Advertisement

The Anti-inflammatory Sesquiterpene Lactone Helenalin Inhibits the Transcription Factor NF-κB by Directly Targeting p65*

Open AccessPublished:December 11, 1998DOI:https://doi.org/10.1074/jbc.273.50.33508
      The sesquiterpene lactone helenalin is a potent anti-inflammatory drug whose molecular mechanism of action remains unclear despite numerous investigations. We have previously shown that helenalin and other sesquiterpene lactones selectively inhibit activation of the transcription factor NF-κB, a central mediator of the human immune response. These drugs must target a central step in NF-κB pathway, since they inhibit NF-κB induction by four different stimuli. It has previously been reported that sesquiterpene lactones exert their effect by inhibiting degradation of IκB, the inhibitory subunit of NF-κB. These data contradicted our report that IκB is not detectable in helenalin-treated, ocadaic acid-stimulated cells. Here we use confocal laser scanning microscopy to demonstrate the presence of IκB-released, nuclear NF-κB in helenalin-treated, tumor necrosis factor-α stimulated cells. These data show that neither IκB degradation nor NF-κB nuclear translocation are inhibited by helenalin. Rather, we provide evidence that helenalin selectively alkylates the p65 subunit of NF-κB. This sesquiterpene lactone is the first anti-inflammatory agent shown to exert its effect by directly modifying NF-κB.
      In traditional medicine, alcoholic preparations from flowers ofArnica montana and Arnica chamissonis ssp.foliosa are applied externally to treat hematomas, contusions, sprains, rheumatic diseases, and superficial inflammations of the skin. The secondary metabolites that mediate the anti-inflammatory effects are sesquiterpene lactones of the 10α-methylpseudoguaianolide type like helenalin, 11α,13-dihydrohelenalin, chamissonolid, and their ester derivatives (
      • Willuhn G.
      ,
      • Hall I.H.
      • Lee K.H.
      • Starnes C.O.
      • Sumida Y.
      • Wu R.Y.
      • Waddell T.G.
      • Cochran J.W.
      • Gerhart K.G.
      ,
      • Hall I.H.
      • Starnes C.O.
      • Lee K.-H.
      • Waddell T.G.
      ). Several studies have investigated how these natural compounds exert their anti-inflammatory effect. Sesquiterpene lactones have been shown to modulate many processes that influence inflammatory reactions, for example, oxidative phosphorylation, platelet aggregation, histamine and serotonin release (
      • Hall I.H.
      • Starnes C.O.
      • Lee K.-H.
      • Waddell T.G.
      ,
      • Schröder H.
      • Lösche W.
      • Strobach H.
      • Leven W.
      • Willuhn G.
      • Till U.
      • Schröer K.
      ). These activities are mediated chemically by α,β-unsaturated carbonyl structures, such as an α-methylene-γ-lactone or an α,β-unsubstituted cyclopentenone. These structure elements react with nucleophiles, especially cysteine sulfhydryl groups, by a Michael-type addition (
      • Picman A.K.
      • Rodriguez E.
      • Towers C.H.N.
      ,
      • Schmidt T.J.
      ). Therefore, exposed thiol groups, such as cysteine residues in proteins, appear to be the primary targets of sesquiterpene lactones. Some sesquiterpene lactones, for example, helenalin, possess two alkylant structure elements which may be responsible for their remarkably high activity (
      • Picman A.K.
      • Rodriguez E.
      • Towers C.H.N.
      ,
      • Schmidt T.J.
      ,
      • Kupchan S.M.
      • Eakin M.A.
      • Thomas A.M.
      ,
      • Lee K.H.
      • Meck R.
      • Piantadosi C.
      ). However, other factors, such as lipophilicity, molecular geometry, and the chemical environment of the target sulfhydryl may also influence the activity of sesquiterpene lactones (
      • Schmidt T.J.
      ).
      Despite many studies which prove the anti-inflammatory activity of sesquiterpene lactones, the molecular mechanism by which they exert their anti-inflammatory effect has not been sufficiently explained. Recently, we were able to show that helenalin, 11α,13-dihydrohelenalin, and chamissonolid potently inhibit induction of the transcription factor NF-κB (
      • Lyβ G.
      • Schmidt T.J.
      • Merfort I.
      • Pahl H.L.
      ). This protein is a central mediator of the human immune response (
      • Baeuerle P.
      • Henkel T.
      ). In almost all cell types, NF-κB, composed of a p50 and p65 subunit, is retained in an inactive cytoplasmic complex by binding to a third, inhibitory subunit, IκB. A large variety of inflammatory conditions, such as bacterial and viral infection as well as inflammatory cytokines, rapidly induce NF-κB activity. Active NF-κB is released from the cytoplasmic complex by phosphorylation and proteolytic degradation of the IκB subunit. The activated factor then translocates to the nucleus, where it stimulates the transcription of its target genes. NF-κB regulates the transcription of various inflammatory cytokines, such as interleukin-1, -2, -6, and -8 and TNF-α,
      The abbreviations used are: TNF, tumor necrosis factor; DTT, dithiothreitol; EMSA, electrophoretic mobility shift assay; GSH, glutathione; MG132, carbobenzoxyl-leucinyl-leucinyl-leucinal-H; PBS, phosphate-buffered saline; poly(dI-dC), polydeoxyinosinic deoxycytidylic acid, double-stranded alternating copolymer.
      1The abbreviations used are: TNF, tumor necrosis factor; DTT, dithiothreitol; EMSA, electrophoretic mobility shift assay; GSH, glutathione; MG132, carbobenzoxyl-leucinyl-leucinyl-leucinal-H; PBS, phosphate-buffered saline; poly(dI-dC), polydeoxyinosinic deoxycytidylic acid, double-stranded alternating copolymer.
      as well as genes encoding cyclooxygenase-II, nitric oxide synthase, immunoreceptors, cell adhesion molecules, hematopoetic growth factors, and growth factor receptors (see Refs.
      • Baeuerle P.
      • Henkel T.
      and
      • Baeuerle P.
      • Baltimore D.
      ). Pharmacological inhibition of NF-κB in vivo may thus substantially attenuate inflammatory processes.
      We have recently shown that the sesquiterpene lactones helenalin, 11α,13-dihydrohelenalin, and chamissonolid inhibit DNA binding activity of NF-κB (
      • Lyβ G.
      • Schmidt T.J.
      • Merfort I.
      • Pahl H.L.
      ). However, we were not able to identify the molecular target through which these natural compounds exert their effect. Since NF-κB activation by many different inducers is inhibited by helenalin, the sesquiterpene lactone must target a common step in these signal transduction pathways. In addition to the physiological induction of NF-κB by inflammatory mediators, the transcription factor can be activated in vitro by treatment of cell extracts with the detergent desoxycholate. This dissociates the inhibitor IκB from the p50 and p65 subunits, releasing an active NF-κB heterodimer. We were able to show that in cell extracts generated from helenalin-treated cells, desoxycholate treatment no longer activates NF-κB-DNA binding activity (
      • Lyβ G.
      • Schmidt T.J.
      • Merfort I.
      • Pahl H.L.
      ). These data suggest that helenalin either modifies the NF-κB·IκB complex in such a way that desoxycholate cannot dissociate the complex, or that it attacks the desoxycholate-released NF-κB so that DNA binding is impaired. We were subsequently able to demonstrate that the homodimer of subunit p50, which contains a redox sensitive cysteine residue in its DNA-binding domain (
      • Kumar S.
      • Rabson A.B.
      • Gélinas C.
      ,
      • Matthews J.R.
      • Wakasugi N.
      • Virelizier J.-L.
      • Yodoi J.
      • Hay R.T.
      ,
      • Toledano M.B.
      • Leonhard W.J.
      ), is not affected by helenalin. Helenalin-treated stimulated cells, which show no NF-κB binding activity in the EMSA, nonetheless show a decrease in the amount of IκB-α protein in Western blot analysis. These data allow two interpretations. Either, helenalin may modify the IκB-α protein so that it is no longer recognized by the anti-IκB-α antibody. Alternatively, IκB is degraded in helenalin-treated cells, releasing NF-κB, but the transcription factor is modified, so that it no longer binds DNA.
      Hehner and colleagues (
      • Hehner S.P.
      • Heinrich M.
      • Bork P.M.
      • Vogt M.
      • Ratter F.
      • Lehmann V.
      • Schulze-Osthoff K.
      • Dröge W.
      • Schmitz M.L.
      ) have recently presented evidence that the sesquiterpene lactone parthenolide inhibits degradation of IκB-α and IκB-β. From their data, these authors conclude that sesquiterpene lactones inhibit NF-κB activation by preventing the degradation of IκBs. These results contradict our previous report that IκB-α immunoreactivity decreases significantly in helenalin-treated, ocadaic acid-stimulated cells (
      • Lyβ G.
      • Schmidt T.J.
      • Merfort I.
      • Pahl H.L.
      ). In this study we show that in helenalin-treated cells, degradation of IκB, and NF-κB translocation to the nucleus are not inhibited. Using a confocal laser scanning microscope, we demonstrate the presence of IκB-released NF-κB in the nucleus of helenalin-treated, TNF-α-stimulated cells. Moreover, we show that helenalin directly interferes with NF-κB DNA binding due to its alkylating activity. Helenalin selectively modifies the p65 subunit of the transcripition factor, thereby inhibiting its DNA binding. Most importantly we show that, unlike antioxidants such as acetylsalicylic acid, helenalin can inactivate the active NF-κB complex. This property is crucial for the treatment of inflammation, where previously activated NF-κB is sustaining the process of inflammation and needs to be inactivated. Helenalin is the first anti-inflammatory agent shown to exert its effect by directly modifying the transcription factor NF-κB. These data reinforce the view of Willuhn (
      • Willuhn G.
      ) that sesquiterpene lactones could serve as lead compounds for the development of pharmaceutically used anti-inflammatory remedies.

      DISCUSSION

      Sesquiterpene lactones represent an active principle of many drugs used in traditional medicine as anti-inflammatory remedies. Flowerheads from A. montana and A. chamissonis ssp.foliosaare prominent examples (
      • Willuhn G.
      ). Their extracts are used to treat hematomas, contusions, sprains, rheumatic diseases, and superficial inflammations of the skin (see for Ref.
      • Willuhn G.
      ). The anti-inflammatory effect of this remedy is mediated by esters of helenalin, 11α,13-dihydrohelenalin and chamissonolid. In several studies it was shown that these compounds interfere with cellular processes including oxidative phosphorylation, platelet aggregation, histamine and serotonin release (
      • Hall I.H.
      • Starnes C.O.
      • Lee K.-H.
      • Waddell T.G.
      ,
      • Schröder H.
      • Lösche W.
      • Strobach H.
      • Leven W.
      • Willuhn G.
      • Till U.
      • Schröer K.
      ). However, the molecular mechanism by which sesquiterpene lactones exert their anti-inflammatory effect could not be sufficiently explained. Recently we have shown that helenalin selectively inhibits DNA binding activity of the transcription factor NF-κB, and prevents the induction of κB-dependent gene expression (
      • Lyβ G.
      • Schmidt T.J.
      • Merfort I.
      • Pahl H.L.
      ). Since various inflammatory processes are directly dependent on NF-κB activity, we suggested that this is the molecular mechanism by which helenalin exerts its anti-inflammatory effect. However, at that time we were not able to explain in detail the mechanism by which helenalin inhibits NF-κB activity. Two experiments pointed to possible models. First, in helenalin-treated cells, the anti-IκB-α immunoreactivity decreases after ocadaic acid stimulation, even though no active NF-κB can be detected. Second, the detergent desoxycholate, which can release active NF-κB from IκB in untreated cell extracts, is no longer able to do so in helenalin-treated cells. These observations lead us to formulate two hypotheses. Either, helenalin may modify the NF-κB·IκB-α complex so that it cannot be dissociated by desoxycholate and IκB is no longer recognized by the anti-IκB-α antibody. Alternatively, IκB is degraded in helenalin-treated cells releasing NF-κB, but the transcription factor is modified so that it no longer binds DNA.
      Here we show that the helenalin-mediated inhibition of NF-κB is due to alkylation of the active transcription factor. Interestingly, our data indicate that helenalin targets the cysteine sulfhydryl groups in the p65 subunit. Another alkylating agent shown to directly modify NF-κB, N-ethylmaleimide, reacts with cysteine 62 in the p50 subunit (
      • Toledano M.B.
      • Ghosh D.
      • Trinh F.
      • Leonard W.J.
      ). p65 alkylation prevents DNA binding of active NF-κB. This model is supported by the recent elucidation of the x-ray structure of the p65 homodimer, which shows that cysteine 38 is located within the DNA-binding domain and another cysteine, cysteine 120, is found in a proximal loop (
      • Chen Y.-Q.
      • Ghosh S.
      • Ghosh G.
      ). It is possible that helenalin modifies either one or both of these residues. In addition, a single helenalin molecule may cross-link these two cysteines. In a computer based study, we are currently investigating possible alkylation sites in the p65 subunit.
      We substantiate the observation that helenalin does not prevent IκB degradation. Using immunocytochemistry, we show that in helenalin-treated cells NF-κB is translocated to the nucleus, which can only occur after IκB-α degradation. Therefore, the recent report that the sesquiterpene lactone parthenolide prevents the induced degradation of IκB-α and IκB-β should be reinvestigated (
      • Hehner S.P.
      • Heinrich M.
      • Bork P.M.
      • Vogt M.
      • Ratter F.
      • Lehmann V.
      • Schulze-Osthoff K.
      • Dröge W.
      • Schmitz M.L.
      ). Parthenolide possesses a germacranolide skeleton with two potentially reactive centers in the form of an exocyclic methylene group and an epoxide ring. Since we have shown here that sesquiterpene lactones with two reactive centers interfere directly with NF-κB DNA binding, the statement that parthenolide does not do so may also require revision. These discrepancies between the results of Hehner et al.(
      • Hehner S.P.
      • Heinrich M.
      • Bork P.M.
      • Vogt M.
      • Ratter F.
      • Lehmann V.
      • Schulze-Osthoff K.
      • Dröge W.
      • Schmitz M.L.
      ) and our observations may be due to the different experimental conditions. We had also previously reported that helenalin does not interfere with NF-κB DNA binding (
      • Lyβ G.
      • Schmidt T.J.
      • Merfort I.
      • Pahl H.L.
      ). However, we realized that the conditions used in this in vitro experiment are not comparable to the conditions under which helenalin reacts in intact cells. Specifically, the reaction time allowed in vitro was shorter (20 min versus 1 h in vivo) and the temperature lower (room temperature versus 37 °C). Our attempts to adapt the in vitro conditions to those presentin vivo were not successful. The higher temperature and longer incubation time led to degradation of the protein extracts. Therefore, we designed new experiments (Figs. 3 and 4) to investigate whether sesquiterpene lactones directly modify NF-κB. Using this approach, we were able to show that the direct inhibitory effect on NF-κB binding is not restricted to sesquiterpene lactones of the pseudoguaianolide type but is observed also with the germacranolides diversifolin, tirotundin, and diversifolin methylether (
      • Rüngeler P.
      • Lyβ G.
      • Castro V.
      • Mora G.
      • Pahl H.L.
      • Merfort I.
      ).
      We have investigated the effects of lipophilicity and molecular geometry on NF-κB inhibition. Our results suggest that neither property influences sesquiterpene lactone activity. Helenalin isobutyrate is more lipophilic than helenalin, but both compounds have nearly the same inhibitory activity. In contrast to helenalin and helenalin isobutyrate, Mexicanin I possesses a trans-configurated lactone ring. We could show that Mexicanin I has the same inhibitory activity as the two cis-configurated lactones. Thus, lipophilicity and molecular geometry appear to be of marginal importance among these bifunctional alkylants. In contrast, the monofunctional 2,3-dihydroaromaticin is significantly less effective at inhibiting NF-κB (50 μm versus 20 μm in Mexicanin I). Therefore, the most important structural element is the occurrence of two reactive centers, one as an exocyclic methylene group and the other as an α,β-unsaturated cyclopentenone ring. However, the small number of compounds used do not allow us to draw final conclusions. We are currently investigating additional sesquiterpene lactones in order to define the structural requirements for NF-κB inhibition in more detail.
      Taken together, we have shown that sesquiterpene lactones inhibit the NF-κB signaling cascade by directly targeting the active heterodimer. By selectively modifying the p65 subunit they use a unique mechanism of NF-κB inactivation, which is quite different from that of other anti-inflammatory agents (
      • Baeuerle P.A.
      • Baichwal V.R.
      ). Based on our results, sesquiterpene lactones could serve as lead compounds for the development of novel, potent anti-inflammatory drugs for the treatment of inflammatory disorders such as rheumatoid arthritis or inflammatory bowel diseases. These drugs could also be important in the treatment or prevention of adult respiratory distress syndrome or systemic immune response syndrome.

      Acknowledgments

      We are grateful to Dr. M. Follo, Core facility, Department of Medicine I, University Hospital, Freiburg, for help with the confocal laser scanning microscope.

      REFERENCES

        • Willuhn G.
        Pharm. Unserer Zeit. 1981; 10: 1-7
        • Hall I.H.
        • Lee K.H.
        • Starnes C.O.
        • Sumida Y.
        • Wu R.Y.
        • Waddell T.G.
        • Cochran J.W.
        • Gerhart K.G.
        J. Pharmacol. Sci. 1979; 68: 537-541
        • Hall I.H.
        • Starnes C.O.
        • Lee K.-H.
        • Waddell T.G.
        J. Pharmacol. Sci. 1980; 69: 537-543
        • Schröder H.
        • Lösche W.
        • Strobach H.
        • Leven W.
        • Willuhn G.
        • Till U.
        • Schröer K.
        Thromb. Res. 1990; 57: 839-845
        • Picman A.K.
        • Rodriguez E.
        • Towers C.H.N.
        Chem. Biol. Interact. 1979; 28: 83-89
        • Schmidt T.J.
        Bioorg. Med. Chem. 1997; 5: 645-653
        • Kupchan S.M.
        • Eakin M.A.
        • Thomas A.M.
        J. Med. Chem. 1971; 14: 1147-1152
        • Lee K.H.
        • Meck R.
        • Piantadosi C.
        J. Med. Chem. 1973; 16: 299-301
        • Lyβ G.
        • Schmidt T.J.
        • Merfort I.
        • Pahl H.L.
        Biol. Chem. 1997; 378: 951-961
        • Baeuerle P.
        • Henkel T.
        Annu. Rev. Immunol. 1994; 12: 141-179
        • Baeuerle P.
        • Baltimore D.
        Cell. 1996; 87: 13-20
        • Kumar S.
        • Rabson A.B.
        • Gélinas C.
        Mol. Cell. Biol. 1992; 12: 3094-3106
        • Matthews J.R.
        • Wakasugi N.
        • Virelizier J.-L.
        • Yodoi J.
        • Hay R.T.
        Nucleic Acids Res. 1992; 20: 3821-3830
        • Toledano M.B.
        • Leonhard W.J.
        Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 4328-4332
        • Hehner S.P.
        • Heinrich M.
        • Bork P.M.
        • Vogt M.
        • Ratter F.
        • Lehmann V.
        • Schulze-Osthoff K.
        • Dröge W.
        • Schmitz M.L.
        J. Biol. Chem. 1998; 273: 1288-1297
        • Graham F.L.
        • Van Der Eb A.J.
        Virology. 1973; 52: 456-467
        • Ruben S.M.
        • Dillon P.J.
        • Schreck R.
        • Henkel T.
        • Chen C.-H.
        • Maher M.
        • Baeuerle P.A.
        • Rosen C.A.
        Science. 1991; 251: 1490-1493
        • Schmitz M.L.
        • Baeuerle P.A.
        EMBO J. 1991; 10: 3805-3817
        • Willuhn G.
        • Kresken J.
        • Leven W.
        Planta Med. 1990; 56: 111-114
      1. Leven, W. (1988) Thesis, Heinrich-Heine-Universität, Düsseldorf.

        • Merfort I.
        • Wendisch D.
        Phytochemistry. 1993; 34: 1436-1437
        • Chen Y.-Q.
        • Ghosh S.
        • Ghosh G.
        Nat. Struct. Biol. 1998; 5: 67-73
        • Schreck R.
        • Meier B.
        • Männel D.N.
        • Dröge W.
        • Baeuerle P.A.
        J. Exp. Med. 1992; 175: 1181-1194
        • Wiebcke M.
        • Kresken J.
        • Mootz D.
        • Willuhn G.
        Tetrahedron. 1982; 38: 2709-2714
        • Schmidt T.J.
        J. Mol. Struct. 1996; 385: 99-112
        • Ganchi P.A.
        • Sun S.S.
        • Green W.C.
        • Ballard D.W.
        Mol. Cell. Biol. 1993; 13: 7826-7835
        • Appendino G.
        • Calleri M.
        • Chiari G.
        • Gariboldi P.
        • Menichini F.
        Gazz. Chim. Ital. 1986; 116: 637-641
        • Toledano M.B.
        • Ghosh D.
        • Trinh F.
        • Leonard W.J.
        Mol. Cell. Biol. 1993; 13: 852-860
        • Rüngeler P.
        • Lyβ G.
        • Castro V.
        • Mora G.
        • Pahl H.L.
        • Merfort I.
        Planta Med. 1998; 64: 588-593
        • Baeuerle P.A.
        • Baichwal V.R.
        Adv. Immunol. 1997; 65: 111-137