Identification of a Dual Inhibitor of Janus Kinase 2 (JAK2) and p70 Ribosomal S6 Kinase1 (S6K1) Pathways

Background: An underlying cause of cancer therapeutic resistance is the hyperactivation of endogenous overlapping or redundant signaling pathways in cancer cells.

Results: Gingerenone A selectively kills cancer cells through dual inhibition of JAK2 and S6K1.

Conclusion: Co-targeting JAK2 and S6K1 induces synergistic anti-cancer effects.

Significance: Investigating the targets of bioactive compounds can lead to suggestions for novel therapeutic strategies.

Abstract

Bioactive phytochemicals can suppress the growth of malignant cells, and investigation of the mechanisms responsible can assist in the identification of novel therapeutic strategies for cancer therapy. Ginger has been reported to exhibit potent anti-cancer effects, although previous reports have often focused on a narrow range of specific compounds. Through a direct comparison of various ginger compounds, we determined that gingerenone A selectively kills cancer cells while exhibiting minimal toxicity toward normal cells. Kinase array screening revealed JAK2 and S6K1 as the molecular targets primarily responsible for gingerenone A-induced cancer cell death. The effect of gingerenone A was strongly associated with relative phosphorylation levels of JAK2 and S6K1, and administration of gingerenone A significantly suppressed tumor growth in vivo. More importantly, the combined inhibition of JAK2 and S6K1 by commercial inhibitors selectively induced apoptosis in cancer cells, whereas treatment with either agent alone did not. These findings provide rationale for dual targeting of JAK2 and S6K1 in cancer for a combinatorial therapeutic approach.

Introduction

Recent progress in the understanding of signaling pathways in cancer have led to the development of target-based approaches for novel therapeutics with greater efficacy and selectivity. The survival and proliferation of cancer cells often depends on key oncogenic signaling intermediates, which provides an opportunity to inhibit such targets for the selective killing of cancer cells (1). However, the suppression of single cellular targets is unlikely to significantly extend disease-free survival for many cancer patients, as the amplification of redundant signaling pathways may allow for a negation of the inhibitory effect (2, 3). This observation has prompted the development of combinatorial targeting strategies involving multiple oncogenic pathways to achieve more desirable outcomes (3).

Dysregulation of the JAK2 signaling cascade has been implicated in multiple oncogenic processes including proliferation, metastasis, angiogenesis, inflammation, and evasion of immune processes (4). Stimulation of JAK2 by cytokines and growth factors subsequently leads to activation of STAT 3/5, promoting transcription of STAT-dependent genes (5). In addition, recent findings have shown that JAK2-mediated STAT-independent epigenetic up-regulation promotes tumor formation (6, 7). Various solid and hematological cancers have been reported to exhibit constitutive activation of JAK2/STAT3 signaling due in part to JAK2 activating mutations, elevated cytokine/growth factor levels, and inactivation of endogenous suppressors of the pathway (4). There are currently a number of JAK pathway inhibitors at various stages of clinical development (4, 8). However, despite the significant efficacy exhibited by JAK2 inhibitors, some patients fail to adequately respond or develop resistance, necessitating the development of improved approaches (5).

Another promising target for cancer therapy is the p70 ribosomal S6 kinase (S6K),⁵ a major downstream effector of the mTOR complex 1 pathway that integrates nutrient and growth factor signaling to enable coordinated cellular responses (9). S6K plays critical roles in the regulation of translation, cell growth, metastasis, and survival (10), and its overexpression has been linked to aggressive malignant phenotypes (9, 11–13). Because of its pivotal role in these processes, S6K has been recognized as a promising therapeutic target for a number of human cancers (10, 12, 14, 15). Genetic and pharmacological suppression of the S6K pathway has been shown to prevent malignant transformation and block cancer progression (9, 15–17). However, recent reports have exposed some limitations to the singular targeting of the S6K pathway alone, as such approaches can primarily induce cell cycle arrest with minimal induction of cell death (10, 18).

A significant number of current anticancer drugs have natural origins or were generated through the addition of structural modifications to existing natural compounds (19). Natural bioactive compounds provide a wealth of structural diversity, and their investigation for targeted therapies can offer promising opportunities to identify novel anticancer agents and aid in the understanding of cancer cell signaling (20, 21). Ginger (Zingiber officinale) has been reported to exhibit strong chemopreventive and chemotherapeutic properties in various cancer models, underlining its potential for clinical application (22–26). Gingerols are the most abundant form of aromatic compound present in ginger root and have been extensively studied for their anti-cancer effects (27–31). Shogaols are present at low concentrations in fresh ginger and can be found in dried and thermally treated ginger roots (32, 33). Shogaols generally exhibit anti-cancer effects at relatively low concentrations compared with gingerols (34–36). However, other bioactive compounds are present in ginger and may be responsible for the enhanced anticancer effects observed with whole ginger extract compared with ginger compound research mixtures (e.g. gingerols and shogaols) (22, 23, 37, 38). In the present study we examined a number of structurally diverse ginger compounds and discovered that gingerenone A exhibits potent and selective toxicity toward specific types of cancer cells. As the anticancer effects and mechanism of action of gingerenone A has not been reported, we sought to explore the underlying mechanism of action to identify its cellular target and further understand specific vulnerabilities present in malignant cells.

Experimental Procedures

Materials

Gingerenone A used in the study was synthesized based on a previous report (39), and the purity was >98% as measured by HPLC (1260 Infinity, Agilent Technologies, CA). 6-Gingerol, 8-gingerol, 10-gingerol, 6-shogaol, curcumin, and the antibody against β-actin were purchased from Sigma. Zingerone was purchased from Alfa Aesar. Tetrahydrocurcumin was ordered from Carbosynth LLC. 1,7-bis(4-hydroxy-3-methoxyphenyl)heptane-3,5-diol was provided by Dr. Jaebong Jang (Seoul National University, Seoul, Korea). PF-4708617 and NVP-BSK805 were obtained from Selleckchem. Antibodies against JAK2, STAT3, STAT1, STAT6, S6K1, S6, Akt, ERK, Bim, Bid, and PUMA and phosphorylated JAK2, STAT3, STAT1, STAT6, S6K1, S6, Akt, and ERK were purchased from Cell Signaling. Anti-Bad antibody was purchased from Santa Cruz Biotechnology Inc.

Cell Culture

Primary cultured human dermal fibroblast (HDF), IMR-90, SW480, A549, MDA-MD-468, HCT116, and EJ cells were grown in DMEM (Cellgro) supplemented with 10% FBS and penicillin/streptomycin. RPMI 1640 (Cellgro) supplemented with 10% FBS and penicillin/streptomycin was used for HuCCT1 and OVCAR-8 cells. CCD-18co cells were grown in MEM (Cellgro) supplemented with 10% FBS and penicillin/streptomycin. All cells were maintained as monolayer cultures at 37 °C in an incubator with an atmosphere of 5% CO₂.

Cell Viability Assay

Cell viability was determined using the sulforhodamine B-Based In Vitro Toxicology Assay kit (Sigma). Cells were plated in 6-well plates and on the next day were treated with compounds at the indicated concentrations. Staining and quantitative analyses were performed according to the manufacturer's instructions. All experiments were performed in triplicate.

High-throughput Kinase Profiling

Gingerenone A was sent out for screening against 413 kinases using Z′-LYTE® and Adapta® kinase activity assays and LanthaScreen® Eu Kinase Binding Assays in dry ice to a service provider (SelectScreen® Kinase Profiling Services; Life Technologies). Gingerenone A was dissolved in DMSO at 10 mm, and a final concentration of 10 μm was used for screening. Each kinase assay was performed in duplicate.

Apoptosis Analysis

Apoptosis was measured using the Annexin V-FITC Apoptosis Detection kit from MBL International Corp., Watertown, MA. Cells were collected with 0.025% trypsin + 5 mm EDTA in PBS, and 2.5% FBS in PBS + 5 mm EDTA was added as soon as the cells were released from the dish. Cells were washed with PBS and incubated for 5 min at room temperature with Annexin V-FITC plus propidium iodide following the protocol included in the kit. Cells were analyzed on a BD Biosciences FACSCalibur flow cytometer.

Animals

All animal procedures were conducted in accordance with animal care guidelines provided by Seoul National University (Seoul, Korea). Male nude mice (6-week-old) were purchased from the Institute of Laboratory Animal Resources at Seoul National University. Animals were acclimated for 1 week before the study and had free access to food and water. The animals were housed in climate-controlled quarters with a 12-h light/dark cycle.

Xenograft Model

HCT-116 cells in 100 μl were mixed with 100 μl of BD Matrigel^TM Basement Membrane Matrix (BD Biosciences). 2 × 10⁶ cells were implanted subcutaneously in the hind flank of each mouse. Mice were treated when their tumor volume reached ∼50 mm³ as measured using calipers, and the volume was estimated using the equation V = π/6(l × h × w). Gingerenone A was administered intraperitoneally for 12 days. All procedures were conducted in accordance with accepted guidelines for the use and care of laboratory animals.

Kinase Assays

Active JAK2 and p70S6K1 proteins and assays were purchased from Life Technologies and Millipore, respectively. Kinase assays for JAK2 or p70S6K1 were performed in accordance with the instructions provided by the manufacturer. GinA and active JAK2 or p70S6K1 were mixed and incubated at 30 °C for 10 min. After incubation, substrate peptide of JAK2 or p70S6K1 and 10 μl of diluted [γ-³²P]ATP solution was added. After additional incubation at 30 °C for 10 min, 25-μl aliquots were transferred onto p81 filter paper and washed 3 times with 0.75% phosphoric acid and 1 time with acetone for 5 min. The radioactive incorporation was determined using a scintillation counter (LS6500; Beckman Coulter). Each experiment was performed three times.

Structural Analysis Using Electron Microscopy

For negative staining, each kinase (JAK2 and S6K1) was diluted to 100 nm in 40 mm MOPS/NaOH (pH 7.0), 1 mm EDTA. To study structural changes after interaction with GinA, GinA was mixed with each kinase at a 1:100 ratio in 30 μm ATP. A 5-μl droplet was applied to a carbon-coated grid that had been freshly (within 30 min) irradiated with UV light (Mineralight Lamp, R-52G (ozone-producing); UVP, Upland, CA) for 40 min (40) and the grid stained using 1% uranyl acetate (41). Grids were examined at 120 kV using an FEI Tecnai 12 (FEI, Hillsboro, OR) transmission electron microscope, and digital images were recorded with an FEI Eagle 4K × 4K CCD camera. For image processing, particles were selected interactively from raw images using the display program JWEB associated with SPIDER (42), and the program was used for all subsequent image processing steps. Selected particles were windowed into individual images and subjected to 10 rounds of alignment and classification.

Immunoblotting

After cells were seeded in 100- or 60-mm dishes overnight the indicated compounds were treated. Harvested cells were disrupted with cell lysis buffer (0.5% Nonidet P-40, 50 mm Tris-HCl (pH 8.0), 100 mm NaCl, 5 mm EDTA, 2% β-mercaptoethanol), with protease inhibitor mixture (Roche Applied Science), and with a phosphatase inhibitor mixture (Roche Applied Science). The protein concentration was determined using a dye-binding protein assay kit (Bio-Rad) as described in the manufacturer's manual. Protein lysate was subjected to SDS-PAGE and electrophoretically transferred to a nitrocellulose membrane (Bio-Rad). The membrane was incubated with a specific primary antibody at 4 °C overnight. After incubation with a suitable secondary antibody, protein bands were visualized using Western Lightning Plus ECL (PerkinElmer Life Sciences).

Quantitative Analysis of Gingerenone A by LC-MS

The detection of gingerenone A was analyzed by an LC-MS Agilent 1200 Series analytical system equipped with a photodiode array (PDA) detector combined with a 6130 Series ESI mass spectrometer. Briefly, fresh ginger (Z. officinale) rhizomes were purchased from a market in Boston. 1 kg of ginger rhizomes was chopped and soaked in 1.5 liters of methanol (MeOH) at room temperature for 7 days, and the extract was filtered using filter paper and concentrated to dryness on a rotary evaporator at 31 °C to afford 12.3 g of sticky extract (G-R). To compare with a different condition of extraction, another 1 kg of the ginger rhizomes was extracted with 1.5 liter of methanol (MeOH) at 60 °C for 7 days, and the extract was evaporated under reduced pressure to obtain 14.1 g of sticky extract (G-H). Each extract was dissolved in 50% aqueous MeOH and filtered through a 0.50-μm syringe filter. The filtered samples were analyzed using a Phenomenex phenyl hexyl analytical column (4.6-mm × 100-mm inner diameter, particle size = 5.0 μm) with a Security guard cartridge (3.0-mm × 4-mm inner diameter, particle size = 5.0 μm) set at 25 °C. The mobile phase was a gradient program from mixtures of 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B), which was as follows: 0–40 min from 2% to 98% B, 40–45 min at 98% B, followed by a rapid drop to 2% B at 45.5 min, and then isocratic conditions with 2% B to 55 min (total 55 min). The flow rate was set at 0.5 ml/min, and the injection volume was 10 μl. Gingerenone A was detected at 27.1 min of retention time. Calibration curves and linear regression equations were generated for the external standard, gingerenone A. Quantification of gingerenone A was based on the peak area obtained from the MS detection and calculated as equivalents of the standard. All contents are expressed as μg per 100 g of extract weight. Each result shown in the table is the mean of three replicated measurements.

Synergy Assessment

Combination index (CI) was used to quantify synergism or antagonism for two drugs (43),

where CI < 1, =1, and > 1 indicate synergism, independence, and antagonism, respectively. In the denominators, (D_x)₁ represents D₁ “alone” that inhibits a system x%, and (D_x)₂ is for D₂ “alone” that inhibits a system x%. In the numerators, (D)₁ and (D)₂ “in combination” also inhibit x%. CI was calculated for every dose of two drug pairs. Fraction affected (F_a) is the fractional inhibition of a phenotype by a compound treatment(s). F_a of a group was calculated as F_a = percent inhibition of cell viability/100. Fraction affected-combination index plot was drawn with every group treated with more than one compound.

Statistical Analysis

Graphs are presented as the means of technical replicates with the error range indicated. Experiments shown are representative and have been repeated a minimal of three times. Software employed was GraphPad Prism v.6. 2-Tailed unpaired Student's t test was used to compare between two groups. As indicated in each figure legend, data are presented as the mean values ± S.D. p values are indicated in each case.

Results

Gingerenone A Selectively Kills Cancer Cells

To draw direct comparisons among anti-cancer effects of various ginger compounds, cell viability was measured after HCT116 cells were treated with a number of candidates (Fig. 1, A and B). Treatment with gingerenone A (GinA) caused the greatest reduction in cell viability among the compounds tested (Fig. 1B). GinA also significantly suppressed the growth of EJ, HCT116, OVCAR-8, MDA-MB-468, and A549 cancer cells whereas exhibiting relatively little toxicity toward SW480 and HuCCT1 cancer cells (Fig. 1C). The effect of GinA on the non-malignant cells, CCD-18co, primary cultured HDF, and IMR-90 was also relatively low (Fig. 1C), suggesting a selective cytotoxic effect toward certain types of cancer cells. To determine whether the inhibitory effect of GinA was due to an induction of cell death, we assessed apoptosis levels in cells exposed to GinA. Although the normal cell HDF showed no significant induction of apoptosis, EJ and HCT116 showed a marked induction of apoptosis after GinA treatment, suggesting that GinA selectively induces apoptosis in cancer cells (Fig. 1D).

FIGURE 1.

Gingerenone A selectively targets cancer cells. A, structures of ginger compounds. B, effect of ginger compounds on HCT116 cell viability. Viability was measured after HCT116 cells were treated with various ginger compounds for 48 h. GinA, gingerenone A; 6-Shoga, 6-shogaol; Curcu, curcumin; 8-Gin, 8-gingerol; 6-Gin, 6-gingerol; Zinger, zingerone; BH-diol, 1,7-bis(4-hydroxy-3-methoxyphenyl)heptane-3,5-diol; Tet-hyd curcu, tetrahydrocurcumin; 10-Gin, 10-gingerol. All data are presented as the mean ± S.D. C, effect of GinA on the viability of various normal and cancer cells. Cells were treated with GinA for 48 h before measuring viability. HDF, primary cultured human dermal fibroblasts. All data are presented as the mean ± S.D. D, GinA induces apoptosis in cancer cells. Primary cultured HDF, EJ, and HCT116 cells were treated with GinA for 48 h and analyzed for apoptosis. The right panel shows a quantification of apoptosis for which early apoptosis was measured using the Annexin V-only-positive population, and late apoptosis was measured using the Annexin V and propidium iodide (PI)-positive populations. All data are presented as the mean ± S.D. from three separate dishes measured simultaneously. *** p < 0.001, significant difference between DMSO-treated and GinA-treated cells.

JAK2 and S6K1 Are Targets of Gingerenone A

We next sought to understand the underlying mechanism responsible for the selective effects of GinA by identifying its cellular targets. We performed a high throughput screening process for inhibition and binding affinity of GinA with various kinases. We identified several kinases that were inhibited by GinA(Fig. 2A and supplemental Table 1) and selected JAK2 and S6K1 for further verification and analysis due to their previously reported roles in cancer cell survival and proliferation. GinA attenuated JAK2 kinase (Fig. 2B) and S6K1 activity (Fig. 2C) in a dose-dependent manner. Structural analysis using electron microscopy revealed that incubation with GinA displayed enlarged JAK2 and S6K1 structures (Fig. 2, D and E).

FIGURE 2.

Gingerenone A targets JAK2 and S6K1. A, target kinase screening with GinA. Kinase assay screening was performed using SelectScreen® Biochemical Kinase Profiling. B, JAK2 kinase assay. NVP-BSK805 was used as a positive control. All data are presented as the mean ± S.D. (***, p < 0.001, significant difference between DMSO-treated control and GinA- or NVP-BSK805-treated group) (C) S6K1 kinase assay. Staurosporine (STS) was used as a positive control. All data are presented as the mean ± S.D. ***, p < 0.001, significant difference between DMSO-treated control and GinA or STS-treated group (D). E, electron microscopy analysis of JAK2 and S6K1 structure with GinA. The top panels show raw images of the kinases in the presence or absence of GinA, and the bottom panel shows representative images of single kinase particles. The GinA reaction showed an increase in overall size.

To further examine the mechanism of GinA, the effect of GinA on JAK2 and S6K1 signaling pathways was investigated. Treatment of GinA in EJ and HCT116 cells caused a significant reduction in STAT3 phosphorylation while showing little or no effect against the phosphorylation levels of STAT1 and STAT6, implying a specific targeting of the JAK2-STAT3 pathway (Fig. 3A). GinA did not reduce the phosphorylation of JAK2 itself (Fig. 3A). Multiple JAK2 inhibitors have been previously reported to show similar results where they directly target JAK2 but increase or do not affect JAK2 phosphorylation levels (44, 45). GinA markedly inhibited S6 phosphorylation dose-dependently in both EJ and HCT116 cells (Fig. 3B) but did not reduce MAPK signaling. Treatment caused an increase in Akt phosphorylation in HCT116 cells, which may be due to negative feedback between the S6K1 and Akt pathways as previously reported in certain cell lines (Fig. 3B) (46). Further investigation into the selectivity of GinA revealed that cancer cells sensitive to GinA harbor relatively higher levels of JAK2 and S6K1 phosphorylation, demonstrating an association between the cytotoxic effect of GinA and activation levels of the target kinases (Fig. 3C). In addition to non-malignant cells (CCD-18co, HDF, and IMR-90), cancer cells relatively resistant to GinA (SW480 and HuCCT1) also had low levels of JAK2 and S6K1 phosphorylation, further supporting the conclusion that the sensitivity of GinA is dependent on JAK2 and S6K1 (Fig. 3C). Moreover, a comparison with 6-shogaol, curcumin, tetrahydrocurcumin, and zingerone showed that GinA more potently suppresses the STAT3 and S6 phosphorylation than the other ginger compounds, which also correlated well with their relative cytotoxic capacities (Fig. 3D).

FIGURE 3.

Gingerenone A inhibits JAK2 and S6K1 signaling. A, analysis of JAK2-STAT3 signaling. Immunoblots of lysates from EJ and HCT116 cells treated with GinA for 3 h. B, analysis of S6K1 signaling. Immunoblots of lysates from EJ and HCT116 cells treated with GinA for 3 h. C, expression levels of phosphorylated JAK2 and S6K1 in various cells. D, effect of gingerenone A, 6-shogaol (5-Sho), curcumin (Curc), tetrahydrocurcumin (TetCu), and zingerone (Zing) on phosphorylation of JAK2, STAT3, and S6. EJ cells were treated with compounds for 3 h and analyzed by immunoblot.

Gingerenone A Suppresses Tumor Growth in Vivo

Next, we examined if GinA was capable of suppressing tumor growth in vivo. Administration of GinA at 5 and 20 mg/kg in xenografted nude mice harboring HCT116 tumors resulted in a significant inhibition of tumor growth (Fig. 4A). Because GinA was found to target the JAK2-STAT3 and S6K1 pathways, we further examined whether the GinA-induced anti-cancer effect in vivo was associated with alteration in these signaling pathways. Analysis of tumor tissues obtained from mice showed marked inhibition in phosphorylation of STAT3 and S6 in the GinA-treated groups (Fig. 4B).

FIGURE 4.

Effect of gingerenone A in vivo. A, GinA suppresses tumor growth in an HCT116 xenograft model. Representative pictures of tumors on the top panel and quantified tumor weights. **, p < 0.01; ***, p < 0.001, significant difference between groups administrated with vehicle only and GinA treated group. B, effect of GinA on phosphorylation of JAK2, STAT3, and S6 in tumor tissues. HCT116 tumor tissues were homogenized and proteins were extracted for immunoblotting.

Co-inhibition of JAK2 and S6K1 Results in Synergistic Cytotoxicity in Cancer Cells

Because GinA was observed to specifically target JAK2 and S6K1, we questioned whether simultaneous targeting of these pathways could result in more potent killing of cancer cells. We treated EJ, HCT116, and HDF cells with NVP-BSK805, a well known commercial JAK2 inhibitor, and the S6K1 inhibitor PF-4708671. Interestingly, although individual treatment of NVP-BSK805 or PF-4708671 did not show a strong effect on cell viability, combined treatment with these two compounds elicited a significant synergistic effect on cancer growth suppression (Fig. 5, A and B). However, single as well as combinatorial treatment of NVP-BSK805 and PF-4708671 had little effect on the viability of non-malignant HDF cells (Fig. 5C). Interestingly, calculation of the combination index between NVP-BSK805 and PF-4708671 demonstrated that although co-treatment led to synergism in EJ and HCT116 cells, no synergism was observed in HDF normal cells (Fig. 5, A–C). These results show that simultaneous combination therapy targeting JAK2 and S6K1 can selectively and synergistically inhibit cancer cells. We further examined whether this co-treatment induced apoptosis. Although neither NVP-BSK805 nor PF-4708671 alone induced high levels of cell death, combination of the two compounds caused a significant increase in apoptosis in EJ and HCT116 cells (Fig. 5, D and E).

FIGURE 5.

Co-targeting JAK2 and S6K1 synergistically suppresses cancer cell viability. A–C, cell viability of HCT116, EJ, and HDF cells treated with NVP-BSK805 (BSK) and PF-4708671 (PF). Cells were treated at the indicated concentrations of NVP-BSK805 and PF-4708671 for 72 h. Color intensity represents relative cell viability compared with DMSO-only-treated control. The bottom panel shows combination indices for NVP-BSK805 and PF-4708671 in various cell lines. Combination indices were calculated based on formulas previously reported (43). Faction affected (F_a) of a group was calculated as F_a = percent inhibition of cell viability/100. D and E, co-treatment of NVP-BSK805 (BSK) and PF-4708671 (PF) induces apoptosis in EJ and HCT116 cells. Cells were treated with BSK and PF for 72 h and were analyzed for apoptosis. All data are presented as the mean ± S.D. F, co-treatment of NVP-BSK805 (BSK) and PF-4708671 (PF) induces Bim and PUMA expression. HCT116 and EJ cells were treated with 6 μm of BSK and 30 μm of PF for 36 h. G, GinA treatment induced Bim and PUMA expression. GinA was treated for 36 h.

Because dual targeting JAK2 and S6K1 significantly increased apoptosis in cancer cells, we examined whether the expression of apoptotic molecules was altered. We observed that subtypes of the proapoptotic protein Bim (BimEL, BimL, and BimS) were substantially induced when NVP-BSK805 and PF-4708671 were treated concomitantly, compared with single treatment with either compound alone (Fig. 5F). In addition, expression of the proapoptotic protein PUMA was specifically induced, whereas the expression of a number of other apoptotic proteins tested did not change upon NVP-BSK805 and/or PF-4708671 treatment (Fig. 5F). Treatment with GinA elicited a similar response pattern to that of NVP-BSK805 and PF-4708671 combination treatment, in terms of the induction of Bim (BimEL, BimL, and BimS) and PUMA expression in EJ and HCT116 cells (Fig. 5G). Taken together, these observations suggest that simultaneous targeting of JAK2 and S6K1 can lead to selective apoptotic responses in cancer cells.

Discussion

Recent advancements in molecular medicine have improved our understanding of the critical pathways that cancer cells are dependent upon for survival. However, it is becoming increasingly clear that these signaling cascades form complex interacting networks that incorporate redundant signaling pathways, limiting clinical efficacy for single target-based drugs and promoting resistance to therapy (3, 47). The optimal employment of combination therapies can enhance therapeutic responses in such scenarios by modulating the apoptotic balance to promote cell death. In our study, we found that a subset of cancer cells exhibited higher levels of JAK2 and S6K1 pathway activation compared with non-malignant cells and were strikingly sensitive to dual inhibition of these two kinases. Although JAK2 inhibitors are indicated for myeloproliferative neoplasms and are being evaluated for other cancers (5, 8), certain types of cancers are likely to have limited responses to JAK2 inhibition alone (48, 49). S6K1 is a primary downstream effector of mTORC1 signaling, and inhibition of this pathway has been known to cause a cytostatic effect in cancer cells rather than inducing cell death (18, 50). We have found that although JAK2 and S6K1 pathways are frequently up-regulated in cancers, in cases where single inhibition of either pathway has minimal effects, dual targeting can lead to a better therapeutic response (Fig. 6).

FIGURE 6.

Proposed mechanism of gingerenone A. A redundancy relationship between JAK and S6K1 for survival pathways in some cancer cell types suggests that their dual inhibition may induce higher levels of cell death.

Based on kinase assays and cell signaling results, there was clear evidence that gingerenone A inhibited JAK2; however, gingerenone A treatment did not reduce the phosphorylation of JAK2. Previous studies have also reported multiple JAK2 inhibitors that maintained or increased the phosphorylation levels of JAK2 (44, 45). In addition, inhibitors for other kinases have been shown to not affect the phosphorylation levels of the target kinase, whereas the activity was inhibited by the compounds (51–53). A possible explanation might be that a negative feedback pathway occurs after inhibition of JAK2 such that a hyperactivated upstream kinase could increase the phosphorylation level. Another possibility could be that if the compound merely inhibits the binding of ATP or hinders the binding between JAK2 and the substrate it could not affect the autophosphorylation site of JAK2 even after binding. These issues can be further studied in a more detailed analysis of the structural relationship between gingerenone A and JAK2.

A significant increase in the proapoptotic proteins Bim and PUMA was observed after GinA treatment and when JAK2 and S6K1 inhibitors were co-treated. Bim is a very potent inducer of apoptosis, and there are three major Bim isoforms, Bim short (BimS), Bim long (BimL), and Bim extra long (BimEL), each of which is generated by alternative splicing (54). BimS is the strongest inducer of cell death followed by BimL, with BimEL being the least potent inducer (55). Additionally, PUMA is known to antagonize all known anti-apoptotic Bcl-2 family members to induce mitochondrial dysfunction and apoptosis (56). Bim and PUMA in particular have been recognized for their pivotal role in oncogene inhibition-induced apoptosis (57–60). The marked induction of all isoforms of Bim and PUMA is likely to be at least partially responsible for the strong anti-cancer effect observed after JAK2 and S6K1 inhibition.

Various compounds in ginger have been reported to possess anti-cancer effects (27, 28, 36). We found gingerenone A and 6-shogaol to be the most potent among the tested ginger compounds. Interestingly, like shogaols (32, 33), the concentration of GinA in ginger extract also increased after exposure to heat (Table 1). This implies that precursors for both of these compounds are present in raw ginger, and the addition of heat may accelerate their conversion. It also suggests that the chemotherapeutic and chemopreventive effects of ginger may be enhanced by cooking before consumption. As ginger compounds exhibit promising anticancer therapeutic effects, further identification of the mechanisms involved could lead to development of novel anti-cancer drugs.

TABLE 1.

Gingerenone A content (μg/100 g of extract weight) in ginger (Zingiber officinale) rhizome extract

Each value represents the mean values derived from three independent replicates ± S.D. G-R, ginger extract at room temperature; G-H, ginger extract at 60 °C.

Sample	Gingerenone A
G-R	66.23 ± 9.69
G-H	447.57 ± 13.53

The continual development of new and better therapies for cancer treatment has been hindered in part by the considerable complexity of the intracellular signaling networks that govern processes responsible for the hallmarks of cancer. Against this backdrop, pharmaceutical research aims to identify appropriate targets and new combinational therapies with acceptable safety margins. The comprehensive investigation of redundancies and feedback mechanisms in cell signaling pathways requires a significant investment of resources, which has led to reduced pipeline diversity and longer developmental timelines. An alternative and potentially efficient strategy for determining new therapeutic combinations or simultaneous targets for synergistic effects may be to investigate the mechanisms responsible for the effects of natural bioactive compounds. As has been the case with GinA, this may reveal new potential weaknesses in the diverse signaling networks that cancer cells rely on for proliferation and survival.

Author Contributions

S. B. conducted most of the experimental work. S. L. participated in the experiments in Figs. 1B, 2C, and 4. J. Y. M. conducted the experiments in Fig. 2, D–E. K. H. K. conducted the experiments in Table 1. T. R. R. participated in the experiments in Fig. 1 and Table 1. L. F. participated in the experiments in Fig 1 and in writing the manuscript. S. H. S. participated in the experiments in Figs. 1D and 5. N. R. T. and K. H. K. participated in the synthesis of compounds. S. B., S. L., H. J. L., D. A. F., J. C., S. W. L., and K. W. L. designed the experimental plans, analyzed, and interpreted the data. J. C., S. W. L., and K. W. L. designed and directed the project. S. B., L. F., J. C., S. W. L., and K. W. L. wrote the manuscript.