Quantitative Approaches in Studying Cancer Metabolism With Therapeutic Applications

Abstract

Modern drug discovery often involves screening libraries of structurally diverse small molecules (“chemical space”) and evaluating their ability to interact with defined protein targets or to perturb biological pathways of interest in a model system. However, the therapeutic targets and the model systems themselves can have stunning diversity at various levels including single cell heterogeneity, cell line disparity, and patient-to-patient variation. This dissertation explores on how the intrinsic diversity in the “cell space” influences chemical sensitivity and genetic dependency, and therefore provides clues for future therapeutic development. In particular, an emphasis was given on the “cell metabolism space” given its variation across different biological samples has not been thoroughly explored especially in the context of cancer. To study the observation of such diversity in a unified framework, I adopted statistical approaches that account for the “diversity” between different subjects (e.g., cell lines, patient samples) and quantitatively explored potential cancer dependencies that might serve as a first step towards therapeutic design and optimization. In Chapter 1, I examined the diversity of metabolism in model cancer cell lines with various examples highlighting the interactions with genetic and transcriptional programs. To characterize metabolic diversity, two complementary approaches are introduced in this chapter. First, baseline metabolic phenotypes of cancer cell lines can be characterized by comprehensive profiling of metabolite abundance. Second, metabolic dependencies of cancer cell lines can be screened in a pooled form under different nutrient conditions. As a proof-of-concept, I studied the intrinsic hypermethylation of asparagine synthetase observed in subsets of hepatic and gastric cancers. This epigenetic alteration created a selective metabolic dependency that can be exploited as a therapeutic target. In Chapter 2, I further investigated the metabolic dysregulation of pentose phosphate pathway revealed in quantitative profiling from Chapter 1. My results demonstrate that 6-phosphogluconate dehydrogenase links cytosolic carbohydrate metabolism to protein secretion. I also show an unbiased computational approach to unveil distinct genes involved in inherently related biological processes. In Chapter 3, I expanded the kynurenine study reported in Chapter 1 to a clinical setting and explored how metabolic diversity can inform therapeutic design. In particular, I focused on the metabolic alterations in response to immune checkpoint blockade. The results convincingly advocate for patient stratification and metabolic monitoring in designing the next generation of immunotherapy clinical trials. To summarize, these results prove the value of comprehensive, quantitative approaches in understanding cancer metabolism and open up therapeutic opportunities tailored to the metabolic diversity.