PROJECT SUMMARY Cancer tissues exhibit a high degree of phenotypic heterogeneity and plasticity, with cancerous tissues comprising many different subpopulations of cells in various states. Quantifying this heterogeneity at the single- cell level and with molecular depth across large numbers of cells provides information that cannot be obtained from bulk studies and that will ultimately lead to improved diagnostics and more effective treatments. While single-cell sequencing approaches are having a significant impact on cancer research, proteins mediate the bulk of cellular function and are the targets of most therapeutics. Given that a compelling body of literature has shown that the correlation between RNA and protein abundance is at best poor to moderate, there is an urgent need to develop new technologies for large-scale unbiased direct proteome profiling at the single-cell level. To fill this gap, mass spectrometry (MS)-based profiling of protein expression in single cells has very recently become a reality due to more efficient sample processing workflows, novel experimental designs and improved instrument sensitivity. Label-free MS-based proteomics can currently quantify up to 1500 protein groups per cell across >4 orders of magnitude of dynamic range, but throughput has been limited to ~24 samples per day. This low throughput is inadequate to perform the large-scale statistically powered studies required to characterize heterogeneity in cancer cell populations. To increase measurement throughput, multiplexed workflows have been developed based on isobaric tandem mass tags (TMTs) that enable >10 single cells to be measured in an LC-MS analysis, but these suffer from a number of significant drawbacks including isotopic contamination, degraded quantitative accuracy when employing a carrier channel, precursor coisolation with concomitant ratio compression, chemical noise resulting from cross-reactivities of TMT reagents with contaminants, etc. The overall objective is to develop a platform that exceeds the throughput of current TMT-based workflows while preserving the depth of coverage and dynamic range of label-free workflows. We hypothesize that a robust multicolumn ultra-high-performance nanoLC system with a 5-minute peptide elution window and a 100% duty cycle, combined with novel MS1-level protein identification and quantification, will enable label-free profiling of >2000 protein groups per cell at a throughput of up to 288 samples per day, thus providing a providing a capability for direct, in-depth and large-scale protein quantification that is analogous to single-cell RNA-seq. Studies in Aim 1 will focus on developing high-peak-capacity fast nanoLC separations, as well as a novel sorbent-coated sample-loop providing desalting and debris removal for robust long-term operation. In Aim 2 we will develop a 4-column LC platform based on these rapid separations and a primarily MS1-based acquisition workflow to increate duty cycle to 100% and...