The JAK/STAT (Janus kinase/signal transducer and activator of transcription) signaling pathways are major pathways for cytokine signaling and regulate important cellular events, such as hematopoiesis and immune development [1
]. Aberrant STAT3 signaling is frequently linked to cancer cell proliferation, survival, metastasis, tumor immunosuppression, and angiogenesis [3
]. JAK2 is the major kinase of STAT3 [8
] and has been reported to be constitutively active in many cancers and other proliferative diseases [9
]. The activating mutation V617F in JAK2 is considered as a driving factor for myeloproliferative disorders [10
]. Therefore, JAK2 has become an important target for the development of new drugs to treat these diseases [12
Thus far, most of the JAK2 inhibitors developed are ATP-competitive kinase inhibitors, such as CYT387, BMS911543, TG101348, and Ruxolitinib [14
]. Ruxolitinib was the first FDA-approved JAK inhibitor to treat myelofibrosis [17
]. However, it has not achieved significant reductions in disease burden in most patients with myelofibrosis [20
]. This failure is caused by the heterodimerization between activated JAK2 and other JAKs, resulting in the reactivation of this pathway. Moreover, CYT387, BMS911543, or TG101348 were cross-persistent to ruxolitinib [21
]. These findings suggest that more JAK2 inhibitors with different mechanisms are needed to improve efficacy. In this regard, covalent inhibitors have their unique advantages over ATP-competitive kinase inhibitors, because they can irreversibly interact with their targets and prevent the reactivation of JAK-STAT signaling described above. Furthermore, covalent inhibitors can dissociate drug pharmacodynamics from pharmacokinetics, resulting in desired drug efficacy with short systemic exposure to decrease drug interactions with off-targets. It can also achieve prolonged effects, resulting in less-frequent drug dosing [22
Parthenolide (PN) is an abundant sesquiterpene lactone found in the medicinal herb Feverfew (Tanacetum parthenium
). It has been used to treat arthritis, fever, headache for centuries in Europe. PN has been reported to exhibit inhibitory activity on the IL-6-induced STAT3 activation, which contributes to its anti-inflammation and anti-cancer properties [23
]. However, the mechanisms of PN as an inhibitor of JAK-STAT3 signaling are still unknown. In the present report, we investigated the molecular mechanisms of PN in regulating JAK-STAT3 signaling. Our study provides a covalent strategy to develop a JAK inhibitor and suggests PN as a promising anti-inflammation and anticancer drug candidate.
Although PN has been reported to inhibit JAK-STAT3 signaling pathway, its precise molecular mechanism has not been understood [23
]. We presented evidence to demonstrate that PN was a covalent pan-JAK inhibitor. PN directly interacted with JAKs by covalently binding to specific cysteines of JAKs and inactivated their enzymatic activities.
The cysteines targeted by PN were all located in the FERM-SH2 domain of JAK2, which is necessary for the interaction between JAKs and receptors and is important for JAK activation [28
]. The physical interactions between JAK2 and Gp130, however, did not seem to be affected by PN (data not shown), suggesting that PN may inactivate the JAKs mainly by changing their conformations.
LC/MS analysis demonstrated that three cysteines, Cys178, Cys243, and Cys480, of JAK2 were covalently modified by PN. We noticed from the protein sequence alignment that Tyk2 seemed to have only one of the three PN-targeted cysteines, the Cys536, the JAK2-Cys480 equivalent. It is, therefore, possible that a Tyk2 mutant at Cys536 may lose the inhibitory effect of parthenolide.
The binding of PN to the JAKs were quite selective, although not specific. The human JAK2 contains 27 cysteines and PN only bound to three of them. The computational analyses of the PN-JAK2 interaction predicted the binding preference for Cys243 and Cys480, which is consistent with our experimental data. The binding of PN to Cys178 is somewhat unexpected because cys178 is buried inside of the protein according to the crystal structure of JAK2. There is a possibility that a protein misfolding occurred during the overexpression and preparation of JAK2 so that certain cysteines, such as cys178, might become accessible to PN. PN also preferred JAKs over abundant proteins, such as tubulin and actin, and had little inhibitory effects on the kinase activities of PI3K, c-Src, MAPK1, EGFR, VEGFR3, and IGF-IR. Therefore, an appropriate tertiary structure of JAKs may be required for PN to interact with specific cysteines [40
]. We also noticed that the phosphorylation of EGFR was slightly up-regulated by PN. It was reported that activation of EGFR was associated with ROS production, which transiently inactivates protein tyrosine phosphatases to enhance or prolong EGFR activation [41
]. Thus, it is possible that the PN-induced up-regulation of EGFR phosphorylation was the result of PN-induced ROS.
Nucleophilic methylene-γ-lactone rings of natural compounds have often been reported to induce ROS in cells. PN also contains the methylene-γ-lactone ring and has the ability to induce ROS, which has been reported to regulate STATs signaling [42
]. We, therefore, analyzed the roles of the PN-induced ROS in the inhibition of JAK/STAT3 signaling by PN. We found that the PN-induced ROS did not contribute to the PN inhibition of STAT3 phosphorylation in the MDA-MB-231 cells. This discrepancy may be due to different cellular contexts. Similar data has been reported that an oxidative stress caused by hydrogen peroxide treatment resulted in the inhibition of STAT phosphorylation in neuronal cells, but not in non-neuronal cells. The activation of Src in the non-neuronal cells could be a possible mechanism [44
In summary, our study demonstrated PN as a novel covalent pan-JAK inhibitor. Elucidating the molecular mechanism of PN in inhibiting the JAK/STAT signaling pathway will contribute to therapeutic developments of JAK inhibitors and will help to make better use of PN in anti-inflammation and anti-cancer therapy.
4. Materials and Methods
4.1. Cell Culture
HepG2/STAT3 cells are gifts from Professor Xin-Yuan Fu (National University of Singapore, Singapore), which were stably transfected with STAT3-responsive firefly luciferase reporter plasmid. All other cell lines were obtained from the American Type Culture Collection. MEF, HEK293, Hela, Hs578t, HBE, H4, MDA-MB-453 cells were cultured in DMEM (Gibco, Grand Island, NY, USA) supplemented with 10% FBS (Gibco, Grand Island, NY, USA). MDA-MB-231, MDA-MB-468, HCT116, HT-29, Lovo, NCI-H1299, Colo205, BGC, H460 cells were grown in RPMI 1640 medium (Gibco, Grand Island, NY, USA) supplemented with 10% FBS. NCI-H1975 and Du145 were grown in RPMI 1640 medium (Gibco, Grand Island, NY, USA) supplemented with 10% FBS and 2 mmol/L glutamines. HepG2/STAT3 cells were cultured in MEMα medium (Gibco, Grand Island, NY, USA) supplemented with 10% FBS. All cell lines were cultured in 37 °C, 5% CO2, and a humidified atmosphere of 95% air.
4.2. Chemicals and Reagents
The following chemicals and reagents were used: parthenolide (Sigma-Aldrich, Saint Louis, MO, USA); AZD1480 (Selleck, Shanghai, China); DTT and MTT (Genebase, Shanghai, China); GSH (Shanghai Sibas Bioscience, Shanghai, China); IL-6 and IFNα (Peprotech, MN, USA); streptavidin agarose and DCFH-DA probe (Thermo Fisher Scientific, Waltham, CA, USA); total glutathione assay kit (Beyotime, Shanghai, China).
Plasmids encoding JAK2-FLAG were a gift from Prof. David E. Levy (New York University).
The following antibodies were from Cell Signaling Technology (Boston, MA, USA): phospho-EGFR (1:2000, #3777), EGFR (1:2000, #4267), phospho-InsR/IGF1R (1:2000, #3024), IGF1R (1:2000, #3018), phospho-Y705-STAT3 (1:3000, #9145), STAT3 (1:3000, #9139), STAT1 (1:2000, #9172), phospho-Y1007/1008-JAK2 (1:2000, #3776), JAK2 (1:2000, #3230), phospho-Y1022/1023-JAK1 (1:1000, #3331), JAK1 (1:2000, #3332), phospho-Y1054/1055-TYK2 (1:1000, #9321), TYK2 (1:2000, #9312), and Cofilin (1:2000, #5175). The antibodies for GP130 (1:1000, #sc-656) and α-Tubulin (1:3000, #SC-5286) were from Santa Cruz Biotechnology (Dallas, TX, USA). The antibodies for β-Actin were from Abmart (1:3000, #P30002M). The antibodies for phospho-Y1230/1231-VEGFR3 (1:2000, #CY1115) were from Cell Applications. Secondary HRP-conjugated antibodies were from Multi Sciences Biotech (1:5000, Hangzhou, China). Anti-Flag affinity gel was purchased from Bimake (1:50, #B23101, Shanghai, China).
4.4. Luciferase Assay
HepG2/STAT3 cells were seeded onto 96-well cell culture plates and grew to 90% confluence. Cells were then treated with PN for 1 h followed by stimulation with 10 ng/mL IL-6 for 4 h. Luciferase activity was determined using Promega luciferase kits according to the manufacturer’s instruction (Promega, Madison, WI, USA).
4.5. Western Blot Analysis
Western blotting was performed as previously described [45
4.6. JAK2 In Vitro Kinase Assay
The JAK2 in vitro kinase assay was performed using JAK2 immunoprecipitants, an HTScan JAK2 Kinase Assay Kit (Cell Signaling Technology, Beverly, MA, USA) and streptavidin-coated 96-well plates (#22351, Beaverbio, Suzhou, China). HEK293 cells were transfected with plasmids encoding JAK2-FLAG for 24 h by Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) and then lysed with 1 mL lysis buffer (50 mmol/L HEPES (pH 7.4), 150 mmol/L, 0.15% Triton X-100, NaCl, 0.5 mmol/L DTT, 2 mmol/L Na3
, 2 mmol/L NaF, 1 mmol/L PMSF, and protease inhibitor cocktail (Sigma, 1:1000)) on ice for 30 min. Lysates were centrifuged and the supernatants were immunoprecipitated with anti-Flag affinity gel. The immunoprecipitates were rinsed with kinase reaction buffer (60 mmol/L HEPES (pH 7.5), 5 mmol/L MnCl2
, 5 mmol/L MgCl2
, 25 μmol/L Na3
, 200 μmol/L ATP). JAK2 protein was incubated with PN for 30 min in kinase reaction buffer. The reaction was started by adding 1.5 μmol/L FLT3 substrate into the reaction buffer. The reaction was incubated at 25 °C for 30 min. In vitro kinase activity was conducted as described [46
]. We chose to use HEK 293T cells because they are amenable to liposome-mediated transformations.
4.7. HPLC-MS Analysis of GSH-PN Adduct
A total of 0.5 mmol/L PN was incubated with 2 mmol/L of GSH in 20 mmol/L Tris-HCl (pH 7.5) for 1 h at 37 °C. The products were subjected to LC-MS system (ESI-positive mode; mobile phase: methanol/water (1:1 v/v); rate: 0.2 mL/min). The Agilent HP 1100 series HPLC system was from Agilent Technologies (Palo Alto, CA, USA). The LCQ Deca ion trap mass spectrometer was from Thermo Finnigan (San Jose, CA, USA).
4.8. Syntheses of Biotinylated Parthenolide
Biotinylated parthenolide (Bio-PN) was prepared by oxidation of parthenolide first with selenium dioxide and tert-butylhydroperoxide to furnish the allylic alcohol, as previously described [47
4.9. Pull-Down Assay
MDA-MB-231 cells were grown on 100 mm dishes to 100% confluence. Then cells were pretreated with PN or DMSO for 1 h and incubated with Bio-PN or DMSO for 1 h. After that, cells were lysed with 1 mL lysis buffer (pH 7.4, 50 mmol/L Tris-HCl, 150 mmol/L NaCl, 1 mmol/L EDTA, 1% NP-40, 1 mmol/L PMSF, and protease inhibitor cocktail) on ice for 0.5 h. Lysates were centrifuged at 12,000× g at 4 °C for 10 min. The supernatant was incubated with 5% (v/v) streptavidin agarose beads for 2 h at room temperature. The precipitates were rinsed five times with lysis buffer and lysed with Laemmli buffer, followed by Western blot analysis.
4.10. In Vitro Kinase Assay
4.11. HPLC–MS/MS Analysis of JAK2
HEK293 cells were transfected with plasmids encoding mouse JAK2-FLAG for 24 h by Lipofectamine 2000 (Invitrogen). Cells were then treated with 20 mmol/L PN for 1 h and lysed as described above. JAK2 protein was immunoprecipitated by anti-Flag affinity gel and harvested by Laemmli buffer. JAK2 was pre-separated by SDS-PAGE and was cut off from the PAGE and digested in gel [37
]. The tryptic peptides were desalted and dried in a Speed-Vac. The dried peptides were processed as described previously [48
]. Peak lists of the HPLC/MS/MS data were generated by Proteome Discoverer software (version 1.4, Thermo Fisher) and searched against the UniProt Human database by Mascot (v2.3, Matrix Science Ltd., London, UK). The protease was set to trypsin/P allowing for a maximum of two missed cleavage sites. The fixed modification of cysteine residues was set to carbamidomethylation. The variable modification was set to nature product derivatization, protein N-terminal acetylation, and methionine oxidation. The drug-modified peptide spectra with a Mascot ion score of more than 20 were manually inspected with stringent criteria, as described [38
4.12. Docking Study
Docking study was conducted by Maestro 10.1. Crystal structure of JAK2 (4Z32) was downloaded from RCSB (Protein Date Bank). Protein Preparation Wizard Workflow (Schrödinger program suite) was used to prepare the protein and Ligand Preparation was used to prepare the compound. PN was docked into the defined binding site without constraint. Results were generated by Pymol based on the Prime-score.
4.13. Determination of Cellular ROS
Accumulation of intracellular ROS was detected with the probe DCFH-DA as described [49
]. Briefly, after drug treatment, cells were incubated with 10 μmol/L DCFH2-DA in the cell culture incubator for 20 min. The labeled cells were washed and harvested. To quantify ROS, the fluorescence intensity was measured by flow cytometry (FACSCalibur, BD Biosciences, San Diego, CA, USA).
4.14. Determination of Cellular GSH Level
Cellular glutathione level was determined by a Total Glutathione Assay Kit (Beyotime, Shanghai, China) according to the instruction of the manufacturer.
4.15. MTT Assay
MTT assay was performed as previously described [45
4.16. Fluorimetric Method to Determine the Viability of MDA-MB-231
MDA-MB-231 cells were seeded at 20,000 cells per well in serum-free 1640 medium with or without 50 ng/mL IL-6 in a 96-well plate. Cells were then treated with PN. Twenty microliters of 10% alamar blue was added into each well. Fluorescence was measured 4, 6, 8, 12 h later.
4.17. Migration Assay
MDA-MB-231 cells were seeded at 20,000 cells per well in serum-free 1640 medium with or without 50 ng/mL IL-6 in the upper chamber of CIM-plate16. And 150 µL 1640 medium with 10% FBS was added in the lower chamber of CIM-plate16. Cells were then treated with PN. Insert the CIM-plate 16 into the RTCA DP Analyzer (Roche Applied Science, Penzberg, Germany). The following steps were performed as described [50