11, PDB: 3P37) (Supporting Information Figure S13).16 Peptide 11 represents an extended version of 1 encompassing residues 71–79 of the T78 region within the PBIP1 protein.9 For PBD complexes of 7, 10 and 11, important protein residues lining the binding channel (V415, Y417, Y421, L478, Y481, F482 and Y485) provide a highly hydrophobic environment (Supporting Information Figure S13 and Table S4). Hydrophobic interactions with the residues within the channel are similar for 7, 10 and 11, with the exceptions that 11 has significantly more contact with Y417 (due to face-on binding with the peptide P3 residue) and 7 and 10 make greater contact with the L478 residue than 11 does. These residues are oriented similarly to those found in the PBD•1 complex (PDB: 3HIK), with a critical exception arising from the Y481 side chain, which occludes the binding channel in the unliganded protein (PDB: 1Q4O) and in the PBD•1 complex. Dramatic rotation of the Y481 side chain is required to reveal the binding channel and open access to the αB helix-proximal aryl-binding pocket in the PBD complexes of 7, 10 and 11 (Supporting Information Figure S14).15, 16

By application of a facile oxime-based post solid-phase peptide diversification protocol, we have realized a 2 orders-of-magnitude improvement in affinity over the starting WT peptide. This was achieved while maintaining good Plk1 PBD selectivity relative to the PBDs of Plk2 and Plk3. More importantly, an X-ray cocrystal structure of the most potent analogue revealed new binding interactions in a hydrophobic channel of the PBD that is occluded in the unliganded protein or in the protein complexed with parental 5-mer peptide . Our work reinforces the value of oxime-based post solid-phase ligation for the discovery of new binding interactions that would not otherwise be apparent from the binding of peptides containing naturally encoded amino acids. Viewed as a “tethered fragment” approach, oxime ligation has distinct advantages over traditional fragment-based screening, where the low binding affinities of fragment libraries present inherent limitations.17 The potential power of the directed approach employed in the current study is shown by its ability to identify binding interactions highly similar to those previously identified through serendipity. Our findings expand the design parameters available for development of Plk1 PBD-binding antagonists, and provide direction that may be useful for optimizing PPI inhibitors in general.

METHODS

Peptide Synthesis

Peptides were synthesized and purified as described in the Supporting Information.

Post Solid-phase Diversification: Preparation of Oxime-containing Peptides 4a–p and 5a–p (5 mM in DMSO)

A mixture of HPLC-purified aminooxy-proline containing peptide (2 or 3) (15 mM in DMSO, 10 µL), aldehyde (a–p, Supporting Information Figure S2) (15 mM in DMSO, 10 µL) and acetic acid (70 mM in DMSO, 10 µL) was gently agitated at room temperature (over night). Crude reaction mixtures were used directly for biological evaluation. 4b and 5b were scaled up and purified by preparative HPLC.

ELISA-based PBD-binding inhibition assays

Peptide pull-down assays were carried out essentially as described previously.10, 15 A biotinylated p-T78 peptide was first diluted with 1X coating solution (KPL Inc.) to a final concentration of 0.3 µM, and then 100 µL of the resulting solution was immobilized onto a 96-well streptavidin-coated plate (Nalgene Nunc). The wells were washed once with PBS plus 0.05% Tween20 (PBST), and incubated with 200 µL of PBS plus 1% BSA (blocking buffer) for 1 h to prevent non-specific binding. Mitotic 293A lysates expressing HA-EGFP-Plk1 were prepared in TBSN buffer (~ 60 µg total lysates in 100 µL buffer), mixed with the indicated amount of peptide ligands and applied immediately onto the biotinylated p-T78 peptide-coated ELISA wells, and then incubated with constant rocking for 1 h at 25 °C. Following incubation, the ELISA plates were washed 4 times with PBST. To detect bound HA-EGFP-Plk1, the plates were probed for 2 h with 100 µL well−1 of anti-HA antibody at a concentration of 0.5 µg mL−1 in blocking buffer and then washed 5 times. The plates were further probed for 1 h with 100 µL well−1 of HRP-conjugated secondary antibody (GE Healthcare) at a 1:1,000 dilution in blocking buffer. The plates were washed 5 times with PBST and incubated with 100 µL well−1 of 3,3’,5,5’-tetramethylbenzidine (TMB) substrate solution (Sigma) until a desired absorbance was achieved. The reactions were stopped by the addition of 100 µL well−1 of stop solution (Cell Signaling Technology) and the optical densities (O.D.) were measured at 450 nm using an ELISA plate reader (Molecular Devices). Data are shown in Supporting Information Figures S6 – S10.

PBD Fluorescence Polarization Competition Binding Assays for Plk1

Competition assays for the Plk1 PBD were also performed essentially as described.14, 18, 19 In brief, 5-carboxyfluorescein-GPMQSpTPLNG-OH (5-CF-9) (final concentration: 2 nM) was incubated with the Plk1 PBD (final concentration: 20 nM) in the presence of the test peptides (final concentrations of buffer components: 50 mM NaCl, 10 mM Tris (pH 8.0), 1 mM EDTA, 0.1% (v/v) Nonidet P-40 substitute, 1 mM dithiothreitol, and 2% (v/v) DMSO.) Fluorescence polarization was analyzed after 60 min. Inhibition curves were fitted using SigmaPlot (SPSS) and shown in Supporting Information Figure S11. All experiments were performed in triplicate. Numerical IC50 values are shown in Table 1.

Peptide Pull-down Assays

Peptide pull-down assays were carried out essentially as described previously.10, 15 To study Plk1 PBD-binding specificity, test peptides were covalently conjugated to beads using SulfoLink Coupling Gel (Pierce) via either an N-terminal Cys-(CH2)6−CO linker. Mitotic lysates expressing Plk1-3 were prepared from 293T cells transfected with Flag-Plk1 (K82M), Flag-Plk2 (K108M) or Flg-Plk3 (K52R) (a gift of Wei Dai, New York University School of Medicine, NY) and treated with 200 ng mL−1 of nocodazole for 16 h. Cell lysates were then prepared in TBSN buffer {20 mM Tris-Cl (pH8.0), 150 mM NaCl, 0.5% Np-40, 5 mM EGTA, 1.5 mM EDTA, 20 mM p-nitrophenylphosphate and protease inhibitor cocktail (Roche)} and incubated with bead-immobilized peptides for 2 h at 4 °C. The peptide-associating proteins were precipitated, washed, boiled in sodium dodecyl sulfate (SDS) sample buffer, separated by 8% SDS-polyacrylamide gel electrophoresis (PAGE), and then subjected to immunoblotting analysis with anti-Flag antibody and the enhanced chemilunimescence (ECL) detection system (Pierce). The same membrane was also stained with Coomassie (CBB). Signal intensities were quantified using Image J program. Results are shown in Figure 2b.

X-ray Crystallography of Peptide 7 in Complex with Plk1 PBD Protein

Experimental details and refinement data are provided in the Supporting Information. The graphics in Figure 3 Graphics generated using ICM Chemist Pro (v3.7) by Molsoft, Inc. (www.molsoft.com). Final graphics were assembled using Adobe Photoshop CS4.

Supplementary Material

1_si_001

ACKNOWLEDGMENT

This work was supported in part by the Intramural Research Program of the NIH, Center for Cancer Research, NCI-Frederick and the National Cancer Institute, National Institutes of Health (F. L., J.-E. P., W.-J. Q., K. S. L. and T. R. B.) and National Institutes of Health grant R01 GM60594 and ES015339 (M. B. Y.) and the Deutsche Forschungsgemeinschaft grant BE 4572/1-1 (T. B). The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.

Footnotes

ASSOCIATED CONTENT

Accession Codes. Protein Data Bank: Coordinates for PBD in complex with 7 have been deposited under accession code 4DFW.

Supporting Information. Synthetic for the preparation of proline analogues employed in peptide synthesis and synthesis of peptides and associated analytical data; X-ray crystallographic crystallization and refinement data for Plk1 PBD complexed with 7; Plk1 PBD ELISA binding curves; Plk1 PBD fluorescence polarization binding curves and graphics related to the Plk1 PBD binding of 1, 7, 10 and 11. NMR spectra for synthetic proline analogues and HPLC traces for peptides 2, 3, 4b, 5b, 7 and 8. This material is available free of charge via the Internet at http://pubs.acs.org.

REFERENCES

1. Strebhardt K, Ullrich A. Targeting polo-like kinase 1 for cancer therapy. Nat. Rev. Cancer. 2006;6:321–330. [PubMed]
2. Lee KS, Grenfell TZ, Yarm FR, Erikson RL. Mutation of the polo-box disrupts localization and mitotic functions of the mammalian polo kinase plk. Proc. Natl. Acad. Sci. U. S. A. 1998;95:9301–9306. [PubMed]
3. Cheng K-Y, Lowe ED, Sinclair J, Nigg EA, Johnson LN. The crystal structure of the human polo-like kinase-1 polo box domain and its phospho-peptide complex. EMBO J. 2003;22:5757–5768. [PubMed]
4. Elia AE, Rellos P, Haire LF, Chao JW, Ivins FJ, Hoepker K, Mohammad D, Cantley LC, Smerdon SJ, Yaffe MB. The molecular basis for phosphodependent substrate targeting and regulation of plks by the polo-box domain. Cell. 2003;115:83–95. [PubMed]
5. Liu F, Stephen AG, Fisher RJ, Burke TR. Protected aminooxyprolines for expedited library synthesis: Application to tsg101-directed proline-oxime containing peptides. Bioorg. Med. Chem. Lett. 2008;18:1096–1101. [PMC free article] [PubMed]
6. Liu F, Stephen AG, Waheed AA, Aman MJ, Freed EO, Fisher RJ, Burke TR., Jr Sar by oxime-containing peptide libraries: Application to tsg101 ligand optimization. ChemBioChem. 2008;9:2000–2004. [PMC free article] [PubMed]
7. Kim S-E, Liu F, Im YJ, Stephen AG, Fivash MJ, Waheed AA, Freed EO, Fisher RJ, Hurley JH, Burke TR., Jr Elucidation of new binding interactions with the human tsg101 protein using modified hiv-1 gag-p6 derived peptide ligands. ACS Med. Chem. Lett. 2011;2:337–341. [PMC free article] [PubMed]
8. Kalia J, Raines RT. Hydrolytic stability of hydrazones and oximes. Angew. Chem. Int. Ed. 2008;47:7523–7526. [PMC free article] [PubMed]
9. Kang YH, Park J-E, Yu L-R, Soung N-K, Yun S-M, Bang JK, Seong Y-S, Yu H, Garfield S, Veenstra TD, Lee KS. Self-regulated plk1 recruitment to kinetochores by the plk1-pbip1 interaction is critical for proper chromosome segregation. Mol. Cell. 2006;24(3):409–422. [PubMed]
10. Yun S-M, Moulaei T, Lim D, Bang Jeong K, Park J-E, Shenoy Shilpa R, Liu F, Kang Young H, Liao C, Soung N-K, Lee S, Yoon D-Y, Lim Y, Lee D-H, Otaka A, Appella E, McMahon James B, Nicklaus Marc C, Burke TR, Jr, Yaffe Michael B, Wlodawer A, Lee Kyung S. Structural and functional analyses of minimal phosphopeptides targeting the polo-box domain of polo-like kinase 1. Nat. Struct. Mol. Biol. 2009;16:876–882. [PMC free article] [PubMed]
11. Garcia-Alvarez B, de CG, Ibanez S, Bragado-Nilsson E, Montoya G. Molecular and structural basis of polo-like kinase 1 substrate recognition: Implications in centrosomal localization. Proc. Natl. Acad. Sci. U. S. A. 2007;104:3107–3112. [PubMed]
12. Coan KED, Maltby DA, Burlingame AL, Shoichet BK. Promiscuous aggregate-based inhibitors promote enzyme unfolding. J. Med. Chem. 2009;52:2067–2075. [PMC free article] [PubMed]
13. Liu F, Park J-E, Lee KS, Burke TR., Jr Preparation of orthogonally protected (02s,3r)-2-amino-3-methyl-4-phosphonobutyric acid (pmab) as a phosphatase-stable phosphothreonine mimetic and its use in the synthesis of polo-box domain-binding peptides. Tetrahedron. 2009;65:9673–9679.
14. Reindl W, Strebhardt K, Berg T. A high-throughput assay based on fluorescence polarization for inhibitors of the polo-box domain of polo-like kinase 1. Anal. Biochem. 2008;383:205–209. [PubMed]
15. Liu F, Park J-E, Qian W-J, Lim D, Graber M, Berg T, Yaffe MB, Lee KS, Burke TR., Jr Serendipitous alkylation of a plk1 ligand uncovers a new binding channel. Nat. Chem. Biol. 2011;7:595–601. [PMC free article] [PubMed]
16. Śledź P, Stubbs CJ, Lang S, Yang Y-Q, McKenzie GJ, Venkitaraman AR, Hyvönen M, Abell C. From crystal packing to molecular recognition: Prediction and discovery of a binding site on the surface of polo-like kinase 1. Angew. Chem., Int. Ed. 2011;50:4003–4006. [PMC free article] [PubMed]
17. Rees DC, Congreve M, Murray CW, Carr R. Fragment-based lead discovery. Nat. Rev. Drug Discov. 2004;3:660–672. [PubMed]
18. Reindl W, Yuan J, Kraemer A, Strebhardt K, Berg T. Inhibition of polo-like kinase 1 by blocking polo-box domain-dependent protein-protein interactions. Chem. Biol. 2008;15:459–466. [PubMed]
19. Reindl W, Graeber M, Strebhardt K, Berg T. Development of high-throughput assays based on fluorescence polarization for inhibitors of the polo-box domains of polo-like kinases 2 and 3. Anal. Biochem. 2009;395:189–194. [PubMed]