Electron Paramagnetic Resonance Implemented with Multiple Harmonic Detections Successfully Maps Extracellular pH In Vivo

Abstract

Extracellular acidification indicates a metabolic shift in cancer cells and is, along with tissue hypoxia, a hallmark of tumor malignancy. Thus, non-invasive mapping of extracellular pH (pHe) is essential for researchers to understand the tumor microenvironment and to monitor tumor response to metabolism-targeting drugs. While electron paramagnetic resonance (EPR) has been successfully used to map pHe in mouse xenograft models, this method is not sensitive enough to map pHe with a moderate amount of exogenous pH-sensitive probes. Here, we show that a modified EPR system achieves twofold higher sensitivity by using the multiple harmonic detection (MHD) method and improves the robustness of pHe mapping in mouse xenograft models. Our results demonstrate that treatment of a mouse xenograft model of human-derived pancreatic ductal adenocarcinoma cells with the carbonic anhydrase IX (CAIX) inhibitor U-104 delays tumor growth with a concurrent tendency toward further extracellular acidification. We anticipate that EPR-based pHe mapping can be expanded to monitor the response of other metabolism-targeting drugs. Furthermore, pHe monitoring can also be used for the development of improved metabolism-targeting cancer treatments.

Figure 1

EPR-based pH measurement setup. (A) Unprotonated and protonated forms of the pH-sensitive free radical probe dR-SG exhibit different hyperfine splitting constants. SG stands for glutathione residue. (B) The first-derivative EPR absorption spectra for dR-SG at 5.6 and 7.6 pH and 37 °C. (C) EPR spectrometer setup with the MHD receiver system (for simplicity, magnetic field gradient coils and power supplies for the magnetic field gradients are not shown). (D) The signal acquisition and spectral data processing scheme. To record 8192 spectral data points for the nth harmonics, the voltage signal of the low-frequency (LF) amplifier output was digitized over a period of 100 ms. The details of the radiofrequency (RF) and electronic components used in the EPR spectrometer are listed in Supplementary Table 1.

Figure 2

Multiple harmonic detections and spectral reconstruction. (A) First-derivative EPR absorption spectra recorded with phase-sensitive detection (PSD) and multiple harmonic detection (MHD) schemes (modulation ratio a = 1.0). (B) First to fifth harmonic spectra of dR-SG (modulation ratio a = 2.0). (C) Signal-to-noise ratio as a function of the number of harmonics involved in MHD (a = 1.0 and 2.0). (D) Peak-to-peak linewidth of the central spectral peak (a = 1.0 and 2.0). (E) Comparison of the signal-to-noise ratios between EPR spectra obtained with PSD and MHD (sample size n = 300). In (C) and (D), plots and error bars represent the mean ± standard error (SE). Circles in (E) are outliers. The pH-sensitive probe dR-SG was dissolved in 2 mL of phosphate-buffered saline, and its concentration was 2 mmol L^–1.

Figure 3

pH mapping of solution samples. (A) Photograph of a 2 mL vial containing dR-SG radical solution (the ruler is in centimeters). (B) The surface-rendered image of EPR signal intensities of the dR-SG vial (image matrix 48 × 48 × 72 with a field of view of 25 mm × 25 mm × 37.5 mm). (C) EPR signal intensity map at the center slice of the visualized space. (D) Reconstructed pH maps for the solution samples with different pH conditions (6.2 to 7.4 pH). (E) Linearity test between the measured and corrected pH values obtained by eq 2. Closed circles and error bars represent the mean ± standard deviation (SD) of the measured (black) and corrected (blue) pH values. The curve in red shows the cubic polynomial that best approximates the measured pH values. The scale bar represents 5 mm. In vitro pH mapping was performed at 37 °C. The concentration of dR-SG was 2 mmol L^–1.

Figure 4

Tumor growth delay due to U-104 treatment. (A) Chemical structure of the CAIX inhibitor U-104 and its administration plan for this experiment. Approximately 10 million MIA PaCa-2 cells were subcutaneously inoculated into the right hind legs of each mouse. (B) Tumor growth curves for the three dose groups. Closed squares and error bars represent the mean ± standard error (SE). Individual tumor volumes are represented as closed circles colored to match the lines. (C) Time at which a fivefold increase in tumor volume (200 to 1000 mm³) occurred. Individual times are shown in matched color-filled circles for each group. The sample sizes during the tumor growth monitoring experiment were 6 for the vehicle control group, 5 for the group treated with 20 mg per kg body weight U-104, and 5 for the group treated with 40 mg per kg body weight U-104.

Figure 5

Influence of U-104 treatment on extracellular pH (pHe) in MIA PaCa-2 mouse xenograft models. (A) Experimental plan for extracellular pH (pHe) mapping and single U-104 administration. (B) Surface-rendered image of the EPR signal intensities for a representative tumor-bearing leg (image matrix 48 × 48 × 48 with FOV of 25 mm × 25 mm × 25 mm), (C, G) MR anatomical maps, (D, H) EPR signal intensity maps, (E, I) pHe maps, and (F, J) histograms of tumor pHe before and after U-104 treatment, respectively. The 2D EPR and pHe maps have an image matrix of 48 × 48 with an FOV of 25 mm × 25 mm. The scale bar represents 5 mm. At the time of both scans, the mice weighed 20 to 26 g. The pHe maps in (E) and (I) were obtained with 20% maximum intensity thresholding in the corresponding EPR signal intensity maps.