Background
Tumor cells employ multiple tactics to evade immune cell activities. Programmed cell death ligand 1 (PD-L1) is one of the key antigens presented on tumor cells that bind to programmed cell death protein 1 (PD-1) and block immune response. Multiple immunotherapies targeting PD-1/PD-L1 blockade have been approved by FDA. Currently, patients are tested for PD-L1 protein expression to be considered for PD-1/PD-L1 checkpoint inhibitor immunotherapies. Clinical outcomes of these immunotherapies, however, do not always correlate to the expression level of PD-L1. There is a need for better biomarkers that are more predictive of clinical outcomes. Direct detection of PD-1/PD-L1 interactions in patient tissues is likely to have better correlation to the therapeutic effect of checkpoint inhibitors than PD-L1 test alone.
We have developed an assay that enables visualization of protein-protein interaction with multiomic context of tumor immune microenvironment (TIME). We observed PD-1/PD-L1 interaction, individual proteins, and mRNA at high spatial resolution in various tumor tissues using new workflow enabled by high sensitivity and specificity of RNAscopeTM technology.
Methods
To visualize protein-protein interactions in situ, we developed a novel technique enabling multiplexed detection of protein and mRNA targets on a single slide. Oligonucleotide-conjugated antibodies were prepared to integrate the detection of protein-protein interaction and proteins into semi-automated co-detection workflow based on RNAscopeTM technology, performed on Leica BOND instrument. The protocol allows imaging of up to 4 targets per cycle, and three cycles of detection. Protein and highly expressed mRNA targets were imaged first, followed by the detection of protein-protein interaction and mRNAs with lower expression level, with tyramide signal amplification.
Results
New assay for co-detection of protein-protein interaction, protein, and mRNA (up to 12-plex detection) was developed. First, we verified the PD-1/PD-L1 interaction signals generated with this assay, are detected where PD-1 and PD-L1 signals overlap using sequential immunofluorescence. Next, to characterize TIME surrounding the interaction-positive region, the panel was expanded to include cell phenotyping protein markers for both immune and tumor cells such as CD3, CD4, CD8 and PanCK, as well as mRNA markers for chemokines and cytokines such as CXCL10, IFNG and TNFA. The interaction signals appear as a group of punctate dots implying possibility for semi-quantitative analysis.
Conclusions
New RNAscopeTM multiomics workflow is a powerful technique to resolve PD-1/PD-L1 interaction in the context of TIME. Spatial multiomic analyses of ligand-receptor interaction will expand our knowledge of tumor immune evasion strategies and potentially offer new patient stratification strategy for checkpoint inhibitor immunotherapies.