Therapeutic potential of pentamidine for glioma‐initiating cells and glioma cells through multimodal antitumor effects

Abstract

Glioma‐initiating cells, which comprise a heterogeneous population of glioblastomas, contribute to resistance against aggressive chemoradiotherapy. Using drug reposition, we investigated a therapeutic drug for glioma‐initiating cells. Drug screening was undertaken to select candidate agents that inhibit proliferation of two different glioma‐initiating cells lines. The alteration of proliferation and stemness of the two glioma‐initiating cell lines, and proliferation, migration, cell cycle, and survival of these two differentiated glioma‐initiating cell lines and three different glioblastoma cell lines treated with the candidate agent were evaluated. We also used a xenograft glioma mouse model to evaluate anticancer effects of treated glioma cell lines. Among the 1301 agents, pentamidine—an antibiotic for Pneumocystis jirovecii—emerged as a successful antiglioma agent. Pentamidine treatment suppressed proliferation and stemness in glioma‐initiating cell lines. Proliferation and migration were inhibited in all differentiated glioma‐initiating cells and glioblastoma cell lines, with cell cycle arrest and caspase‐dependent apoptosis induction. The in vivo study reproduced the same findings as the in vitro studies. Pentamidine showed a stronger antiproliferative effect on glioma‐initiating cells than on differentiated cells. Western blot analysis revealed pentamidine inhibited phosphorylation of signal transducer and activator of transcription 3 in all cell lines, whereas Akt expression was suppressed in glioma‐initiating cells but not in differentiated lines. In the present study, we identified pentamidine as a potential therapeutic drug for glioma. Pentamidine could be promising for the treatment of glioblastomas by targeting both glioma‐initiating cells and differentiated cells through its multifaceted antiglioma effects.

Using drug repositioning, pentamidine—an antibiotic for Pneumocystis jirovecii— was identified as an antiglioma agent. Pentamidine suppressed proliferation of both glioma‐initiating cells and differentiated cells. Pentamidine inhibited signal transducer and activator of transcription 3 activity in glioma cells, and Akt activity was also suppressed in glioma‐initiating cells.

List of abbreviations

BBB

blood–brain barrier

EC₅₀

half maximal effective concentration

GBM

glioblastoma

GIC

glioma‐initiating cell

HPF

high‐power field

IHC

immunohistochemistry

PE

pentamidine

PI

propidium iodide

SOX2

sex‐determining region Y box 2

STAT3

signal transducer and activator of transcription 3

1. INTRODUCTION

Glioblastoma, one of the most common primary brain tumors, is highly malignant, and most patients with GBM die within 2 years of onset. ¹ , ² , ³ , ⁴ , ⁵ Temozolomide, used as the standard chemotherapy for GBM, improves patient survival. ⁶ However, patients with GBM eventually experience tumor relapse due to chemotherapy resistance. ⁷ Previous research has revealed that GBM comprises heterogeneous cell populations, including GICs, ⁸ which are highly resistant to treatments ⁹ , ¹⁰ and contribute to heterogeneity and tumor growth. ¹¹ Therefore, treatment against GICs is urgently needed.

Drug repositioning, the application of known drugs to discover new therapeutic targets for different diseases, ¹² could reduce the cost and time required to develop new drugs. These drugs are promptly used for patients because the appropriate usage rules and side‐effects are already known. Previously, we and others identified several compounds that had some therapeutic effects on glioma. ⁹ , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ However, none of them have been applied to current clinical treatments. ¹⁸ , ¹⁹

Pentamidine, an antiprotozoal drug for pneumonia caused by Pneumocystis jirovecii, suppresses glucose metabolism, amino acid transport, and RNA synthesis. ²⁰ However, the detailed molecular biological mechanisms of PE have not yet been clarified. Only a few studies have mentioned that PE has anticancer effects. ²¹ , ²² To date, no reports have indicated the therapeutic potential of PE against gliomas. Pentamidine can be given during the clinical course of GBM because opportunistic infections caused by lymphopenia are one of the most common side‐effects of temozolomide. ⁶ , ²³

In this study, 1301 candidates were investigated for the treatment of GBM, and PE was extracted as a curative drug for both glioma cells and GICs. We analyzed the anticancer effects of PE on glioma in vitro and in vivo. Detailed analysis of the mechanism of PE revealed multimodal therapeutic effects for glioma.

2. MATERIALS AND METHODS

2.1. Cell culture

Human patient‐derived GICs KGS01 and KGS07 were established at Kanazawa University. Previous reports have confirmed these cell lines as tumor‐initiating cells, as cultured cells have self‐renewal ability in vitro and recapitulate the original tumor in a mouse xenograft model. ⁹ , ²⁴ , ²⁵ These GICs were cultured as previously described. ⁹ , ²⁵

Human GBM cell lines A172, T98G, and U87, were purchased from the European Collection of Cell Culture and were cultured as previously described. ²⁶ In this study, we established differentiated KGS01 (dKGS01) and differentiated KGS07 (dKGS07). These two differentiated GICs were cultured under the same conditions as human GBM cell lines. To confirm the differentiation, western blot analysis revealed that the expression levels of SOX2, which is known as the initiating marker, were decreased in both dKGS01 and dKGS07 as observed in A172, T98G, and U87 cell lines (Figure S1).

2.2. Drug screening

We undertook drug screening of the drug libraries as previously reported. ⁹ , ¹⁴ , ¹⁵ There were three steps in screening against GICs to detect potential agents. First, the GIC line (KGS01) in a 384‐well plate (Corning) was treated with each compound provided from the libraries at three concentrations (1, 5, and 20 μM) for 48 h, followed by a WST‐8 cell viability assay (Dojindo). Compounds showing a 25% dose‐dependent reduction in cell viability were selected. Second, we detected candidate compounds that were already in regulatory approval in Japan and whose effect on gliomas had never been reported. Third, the alamarBlue assay (Bio‐Rad Laboratories) was carried out using two different GIC lines (KGS01 and KGS07) with the following concentrations: middle, EC₅₀; low, half of EC₅₀; and high, twice of EC₅₀. Candidate drugs were selected based on the demonstration of valid cell rate reduction, even at lower concentrations (Figure 1A).

FIGURE 1.

Schema of drug screening and results of analysis using glioma‐initiating cells (GICs). (A) In this study, 1301 compounds were gathered and a three‐step procedure was applied to detect the candidate agents. First, the compounds that showed a 25% greater reduction in viability of GICs in a dose‐dependent manner were selected. Second, candidate compounds with regulatory approval in Japan and with no previously reported effects on glioma were included. Finally, an alamarBlue assay was carried out using two GIC lines with a more detailed concentration distribution. (B) Two different GICs were treated with pentamidine (PE) at 0, 1, and 5 μM, and each viability curve was analyzed. (C) Bar graphs indicate the number of spheres with diameters larger than 100 μm in each GIC line. These data were obtained 10 days after PE treatment. (D) Western blot analysis indicated the expression of sex‐determining region Y box 2 (SOX2) in PE‐treated glioma cell specimens. β‐Actin was used as a loading control in western blot analysis. Bars represent the mean values ± SD (B, C). **p < 0.01, ***p < 0.001.

2.3. Cell viability assay

Cell viability was assessed using an alamarBlue assay (Bio‐Rad Laboratories). Briefly, the GIC spheres were dissociated into single cells using StemPro Accutase (Thermo Fisher Scientific). The cells were then seeded into a 96‐well Costar ultralow attachment flat‐bottomed plate (Corning) at a density of 3 × 10³ cells per 200 μL. Glioma cell lines were dissociated into single cells using 0.05% trypsin–EDTA (Thermo Fisher Scientific). These cells were then seeded into a 96‐well flat‐bottomed tissue culture plate (Falcon) at a density of 5 × 10² cells (dKGS01) or 1 × 10³ cells (dKGS07, A172, T98G, and U87) per 200 μL. All cells were treated with PE (1 or 5 μM) (Sigma‐Aldrich), and DMSO (Sigma‐Aldrich) was used as the control. The relative number of viable cells was determined by measuring the absorbance using a microplate reader (Bio‐Rad Laboratories). The average optical density values from the three wells in each group were calculated and plotted.

2.4. Tumor sphere‐forming assay

First, we dissociated the spheres of GICs (KGS01 and KGS07) into single cells using StemPro Accutase (Thermo Fisher Scientific). The cells were then seeded into a 96‐well Costar ultralow attachment round‐bottomed plate (Corning) at a density of 3 × 10³ cells per 200 μL. The GICs were treated with PE (1 or 5 μM), and DMSO was used as a control. Ten days after incubation in the neurosphere medium with 1% methylcellulose, images of these spheres were acquired. The average number of spheres with diameters greater than 100 μm from the three wells in each group was calculated and plotted.

2.5. Western blot analysis

Following the treatment of cells with the indicated reagents, protein samples were extracted using ice‐cold lysis buffer (Sigma‐Aldrich) containing protease inhibitors (Sigma‐Aldrich) and phosphatase inhibitors (Sigma‐Aldrich). Extracted proteins were analyzed by western blotting to determine the expression levels of the proteins of interest with the corresponding specific Abs (Table S1), as previously described. β‐Actin was used as a loading control. Three independent immunoblotting analyses were carried out. Immunoblot signals were measured using a CS analyzer (version 2.0; ATTO), and the average of band intensity was calculated.

2.6. Cell morphometry

Five different glioma cell lines (A172, T98G, U87, dKGS01, and dKGS07) were seeded into a 6‐well flat‐bottomed tissue culture plate (Falcon) at a density of 5 × 10⁴ cells per 1000 μL. All cells were treated with PE (5 μM), and DMSO was used as a control. After incubation for 24 h, all the culture plates were stained with Alexa Fluor 568 phalloidin (1:1000 dilution: Thermo Fisher Scientific) to evaluate morphological changes induced by PE treatment. Cells were stained for phalloidin and phosphorylated paxillin at tyrosine 118 (pPaxillinY118) and identified double‐stained edge as pseudopodia, which associates with cell mobility, according to previous reports. ²⁷ The average number of double‐stained pseudopodia per cell for 30 independent cells in each group was manually counted and plotted.

2.7. Cell migration assay

To investigate the effect of PE on the migration of glioma cells, we undertook a scratch assay. First, glioma cells (A172, T98G, U87, dKGS01, and dKGS07) were seeded into a 24‐well flat‐bottomed tissue culture plate (Falcon) at a density of 1 × 10⁵ cells per 500 μL. All cells were treated with PE (5 μM), and DMSO was used as a control. After incubating them for 24 h, we checked whether all culture plates were confluent and scratched the center of all plates with a 200 μL pipette to create a cell‐free zone (wound). Thereafter, all culture plates were washed twice using PBS and the cells were treated with PE (5 μM) with DMSO as a control. Finally, 12 h after scratching the culture plates, we measured the distance of the wounds. The average wound distance was calculated from the five micrographs in each group and plotted graphically.

2.8. Flow cytometry analysis

Cellular DNA content was analyzed using a FACSCanto II (Becton, Dickinson and Company) with FlowJo software (Becton, Dickinson and Company).

2.9. Cell apoptosis assay

Apoptosis in glioma cell lines was assessed using fluorescent immunostaining. First, glioma cells (A172, T98G, U87, dKGS01, and dKGS07) were seeded into a 6‐well flat‐bottomed tissue culture plate (Falcon) at a density of 1 × 10⁵ cells per 1000 μL. Each cell line was treated with PE (5 μM) or DMSO, used as a control. After incubating them for 48 h, all culture plates were washed twice using PBS. Cells were stained with Hoechst 33,258 (1:500 dilution; Dojindo) and PI (1:500 dilution; Dojindo) for 30 min in room air. The average number of cells undergoing apoptosis from five HPFs in each group was calculated and plotted. To investigate the involvement of caspase 9 in apoptosis induced by PE, we carried out a cell apoptosis assay with pretreatment of irreversible caspase 9 inhibitor, Z‐LEHD‐FMK (MBL), as previously described. ²⁸ , ²⁹ , ³⁰ Z‐LEHD‐FMK was used at a final concentration of 20 μM and was added 2 h prior to PE (5 μM) treatment.

2.10. Target silencing by siRNA

Two different sequences of siRNAs for STAT3 and negative control (Qiagen) were used. The target sequence‐1 of human STAT3 (siSTAT3‐1) (SI00048377) was 5′‐CTGGTCTTAACTCTGATTGTA‐3′; the target sequence‐2 of human STAT3 (siSTAT3‐2) (SI02662338) was 5′‐CAGCCTCTCTGCGCAGAATTCAA‐3′. First, all cells were seeded into a 6‐well flat‐bottomed tissue culture plate at the density of 5 × 10⁴ cells per 2 mL DMEM for 24 h. Subsequently, STAT3 siRNAs and control siRNA were transfected into A172, T98G, U87, dKGS01, and dKGS07 using Lipofectamine 2000 reagent (Invitrogen), according to the manufacturer's protocol. Cells were incubated for 72 h after transfection and used for the experiments. The downregulation of STAT3 was confirmed by western blotting.

2.11. Xenograft glioma model treatment

2.11.1. Xenograft mouse model of GIC

Based on previous studies, ¹⁴ we generated a xenograft mouse brain tumor model of human GBM by intracranial implantation of KGS07 cells into nude mice (BALB/scSlc‐nu/nu; Sankyo Labo Service Corporation). After appropriate anesthetic treatment, a burr hole was perforated in the skull 2–3 mm right lateral to the bregma, using a drill. Next, 1 × 10⁵ KGS07 cells were stereotactically injected at depth of 3 mm below the dura mater. The mice were randomly assigned to two groups and treated with PE (n = 5) or DMSO (n = 5). The PE‐treated group was given an intramuscular injection of 0.15 mg PE dissolved in 100 μL PBS with 5% DMSO every 2 days. The control group was treated with DMSO and given an intramuscular injection of 100 μL PBS with 5% DMSO every 2 days. After 8 weeks of treatment, all 10 mice were killed. The treatment method and optimum drug dose for xenograft mouse models were determined based on previous reports. ²¹ , ²² The brain and other four organs (kidney, liver, lung, and spleen) in the xenograft mouse in each treatment group were extracted.

2.11.2. Analysis of xenograft mouse model specimens

The extracted mouse tissues were dissected, embedded in paraffin, and cut into 4 μm serial coronal sections. The sectioned tissues were stained using the standard technique for H&E and TUNEL staining. Slides were also immunostained for nestin, Ki‐67, STAT3, and pSTAT3^Y705 (Table S1). Immunohistochemistry (IHC) was performed as previously described. ³¹ Positive and negative controls were used as the markers. To analyze the size of brain tumor in the xenograft mouse model, we calculated the average size of brain tumor in three coronal sections (anterior, middle, and posterior sections) per mouse. The size of the brain tumor tissue was measured using ImageJ (NIH). Three HPFs of brain tumor sections were randomly selected per mouse. The percentage of Ki‐67 labeling index, number of migrating cells, number of pSTAT3^Y705 spots, and number of apoptotic cells, as evaluated by TUNEL staining, were calculated manually. Every examination was evaluated from five HPFs in each group of specimens. All images were acquired using a BZ‐X700 microscope (Keyence).

2.12. Statistical analysis

All data are presented as the mean ± SD. Significant differences between the means of two groups were assessed using Student's t‐test. All statistical analyses were undertaken using Prism 8 software (GraphPad Software).

3. RESULTS

3.1. Drug screening to identify candidate agents against GICs

Drug screening was undertaken as previously reported. ⁹ , ¹⁴ , ¹⁵ After the first screening of 1301 compounds, 89 compounds showed inhibition of the two different GICs in a dose‐dependent manner. Fourteen candidates were selected after the second screening. Finally, PE was extracted as a candidate agent for glioma treatment (Figure 1A).

3.2. Attenuation of stemness and proliferation in GICs and its molecular signal changing by PE treatment

The proliferation assay revealed that PE treatment inhibited the viability of two GICs, KGS01 and KGS07 (Figure 1B). Compared with the control, 41% and 48% of the proliferation activities in KGS01 were inhibited at 1 and 5 μM PE, respectively. KGS07 also inhibited 77% and 82% of proliferation activity at each concentration of PE. There was no statistically significant difference in the inhibitory effects of 1 and 5 μM PE in either GIC. Additionally, PE treatment significantly reduced the number of spheres in GICs (Figure 1C). Compared to the control, 67% and 89% of the number of spheres in KGS01 were reduced at treatment with 1 and 5 μM PE, respectively. The number of spheres in KGS07 were also reduced by 78% and 96%, at 1 and 5 μM of PE, respectively. Western blot analysis showed that the expression of SOX2, which is known to play an important role in maintaining stem cell properties, was suppressed in GICs by PE in a dose‐dependent manner (Figure 1D).

3.3. Effect of PE on morphology and migration in glioma cell lines

Following the study of GICs, further analysis was carried out using glioma cells and differentiated GICs. The differentiation of GICs was confirmed by suppression of SOX2 expression (Figure S1). To evaluate the morphological changes induced by PE treatment of glioma cells, we focused on pseudopodia formation (Figure S2A). Fluorescent immunostaining revealed the localization of phosphorylated paxillin at tyrosine 118 (pPaxillin^Y118) (green) and phalloidin (red). The pseudopodia, in which paxillin is localized (yellow), are associated with glioma migration (Figure 2A). ²⁷ Treatment with PE suppressed the number of double stained pseudopodia in the glioma cells (Figures 2A,B and S2B). Consistent with these morphological changes, the results of the scratch assay indicated that the migration ability of glioma cells was suppressed by PE treatment (Figures 2C and S2C).

FIGURE 2.

Pentamidine (PE) treatment changed morphology and suppressed migration of glioma cells. (A) Images of A172 glioma cells using fluorescent immunostaining with phosphorylated paxillin at tyrosine 118 (pPaxillin^Y118) (green), phalloidin (red), and DAPI (blue) after 24 h of treatment. Enlarged images show the edge of the double stained pseudopodia (yellow). Scale bar, 20 μm. (B) Bar graphs indicate the average number of double‐stained pseudopodia in five different glioma cell lines. (C) Line graphs indicate trends of the scratch assay. The vertical axis indicates the average distance of the wound and the horizontal axis indicates time. Bars represent the mean values ± SD. *p < 0.05, **p < 0.01, ***p < 0.001 (B, C).

3.4. Effect of PE on proliferation of glioma cell lines and their cell cycle

The results of the proliferation assay showed the same trends as those of a previous study using GICs. Treatment with PE inhibited the viability of all glioma cell lines in a dose‐dependent manner (Figure 3A). Compared to the control, dKGS01, dKGS07, A172, T98G, and U87 showed 23%, 18%, 32%, 18%, and 15% inhibition of proliferation activities at 1 μM of PE, respectively. A comparison of the proliferation assay of glioma cell lines revealed that the PE treatment potential was higher for GICs than for differentiated GICs. Interestingly, the proliferation of differentiated GICs following PE treatment was similar to that of the glioma cell lines (Figures 1A and 3A). The percentage of cell population, analyzed by flow cytometry, revealed that a number of cells in the PE treated group were arrested in the G₁ phase (Figure 3B). Western blot analysis indicated that almost all PE‐treated glioma specimens displayed suppression of cyclin‐dependent kinase 4, which subserves the cell cycle during the G₁‐to‐S phase transition and proliferation ³² (Figure 3C).

FIGURE 3.

Analysis of the effect of pentamidine (PE) on the cell cycle of glioma cells. (A) Two differentiated glioma‐initiating cell lines (differentiated KGS01 [dKGS01] and differentiated KGS07 [dKGS07]) and three glioma cell lines (A172, T98G, and U87) were treated with PE at 0, 1, and 5 μM, and their viability curves were analyzed. (B) Cell cycle analysis in each treatment group using flow cytometry. (C) Western blot analysis shows the expression of cyclin‐dependent kinase 4 (CDK4). β‐Actin was used as a loading control. Bars represent mean values ± SD. *p < 0.05, **p < 0.01, ***p < 0.001 (A, B).

3.5. Effect of PE on apoptosis and its molecular signal alteration

Next, the ability of PE to induce apoptosis was investigated using fluorescent immunostaining. Normal nuclei of viable cells were stained with Hoechst 33258 (blue), and only the cytoplasm was stained with PI (red). In contrast, the nuclei of apoptotic cells were aggregated and double‐stained with Hoechst 33258 and PI (pink) (Figure 4A). ³³ , ³⁴ , ³⁵ , ³⁶ Fluorescent immunostaining of glioma cell lines revealed that the number of apoptotic cells increased in the PE‐treated group. In addition, pretreatment of Z‐LEHD‐FMK abrogated the increase of apoptotic cells by PE treatment (Figures 4B and S3). The induction of apoptosis in glioma cells was assessed by western blot analysis. The results showed that all treated cell lines displayed higher expression of cleaved‐caspase 9, which is a known marker of apoptosis. ³⁷ , ³⁸ , ³⁹ The increase of cleaved‐caspase 9 by PE treatment was canceled by Z‐LEHD‐FMK pretreatment (Figure 4C), which is consistent with the results of apoptosis assay.

FIGURE 4.

Pentamidine (PE) treatment induces apoptosis in glioma cell lines. (A) Images of nuclear staining of glioma cell line (A172) with Hoechst 33258 (blue) and propidium iodide (PI) (red) after 48 h of DMSO (control) or PE treatment. The enlarged image shows nuclear staining of PI and Hoechst 33258 with chromatin condensation in cells (pink). Scale bar, 100 μm. (B) Cell counting and (C) western blot analysis. Left column, DMSO (control); middle column, PE treatment; and right column, Z‐LEHD‐FMK, caspase 9 inhibitor (C9I) prior to PE treatment (PE + C9I). Bar graph indicates the average number of double‐stained cells per high‐power field. Bars represented mean values ± SD. *p < 0.05, **p < 0.01, ***p < 0.001. Western blot analysis shows the expression of cleaved‐caspase 9. β‐Actin was used as a loading control.

3.6. Main effect of PE treatment for glioma cell lines is suppressing activation of STAT3

To analyze the signal transduction pathways in GBM, western blotting was carried out. Previous studies have investigated the activation of Akt, ERK, and STAT3, which contribute to glioma cell proliferation. ⁹ , ²⁵ The results revealed that PE suppressed the expression of pSTAT3^Y705 in GICs and GBM cell lines (Figure 5A). Production and activation of Akt were suppressed in GICs, but not in differentiated GICs or GBM cell lines. Activation of ERK was sustained in all glioma cell lines used (Figures 5B and S4). Dual suppression of STAT3 and Akt could contribute to stronger inhibition of proliferation in GICs.

FIGURE 5.

Suppression of signal transducer and activator of transcription 3 (STAT3) activation is observed in glioma cell lines by pentamidine (PE) treatment. (A) Expression levels of phosphorylated signal transducer and activator of transcription 3 at tyrosine 705 (pSTAT3^Y705) and total STAT3 in glioma‐initiating cells (KGS01 and KGS07). These differentiated lines (dKGS01 and dKGS07) and three glioblastoma cell lines (A172, T98G, and U87) treated by PE were analyzed using western blotting. (B) Expression of phosphorylated and total Akt and ERK in all cell lines. β‐Actin was used as a loading control.

3.7. Treatment with PE suppresses brain tumor growth in xenograft mouse model

To validate these in vitro findings in vivo, we injected the human patient‐derived GBM cell line KGS07 (1 × 10⁵ cells) into mouse brains and treated them with PE or control (DMSO). The tumor size in the control group tended to be larger than that in the PE treatment group with statistical significance (Figure 6A). We investigated the differences in the migration ability in each treatment group. Slide sections stained with nestin showed that the number of migrated glioma cells treated with PE was higher than that of the control cells (Figure 6B). We also investigated the proliferative ability of the brain tumor, and slide sections from each group were stained with Ki‐67. The results revealed that the Ki‐67 labeling index in the PE treatment group was significantly lower than that in the control group (Figure 6C). For the evaluation of apoptotic cells, the tumor tissues were subjected to TUNEL staining. The results revealed that the number of the PE treated group tumor cells in the PE‐treated group, which was stained black (red dotted circle), was significantly higher than that in the control group (Figure 6D). To assess molecular signal changes following PE treatment, the activation levels of STAT3 in each treatment group were investigated. The pSTAT3^Y705 was spotty positive, and the result showed that the number of pSTAT3^Y705 positive spots in the PE treatment group was significantly higher than that in the control group (Figure 6E).

FIGURE 6.

Effect of pentamidine (PE) treatment on the xenograft mouse model and the putative mechanism on antiglioma effect of PE. (A) H&E staining (right column) and immunohistochemistry of nestin (left column) in serial coronal sections of xenograft mouse brain. Bar graph indicates the tumor size of the mice in each treatment group. (B) Immunohistochemistry for nestin shows a border between the tumor and adjacent normal brain tissue. Bar graph indicates the cells that migrated beyond the boundary in each treatment group. (C) Macrographs showing immunostaining for Ki‐67. Bar graph indicates the immunoactivity of Ki‐67 in each treatment group. (D) Micrographs of TUNEL staining reveals apoptotic cells. Inset shows an enlarged image of cells undergoing apoptosis, stained black. Red dotted circle indicates the cells in the apoptotic state. Bar graph indicates the number of apoptotic cells in each treatment group. (E) Macrographs showing immunostaining for phosphorylated signal transducer and activator of transcription 3 at tyrosine 705 (pSTAT3^Y705). Bar graph indicates the number of pSTAT3^Y705 positive spots in each treatment group. (F) Schematic representation of PE effect in glioma‐initiating cells (GICs) and glioma cells. Bars represent the mean values ± SD (A–E). Scale bar, 100 μm (B–E). *p < 0.05, **p < 0.01. CDK4, cyclin‐dependent kinase 4; SOX2, sex‐determining region Y box 2.

4. DISCUSSION

We have identified PE as a therapeutic drug for GBM for the first time. Pentamidine suppresses migration and proliferation through G₁ arrest and induction of apoptosis. It should be emphasized that all five glioma cell lines showed the same trend in vitro. Inhibition of STAT3 was observed in all glioma cell lines and GIC lines by western blotting. We also undertook an in vivo study by generating a xenograft mouse model of human GBM, and IHC staining of specimens reproduced these findings in the in vitro study. Interestingly, our results indicated that PE has a strong antitumor effect in GICs. ¹⁶ Additionally, we successfully generated differentiated GICs and compared the influence of PE on intracellular signaling between GICs and differentiated GICs.

In GICs, PE inhibits the phosphorylation of both pSTAT3^Y705 and pAkt^S473. Each molecular signaling pathway is independent and contributes to cell proliferation, ⁴⁰ , ⁴¹ , ⁴² suggesting that dual inhibition causes a powerful deterioration of proliferation in GICs. Previous studies have reported an immense effect of STAT3 on malignant tumors, including GBM. High expression of pSTAT3^Y705 in tumor specimens is correlated with poor prognosis in patients with GBM. JAK/STAT3/SOX2 signaling contributes to the enhancement of cancer‐initiating cell properties. ⁴³ , ⁴⁴ STAT3 contributes to cell proliferation in the G₁ phase of the cell cycle. ⁴⁵ Previous studies have also indicated that inhibition of STAT3 contributes to the expression of cleaved‐caspase 9 and induced apoptosis. ⁴⁶ Additionally, STAT3 is associated with the activation of paxillin, which is expressed in pseudopodia that facilitate cell migration. ⁴⁷ , ⁴⁸ , ⁴⁹ We revealed the possible involvement of STAT3 in glioma cell migration activity through pseudopodia. Therefore, the inhibition of STAT3 could suppress stemness, proliferation, and migration, and induce apoptosis in glioma cells. ³³ , ⁴⁰ , ⁴¹ , ⁴⁹ To evaluate the importance of STAT3 inhibition by PE in glioma cell lines, we carried out a proliferation assay using two different siSTAT3s (Figure S5A). The results showed that PE treatment had a potential additive effect on glioma viability in combination with selective STAT3 inhibition by siSTAT3s (Figure S5B). Although the major mechanism of PE in glioma appears to be the inhibition of STAT3 activation, other target molecules might be involved in PE treatment (Figure 6F). Analysis of detailed molecular machinery of PE in glioma cells is ongoing in our laboratory.

Although the optimal concentration of PE for patients with GBM was not assessed in this study, some clinical data were available from previous studies. The PE concentration in the choroid plexuses of patients with human African trypanosomiasis was much higher than that verified in the present study. ⁵⁰ This result suggests that the concentration of PE at this time would be sufficient to treat patients with GBM. The permeability of PE to the human BBB in GBM patients is unclear, but there is a possibility that the BBB of patients with GBM is damaged, and thus the drug concentration of the tumor in the cerebrospinal fluid might be similar to that in the blood. ¹⁸ Regarding safety of PE treatment, pathology of xenograft organs showed no obvious morphological changes after treatment (Figure S6A), suggesting that PE treatment causes no obvious adverse events. We also undertook proliferation assay, migration assay, and western blotting analysis using HFF‐1, a normal human fibroblast. No significant difference was observed in proliferation, migration, or phosphorylation of STAT3 by PE treatment (Figure S6B–D). Both in vitro and in vivo experiments suggested the safety of PE treatment. In conclusion, the present study indicates that PE is a high‐priority medicine that shows therapeutic effects on both glioma cells and GICs. Some GBM patients already receive PE as prophylaxis for pneumocystis pneumonia, and further clinical analysis would lead to successful GBM therapy in the future.

AUTHOR CONTRIBUTIONS

S. Tamai and MN designed the experiments. S. Tamai performed the experiments. TI, SJ, NH, HS, S. Tanaka, M. Kinoshita, and M. Kobayashi contributed to providing technical advice. S. Tamai wrote the manuscript. MN supervised the manuscript. All authors have read, revised, and agreed the final version of the manuscript.

FUNDING INFORMANTION

This work was supported by JSPS KAKENHI (20 K17956 to ST), Kobayashi International Scholarship Foundation (to MN), the Extramural Collaborative Research Grant of the Cancer Research Institute, Kanazawa University (to MN and AH), and an intramural clinical research grant from Kanazawa University (to MN).

CONFLICT OF INTEREST STATEMENT

Atsushi Hirao and Mitsutoshi Nakada are editorial board members of Cancer Science. The other authors have no conflict of interest.

ETHICS STATEMENTS

Approval of the research protocol by an institutional review board: The experimental protocol was developed in strict accordance with the Ethical Guidelines for Medical and Health Research Involving Human Subjects and was approved by the Ethics Committees of Kanazawa University (approval numbers: 209, 2080, and 3581).

Informed consent: All methods were performed in accordance with the relevant guidelines and regulations. Written informed consent was obtained from all participants or their representatives related to this research.

Animal studies: The animal experimental protocol of this study was approved by the appropriate research ethics committee (approval number: AP‐204197). All animal experiments were performed in accordance with the relevant guidelines and regulations.

Registry and registration no. of the study/trial: N/A.

Supporting information

Figure S1.

Figure S2.

Figure S3.

Figure S4.

Figure S5.

Figure S6.

Table S1.

Tamai S, Ichinose T, Jiapaer S, et al. Therapeutic potential of pentamidine for glioma‐initiating cells and glioma cells through multimodal antitumor effects. Cancer Sci. 2023;114:2920‐2930. doi: 10.1111/cas.15827