Evaluation of hyperpolarized [1-¹³C]-pyruvate by magnetic resonance to detect ionizing radiation effects in real time

Abstract

Ionizing radiation (IR) cytotoxicity is primarily mediated through reactive oxygen species (ROS). Since tumor cells neutralize ROS by utilizing reducing equivalents, we hypothesized that measurements of reducing potential using real-time hyperpolarized (HP) magnetic resonance spectroscopy (MRS) and spectroscopic imaging (MRSI) can serve as a surrogate marker of IR induced ROS. This hypothesis was tested in a pre-clinical model of anaplastic thyroid carcinoma (ATC), an aggressive head and neck malignancy. Human ATC cell lines were utilized to test IR effects on ROS and reducing potential in vitro and [1-¹³C] pyruvate HP-MRS/MRSI imaging of ATC orthotopic xenografts was used to study in vivo effects of IR. IR increased ATC intra-cellular ROS levels resulting in a corresponding decrease in reducing equivalent levels. Exogenous manipulation of cellular ROS and reducing equivalent levels altered ATC radiosensitivity in a predictable manner. Irradiation of ATC xenografts resulted in an acute drop in reducing potential measured using HP-MRS, reflecting the shunting of reducing equivalents towards ROS neutralization. Residual tumor tissue post irradiation demonstrated heterogeneous viability. We have adapted HP-MRS/MRSI to non-invasively measure IR mediated changes in tumor reducing potential in real time. Continued development of this technology could facilitate the development of an adaptive clinical algorithm based on real-time adjustments in IR dose and dose mapping.

Figure 1. IR decreases but does not arrest ATC growth.

A) ATC tumors were allowed to grow under control conditions or following irradiation (IR) (5 Gy single fraction) administered at Day 0. Tumors were imaged using Xenogen imaging prior to and following irradiation (please note y-scale is logarithmic). B) ATC tumors were imaged at multiple points during the experimental period and tumor volume calculated as a product of the largest dimensions in the axial, sagittal and coronal planes. Error bars indicate standard deviation and *denotes p-value <0.05 compared to control for the specific time point using two-tailed Student’s t-test.

Figure 2. IR cytotoxicity in ATC is driven by changes in ROS levels.

A) ATC (U-HTH83) intra-cellular ROS levels can be manipulated through the addition of exogenous ROS sources (H₂O₂) or ROS scavenging NAC. B) IR induces a dose dependent increase in intra-cellular ROS levels, which is neutralized by the addition of NAC. C) IR cytoxicity as measured using surviving fraction can be potentiated by the addition of H₂O₂ or reversed by NAC. Data are presented as averages with error bars representing standard deviation. Each experiment was performed at least in duplicate. *indicates p-value <0.05 compared to corresponding control condition unless otherwise indicated as in panel C. All experiments were conducted using the U-HTH83 cell line. (CNT = control, DCFDA = 2′,7′-dichlorodihydrofluorescein diacetate).

Figure 3. Glucose catabolism controls ATC reducing potential and ROS levels.

A) ATC (U-HTH7) cellular reducing potential is maintained largely through glucose catabolism. B) ATC (U-HTH83) inhibition of glucose catabolism using 2-deoxyglucose (2-DG) increases intra-cellular ROS levels in a dose dependent fashion. C) ROS perturbations trigger changes in cellular reducing potential. D) Inhibition of glucose catabolism radiosensitizes ATC cells (U-HTH83) in a dose dependent manner. These effects are reversed by NAC. Data are presented as averages with error bars representing standard deviation. Each experiment was performed at least in duplicate. *indicates p-value <0.05 compared to corresponding control condition unless otherwise indicated. (CNT = control, DCFDA = 2′,7′-dichlorodihydrofluorescein diacetate).

Figure 4. HP-MR detects changes in ROS driven reducing equivalent levels.

A) IR increases metabolically driven ROS, which are scavenged by thiol moieties. B) Thiol moieties are regenerated through utilization of secondary reducing equivalents (NADH/NADPH). C) Conversion of endogenous pyruvate into lactate (ubiquitous in tumor cells) requires reducing equivalents. The rate of conversion therefore indirectly reflects global levels of reducing equivalents. HP-MR piggybacks on this endogenous reaction by allowing measurement of conversion of ¹³C pyruvate (PYR) into ¹³C (LAC). The rate of conversion is therefore directly related to reducing equivalent levels and indirectly related to intra-cellular ROS levels. Inset illustrates the dependence of the PYRLAC reaction on available NADH concentration (experiment performed in vitro using purified enzyme, NADH and ¹³C HP-PYR).

Figure 5. Perturbations in oxidative stress and reducing potential are reflected in altered lactate generation.

A) Endogenous reactive oxygen species (ROS) are generated by multiple cellular processes in both the mitochondria and cytoplasm. Exogenous ROS can increase the free radical burden inside the cell. Reducing equivalents in the form of NAD and NADP moieties are generated through multiple metabolic pathways and can cycle rapidly throughout various cellular compartments. Reducing equivalents are utilized by the cell to neutralize ROS. The conversion of pyruvate into lactate requires the presence of NADH; the conversion rate of labeled pyruvate into lactate therefore is an indirect measure of global cellular reducing potential and ROS stress. B) ATC cells were exposed to varying dose of radiation. Cells were harvested at indicated time points and lactate production was assayed biochemically. Data are presented as means, with error bars indicating standard deviation. *indicates p-value compared to control time point <0.05 using a two-tailed Student’s t test.

Figure 6. HP-MRS can detect both acute and chronic IR induced changes in ATC reducing potential.

A) Control (n = 4) and irradiated (5 Gy) (n = 3) tumors were imaged pre- and post- IR (or sham). Generation of labeled lactate (nLac) and conversion rate constants (Kpl) were calculated and changes from first to second measurement were recorded. B) Control and irradiated ATC tumors demonstrating difference in size using anatomic imaging (T2 weighted sequential slices). C) Reducing potential levels in control (n = 3) and irradiated (n = 3) tumors at 2 weeks post irradiation (5 Gy).

Figure 7. ATC tumors display significant metabolic heterogeneity.

A) Control and irradiated tumors (IR) were serially sectioned and H&E staining was used to evaluate overall tumor architecture. Control tumors were substantially larger, but the majority of the tumor core consisted of non viable tissue (NVT). In contrast the majority of the irradiated tumor volume consisted of apparently viable tissue (VT). Irradiated tumors exhibited a high degree of aberrant cellular morphology as illustrated in the right lower panel inset. B) ATC tumor imaged at 2 weeks following tumor cell injection. Single snapshot imaging was performed 20 seconds after injection of labeled pyruvate. Spatial heat maps were generated from raw data and superimposed onto T2 weighted anatomic images for both pyruvate and lactate. Chemical spectra obtained in two separate voxels demonstrating differential conversion of pyruvate into lactate are shown.