Abstract 5305: Detection and quantification of proteins using protein identification by short-epitope mapping (PrISM)

Abstract

Abstract Introduction: Here, we demonstrate Protein Identification by Short-epitope Mapping (PrISM), which aims to provide comprehensive proteome analysis with broad dynamic range at single-molecule resolution by interrogating immobilized, intact proteins in parallel using multi-affinity probes. Improving the dynamic range and scale for protein analyses will enable a deeper understanding of low abundant protein in all stages of cancer progression. In addition, enabling the interrogation of single-molecules will provide a deeper understanding of protein diversity, such as the proteoforms resulting from non-canonical post-translational modifications inherent to cancer1. The combination of single-molecule sensitivity with comprehensive proteome coverage could also open the door for highly sensitive and specific diagnostics. Methods: PrISM uses non-traditional affinity reagents with high affinity and low specificity that bind to short epitopes in multiple proteins. Sample proteins were conjugated to DNA nanoparticles and deposited on a high-density patterned flow cell at optically resolvable locations. Multi-affinity probes were applied to sample proteins over multiple cycles to generate binding patterns for each single-molecule protein, which are translated to protein identifications and quantities using a custom machine learning approach. We acquired PrISM data on native biological and control samples using dozens of multi-affinity probes targeting trimer or tetramer sequences. Results: We report single-molecule deposition of over 1 billion DNA nanoparticle complexes on a flow cell. We demonstrate how the PrISM methodology identifies individual protein molecules through iterative probing with our multi-affinity probes. Further, we provide an analytical assessment of the sensitivity and specificity of PrISM and demonstrate the ability to accurately estimate the false identification rate of these proteins using a target-decoy based statistical approach. Conclusions: Combining single-molecule analysis, intact (non-digested) proteins, and iterative probing, PrISM provides a new tool for quantitative proteomics. We demonstrate linear and reproducible quantification of proteins using PrISM, potentially enabling detection of low abundant proteins and proteoforms associated with cancer. The ability to make comprehensive measurements of intact proteins at single-molecule resolution could accelerate basic cancer research through to the clinic.