Alternative magnetic field exposure suppresses tumor growth via metabolic reprogramming

Abstract

Application of physical forces, ranging from ultrasound to electric fields, is recommended in various clinical practice guidelines, including those for treating cancers and bone fractures. However, the mechanistic details of such treatments are often inadequately understood, primarily due to the absence of comprehensive study models. In this study, we demonstrate that an alternating magnetic field (AMF) inherently possesses a direct anti‐cancer effect by enhancing oxidative phosphorylation (OXPHOS) and thereby inducing metabolic reprogramming. We observed that the proliferation of human glioblastoma multiforme (GBM) cells (U87 and LN229) was inhibited upon exposure to AMF within a specific narrow frequency range, including around 227 kHz. In contrast, this exposure did not affect normal human astrocytes (NHA). Additionally, in mouse models implanted with human GBM cells in the brain, daily exposure to AMF for 30 min over 21 days significantly suppressed tumor growth and prolonged overall survival. This effect was associated with heightened reactive oxygen species (ROS) production and increased manganese superoxide dismutase (MnSOD) expression. The anti‐cancer efficacy of AMF was diminished by either a mitochondrial complex IV inhibitor or a ROS scavenger. Along with these observations, there was a decrease in the extracellular acidification rate (ECAR) and an increase in the oxygen consumption rate (OCR). This suggests that AMF‐induced metabolic reprogramming occurs in GBM cells but not in normal cells. Our results suggest that AMF exposure may offer a straightforward strategy to inhibit cancer cell growth by leveraging oxidative stress through metabolic reprogramming.

We demonstrate that an alternating magnetic field (AMF) inherently possesses a direct anti‐cancer effect by enhancing oxidative phosphorylation (OXPHOS) and thereby inducing metabolic reprogramming.

1. INTRODUCTION

Glioblastoma multiforme (GBM) is the most common and aggressive brain tumor in adults, with a median survival of about 15 months. ¹ Recent advances in genomic technology have led to a better understanding of key molecular alterations that underlie GBM, suggesting possible therapeutic targets. However, Its prognosis remains poor due to factors such as its invasiveness, resistance to chemotherapy, and the challenges associated with achieving complete surgical resection. ² The standard treatment for GBM involves Temozolomide (TMZ) combined with extended focal radiotherapy. ³ However, this combined approach, consisting of surgery, radiotherapy, and TMZ, has only marginally increased the median survival for patients newly diagnosed with GBM. Consequently, there is a significant need for novel treatment strategies for GBM.

Biological processes are universally governed by physical and/or chemical reactions, suggesting that extracorporeal application of physical forces can modulate cellular functions. Heat and ultrasounds, for instance, have been widely employed in treating various medical conditions under both physiological and pathophysiological circumstances. An example is repetitive transcranial magnetic stimulation (rTMS), a non‐invasive brain stimulation method using a low‐frequency AMF (1–20 Hz). rTMS has gained widespread use and is recommended in clinical practice guidelines for treating mental depression and Parkinson's disease by major medical societies in the US, EU, and other regions. ⁴ , ⁵ Despite the accumulation of clinical evidence, elucidation of the molecular mechanisms underlying such physical treatments has been challenging, owing to the cellular complexity and the multitude of chemical and/or physical reactions involved. Only recently has it been convincingly demonstrated that a classic organic transformation, i.e., a simple chemical reaction, can be modulated using an electric field in a field strength‐ and polarity‐dependent manner. ⁶ Nevertheless, it is widely recognized that living organisms, ranging from bacteria to migratory birds, are influenced by physical forces. ⁷ , ⁸ Magnetic material, for example, is found in a variety of tissues, including the human brain, ⁹ as well as in cellular organelles such as mitochondrial respiratory complex enzymes.

In everyday life, alternating magnetic fields (AMFs) have a wide range of applications. For instance, AMFs in the 20–50 kHz range are utilized in induction cooking heaters. On the other hand, electromagnetic waves spanning from megahertz to gigahertz bands are employed in cellular phones and microwave ovens. In our experiment, the frequency of the AMF used was around 200 kHz. This frequency is higher than that used in induction cooking heaters but lower than the range for microwaves. There is a notable scarcity of research and reports concerning frequencies near 200 kHz. This suggests that this particular area of AMF research might be considered a ‘black box’ within the field.

In this study, we have identified that an AMF inherently exhibits direct anti‐cancer effects by enhancing oxidative phosphorylation (OXPHOS) and thus inducing metabolic reprogramming. Acute and excessive production of reactive oxygen species (ROS) was found to suppress cancer cell proliferation. Our approach involves using much higher electricity through the coil to generate an AMF compared with other devices. Due to the high electricity requirement, our device is not portable. However, the anti‐cancer efficacy of AMF is notable in various cancer cells, including GBM cells. We provide evidence of the anti‐cancer effects of AMF and elucidate its mechanism. Herein, we demonstrate the metabolic reprogramming of cancer cells through physical stimulation alone, without the use of any drugs or other media. Looking forward, this concept could be pivotal in developing novel devices for treating GBM and other types of cancer.

2. MATERIALS AND METHODS

2.1. Electric devices

An AMF was driven by a transistor inverter (Hot Shot, Ameritherm Inc., NY, USA) and generated by a solenoid copper coil (resistivity: 1.673 × 10⁻⁸ Ωm) wrapped in insulation tape with an inner diameter of 39 mm, outer diameter of 51 mm. The coil is a hollow pipe, with a pipe diameter of 6 mm and a spiral pitch of 9 mm. Experiments were performed at a frequency of 160–320 kHz and a current of 0–250 ampere root mean square (Arms). ¹⁰ , ¹¹ , ¹² , ¹³ Although an electric current is applied to the coil, the coil is cooled by the water flowing through the pipe, thus the coil temperature rise is suppressed. Insulating tape is wrapped around the coil surface to prevent short circuits, but it has no effect on the magnetic field generated. Magnetic flux density was simulated with JMAG Designer software, version 22.1 (JSOL Corporation, Tokyo, Japan), employing the 3D finite element method (simulation type: magnetic field analysis (frequency response)). Coil temperature was measured with an optical thermometer and thermograph. A thermometer (fiber optic thermometer FL‐2400, Anritsu Meter Co., Tokyo, Japan) or a thermograph (InfraRed camera, Nippon Avionics Co., Ltd., Tokyo, Japan) was used to determine temperature both in vivo and in vitro. ¹²

2.2. Reagents and cell culture

Human GBM cell lines, U251 MG‐Luc (U251, JCRB1386) and A‐172 (A172, JCRB0228) were purchased from the Japanese Collection of Research Bioresources (JCRB) Cell Bank. T98G (T98, CRL‐1690), U‐87 MG (U87, HTB‐14), and LN‐229 (LN229, CRL‐2611) were purchased from American Type Culture Collection (ATCC, VA, USA). The human pancreatic cancer cell line PANC‐1 (PANC1, RCB2095) and human breast cancer cell lines, MDA‐MB‐453 (MDAMB453, RBC1192) were purchased from RIKEN BioResource Research Center (RIKEN BRC). human breast cancer cell line, MDA‐MB‐231 (MDAMB231, HTB‐26) was purchased from ATCC. d‐Luciferin was purchased from Promega (WI, USA). N‐Acetylcysteine (NAC) was purchased from Sigma‐Aldrich (MO, USA). Potassium cyanide (KCN) was purchased from Fujifilm Wako (Tokyo, Japan). For all cell culture experiments, the control samples (non‐AMF‐stimulated group) were placed in the same incubator as the AMF‐stimulated group samples.

2.3. Sodium 2,3,‐bis(2‐methoxy‐4‐nitro‐5‐sulfophenyl)‐5‐[(phenylamino)‐carbonyl]‐2H‐tetrazolium inner salt (XTT) assay

Cell proliferation assay was performed using a commercial kit, XTT Cell Proliferation Assay Kit (ATCC, VA, USA), as previously described. ¹⁴

2.4. Real‐time cell growth assay

In vitro cell proliferation was measured using the xCELLigence (Agilent, CA, USA) Real‐TimeCellular Analysis system. ¹¹

2.5. Cell cycle analysis

Cell cycle analysis was performed using the Cycletest™ Plus DNA Reagent Kit (BD Biosciences, CA, USA) according to the manufacturer's protocol. ¹⁵ , ¹⁶

2.6. Western blotting

Western blot analyses were performed as previously described. ¹⁰ , ¹¹ Briefly, cells were lysed and sonicated in RIPA buffer (Thermo Scientific, IL, USA). Equal amounts of protein were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS‐PAGE). After electrophoretic separation, protein bands were transferred to Millipore Immobilon‐P membrane followed by immunoblotting with antibodies against molecules of interest. The following primary antibodies were used for immunoblotting: GAPDH (Cell Signaling, MA, USA), PGC1α (abcam, Cambridge, UK), p‐p53 (Ser15) (Cell Signaling), p53 (Cell Signaling), p21 Waf1/Cip1 (12D1) (Cell Signaling), p‐CDK‐2 (Thr160) (Cell Signaling), CDK2 (78B2) (Cell Signaling), Cyclin A (Santa Cruz, Texas, USA), Cyclin B1 (Cell Signaling), Cyclin D1 (DCS6) (Cell Signaling), Cyclin E1 (Cell Signaling), MnSOD (SOD‐2 (E‐10), Santa Cruz), and Cu/ZnSOD (SOD1, Cell Signaling).

2.7. Evaluation of AMF‐induced anti‐cancer effects in a mouse subcutaneous tumor model

We created a mouse subcutaneous tumor model to examine the AMF‐induced anti‐cancer effect. ¹³ U87 cells (1.0 × 10⁷ cells/body) were injected into the back of each mouse (n = 5). LN229 cells (5.0 × 10⁶ cells/body) were injected into the back of each mouse (n = 4). At 7 days after injection of tumor cells, 227 kHz of AMF (250 Arms) was exposed to the tumor. Tumor size was measured manually for 2 weeks. Manual measurement was performed every day. Tumor volume and regression rate were calculated using the following formula ¹⁷ :

$$\text{Tumor volume}\left(\mathrm{TV}\right)\left({\mathrm{mm}}^3\right)=\text{length}\times {\left(\text{width}\right)}^2/2$$

$$\text{Regressionrate} \text{=} \text{TV}/{\mathrm{TV}}_{\left(\text{day0}\right)}\times 100$$

2.8. Evaluation of AMF‐induced anti‐cancer effects in a mouse brain tumor model

Female BALB/c nu/nu mice (17–23 g) (Japan SLC, Inc., Shizuoka, Japan), 6 weeks old, were anesthetized with isoflurane. ¹³ In the case of brain injection, a burr hole was made in the skull 0.5 mm anterior and 2.0 mm lateral to the bregma before tumor cells were stereotactically injected by a 30‐gauge injection cannula to a depth of 4.0 mm. The injection volume was 10 μL per body. ¹⁷ , ¹⁸ U87 cells expressing the luciferase gene were injected stereotactically into the brain (1.0 × 10⁶ cells/body). At 7 days after the injection of tumor cells, 227 kHz of AMF (250 Arms) was exposed to the same location of the brain (n = 6). Exposure time was 30 min per body and per day. Regarding the control, the observation continued outside for the same duration and in the same environment as the AMF treatment group under anesthesia. Tumor size was determined from the luciferin‐induced photon flux, which was measured once a week for 3 weeks. When mice carrying U87 cells were injected intraperitoneally with d‐luciferin (4.7 mg/body), the tumors emitted a visible light signal that could be monitored using an in vivo imaging system (IVIS, Xenogen, Alameda, CA, USA), as previously described. ¹³

2.9. Microarray analysis

The effect of AMF on mRNA expression profiling was evaluated by microarray analysis. ¹⁹ U87 and NHA cells were maintained in control or AMF for 30 min followed by total RNA extraction from U87 cells 30 min after AMF exposure (227 kHz, 250 Arms). Microarray experiments were carried out using SurePrint G3 Human GE 8 × 60 K Ver. 3 Microarray (Agilent Technologies, CA, USA) with total RNA as starting material according to the manufacturer's protocol. To identify significantly enriched functional pathways, the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were analyzed using the KEGG Orthology Based Annotation System 3.0. ²⁰

2.10. iTRAQ labeling and LC/MS analysis

iTRAQ labeling with iTRAQ® Reagent multiplex Assay Kit (SCIEX, Tokyo, Japan) was performed according to the manufacturer's protocol (Filgen, Nagoya, Japan). The labeled peptides mixture was concentrated by Titansphere® Phos‐Tio Kit (GL Sciences Inc., Tokyo, Japan). LC–MS/MS analysis was performed on a SCIEX TripleTOF™ 5600 System and DiNa system (LC) (SCIEX, Tokyo, Japan). Raw MS data were acquired using a data‐dependent top 10 method. Finally, standard protein identification and database search were performed using ProteinPilot™ Software 4.5 (SCIEX). The peptide false discovery rate was set at 0.05. Protein quantification was calculated by the median protein ratio of unique peptides of the protein.

2.11. Mitochondrial test

Measurement of mitochondrial potential was carried out using MitoTracker Red CMXRos (Thermo Fisher Scientific) as previously described. ¹¹ Staining intensity of the mitochondria was analyzed by NIS Element software (Nikon, Tokyo, Japan).

2.12. Measurement of ROS

Measurement of ROS was performed as reported. ²¹ GB cells (U87, LN229 and NHA cells) were incubated overnight in 96‐well plates (1.0 × 10⁴ cells per well) and then exposed to AMF (227 kHz, 250 Arms) at 37°C for 24 h. Intracellular ROS was measured using a fluorescent dye, 2′,7′‐dichlorofluorescein diacetate (DCFH‐DA; Sigma, Japan). ROS production was measured using a microplate reader equipped with a spectrofluorometer (ARVO‐Mx, PerkinElmer, MA, USA).

2.13. Reverse transcriptase‐polymerase chain reaction (RT‐PCR)

Isolation of total RNA, generation of cDNA and RT‐PCR analysis were done as previously described. ²² The sequences of the specific primers are as follows: human mtDNA (forward, 5′‐CACCCAAGAACAGGGTTTGT‐3′; reverse, 5′‐TGGCCATGGGTATGTTGTTA‐3′) and GAPDH (forward, 5′‐CCCATCACCATCTTCCAGGAGCG‐3′; reverse, 5′‐GGCAGGGATGATGTTCTGGAGAGCC‐3′). The abundance of each gene was determined relative to that of the GAPDH transcript.

2.14. Extracellular acidification rate (ERCA)

Extracellular acidification was performed using Glycolysis Assay (ab197244, abcam) according to the manufacturer's protocol.

2.15. Oxygen consumption rate (OCR)

OCR was performed using the Mitopress Xtra HS Method (No. 600800; Cayman Chemical, MI, USA) according to the manufacturer's protocol.

2.16. Data analysis and statistics

Statistical comparisons among groups were performed using Student's t‐test or one‐factor analysis of variance (ANOVA) with the Bonferroni post hoc test. A p‐value of less than 0.05 was considered statistically significant. Unless otherwise indicated, significance levels are denoted as follows: ns, not significant; *p < 0.05; **p < 0.01; ***p < 0.001.

3. RESULTS

3.1. AMF condition

The frequency and intensity range of the current AMF (227 kHz and 250 A, 18 mT) is unusual and not commonly generated in laboratory environments. Consequently, there have been few biological studies conducted using these parameters in the past. ¹⁰ , ¹³ , ²¹ We have indicated the magnetism condition of the AMF through both actual and simulated measurements (Figure 1A–D). The experimental setup involved a coil designed to provide a high‐frequency alternating current (AC) magnetic flux. This AC current was generated by a predetermined resonant frequency, created by a coil with specifically designed inductance, resistance, and capacitance (data not shown). The AC magnetic field was then produced in accordance with Ampere's circuital law. We measured the magnetic flux density distribution in the axial direction of the coil with a search coil. Our experimental data on the generated magnetic flux density were consistent with the simulation data (Figure 1C,D). Within the coil, the magnetic flux density measured 18 mT at the center and increased progressively closer to the coil (Figure 1E).

FIGURE 1.

Configuration and measured and simulated magnetic flux density B in the coil. (A) Photograph of a solenoid coil and the AMF generator (left). Photograph showing a solenoid coil wrapped in insulation tape (right). (B) Details of the shape and size of the solenoid coil. (C) Simulated magnetic flux density distribution in the coil cross‐section (AMF: 227 kHz, 250 Arms). The color scale represents magnetic flux density values, with red indicating high density and blue indicating low density. (D) Comparison of measured and simulated magnetic flux density B along the coil axis (cm). This schematic represents simulation values, with solid lines indicating the simulated data. The red dots denote actual measured values. (E) Simulated magnetic flux density B in the radial direction (cm) of the coil. This schematic represents simulation values, with solid lines indicating the simulated data.

3.2. AMF inhibits cancer cell growth

We discovered that an AMF of 227 kHz effectively inhibited the proliferation of various cancer cells, including GBM cells and other cancer cell lines. The frequency range demonstrating this inhibitory effect was relatively narrow, specifically between 198 and 227 kHz. Frequencies lower than 160 kHz or higher than 320 kHz, did not show significant inhibition in human GBM cells (U87 and LN229) (Figure 2A). The extent of inhibition was also dependent on field intensity (Arms), with the maximum intensity tested being 250 Arms due to equipment limitations (Figure&amp;#x000A0;2B). The inhibitory effect increased with AMF exposure time (min) but plateaued after 30 min (Figure 2C). AMF also reduced cell viability in a variety of tumor cell lines, including other GBM cells (U251, T98, and A172), and other tumor types such as the human pancreatic cell line (PANC1) and human breast cancer cell lines (MCF7, MDA‐MB‐231, MDA‐MB‐453) (Figure 2D). However, it did not affect non‐cancerous human cells [normal human astrocyte (NHA), human cardiac fibroblast (HCF), human umbilical vein endothelial cells (HUVEC)].

FIGURE 2.

Suppression of cancer cell proliferation by AMF at 227 kHz for more than 30 min. (A) The effect of different frequencies (kHz) of AMF (250 Amrs) on the proliferation of GB cell lines (U87 and LN229). XTT cell proliferation assays were conducted at various AMF frequencies (kHz) for 30 min, with evaluation occurring 24 h post‐AMF exposure (n = 4, *p < 0.05, **p < 0.01, ***p < 0.001 vs. 0 kHz). (B) The impact of varying electric current intensities (Arms) in AMF (227 kHz) on the proliferation of GBM cell lines (U87 and LN229) (n = 4, ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001 vs. 0 Arms). (C) The effect of different exposure durations (min) to AMF (227 kHz, 250 Amrs) on the proliferation of GBM cell lines (LN229, U251) (n = 4, ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001 vs. 0 min). (D) The influence of AMF (227 kHz, 250 Amrs) on other GB cell lines (U251, T98, and A172), a pancreatic cell line (PANC1), human breast cancer cell lines (MCF7, MDA‐MB‐231, MDA‐MB‐453), normal human astrocyte (NHA), human cardiac fibroblast (HCF), and human umbilical vein endothelial cells (HUVEC) (n = 4, ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001 vs. CTRL). (E, F) Continuous monitoring of cell growth with and without a 30‐min AMF exposure (227 kHz, 250 Arms) in U251 and LN229 cell lines. In vitro cell proliferation was measured using the xCELLigence Real‐Time Cellular Analysis system. (G) Cell cycle analysis 3 and 24 h post‐AMF exposure (227 kHz, 250 Arms, 30 min), revealing the inhibitory effect of AMF, notably the induction of S and G2 phase arrest (n = 4, ns, not significant, **p < 0.01, ***p < 0.001 vs. CTRL). (H) Immunoblot analysis of phosphorylated and unphosphorylated forms of p53, p21, CDK2, Cyclin A, Cyclin B1, Cyclin D1, Cyclin E, and GAPDH 24 h after a 30‐min AMF exposure (227 kHz, 250 Arms) (n = 4, ns, not significant, **p < 0.01, ***p < 0.001 vs. CTRL).

The impact of exposing human GBM cells (U87 and LN229) to AMF was further investigated using real‐time, label‐free analysis with the xCELLigence system over a period of 96 h. As anticipated, control cells demonstrated rapid proliferation. However, both cells exposed to AMF for a continuous period displayed significantly slower proliferation rates (Figure 2E,F). This observation was notable, particularly since the cells were exposed to AMF for only 30 min prior to the assay. Subsequently, we assessed the effect of AMF on the cell cycle. We observed that AMF exposure led to an accumulation of cells in the S and G2 phases of the cell cycle 24 h after exposure (Figure 2G). As expected, AMF did not alter the cell cycle in NHA (Figure S1A). The AMF exposure resulted in decreased phosphorylation of cyclin‐dependent kinase 2 (CDK2) and reduced protein expression of cyclin A and cyclin B1. Conversely, AMF increased the phosphorylation of p53 and the protein expression of p21, cyclin D1, and cyclin E1 (Figure 1H, Figure S13).

3.3. AMF inhibits GBM growth in mice

Next, we evaluated the efficacy of AMF in vivo using mouse models. The anti‐cancer effects of AMF were observed in both a subcutaneous GBM model and an intracranial GBM model. Mice implanted with GBM cells (U87 and LN229) in the subcutaneous tissue of the thigh were placed within a coil to expose them to AMF at 227 kHz and 250 Arms. ¹³ The tumor was positioned 5 mm from the edge of the coil, with an estimated magnetic flux density of 18 milli tesla (mT) at the tumor site. Mice were exposed to AMF for 30 min daily, for a total of five sessions. After 2 weeks, tumor sizes in both AMF‐exposed groups were significantly reduced compared with the control group (Figure 3A,B). Immunohistochemical analysis with Ki‐67 staining indicated that AMF significantly suppressed GBM (U87) tumor growth (Figure 3C). To verify the antitumor effect of AMF in different types of cancer, breast cancer (MDA‐MB‐231) and pancreatic cancer (PANC‐1) transplant models were developed and investigated. The results showed a significant treatment effect, as assessed by tumor volume, after conducting AMF treatment five times a week for 2 weeks (Figure S2A,B).

FIGURE 3.

Anti‐cancer effects of AMF in mice subcutaneous and brain GBM models. (A) Changes in the volume (mm³) of subcutaneous tumors (U87 and LN229 cells) over 14 days in the control group versus the AMF treatment (227 KHz, 250 Amrs, 30 min per session) group. (B) Photographs of tumors from subcutaneous implantation of U87 cells (control and AMF treatment groups). (C) Representative images of brain tumor sections following H&E staining and Ki‐67 staining. The left images are from the control group, and the right images are from the AMF treatment group (n = 4). The graph shows the ratio of Ki‐67 positive cells in both the control and AMF treatment groups. White arrows indicate positive area. Calibration bar: 500 μm. (D) Schedule of AMF treatment for the mouse brain GBM model: 227 kHz, 250 Arms, 30 min per session, 5 times per week for a duration of 2 weeks. (E) Representative images from the in vivo imaging system (IVIS) images of mouse brains at 6, 11, 16, and 21 days post‐implantation of U87 cells. (F) Luminescent intensity comparison between the control and AMF treatment groups. The graph depicts the time course of tumor volume changes (n = 6). (G) Overall survival curve of mice in the study. The blue line represents the survival percentage (%) in the AMF‐treated group, while the black line represents survival in the control group (without AMF treatment).

A similar protocol was applied in a mouse brain tumor model, using U87 cells transfected with a luciferase‐encoding vector. ¹³ Tumor growth was monitored by measuring photon flux using an in vivo imaging system (IVIS). Mice underwent 30‐min AMF treatments five times a week for 2 weeks, starting 1 week after cell implantation (Figure 3D). While the control group exhibited significant tumor growth, the tumors in the AMF‐treated group, which received treatments five times a week remained stable even after 3 weeks (Figure 3E,F, Figure S3). Most importantly, AMF treatment significantly extended the overall survival of the mice without causing major side effects (Figure 3G, Figure S4).

3.4. AMF induces metabolic reprogramming

To gain deeper mechanistic insights into the effects of AMF exposure on GBM cells and tumor growth, we conducted a microarray analysis 30 min post a 30‐min AMF exposure (Figure 4A). The microarray data, analyzed through the KEGG pathway analysis, revealed significant differences in downregulated signals (log < −1.5) between AMF‐exposed and control samples in U87 cells (Figure 4B). These results indicated that AMF exposure led to a decrease in various metabolic pathways, including glycolysis. Interestingly, in NHA cells, microarray analysis also showed changes in metabolic pathways, including glycolysis (Figure S5A,B).

FIGURE 4.

Comprehensive analysis of protein expression and phosphorylation induced by AMF in GBM cells. (A) Timeline for conducting the microarray analysis. (B) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis illustrating pathways that are enriched with differentially expressed genes. Comparison made between proteins collected from U87 cells exposed to 30 min of AMF (250 kHz, 250 Arms) and those from U87 control cells (without AMF). (C) Timeline for the iTRAQ phosphorylated protein analysis. (D) Changes in protein expression induced by AMF (250 kHz, 250 Arms) in U87 cells. Histograms display the count of proteins whose phosphorylation levels were either upregulated (red) (>1.5‐fold) or downregulated (blue) (<1.5‐fold) following AMF exposure at 0.5, 1, and 2 h. (E) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis showing pathways enriched with differentially expressed genes. Comparison made between proteins collected from U87 cells 0.5 h post‐AMF exposure and proteins from GBM cells not exposed to AMF.

Further, we utilized iTRAQ‐based proteomics analysis via liquid chromatography‐mass spectrometry (LC–MS/MS) and conducted a corresponding KEGG pathway analysis (Figure 4C). This analysis identified changes in over 600 proteins 30 min after AMF exposure (Figure 4D). KEGG pathway analysis revealed that AMF exposure affected protein phosphorylation in several pathways, notably glycolysis/gluconeogenesis (Figure 4E). In NHA, LN229, MCF7, and PANC‐1 cells, the application of AMF was observed to cause changes in protein phosphorylation over time (Figures S5C,D–S7). Additionally, through KEGG pathway analysis, changes in the metabolic pathway observed in NHA, LN229, MCF7, and PANC‐1 cells were also detected in these cells. We extracted proteins that changed commonly in U87 and LN229 cells. Among the individual proteins related to metabolism, pyruvate kinase M2 (PKM2) was found to be inhibited in both. As this enzyme plays an important role in shifting the energy production pathway to glycolysis, it seems to suggest that the AMF may induce metabolic reprogramming in cancer cells. The results from mass spectrometry showed that the number of proteins altered by the AMF in NHA cells was significantly less compared with U87 cells. Additionally, although we extracted proteins with common changes in both U87 and NHA cells using iTRAQ results, no individual proteins, including those related to metabolism, were found to be common between them. This may indicate differences in the effects of AMF on cancer cells and normal cells.

These findings suggest that while there are differences depending on the type of cancer and whether the cells are cancerous or normal, AMF can influence metabolism. Both microarray and mass spectrometry analyses consistently showed a downregulation in the glycolysis/gluconeogenesis pathway. Consequently, we decided to focus primarily on the effects of AMF on glycolysis in GBM cells.

3.5. AMF promotes ROS production and increases MnSOD expression

Given that AMF promptly increased oxygen consumption, next we assessed its impact on mitochondrial function. We found that AMF exposure elevated mitochondrial membrane potential after 6 h (Figure 5A). As mitochondria are major contributors to endogenous ROS production, ²³ with approximately 1% of oxygen (O₂) used for ROS generation, we further investigated this aspect. To evaluate ROS production, we used 2′, 7′‐dichlorodihydrofluorescin diacetate (DCFH‐DA). A 30‐min AMF exposure in U87 cells led to a time‐dependent increase in ROS production, measured at intervals of 0, 15 min, and 3, 6, 12, 24, and 48 h post‐exposure (Figure 5B). A similar trend was observed in LN229 cells (Figure S8A). In contrast, AMF exposure did not increase ROS production in NHA cells (Figure S1D). Both KCN, a mitochondrial complex IV inhibitor and NAC, an ROS scavenger counteracted the AMF‐induced ROS production in U87 cells (Figure 5C,D). Similar results were observed in LN229 cells, consistent with the U87 data (Figure S8B,C).

FIGURE 5.

AMF promotes ROS production and increases MnSOD. (A) Increase in mitochondrial membrane potential in U87 cells following 30 min of AMF exposure (n = 4, ns, not significant, *p < 0.05, ***p < 0.001 vs. CTRL). (B) Measurement of ROS production in U87 cells at 0, 24, and 48 h post 30‐min AMF exposure (n = 4, ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001 vs. CTRL). (C) ROS production in U87 cells treated with 10 μM potassium cyanide (KCN) following 30 min of AMF exposure (n = 4). (D) ROS production in U87 cells treated with 5 mM NAC following 30 min of AMF exposure (n = 4). (E) Immunoblot analysis of MnSOD, Cu/ZnSOD, and cytochrome c protein expression in GBM cells (LN229) exposed to AMF for 30 min at various time points (0, 5, 15, 30 min, 1, 6, 12, 24, 48 h) (n = 4). (F) Immunoblot analysis of MnSOD phosphorylation and protein expression 6 h post‐AMF exposure for 30 min, with or without 10 μM KCN in LN229 cells (n = 4). (G) Immunoblot analysis of MnSOD phosphorylation and protein expression 6 h post‐AMF exposure for 30 min, with or without 5 mM NAC in LN229 cells (n = 4). In all these experiments, the AMF was conducted under conditions of 227 kHz and 250 Arms.

Indeed, we found that the expression of manganese superoxide dismutase (MnSOD; SOD2 or mitochondrial SOD) and cytochrome c, but not copper/zinc superoxide dismutase (Cu/ZnSOD; SOD1), increased following AMF exposure (Figure 5E). However, AMF did not elevate the expression of either MnSOD or Cu/ZnSOD in NHA (Figure S9). Furthermore, when U87 cells were exposed to AMF in the presence of KCN or NAC, the expression of MnSOD was reduced (Figure 5F,G). In contrast, Cu/ZnSOD did not significantly change.

Furthermore, when U87 cells were exposed to AMF in the presence of KCN or ROS scavengers (NAC or ascorbic acid/vitamin C), the anti‐cancer effect of AMF was reduced, as shown by dose‐dependent decreases in cell viability using XTT assays in both U87 (Figure S10A–C) and LN229 cells (Figure S11A–C). Therefore, ROS production following AMF exposure may contribute to its anti‐cancer effects.

3.6. AMF promotes mitochondrial biogenesis

To further understand the molecular mechanism by which AMF‐induced ROS promote mitochondrial biogenesis and enhance oxidative metabolism, we investigated the expression of genes, including transcription factors and coactivators that regulate mitochondrial biogenesis. PGC1α, a key regulator of mitochondrial function, oxidative metabolism, and energy homeostasis in various cells, ²⁴ , ²⁵ is known to increase under oxidative stress, thereby regulating components of the ROS defense system. ²⁵ Consistent with this observation, an increase in mitochondrial DNA (mtDNA) was noted following exposure to AMF (Figure 6A). Corresponding to the rise in ROS, there was also a time‐dependent increase in PGC1α protein expression post‐AMF exposure (Figure 6B). In contrast, AMF exposure did not increase PGC1α protein expression in NHA cells (Figure S1E). Notably, both KCN and NAC reversed the AMF‐induced increase in PGC1α expression (Figure 6C,D).

FIGURE 6.

AMF promotes mitochondrial biogenesis. (A) Real‐time PCR analysis of mitochondrial DNA (mtDNA) at 6, 24, and 48 h post 30‐min AMF exposure in LN229 (n = 4, ns, not significant, *p < 0.05 vs. CTRL). (B) Immunoblot analysis of PGC1α protein levels at 15 min, 3, 6, 12, and 24 h following 30 min of AMF exposure in LN229 (n = 4, *p < 0.05, ***p < 0.001 vs. CTRL). (C) Immunoblot analysis of PGC1α protein levels in the presence of 10 μM potassium cyanide (KCN) 6 h post 30‐min AMF exposurein LN229 (n = 4). (D) Immunoblot analysis of PGC1α protein levels in the presence of 5 mM NAC 6 h after 30 min of AMF exposure in LN229 (n = 4). In all these experiments, the AMF was conducted under conditions of 227 kHz and 250 Arms.

3.7. AMF‐induced metabolic reprogramming occurs in GBM cells but not in normal cells

To assess glycolysis with and without AMF exposure, we measured the ECAR and the OCR in the supernatant medium of U87 cells for 5 h following a 30‐min AMF treatment. The AMF exposure gradually decreased ECAR and increased OCR, suggesting that AMF inhibited glycolysis and enhanced OXPHOS (Figure 7A,B). The effects of AMF on ECAR and OCR in U87 cells were found to be frequency dependent (Figure 7C,D) and intensity dependent (Figure 7E,F). Similar results were observed in LN229 cells, consistent with the U87 data (Figure S12A–D). Notably, AMF exposure did not alter ECAR and OCR in NHA (Figure S1B,C), indicating a differential response to AMF in normal and cancer cells.

FIGURE 7.

Metabolic reprogramming in GBM cells induced by AMF. (A) Extracellular acidification rate (ECAR) levels of GBM cells (LN229) measured after 30 min of AMF exposure. (B) Oxygen consumption rate (OCR) levels of GBM cells following 30 min of AMF exposure (C) ECAR levels of GBM cells under different kHz conditions immediately after 30 min of AMF exposure. (D) OCR levels of GBM cells at various kHz conditions immediately following 30 min of AMF exposure . (E) ECAR levels of GBM cells at different intensities (Arms) immediately after 30 min of AMF exposure. (F) OCR levels of GBM cells at various intensities (Arms) immediately following 30 min of AMF exposure. (G) ECAR levels of GBM cells treated with 10 μM KCN after 30 min of AMF exposure . (H) OCR levels of GBM cells in the presence of 10 μM KCN following 30 min of AMF exposure. (I) ECAR levels of GBM cells treated with 5 mM NAC after 30 min of AMF exposure. (J) OCR levels of GBM cells in the presence of 5 mM NAC following 30 min of AMF exposure (n = 4, ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001 vs. CTRL). (K) Proposed mechanism of AMF‐mediated cell cycle alteration in GBM cells. In all these experiments, the AMF was conducted under conditions of 227 kHz and 250 Amrs,

Interestingly, the changes in ECAR and OCR were negated by potassium cyanide (KCN), a mitochondrial complex IV inhibitor, in both U87 and LN229 cells (Figure 7G,H, Figure S12E,F). In contrast, the ROS scavenger (NAC) did not reverse these changes (Figure 7I,J, Figure S12G,H). This suggests that AMF impacts mitochondrial function, leading to metabolic reprogramming. Such reprogramming is reminiscent of the anti‐Warburg effect, a shift from anaerobic glycolysis to OXPHOS, known to decrease cancer cell viability. ²⁶ , ²⁷

4. DISCUSSION

The current study has demonstrated that AMF exposure alone inhibits GBM cell growth. This mechanism likely involves enhancing mitochondrial oxidative phosphorylation (OXPHOS), leading to increased ROS production (Figure 7K). Excess ROS results in the consequent inhibition of cancer cell growth and therefore promotes mitochondrial biogenesis. A previous report that showed AM alternating electric fields to have antitumor effects against cancer. ²⁸ Alternating electric fields are generated between two electrodes and are applied to biological tissues by contacting the tissues with these electrodes. The electric field was affected by the dielectric properties, which varied among different biological tissues, leading to a distribution of electric fields within the body. In contrast, AMFs were generated in a non‐contact manner within a coil and are applied to biological tissues without direct contact. The magnetic field was affected by the magnetic permeability, which is the same for biological tissues as it is for air, resulting in a uniform magnetic field within the body. This allows the AMF to easily affect areas deep within the body, among other characteristics.

GBM is the most common primary malignant brain tumor in adults. Median overall survival from diagnosis of 14.6–16.7 months has been reported in clinical trials because of its invasiveness and treatment resistance due to tumor heterogeneity and microenvironment, even with maximal safe surgical resection followed by advanced chemotherapy and extended focal radiotherapy. Recently, it was shown that alternating electric fields (200 kHz, 1–3 V/cm) are effective in treating GBM (NovoTTF‐100A system, Novocure Ltd., Haifa, Israel), ²⁹ , ³⁰ although these employ much smaller fields compared with our study. Electric field exposure is thought to selectively destroy GBM cells via mitotic arrest and apoptosis. ³¹ While the exact molecular mechanisms of electric field therapy remain largely unknown, multiple large‐scale clinical trials have demonstrated its effectiveness, leading to recommendations in clinical guidelines. However, this therapy only modestly increases median overall survival in recurrent GBM by 0.6 months, highlighting the need for more effective and less harmful therapies using physical energy. ³²

The AMF device comprises a magnetic coil, a generator producing high amplitude electric current, and a cooling system for the coil. The two main benefits of AMF are its short treatment duration per session and its non‐contact nature. Only a 30 min of daily AMF exposure is required, allowing shared use of the device among multiple patients, potentially reducing costs. Additionally, AMF is advantageous in terms of cleanliness, as the device does not contact the coil. Unlike electric fields, which may not be uniform due to the varying electrical conductivity of biomaterials, the magnetic field is uniform as most biomaterials have similar permeability. No adverse effects of AMF were identified in our study.

The exact mechanism by which AMF increases OXPHOS is not fully understood. In the oxidation of cytochrome c, for example, the antimagnetic Fe²⁺ and paramagnetic Fe³⁺ ions alternate their ionic states to produce ATP. It is speculated that AMF at this specific frequency may interfere with this reaction. Further exploration is needed to examine AMF's effect on each chemical/biochemical reaction involved in this process. Synthesizing these findings, it is tentatively hypothesized that AMF exposure initially increases OXPHOS and ROS. However, this increase appears to be transient due to mitochondrial damage caused by heightened oxidative stress, leading to consequent inhibition of cancer cell growth. In this experiment, AMF showed an antitumor effect on cancer cells but not on normal cells. Furthermore, in cancer cells, AMF decreased ECAR and increased OCR, while there were no changes in normal cells. This suggests that AMF caused the metabolic mode of cancer cells to shift from glycolysis, characteristic of cancer cells, toward OXPHOS, which is more typical of normal cells. This implies that the difference in energy metabolism between cancer cells and normal cells, known as the Warburg effect, is reflected in the differential antitumor effects of AMF. ²⁶ , ²⁷ , ³³

The anti‐cancer effect through oxidative stress is, at least partly, similar to that of quinone chemotherapeutic compounds such as mitomycin C or Adriamycin, which exert cytotoxicity by increasing oxidative stress. These compounds are effective against many cancer cell types, but often come with undesirable side effects. Upon their manifestation, it is challenging to immediately terminate their toxicity due to the time required for drug metabolism and excretion. For these reasons, unlike drugs, AMF, being solely a physical stimulus, not only allows for easier initiation of antitumor effects but also presents the possibility of quickly stopping its action by ceasing exposure.

In summary, we have shown that a specific frequency of AMF (227 kHz) induces a metabolic shift from glycolysis to OXPHOS, leading to suppressed GBM growth in culture and in vivo. The intensity is unusually high, making it previously unexplored for biological effects. Our strategy of inducing oxidative phosphorylation may offer a safe and cost‐effective therapy option. Future applications may extend to other cancer therapies, as preliminary data suggest sensitivity to AMF exposure in other cancer cell lines as well.

AUTHOR CONTRIBUTIONS

Taisuke Akimoto: Data curation; formal analysis; methodology; validation. Md Rafikul Islam: Data curation; formal analysis; methodology. Akane Nagasako: Data curation; methodology; validation. Kazuhito Kishi: Data curation; formal analysis; methodology; writing – review and editing. Rina Nakakaji: Data curation; writing – review and editing. Makoto Ohtake: Methodology; validation; writing – review and editing. Takashi Yamaguchi: Resources; software; supervision. Shigeki Yamada: Data curation; formal analysis; writing – review and editing. Tetsuya Yamamoto: Validation; writing – review and editing. Yoshihiro Ishikawa: Conceptualization; investigation; methodology; writing – review and editing. Masanari Umemura: Conceptualization; data curation; investigation; methodology; project administration; resources; supervision; validation; visualization; writing – original draft; writing – review and editing.

FUNDING INFORMATION

This study was supported in part by the Japan Agency for Medical Research and Development (AMED) (19191258, 23810577), the Japan Society for the Promotion of Science (22K06928), Ricoh Co., Ltd., and the Japan Science and Technology Agency (JST) (JPMJFR205A) (to M.U.).

CONFLICT OF INTEREST STATEMENT

We have received funding for a portion of our research from Ricoh Company, Ltd.

ETHICS STATEMENT

Approval of the research protocol by an Institutional Review Board: Animal experiments were performed according to the Yokohama City University guidelines for experimental animals. All experimental protocols were approved by the Animal Care and Use Committee at Yokohama City University School of Medicine (F‐A‐22‐044).

Informed Consent: N/A.

Registry and the Registration: N/A.

Animal Studies: N/A.

Supporting information

Figure S1.

Akimoto T, Islam MR, Nagasako A, et al. Alternative magnetic field exposure suppresses tumor growth via metabolic reprogramming. Cancer Sci. 2024;115:2686‐2700. doi: 10.1111/cas.16243