3-D High-Resolution Mapping of the Heterogeneity in Mitochondrial Redox State of Human Breast Tumor Xenografts

Abstract

Intratumor heterogeneity is of great research interest for understanding cancer pathology and developing diagnostic and therapeutic methods. The low-temperature 3D NADH/Fp (oxidized flavoproteins including FAD) fluorescence imaging or the redox scanning provides an effective tool for imaging intratumor heterogeneity in mitochondrial redox state at a high spatial resolution (down to 50 × 50 × 20 μm3). Previously, we have shown mitochondrial redox state and its heterogeneity in tumor tissue provide sensitive and potentially diagnosis-useful characteristics for differentiating among five human melanoma and two breast cancer mouse xenografts of different metastatic potential. Here, we report the preliminary results of imaging the in vivo mitochondrial redox state of the entire tumor for three human breast cancer lines having ascending order of aggressiveness, i.e., MCF-7 < MDA-MB-468 < MDA-MB-231 xenografted in athymic nude mice. The tumor-bearing mice were anesthetized and snap-frozen in liquid N2 so that the in vivo mitochondrial redox state was maintained for ex vivo redox scanning. The entire excised tumors were scanned section by section at different depths with 400 μm spacing, total of 11–14 sections per tumor. We obtained the 3D distribution of nominal concentrations of oxidized flavoproteins (Fp) and NADH in tissue and calculated the Fp redox ratio Fp/(Fp + NADH) representing the mitochondrial redox state. The results revealed that both aggressive tumors (MDA-MB-231 and MDA-MB-468) displayed heterogeneity in the distributions of NADH, Fp and Fp redox ratio, with a localized area exhibiting significantly higher Fp redox ratio than other regions; the indolent MCF-7 tumor displayed a relatively uniform distribution in both Fp and NADH, thus Fp redox ratio. The results suggest possible novel imaging biomarkers on the basis of NADH, Fp and Fp redox ratio images to differentiate among these tumors. Potentially, these biomarkers may be useful for cancer diagnosis and therapy.