Application of Ultra-low Circulating Biomarkers and Single Cell Analysis from Solid Tumors with Single Molecule Resolution

Circulating tumor cells (CTCs) and extracellular vesicles (EVs) carry tumor-derived cargo in the circulation and provide the potential for minimally invasive patient monitoring. These circulating biomarkers have both been shown to be critical for cancer progression and metastasis. However, their clinical utility has been limited due to the rarity of CTCs in the blood, whereas EVs, while abundant, only carry miniscule amounts of cancer-derived cargo. This thesis shows the development and application of novel characterization methods for profiling these two circulating biomarkers, which both require ultra-sensitive approaches for characterization. First, a single cell analysis workflow was developed for mutation and gene expression profiling from the same cell. This multistep process was subsequently used for mutation detection using droplet digital PCR (ddPCR) and gene expression profiling using the Fluidigm Biomark HD. Epidermal growth factor receptor (EGFR) mutations were used as a model to optimize and validate this workflow. EGFR mutations confer either sensitivity or resistance to tyrosine kinase inhibitors (TKIs). Significant cell-to-cell variability of total EGFR expression and relative mutant-to-wildtype allele expression were observed, even within a clonal cell line. This workflow was applied to CTCs from non-small cell lung cancer (NSCLC) patients. CTCs were isolated using the Labyrinth, a high-throughput microfluidic CTC isolation technology. CTCs showed variable mutation status and revealed intra-patient heterogeneity. All CTCs carrying point mutations showed higher wildtype than mutant allele expression, which may be a mechanism for avoiding cell death from TKIs. Next, I present the first longitudinal patient monitoring for mutations carried in EVs. Mutation profiling in EVs is a new field, with most groups reporting combining EV-derived RNA and DNA with circulating tumor DNA. Here I screened EV-derived RNA (EV-RNA) using ddPCR and protein (EV-protein) using western blot for the presence of EGFR mutations from NSCLC patients. The change in EGFR mutation carried in EV-RNA compared to the previous visit mirrored disease trajectory. Neither EV-protein presence nor change showed a correlation with disease trajectory or detection with EV-RNA. Then, I applied a multi-pronged CTC profiling method to compare patient survival in metastatic pancreatic adenocarcinoma. CTCs were isolated from treatment naïve patients as well as after beginning therapy. CTCs were screened for their abundance and phenotype, KRAS mutation status and expression, and transcriptome profiles. Patients with poor survival showed higher levels of total CTCs and epithelial-to-mesenchymal transition (EMT)-CTCs at pre-treatment samples. Mutant KRAS was higher at both time points in deceased patients. Significant deregulation of genes and related pathways was observed including those known to be involved in RAS signaling. Lastly, I evaluated the expression of natural killer (NK) cell activating and inhibiting ligands on CTC-derived cell lines and freshly isolated CTCs. NK cells have been implicated in controlling and halting metastasis, however, little is known about the regulation balancing activating and inhibiting signals. I showed the NK-sensitivity varied based on the relative expression of these ligands. In freshly isolated CTCs, I showed that increased expression of human leukocyte antigen (HLA)-A/B/C and decreased expression of CADM1 were correlated with disease progression. Overall, in this thesis I developed novel circulating biomarker analysis methods. Using these methods I showed that patients can be longitudinally monitored for signatures of treatment resistance and patient survival, which I posit can lead to improved patient care.PHDChemical EngineeringUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/168139/1/snowen_1.pd