Identification of novel potent peptide inhibitors targeting the polo-box domain of PLK1: structure-based pharmacophore modelling, virtual screening, molecular docking, molecular dynamics study and biological evaluation

Abstract

Multiple myeloma, a hematological malignancy, shows PLK1 overexpression in cells correlates with poor prognosis, suggesting PLK1 as a potential therapeutic target. In this study, we discovered five peptides (PLs 1–5) targeting the polo box domain (PBD) of PLK1 through an integrated virtual screening strategy. MST assays confirmed that PLs 1–5 had strong binding affinity for PLK1, especially PL-1 (K_d = 3.11 ± 0.05 nM). Meanwhile, the kinase selectivity assay showed that the PL-1 had no significant inhibitory effects on a panel of other kinases. Molecular dynamics simulation further demonstrated the structural stability of PL-1 and PLK1 complex. Notably, PL-1 displayed potent antiproliferative efficacy against U266 multiple myeloma cells (IC₅₀ = 0.09 ± 0.01 µM). PL-1 showed high intracellular uptake capacity. In addition, PL-1 exhibited good biostability in human serum and liver microsomes. Taken together, PL-1 is a potent and highly selective antitumor agent with considerable therapeutic promise for multiple myeloma.

Introduction

The escalating global incidence of cancer has imposed a substantial health burden^1,2. Multiple myeloma, the second most common hematological malignancy, is a malignancy caused by the abnormal proliferation of plasma cells^3,4. Multiple myeloma is defined by the abnormal growth of plasma cells within the bone marrow, resulting in destructive bone damage, anaemia, hypercalcemia and renal dysfunction^5–7. Currently, a variety of non-targeted therapeutic agents have been approved for the treatment of multiple myeloma, including alkylating agents, corticosteroids, and proteasome inhibitors^8,9. Although these therapeutic agents are effective in controlling disease progression and prolonging survival in multiple myeloma patients, their non-specific mechanisms of action often result in severe side effects^3,10,11. Therefore, targeting specific proteins involved in the survival and proliferation of myeloma cells represents a promising strategy to enhance therapeutic effectiveness and reduce treatment-related toxicity.

Polo-like kinase 1 (PLK1) is a member of the serine/threonine kinase family that regulates a variety of cellular functions including cell cycle progression, differentiation and survival^12,13. PLK1 was shown to be highly expressed in mitotically active cells and to have a critical function in the mitotic process¹⁴. Thus, PLK1 is often expressed at high levels in the majority of human cancers, and its overexpression correlates with poor prognosis for cancer patients^15,16. Studies have shown that the upregulation of PLK1 expression in multiple myeloma is significantly associated with poor prognosis, and that some PLK1 inhibitors can effectively suppress multiple myeloma cell proliferation by inhibiting PLK1 kinase activity^17,18. Thus, targeting PLK1 offers a potential therapeutic strategy for the treatment of multiple myeloma.

PLK1 consists of a highly conserved kinase catalytic domain (KD) and a polo box domain (PBD)¹⁹. This PBD acts as a distinctive phosphopeptide-binding module and is critical for the identification of PLK1 substrates^20,21. In particular, PLK1 localises subcellularly via its PBD at different stages of the cell cycle and exerts regulatory functions at multiple stages of mitosis^22,23. Studies have shown that phosphate-dependent substrates can regulate PLK1 function by binding PBD, stimulating PLK1 kinase activity and regulating the cell cycle^24,25. Therefore, the PBD with specificity for PLKs is a suitable target for the design of selective PLK1 inhibitors²⁶. Several reports have shown that various small molecule compounds, such as Thymoquinone²⁷, Poloxin²⁸, and MCC1019²⁹ were identified as PLK1-PBD inhibitors (Figure 1). However, these compounds typically have limited specificity for PLK1-PBD and require higher doses to induce anti-PLK1 effects, restricting their potential for future drug development³⁰. Compared to small molecules, therapeutic peptides have great potential in cancer therapy due to their unique advantages of low molecular weight, low toxicity and specific targeting of tumour cells^31,32. Meanwhile, peptides can better adapt to the large interaction surfaces of proteins due to their distinct chemical and physical properties^33,34. Therefore, the discovery of peptide drugs is expected to offer new opportunities for targeted cancer treatment. In the previous study, a peptide inhibitor (peptide 5) targeting the PLK1-PBD was reported³⁵, but subsequent investigations revealed several limitations of peptide 5 that have limited its further therapeutic application prospects: (1) poor stability in serum; (2) low selectivity for PLK1; (3) weak binding affinity to PLK1. To overcome these limitations, we aimed to develop new peptide inhibitors targeting PLK1-PBD with enhanced therapeutic potential through structure-based virtual screening.

Figure 1.

Reported PLK1-PBD inhibitors.

Computer-aided drug design is widely used to screen large databases of biologically active compounds against specific protein targets due to its cost-effectiveness and efficiency³⁶. Currently, structure-based virtual screening is increasingly influencing the identification and optimisation phases of lead discovery^37,38. Pharmacophore models depict the essential chemical features necessary for ligands to bind to target proteins and can be used to identify novel active compounds within large compound databases^39–41. Molecular docking is a computational method used to predict ligand-receptor interactions, identify ligand binding patterns within the binding site and assess the binding affinity between the ligand and receptor^42,43. Thus, the combined application of pharmacophore modelling and molecular docking can complement the strengths of each method, thereby improving the accuracy of virtual screening^44,45. In our previous studies, several novel inhibitors have been successfully identified by structure-based virtual screening approach^46–51. In this study, we identified a potent peptide inhibitor (PL-1) targeting the PLK1-PBD through an integrated virtual screening approach. PL-1 had significant binding affinity for PLK1-PBD and exhibited high selectivity towards PLK1. Notably, PL-1 showed significant antiproliferative activity against U266 multiple myeloma cells and high intracellular uptake capacity. Additionally, it showed good biostability in human serum and liver microsomes. These results suggest that PL-1 may be a novel therapeutic approach for multiple myeloma and warrant further investigation.

Materials and methods

General chemistry

PLs 1–5, peptide 5, and BI 2536 were purchased from WuXi AppTec (Shanghai, China), and all peptides had a purity of greater than 98% (Supplementary materials). The PLK1-PBD was purchased from Abcam (Cambridge, MA, USA). Roswell Park Memorial Institute (RPMI)-1640, foetal bovine serum (FBS), penicillin/streptomycin, MTT reagent, human serum and human liver microsomes were obtained from Gibco (Grand Island, New York, USA). The MO-L011 Protein Labelling Kit RED-NHS 2nd Generation was purchased from NanoTemper Technologies (Munich, Germany). Rhodamine-labelled dextran and Hoechst 33342 were purchased from Sigma (St. Louis, MO, USA). Dimethyl sulfoxide (DMSO), Tris, NaCl, and Tween 20, and other chemical reagents were obtained from Aladdin Reagent (Shanghai, China).

Cell lines and culture conditions

The human U266 multiple myeloma cells (catalog number: TIB-196) were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were cultured in RPMI-1640 medium supplemented with 10% FBS, 80 U/L penicillin, and 0.08 mg/mL streptomycin, then incubated at 37 °C in a humidified atmosphere containing 5% CO₂.

Pharmacophore construction and validation

The crystal structure of the PLK1 and ligand complex (PDB ID: 3P37) was retrieved from the Protein Data Bank (PDB). The virtual screening procedures were conducted using the Molecular Operating Environment (MOE) software. First, initial structure preparation was performed through the QuickPrep module, including the elimination of unbound water molecules, assignment of partial charges, and energy minimisation. Then, molecular interactions between PLK1 and ligand were analysed using the Ligand Interaction module to identify the critical binding sites. Following this, the key interaction point in the ligand was identified and pharmacophore features were generated using the Pharmacophore Query Editor. These features included hydrogen bond acceptors, hydrogen bond donors and hydrophobic centroid.

The Gunner-Henry scoring method was employed to verify the reliability of the constructed pharmacophore model. A database containing 384 inactive peptides and 8 active peptides was established, with the 8 active peptides derived from published literature^47,52. Subsequently, the pharmacophore model was used as a three-dimensional retrieval criterion, and the database was screened using the Pharmacophore Search module in MOE software. Finally, the key validation parameters were calculated. The Goodness of hit score (GH) ranges from 0 to 1, with 0 representing an invalid model and 1 representing an ideal model.

Virtual screening

An in-house database comprising 15,625 peptides was created using combinatorial chemistry methods⁵³. The 2D structures of the peptides were converted into 3D conformations using the Conformation Import tool in MOE. The MMFF94x force field was utilised for conformation generation, and the Energy Minimisation tool was employed to ensure structural stability. Pharmacophore models were utilised to conduct virtual screening of the database via Pharmacophore Search panel. Docking validation was conducted using the pose selection method. The ligand was removed from the PLK1 structure (PDB ID: 3P37), then re-docked into the active site of PLK1-PBD via Dock tool in MOE. Subsequently, the re-docked complex was superimposed onto the original co-crystallized complex to verify the reliability of the molecular docking protocol. Next, molecular docking was carried out on the identified hits using the Dock tool in MOE. The binding free energy of these peptides to PLK1-PBD was evaluated using the Docking Scoring tool, with docking parameters set to the Triangle Matcher method and the London dG scoring algorithm. A lower docking score was indicative of a higher level of binding affinity between the peptides and the target protein.

MST assays

Microscale thermophoresis (MST) assays were conducted as previously reported⁵⁴. The reaction was carried out in an assay buffer (pH 8.0) consisting of 20 mM Tris (pH 8.0), 200 mM NaCl, and 0.1% (v/v) TWEEN20. PLK1-PBD was labelled using the MO-L011 Protein Labelling Kit RED-NHS 2nd Generation according to the manufacturer’s protocol. The concentration of PLK1-PBD was adjusted according to the fluorescent counts from the instrument to ensure stable and detectable signals, which typically ranged from 1200 to 1500 at 20% LED power. The test compounds were serially diluted to concentrations of 0.8, 1.6, 3.2, 6.4, 12.8, 25.6, 51.2, 102.4, and 204.8 nM. The labelled PLK1-PBD protein was then mixed at a ratio of 1:1 with serially diluted compounds. After five minutes incubation at room temperature, the mixture was filled into the capillaries for detection using the Monolith NT.115 instrument. Measurements were performed in temperature jump analysis mode with 20% MST power. The dissociation constant (K_d) was calculated using MO. Affinity Analysis software with the K_d formula: f(c) = unbound + (bound – unbound)/2 × (FluoConc + c + K_d – Sqrt ((FluoConc + c + K_d)² – 4FluoConc × c)).

Selectivity assays

The kinase selectivity profile of PL-1 was assessed by examining its inhibitory activity against a panel of kinases, including other members of the PLKs family. The experiment was carried out by ICE Bioscience Inc. (Beijing, China). Human kinases were incubated in a solution containing 50 mM Tris (pH 7.5), 0.1 mM ethylene glycol-bis(β-aminoethyl ether)-N,N,N’,N’-tetraacetic acid, 0.1% β-mercaptoethanol, 250 μM peptide (3 μM ATF2), 10 mM magnesium acetate, [γ-³³P-ATP], and 10 μM test compounds. The reaction was activated through the addition of MgATP mixture, then allowed to incubate for 40 min, and eventually terminated by 3% phosphoric acid solution. Subsequently, 10 μL of the reaction solution was spotted onto a P30 filter mat, rinsed three times with 75 mM phosphoric acid and once with methanol, dried, and quantified using scintillation counting. The relative kinase activity was calculated by taking the kinase activity without inhibitors (set as 100%) as a benchmark. Inhibition rate (%) = 1 - kinase activity (%).

Molecular dynamics (MD) simulation

The crystal structure of PLK1 (PDB ID: 3P37) was retrieved from the PDB. The ligands PL-1 and PL-2 were individually docked into the PLK1-PBD to generate complex structures using MOE. MD simulations were performed using GROMACS (version 2022). The topological modelling of the complex was carried out under the AMBER99SB-ILDN force field. The system was solvated using the SPC/E water model in a 1.0 nm cubic box, with system charge neutralisation achieved by appropriate ion placement (Na⁺/Cl^-). The steepest descent algorithm was utilised to perform energy minimisation for 5000 steps. The system temperature was stabilised at 300 K during a 100 ps NVT simulation and then a 100 ps NPT simulation was performed to achieve pressure equilibrium at 1 bar. Ultimately, 100 ns MD simulations were performed. The resulting data were analysed using Prism 8.0 software.

MTT assay

Cells were plated in 96-well culture plates at a seeding density of 5 × 10⁴ cells/well and incubated overnight. The test compounds were then added at specified concentrations (0.02–2.56 µM) and the plates were maintained at 37 °C for 72 h. After incubation, the supernatant was carefully removed, and the cells were treated with MTT solution (5 mg/mL) for 4 h. After removal of the supernatant, DMSO was added to dissolve the formazan crystals. Absorbance was measured at 570 nm using a microplate reader. Dose-response curves were generated using non-linear regression analysis, and IC₅₀ values were calculated. Data processing and analysis were conducted using Prism 8.0 software.

Confocal microscopy

The cells were seeded in glass-bottomed microwell plates and cultured overnight. After two gentle washes with PBS, the cells were incubated at 37 °C for 2 h in complete medium supplemented with 1 μM FITC-labelled peptide and 1 mg/mL rhodamine-labelled dextran (dextran^Rho). The medium was then removed and the cells were washed three times with PBS. Subsequently, the cells were incubated for 30 min in PBS containing 10 μg/mL Hoechst 33342. Finally, the cells were imaged using an LSM 700 laser scanning confocal microscope.

Serum stability test

Diluted human serum (25%) supernatant was prepared by centrifugation. PL-1 was added to a final concentration of 5 μM and incubated at 37 °C for 6 h. Subsequently, 200 μL of the reaction solution was taken at different time points and mixed with an equal volume of acetonitrile, and allowed to stand at 4 °C overnight. After centrifugation, the supernatant was collected for HPLC analysis, and the residual rate was calculated by peptide characteristic peak area integration.

Microsomal stability test

The test compounds (at a concentration of 1 μM) were mixed with 0.5 mg/mL human liver microsomes, 1 mM MgCl₂, and 0.1 M phosphate buffer (pH 7.4). The reaction was initiated by adding 1 mM reduced NADPH, after which the mixture was incubated at 37 °C. Samples were collected at 0, 5, 10, 20, 40, and 60 min, and the reaction was terminated by extraction with cold acetonitrile. Following centrifugation, the samples were analysed via liquid chromatography-tandem mass spectrometry (LC-MS/MS).

Results and discussion

Pharmacophore construction and validation

In this study, to ensure that the PLK1-PBD crystal structure used for pharmacophore construction and molecular docking possesses sufficient structural accuracy and reliability, we have established reasonable screening criteria: resolution < 3 Å; R-value < 0.3; Ramachandran outliers = 0; and ligand-PLK1-PBD binding affinity not exceeding 0.3 μM (to screen for high-affinity ligands). A systematic evaluation showed that the PLK1-PBD crystal structure (PDB ID: 3P37) meets all criteria, making it suitable for pharmacophore construction and molecular docking. The detailed information of the crystal structure is shown in Table S1. Based on the structure of PLK1-PBD (PDB ID: 3P37), the interaction between PLK1 and ligand was analysed using MOE. The critical amino acids involved in PLK1-ligand binding interface offered valuable insights for the rational design of pharmacophore features targeting the PLK1-PBD. As shown in Figure 2(A), the Lys540, His538 and Trp414 residues of PLK1 formed hydrogen bond interactions with the ligand, leading to the identification of the hydrogen-bond acceptor features (F1, F2) and a hydrogen-bond donor feature (F3). Additionally, Tyr417, Tyr421, Leu478, Tyr481, Phe482, Tyr485, and Leu490 on the PLK1 binding site engaged in hydrophobic interactions with the ligand, which facilitated the establishment of a hydrophobic centroid feature (F4). Finally, the interactions of amino acids at the PLK1 binding site with the ligand were analysed, allowing the construction of four key pharmacophore features: (i) two hydrogen-bond acceptor features (F1 and F2: Acc), (ii) one hydrogen-bond donor feature (F3: Don), (iii) one hydrophobic centroid feature (F4: Hyd) (Figure 2A). These pharmacophore features represent the essential chemical characteristics necessary to screen the lead peptide against the PLK1-PBD interface.

Figure 2.

(A) The constructed pharmacophore features on the PLK1-PBD binding site. F1 and F2: hydrogen-bond acceptor (Acc, red color); F3: hydrogen-bond donor (Don, green color); F4: hydrophobic centroid (Hyd, yellow color). (B) The multi-step virtual screening workflow.

To verify the reliability of the constructed pharmacophore model, the model was evaluated using a test set database containing 392 peptides (including 8 active PLK1 inhibitors) with the GH score as a metric. As shown in Table 1, a series of validation parameters were calculated, including total hits (Ht), % yield of actives (recall), % ratio of actives (precision), enrichment factor (E), false negatives, false positives, and goodness of hit score (GH). From a total of 392 peptides, 9 were successfully identified as hits, with 8 of these being active compounds. The recall rate was 89%, while the precision rate reached 100%. A pharmacophore model is considered to exhibit good performance when its GH score surpasses 0.7. The GH score of the model was 0.91, indicating that the constructed pharmacophore model can effectively distinguish between active and inactive molecules.

Table 1.

Validation of pharmacophore model using the GH score method.

Parameter	Pharmacophore model
Total molecules in database (D)	392
Total number of actives in database (A)	8
Total hits (Ht)	9
actives hits (Ha)	8
% Yield of actives ((Ha/Ht) × 100)	89%
% Ratio of actives ((Ha/A) × 100)	100%
Enrichment factor (E) ((Ha × D)/(Ht × A))	44
False negatives (A − Ha)	0
False positives (Ht − Ha)	1
Goodness of hit score (GH)	0.91

Virtual screening

In this study, an integrated virtual screening strategy encompassing pharmacophore modelling, pharmacophore screening, docking screening, MD simulations and biological validation was performed to identify promising peptides from a peptide database (Figure 2B). First, the peptide database of 15,625 peptides underwent structural conversion from 2D to 3D conformation. Subsequently, a virtual screening was implemented using the established pharmacophore features to systematically evaluate the entire peptide database. This computational screening process identified 54 candidate peptides with root mean square deviation (RMSD) value of less than 1.0 Å that matched the F1-F4 pharmacophore features. Prior to performing molecular docking, docking validation was conducted using the pose selection method based on the crystal structure of PLK1-PBD (PDB ID: 3P37). As shown in Figure 3(A), a high degree of overlap was observed between the native and docked conformations, thereby confirming the reliability of this docking protocol. Following this, each of the 54 peptides was subjected to a molecular docking screen targeting the PLK1-PBD to identify the optimal peptide candidate. Binding affinity was assessed using the docking scores, with lower scores corresponding to higher affinity. As depicted in Figure 3(B) and Table 2, the top five peptides (PLs 1–5) with the lowest docking scores (ranging from −16.19 to −11.49 kcal/mol) were selected as optimal candidates from 54 peptides. Specifically, PL-1 had the lowest docking score among the five peptides, suggesting that PL-1 showed the strongest binding to PLK1-PBD. The sequences of PLs 1–5 are shown in Table 3 and the chemical structures of PLs 1–5 are shown in Figure 4.

Figure 3.

(A) The docking validation between the native pose (cyan) and the docked pose (purple). (B) The binding free energy (kcal/mol) of PLs 1–5.

Table 2.

The docking scores and structure-activity relationship analysis of PLs 1–5 with PLK1-PBD.

Name	Docking scores (kcal/mol)	Hydrogen bonding interactions		Hydrophobic interactions
		Amino acid involved	Hydrogen bond numbers
PL-1	−16.19 ± 0.19	Trp414	3	Tyr417, Tyr421, Leu478, Tyr481, Phe482, Tyr485, Leu490
		Asp416	2
		Arg516	1
		His538	1
		Lys540	1
PL-2	−13.94 ± 0.25	Trp414	3	Tyr417, Tyr421, Leu478, Tyr481, Phe482, Tyr485, Leu490
		Asp416	2
		Arg516	1
		His538	1
		Lys540	1
PL-3	−13.85 ± 0.17	Trp414	3	Tyr417, Tyr421, Leu478, Tyr481, Phe482, Tyr485, Leu490
		Asp416	2
		His538	1
		Lys540	1
PL-4	−13.20 ± 0.24	Trp414	3	Tyr417, Tyr421, Leu478, Tyr481, Phe482, Tyr485, Leu490
		Asp416	2
		His538	1
		Lys540	1
PL-5	−11.49 ± 0.33	Trp414	3	Tyr417, Tyr421, Leu478, Tyr481, Phe482, Tyr485, Leu490
		Asp416	2
		His538	1
		Lys540	1

The data are presented as the mean ± SD, n = 3.

Table 3.

The binding affinity of PLs 1–5 to PLK1-PBD and the antiproliferative activity of PLs 1–5 on U266 cells.

Name	Sequencea	PLK1-PBD (Kd, nM)	U266 (IC50, µM)
PL-1	Ac-XAPPEHSpTA-NH2	3.11 ± 0.05	0.09 ± 0.01
PL-2	Ac-FAPPEHSpTA-NH2	14.37 ± 0.12	0.35 ± 0.03
PL-3	Ac-WAPPLHSpTA-NH2	29.64 ± 1.13	0.72 ± 0.02
PL-4	Ac-HAPPLHSpTA-NH2	33.25 ± 1.42	0.84 ± 0.05
PL-5	Ac-LAPPLHSpTA-NH2	40.42 ± 2.07	0.97 ± 0.06
peptide 5	ΦNPPLHSpTA	67.03 ± 2.82	2.25 ± 0.11

^a PLs 1–5 undergo N-terminal acetylation and C-terminal amidation modifications. X represents 2-amino-3–(4-bromo-3-chlorophenyl) propanoic acid. Φ represents L-3,4-dichlorophenylalanine. The data are presented as the mean ± SD, n = 3.

Figure 4.

The chemical structures of PLs 1–5.

Structure-activity relationship analysis

To validate the screening results, a detailed analysis of the molecular interactions between PLs 1–5 and PLK1-PBD was performed. Figure 5 illustrates the 2D and 3D interactions of PLs 1–5 with PLK1-PBD. In addition, the docking scores and structure-activity relationships between PLs 1–5 and PLK1-PBD are detailed in Table 2. The amino acids Trp414, Asp416, Arg516, His538, and Lys540 in PL-1 and PL-2 formed eight hydrogen bond interactions with PLK1-PBD, while Trp414, Asp416, His538, and Lys540 in PLs 3–5 created seven hydrogen bonds with PLK1-PBD. These hydrogen bond interactions play a crucial role in anchoring PLs 1–5 to the PLK1-PBD binding site. In addition, structural analysis revealed that replacing Leu in peptide 5 with Glu (as in PL-1 and PL-2) (Table 3) formed one new hydrogen bond between PL-1 and PLK1-PBD, potentially increasing the binding affinity. Meanwhile, the PLs 1–5 engaged in hydrophobic contacts with Tyr417, Tyr421, Leu478, Tyr481, Phe482, Tyr485, and Leu490 of PLK1-PBD, which is crucial factor for binding affinity. Taken together, these findings demonstrate that the binding affinity between PLs 1–5 and PLK1-PBD is stabilised by hydrogen bonds and hydrophobic interactions, facilitating the embedding of PLs 1–5 into the PLK1-PBD binding site. Notably, PL-1 and PL-2 show relatively stronger binding to PLK1-PBD than other peptides, likely resulting from their increased hydrogen bond formation and more effective engagement with the hydrophobic region. Meanwhile, the results of the structure-activity relationship analysis are consistent with the docking scores. Specifically, PL-1 exhibits strong binding interactions with PLK1-PBD and has the lowest docking score.

Figure 5.

The 2D and 3D interactions of PLs 1–5 within the PLK1-PBD binding site. (A) PL-1 is shown in yellow; (B) PL-2 is shown in cyan; (C) PL-3 is shown in pink; (D) PL-4 is shown in red; (E) PL-5 is shown in blue. Active site residues are shown as green sticks; hydrogen bonds are represented in black dashed lines.

Binding affinity assessment

The binding affinities of PLs 1–5 to PLK1-PBD were evaluated using the MST assay with the PLK1-PBD inhibitor peptide 5 and PLK1-KD inhibitor BI 2536 as controls. As indicated in Table 3, PLs 1–5 showed substantially strong binding affinities for PLK1-PBD with K_d values in the range of 3.11–40.42 nM. Notably, PL-1 showed the highest binding affinity for PLK1-PBD (K_d = 3.11 ± 0.05 nM) among the tested peptides. Meanwhile, the affinity of PL-1 for PLK1-PBD was approximately 21-fold higher than that of peptide 5 (K_d = 67.03 ± 2.82 nM), suggesting that the hydrogen bond formed by Glu in PL-1 enhances its binding affinity with PLK1-PBD compared to peptide 5. The fitting curves of MST are shown in Figure S1. Meanwhile, the binding affinity between BI 2536 and PLK1 was detected. As shown in Table S2, the K_d value of BI 2536 for the PLK1-KD was 8.26 ± 0.44 nM, suggesting its binding affinity was lower than that of PL-1 for the PLK1-PBD (K_d = 3.11 ± 0.05 nM). In addition, we compared the K_d values determined by MST assays of previously reported PLK1-PBD inhibitors (Table 4)²⁶ with PLs 1–5, demonstrating that PLs 1–5 exhibited significantly higher binding affinity to PLK1-PBD than these inhibitors. These results suggest that PL-1 is a promising candidate for targeting PLK1-PBD and that further biological experiments are warranted to explore its potential anticancer effects.

Table 4.

Binding affinities of reported PLK1-PBD inhibitors detected by MST assay.

PBD inhibitor	3v	ZINC20503376	MCC1019	AN-329/10,077,003
PLK1-PBD (aKd, μM)	2.3	0.28	13	26

^a K_d value is derived from the reference with DOI: 10.1080/17460441.2023.2159942.

Kinase selectivity assay

The kinase selectivity experiments were conducted on other PLKs members and a panel of kinases to assess the selective inhibitory ability of PL-1, with the PLK1-PBD inhibitor peptide 5 and PLK1-KD inhibitor BI 2536 used as controls. As shown in Table 5, PL-1 exhibited low inhibitory activities against PLKs 2–5. In contrast, peptide 5 showed low inhibitory activities against PLKs 2–4 but demonstrated significantly higher potency against PLK5, suggesting that it lacks strong selectivity for PLK1. Meanwhile, BI 2536 showed high inhibitory rates against all PLKs 2–5, indicating low selectivity for PLK1 (Table S3). In addition, PL-1 showed low inhibition rates against a panel of kinases at a concentration of 10 μM (Table 6). These results demonstrate that PL-1 is highly selective for PLK1-PBD with minimal inhibitory effect against other kinases.

Table 5.

Selectivity profile of PL-1 at a concentration of 10 µM against other PLK family members.

Target	% inhibition at 10 μM
	PLK2	PLK3	PLK4	PLK5
PL-1	4.09 ± 0.23	2.27 ± 0.16	3.18 ± 0.19	1.46 ± 0.07
peptide 5	8.21 ± 0.29	6.92 ± 0.14	5.04 ± 0.18	91.24 ± 3.37

The data are presented as the mean ± SD, n = 3.

Table 6.

Kinase selectivity profile of PL-1 at a concentration of 10 µM on a panel of kinases.

Target	% inhibition at 10 μM	Target	% inhibition at 10 μM	Target	% inhibition at 10 μM
ABL1	0.39 ± 0.02	DAPK1	0.23 ± 0.01	p38α	2.17 ± 0.11
AKT1	0.08 ± 0.01	DCAMKL2	0.48 ± 0.03	p70S6K	0.29 ± 0.02
ALK 4	5.63 ± 0.18	DYRK1B	6.83 ± 0.25	PAK2	0.15 ± 0.01
AMPKα1	0.16 ± 0.01	EGFR	0.14 ± 0.01	PDGFRα	4.04 ± 0.13
MAPKAPK2	0.34 ± 0.02	EPHB4	0.5 ± 0.16	PIM1	0.37 ± 0.01
MET	2.82 ± 0.14	Erk2	0.69 ± 0.03	PKACα	0.79 ± 0.04
ZAP70	0.71 ± 0.03	FGFR1	0.74 ± 0.02	PKCα	1.58 ± 0.08
TYRO3	0.25 ± 0.01	GSK3β	0.81 ± 0.04	PKD2	0.95 ± 0.05
AurB	1.6 ± 0.08	HGK	5.45 ± 0.19	CK2a1	0.31 ± 0.01
AXL	0.13 ± 0.01	HIPK1	0.99 ± 0.04	CK1γ3	0.42 ± 0.02
BRAF	0.16 ± 0.01	HPK1	0.82 ± 0.03	ROCK1	5.54 ± 0.24
BRSK2	0.42 ± 0.03	IGF1R	0.06 ± 0.01	RSK1	0.32 ± 0.02
CDK1/CycE1	0.77 ± 0.04	IKK-alpha	4.03 ± 0.11	SGK	0.19 ± 0.01
CDK18	5.59 ± 0.21	ITK	0.6 ± 0.14	SRC	1.67 ± 0.08
CDK2/CycE1	0.85 ± 0.09	JAK3	0.58 ± 0.03	SYK	0.03 ± 0.001
CDK5/p35NCK	0.49 ± 0.04	JNK1	0.32 ± 0.01	TIE2	0.35 ± 0.02
CHK1	4.73 ± 0.22	KDR	5.74 ± 0.18	TRKB	2.24 ± 0.07
CHK2	0.07 ± 0.01	LCK	0.64 ± 0.06	TSSK1	0.18 ± 0.01
CK1α	0.94 ± 0.05	MAP2K1	4.15 ± 0.13
CK1γ2	3.73 ± 0.16	NEK1	4.42 ± 0.17
CLK1	0.54 ± 0.04	NEK2	0.08 ± 0.01

The data are presented as the mean ± SD, n = 3.

MD Simulation

As PL-1 and PL-2 exhibited higher binding affinities with PLK1-PBD compared to other peptides, 100 ns MD simulations were conducted using GROMACS to evaluate their binding stability with PLK1-PBD. The smaller RMSD values indicate lower conformational fluctuations of the complexes, thus reflecting their higher binding stability. Figure 6(A,B) shows that the RMSD of both PLK1-PL-1 and PLK1-PL-2 complexes reached a stable value with slight fluctuations around 0.22 nm. This indicates that the binding of PL-1 and PL-2 to PLK1 is stable. Root mean square fluctuations (RMSF) values were used to analyse the conformational flexibility of amino acid residues involved in ligand binding. As shown in Figure 6(C,D), the RMSF values of key residues Trp414, Asp416, Tyr417, Arg516, His538, Lys540, Tyr421, Leu478, Tyr481, Phe482, Tyr485, and Leu490 in the PLK1 with PL-1 and PL-2 binding site were below 0.15 nm and 0.16 nm, respectively. These low RMSF values indicate the stable binding of the key amino acids in PLK1 to PL-1 and PL-2. In addition, Figure 6(E,F) shows the radius of gyration (Rg) values of the PLK1-PL-1 and PLK1-PL-2 complexes. The complexes were kept stable at approximately 1.84–1.87 nm during the simulation, indicating that the two complexes maintained good structural integrity. Meanwhile, Figure 6(G,H) indicates that the secondary structure of the complexes of PLK1 with PL-1 and PL-2 did not change significantly throughout the simulation, further confirming the stability of the complex structure. In conclusion, PL-1 and PL-2 showed stable binding to the PLK1-PBD active site and maintained the conformational integrity of the complex structure throughout the MD simulation process.

Figure 6.

MD simulations of PLK1 in complex with PL-1 and PL-2. (A, B) RMSD of PLK1-PL-1 and PLK1-PL-2 complexes, respectively. (C, D) RMSF of PLK1 in the PLK1-PL-1 and PLK1-PL-2 complexes, respectively. (E, F) Rg of PLK1 in the PLK1-PL-1 and PLK1-PL-2 complexes, respectively. (G, H) The secondary structures of PLK1 in the PLK1-PL-1 and PLK1-PL-2 complexes, respectively.

Antiproliferative activity

The cytotoxic potential of PLs 1–5 against multiple myeloma cells U266 was assessed through MTT assays. As shown in Table 3, PLs 1–5 exhibited varying degrees of antiproliferative effects on the U266 cell line, with their cytotoxicity being significantly stronger than that of the positive control peptide 5 (IC₅₀ = 2.25 ± 0.11 µM). Notably, PL-1 demonstrated the highest potency in suppressing U266 cell growth (IC₅₀ = 0.09 ± 0.01 µM). The cell inhibitory rates (%) are shown in Figure S2. In addition, the PLK1-KD inhibitor BI 2536 exhibited lower antiproliferative activity (IC₅₀ = 0.28 ± 0.03 µM) against U266 cells than PL-1 (Table S2). The experimental data suggest that PL-1 possesses significant therapeutic potential, highlighting its promise as a candidate for multiple myeloma therapy.

Intracellular uptake ability

To evaluate the intracellular uptake ability of PL-1, live-cell confocal microscopy was performed on U266 cells treated with FITC-PL-1. As shown in Figure 7, U266 cells treated with FITC-PL-1 exhibited strong green fluorescent signals in the cytoplasm and nucleus, indicating that it can efficiently escape from endosomes into the cytoplasm and nucleus of U266 cells. This suggests that FITC-PL-1 exhibits high intracellular uptake ability.

Figure 7.

Confocal laser scanning microscopy images of U266 cells incubated with FITC-labelled PL-1 for 2 h. Rhodamine-labelled dextran (dextran^Rho) was used as an endocytosis marker, and cell nuclei were stained with Hoechst 33342.

Biostability evaluation

Good biostability is an important factor in the anticancer effect of peptide drugs. Compared to peptide 5, PL-1 was modified with N-terminal acetylation and C-terminal amination to improve serum stability. To evaluate the serum stability of PL-1 and peptide 5, a human serum stability assay was conducted. As shown in Figure 8, peptide 5 was almost completely degraded in human serum after 240 min. In contrast, approximately 75% of PL-1 remained stable after 120 min. This result suggests that PL-1 is significantly more biologically stable than peptide 5. In addition, we further evaluated the metabolic stability of PL-1 and peptide 5 in the human liver microsomes. As shown in Table S4, compared with peptide 5, PL-1 exhibited a longer half-life (T_1/2 = 36 ± 0.5 min) and a lower intrinsic clearance (CL_int = 39 ± 0.4 μL/min/mg). These data indicate that PL-1 has superior metabolic stability, a property that is beneficial for prolonging its in vivo duration of action and maintaining effective plasma concentrations.

Figure 8.

The biostability of PL-1 in human serum. Results are expressed as mean ± SD, n = 3.

Conclusion

Multiple myeloma is a systemic hematological malignancy and presents significant therapeutic challenges due to its mechanistic complexity. PLK1, a key regulator of mitotic progression, is overexpressed in multiple myeloma cells and shows a strong association with poor clinical outcomes. This suggests that PLK1 may represent a potential therapeutic target for multiple myeloma. In our research, five novel peptides (PLs 1–5) targeting the PLK1-PBD were successfully identified using an integrated virtual screening strategy. First, the accurate pharmacophore features were constructed by systematically analysing the key interaction patterns between PLK1-PBD and its ligand (PDB ID: 3P37). Based on these pharmacophore features, a large-scale virtual screening of the peptide database was performed to identify potential peptides that matched these features. Subsequently, the peptides selected from the screening were subjected to molecular docking against the PLK1-PBD active site, and the five peptide candidates (PLs 1–5) with optimal affinity were selected on the basis of binding free energy scores. The binding affinities of PLs 1–5 to PLK1-PBD were confirmed by MST assay. The results showed that these peptides exhibited strong binding affinities to PLK1-PBD, with PL-1 having the highest affinity. Meanwhile, kinase selectivity assays showed that PL-1 had no significant inhibitory activity against non-target kinases, including other members of the PLKs. This finding suggests that PL-1 has a high degree of selectivity for PLK1, which is crucial to minimise potential off-target effects and enhance therapeutic specificity. In addition, MD simulations showed that the PLK1-PL-1 complex maintained conformational stability throughout the simulation period. This stability demonstrates the strong binding affinity and structural integrity of the complex, supporting the potential of PL-1 as a selective PLK1 inhibitor. Notably, PL-1 exhibited significant antiproliferative activity against U266 multiple myeloma cells and high intracellular uptake capacity. The high binding affinity and selectivity of PL-1 for PLK1-PBD suggests that PL-1 may interfere with the key regulatory functions of PLK1 in the cell cycle and inhibit the proliferation of multiple myeloma cells.

Compared with the previously reported PLK1-PBD inhibitor peptide 5, PL-1 exhibits enhanced binding affinity, high selectivity, strong biostability, and antiproliferative activity. Structural analysis revealed that the substitution of Leu in peptide 5 with Glu (as in PL-1) facilitated the formation of an additional hydrogen bond with PLK1, substantially enhancing binding affinity. MST assay showed that the affinity of PL-1 for PLK1 was approximately 21-fold higher than that of peptide 5. Meanwhile, PL-1 exhibited lower inhibitory activity against PLKs 2–5, but peptide 5 showed a stronger inhibitory activity against PLK5, which indicated that PL-1 was more selective for PLK1 than peptide 5. In addition, PL-1 underwent N-terminal acetylation and C-terminal amination modifications, resulting in enhanced serum stability and liver microsomal stability compared to peptide 5. Furthermore, PL-1 displayed stronger antiproliferative effects against U266 multiple myeloma cells than peptide 5, suggesting its greater therapeutic potential as a selective PLK1-PBD inhibitor for cancer treatment. In conclusion, PL-1 is a highly promising candidate that warrants further investigation to fully explore its therapeutic potential.

Funding Statement

This work was supported by the research project of Huai’an Science and Technology Bureau (HAB202323), Huai’an Health Commission Research General Project (HAWJ2024017), Jiangsu Commission of Health scientific research project (Geriatrics) (LKM2023045) and Xuzhou Medical University Affiliated Hospital Development Fund General Project (XYFM202246).

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author/s.