Dual Inhibition of NOX2 and Receptor Tyrosine Kinase by BJ-1301 Enhances Anticancer Therapy Efficacy via Suppression of Autocrine-Stimulatory Factors in Lung Cancer

Lung cancer is the leading cause of cancer-associated deaths (1), and non–small cell lung cancer (NSCLC) constitutes over 80% of lung cancers. With advancements in the study of cancer biomarkers, NSCLC treatments have gone from platinum-based chemotherapy (associated with a low response rate) to targeted therapy using inhibitors of EGFR tyrosine kinase based on the identification of sensitizing mutations within EGFRs (2, 3). However, the number of cases suitable for such EGFR-targeted therapy is small (4), and the overall 5-year survival rate is still low (5).

NADPH oxidase (NOX) is a multicomponent enzyme that consists of plasma membrane-spanning catalytic (i.e., gp91phox and p22phox) and regulatory/cytosolic (i.e., p47phox, p67phox, and GTPase Rac1) components (6). Functionally distinct homologs of gp91phox (denoted NOX1 to NOX5, DUOX1, and DUOX2) are expressed in a tissue-specific manner (7); NOX2 is the major catalytic isoform in endothelial cells (8), and NOX4 is the major catalytic isoform in renal and vascular smooth muscle cells (9). Furthermore, cancer cells originating from different tissues overexpress different isoforms of the catalytic unit. NOX1 and DUOX2 are overexpressed in colorectal and pancreatic cancers, NOX4 is overexpressed in premalignant fibrotic lung and liver tissues, and NOX5 is overexpressed in melanoma and prostate cancers (10). Such hyperactivation or overexpression of NOX is implicated in the pathogenesis of cancers (10, 11). NOX2 activation has been shown to enhance angiogenesis, tumor growth, aggressiveness, and metastases (12). Furthermore, tumor promoter–induced isotype switching from NOX1 to NOX2 is associated with malignant progression of colon cancer (13).

Activation of NOX is coupled with stimulation of various cell membrane receptors as growth factors (14, 15). Upon binding to growth factors, receptor tyrosine kinase (RTK) phosphorylates tyrosine residues in its intracellular domain, which then activates downstream signaling molecules such as PI3K/protein kinase B (Akt) and MAPKs. This leads to phosphorylation of NOX cytosolic components (i.e., p47phox and p67phox) and triggers translocation of these subunits along with Rac1 to the cell membrane (16), activating NOX and subsequently generating superoxide anion. Superoxide anion is dismutated to hydrogen peroxide and subsequently converted into other reactive oxygen species (ROS). ROS, in turn, enhances the activity of the redox-dependent transcription factor, NF-κB (16), upregulating its target genes, including NOX2 and growth factors (13, 17). Following the initial activation of RTK, the function of growth factors and NOX is reinforced by a positive feedback mechanism. Inhibition of NOX by pharmacologic inhibition or knockdown using siRNAs leads to decreased lung cancer cell invasion and metastases (18, 19), suggesting that NOX is a promising target for novel anticancer drug development (16, 20).

In agreement with findings that ROS promote angiogenesis and tumor growth (21, 22), recent studies have shown that disruption of redox-sensitive signaling pathways by free-radical scavengers, such as vitamin E (23), suppresses lymphocyte- and VEGF-induced angiogenesis (24, 25). Among members of the vitamin E family, tocotrienols, which possess better antioxidative activity than tocopherols (26, 27), have been shown to exert cancer preventive and antiangiogenic activities (28, 29). However, the use of tocopherols and tocotrienols at high dosages is potentially associated with an increased incidence of heart failure and mortality (30, 31), necessitating the development of potent agents that inhibit the redox-sensitive signaling pathways.

In this study, the mode of anticancer activity was characterized for a synthetic analogue of α-tocopherol (α-TOH), BJ-1301. BJ-1301 (Fig. 1A) is an aminopyridinol derivative of α-TOH that was previously shown to exhibit strong antioxidant activity (32). Our results show that the anticancer activity of BJ-1301 is mediated through inhibition of RTK signaling in association with NOX2.

The endothelial growth medium (EGM)-2 bullet kit was purchased from Cambrex. Matrigel was obtained from BD Biosciences. α-TOH was purchased from Sigma-Aldrich. Sunitinib malate was obtained from Selleckchem (Selleckchem). BJ-1301 (Fig. 1A) was prepared using a previously reported synthetic method (32) in the laboratories of Prof. Byeong-Seon Jeong (College of Pharmacy, Yeungnam University, Republic of Korea) and Dr. Jin-Mo Ku (Bio-Center, Gyeonggi Institute of Science and Technology Promotion, Republic of Korea). Cisplatin was obtained from Shandong Boyuan Pharmaceutical Co. Ltd.. SR11302 and SU4312 were obtained from Tocris Bioscience (Tocris House). Apocynin, VAS2870, SB203580, PD98059, BAY-117085, and pyrrolidine dithiocarbamate (PDTC) were purchased from Sigma Aldrich. SP600125 was received from Calbiochem. Antibodies directed against phospho-p38, p38, phospho-ERK, ERK, phospho-JNK, JNK, p65NF-κB, phospho-p85-PI3K (at Tyr488), p85-PI3K, phospho-AKT (at T308), AKT, and p-IκB were purchased from Cell Signaling Technology Inc.. NOX-2 mAbs, as well as phospho-p47phox, stem cell factor (SCF), TGF-α, NOX-1, ras, lamin B, β-actin, and NOX-4 antibodies were obtained from Abcam. Antibodies against IκB, p47phox, Rac, and p67phox were purchased from Santa Cruz Biotechnology.

Human umbilical vein endothelial cells (HUVEC) were purchased from Lonza in 2010. Bovine Aortic Endothelial Cells (BAOEC) were received as a kind gift from Dr. You Mie Lee (Kyungpook National University, Daegu, Republic of Korea) in 2017. The human lung cancer cell line, A549, was obtained from ATCC in 2008 and H1299 was purchased from Korean cell line bank (Seoul, Korea) in 2009. BAOECs were grown in DMEM obtained from Hyclone supplemented with 10% FBS and 1% penicillin/streptomycin. HUVECs were grown on a 0.2% gelatin-coated flask and supplemented with endothelial cell growth supplement (EGM-2 SingleQuots). HUVECs between passage number 2 and 6 were used in our experiments. A549 and H1299 human lung cancer cells were grown in DMEM and RPMI1640, respectively, supplemented with 10% FBS, 100 IU/mL penicillin, and 100 μg/mL streptomycin in a humidified incubator at 5% CO₂/95% air.

Cells were seeded in 96-well plates and serum-starved for 24 hours, followed by treatment with VEGF (20 ng/mL) or FBS (10%) in the absence or presence of test compounds for 72 hours. After incubation, the number of viable cells was measured using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Color intensities were measured using a microplate reader (Versamax, Molecular Devices, Inc.).

A kinase insert domain receptor (KDR; Promega #V2681), platelet-derived growth factor receptor beta (PDGFRβ; Promega #V3731), and c-KIT (Promega #V4498) kinase enzyme systems, along with the adenosine diphosphate (ADP)-glow kinase assay kit (Promega #V9101) were used to perform kinase assays following the manufacturer's instructions. Briefly, the reactions were started after adding 50 μmol/L adenosine triphosphate (ATP, final concentration) to a 0.2 μg/μL mixture of poly (Glu4, Tyr1), test drug, and 1.5 ng KDR, 10 ng PDGFRβ, or 20 ng c-KIT enzyme. The reaction was carried out at room temperature for a specified period of time. After adding detection reagents, luminescence was measured using a Fluostar Optima microplate reader (BMG Labtech GmbH).

Cytoplasmic and nuclear proteins were extracted from cells using a NE-PER kit (Pierce-Thermo). To determine total protein, cells and tissues were lysed using radioimmunoprecipitation buffer containing a 1× protease and phosphatase inhibitor cocktail. The lysates were centrifuged at 12,000 rpm for 10 minutes and the supernatant portion was collected. HUVECs were stimulated with VEGF for 5 minutes following pretreatment with the compounds for 1 hour; 100 μL of subcellular fractionation (SF) buffer was added to the cell pellet. The cell suspension was passed through a 30 G needle 35 times and centrifuged at 700 × g for 10 minutes at 4°C. The supernatant portion was centrifuged at 100,000 rpm (Sorvall RC M120 EX ultracentrifuge) for 1 hour at 4°C. Supernatant containing the cytosol fraction was collected and the pellet was resuspended in 60 μL SF buffer. The suspension was sonicated for 30 seconds four times. After centrifuging samples at 100,000 rpm for 1 hour at 4°C, supernatants containing the membrane fraction were saved.

Protein concentrations were determined using a BCA protein reagent kit (Pierce-Thermo). Equal amounts of protein were loaded and separated using SDS-PAGE. After incubation with respective antibodies, proteins were detected using a LAS-4000 mini imager (Fujifilm).

Angiogenesis was examined using previously published methods (33, 34). VEGF (20 ng/CAM) was dissolved in PBS containing 0.1% BSA and added topically to the sterile disk on top of 9-day-old chick embryo CAMs. α-TOH, BJ-1301, or vehicle was then applied topically to the chick chorioallantoic membranes (CAM). The CAM tissue directly beneath the disk was resected from the embryo after 72 hours and harvested under light microscopy (Leica). The number of new blood vessel branch points contained in a circular region equal to the area of the filter disk was then counted for each section. The resulting angiogenesis index is expressed as the mean ± SEM of the new branch points for each set of samples. Inhibition ratio (%) of compounds was calculated according to the following equation:

where [A] and [B] represent angiogenesis after a 3-day incubation with VEGF in the presence or absence of test compounds, respectively.

In the tumor angiogenesis experiments, all procedures were the same as described above, except that H1299 or A549 human lung cancer cells (2 × 10⁶ cells/CAM) were inoculated onto the CAM instead of VEGF (35). The number of vessel branch points contained in a tumor region was counted by two double-blinded observers. The tumor tissue growing on top of the CAM was isolated from the membrane and weighed.

Intracellular ROS generation was measured using the dichlorofluorescin diacetate (DCF-DA) fluorescent dye. Serum-starved cells were cotreated with VEGF in the presence or absence of compounds. After incubation for 5 minutes, DCF-DA (5 μmol/L) was added to the cells at 37°C, and the cells were imaged using an inverted fluorescent microscope (TE2000-U; Nikon).

Superoxide anions were measured as previously published with slight modifications (36). Briefly, A549 and H1299 (1 × 10⁵ cells/well) cells and HUVECs (0.5 × 10⁵ cells/well) were seeded in white opaque 96-well plates. On the following day, cells were pretreated with drugs for 1 hour and then a stimulator for 3 hours. Chemiluminescence was measured using lucigenin (400 μmol/L) in a Fluostar Optima microplate reader.

HUVECs were seeded on a confocal dish at a density of 5 × 10⁴ cells/cm². Serum-starved cells were pretreated with test compounds for 1 hour and stimulated with VEGF or 12-O-tetradecanoylphorbol-13-acetate (TPA) for 2 minutes. Cells were fixed with paraformaldehyde (3.7%) in PBS, permeabilized with Triton X-100 (0.1%) for 10 minutes, and blocked with BSA (3%) for 1 hour. Samples were incubated overnight at 4°C with primary antibody, then with secondary antibody (IgG-FITC), and stained with 4′,6-diamidino-2-phenylindole (DAPI) for 10 minutes and photographed using a confocal A1 imaging system (400×, Nikon Corp.).

Total RNA from HUVECs was isolated using TRIzol reagent (Life Technologies Inc.). Quantitative analysis of mRNA was performed using the QuantiTect SYBR Green PCR kit (Qiagen). The primer sequences are specified in Supplementary Table S1.

Copy number of NOX1, NOX2, and NOX4 was determined according to the established protocols with some modifications (13). Briefly, human NOX1, NOX2, and NOX4 cDNAs (Invitrogen) were cloned into the pcDNA5/FRT/TO vector (Invitrogen). Copy number was calculated using the following equation:

Through serial dilution of NOX1, NOX2, and NOX4 plasmids (from 10¹ to 10⁵), standard curves of each were generated using Quantitect Probe PCR kit (Qiagen) and probe and primer mix for NOX1 (Hs01071088_m1), NOX2 (Hs00166163_m1), and NOX4 (Hs01379108_m1; Thermo Fisher Scientific Corporation). cDNA synthesized from the isolated RNA was subjected for qRT-PCR using specific probe and primer mix for NOX1, NOX2, and NOX4 (Thermo Fisher Scientific) on Corbett Rotor-Gene (Qiagen). Copy number of NOX1, NOX2, and NOX4 in the cells were determined from the standard curve.

A549 cells (1 × 10⁷ cells) were injected subcutaneously in the right flank of 6-week old BALB/C female nude mice purchased from Orient Co. Ltd.. Tumor volume was calculated according to the following equation:

When tumor volume reached 200 mm³, mice were divided into different groups and injected intraperitoneally with drugs once daily. At the end of treatment, mice were sacrificed and the excised tumors were photographed and weighed.

Animal experiments were conducted in accordance with the institutional guidelines of the Institute of Laboratory Animal Resources (1996) and of Yeungnam University for the Care and Use of Laboratory Animals (2009).

Data are presented as the means ± SEM. Statistical analysis was performed using one-way ANOVA followed by the Student–Newman–Keuls comparison (GraphPad Prism 5.0 software) for calculating differences between groups. P values <0.05 were considered statistically significant.

Considering the previously reported antiangiogenic activity of vitamin E, we first examined the inhibitory effects of BJ-1301 on angiogenesis in vitro in HUVECs. We included α-TOH as a reference compound. HUVECs were treated with BJ-1301 or α-TOH along with VEGF for 48 and 72 hours, and then cell proliferation was measured. VEGF-induced proliferation was significantly inhibited by BJ-1301 in a concentration-dependent manner (Fig. 1B), and the antiproliferative effect of BJ-1301 was greater than that of α-TOH (Fig. 1B). BJ-1301 alone without VEGF treatment did not induce viability decrease in endothelial cells (Supplementary Fig. S1A) or HUVEC proliferation (Supplementary Fig. S1B). To further assess inhibitory effect of BJ-1301 on VEGF-induced angiogenesis, tube formation assays were performed. VEGF-induced endothelial tube formation was significantly inhibited by BJ-1301 in a concentration- (Supplementary Fig. S1C, top) and time-dependent (Supplementary Fig. S1D) manner. Similarly, VEGF-induced endothelial invasion was also inhibited (Supplementary Fig. S1C, bottom). The inhibitory effects of BJ-1301 on tube formation and invasion were greater than those of α-TOH at a concentration of 1 μmol/L. Differences in the antiangiogenic effects of the two compounds correspond to their inhibitory activities against VEGFR2 tyrosine kinase, the receptor isoform mediating major growth effects of VEGF (37), as measured by a KDR kinase activity assay. KDR kinase activity was inhibited by BJ-1301 and α-TOH (maximally ∼80% and ∼30%, respectively; Fig. 1C). Furthermore, IC₅₀ values for KDR kinase activity were estimated to be 0.51 and >10 μmol/L for BJ-1301 and α-TOH, respectively (Fig. 1C; Supplementary Table S2), indicating greater potency of BJ-1301 as an inhibitor of VEGFR2 tyrosine kinase activity. Consistent with the inhibitory activities of BJ-1301 and α-TOH on VEGFR2 tyrosine kinase activity, BJ-1301 was more efficient at suppressing downstream signaling pathways of VEGFR2-induced MAPK phosphorylation, p38, ERK1/2, and JNK than that of α-TOH in HUVECs (Fig. 1D). Similarly, VEGF-induced cytosolic IκB phosphorylation and NF-κB nuclear translocation were significantly suppressed by BJ-1301, with stronger activity than that of α-TOH at a concentration of 1 μmol/L (Fig. 1E). As observed in HUVEC proliferation, BJ-1301 alone treatment had no effect on basal status of MAPK (Supplementary Fig. S2A), NF-κB nuclear translocation (Supplementary Fig. S2B), and ROS production in HUVECs (Supplementary Fig. S2C). The in vivo antiangiogenic activity of α-TOH and BJ-1301 was also compared using CAM assay. VEGF-induced blood vessel formation in the growing CAM was significantly suppressed by administration of α-TOH and BJ-1301 (Fig. 1F). The BJ-1301 ID₅₀ was 0.007 pmol/CAM (Supplementary Table S3), which was approximately 200-fold more potent than that of α-TOH (1.1 pmol/CAM).

Next, we examined whether BJ-1301 exerts inhibitory effects on other RTKs, such as PDGFRβ and c-KIT, similar to sunitinib that has been approved to treat several cancers. BJ-1301 significantly inhibited PDGFRβ (Fig. 2A) and c-KIT (Fig. 2B) tyrosine kinases with IC₅₀ values of 0.45 and 0.38 μmol/L, respectively (Supplementary Table S2). These data indicate that BJ-1301 is a less potent multi-RTK inhibitor than that of sunitinib (38). To determine whether the antiangiogenic activities of α-TOH and BJ-1301 result in inhibition of tumor growth similar to that of sunitinib, the compounds were tested in CAMs inoculated with H1299 human lung cancer cells. Implanted tumor-induced angiogenesis was significantly suppressed by α-TOH, sunitinib, and BJ-1301 in a dose-dependent manner; BJ-1301 showed the strongest inhibitory effects (Fig. 2C and D). The implanted tumor growth was suppressed by the compounds, with BJ-1301 being the most efficacious (Fig. 2D). Similar antiangiogenic and antitumor activity of BJ-1301 was also observed in CAMs inoculated with A549 human lung cancer cells (Fig. 2E and F).

Next, we examined whether the strong antiangiogenic and antitumor activities of BJ-1301 were associated with its strong antioxidant activity. Activation of growth factor receptor (e.g., VEGFR) upon binding of its ligand (e.g., VEGF) leads to NOX2 activation and subsequent ROS generation (39). To determine whether the antiangiogenic activity of BJ-1301 is associated with its antioxidant activity, ROS generated by VEGF were measured in HUVECs; α-TOH, sunitinib, SU4312 (a specific VEGFR2 tyrosine kinase inhibitor; ref. 40), apocynin (41), and VAS2870 (42) were included as controls. BJ-1301 inhibited VEGF-induced production of ROS in a concentration-dependent manner, and its potency was better than that of α-TOH, apocynin, or VAS2870 (Fig. 3A), but similar to that of sunitinib and SU4312 (IC₅₀ values of 0.151, 0.707, 0.102, 0.120, 21.37, and 5.75 μmol/L for BJ-1301, α-TOH, sunitinib, SU4312, apocynin, and VAS2870 respectively, Supplementary Table S4). We observed a linear dose–response curve for α-TOH, whereas BJ-1301 showed a sigmoidal dose–response curve, indicating that BJ-1301 suppresses ROS formation in a different manner from that of α-TOH. To determine whether the inhibitory action of BJ-1301 on ROS production is attributable to inhibition of NOX activity, we examined whether BJ-1301 inhibits NOX-associated ROS generation using chemical stimulators of NOX that do not activate VEGFR2. TPA activates NOX through protein kinase C, whereas geranylgeranyl pyrophosphate (GGPP) and mevalonate (Mev) directly activate NOX through posttranslational modification of Rac1 (43, 44). BJ-1301 inhibited TPA-induced superoxide production in a concentration-dependent manner and was more potent than α-TOH, sunitinib, SU4312, apocynin, or VAS2870 (Fig. 3C; Supplementary Table S4). Furthermore, BJ-1301 inhibited GGPP- or Mev-induced superoxide production (Fig. 3D and E) to a greater extent than that of α-TOH. To verify the direct inhibitory activity of NOX2 activation by BJ-1301, membrane translocation of cytosolic components was examined in HUVECs. In an immunocytochemistry assay, BJ-1301, but not α-TOH, significantly blocked VEGF- and TPA-induced Rac1 translocation to the plasma membrane (Fig. 3F and G), which was further confirmed by quantitating Rac1 level in the cytosol and plasma membrane (Fig. 3H). Western blot results also show that BJ-1301 inhibits VEGF-induced phosphorylation of p47phox and its translocation to the membrane in HUVECs in a concentration-dependent manner (Fig. 3H), whereas α-TOH had no effect on p47phox phosphorylation or translocation of p47phox and p67phox. Results from the immunocytochemistry assay using anti-p67phox antibodies also show that BJ-1301 inhibits VEGF- or TPA-induced translocation (Fig. 3I and J) of p67phox to the cell membrane in a concentration-dependent manner. Such inhibition was not observed for α-TOH.

Binding of growth factors to RTK triggers MAPK and NF-κB activation, leading to upregulation of multiple target gene expression including NOX2. VEGF enhanced the mRNA expression of NOX2, p47phox, and p67phox, with no changes in p22phox or p40phox expression in HUVECs, in a time- and concentration-dependent manner starting as early as 30 minutes after VEGF treatment (Fig. 4A). Also, VEGF-treated HUVECs induced 4-fold increase in ROS production within 5 minutes and subsequent modest increase reaching maximum level at 1 hour (Fig. 4B). VEGF-induced upregulation of NOX2, p47phox, and p67phox subunits was blocked by inhibitors of MAPK (SB203580, a p38 inhibitor (45); SP600125, a JNK inhibitor (46); and PD98059, an ERK inhibitor (47), NF-κB (PDTC) (48), and IκB kinase (BAY-117085; ref. 49), but not by the activator protein-1 (AP-1) inhibitor SR11302 (50; Fig. 4C). NOX isoforms were differentially expressed in HUVECs, as shown in copy number measurement of NOX1, NOX2, and NOX4 mRNAs (Fig. 4D). NOX4 is expressed more than 100 times than NOX2, whereas NOX1 expression level was very low in HUVECs. Treatment with VEGF enhanced NOX2, but not of NOX1 or NOX4 mRNA expressions, and BJ-1301 suppressed VEGF-induced NOX2 increase (Fig. 4D). Furthermore, BJ-1301 inhibited VEGF-induced upregulation of NOX2, p47phox, and p67phox mRNA (Fig. 4E) and protein (Fig. 4F) expressions in HUVECs, which is similar to that observed from NF-κB and IKK inhibitors. NOX2-selective action of VEGF and BJ-1301 was further confirmed by immunoblot assay including positive and negative controls (Supplementary Fig. S3).

To determine whether NOX2 inhibition by BJ-1301 in cancer cells leads to altered lung cancer cell proliferation, we examined A549 (wild-type p53 gene) and H1299 (p53 null) human NSCLC proliferation in the presence of BJ-1301. α-TOH (a compound that does not inhibit NOX activity), cisplatin (a component of the nontargeted NSCLC chemotherapy regimen), and sunitinib (a multi-RTK inhibitor and component of targeted NSCLC therapy) were included as controls. Both A549 (Fig. 5A) and H1299 (Fig. 5B) cells showed resistance to cisplatin (for 72 hours treatment, IC₅₀ = 11.22 μmol/L and 19.05 μmol/L for A549 and H1299 cells, respectively). α-TOH showed very weak inhibitory effects on cancer cell proliferation (IC₅₀ values of 74.1 μmol/L and 87.9 μmol/L for A549 and H1299 cells, respectively), while BJ-1301 showed IC₅₀ values of 0.58 and 0.87 μmol/L in A549 and H1299 cells, respectively. Sunitinib showed less potent IC₅₀ values (1.31 and 1.86 μmol/L in A549 and H1299 cells, respectively) than those of BJ-1301. Interestingly, BJ-1301 inhibited serum (1%)-, TPA- or VEGF-induced superoxide radical production in A549 and H1299 cancer cells to a greater extent than that of sunitinib (Fig. 5C). To determine whether the increased inhibition of ROS production by BJ-1301 (despite the much weaker RTK inhibitory activity than that of sunitinib) is connected to RTK signal transduction, we examined activation of downstream signaling molecules by phosphorylation in lung cancer cells. BJ-1301 inhibited phosphorylation of PI3K/Akt and Ras/ERK signaling pathways downstream of growth factor receptors in A549 and H1299 cells, and the effects were stronger than those of sunitinib (Fig. 5D). Consequently, downregulation of RTK downstream targeted gene expression, including TGF-α (an EFGR ligand) and SCF (a c-KIT ligand), by BJ-1301 was greater than that of sunitinib in lung cancer cells (Fig 5E). NOX isoforms were differentially expressed in two cancer cell lines; NOX1 and p67phox mRNA levels in A549 cells were higher than H1299, while NOX2, NOX4, and p47phox mRNA levels were higher in H1299 cells than in A549 cells (Supplementary Fig. S4; Supplementary Table S5). As in HUVECs, BJ-1301 treatment downregulated expression of the major catalytic subunit NOX2 (but not NOX1 or NOX4) in both A549 and H1299 cells to a greater extent than that of sunitinib (Fig. 5E).

To confirm the anticancer activity of BJ-1113 in vivo, we examined its antitumor effects in mice xenografted with A549 cells and compared them with the effects of α-TOH, sunitinib, or cisplatin. On the basis of preliminary results showing that cisplatin (3 mg/kg) treatment is highly toxic in mice, a lower dose (i.e., 1 mg/kg) was used. BJ-1301 dose-dependently suppressed tumor growth to a greater extent than that of cisplatin (Fig. 6A). Furthermore, BJ-1301 at a dose of 5 mg/kg inhibited tumor growth in mice to a greater extent than that of sunitinib at a dose of 40 mg/kg, with tumor volumes being 80.6% and tumor weights being 88.9% lower in the BJ-1301–treated group (mean tumor volume 603 ± 38 mm³ and tumor weight 0.70 ± 0.15 mg) than those of the vehicle-treated controls (mean tumor volume 3,110 ± 340 mm³ and tumor weight 6.28 ± 0.51 mg; Fig. 6A and B). While cisplatin-treated mice showed progressive losses in body weight (a marker of toxicity), body weights of BJ-1301–treated mice did not decrease (Fig. 6C), suggesting a lack of toxicity. BJ-1301 decreased the expression of TGF-α, SCF, and NOX2 in the tumor tissues in a dose-dependent manner (Fig. 6D).

Previous studies have shown that natural or synthetic antioxidants including vitamin E inhibit angiogenesis via inhibiting one of multiple downstream pathways triggered upon RTK activation (i.e., PI3K/PDK/Akt). In the current study, we report that BJ-1301 (an aminopyridinol derivative of α-TOH) inhibits the regulatory network composed of RTK and NADPH oxidase and thus effectively blocks overall RTK activation and angiogenesis, presenting a potent and efficacious therapeutic agent against lung cancer.

Our results show that BJ-1301 and α-TOH inhibit ROS generation triggered by various stimuli at different efficacy. BJ-1301 inhibited the rise in intracellular ROS level which triggers the activation of redox-sensitive MAPKs and NF-κB as well as upregulation of NOX2, p47phox, and p67phox expression in both endothelial cells and lung cancer cells, to a greater extent than α-TOH did. The higher efficacy of BJ-1301 can be attributed to different modes of action for inhibiting NADPH oxidase activation and subsequent ROS generation. Of note, NADPH oxidase activation requires membrane localization of cytosolic subunits through either phosphorylation of Rac and p47phox via VEGF- or TPA-induced phosphokinase C (PKC) activation (51) or prenylation of Rac by GGPP and Mev (52). α-TOH blocked ROS production induced by VEGF and TPA, but not GGPP or Mev, suggesting that α-TOH suppresses the activation of NADPH oxidase via inhibition of PKC. These results are consistent with those of previous studies in monocytes and microglia showing that α-TOH inhibits the assembly and activation of NADPH oxidase through inhibition of PKC (53, 54). In contrast, BJ-1301 inhibited ROS production triggered by any of the stimuli (VEGF, TPA, GGPP, and Mev), suggesting that BJ-1301 inhibits NADPH oxidase activation suppressing via both p47phox phosphorylation and Rac activation and translocation. Indeed, results from Western blots and confocal microscopy showed that BJ-1301 inhibits translocation of NADPH oxidase induced by VEGF and TPA. These effects were obvious starting at the concentration of 0.1 μmol/L, and the inhibitory effect reached maximum at the concentration of 1 μmol/L. Although our current study showed that α-TOH at concentration of 1 μmol/L failed to inhibit translocation of p67phox to cell membrane, it has been shown that α-TOH could also block p67phox translocation at higher concentration (100 μmol/L; ref. 55). This further demonstrates better efficacy of BJ-1301 than that of α-TOH.

Our current results clearly demonstrate that NOX2, but not NOX1 or NOX4, is the target gene of RTK and NOX activation in both HUVECs and lung cancer cells. This is in line with our previous study that the feed-forward effects of NOX2-derived ROS on induction of NOX2 and other target genes (e.g., matrix metalloproteinase-7) in colon cancer, leading to a highly metastatic phenotype (13). The current study also reveals that decreased NOX2 expression is associated with lower expression of TGF-α and SCF in lung cancers. TGF-α is a member of the EGF family. It is excreted to the microenvironment of cancer cells, binds to EGFR, and activates mitogenic signaling, serving as an autocrine factor for tumor growth. SCF, a ligand of c-KIT, is another autocrine factor that has a mitogenic and angiogenic activity (56). While c-KIT expression is normally restricted in primitive hematopoietic cells, endothelial cells, and interstitial cells of Cajal, many carcinoma cells, including lung cancer cells, express functional c-KIT (57, 58). STI-571 (Gleevec), a c-KIT inhibitor, suppresses growth of cisplatin-resistant A549 lung cancer cells, and combination treatment with STI-571 and cisplatin results in synergistic effects on cytotoxicity in cells (59). Together, these results indicate that the inhibition of these autocrine factors serves as an effective approach to suppress tumor growth and angiogenesis, by blocking the autoregulatory loop that is critical for lung tumor growth. Downregulation of NOX2 by BJ-1301 and subsequent suppression of TGF-α and SCF expression thus should serve as a promising therapeutic agent against lung cancer.

Our current results clearly demonstrate that BJ-1301 possesses inhibitory activity against RTKs (i.e., VEGFR2, PDGFRβ, and c-KIT). Notably, BJ-1301 exhibited better RTK inhibitory activity as compared with α-TOH, which has an IC₅₀ greater than 10 μmol/L). Interestingly, however, BJ-1301 inhibited proliferation, signaling molecule activation, and RTK target gene expression in cisplatin-resistant lung cancer cells to a much greater extent than that of sunitinib. These superior effects of BJ-1301 were also prominent in in vivo tumor models; BJ-1301 treatment (5 mg/kg) completely blocked NOX2, TGF-α, and SCF expression in tumor-bearing mice while the inhibition by sunitinib (at 40 mg/kg) was incomplete. The better efficacy of BJ-1301 is explained in part by its dual inhibition of RTK and NADPH oxidase. RTK and NADPH oxidase activate each other and synergistically enhance expression of autocrine factors for tumor growth (e.g., TGF-α and SCF). Dual inhibition of both RTK and NADPH oxidase thus inhibit tumor growth more effectively than the inhibitors of single pathway.

In conclusion, our current results demonstrate that BJ-1301 is a promising candidate for the management of tumor angiogenesis and growth via inhibition of RTK and NADPH oxidase activities.

No potential conflicts of interest were disclosed.

Conception and design: T.-G. Nam, B.-S. Jeong, J.-A. Kim

Development of methodology: B.-S. Jeong, J.-A. Kim

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J. Gautam, J.-M. Ku, S.C. Regmi, Y. Wang, S. Banskota, M.-H. Park

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J. Gautam, J.-M. Ku, S.C. Regmi, Y. Wang, S. Banskota, M.-H. Park, J.-A. Kim

Writing, review, and/or revision of the manuscript: J. Gautam, H. Jeong, B.-S. Jeong, J.-A. Kim

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J.-M. Ku, B.-S. Jeong

Study supervision: B.-S. Jeong, J.-A. Kim

This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Korean government (MSIP; NRF-2014R1A2A2A01006833 and NRF-2014R1A4A1071040; through J.A. Kim).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.