Abstract
The breakthrough in the discovery of immune checkpoint PD-1/PD-L1 inhibitors, such as the series of Bristol Myers Squibb synthetic compounds, boosted the research of small molecules with blockade effects on the interaction of PD-1/PD-L1. However, the search for natural products derived PD-1/PD-L1 inhibitors can be impeded by the false positive and/or negative results from the screening assays. Herein, we combined a PD-1/PD-L1 blockade assay (pair ELISA) and a PD-L1/PD-L1 binding assay (surface plasmon resonance; SPR) to evaluate a panel of natural compounds previously reported to show anti-PD-1/PD-L1 activity. The test compounds included kaempferol, cosmosiin, tannic acid, pentagalloyl glucose, ellagic acid, resveratrol, urolithin A, and rifubutin. Based on the analyses of their responses to the combined screening assays, these compounds were categorized into four groups: I) PD-1/PD-L1 inhibitors that can bind to PD-1 and PD-L1; II) PD-1/PD-L1 inhibitors selectively bind to PD-L1 protein; III) PD-1/PD-L1 inhibitors without binding capacity, and IV) PD-1/PD-L1 binders without blockade effect. Discrimination of positive responders in the PD-1/PD-L1 blockade and binding assays can provide useful insights to avoid false outcomes. Examples demonstrated in this study suggest that it is crucial to adopt proper evaluation methods (including using multiple-facet functional assays and target binding techniques) for the search for natural products derived PD-1/PD-L1 inhibitors.