In cancer immunotherapy, an emerging approach is to block the interactions of programmed cell death-1 (PD-1) and programmed cell death-ligand 1 (PD-L1) using small-molecule inhibitors. The food-derived polyphenols curcumin (CC), res (RSV) and epigallocatechin gallate (EGCG) have anticancer immunologic functions, which, recently, have been proposed to act via the downregulation of PD-L1 expression. However, it remains unclear whether they can directly target PD-L1 dimerization and, thus, interrupt the PD-1/PD-L1 pathway. To elucidate the molecular mechanism of such compounds on PD-L1 dimerization, molecular docking and nanosecond molecular dynamics simulations were performed. Binding free energy calculations show that the affinities of CC, RSV and EGCG to the PD-L1 dimer follow a trend of CC > RSV > EGCG. Hence, CC is the most effective inhibitor of the PD-1/PD-L1 pathway. Analysis on contact numbers, nonbonded interactions and residue energy decomposition indicate that such compounds mainly interact with the C-, F- and G-sheet fragments of the PD-L1 dimer, which are involved in interactions with PD-1. More importantly, nonpolar interactions between these compounds and the key residues Ile54, Tyr56, Met115, Ala121 and Tyr123 play a dominant role in binding. Free energy landscape and secondary structure analyses further demonstrate that such compounds can stably interact with the binding domain of the PD-L1 dimer. The results provide evidence that CC, RSV and EGCG can inhibit PD-1/PD-L1 interactions by directly targeting PD-L1 dimerization. This provides a novel approach to discovering food-derived small-molecule inhibitors of the PD-1/PD-L1 pathway with potential applications in cancer immunotherapy.

Inhibitors blocking the PD-1/PD-L1 immune checkpoint demonstrate impressive anti-tumor immunity, and small molecule inhibitors disclosed by the Bristol-Myers Squibb (BMS) company have become a hot topic. In this work, by modifying the carbonyl group of BMS-202 into a hydroxyl group to achieve two enantiomers (MS and MR) with a chiral center, we found that this is an effective way to regulate its hydrophobicity and thus to reduce the negative effect of polar solvation free energy, which enhances the stability of PD-L1 dimer/inhibitor complexes. Moreover, we studied the binding modes of BMS-200 and BMS-202-related small molecule inhibitors by molecular dynamics simulation to explore their inhibitory mechanism targeting PD-L1 dimerization. The results showed that the size exclusion effect of the inhibitors triggered the rearrangement of the residue ATyr56, leading to the formation of an axisymmetric tunnel-shaped pocket, which is an important structural basis for improving the binding affinity of symmetric inhibitors with PD-L1. Furthermore, after inhibitor dissociation, the conformation of ATyr123 and BMet115 rearranged, which blocked the entrance of the binding pocket, while the reverse rearrangements of the same residues occurred when the PD-L1 monomer was complexed with the inhibitors, preparing PD-L1 for dimerization. Overall, this study casts a new light on the inhibitory mechanism of BMS inhibitors targeting PD-L1 dimerization and provides an idea for designing novel small molecule inhibitors for future cancer immunotherapy.

Using small molecules to inhibit the PD-1/PD-L1 pathway is an important approach in cancer immunotherapy. Natural compounds such as capsaicin, zucapsaicin, 6-gingerol and curcumin have been proposed to have anticancer immunologic functions by downregulating the PD-L1 expression. PD-L1 dimerization promoted by small molecules was recently reported to be a potential mechanism to inhibit the PD-1/PD-L1 pathway. To clarify the molecular mechanism of such compounds on PD-L1 dimerization, molecular docking and molecular dynamics simulations were performed. The results evidenced that these compounds could inhibit PD-1/PD-L1 interactions by directly targeting PD-L1 dimerization. Binding free energy calculations showed that capsaicin, zucapsaicin, 6-gingerol and curcumin have strong binding ability with the PD-L1 dimer, where the affinities of them follow the trend of zucapsaicin > capsaicin > 6-gingerol ≈ curcumin. Analysis by residue energy decomposition, contact numbers and nonbonded interactions revealed that these compounds have a tight interaction with the C-sheet, F-sheet and G-sheet fragments of the PD-L1 dimer, which were also involved in the interactions with PD-1. Moreover, non-polar interactions between these compounds and the key residues Ile54, Tyr56, Met115 and Ala121 play a key role in stabilizing the protein–ligand complexes in solution, in which the 4′-hydroxy-3′-methoxyphenyl group and the carbonyl group of zucapsaicin, capsaicin, 6-ginger and curcumin were significant for the complexation of small molecules with the PD-L1 dimer. The conformational variations of these complexes were further analyzed by free energy landscape (FEL) and principal component analysis (PCA) and showed that these small molecules could make the structure of dimers more stable. This work provides a mechanism insight for food-derived small molecules blocking the PD-1/PD-L1 pathway via directly targeting the PD-L1 dimerization and offers theoretical guidance to discover more effective small molecular drugs in cancer immunotherapy.