Small molecule inhibitors of IκB kinase β: A chip-based screening and molecular docking simulation
Yong Wan Choa,b,c, Hye Jin Limb, Moon Hi Hanb, Byung-Chul Kima, Sanghwa Hana,⁎
a Department of Biochemistry, Kangwon National University, Chuncheon 24341, Republic of Korea
b Proteogen Inc., Seoul, Republic of Korea
c Shebah Biotech Inc., G-Tech Village, Chuncheon, Republic of Korea

A R T I C L E I N F O

Keywords:
Inhibitor IκB kinase β IKKβ
Screening
Docking simulation
A B S T R A C T

A chip-based screening system for IκB kinase β (IKKβ) has been developed by physically immobilizing the substrate IκBα on a glass matrix using a calixarene linker. Phosphorylation of IκBα by IKKβ and ATP was quantitated using a fluorescently labeled antibody. Using this efficient assay system a chemical library of 2000 bioactive compounds was screened against IKKβ and four were identified as good inhibitors, namely, aurin- tricarboxylic acid, diosmin, ellagic acid, and hematein. None of them have been reported to be an inhibitor of IKKβ although they were implicated in various NFκB-mediated biological processes. Our enzyme-based assay showed that IC50 of the four inhibitors is comparable with that of IKK-16, a previously known strong inhibitor. Molecular docking simulation shows that the hydrophobic moiety of an inhibitor interacts with the four hy- drophobic residues (Leu21, Val29, Val152, and Ile165) of the active site. The MM-PBSA calculation suggests that these hydrophobic interactions appear to be the predominant contributor to the binding free energy. As IKKβ is ubiquitously expressed in various cell types and executes many biological functions, the enzyme and cell spe- cificity of the four inhibitors need to be rigorously tested before accepted as a drug candidate.

Nuclear factor κB (NF-κB) is localized in the cytosol when com- plexed with IκBα, the inhibitor of NF-κB. Pro-inflammatory stimuli phosphorylate the IKKβ subunit of IκBα kinase (IKK) complex leading to its activation. Activated IKK now phosphorylates IκBα causing its degradation and subsequent release of NF-κB that is translocated to the nucleus to function as a transcription factor.1 IKKβ plays an important role in ischemia-induced brain damage.2 It also is a regulator of mRNA stability3 and serves as a therapeutic target for inflammatory diseases and cancer.4,5 Therefore much effort has been devoted to discovery of small molecules that inhibit IKKβ.6–10 Many inhibitors with a small IC50 value are available to date but none have passed the clinical trials.11 Search for new inhibitors of IKKβ is still in progress.
In this study we have screened a large chemical library against IKKβ using a chip-based assay system12 that allows random immobilization of the protein substrate (IκBα) on a glass chip. This technique has been successfully applied to screening of inhibitors of other kinases.13,14 IC50 of the newly found inhibitors were measured on the enzyme level. Structures of the enzyme-inhibitor complexes were modeled by mole- cular docking simulation. Binding free energy and inhibition constant (Ki) were also calculated for each inhibitor.
Procedure for the chip-based assay is described briefly here for IKKβ
(see Fig. 1a). A protein substrate, IκBα (5~10 ng/μL, 1 μL) in a buffer (10% PEG-200 in PBS, pH 7.4) was spotted in each well (1.5 mm dia- meter) formed on the glass plate coated with ProLinker™. It was in- cubated overnight at 4 °C in a humidity chamber and then rinsed with water and with PBST (PBS containing 0.5% Tween-20). The chip was blocked with a PBS buffer containing 1% BSA for 1 h at room tem- perature, rinsed extensively, and dried with nitrogen gas. An inhibitor solution (0.5 μL of a desired concentration) was added to the well and then mixed with 0.5 μL of a kinase assay buffer containing an IKKβ (2 ng/μL) in a mixture of ADBI and ATP/Mg buffer. The kinase reaction was allowed to proceed for 1 h. The chip was then washed and dried again and incubated for 30 min with 1 μL of an IκBα-specific anti- phospho Ab that was diluted to 1:100 with 1% BSA and 10% PEG-200 in PBS. After washing and drying, a Cy5-labeled Ab (diluted to 1:100) was added to each well and allowed to bind to the phosphorylated substrate for 30 min. After rinsing and drying the Cy5 fluorescence was read using GenePix 4000B and 4300A scanners (Molecular Devices, USA). The scanned images were analyzed using a GenePix Pro 6.0 (Molecular Devices, USA). Screening was carried out against a chemical library of 2000 compounds on an IKKβ assay chip with 96 wells (1.5 mm diameter). Concentration of an inhibitor for primary screening

⁎ Corresponding author.
E-mail address: [email protected] (S. Han).

https://doi.org/10.1016/j.bmc.2020.115440

Received 22 October 2019; Received in revised form 7 March 2020; Accepted 10 March 2020
0968-0896/©2020ElsevierLtd.Allrightsreserved.

Fig. 1. Preparation of an IKKβ assay chip. (a) A cover slip was treated with ProLinker™ to which a protein substrate IκBα was bound. The immobilized IκBα was phosphorylated by IKKβ in the presence of ATP. Anti-phospho specific antibody was attached to the phosphorylated substrate. Finally the antibody was bound to a fluorescent Cy5-labeled secondary antibody whose fluorescence was measured to quantitate phosphorylation. (b) A rainbow color display of fluorescence intensity of the Cy5-labeled antibody as a function of inhibitor concentration. Data were triplicated for each inhibitor.

was 50 μM. Active compounds were selected and screened again in a dose-dependent manner.
Ligand K252a (named KSA in 4KIK) was removed from the chain B of 4KIK. Hydrogen atoms were added to the inhibitors and the protein. AutoDockTools (ADT) was used to obtain the pdbqt (pdb with charge and torsion) file of the protein and to set the center and size of the grid box. The pdbqt of a ligand was generated by Open Babel.15 Docking simulation was performed using Smina,16 a fork of AutoDock Vina, treating the protein as rigid with an exhaustiveness of 16. Inhibition constant Ki was calculated using the relation ΔG = -RTlnKi. Contribu- tion of each residue to the binding free energy was calculated using the molecular mechanics Poisson-Boltzmann surface area (MM-PBSA) method.17 Structures were visualized with UCSF Chimera.18
In general a chip-based assay system requires immobilization of a substrate (or enzyme) on a matrix via chemical or physical binding. Chemical immobilization involves formation of chemical bonds be- tween reactive residues of a protein and a functional group of deriva- tized glass surface. Such covalent binding is robust but sometimes causes conformational changes in the bound protein that may lead to inactivation. Physical binding on the other hand is weaker than che- mical binding and therefore is less likely to suffer from inactivation caused by a large conformation change. However, the substrate can be lost by desorption from the surface. The system used in this study provides a strong physical binding via a host-guest interaction without loss of activity. Furthermore the substrate molecules are attached with a random orientation as a protein often possesses multiple lysine re- sidues. This guarantees that at least some molecules are available for the incoming enzyme.
In search of small molecule inhibitors of IKKβ, we screened a large chemical library of bioactive and natural compounds using a chip-based assay system (Fig. 1a). The protein substrate, IκBα, is immobilized by a calixarene linker (ProLinker™) via a host–guest chemistry, i.e., the protonated amino group of a lysine residue of the substrate is
accommodated by the crown moiety of a calixarene (see Fig. S1). IKKβ and ATP phosphorylate the immobilized IκBα to which a phospho- specific antibody is bound. Finally the antibody was bound to a Cy5- labeled secondary antibody and the level of phosphorylation can be quantitated by the intensity of fluorescence. Fig. 1b is a rainbow color display of fluorescence intensity which is proportional to the amount of phosphorylated substrate.
Using the method described above, we screened 2000 compounds against IKKβ to identify four good inhibitors, namely, aurin- tricarboxylic acid, diosmin, ellagic acid, and hematein. Their structures are shown in Fig. 2 along with the two previously known inhibitors, K252a and IKK-16. Quantitation of the fluorescence intensity of Fig. 1b yielded IC50 values as summarized in Table 1. They are in a few μM range which appears to be much larger than the reported values for the previously known inhibitors that have IC50 in nM range.11 It is noted, however, that our assay system is different from the conventional assay method: the substrate is confined in 2D surface rather than in 3D space and the conditions (concentrations, etc) under which IC50 is measured are different. So we measured IC50 of a known inhibitor IKK-16 (also known as Inhibitor VII) using our chip assay method to find a value of
6.46 μM, which is comparable with the four inhibitors of the present study. A reported IC50 of IKK-16 is 40 nM19 suggesting that our four inhibitors would have IC50 in nM range if measured using a conven- tional method.
Binding modes of the inhibitors were predicted by molecular docking simulation. For human IKKβ only two structures, a ligand-free (4E3C)20 and ligand-bound (4KIK)21 enzyme, are available to be used as a target for ligand docking. Examination of the two structures revealed that ligand binding induces a large conformational change around the ligand binding site. The ligand-free enzyme (4E3C) has a loop that prevents a ligand from accessing the ligand binding site (Fig. S2, arrow). Docking simulation using Smina,16 a version of AutoDock Vina, showed that none of the inhibitors were bound to the active site when

Fig. 2. Structures of the inhibitors.

Table 1
Experimental IC50, theoretical free energy of binding, and theoretical inhibition constant for the screened and known inhibitors.

Inhibitors IC50 (μM)a ΔG (kJ/mol) Ki (nM)b
aurintricarboxylic acid 5.40 −48.03 ± 0.04 4.34

diosmin 3.89 −52.40 ± 0.29 0.753
ellagic acid 1.98 −44.76 ± 0.01 16.1
hematein 2.18 −44.65 ± 0.04 16.8
IKK-16 6.46 −44.52 ± 0.06 17.7
K252a – −60.96 ± 0.08 0.0243

a Estimated by the chip assay method of this work.
b Calculated according to the equation, ΔG = -RTlnKi.

4E3C was used as a target protein (not shown). Therefore the kinase domain of the ligand-bound structure 4KIK was used in docking simu- lation.
We first tested the performance of Smina using an experimentally determined structure of an IKKβ-K252a complex.21 K252a, an analog of staurosporin, is a strong inhibitor that binds to the ATP-binding site of IKKβ. A grid box large enough to encompass the whole IKKβ molecule was used to ensure an unbiased blind docking although the binding site of K252a is known. Out of ten independent runs, eight produced a structure with the binding free energy (ΔGbind) of −60.96 ± 0.08 kJ/ mol and the other two −46.22 kJ/mol. In Fig. S3 the conformation of K252A with the lower ΔGbind is compared with the experimentally determined structure of 4KIK. It is clear that docking simulation using Smina accurately predicted the structure of the bound K252a. The RMSD of the experimental and docked structures was only 0.043 nm. Structures of the other two having the higher binding free energy differ only in the methyl ester group, which was flipped over to the other side. Using Smina the inhibitors were subject to docking simulation and
the result is shown in Fig. 3. All inhibitors were docked to the ATP- binding site21 although the blind docking protocol allowed the ligands to bind anywhere on and within the enzyme. Ten runs of docking

Fig. 3. (a) Structures of aurintricarboxylic acid (orange), diosmin (blue), ellagic acid (green), and hematein (yellow) docked to IKKβ. (b) Hydrophobicity sur- face showing that the inhibitors are bound to a hydrophobic pocket. Hydrophobic and hydrophilic surfaces are colored red and blue, respectively.

simulation for each inhibitor produced almost identical conformations for all the inhibitors except diosmin (see below). The binding pocket is hydrophobic primarily due to Leu21, Val29, Val152, and Ile165. Hy- drophobic moiety of the inhibitors is accommodated in the pocket while the hydrophilic moiety protrudes from the active site out to the protein surface.
Fig. 4 shows hydrogen bonding and hydrophobic interactions of the inhibitors with IKKβ. K252a is fit in the binding pocket by the four hydrophobic residues and four hydrogen bonds (Fig. 4a). Leu21 and Val29 are positioned above the flat K252a molecule whereas Val152 and Ile165 interact from below. Polar groups (NeH, C]O, and OeH) participate in hydrogen bonding with the nearby residues. K252a is an inhibitor derived from staurosporin22 and the only ligand complexed with IKKβ whose X-ray structure has been determined.21
Aurintricarboxylic acid is a compound with multiple bioactivities. Interestingly it was suggested as an inhibitor of IKK in LPS-mediated cell death of macrophages.23 In aqueous media it is deprotonated car- rying a negative charge. However, the fully protonated neutral form was used in docking simulation as the binding pocket of IKKβ is hy- drophobic. Ten runs of simulation produced only one conformation

Fig. 4. Hydrogen bonding and hydrophobic interactions of the inhibitors and IKKβ. (a) K252a, (b) aurintricarboxylic acid, (c) diosmin, (d) ellagic acid, (e) hematein, and (f) IKK-16. Side chains of the four hydrophobic residues in the ligand binding pocket are shown in green. Blue dotted lines represent hydrogen bonds.

with ΔGbind of −48.03 ± 0.04 kJ/mol (Ki of 4.34 nM). The peripheral carboxyl groups form hydrogen bonds with Cys99 and Asn150. In the docked state, two of the three rings are more or less coplanar and the four hydrophobic residues are located above and below these two hy- drophobic rings (Fig. 4b). Structure of aurintriboxylic acid alone or in complex with a protein has not been determined experimentally. We therefore performed a quantum mechanical geometry optimization at the level of HF/6-31G(g) to find that the rings are not coplanar to each other (not shown). This means that the interaction of the bound ligand with the protein should compensate for the unfavorable torsional en- ergy due to twisting the rings.
Diosmin is a glycoside of a flavone with pharmacological activities24
but its action on IKKβ has not been documented. The glycosidic moiety is very hydrophilic due to six hydroxyl groups. When bound to the enzyme the glycosidic moeity sticks out to the surface while the more hydrophobic flavone moiety is fitted in the binding pocket. Strong in- teractions of the hydroxyl groups with the polar residues allowed two binding modes. Ten runs of simulation produced five conformations with ΔGbind of −52.40 ± 0.29 kJ/mol (Ki of 0.753 nM) and the other five with −50.58 ± 0.18 kJ/mol. A large standard deviation was observed only for the long glycosidic group which forms multiple hy- drogen bonds with the protein polar groups. Nevertheless the planar flavone moiety is sandwiched by the four hydrophobic residues per- mitting much less variation than the hydrophilic moiety (Fig. 4c).
Ellagic acid is a natural antioxidant with pharmacological effects on many diseases.25 It is known to inhibit NF-kB signaling although direct inhibition of IKKβ has not been reported.26,27 Docking simulation shows that the dilactone of four rings interact with the four hydro- phobic residues of the binding pocket yielding ΔGbind of
−44.76 ± 0.01 kJ/mol (Ki of 16.1 nM). This planar molecule has four hydroxyl groups, one of which forms a hydrogen bond with the car- bonyl oxygen of Glu97 (Fig. 4d). The hydroxyl groups stay away from
the four hydrophobic residues as expected.
Hematein is a planar molecule derived from hematoxylin, a dye extracted from the logwood tree. It is primarily used as a histochemical stain. Docking simulation predicted a unique structure (Fig. 4e) for the bound hematein and shows that hematein is a strong inhibitor of IKKβ with ΔGbind of −44.65 ± 0.04 kJ/mol (Ki of 16.8 nM). Again the hydrophobic rings are sandwiched by the four hydrophobic residues and the carbonyl group of hematein is hydrogen bonded to the NeH of Cys99. Biological activity of hematein has been reported only for a few cases. It is an inhibitor of casein kinase II28 and prion replication.29 In relation to our finding, a special attention needs to be paid to the report on inhibition of NFκB activation by hematein.30 Based on the finding of this work, it is tempting to propose that the inhibition of IKKβ by he- matein is responsible for the inactive NFκB. However, such an inter- pretation has not been given in their report.30
IKK-16 is known as a selective inhibitor of IKKβ with IC50 of 40 nM.19 However, the structure of IKK-16 in a complex with IKKβ has not been determined experimentally. Docking simulation yielded the structure shown in Fig. 4f. Hydrophobic moiety of the long IKK-16 in- teracts with only Val29 and Ile165. The other two, Leu21 and Val152, do not contribute significantly to the hydrophobic interactions. More hydrophilic rings of IKK-16 extend out to the surface through a hy- drophilic crevice. Calculated binding free energy was
−44.52 ± 0.06 kJ/mol (Ki of 17.7 nM).
Finally contribution of each residue to the binding free energy was estimated by the MM-PBSA method17 which is implemented in Gromacs.31 Table S1 shows that the four hydrophobic residues of the binding pocket are among the top five contributors for all the inhibitors except IKK-16, for which only two (Val29 and Ile165) are in the list. This is consistent with the hydrophobic interactions discussed above. It appears that the interaction between the hydrophobic moiety of an inhibitor and the hydrophobic pocket of the enzyme is prerequisite for a

strong inhibition.
In conclusion, we used an efficient chip-based assay system to screen 2000 bioactive compounds against IKKβ and identified four in- hibitors, namely aurintricarboxylic acid, diosmin, ellagic acid, and hematein. None of them have previously been reported as an inhibitor
of IKKβ although they were implicated in many biological processes

Zheng J, Park MH, Lee HP, et al. A small molecule, (E)-2-metho⦁ x⦁ y-4-(3-(4-metho⦁ x⦁ - ⦁ yphenyl)⦁ ⦁ prop-1-en-1-yl)⦁ ⦁ phenol⦁ ⦁ suppresses⦁ ⦁ tumor⦁ ⦁ growth⦁ ⦁ via⦁ ⦁ inhibition⦁ ⦁ of⦁ ⦁ IkappaB
kinase β in colorectal cancer in vivo and in vitro. Oncotarget. 2017;8:91258–91269.
Xie X, Tu J, You H, Hu B. Design, synthesis, and biological evaluation of novel EF24 ⦁ and EF31 analogs as potential I⦁ κ⦁ B kinase β inhibitors for the treatment of pancreatic ⦁ cancer. ⦁ Drug Des Devel Ther. ⦁ 2017;11:1439–1451.
Choi⦁ ⦁ JH,⦁ ⦁ Park⦁ ⦁ SH,⦁ ⦁ Jung⦁ ⦁ JK,⦁ ⦁ et⦁ ⦁ al.⦁ ⦁ Caffeic⦁ ⦁ acid⦁ ⦁ cyclohexylamide⦁ ⦁ rescues⦁ ⦁ lethal⦁ ⦁ in-

involving NFκB. Our enzyme-based assay shows that IC50 of the four
flammation in septic mice through inhibition of IκB kinase in innate immune process.
Sci Rep. 2017;7:41180.

inhibitors is comparable with IC50 of the previously known IKK-16. Molecular docking simulation and MM-PBSA calculation suggest that the predominant contribution to the binding free energy is from the hydrophobic interactions between the hydrophobic moiety of an in- hibitor and the hydrophobic residues of the active site. As IKKβ is ubiquitously expressed in various cell types and executes multiple biological functions, the enzyme and cell specificity of the four in- hibitors need to be rigorously examined before accepted as a drug candidate.
Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influ- ence the work reported in this paper.
Acknowledgments
This research was supported by a Grant (M10415010004-05N1501- 00410) from Korea Biotech R&D Group of Next-generation growth engine project of the Ministry of Education, Science and Technology, Republic of Korea.
Appendix A. Supplementary material
Supplementary data to this article can be found online at https:// doi.org/10.1016/j.bmc.2020.115440.
References

Kari⦁ n M. How NF-kappaB is activated: the role of the IkappaB kinase (IKK)⦁ ⦁ complex.
Oncogene. 1999;18:6867–6874.
Herrmann O, Baumann B, de Lorenzi R, et al. IKK mediates ischemia-induced neu- ⦁ rona⦁ l death. ⦁ Nat Med.⦁ ⦁ 2005;11:1322–1329.
Mikuda N, Kolesnichenko M, Beaudette P, et al. The I⦁ κ⦁ B kinase complex is a regulator ⦁ of⦁ mRNA stability. ⦁ EMBO J.⦁ ⦁ 2018;37:e98658.
Llona-Minguez S, Baiget J, Mackay SP. Small-molecule inhibitors of I⦁ κ⦁ B kinase (IKK) ⦁ and IKK-related kinases. ⦁ Pharm Pat Anal.⦁ ⦁ 2013;2:481–498.
Cao H, Jiang S, Yuan R, et al. Inhibition of I⦁ κ⦁ B kinase 2 attenuated the proliferation ⦁ and⦁ induced apoptosis of gastric cancer.⦁ ⦁ Dig Dis Sci. ⦁ 2019;64:1204–1216.
Shin⦁ ⦁ Y,⦁ ⦁ Lim⦁ ⦁ SM,⦁ ⦁ Yan⦁ ⦁ HH,⦁ ⦁ et⦁ ⦁ al.⦁ ⦁ Optimization⦁ ⦁ and⦁ ⦁ biological⦁ ⦁ evaluation⦁ ⦁ of⦁ ⦁ amino-
pyrimidine-based IκB kinase β inhibitors with potent anti-inflammatory effects. Eur J Med Chem. 2016;123:544–556.
Zhang⦁ ⦁ Y,⦁ ⦁ Lapidus⦁ ⦁ RG,⦁ ⦁ Liu⦁ ⦁ P,⦁ ⦁ et⦁ ⦁ al.⦁ ⦁ Targeting⦁ ⦁ I⦁ κ⦁ B⦁ ⦁ kinase⦁ ⦁ β/NF-⦁ κ⦁ B⦁ ⦁ signaling⦁ ⦁ in⦁ ⦁ human
prostate cancer by a novel IκB kinase β inhibitor CmpdA. Mol Cancer Ther.
2016;15:1504–1514.
⦁ Prescott JA, Cook SJ. Targeting IKKβ in Cancer: Challenges and opportunities for the ⦁ therapeutic⦁ utilisation of IKKβ inhibitors. ⦁ Cells.⦁ ⦁ 2018;7:115.
Lee⦁ ⦁ Y,⦁ ⦁ Lee⦁ ⦁ EK,⦁ ⦁ Cho⦁ ⦁ YW,⦁ ⦁ et⦁ ⦁ al.⦁ ⦁ ProteoChip:⦁ ⦁ a⦁ ⦁ highly⦁ ⦁ sensitive⦁ ⦁ protein⦁ ⦁ microarray
prepared by a novel method of protein immobilization for application of protein- protein interaction studies. Proteomics. 2003;3:2289–2304.
Kim⦁ ⦁ EY,⦁ ⦁ Choi⦁ ⦁ YJ,⦁ ⦁ Park⦁ ⦁ CW,⦁ ⦁ Kang⦁ ⦁ IC.⦁ ⦁ Erkitinib,⦁ ⦁ a⦁ ⦁ novel⦁ ⦁ EGFR⦁ ⦁ tyrosine⦁ ⦁ kinase⦁ ⦁ inhibitor ⦁ screened using a ProteoChip system from a phytochemical library. ⦁ Biochem Biophys ⦁ Res Commun.⦁ ⦁ 2009;389:415–419.
Cho⦁ ⦁ YW,⦁ ⦁ Lim⦁ ⦁ HJ,⦁ ⦁ Han⦁ ⦁ MH,⦁ ⦁ Kim⦁ ⦁ B-C,⦁ ⦁ Han⦁ ⦁ S.⦁ ⦁ Inhibitors⦁ ⦁ of⦁ ⦁ aurora⦁ ⦁ kinases⦁ ⦁ screened⦁ ⦁ by⦁ ⦁ a
chip-based assay system. Bull Korean Chem Soc. 2019;40:1202–1207.
O⦁ ‘⦁ Boyle NM, Banck M, James CA, Morley C, Vandermeersch T, Hutchison GR. Open ⦁ Babel:⦁ ⦁ An⦁ ⦁ open⦁ ⦁ chemical⦁ ⦁ toolbo⦁ x⦁ .⦁ ⦁ J.⦁ ⦁ Cheminf.⦁ ⦁ 2011;3:33.
Koes DR, Baumgartner MP, Camacho CJ. Lessons learned in empirical scoring with ⦁ smina from the CSAR 2011 benchmarking exercise. ⦁ J Chem Inf Model. ⦁ 2013;53:1893–1904.
Kumari R, Kumar R. Open Source Drug Discovery Consortium, Lynn A. g_mmpbsa-a ⦁ GROMACS⦁ ⦁ tool⦁ ⦁ for⦁ ⦁ high-throughput⦁ ⦁ MM-PBSA⦁ ⦁ calculations.⦁ ⦁ J⦁ ⦁ Chem⦁ ⦁ Inf⦁ ⦁ Model.
2014;54:1951–1962.
Pettersen EF, Goddard TD, Huang CC, et al. UCSF Chimera-a visualization system for ⦁ explorator⦁ y research and analysis.⦁ ⦁ J Comput Chem. ⦁ 2004;25:1605–1612.
Waelchli R, Bollbuck B, Bruns C, et al. Design and preparation of⦁ ⦁ 2-benzamido-
pyrimidines as inhibitors of IKK. Bioorg Med Chem Lett. 2006;16:108–112.
Polley S, Huang DB, Hauenstein AV, et al. A structural basis for I⦁ κ⦁ B kinase 2 acti- ⦁ vation via oligomerization-dependent trans auto-phosphorylation. ⦁ PLoS Biol. ⦁ 2013;11:e1001581.
Liu S, Misquitta YR, Olland A, et al. Crystal structure of a human I⦁ κ⦁ B kinase β ⦁ asymmetric⦁ dimer. ⦁ J Biol Chem. ⦁ 2013;288:22758–22767.
Wisniewski D, LoGrasso P, Calaycay J, Marcy A. Assay for IkappaB kinases using an ⦁ i⦁ n vivo biotinylated IkappaB protein substrate. ⦁ Anal Biochem.⦁ ⦁ 1999;274:220–228.
Tsi CJ, Chao Y, Chen CW, Lin WW. Aurintricarbo⦁ x⦁ ylic acid protects against cell death ⦁ caused by lipopolysaccharide in macrophages by decreasing inducible⦁ ⦁ nitric-o⦁ x⦁ ide
synthase induction via IkappaB kinase, extracellular signal-regulated kinase, and p38
mitogen-activated protein kinase inhibition. Mol Pharmacol. 2002;62:90–101.
Patel K, Gadewar M, Tahilyani V, Patel DK. A review on pharmacological and ana- ⦁ lytical⦁ ⦁ aspects⦁ ⦁ of⦁ ⦁ diosmetin:⦁ ⦁ a⦁ ⦁ concise⦁ ⦁ report.⦁ ⦁ Chin⦁ ⦁ J⦁ ⦁ Integr⦁ ⦁ Med.⦁ ⦁ 2013;19:792–800.
Ríos JL, Giner RM, Marín M, Recio MC. A pharmacological update of ellagic⦁ ⦁ acid.
Planta Med. 2018;84:1068–1093.
Lee J, Won J, Choi J, et al. Protective effect of ellagic acid on concanavalin A induced ⦁ hepatitis via toll-like receptor and mitogen-activated protein kinase/nuclear factor ⦁ κ⦁ B signaling pathways. ⦁ J Agric Food Chem.⦁ ⦁ 2014;62:10110–10117.
Zhou⦁ ⦁ B,⦁ ⦁ Li⦁ ⦁ Q,⦁ ⦁ Wang⦁ ⦁ J,⦁ ⦁ Chen⦁ ⦁ P,⦁ ⦁ Jiang⦁ ⦁ S.⦁ ⦁ Ellagic⦁ ⦁ acid⦁ ⦁ attenuates⦁ ⦁ streptozocin⦁ ⦁ induced
diabetic nephropathy via the regulation of oxidative stress and inflammatory sig- naling. Food Chem Toxicol. 2019;123:16–27.
Hung⦁ ⦁ MS,⦁ ⦁ Xu⦁ ⦁ Z,⦁ ⦁ Chen⦁ ⦁ Y,⦁ ⦁ et⦁ ⦁ al.⦁ ⦁ Hematein,⦁ ⦁ a⦁ ⦁ casein⦁ ⦁ kinase⦁ ⦁ II⦁ ⦁ inhibitor,⦁ ⦁ inhibits⦁ ⦁ lung
cancer tumor growth in a murine xenograft model. Int J Oncol. 2013;43:1517–1522.
⦁ Biggi S, Pancher M, Stincardini C, et al. Identification of compounds inhibiting prion replication and toxicity by removing PrPC from the cell surface. J Neurochem. 2019. https://doi.org/10.1111/jnc.14805⦁ .
Choi JH, Jeong TS, Kim DY, et al. Hematein inhibits atherosclerosis by inhibition of ⦁ reactive o⦁ x⦁ ygen generation and NF-kappaB-dependent inflammatory mediators in ⦁ hyperlipidemi⦁ c⦁ ⦁ mice.⦁ ⦁ J⦁ ⦁ Cardiovasc⦁ ⦁ Pharmacol.⦁ ⦁ 2003;42:287–295.
Hess B, Kutzner C, van der Spoel D, Lindahl E. GROMACS 4: Algorithms for Highly ⦁ Efficient, Load-Balanced, and Scalable Molecular Simulation. ⦁ J Chem Theory Comput. ⦁ 2008;4:435–447.