Xanthatin inhibits STAT3 and NF‐κB signalling by covalently binding to JAK and IKK kinases

Abstract Aberrant activation of the signal transducer and activator of transcription 3 (STAT3) and the nuclear factor‐κB (NF‐κB) signalling pathways is associated with the development of cancer and inflammatory diseases. JAKs and IKKs are the key regulators in the STAT3 and NF‐κB signalling respectively. Therefore, the two families of kinases have been the major targets for developing drugs to regulate the two signalling pathways. Here, we report a natural compound xanthatin from the traditional Chinese medicinal herb Xanthium L. as a potent inhibitor of both STAT3 and NF‐κB signalling pathways. Our data demonstrated that xanthatin was a covalent inhibitor and its activities depended on its α‐methylene‐γ‐butyrolactone group. It preferentially interacted with the Cys243 of JAK2 and the Cys412 and Cys464 of IKKβ to inactivate their activities. In doing so, xanthatin preferentially inhibited the growth of cancer cell lines that have constitutively activated STAT3 and p65. These data suggest that xanthatin may be a promising anticancer and anti‐inflammation drug candidate.

receptor oligomerization leads to transphosphorylation and activation of JAKs and phosphorylation of the receptors and their substrates STATs. Phosphorylated STATs are then dimerized and translocate to the nucleus and function as transcription factors. 2,4,5 Accumulating evidence indicate that abnormalities in the JAK/STAT signalling pathways are associated with inflammation and cancer. 7,8 JAKs are the key enzymes in regulating these pathways and are promising drug targets. A pan-JAK inhibitor ruxolitinib has already been approved by FDA for the treatment of myeloproliferative neoplasms. 17 A JAK3 inhibitor tofacitinib has also been approved for the treatment of rheumatoid arthritis and is being evaluated for the treatment of psoriasis and inflammatory bowel disease. 18,19 The transcription factor nuclear factor-κB (NF-κB) regulates the expression of a wide range of genes vital for immune responses and cell survival and proliferation. 20 In unstimulated cells, the NF-κB dimers are sequestered in the cytoplasm by inhibitory subunit IκBs (Inhibitor of κB). In the canonical NF-κB pathway, an IKK (IκB kinase) complex composed of IKKα, IKKβ and IKKγ regulates the activation of NF-κB. Upon stimulation, the IKK complex phosphorylates IκBα, leading to the degradation of IκBα and the release of NF-κB complex.
NF-κB can be further phosphorylated by IKK complex, resulting in enhanced transcriptional activity. 21,22 Constitutive NF-κB activation occurs in chronic inflammation and in a wide range of haematological and solid tumours, making NF-κB signalling pathway an attractive target for the development of anti-inflammatory and anticancer drugs. 25,26 Because of the indispensable role of IKK in the activation of NF-κB signalling pathway, development of IKK inhibitors is an effective approach to block NF-κB signalling. 31,32 Xanthatin (Xa) is a bioactive compound identified from the plant Xanthium L., which has been used as an anti-inflammatory herb in traditional Chinese medicine to treat diseases such as nasal sinusitis and arthritis. Xanthatin has been reported to inhibit both signal transducer and activator of transcription 3 (STAT3) and NF-κB activation, but the molecular mechanisms of action are still unknown. 34,35 Here, we report that xanthatin is a covalent and selective inhibitor of JAKs and IKKs. It preferred JAKs and IKKs over abundant proteins, such as tubulin and actin and had no effects on many other kinases. Moreover, xanthatin selectively inhibited the growth of cancer cells with constitutively activated STAT3. These data explain how Xanthium L. plant may act as an anti-inflammatory herb and suggest that xanthatin may be a promising anti-inflammation and anticancer drug candidate.

| The preparation of Xa, Xa-2 and Xa-3
The xanthatin (Xa), 11α, 13-dihydroxanthatin (Xa-2) and xanthinosin (Xa-3) were isolated from the aerial parts of Xanthium mogolium Kitag, as described previously. 37 Their structures were determined by LC-MS, 1 H NMR and 13 C NMR and are in accordance with the published data.

| The preparation of Xa-1
Xanthatin (100 mg, 0.41 mmol) was dissolved in methanol (10 mL), followed by addition of Pd/C (500 mg) under N 2 protection. H 2 (~1 atm) was then filled into the flask and the mixture was incubated with stirring for 1 hour at room temperature. The mixture was then filtrated with celite and the filtrate was concentrated under reduced pressure. The concentrated residue was purified with silica chromatograph to yield the colourless oil Xa-1 quantitatively.  The reaction was then transferred to a 50°C oil bath and stirred for 2 hours. Then the solvent and the redundant oxalyl chloride were removed under reduced pressure. The residue was dissolved in 3 mL anhydrous DCM and was stirred in ice-water bath under N 2 protection, followed by addition of 13 μL 2-propynylamine (0.20 mmol) and 55 μL triethylamine (0.40 mmol). Thirty minutes later, the mixture was diluted with 40 mL DCM and washed with 30 mL saturated NaHCO 3 aqueous solution and 30 mL saturated NaCl aqueous solution sequentially and dried with anhydrous Na 2 SO 4 . The organic solvent was removed under reduced pressure and the remaining residue was purified with silica gel chromatograph to produce 38 mg Alk-Xa with a yield of 67%.

| Antibodies
The following antibodies were purchased from Cell Signaling

| Luciferase assays
HEK293/NF-κB or HepG2/STAT3 cells were seeded into 96-well cell culture plates and cultured to 90% confluence. Cells were then treated with xanthatin for 1 hour followed by stimulation with 1 ng/ mL TNFα or 10 ng/mL IL-6 for 4 hours. Luciferase activities were determined using the Promega luciferase kits according to the manufacturer's instructions (Promega, Madison, WI, USA).

| In vitro kinase assays
The JAK2 or IKKβ in vitro kinase assay was performed with JAK2

| HPLC-MS analyses of adduct of xanthatin and GSH
About 20 μmol/L xanthatin was incubated with 1 mmol/L of GSH in 20 mM Tris-HCl (pH 7.4) for 1 hour at 37°C. Then the sample was injected into the liquid chromatography-mass spectrometry (LC-MS) system using methanol/water (1:1) as the mobile phase at a rate of 0.2 mL/min. at 4°C and the pellet was washed twice with pre-chilled acetone, airdried for 10 minutes and resuspended in the cell lysis buffer, followed by incubation with 5% (v/v) streptavidin agarose beads for 2 hours at room temperature. The precipitates were washed five times with the lysis buffer and dissolved in Laemmli buffer, followed by Western blot analyses.

| In-gel fluorescence
Overexpressed JAK2 and IKKβ proteins were immunoprecipitated as described above. The immunoprecipitates were incubated with DMSO or Alk-Xa for 30 minutes. About 100 μmol/L rhodamineazide and freshly premixed click chemistry reaction cocktail described above were added. The mixtures were incubated at room temperature for 2 hours with gentle mixing. Five volumes of pre-chilled acetone was added and the mixtures were incubated at −20°C overnight and were then centrifuged at 14 000 g for 10 minutes at 4°C. The pellets were air-dried for 10 minutes and dissolved in Laemmli buffer. The proteins were resolved by 8% denaturing PAGE and the fluorescence was detected by a GE Typhoon scanner.
F I G U R E 1 Xanthatin blocked signal transducer and activator of transcription 3 (STAT3) signalling by directly inhibiting JAK2 kinase activity. A, Structure of xanthatin. B, Effects of xanthatin on the IL-6-induced luciferase activity. HepG2/STAT3 cells were pre-treated with xanthatin at indicated concentrations for 1 h and luciferase activity was measured following stimulation of IL-6 (10 ng/mL) for 4 h. C, Xanthatin dose-dependently inhibited IL-6-induced STAT3 phosphorylation. MDA-MB-231 cells were pre-treated with xanthatin at various concentrations for 1 h before stimulation by IL-6 (10 ng/mL) for 10 min. Cell lysates were processed for Western blot analysis. D, Xanthatin time-dependently inhibited IL-6-induced STAT3 phosphorylation. MDA-MB-231 cells were incubated with 5 μmol/L xanthatin for various durations before stimulation by IL-6 (10 ng/mL) for 10 min. Cell lysates were processed for Western blot analysis. E, Xanthatin dosedependently inhibited IL-6-induced JAK2 phosphorylation. MDA-MB-231 cells were pre-treated with xanthatin at various concentrations for 1 h before stimulation by IL-6 (10 ng/mL) for 10 min. Cell lysates were processed for Western blot analysis. F, Effects of xanthatin on the IFNα-induced JAK1 and TYK2 phosphorylation. MDA-MB-231 cells were pre-treated with xanthatin at indicated concentrations for 1 h before stimulation by IFNα (1000 IU/mL) for 10 min. Cell lysates were processed for Western blot analysis. G, JAK2 in vitro kinase assay. The overexpressed JAK2 protein immunoprecipitated from the HEK293 was subjected to in vitro kinase assay as described in Materials and Methods. (n = 3, * P < 0.05, **P < 0.01, ***P < 0.001 compared with IL-6-induced cells)

| Coomassie G-250 staining
The coomassie G-250 staining was performed as described. 38 39 The drug modified peptide spectra with a Mascot ion score of more than 20 were manually inspected using stringent criteria as previously described. 40

| Statistical analyses
The statistical analyses were performed with the Graphpad Prism software. IC50 values were calculated by the Graphpad Prism 7. For comparisons between and within more than two groups, one-way Analysis of Variance (ANOVA) and two-way ANOVA were used, followed by the Dunnett's multiple comparisons test. All values are reported as mean ± SD.

| Xanthatin inhibited the IL-6-induced STAT3 activation by directly inactivating JAK2 kinase activity
We investigated the effects of xanthatin ( Figure 1A JAK2 is the major kinase of STAT3. 41 To explore whether the inhibition on STAT3 phosphorylation was the result of inactivating JAK2 by xanthatin, we examined the effects of xanthatin on the phosphorylation/activation of JAK2. Xanthatin inhibited the IL-6induced JAK2 phosphorylation in a similar fashion as that on the STAT3 phosphorylation ( Figure 1E). Therefore, JAK2 appeared to be the direct target of Xa.
To investigate the specificity of xanthatin, we analysed the effects of xanthatin on the IFNα-induced phosphorylation of JAK1 and TYK2, two other members of the JAK family. The IFNα-induced phosphorylation of JAK1 and TYK2, as well as STAT3, was also inhibited by xanthatin ( Figure 1F). It appeared that xanthatin was a pan-JAK inhibitor.
To confirm that xanthatin is a direct inhibitor of JAK kinases, we overexpressed and immunoprecipitated JAK2 protein from HEK293 cells and performed an in vitro kinase assay. Xanthatin inhibited JAK2 kinase directly with an IC50 of 4.078 μmol/L ( Figure 1G). These data suggested that xanthatin inhibited the IL-6-induced STAT3 activation by directly inhibiting the JAK2 kinase activity.

| Xanthatin inhibited NF-κB signalling by blocking IKKβ kinase activity
We next investigated the effects of xanthatin on the NF-κB signalling using a Hek293 cell line transfected with a NF-κB-responsive luciferase reporter gene and found that the TNFα-induced NF-κBresponsive luciferase activity was inhibited by xanthatin in a dosedependent manner (IC50 = 9.607 μmol/L) (Figure 2A).
To understand the mechanisms of xanthatin in inhibiting the activation of NF-κB, we examined the effects of xanthatin on IKKα/β. Figure 2B-C, the TNFα-induced phosphorylation of IKKα/β was inhibited by xanthatin in a dose-and time-dependent manor, leading to the inhibition of IκBα degradation as well as p65 phosphorylation.

As shown in
We then investigated the effects of xanthatin on the activity of IKKβ, the key kinase of the NF-κB pathway, by overexpressing and immunoprecipitating the IKKβ protein from the HEK293 cells to perform an in vitro kinase assay. Xanthatin directly inhibited IKKβ kinase activity with an IC50 of 11.315 μmol/L ( Figure 2D). Hence, xanthatin blocked NF-κB signalling by directly inhibiting IKKβ kinase.

| The activity of xanthatin was blocked by GSH and was dependent on its α-methylene-γbutyrolactone group
The α, β-unsaturated carbonyl group of xanthatin can serve as a site for Michael addition, react with protein thiols and covalently bind to the proteins. 42,43 To find out whether xanthatin is a covalent inhibitor, we pre-incubated xanthatin with the thiol-containing glutathione (GSH) and then to examine whether this pre-incubation could alleviate the inhibitory effects of xanthatin on the two signalling pathways. The inhibitory effects of xanthatin were completely abrogated by GSH ( Figure 3A). We next analysed the incubation products of xanthatin and GSH by LC-MS. As xanthatin possesses two α, β-unsaturated carbonyl groups, there should be three products if both α, β-unsaturated carbonyl groups were reactive. However, there was only one major product with a molecular weight of 553, suggesting an addition of one molecule of GSH to one molecule of xanthatin ( Figure 3B).
We then assessed the reactivity of the two α, β-unsaturated carbonyl groups of xanthatin to understand their contributions to the inhibitory activity of xanthatin by structure modifications of the two α, β-unsaturated carbonyl groups. As shown in Figure 3D-E, the major α, β-unsaturated carbonyl group that contributed to the inhibition of JAK/STAT and NF-κB signalling pathways was the α-methylene-γbutyrolactone group. Modification of the other group had very little effects. Taking together, these data suggested that xanthatin was a covalent inhibitor and its inhibitory activity mainly relied on its α-methylene-γ-butyrolactone group.

| Xanthatin covalently modified Cys243 of JAK2 and Cys412, Cys464 of IKKβ
Above data suggested that xanthatin was a covalent inhibitor.
To verify whether xanthatin covalently modified JAK2 and IKKβ, we synthesized an alkyne-containing xanthatin analogue Alk-Xa ( Figure 4A). Alk-Xa maintained the biological activities of xanthatin ( Figure 4B-C). We then overexpressed and immunoprecipitated the F I G U R E 2 Xanthatin blocked NF-κB signalling by directly inhibiting IKKβ kinase activity. A, Effects of xanthatin on the TNFα-induced luciferase activity. HEK293/NF-κB cells were pre-treated with xanthatin at indicated concentrations for 1 h and luciferase activity was measured following stimulation of TNFα (1 ng/mL) for 4 h. B, Xanthatin dose-dependently inhibited TNFα-induced NF-κB pathway. MDA-MB-231 cells were pre-treated with xanthatin for 1 h before stimulation by TNFα (1 ng/mL) for 10 min. Cell lysates were processed for Western blot analysis. C, Xanthatin time-dependently inhibited TNFα-induced NF-κB pathway. MDA-MB-231 cells were pre-treated with 20 μmol/L xanthatin for various durations (0-120 min) before stimulation by TNFα (1 ng/mL) for 10 min. Cell lysates were processed for Western blot analysis. D, IKKβ in vitro kinase assay. The overexpressed IKKβ protein immunoprecipitated from the HEK293 was subjected to in vitro kinase assay as described in Materials and Methods. (n = 3, *P < 0.05, **P < 0.01, ***P < 0.001 compared with TNFα-induced cells)

JAK2 and IKKβ proteins from HEK293 cells and incubated them with
Alk-Xa. The incubation products were subsequently subjected to click chemistry with rhodamine-azide. The in-gel fluorescence scanning showed that the fluorescence intensity of JAK2 and IKKβ was proportional to the dosage of Alk-Xa, indicating that Alk-Xa covalently bound to JAK2 and IKKβ ( Figure 4D).  Figure 4E).
We then analysed the location of Cys243 in the crystal structure of JAK2. 46 As shown in Figure 4F Figure 4G).

F I G U R E 3
Xanthatin had the potential to react with protein thiols and its activity depended on its α-methylene-γ-butyrolactone moiety. A, GSH blocking assay. HepG2/STAT3 cells were pre-treated with 1 mmol/L GSH, xanthatin or their mixture for 1 h and then stimulated with IL-6 for 4 h. Cells were harvested for luciferase assay. B, LC-MS analysis of the incubation product of xanthatin and GSH. About 20 μmol/L xanthatin was incubated with 1 mmol/L GSH for 1 h at 37°C and the mixture was resolved by LC-MS as described in Materials and Methods. Molecular weights of the molecules are indicated. C, Xanthatin and its derivatives. D, The effects of xanthatin and its derivatives on STAT3 activation were measured by luciferase assay. HepG2/STAT3 cells were pre-treated with Xa, Xa-1, Xa-2 and Xa-3 for 1 h and luciferase activity was measured following stimulation of IL-6 (10 ng/mL) for 4 h. E, The effects of xanthatin and its derivatives on NF-κB activation were measured by luciferase assay. HEK293/NF-κB cells were pre-treated with Xa, Xa-1, Xa-2 and Xa-3 for 1 h and luciferase activity was measured following stimulation of TNFα (1 ng/mL) for 4 h. (n = 3, *P < 0.05, **P < 0.01, ***P < 0.001 compared with cytokine-induced cells) We next performed similar analysis for the interaction between xanthatin and IKKβ and found that the Cys412 and Cys464 of IKKβ were covalently modified by xanthatin ( Figure 4H-I). The locations of Cys412 and Cys464 in the crystal structure of IKKβ were also examined. 47 We found that the Cys464 was on the surface, while the Cys412 was located deep in a cavity of IKKβ ( Figure 4J). In the molecular docking study, only the Cys464 could be successfully docked and the binding mode of xanthatin to the Cys464 showed that the carbonyl group of the side chain of xanthatin formed an H-bond with the Asn457 of IKKβ. The estimated free energy of xanthatin to cys464 was much lower, suggesting that cys464 was more easily to be attacked by xanthatin ( Figure 4K).

| Xanthatin preferentially bound to JAKs and IKKs
Many proteins, including abundant proteins like tubulin and actin, contain exposed cysteine residues that can potentially be modified by reactive oxygen/nitrogen species and electrophiles. 48,49 To determine whether xanthatin indiscriminately reacted with cysteine-containing proteins, we used Alk-Xa to pull down the xanthatin-interacting proteins. After incubation with Alk-Xa, the cells were lysed and precipitated with streptavidin resin and subjected to Western blotting analysis. As shown in Figure 5A, JAK1, JAK2, Tyk2, IKKα and IKKβ were pulled down with the Alk-Xa and the interactions could be competed away by excessive unlabelled xanthatin, indicating that Alk-Xa interacted with these proteins. On the contrary, Alk-Xa failed to pull down gp130 and p65, two components in the JAK/STAT and NF-κB pathways respectively. We also analysed several abundant proteins, such as tubulin, actin, cofilin and α-actinin, to examine the selectivity of xanthatin. Our data demonstrated that Alk-Xa could not pull down these proteins even though their abundances were much higher than that of JAKs and IKKs. We F I G U R E 4 Xanthatin covalently modified Cys243 of JAK2 and Cys412, Cys464 of IKKβ. A, Structure of Alk-Xa. B, Effects of Alk-Xa on IL-6-induced luciferase activity. HepG2/STAT3 cells were pre-treated with Alk-Xa for 1 h and luciferase activity was measured following stimulation of IL-6 (10 ng/mL) for 4 h. (n = 3, ***P < 0.001 compared with IL-6-induced cells) C, Effects of Alk-Xa on IL-6-induced JAK2/ STAT3 phosphorylation. MDA-MB-231 cells were treated with Alk-Xa for 1 h before stimulation by IL-6 (10 ng/mL) for 10 min. Cell lysates were processed for Western blot analysis. D, In-gel fluorescence scanning of Alk-Xa-treated JAK2 and IKKβ protein. Purified JAK2 and IKKβ proteins were incubated with Alk-Xa for 30 min and subsequently subjected to click chemistry with rhodamine-alkyne as described in Materials and Methods. In-gel fluorescence scanning was showed. E, Xanthatin covalently bound to Cys243 of JAK2. MDA-MB-231 cells were incubated with 20 μmol/L xanthatin for 1 h. After that, JAK2 protein was immunoprecipitated from cell lysate and subjected to LC-MS/

CO N FLI C T O F I NTE R E S T
No potential conflicts of interest are disclosed.