Non-invasive electron paramagnetic resonance imaging detects tumor redox imbalance induced by ferroptosis

Abstract

Targeting ferroptosis, cell death caused by the iron-dependent accumulation of lipid peroxides, and disruption of the redox balance are promising strategies in cancer therapy owing to the physiological characteristics of cancer cells. However, the detection of ferroptosis using in vivo imaging remains challenging. We previously reported that redox maps showing the reduction power per unit time of implanted tumor tissues via non-invasive redox imaging using a novel, compact, and portable electron paramagnetic resonance imaging (EPRI) device could be compared with tumor tissue sections. This study aimed to apply the EPRI technique to the in vivo detection of ferroptosis. Notably, redox maps reflecting changes in the redox status of tumors induced by the ferroptosis-inducing agent imidazole ketone erastin (IKE) were compared with the immunohistochemical images of 4-hydroxynonenal (4-HNE) in tumor tissue sections. Our comparison revealed a negative correlation between the reducing power of tumor tissue and the number of 4-HNE-positive cells. Furthermore, the control and IKE-treated groups exhibited significantly different distributions on the correlation map. Therefore, redox imaging using EPRI may contribute to the non-invasive detection of ferroptosis in vivo.

Keywords: Electron paramagnetic resonance imaging; Redox imaging‌; cancer therapy; ferroptosis; tumor redox status; ‌GPX4; ‌Imidazole ketone erastin; ‌lipid peroxidation.

Graphical abstract

Figure 1.

Effect of IKE treatment in tumor tissues using redox imaging. Changes in the redox status of tumors before and after IKE treatment (50 mg/kg, i.p.) were examined using EPRI. (A) The experimental plan for IKE treatment, redox imaging, and subsequent analyses. (B) Representative comparison of H&E-and redox-stained images. Images of 3CP signal intensity were obtained immediately after 3CP administration. The relative 3CP intensities, averaged over five pixels, are shown in the redox images. (C) A representative example of the time course of the 3CP EPR signal intensity in the tumor tissue. (D) Comparison of the reduction rate constants of 3CP in redox imaging before and after IKE treatment in three mice. Error bars represent S.E. *p < 0.05.

Figure 2.

Effects of IKE on tumor reductants. After the second redox imaging, the amounts of the reducing substances and GPX4 were quantified in the control (vehicle) and IKE-treated groups. (A) HPLC-ECD chromatograms of both groups. Reduced ascorbic acid, reduced cysteine, N-acetylcysteine (as an internal standard), and reduced glutathione levels were measured using HPLC-ECD. The representative chromatograms are shown. (B – D) Comparison of reduced ascorbic acid, reduced cysteine, and reduced glutathione levels in tumors between the control (n = 5) and IKE-treated (n = 5) groups. (E) Representative example of HCT116 xenografts immunostained for GPX4. (F) Proportions of GPX4-positive areas in all tumors were obtained using ImageJ software for each group of three mice. Error bars represent S.E. n.s. = not significant. *p < 0.05.

Figure 3.

IKE induces lipid peroxidation. After the second redox imaging, the 4-HNE levels in the tumors were assessed using immunohistochemistry. (A, B) Representative images of tumors immunostained for 4-HNE in the control and IKE-treated groups. (C) The 4-HNE-positive cells per mm² were quantified by counting four random microscopic fields per mouse in each group of three mice. The black arrows indicate 4-HNE-positive cells. Error bars represent S.E. ***p < 0.001.

Figure 4.

Correlation between redox status and lipid peroxidation. A combined analysis using redox imaging and immunohistochemical staining with 4-HNE was performed. (A) Representative comparison of the redox maps and 4-HNE immunohistochemical staining images. (B) Pearson’s correlation analysis was performed to analyze the correlation between the rate constant of the redox map and the number of 4-HNE-positive cells in each group of three mice. R²= Pearson’s correlation coefficient.