Dual imaging of pHi and pHe to dissociate glycolytic/oxidative phenotypes for novel GBM therapeutics

About This Grant

Project Summary/Abstract The acidity of the tumor microenvironment (TME), driven by hyperglycolytic tumor metabolism, represents a key mechanism by which cancers develop resistance to therapies. Metabolic reprogramming in glycolytic phenotypes overproduce H+ and lactate, but the function of intracellular machinery essential for glycolysis require an alkaline pHi. To support tumor survival and progression, H+/lactate are exported from glycolytic phenotypes into the TME, but paradoxically H+/lactate may also be imported into oxidative phenotypes for their metabolic needs, a mechanism termed “metabolic symbiosis” between glycolytic and oxidative phenotypes. In addition, CO2 produced during substrate oxidation also leads to extracellular acidification using enzymes like carbonic anhydrases (e.g., CAIX). One mechanism by which H+ and lactate is imported/exported is through proton- coupled transmembrane monocarboxylate transporters (MCTs). Thus, the difference between intracellular pH (pHi) and extracellular pH (pHe), called the transmembrane pH gradient (∆pH=pHi–pHe), is much larger in tumors than in normal tissue. Due to differences in metabolic reprogramming between glycolytic and oxidative phenotypes, we posit that ∆pH will be larger for glycolytic vs. oxidative phenotype. Targeting MCTs (such as MCT1 or MCT4) or CAIX has shown promising results as novel anti-cancer therapeutic strategies. We propose to develop an MR-based platform for high-resolution ∆pH imaging in glioblastoma multiform (GBM), an incurable and aggressive cancer with high resistance to chemotherapy, with increased expression levels of MCT1 and MCT4. Currently, there is a paucity of techniques that can provide simultaneous pHi and pHe imaging. Measuring both pHi and pHe is critical to differentiate between glycolytic and oxidative phenotypes. pHi can be imaged by MRI with Amine and Amide Concentration-Independent Detection (AACID), whereas pHe is imaged with an MRSI method called Biosensor Imaging of Redundant Deviation in Shifts (BIRDS). Our goals are to use unsupervised deep learning to improve the pHe resolution to match that of pHi resolution so that high- resolution ∆pH imaging enables differentiation of glycolytic vs. oxidative phenotypes, and then determine how their metabolic state is affected by treatment with MCT and/or CAIX inhibitors. The primary metabolic measures will be enhanced by clinical measures of tumor perfusion and cellularity, as well as tissue staining. First we will develop high resolution ∆pH imaging in rat brains with patient-derived xenografts, where pHe resolution will be improved using deep learning. Then we will apply high resolution ∆pH imaging to examine the effects of MCTs inhibition. Because Temozolomide (which inhibits DNA synthesis) and MCTs inhibition target different GBM progression mechanisms, we will compare the effect of MCTs inhibition with the standard GBM treatment with Temozolomide. Finally, we will determine if combined MCT/CAIX inhibition is more effective than separate MCT and CAIX treatments. This powerful MR imaging tool has high potential for translation to monitor tumor phenotypes and therapy response to inform treatment decisions for improved outcomes.