Quantitative profiling of the expressed proteome of cancer/cancer-associated cells has potentially transforma- tive applications in cancer research, treatment development, and individualization of therapy, potentially provid- ing (i) a molecular-level understanding of cancer subtypes and how they respond to standard-of-care (SoC) therapy or new agents, (ii) insights into tumor cell/stromal interactions that govern responses to therapy, and (iii) how signaling and drug response networks are `wired' in both genetically-mutated cancer cells and in non- cancer stromal cells. Given the prevalent discordance between genetic/transcriptomic analysis and expressed tissue proteomes, we hypothesize that large scale, quantitative proteomic analysis that interrogates the expressed proteome can provide an in-depth, high-resolution understanding of tumor and associated cells respond to chemotherapy drugs. We hypothesize that detailed understanding of SoC chemotherapy responses will also assist in selection or repurposing of new agents into treatment regimens to enhance response by influencing specific cellular response networks. We will employ a novel workflow, IonStar, that incorporates analytical, technological and informatics advances to enable robust simultaneous quantification of a majority of the expressed tumor proteome at high sensitivity, accuracy, and reproducibility, in large sample cohorts, in order to capture concentration- and time-dependent responses of patient-derived xenograft (PDX) models of pancreatic adenocarcinoma (PDAC) to SoC regimens. IonStar is a robust, well-validated sample processing, analysis, and informatics workflow that currently quantifies 5-7,000 proteins in 80-100 biological sample batches, with >95% of proteins free of missing data, at high accuracy and reproducibility, and a <5% false- positive biomarker discovery rate, without prior sample fractionation. IonStar permits discrimination of tumor (human cell) and stromal (mouse) responses. We will obtain rich data sets that capture responses of multiple PDAC PDX isolates to SoC drugs gemcitabine and paclitaxel (Abraxane) at high temporal resolution, providing a foundation for understanding tumor and stromal responses to these agents, and how PDAC PDX isolates differ in their responses. The phosphoproteome and transcriptome will also be interrogated at specific times to provide overlapping data. Novel mathematical, systems-level pharmacodynamic models integrated with traditional network informatics approaches will accelerate knowledge extraction by imposing quantitative pharmacodynamic response mechanisms upon individual and clustered protein responses. We hypothesize this approach will address the challenge of proteomic analysis that key signaling and response network nodes may not always be represented within datasets, confounding conventional network analysis. With well- validated, sensitive, accurate, and reproducible high-throughput workflows coupled with novel systems pharmacological analysis, we will use the insights garnered to create testable predictions of chemotherapy combinations that are more efficacious in PDAC, or in particular PDAC patient isolates.
`Quantitative global proteomics' provides the capability for extensive quantification of nearly all of the proteins expressed by cells or tissues. Because proteins, in most cases, constitute the `machinery' and structural components of cells and tissues, investigation of the proteome thus reveals a cell's characteristics and behavior, and in the research here, reveals the mechanisms by which tumor cells respond to chemotherapy drugs, and either succumb to treatment or survive and develop cancer drug resistance. We will monitor, in high detail, the proteome-level responses of patient-derived pancreatic cancers to standard-of-care chemotherapy drugs, and employ novel, state-of-the-art mathematical analysis and modeling to accelerate the extraction of knowledge and insight into the mechanisms of tumor response that lead to tumor death or chemotherapy resistance, in order to select, on a mechanistic basis, the best chemotherapy combinations for a particular tumor isolate.
