Author ORCID Identifier:

https://orcid.org/0009-0005-9907-0053

Date of Graduation

8-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Engineering (PhD)

Degree Level

Graduate

Department

Biomedical Engineering

Advisor/Mentor

Rajaram, Narasimhan

Committee Member

Quinn, Kyle

Second Committee Member

Griffin, Robert

Keywords

metabolism; multivariate; optical imaging; oxygenation; tumor; vascular

Abstract

Metabolic reprogramming is a characteristic hallmark of multiple pathologies and promotes disease progression, survival, and severity. Current methods to investigate metabolic reprogramming, however, require reductions of dimensionality both because of instrumentation limitations and analysis tools. This work aims to define a blueprint for high dimensional study of the oxygen metabolism axis to enable more sensitive study metabolic reprogramming. We hypothesized (1) multivariate imaging of metabolic and oxygenation metrics in ex vivo tissue sections would elucidate differences in resistant and sensitive head and neck squamous cell carcinoma (HNSCC) xenografts that traditional uni-variate methods miss and (2) an engineered hyperspectral darkfield microscopy (HSDFM)–autofluorescence intensity (AFI)–fluorescence lifetime imaging microscopy (FLIM) microscopy platform would enable longitudinal in vivo study of oxygenation and metabolism directly, without the use of exogenous labelling. In Chapter 3, we performed metabolic autofluorescence intensity (AFI) imaging and FLIM on unlabeled tissue sections and followed on with multiplexed immunohistochemistry (IHC) for hypoxia and hypoxia inducible factor 1-alpha (HIF-1α) to investigate multivariate relationships on both sides of the oxygen–metabolism axis. We detected significant and non-canonical changes in the optical redox ratio (ORR) and HIF-1α of radiation resistant tumor sections. Notably, we observed increased ORR associated with sections of low hypoxia and elevated HIF-1α following therapy, supporting HIF-1α-mediated treatment resistance, possibly through reactive oxygen species (ROS)-mediated stabilization of HIF-1α (p ≪ 0.0001 at 24 hours post-treatment (hr) and p = 0.03 at 48 hr). We also detected a strong trend at the sub-population level—notably, decreased ORR associated with increased HIF-1α at baseline in the resistant tumors (p = 0.0583)—supporting this histogram-based analysis and motivating future work. Toward the goal of engineering a label-free capable system, in Chapter 4, we engineered the first prototype of a combined HSDFM–MPM. Despite meeting many translation challenges from theory, the system showed promise to be sensitive to multivariate changes along the oxygen—metabolism axis. We provide a loose framework for simulation of and in silico validation of custom Monte Carlo lookup table (MCLUT), for robust FLIM phasor analysis with Monte Carlo-based testing of parameter extraction, and for multivariate analysis of label-free oxygenation and metabolism end points provided from this system. The new instrument achieved < 2 µm resolution, while resolving absorption differences in oxygenated and deoxygenated hemoglobin, all with total image set acquisition times between of around 3 minutes (including HSDFM, AFI, and FLIM). Together, these findings make significant headway in both hardware and software tools to investigate metabolic programming with enhanced sensitivity in vivo, minimizing reductive assumptions, instrumentation, or analysis.

Download

Share

COinS