Capture and Identification of Dual Specificity Mitogen-Activated Protein Kinase Kinase 7 as a Direct Proteomic Target of Berberine to Affect the c-JunN-Terminal Kinase Pathway

Abstract

Open AccessCCS ChemistryCOMMUNICATION1 May 2022Capture and Identification of Dual Specificity Mitogen-Activated Protein Kinase Kinase 7 as a Direct Proteomic Target of Berberine to Affect the c-JunN-Terminal Kinase Pathway Qing-Xuan Zeng†, Wei Wei†, Tian-Yun Fan, Hong-Bin Deng, Xi-Xi Guo, Li-Ping Zhao, Xin-Tong Zhang, Sheng Tang, Jian-Dong Jiang, Ying-Hong Li, Yan-Xiang Wang and Dan-Qing Song Qing-Xuan Zeng† Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 †Q.-X. Zeng and W. Wei contributed equally to this work.Google Scholar More articles by this author , Wei Wei† Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 †Q.-X. Zeng and W. Wei contributed equally to this work.Google Scholar More articles by this author , Tian-Yun Fan Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 Google Scholar More articles by this author , Hong-Bin Deng Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 Google Scholar More articles by this author , Xi-Xi Guo Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 Google Scholar More articles by this author , Li-Ping Zhao Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 Google Scholar More articles by this author , Xin-Tong Zhang Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 Google Scholar More articles by this author , Sheng Tang Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 Google Scholar More articles by this author , Jian-Dong Jiang Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 Google Scholar More articles by this author , Ying-Hong Li *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 Google Scholar More articles by this author , Yan-Xiang Wang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 Google Scholar More articles by this author and Dan-Qing Song *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Institute of Medicinal Biotechnology, Chinese Academy of Medical Sciences & Peking Union Medical College, Beijing 100050 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202100986 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Seven photoaffinity-based and sixteen biotin-based berberine (BBR) probes were constructed and screened for their effects on c-Jun N-terminal protein kinases (JNK) phosphorylation (p-JNK) suppression at the cellular level. Taking active-photoaffinity probe 7c as a chemical tool, we first identified mitogen-activated protein kinase 7 (MAP2K7), an upstream protein on the JNK/stress activated protein kinase (SAPK) pathway, as a direct proteomic target of BBR using activity-based protein profiling (ABPP) and other chemical proteomic techniques. Furthermore, BBR’s inhibitory effect on p-JNK was significantly attenuated in both the MAP2K7-knockdown and models, indicating a MAP2K7-dependent inhibition on the JNK signaling pathway. For the first time, we demonstrate the unique mechanism of BBR that directly targets MAP2K7 to inhibit p-JNK rather than JNK activity with the advantages of multiple activities and a good safety profile. Download figure Download PowerPoint Introduction Berberine (BBR; Figure 1), a Chinese natural product extracted from Coptis chinensis, has been widely used as an over-the-counter (OTC) drug to treat bacterial-caused diarrhea in China for decades.1 Over the past 20 years, its multiple pharmacological applications have been discovered and elucidated, including antiinflammatory, antitumor, hypoglycemic, hypolipidemic, antiviral, and so on,2–11 suggesting the typical multitarget-directed mechanism of BBR. A few classical cellular pathways, such as the NIMA-related kinase 7 (NEK7)/nucleotide-binding domain-like receptor protein 3 (NLRP3) and the c-Jun N-terminal protein kinases (JNK)/stress activated protein kinases (SAPKs), as well as the interleukin-6 (IL-6)/signal transducer and activator of transcription 3 (STATs),8,12–27 have been proven to closely account for BBR’s polypharmacological effects. Furthermore, several direct proteomic targets of BBR, such as NEK7, E3 ubiquitin-protein ligase1 (UHRF1), retinoic acid receptor RXR-alpha (RXRα), and ephrin-B2,18,24,25,28,29 have been identified to explain its synergistic effects.24,30,31 Figure 1 | Structure of BBR and construction of photoaffinity and biotinylated BBR probes. Download figure Download PowerPoint The JNK pathway is mainly activated by various damage-associated molecular patterns (DAMPs), including proinflammatory cytokines [tumor necrosis factor (TNF)-α and IL-1α]. The stimulation of the JNK pathway gives rise to a marked increase of the JNK phosphorylation (p-JNK) level, which leads to the progression of inflammatory, neurodegenerative, and cell proliferative diseases.12,15,25,26 It is worth noting that BBR could suppress p-JNK to alleviate inflammation-related symptoms, such as obesity and insulin resistance, and exert effects on tumorous and neurodegenerative diseases.32–34 However, the direct proteomic targets of BBR upon the JNK signaling pathway are still unknown, so it is necessary to elucidate BBR’s first target in JNK-related diseases so as to fully understand the multitarget pharmacological mechanism of BBR. In this study, several photoaffinity BBR probes containing a diazirin photo cross-linking tag, an alkynyl functional group, and a series of biotinylated BBR probes (Figure 1) were synthesized and screened for their inhibitory activity on p-JNK to obtain the active probes. Then, taking the active probe as a chemical tool, the direct proteomic target of BBR upon the JNK pathway was captured and identified using activity-based protein profiling (ABPP), the cellular thermal shift assay (CETSA), as well as other chemical proteomics techniques (Figure 1) to clarify how BBR affects the JNK pathway. Results and Discussion Chemistry Taking commercially available BBR or thalifendine ( 8, 10-hydroxy BBR) synthesized in our lab24 as the starting materials, we prepared all the target BBR probes, including 7 photoaffinity, as displayed in Schemes 1 and 2, respectively. Scheme 1 | Synthetic scheme for probes 7a–7c. Reagents and conditions: (a) K2CO3, CH3CN, room temperature (r.t.); (b) 195 °C, 30–40 mmHg, 60 min; (c) 3a–3c, K2CO3, CH3CN, reflux; (d) CH3OH/HCl; (e) EDCI, HOBT, Et3N, DMF. Download figure Download PowerPoint As shown in Scheme 1, the varied substituted alkyl alcohols ( 1a– 1d) were reacted with p-toluenesulfonyl chloride ( 2, TsCl) to obtain corresponding p-toluenesulfonates ( 3a– 3d) with higher reactivity in yields of 70–88%. The demethylation of BBR yielded the key intermediate 4 with a 9-hydroxyl,35,36 which went through a nucleophilic substitution reaction with 3a– 3c to achieve the intermediates 5a– 5c in reasonable yields of 55–71%. The tert-butoxycarbonyl (Boc) protection group in compounds 5a– 5c was removed in the presence of HCl to generate compounds 6a– 6c, which were respectively condensed with a photoaffinity fragment in the presence of 1-ethyl-3-(3-dimethylpropylamine)carbodiimide (EDCI) and N-hydroxy benzotrizole (HOBT), as well as triethylamine (Et3N) in dimethylformamide (DMF) to produce the target probes 7a– 7c ( Supporting Information Figures S2–S4) in yields of 40–53%. Scheme 2 | Synthetic scheme for probes 11a–11d. Reagents and conditions: (a) 3a–3d, K2CO3, CH3CN, reflux; (b) CH3OH/HCl; (c) EDCI, HOBT, Et3N, DMF. Download figure Download PowerPoint As depicted in Scheme 2, taking compound 824 with a 10-hydroxyl as the starting material, photoaffinity probes 11a– 11d ( Supporting Information Figure S5–S7) were gained with the overall yields of 23–33% following a similar three-step procedure, including nucleophilic substitution, deprotection, and condensation. Sixteen biotinylated BBR probes 12a– 12d, 13a– 13d, 14a– 14d and 15a– 15d with a biotin fragment attached on position 3, 9, or 10 of BBR were prepared using previously reported methods.20,21,24,35–37 Biological evaluations Inhibitory activity on p-JNK of BBR probes The screening model for p-JNK suppression stimulated by TNF-α was initially constructed to further confirm the regulatory activity of BBR on the p-JNK level. As shown in Figures 2a and 2b, two obvious p-JNK bands, as two main isoforms of JNK1 and JNK2, were detected in the immunoblot assay, which had broad tissue distribution and played a potential role in insulin resistance, inflammation, and cell signaling.17,38 BBR could dose-dependently inhibit both p-JNK1 and p-JNK2 activity at concentrations ranging from 2.5 to 20 μM in Hela and HEK-293 cells. Since BBR exhibited comparable potencies on p-JNK suppression at concentrations of over 10 μM in HEK-293 cell lines, in the next step, the p-JNK inhibitory activity of BBR probes were evaluated in HEK-293 cell lines at the concentration of 10 μM. Figure 2 | p-JNK activity of BBR at different concentrations in Hela (a) and HEK-293 (b) cells. Download figure Download PowerPoint Taking BBR as the positive control, the p-JNK regulating activities of all the aimed BBR probes at the concentration of 10 μM were measured in HEK-293 cells with an immunoblot assay, as shown in Figures 3a–3c. Their chemical structures are shown in Table 1. According to the grey-scanning results, the inhibition ratios were calculated based on the distinct p-JNK1 bands to further analyze the structure–activity relationship (SAR) of the BBR probes, as listed in Table 1. Among the three photoaffinity probes 7a– 7c bearing different linkers on the 9-substituted tag, 7c (Figure 3e) with a polyethylene glycol linker displayed the highest suppressive activity on p-JNK with an inhibition ratio of 47%, superior to that of BBR (19%). All the probes 11a– 11d bearing a tag at the 10-position displayed moderate potencies with inhibition ratios ranging from 13% to 35%. Figure 3 | p-JNK regulatory activity of BBR probes in HEK-293 cells. (a) p-JNK regulatory activity of photoaffinity probes. (b and c) of biotinylated probes. (d) Dose-dependent inhibitory effect on p-JNK of probe 7c stimulated by TNF-α. (e and f) Structures of 7c and 13d. Download figure Download PowerPoint Table 1 | Structures of Target Probes Compound L R Inhibition Ratio (%)a Compound L R Inhibition Ratio (%) 7a (CH2)3 PTb −42 13b (CH2)5 BT 2 7b (CH2)5 PT −24 13c (CH2)2O(CH2)2 BT 5 7c (CH2)2O(CH2)2 PT 47 13d ((CH2)2O)2(CH2)2 BT 35 11a (CH2)3 PT 35 14a (CH2)3 BT −14 11b (CH2)5 PT 13 14b (CH2)5 BT 5 11c (CH2)2O(CH2)2 PT 29 14c (CH2)2O(CH2)2 BT 7 11d ((CH2)2O)2(CH2)2 PT 17 14d ((CH2)2O)2(CH2)2 BT 7 12a (CH2)3 BTc 3 15a (CH2)3 BT 21 aAll the inhibition ratios were calculated as the average values of two independent experiments, based on gray scanning analysis of two bands of p-JNK, as shown in Figure 1. bPhotoaffinity tag. cBiotin tag. Among the biotinylated probes, 12a– 12d and 13a– 13d bearing different linkers at the 9- or 10-position were then evaluated, and the presence of a longer polyethylene glycol linker might be beneficial for the activity. Compound 13d (Figure 3f) with a polyethylene glycol moiety exhibited the most potent inhibitory activity with an inhibition ratio of 35%, higher than that of BBR. While the biotin fragment was attached on position 3, the inhibitory activities of corresponding compounds 14a– 14d decreased dramatically. Then, the methylenedioxy ring was opened, and the obtained O-BBR probes 15a– 15d with a 2,3-dimethoxy moiety displayed comparable or declined activities. The most potent photoaffinity probe 7c was chosen for the next investigation. Furthermore, as depicted in Figure 3d, active photoaffinity probe 7c could dose-dependently decrease the p-JNK level at the concentrations ranging from 5 to 20 μM. Thus, 7c should possess a similar mechanism of action with BBR and was then selected as a chemical tool for further proteomic target fishing. Fishing and identification of direct targets of BBR in HEK-293 cells The overall process of target capture and identification using ABPP assay is presented in Figure 4a. Active probe 7c (10 μM) was incubated with HEK-293 cells to form a 7c-protein complex with the target protein through noncovalent binding. Followed by 365 nm light exposure, the highly active carbene fragment was generated and promptly formed a covalent bond with the adjacent hydroxyl of target protein in situ so as to ensure the stability of the formed complex. Through a click reaction of the alkyne reporter group in the 7c-protein complex with Cy3-azide, the Cy3-labeled complex was obtained for identification of fluorescent differential bands. Taking dimenthyl sulfoxide (DMSO)-treatment as the blank control, the Cy3-labeled proteome was separated by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE), and one distinct band with the molecular weight (MW) of around 50 kDa was detected by fluorescence imaging assay, which could be competitively inhibited by BBR, as depicted in Figure 4b. Figure 4 | BBR labels the direct proteomic target. (a) General workflow of the ABPP approach. (b) Cy3-labeled target proteins were identified using fluorescent gel imaging. Coomassie-blue staining shows protein loading. (c) The MAP2K7 protein pulled down by 7c was immunoblotted and competitively inhibited by BBR. (d and e) The thermal stability of the MAP2K7 (d) and GAPDH (e) protein with or without BBR. Download figure Download PowerPoint Similarly, after a click reaction of the alkyne group with biotin-azide, biotin-labeled complex was formed for further purification, enrichment, and identification through liquid chromatography tandem mass spectrometry (LC-MS/MS) analysis. The biotin-labeled proteome around 50 kDa was analyzed by LC-MS/MS analysis, and 41 proteins were identified, from which 9 were enzymes ( Supporting Information Table S1 and Figure S8), and the rest belonged to other categories, such as spliceosomes or ribosomal proteins. Among the 9 enzymes, an inflammation-related protein named mitogen-activated protein kinase 7 (MAP2K7), an upstream protein of the JNK pathway, was investigated in the first priority. As expected, MAP2K7 was successfully pulled down with its specific antibody by immunoblot assay, implying a possible direct interaction between BBR and MAP2K7 as displayed in Figure 4c. The apparent competitively inhibitory effect of BBR on the interaction between 7c and MAP2K7 was also observed, as shown in Figure 4c. Then, CETSA was conducted in HEK-293 cells to further confirm MAP2K7 as a direct target of BBR. Taking DMSO as the blank control, as displayed in Figure 4d, the thermal stability of the MAP2K7 protein was markedly weakened when the temperature exceeded 40 °C, while the stability of the BBR and MAP2K7 complex was significantly enhanced in the 40–55 °C range after the addition of BBR. No obvious difference was observed in the internal reference glyceraldehyde 3-phosphate dehydrogenase (GAPDH) group with or without BBR (Figure 4e). This result indicated that BBR might serve as a possible substrate of MAP2K7 and enhance the thermostability of MAP2K7, suggesting again a direct interaction between BRR and the MAP2K7 protein. Verification of MAP2K7 as a direct target protein of BBR To further verify the direct interaction between BBR and MAP2K7, pull-down experiments were carried out, as shown in Figure 5a. The recombinant MAP2K7 protein was not fished out in the absence of UV (365 nm) exposure or 7c alone, owing to lack of the corresponding covalent bond between 7c and MAP2K7. After UV exposure, a covalent bond between 7c and MAP2K7 was allowed to form, and the recombinant MAP2K7 protein was successfully pulled down by 7c. Followed by the addition of BBR, the pulled-down band was weakened due to a competitive inhibition of BBR. Meanwhile, the band of 7c-MAP2K7 complex almost vanished after pre-treatment under the condition of 95 °C, which indicated that the active labeling of 7c bonded with MAP2K7 only in the native folded status, rather in the heat-treated unfolded status. Figure 5 | Target protein analysis and verification. (a) The recombinant MAP2K7 protein pulled down by 7c were competitively inhibited by BBR. (b) The recombinant MAP2K7 protein pulled down by 13d (10 μL 13d beads) were competitively inhibited by BBR. (c) SPR sensorgrams obtained on an MAP2K7-coated chip at different concentrations of BBR. (d and e) Molecular docking results for MAP2K7 (PDB code: 6QHO) with BBR. Download figure Download PowerPoint Next, the active biotinylated BBR probe 13d was chosen as another chemical tool to carry out the target verification. As shown in Figure 5b, with the increase of BBR concentration ranging from 5 to 40 μM, the interaction between 13d and recombinant MAP2K7 was gradually weakened due to the competitive inhibition of BBR. Furthermore, BBR also dose-dependently bonded with immobilized MAP2K7 in surface plasmon resonance (SPR) analysis, with a Kd value of 34.8 μM (Figure 5c), hinting at the presence of a moderate interaction between BBR and MAP2K7 via physical bonding. Thus, molecular docking was carried out using Discovery Studio 4.5 software (Waltham, MA, USA). As shown in Figures 5d and 5e, BBR fit well in the active pockets of MAP2K7 (PDB code: 6QHO) with an ideal docking score of 91.3. Three hydrogen bonds between oxygen atoms on 9-methoxy, 10-methoxy, or methylenedioxy, and T217, C218, or L165 residues were formed. Also, molecular docking between BBR and MAP2K4 (PDB code: 3ALN, Supporting Information Figure S1), an upstream modulator of p-JNK to phosphorylate the p38 MAPKs,39 was conducted with a docking score of 76.5, indicating a weaker interaction between BBR and MAP2K4. Therefore, a specific and selective interaction between BBR and MAP2K7 existed, and MAP2K7 as a direct proteomic target of BBR was identified at both the biochemical and cellular levels. Anti-p-JNK effects of BBR in MAP2K7 overexpression and knockdown models Finally, BBR’s inhibitory effects on the p-JNK level were investigated on both MAP2K7 overexpression and knockdown models in HEK-293 cells. As displayed in Figure 6, BBR could significantly inhibit p-JNK induced by TNF-α, while overexpression of MAP2K7 reversed the action of BBR (Figure 6a), and the effect was lost when MAP2K7 was silenced by siRNA (Figure 6b). This result further proved that BBR exerted the p-JNK regulation effect on the JNK signaling pathway in a MAP2K7-dependent manner. Figure 6 | BBR regulates p-JNK by directly targeting MAP2K7. p-JNK regulating effects of BBR while MAP2K7 by transfecting HA-MAP2K7 (a), or silenced by transfecting siRNA targeting MAP2K7 (b). Download figure Download PowerPoint Most of the JNK inhibitors target the highly conserved adenosine 5′-triphosphate (ATP)-binding pocket within the kinome, and thus the poor selectivity might lead to side effects, which might not be tolerable for chronic diseases. Targeting kinases upstream of JNKs might be one appropriate strategy to selectively modulate JNK activity, while still allowing for some JNK basal activity to avoid the potential systemic toxicity.39,40 Some potent MAP2K7 inhibitors have been discovered and proven to selectively regulate the JNK pathway via modulating p-JNK.41–43 We first demonstrate that BBR directly acts on MAP2K7 by suppressing p-JNK rather than directly targeting JNK to display multiple effects with a good safety profile. Our previous studies demonstrate that BBR directly targets the NEK7 protein and specifically blocks the NEK7–NLRP3 interaction and successively inhibits IL-1β release and its related inflammatory diseases.24 MAP2K7 is another direct target of BBR that exerts various effects related to the JNK pathway. As depicted in Figure 7, these results enable us to further understand the BBR’s multitarget synergy effects on inflammation-related diseases at the intersection of JNK and NLRP3 pathways, such as obesity and insulin resistance. Therefore, this study provides forceful evidence that MAP2K7 can be regarded as a druggable target for the development of selective JNK pathway modulators with the advantage of lower side effects. Verification of other direct targets of BBR with other effects is actively ongoing in our lab. Figure 7 | BBR’s network mechanisms. Download figure Download PowerPoint Conclusion Two series of BBR probes were constructed and screened for their p-JNK suppressing effects in Hela and HEK-293 cells, and the most potent probe 7c was chosen as a chemical tool for target capture and identification. Using ABPP and CETSA and other techniques, MAP2K7, an upstream protein of the JNK signaling pathway, was first identified to be a direct target of BBR. The effect on p-JNK suppression of BBR was weakened in both the MAP2K7 overexpression and knockdown models. These results indicate that BBR exerts multiple biological activities related with the JNK pathway owing to its direct targeting of the MAP2K7 protein. Therefore, in this study, we first demonstrate the unique mechanism of BBR by directly targeting MAP2K7 to inhibit p-JNK rather than JNK activity, and propose MAP2K7 as a druggable target for the development of selective JNK pathway modulators. With the advantage of lower side effects, BBR has potential applications for drug screening of multiple conditions related to the JNK pathway, such as obesity, insulin resistance, and neurodegenerative diseases. Supporting Information Supporting Information is available and includes experimental procedure characterization data, NMR spectra of compounds, and molecular docking files for compound 7c and MAP2K4. Conflict of Interest There is no conflict of interest to report. Acknowledgments This work was supported by the CAMS Innovation Fund for Medical Sciences (nos. 2020-I2M-2-010 and 2016-I2M-1-011), the Drug Innovation Major Project (no. 2018ZX09711-001), and the National Natural Science Foundation of China (no. 81974494).