A Bionic-Homodimerization Strategy for Optimizing Modulators of Protein-Protein Interactions: From Statistical Mechanics Theory to Potential Clinical Translation

Abstract

Emerging protein-protein interaction (PPI) modulators have brought out exciting ability as therapeutics in human diseases, but its clinical translation has been greatly hampered by the limited affinity. Inspired by the homodimerize structure of antibody, the homodimerization contributes hugely to generating the optimized affinity is conjectured. Herein, a statistical-mechanics-theory-guided method is established to quantize the affinity of ligands with different topologies through analyzing the change of enthalpy and the loss of translational and rotational entropies. A peptide modulator for p53-MDM2 termed CPAP is used to homodimerize connecting, and this simple homodimerization can significantly increase the affinity. To realize the cellular internalization and tumor accumulation, ^Dimer CPAP and ^Mono CPAP are nanoengineered into gold(I)-CPAP supermolecule by the aurophilic interaction-driven self-assembly. Nano-^Dimer CPAP potently suppressed tumor growth in lung cancer allograft model and a patient-derived xenograft model in more action than Nano-^Mono CPAP, while keeping a favorable drug safety profile. This work not only presents a physico-mechanical method for calculating the affinity of PPI modulators, but also provides a simple yet robust homodimerization strategy to optimize the affinity of PPI modulators.

Keywords: bionic-dimerization; nanomedicine; peptide; protein-protein interactions; statistical mechanics theory.

Figure 1

Statistical mechanics theory for binding affinity of dimeric strategy in the intervention of disease PPIs. A) Illustration of the statistical mechanics theory. B) The schematic diagram of function of ^DimerCPAP and ^MonoCPAP. C,D) ^DimerCPAP and ^MonoCPAP analyzed by HPLC and electrospray ionization mass spectrometry (ESI‐MS), which was performed on a reversed‐phase C18 column (Waters XBridge 3.5 µm, 4.6 × 150 mm) at 40 °C. E) Quantification of the interactions of ^synMDM2 with varying concentrations of ^DimerCPAP and ^MonoCPAP by isothermal titration calorimetry (ITC). Each curve is the mean of 3 independent measurements at 25 °C in 10 × 10⁻³ m HEPES, 150 × 10⁻³ m NaCl, pH 7.4. F) Fluorescence polarization (Fp) binding assay of ^DimerCPAP or ^MonoCPAP to MDM2 protein. G) Competitive FP‐based binding assay of ^DimerCPAP or ^MonoCPAP to MDM2/p53 complex. H) Schematic depiction for nano‐engineering modification of ^DimerCPAP and ^MonoCPAP. I) TEM images of Nano‐^DimerCPAP and Nano‐^MonoCPAP. J) Hydrodynamic diameter and K) ZETA potential of Nano‐^DimerCPAP and Nano‐^MonoCPAP measured in PBS buffer at pH 7.4.

Figure 2

Nano‐^DimerCPAP potently activated p53 signaling cascades beyond Nano‐^MonoCPAP in vitro. A) Apoptosis and necrosis analysis of NCI‐1650 cells incubated with PBS, Nano‐^DimerCPAP (10 µg mL⁻¹), Nano‐^MonoCPAP (10 µg mL⁻¹) or ^CtrlNano via flow cytometry for 48 h (n = 3, mean ± sd). B) Cell cycle analysis of NCI‐1650 cells treated with Nano‐^DimerCPAP, Nano‐^MonoCPAP, ^CtrlNano or PBS control for 48 h by FACS (n = 3, mean ± sd). C) Heat map of RNA‐Seq analysis of NCI‐H1650 cells’ mRNAs which were differentially expressed between Nano‐^DimerCPAP and ^CtrlNano (n = 3). D) GSEA results for the p53 signaling pathway and the p53 downstream pathway. GSEA results for the REACTOME cell cycle checkpoints, E) the REACTOME cell cycle mitotic and F) the KEGG apoptosis. G) Hierarchical clustering of genes differentially expressed in NCI‐H1650 cells after exposure to Nano‐^DimerCPAP for 24 h compared with Nano‐^MonoCPAP (n = 3). H) GSEA analysis of Nano‐^DimerCPAP versus Nano‐^MonoCPAP showing the increased p53 signaling and downstream pathway. I) GSEA showing that the apoptosis of Nano‐^DimerCPAP is superior to Nano‐^MonoCPAP.

Figure 3

Nano‐^DimerCPAP potently suppressed tumor growth superior to Nano‐^MonoCPAP in C57/B6 mice LUAD allograft model. A) Growth curves of LLC homograft model in C57/B6 mice with treatments, following the administration of control (PBS), Nano‐^DimerCPAP (2.5 mg kg⁻¹) and Nano‐^MonoCPAP (2.5 mg kg⁻¹) (n = 5). B) Body weight of mice with the indicated treatments. C) Analysis of red blood cell (RBC), white blood cell (WBC), platelets (PLT), neutrophil, hemoglobin (HGB) in mice whole blood after treatments. D) Representative photographs and E) weight of tumor tissue isolated at the end of experiment. F,G) Representative images of H&E and TUNEL staining in tumor section from mice (scale bar: 50 µm). The immunohistochemical (IHC) staining for H) MDM2, I) MDMX, J) p53, and K) p73 in tumor sections from mice (scale bar: 50 µm).

Figure 4

Nano‐^DimerCPAP enhanced more anticancer activity in LUAD patient‐derived tumor xenograft in NOD/SCID mice. A) Diagrammatic sketch of LUAD‐PDX mice model with indicated treatments. B) Growth curves of LUAD‐PDX mice model after administration of control (PBS), Nano‐^DimerCPAP (2.5 mg kg⁻¹) and Nano‐^MonoCPAP (2.5 mg kg⁻¹) (mean ± sd, n = 5 per group). C) Images and D) weights of tumors excised at the end of treatment. p‐values were calculated by t test (*p < 0.05; **p < 0.01; ***p < 0.001). E) The H&E and F) TUNEL staining in tumor from mice after indicated treatments (scale bar: 200 µm). Representative images of IHC staining of G) Ki67, H) MDM2, I) MDMX, J) p53, and K) p73 in tumor section from mice with indicated treatments (scale bar: 200 µm).