Computational Modeling of Selected Small Molecules Targeting the PLK1 Polo-Box Domain (PBD)

Authors: Monsurat M. Lawal (1), Sathvega Somasundaram (2), Rohan Shah (3), and Anshu Cherukumilli (4)

Editor: Tugba G Kucukkal, Ph.D.

Author Affiliations: (1) MetroTech Institute, (2) Evergreen Valley High School, (3) Poolesville High School, and (4) Montgomery High School

Abstract

Polo-like kinase 1 (PLK1) is a critical regulator of mitotic progression, and its overexpression has been implicated in various cancers. Targeting the polo-box domain (PBD) of PLK1 offers a selective approach to disrupting PLK1 function without directly inhibiting its kinase activity, thereby minimizing off-target effects. This study employs computational modeling techniques to identify and characterize small molecules that selectively bind the PLK1 PBD, aiming to uncover key interactions essential for inhibitor design. A library of 20 small molecules, including first- and second-generation PLK1 PBD inhibitors, was screened against the seven potential binding sites (BS1–BS7) within the PBD. Binding site identification was guided by experimental data and computational predictions using CB-Dock2. Molecular docking enabled ligand interactions with PLK1 and homologous PBDs in PLK2 and PLK3 assessment. PLK3 PBD, lacking an experimentally resolved structure, was modeled using AlphaFold. Post-docking analysis revealed critical hydrogen bonding and hydrophobic interactions driving ligand binding at different sites. Although docking approaches are less exhaustive, this comparative binding assessment across PLK1, PLK2, and PLK3 provided insights into structural determinants governing inhibitor selectivity. This study also provides an introductory computational molecular modeling learning curve for beginners.

Keywords: Polo-like kinase (PLK1, PLK2, PLK3), Polo-box domain (PBD), Multi-pockets (BS1–BS7), Molecular docking, Computational drug discovery, Protein-ligand interactions.

1. Introduction

Polo-like kinase 1 (PLK1) is a serine/threonine kinase that plays a pivotal role in cell cycle regulation, mitotic progression, and genomic stability [1]. Overexpression of PLK1 has been linked to the uncontrolled proliferation of cancer cells, making it a promising therapeutic target in oncology [2]. While many kinase inhibitors target the ATP-binding catalytic domain of PLK1, concerns regarding off-target effects and acquired resistance have driven interest toward alternative regulatory domains. One such region, the polo-box domain (PBD), is essential for substrate recognition and proper intracellular localization of PLK1. Inhibiting the PBD disrupts PLK1 function selectively without directly interfering with its kinase activity, thereby reducing potential toxicities associated with ATP-competitive inhibitors [3].

Despite its therapeutic potential, developing small-molecule inhibitors targeting the PLK1 PBD has been challenging due to its unique structural and biophysical properties. The PBD consists of two polo-box motifs that form a phosphopeptide-binding pocket, allowing it to recognize phospho-serine/threonine-containing motifs in substrate proteins [4]. Several experimental studies have explored peptide-based and small-molecule inhibitors that can bind to this domain; however, many lack specificity, bioavailability, or optimal pharmacokinetic properties [1]. Therefore, rational drug design strategies, particularly computational modeling, are valuable tools for identifying and optimizing small molecules that selectively target the PLK1 PBD.

This study employs computational modeling techniques to identify and characterize small molecules that bind to the PLK1 PBD with high specificity and affinity. The approach integrates molecular docking to (1) predict favorable ligand-PBD interactions and (2) assess cross-binding with homologous PBDs in PLK2 and PLK3. The study focuses on seven key binding sites (BS1–BS7) within the PLK1 PBD to elucidate the interaction modes across these PLKs. While PLK1 is a well-established oncogenic target, PLK2 and PLK3 have distinct cellular functions, including tumor-suppressive roles in certain contexts [5]. With this computational approach, we aim to uncover key molecular determinants of PBD-ligand interactions, providing insights that could inform the rational design of next-generation PLK1 PBD inhibitors.

2. Methods

1.1. Selection of Small Molecules

A diverse library of 20 small molecules was curated from a recent review [1] to evaluate their binding potential to the PBD of PLK1. The selection included known first and second-generation PLK1 PBD inhibitors from the literature (Figure 1). These molecules were retrieved from publicly available chemical databases such as PubChem [6] and RCSB PDB [7] or constructed within Avogadro.

Figure 1: Selected PLK1-PBD molecules compiled from the literature [1].

2.2 Molecular Geometry Optimization

The selected compounds were preprocessed and structurally optimized using Avogadro [8], an open-source molecular editing software. Energy minimization was performed using the MMFF94 force field, ensuring stable and low-energy conformations before docking. Optimized structures were saved in PDBQT format for compatibility with docking software.

2.3 Identification of Binding Pockets

To determine the most relevant binding regions within the PBD, we utilized CB-Dock2 [9], an automated pocket detection and docking tool. CB-Dock2 identifies potential binding sites by analyzing protein surfaces and ligand shape complementarity. The PLK1 and PLK2 PBD structures were retrieved from the Protein Data Bank with ID numbers 4HCO [10] and 4XB0 [11] and prepared for docking. PLK3 PBD was retrieved from AlphaFold [12] since its 3D experimental structure is unavailable. The top seven binding pockets (BS1–BS7, Figure 2) on PLK1 PBD were selected based on experimentally determined pockets (BS1–BS3) [3, 10, 13] and highly populated predicted binding sites (BS1, BS4–BS7) from our computational analysis.

2.4 Molecular Docking

Molecular docking was conducted using AutoDock Vina [14], a widely used docking program that predicts binding poses and affinities of small molecules within a target protein. The first step is protein preparation – remove non-protein molecules like water, ions, and ligands from PLK1 (4HCO) and PLK2 (4XB0), add missing looping residues to 4HCO using MODELLER [15], add hydrogen atoms to all three PLKs using VMD [16, 17], convert protein structures to PDBQT format using AutoDockTools [18]. The second step is docking – grid box coordinates were set for each identified binding site (BS1–BS7) based on CB-Dock2 predictions and experiments (Figure 2). Each ligand was docked separately against the PLK1, PLK2, and PLK3 PBDs, generating multiple binding conformations. The docking scoring function (in kcal/mol) was used to rank the best poses for each ligand. The last step is post-docking analysis – the top-ranked binding conformations were selected for each protein-ligand complex based on the lowest binding energy and interaction profile. Discovery Studio Visualizer [19] was used to analyze hydrogen bonding (HB), hydrophobic contact, electrostatic interaction, and van der Waals (vdW) at each binding site.

2. Results and Discussion

2.1 The PLKs PBD Pockets

The experimentally established binding pockets of PLK1-PBD are BS1, BS2, and BS3, while BS4 to BS7 are the computationally identified pockets in this study. We coupled the selected twenty molecules at these seven binding regions and examined their binding orientations within the PBD pockets. The protocol also involved the PBD of PLK2 and PLK3, yielding 420 PBD-ligand docking simulations. Table 1 presents the key binding pockets of the polo-box domain (PBD) of PLK1, PLK2, and PLK3, highlighting their volume and constituting residues. These binding sites (BS1–BS7) represent the most probable regions where small-molecule inhibitors can bind, influencing the function of the PBD. The results indicate that while PLK1, PLK2, and PLK3 share structural homology, their binding site composition and volume vary, which is crucial in determining ligand selectivity and binding affinity. Some binding sites are highly conserved across all three PLKs, while others exhibit unique structural features that may contribute to inhibitor specificity.

The size of each binding pocket affects ligand accommodation and binding strength. Larger pockets may provide greater flexibility in ligand binding, while smaller ones often require a precise molecular fit. The binding sites BS1, BS2, and BS3 of PLK1 are experimentally validated sites identified in earlier structural studies of PLK1. We predicted BS4–BS7 through docking, suggesting additional potential ligand-binding regions. In PLK1, BS1, known as the phosphopeptide binding pocket [20], contains crucial residues such as Lys413, Trp414, Asp416, Arg516, Phe535, His538, Lys540, and Arg557, which are essential for ligand recognition and stabilization (Table 1). The corresponding residues include Lys506, Trp507, Phe625, His627, His629, Lys631, and Arg650 in PLK2, along with Lys466, Trp467, Phe586, His590, Lys592, and Arg610 in PLK3. The volumes of BS1 in these proteins are 657, 212, and 427 Å3 in PLK1, PLK2, and PLK3, respectively.

The BS2 is the Tyr-rich pocket, an experimentally validated binding pocket available in PLK1–PLK3 [13]. This pocket is cryptic and relatively smaller than BS1, with a volume of 85 ų in PLK1, 65 ų in PLK2, and 46 ų in PLK3, and its ligand accessibility is dependent on the proteins' dynamics [13, 21, 22]. Despite its smaller size, BS2 has hydrophobic residues that enable ligand stabilization, including Tyr417, Tyr421, Asn470, Ser471, Leu472, Met473, Leu478, Tyr481, Tyr485, and His489 in PLK1. PLK2 contains a similar structural Tyr motif with Tyr510, Tyr514, Tyr574, and Tyr578, while PLK3 features Tyr470, Lys473, and Tyr538 at corresponding positions (Table 1).

BS3 (the W-F pocket, Figure 2) is another experimentally validated binding site, with a volume of 102 ų in PLK1, 80 ų in PLK2, and 1014 ų in PLK3. It comprises a series of hydrophobic and polar residues facilitating ligand binding. In PLK1, key residues Ile406, Pro407, Trp410, Pro499, Arg500, Thr551, and Phe559 create a stabilizing environment for ligands. The corresponding residues in PLK2 include Ser500, Trp503, Val592, Thr593, Thr644, Ile646, and Ser652, while PLK3 features Pro458, Leu461, Trp463, Ser481, Arg610, and Ala612. The W-F was coined from Trp410 and Phe559 residues that stabilize the ligand at this allosteric pocket after distorting the PLK1-PBD substrate recognition mechanism [3]. This pocket is specific because the position analogous to the PLK1 Phe559 residue has Ser652 in PLK2 and Ala612 in PLK3 (Table 1). Exploring the W-F could enable the design of selective inhibitors that preferentially target PLK1 over PLK2 and PLK3, offering a promising strategy for isoform-specific therapeutic intervention in PLK1-expressed cancer types.

Table 1: Top pockets and volume of polo-like kinases (PLKs: PLK1, PLK2, and PLK3) polo-box domain (PBD) identified from the literature and predicted through docking approach showing constituting residues per binding site (BS: BS1, BS2, BS3, BS4, BS5, BS6, and BS7)

BS4, a potential binding pocket in this study, overlaps BS3 with shared residues like Ile406 and Pro407 in PLK1. Unique residues in BS4 of PLK1-PBD are Ile408, Arg500, Glu504, Ala506, Arg507, Arg512, Ser526, Asn527, Ser529, Cys544, Leu546, and Met547. Our result is consistent with previously computationally identified [23, 24] prominent pockets (BS4 and BS5) of PLK1-PBD. Phe501, Thr593, Arg597, Arg599, Leu600, Trp605, Gly619, Thr620, Gln622, Gln637, Glu639, and Glu640 form a dynamic environment with potential ligand-binding interactions in PLK2-PBD BS4. Corresponding BS4 residues in PLK3 are Leu461, Glu553, Val557, Ala559, Pro560, Glu565, Asp579, Gly580, Val582, Gly597, Glu559, and Pro600. The large volume of BS4 in PLK1 (788 ų) compared to PLK2 (293 ų) and PLK3 (61 ų) implies that BS4 in PLK1 can accommodate larger or more diverse ligands (Table 1). The unique residue composition in each isoform suggests that ligands designed for PLK1-BS4 may not effectively bind PLK2 or PLK3, offering a potential strategy for selective inhibitor development. The overlap between BS3 and BS4 (e.g., shared residues like Ile406 and Pro407) suggests that BS4 may function as an allosteric site, potentially influencing substrate recognition or protein-protein interactions in PLK1.

BS5 is one of the most structurally variable pockets across the PLKs. In PLK1, key residues include Ser390, Leu394, Val395, Gln397, Glu398, Glu401, Tyr552, Ile553, Asp554, Asp558, Phe559, Arg560, Tyr562, Cys573, Lys574, and Leu576. In PLK2, corresponding residues Leu483, Pro487, Glu488, Asp490, Cys491, Lys494, Tyr645, Ile646, Asn647, Ile651, Ser652, Thr653, Phe655, Ser666, Ser667, and Leu669. Corresponding PLK3 residues for this pocket are Ala443, Pro447, Ala448, Gln450, Asn451, Pro454, Phe605, Val606, Ala607, Ser611, Ala612, Cys613, Tyr615, Ser626, Pro627, and Leu629 (Table 1). The large BS5 volume in PLK1 (152 ų) compared to PLK2 (85 ų) and PLK3 (35 ų) suggests that PLK1 can accommodate larger or more diverse ligands, whereas PLK3's BS5 is much more restricted. The low conservation of residues across the three isoforms implies that ligands targeting BS5 in PLK1 may not effectively bind BS5 in PLK2 or PLK3, offering an opportunity for highly selective inhibitors. Leu576 (PLK1), Leu669 (PLK2), and Leu629 (PLK3) are the only conserved residues, suggesting that interactions involving this residue may be critical for ligand binding in all three isoforms.

BS6 is relatively conserved in PLK1 and PLK2 (408 ų vs. 350 ų) but is significantly smaller in PLK3 (154 ų). In PLK1, the key interacting residues include Cys428 Asp429, Asn430, Asn446, Gly494, Glu501, Gly502, Glu504, and Leu505, which are related to Ser521, Asp522, His523, Pro539, Gly587, Asp594, Ile595, Arg597, and Pro598 in PLK2 (Table 1). The available region corresponding to BS6 is buried in PLK3 with residues like Arg484, Ser498, Asn500, Lys502, Thr503, His505, Ser517, Glu554, and Glu556 available for ligand interaction. The buried nature of BS6 in PLK3 suggests that ligand accessibility is more restricted, potentially making it a less favorable binding site than PLK1 and PLK2.

BS7 is a moderately sized binding site with volumes of 108 ų in PLK1, 350 ų in PLK2, and 197 ų in PLK3. Key residues in PLK1 include Gln426, Val432, Arg483 and Glu488, while PLK2 contains structurally analogous residues such as Gln519, Val525, Ser576, and Glu581, with Gln479, Val485, Ala536, and Gln541 in PLK3 (Table 1). Structural variations indicate that BS7 could be available for noncompetitive inhibitor attachment in PLK1. The smaller BS7 in PLK1 implies that inhibitors targeting this pocket may need higher binding specificity to avoid off-target effects in PLK2 or PLK3.

3.2 Ligand Interaction with PLKs PBD

Ligand interaction at potential binding pockets presents heterogeneity across the three PLKs' polo-box domains. We used some ligands to depict the interaction network of these 20 PLK1-PBD inhibitors (Figures 3 and 4). Compound 3v in BS1 of PLK1 formed HB through its oxygen and sulfur atoms with Arg557 and Lys413 side chain nitrogen atoms, which may contribute to moderate ligand stability. Besides the weak vdW contacts, the mixed hydrophobic interactions between this ligand and BS1 residues are alkyl, π-π stacked, and π-alkyl, including Leu491 forming two different hydrophobic alkyl and π-alkyl interactions (Figure 3a). Critical nonbonded interactions in BS1 of PLK2 involve four hydrogen bonds and five mixed hydrophobic interactions, with Tyr626 initiating three distinct contacts with 3v (Figure 3b). This outcome suggests stronger polar interactions, which may enhance ligand affinity in PLK2 (Figure 5). In PLK3, BS1 residues establish four characteristic carbon-hydrogen bonds through Tyr467, Lys545, Lys592, and Arg610. Additionally, Tyr467 forms a π-donor hydrogen bond and an alkyl hydrophobic contact (Figure 3c). The greater number of hydrogen bonds in PLK2 and PLK3 implies a stronger binding affinity in these isoforms than in PLK1, as depicted from the docking scores of 3v with –7.5 kcal/mol in PLK2 and –7.0 kcal/mol in PLK3 vs –5.7 kcal/mol in PLK1 (Figure 5).

Polotyrin was carefully designed [13] to bind at the Tyr-rich pocket of PLK1-PBD, but it also presents a unique perspective when coupled with PLK2 and 3 (Figures 3d to 3f). Polotyrin formed various electrostatic interactions at the BS2 of all three PLKs, including an attractive charge through Lys474 in PLK1, two salt bridges (via Lys541) and a π-cation (from Lys573) in PLK2, and a π-anion via Tyr538 in PLK3. At least one of the critical tyrosine residues in this pocket forms mixed hydrophobic interactions with Polotyrin (Figures 3d to 3f). Each PLK also forms an HB with this potential inhibitor. Again, our docking analysis (Figure 5) favors PLK2 –6.7 kcal/mol over PLK1 (–6.4 kcal/mol). The drivers of this improved binding affinity in PLK2 could be more electrostatic interactions than others. Although the predicted binding energies do not follow the experimental [13, 25] trends of PLK1-PBD selectivity, our result at the molecular level suggests that several nonbonded interactions contribute significantly to ligand stabilization in PLK2, potentially influencing binding kinetics and residence time. However, additional factors such as conformational dynamics, solvent effects, and induced fit mechanisms may be crucial to driving the experimentally observed PLK1-PBD selectivity through the established pockets (BS1 and BS2).

The conserved Trp410, Trp503, and Trp463 in PLK1, 2, and 3 formed π-π stacked interactions with Allopole-A at the W-F (BS3) allosteric pocket (Figures 3g to 3i). PLK2 also forms two unique π-lone pair interactions with Allopole-A via Ser521 and Leu589 (Figure 3h). The predicted binding affinities of Allopole-A to the W-F pocket are –5.5 kcal/mol in PLK1-PBD, –4.6 kcal/mol in PLK2-PBD, and –5.2 kcal/mol in PLK3-PBD (Figure 5), which is close to the experimental [3] result of PLK1 selectivity over others.

The presence of polar and charged residues (e.g., Arg500, Glu504, Arg507, Figure 4a) in PLK1-PBD BS4 enables strong hydrogen bonding with Kbjk557 [26], corresponding PLK2 residues Thr593, Arg597, and Leu600 form HBs and vdW contacts (Figure 4b). BS4's smallest-sized (61 ų) pocket in PLK3 undergoes structural rearrangements upon Kbjk557 binding, pushing the ligand to interact with non-matching residues Glu459, Asp579, Glu599, Pro600, and Leu602 (Figure 4c). The order of the binding energy values –7.2, –7, and –6 kcal/mol in PLK1, PLK2, and PLK3 (Figure 5) align with the decreasing volume size of BS4 pockets in these PLKs (Table 1).

Figure 1: Zoomed-in interaction network of inhibitors in BS1–BS3 within the polo-box domains of PLK1–PLK3.

Although BS5 has a lower volume size (Table 1), its flat surface area allows several residues to interface ligands at this pocket (Figures 4d to 4f). An-329 [27] interacts in PLK1-PBD BS5 through Tyr552 (vdW), Asp554 (HB), Arg560 (carbon HB), and Cys573 (π-sulfur), the related residues in PLK2 are Tyr645 (vdW), Asn647 (HB), Thr653 (HB), and Ser666 (carbon HB) with Phe605 (vdW), Ala607 (vdW), Cys613 (π-sulfur), and Ser626 (HB) in PLK3. The binding scores of An-329 with PLK1, PLK2, and PLK3 in BS5 are –6.6, –6.3, –6.5 kcal/mol (Figure 5), respectively. BS6 forms several vdW and mixed hydrophobic contacts with Poloppin-II [28], including halogen bonding from Arg500 and one of its fluorine atoms in PLK1-PBD (Figure 4g), π-cation through Arg597 in PLK2 (Figure 4h), and π-anion via Glu556 in PLK3 (Figure 4i). The docking scores of Poloppin-II in BS5 of PLK1, PLK2, and PLK3 are –6.7, –5.3, –4.6 kcal/mol (Figure 5). Note, BS4, BS5, and BS6 have characteristic cysteine residues Cys544, Cys573, and Cys428 in PLK1-PBD (Figure 4a, 4d, and 4g). These Cys are not conserved in PLK2 and 3, so exploring them for selective and specific PLK1 inhibitor designs could be profound.

Figure 1: Zoomed-in interaction network of inhibitors in BS4–BS7 within the polo-box domains of PLK1–PLK3. Legends are the same as in Figure 3.

The replacement of Arg483 in PLK1-PBD BS7 (Figure 4j) with Ser576 in PLK2 (Figure 4k) reduces the interaction strength from π-cation to vdW. In PLK3, Ala536 (vdW) replaces Arg483, reducing electrostatic interactions (Figure 4l) that could impact ligand stabilization in BS7. The docking scores (Figure 5) of Zinc20503376 [29] in BS7 are –6.9 kcal/mol (PLK1), –6.6 kcal/mol (PLK2), and –6.1 kcal/mol (PLK3).

3.3 Docking Scores

The heatmap in Figure 5 shows docking scores across the studied seven binding pockets (BS1–BS7) of PLK1, PLK2, and PLK3 polo-boxes. The high binding affinity of these molecules (Zinc205037376, Polotyrin, and Kbjk557) is likely due to their unique structure with aromatic ring systems and hydrophobic moieties, allowing for stronger interactions with nonpolar residues within the binding sites [30]. In contrast, Thymoquinone (TQ) and Mcc1019 displayed relatively poor binding across most PLK proteins and binding sites. These inhibitors frequently exhibited docking scores around –5 kcal/mol, indicating weaker interactions and less stable binding.

BS1 remains the most favorable binding site, with a larger proportion of inhibitors scoring in the –7 to –9 kcal/mol range due to its well-defined pocket structure, reducing entropy loss upon binding. Interestingly, none of the inhibitors shows favorable binding affinity within the BS1 of PLK1-PBD over PLK2 and 3. Of the twenty inhibitors, Poloppin-II and TQ in the BS2 of PLK1 yielded stronger binding energies than PLK2 and 3. BS3, BS4, and BS5 pockets accommodate strong and moderate binders, with several inhibitors reaching scores in the –5 to –7 kcal/mol range. The binding of these potential inhibitors at the W-F pocket (BS3) results in 15 molecules showing appreciable affinities in PLK1 over PLK2 and 3. The docking output also favors PLK1, with 17 compounds binding strongly at BS4 and BS5. All selected molecules bind strongly at BS6 of PLK1 over corresponding pockets of PLK2 and 3, while 13 molecules form stronger interactions at the BS7 of PLK1 (Figure 5).

4. Conclusion

Beyond the experimentally validated BS1–BS3, our computational approach identified additional binding pockets (BS4–BS7) with varying conservation and structural divergence among the three PLK isoforms. BS4, for example, is a large and dynamic pocket with a volume of 788 ų in PLK1, 293 ų in PLK2, and 61 ų in PLK3. The unique BS4 residues His373, Cys405, Ile408, Arg500, Glu504, and Met547 create additional binding interactions. The presence of distinct residues suggests potential differences in ligand accommodation and binding stability across the PLKs. BS5–BS7 further illustrate the structural and volumetric variations between PLK isoforms. BS7, for instance, is moderately sized, with volumes of 108 ų in PLK1, 350 ų in PLK2, and 197 ų in PLK3. Despite its relatively small size in PLK1, the pocket contains key residues such as Gln426, Val432, Arg483, and Glu488, while PLK2 and PLK3 feature analogous residues with minor differences in their polarity and side-chain orientation. The variations in pocket size and residue composition across these binding sites suggest that ligand binding preferences may be isoform-dependent, influencing the selectivity and efficacy of potential inhibitors.

Our findings highlight the importance of structural differences in the PBD binding pockets of PLK1, PLK2, and PLK3. While some binding sites, such as BS1 and BS2, are highly conserved, others, like BS3 to BS6, show significant variation, which could be explored for designing selective inhibitors. Identifying BS4–BS7 through computational analysis further expands the potential druggable sites in PLK1. These insights provide a structural framework for developing small-molecule inhibitors that selectively target the polo-box domains of PLK isoforms, ultimately contributing to targeted cancer therapeutics advancement.

Figure 1: Heatmap representation of molecular docking scores (kcal/mol) for selected small molecules across the seven primary binding sites (BS1–BS7) of the PLK1–PLK3 polo-box domain (PBD). Protein interaction with each inhibitor is shortened to 1, 2, and 3 for PLK1, PLK2, and PLK3, respectively. Each row represents a protein-ligand complex, while the columns correspond to binding site scores and their averages. Binding affinities are color-coded, with blue indicating stronger binding (more negative values) and red indicating weaker binding (more positive values).

References

1.         Stafford, J.M., Wyatt, M.D., McInnes, C.: Inhibitors of the PLK1 polo-box domain: Drug design strategies and therapeutic opportunities in cancer. Expert Opin. Drug Discov. 18, 65–81 (2023)

2.         Strebhardt, K., Becker, S., Matthess, Y.: Thoughts on the current assessment of Polo-like kinase inhibitor drug discovery. Taylor & Francis (2015)

3.         Park, J.-E., Kirsch, K., Lee, H., Oliva, P., Ahn, J.I., Ravishankar, H., Zeng, Y., Fox, S.D., Kirby, S.A., Badhwar, P., Andresson, T., Jacobson, K.A., Lee, K.S.: Specific inhibition of an anticancer target, polo-like kinase 1, by allosterically dismantling its mechanism of substrate recognition, (2023)

4.         Baxter, M.L.: Selectively targeting Polo-like Kinase 1 (PLK1) using novel inhibitors of the polo-box domain, (2016)

5.         Lee, S.-Y., Jang, C., Lee, K.-A.: Polo-Like Kinases (Plks), a Key Regulator of Cell Cycle and New Potential Target for Cancer Therapy. Dev. Reprod. 18, 65–71 (2014). https://doi.org/10.12717/DR.2014.18.1.065

6.         Kim, S., Chen, J., Cheng, T., Gindulyte, A., He, J., He, S., Li, Q., Shoemaker, B.A., Thiessen, P.A., Yu, B.: PubChem 2023 update. Nucleic Acids Res. 51, D1373–D1380 (2023)

7.         Rose, P.W., Prlić, A., Altunkaya, A., Bi, C., Bradley, A.R., Christie, C.H., Costanzo, L.D., Duarte, J.M., Dutta, S., Feng, Z.: The RCSB protein data bank: integrative view of protein, gene and 3D structural information. Nucleic Acids Res. gkw1000 (2016)

8.         Hanwell, M.D., Curtis, D.E., Lonie, D.C., Vandermeersch, T., Zurek, E., Hutchison, G.R.: Avogadro: an advanced semantic chemical editor, visualization, and analysis platform. J. Cheminformatics. 4, 1–17 (2012)

9.         Liu, Y., Yang, X., Gan, J., Chen, S., Xiao, Z.-X., Cao, Y.: CB-Dock2: Improved protein–ligand blind docking by integrating cavity detection, docking and homologous template fitting. Nucleic Acids Res. 50, W159–W164 (2022)

10.       Yin, Z., Song, Y., Rehse, P.H.: Thymoquinone Blocks pSer/pThr Recognition by Plk1 Polo-Box Domain As a Phosphate Mimic. ACS Chem. Biol. 8, 303–308 (2013). https://doi.org/10.1021/cb3004379

11.       Kim, J.H., Ku, B., Lee, K.S., Kim, S.J.: Structural analysis of the polo-box domain of human Polo-like kinase 2: Structure of the Plk2 Polo-Box Domain. Proteins Struct. Funct. Bioinforma. 83, 1201–1208 (2015). https://doi.org/10.1002/prot.24804

12.       Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A.: Highly accurate protein structure prediction with AlphaFold. Nature. 596, 583–589 (2021)

13.       Sharma, P., Mahen, R., Rossmann, M., Stokes, J.E., Hardwick, B., Huggins, D.J., Emery, A., Kunciw, D.L., Hyvönen, M., Spring, D.R.: A cryptic hydrophobic pocket in the polo-box domain of the polo-like kinase PLK1 regulates substrate recognition and mitotic chromosome segregation. Sci. Rep. 9, 15930 (2019)

14.       Trott, O., Olson, A.J.: AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J. Comput. Chem. 31, 455–461 (2010)

15.       Eswar, N., Eramian, D., Webb, B., Shen, M.-Y., Sali, A.: Protein structure modeling with MODELLER. Struct. Proteomics High-Throughput Methods. 145–159 (2008)

16.       Humphrey, W., Dalke, A., Schulten, K.: VMD: visual molecular dynamics. J. Mol. Graph. 14, 33–38 (1996)

17.       Spivak, M., Stone, J.E., Ribeiro, J., Saam, J., Freddolino, P.L., Bernardi, R.C., Tajkhorshid, E.: VMD as a platform for interactive small molecule preparation and visualization in quantum and classical simulations. J. Chem. Inf. Model. 63, 4664–4678 (2023)

18.       Huey, R., Morris, G.M., Forli, S.: Using AutoDock 4 and AutoDock vina with AutoDockTools: a tutorial. Scripps Res. Inst. Mol. Graph. Lab. 10550, 1000 (2012)

19.       BIOVIA, Dassault Systèmes, Discovery Studio 2021, San Diego: Dassault Systèmes, 2021.

20.       He, D., Huang, L., Xu, Y., Pan, X., Liu, L.: Molecular dynamics directed rational design and fluorescence binding assay of phosphopeptide ligands for PLK polo-box domain. Mol. Simul. 43, 176–182 (2017)

21.       Liu, F., Park, J.-E., Qian, W.-J., Lim, D., Gräber, M., Berg, T., Yaffe, M.B., Lee, K.S., Burke Jr, T.R.: Serendipitous alkylation of a Plk1 ligand uncovers a new binding channel. Nat. Chem. Biol. 7, 595–601 (2011)

22.       Śledź, P., Stubbs, C.J., Lang, S., Yang, Y.-Q., McKenzie, G.J., Venkitaraman, A.R., 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. 50, 4003–4006 (2011)

23.       Liao, C., Park, J.-E., Bang, J.K., Nicklaus, M.C., Lee, K.S.: Probing binding modes of small molecule inhibitors to the Polo-box domain of human Polo-like kinase 1. ACS Med. Chem. Lett. 1, 110–114 (2010)

24.       Bong, S.M., Moon, S., Cho, K.-H., Kim, K.-T., Lee, E.Y., Choi, J.-W., Lee, B.I.: A NCAPG2-Derived Phosphopeptide Selectively Binds to the Polo-Box Domain of PLK1 and Inhibits Cancer Cell Proliferation. Int. J. Pept. Res. Ther. 25, 1397–1403 (2019)

25.       Yun, T., Qin, T., Liu, Y., Lai, L.: Identification of acylthiourea derivatives as potent Plk1 PBD inhibitors. Eur. J. Med. Chem. 124, 229–236 (2016)

26.       Gunasekaran, P., Yim, M.S., Ahn, M., Soung, N.-K., Park, J.-E., Kim, J., Bang, G., Shin, S.C., Choi, J., Kim, M., Kim, H.N., Lee, Y.-H., Chung, Y.-H., Lee, K., EunKyeong Kim, E., Jeon, Y.-H., Kim, M.J., Lee, K.-R., Kim, B.-Y., Lee, K.S., Ryu, E.K., Bang, J.K.: Development of a Polo-like Kinase-1 Polo-Box Domain Inhibitor as a Tumor Growth Suppressor in Mice Models. J. Med. Chem. 63, 14905–14920 (2020). https://doi.org/10.1021/acs.jmedchem.0c01451

27.       Zhou, Y., Yan, F., Huo, X., Niu, M.-M.: Discovery of a potent PLK1-PBD small-molecule inhibitor as an anticancer drug candidate through structure-based design. Molecules. 24, 4351 (2019)

28.       Narvaez, A.J., Ber, S., Crooks, A., Emery, A., Hardwick, B., Almeida, E.G., Huggins, D.J., Perera, D., Roberts-Thomson, M., Azzarelli, R.: Modulating protein-protein interactions of the mitotic polo-like kinases to target mutant KRAS. Cell Chem. Biol. 24, 1017–1028 (2017)

29.       Abdelfatah, S., Berg, A., Böckers, M., Efferth, T.: A selective inhibitor of the Polo-box domain of Polo-like kinase 1 identified by virtual screening. J. Adv. Res. 16, 145–156 (2019). https://doi.org/10.1016/j.jare.2018.10.002

30.       Brylinski, M.: Aromatic interactions at the ligand–protein interface: Implications for the development of docking scoring functions. Chem. Biol. Drug Des. 91, 380–390 (2018)