Anti-Inflammatory Remedy Frankincense the Functional Target

The Journal of Immunology

Identiﬁcation of Human Cathepsin G As a Functional Target of Boswellic Acids from the Anti-Inﬂammatory Remedy Frankincense1

Lars Tausch,* Arne Henkel,† Ulf Siemoneit,† Daniel Poeckel,† Nicole Kather,‡ Lutz Franke,§ Bettina Hofmann,§ Gisbert Schneider,§ Carlo Angioni,¶ Gerd Geisslinger,¶ Carsten Skarke,¶ Wolfgang Holtmeier,ʈ Tobias Beckhaus,* Michael Karas,* Johann Jauch,‡ and Oliver Werz2†

Frankincense preparations, used in folk medicine to cure inflammatory diseases, showed anti-inflammatory effectiveness in animal models and clinical trials. Boswellic acids (BAs) constitute major pharmacological principles of frankincense, but their targets and the underlying molecular modes of action are still unclear. Using a BA-affinity Sepharose matrix, a 26-kDa protein was selectively precipitated from human neutrophils and identified as the lysosomal protease cathepsin G (catG) by mass spectrometry (MALDI- TOF) and by immunological analysis. In rigid automated molecular docking experiments BAs tightly bound to the active center of catG, occupying the same part of the binding site as the synthetic catG inhibitor JNJ-10311795 (2-[3-{methyl[1-(2-naphthoyl)piperidin-4-yl]amino}carbonyl)-2-naphthyl]-1-(1-naphthyl)-2-oxoethylphosphonic acid). BAs potently suppressed the proteolytic ac- ϳ tivity of catG (IC50 of 600 nM) in a competitive and reversible manner. Related serine proteases were significantly less sensitive against BAs (leukocyte elastase, chymotrypsin, proteinase-3) or not affected (tryptase, chymase). BAs inhibited chemoinvasion but not chemotaxis of challenged neutrophils, and they suppressed Ca2؉ mobilization in human platelets induced by isolated catG or by catG released from activated neutrophils. Finally, oral administration of defined frankincense extracts significantly reduced catG activities in human blood ex vivo vs placebo. In conclusion, we show that catG is a functional and pharmacologically relevant target of BAs, and interference with catG could explain some of the anti-inflammatory properties of frankincense. The Journal of Immunology, 2009, 183: 3433–3442.

rankincense, the gum resin derived from Boswellia spe- Whereas 11-keto-␤-BA (KBA) and 3-O-acetyl-11-keto-␤-BA cies, is frequently used in folk medicine to cure inflam- (AKBA) were shown to exhibit numerous biological activities in F matory diseases. Data from animal models of inflam- vitro, the corresponding BAs lacking the 11-oxo moiety (i.e., mation and from clinical trials suggest a therapeutic value of ␤-BA and A␤-BA) are considered as less relevant. In particular, frankincense in the treatment of acute and chronic inflammatory AKBA is thought to be mainly responsible for the pharmacological and allergic disorders (1, 2). The pentacyclic triterpenes boswellic actions of frankincense (2). In search for a molecular basis of the acids (BAs)3 (see Fig. 1A) are major ingredients of frankincense. anti-inflammatory effectiveness, 5- and 12-lipoxygenase (1), cyclooxygenase-1 (3), human leukocyte elastase (HLE) (4), and I␬B kinases (5) have been identified as possible targets of AKBA in *Institute of Pharmaceutical Chemistry, Johann Wolfgang Goethe-University Frank- biochemical and cellular in vitro models. However, due to the high furt, Frankfurt, Germany; †Department of Pharmaceutical Analytics, Pharmaceutical Institute, Eberhard-Karls-University Tuebingen, Tuebingen, Germany; ‡Organic concentrations required to interfere with these targets (IC50 of Chemistry II, University of Saarland, Saarbru¨cken, Germany; §Institut fu¨r Organische 1–50 ␮M) and the low maximal plasma levels of AKBA (Ͻ0.1 Chemie und Chemische Biologie/Zentrum fur Arzneimittelforschung, -Entwicklung ␮M) obtained after oral administration of standard doses of frank- und -Sicherheit, Johann Wolfgang Goethe-University Frankfurt, Frankfurt am Main, Germany; ¶Pharmazentrum Frankfurt, Zentrum fur Arzneimittelforschung, -Entwick- incense extracts (6), the pharmacological relevance of the pro- lung und -Sicherheit, Institut fu¨r Klinische Pharmakologie, Klinikum der Johann posed targets and mechanisms remain questionable (7). Wolfgang Goethe-Universita¨t, Frankfurt, Germany; and ʈDepartment of Gastroenter- ology, Medizinische Klinik 1, University Hospital, Frankfurt, Germany An elegant technique for the identification of a high-affinity Received for publication October 24, 2008. Accepted for publication July 6, 2009. small molecule (ligand)-protein (target) interaction is the protein- fishing approach, using an affinity matrix composed of the small The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance molecule covalently linked to an insoluble resin (8). Application of with 18 U.S.C. Section 1734 solely to indicate this fact. this strategy revealed human histone deacetylase as target for trap- 1 The financial support for this study by Pharmasan (Freiburg, Germany), Medeon oxin (9) or mTOR as target for the immunosuppressant rapamycin (Berlin, Germany), and by the Deutsche Forschungsgemeinschaft is acknowledged. (10). Using a BA-affinity Sepharose matrix and neutrophil lysates 2 Address correspondence and reprint requests to Dr. Oliver Werz, Department of as source of targets, we identified cathepsin G (catG) as a selected, Pharmaceutical Analytics, Pharmaceutical Institute, Eberhard-Karls-University Tu- ebingen, Auf der Morgenstelle 8, 72076 Tuebingen, Germany. E-mail address: high-affinity target for BAs. CatG, a neutral serine protease, is [email protected] mainly expressed in neutrophils, stored in azurophilic granules, 3 Abbreviations used in this paper: BA, ␤-boswellic acid; A␤-BA, 3-O-acetyl-␤- and released upon degranulation (11, 12). After release into the boswellic acid; AKBA, 3-O-acetyl-11-keto-␤-boswellic acid; catG, cathepsin G; JNJ-10311795, 2-[3-{methyl[1-(2-naphthoyl)piperidin-4-yl]amino}carbonyl)-2-naph- plasma, it cleaves extracellular matrix proteins, including laminin, thyl]-1-(1-naphthyl)-2-oxoethylphosphonic acid; HLE, human leukocyte elastase; proteoglycans, collagen, fibronectin, and elastin (13, 14), implying KBA, 11-keto-␤-boswellic acid; RMSD, root mean square deviation; SPR, surface a role in local destruction of connective tissue at sites of injury. plasmon resonance. CatG also processes chemokines (11, 15, 16), functioning as che- Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00 moattractant for T cells and other leukocytes (17), and modulates www.jimmunol.org/cgi/doi/10.4049/jimmunol.0803574 3434 IDENTIFICATION OF CATHEPSIN G AS TARGET OF BOSWELLIC ACIDS

integrin clustering on neutrophils (18). Moreover, catG stimulates In-gel digestion and MALDI-TOF-MS platelets via the protease-activated receptor-4 for aggregation and Protein bands were manually cut out and dissected gel pieces were sub- secretion (19) and acts as chemotactic agonist for the formyl pep- jected to in-gel digestion (24, 25), adapted for use on a Microlab STAR tide receptor on phagocytic leukocytes (20). Accordingly, catG digestion robot (26). Samples were reduced, alkylated, digested overnight inhibitors have been proposed to exhibit potential in treating cer- by trypsin, extracted, and the extracts were dried in a vacuum centrifuge. tain inflammatory disorders such as asthma, chronic obstructive MALDI-TOF-MS experiments were performed on an Ultraflex TOF/TOF mass spectrometer (Bruker Daltonics). The samples were dissolved in 5 ␮l pulmonary disease, emphysema, reperfusion injury, psoriasis, and of water/acetonitrile/TFA (29.5/70/0.5, v/v/v), and ␣-cyano-4-hydroxycin- rheumatoid arthritis (21). Since frankincense extracts showed ben- namic acid (3 mg/ml) was used as matrix. Analyte and matrix were spotted eficial effects in several of these disorders, interference of BAs consecutively in a 1:1 ratio on a stainless steel target and dried under with catG may possess pharmacological relevance. ambient conditions. The dried sample was washed with ice-cold 5% formic acid to reduce salt contamination. Spectra were externally calibrated with a Sequazyme peptide mass standards kit (Applied Biosystems) and inter- Materials and Methods nally calibrated on a tryptic auto digestion peptide (m/z of 2163.0564). The spectra were processed in flexAnalysis version 2.2 (Bruker Daltonics) us- Materials ing the SNAP algorithm (signal to noise threshold, 3; maximal number of BAs were prepared as described previously (22). ␣-Amyrin was from Ex- peaks, 150; quality factor threshold, 40). Proteins were identified by Mas- trasynthe`se; Boswellia serrata gum extract PS0201Bo was standardized to cot (Matrix Science) (peptide mass tolerance, 100 ppm; maximum missed at least 1% KBA and 1% AKBA (quantified by reversed phase HPLC) and cleavages using the NCBInr database: 2,314,886 sequences, 1,066,605,192 was provided by Pharmasan; EAH-Sepharose 4B, GE Healthcare Bio-Sci- total letters, date Jan. 26, 2005). Proteins with a score of 76 or higher were Ͻ ences; Abs against catG, BIOMOL; human catG, human tryptase, human considered significant ( p 0.05). chymase, human chymotrypsin, elastase inhibitor IV (N-(2-(4-(2,2-dimeth- ylpropionyloxy)phenylsulfonylamino)benzoyl)aminoacetic acid) and 2-[3- Purification of catG from neutrophils {methyl[1-(2-naphthoyl)piperidin-4-yl]amino}carbonyl)-2-naphthyl]-1-(1- Human neutrophils (2 ϫ 109) were suspended in ice-cold 0.15 M NaCl and naphthyl)-2-oxoethylphosphonic acid (JNJ-10311795), Calbiochem; human sonicated (five times for 30 s, 65%). Insoluble material was removed by proteinase-3, Elastin Products; matrigel, BD Biosciences; trypsin (sequenc- centrifugation (600 ϫ g, 10 min). The resulting supernatant was centri- ing grade), Roche; all other fine chemicals were obtained by Sigma- fuged (16,000 ϫ g, 30 min) and the granular fraction (pellet) was resus- Aldrich, unless stated otherwise. pended in 1 M NaCl plus 0.005% Triton X-100. The suspension was centrifuged at 16,000 ϫ g for 30 min, and four volumes of water were added Cells to restore isotonicity. Proteins were precipitated by ammonium sulfate (60% saturation) and then resuspended in 40 ml of 0.05 M Tris-HCl (pH For isolation of platelets and neutrophils venous blood was taken from 8.0). After centrifugation (16,000 ϫ g, 30 min), the supernatant was sub- healthy adult donors, with consent. The subjects had no apparent in- jected to an elastin-Sepharose affinity chromatography column (2.5 ϫ 20 flammatory conditions and had not taken anti-inflammatory drugs for at cm) and equilibrated with 0.05 M Tris buffer (pH 8.0). The column was least 10 days before blood collection. Leukocyte concentrates were pre- eluted with the equilibration buffer until the OD returned to baseline and ϫ 280 pared by centrifugation at 4000 g for 20 min at 20°C. Neutrophils then washed with 2 vol of 0.05 M Na acetate, 1 M NaCl (pH 5.0), and were immediately isolated by dextran sedimentation, centrifugation on fractions containing catG activity were eluted with 0.05 M Na acetate, 1 M Nycoprep cushions, and hypotonic lysis of erythrocytes. Neutrophils NaCl, 20% DMSO (pH 5.0). Active fractions were pooled, dialyzed in ϫ 6 Ͼ (7.5 10 cells/ml; purity 96–97%) were finally resuspended in PBS Vivaspin cut-off columns (5000 MWCO) against 20 mM Na acetate, 0.15 (pH 7.4) plus 1 mg/ml glucose. Platelets were isolated from superna- ϫ M NaCl (pH 5.5), and subjected to ion-exchange chromatography (CM tants (800 g, 10 min, room temperature) after centrifugation of leu- Sephadex C-50) column, equilibrated in the same buffer. The sample was kocyte concentrates on Nycoprep cushions (see above) to obtain plate- applied, washed with equilibration buffer, and bound material was eluted let-rich plasma. Platelet-rich plasma was then mixed with PBS (pH 5.9; ϫ by a linear NaCl gradient from 0.15 to 1 M. The total elution volume was 3/2, v/v), centrifuged (2000 g, 15 min, room temperature), and the 300 ml, and fractions of 6 ml were collected at a flow rate of 30 ml/h. The pelleted platelets were resuspended in PBS (pH 5.9)/0.9% NaCl (1/1, ϫ homogeneity of the purified material was assessed by SDS-PAGE and catG v/v), washed by centrifugation (2000 g, 10 min, room temperature) activity assays. and finally resuspended in PBS (pH 7.4). Test compounds were solubilized in ethanol or DMSO, never exceeding 1% (v/v). To exclude Protease activity assays toxic effects of BAs or vehicle (ethanol, 1%) during preincubation, neutrophil viability was analyzed by light microscopy and trypan blue The proteases were mixed with test compounds or DMSO (vehicle con- exclusion. Incubation with 30 ␮M of any of the BAs or ethanol (1%) at trol, Ͻ0.5%) in the respective assay buffer in a 96-well plate (total 37°C for up to 30 min caused no significant change in cell viability. volume 200 ␮l) and preincubated for 20 min at 25°C. The respective chromogenic protease substrate was added and the proteolysis was Immobilization of boswellic acids and protein pull-down assays monitored at 410 nm by spectrophotometric measurement using a Vic- tor2 plate reader (PerkinElmer). The enzymatic activity was determined by KBA was treated with glutaric anhydride to form the half-ester glut-KBA, the progress curve method. The protease activity in the presence of inhib- and analyzed by 1H and 13C nuclear magnetic resonance as well as by mass itor was compared with an uninhibited control (DMSO as vehicle), and spectrometry. Glut-KBA was linked to EAH Sepharose 4B by standard inhibition of the protease is given as the percentage of the control without amide coupling procedures. The carboxylic acid of the KBA core was inhibitor. unlikely to react due to steric crowding. The success of the coupling re- For analysis of isolated catG, the enzyme was either purified from neu- action was determined by two methods. First, glut-KBA was used in de- trophils as described, or commercially available purified catG from neu- ␮ ␮ ␮ fined excess (2 mol of the glut-KBA per 1 mol of NH2 groups of the trophils (Calbiochem) was used. Purified enzyme (0.2 g), diluted in 200 EAH Sepharose 4B). After the coupling reaction, the hypothetical excess ␮l of HEPES 0.1 M, NaCl 0.5 M (pH 7.4), 10% DMSO, was incubated of glut-KBA (1 ␮mol) could be indeed recovered. Second, treatment of with N-Suc-Ala-Ala-Pro-Phe-pNA (Suc-AAPF-pNA, 1 mM) as substrate glut-KBA with KOH in isopropanol under reflux for ϳ3 h cleaved the ester and the absorbance was measured at 410 nm at 25°C. Kinetic studies were bond and gave KBA, analyzed by thin layer chromatography. performed at substrate concentrations ranging from 0.1 to 4 mM using the For protein pull-down experiments, neutrophils were lysed by 1% Triton Lineweaver-Burk method. X-100 and lysates were incubated with the Sepharose slurries in lysis buffer Inhibition of related proteases was performed in analogy to catG. The (50 mM HEPES (pH 7.4), 200 mM NaCl, 1 mM EDTA, 1% Triton X-100, amounts of protease, the assay buffer, and the substrate were individually 2 mM PMSF, 10 ␮g/ml leupeptin, 120 ␮g/ml soybean trypsin inhibitor). adjusted to each type of protease as follows: tryptase, 0.5 ␮g of purified Alternatively, isolated catG (1 ␮g) was solubilized in 425 ␮l of lysis buffer enzyme, Tris-HCl 0.1 M (pH 8.3), 10% DMSO as assay buffer, and N-␣- containing 0.02% BSA and incubated with the Sepharose slurries. The benzoyl-DL-Arg-pNA (1 mM) as substrate; chymase, 0.1 ␮g of purified Sepharose beads were intensively washed, and precipitated proteins were enzyme, Tris-HCl 0.45 M, NaCl 1.8 M (pH 8.0), 10% DMSO, and 0.5 mM finally separated by SDS-PAGE and visualized by silver staining or by Suc-AAPF-pNA as substrate; chymotrypsin, 0.1 ␮g of purified enzyme,

Western blotting, respectively, as described (23). For identiﬁcation of se- Tris-HCl 0.1 M, CaCl2 25 mM (pH 8.3), 10% DMSO, and 0.2 mM Suc- lected proteins of interest, bands were cut out from the gel and proteins AAPF-pNA as substrate; human leukocyte elastase, 0.15 ␮g of puriﬁed were digested and subjected to MALDI-TOF mass spectrometry (MS). enzyme, HEPES 0.1 M, NaCl 0.5 M (pH 7.4), 10% DMSO, and 0.2 mM The Journal of Immunology 3435

N-methoxysuc-Ala-Ala-Pro-Val-pNA as substrate; proteinase-3, 0.5 ␮gof Neutrophil chemoinvasion assay purified enzyme, MOPS 0.1 M, NaCl 0.5 M, 5,5Ј-dithiobis-(2-nitro-ben- zoic acid) 0.1 mM (pH 7.5), 10% DMSO and 1 mM Boc-Ala-Ala-Nva- The migration of neutrophils along a chemotactic (fMLP) gradient through SBzl as substrate. Matrigel was performed according to Steadman et al. (35) with some slight modifications. In brief, freshly isolated neutrophils (2 ϫ 106) were resuspended in 1 ml of HEPES-buffered RPMI 1640 medium with 10% (v/v) Automated molecular docking experiments FCS and preincubated with test compounds or with vehicle (DMSO, Ͻ ␮ Rigid automated molecular docking was performed using GOLD v4.0 0.5%). Cell suspension (150 l) was then placed on the upper chamber ␮ (Cambridge Crystallographic Data Centre, Cambridge, U.K.; www.ccdc. of a two-compartment Boyden chamber (5- m pore size filters) in a 24- cam.ac.uk), which relies on a genetic algorithm (27). We used a known well format. Cells were then allowed to migrate through Matrigel-coated crystal structure of catG (identifier: 1t32, 1.85 Å resolution) (28) from the pore size filters for 40 min into the lower chamber containing buffer (neg- ␮ Protein Data Bank (29). Hydrogens were added, and then energy mini- ative control) or fMLP (0.1 M) as chemoattractant. Cells on the bottom mized using the AMBER99 force field (30) within the software package of the wells were fixed with 3.7% formaldehyde, stained with Gram’s MOE v2008.10 (Chemical Computing Group). For the co-crystallized catG Violet, washed, and the stain was solubilized using acetic acid. The ab- inhibitor JNJ-10311795, hydrogen atoms were added, and energy minimi- sorption of the eluted stain was measured at 570 nm/620 nm. zation was performed using the MMFF94x force field (31). The three- Neutrophil chemotaxis assay dimensional structure of AKBA has recently been determined by our group (N. Kather, L. Tausch, D. Poeckel, O, Werz, E. Herdtweck, and J. Jauch, Freshly isolated neutrophils (2 ϫ 105) were resuspended in 100 ␮lof manuscript in preparation:). GOLD parameter settings for the genetic al- HEPES-buffered RPMI 1640 medium with 10% (v/v) FCS and preincu- gorithm were: number of operations, 100,000; population size, 100; selec- bated with test compounds or with vehicle (1% DMSO) for 20 min at 37°C. tion pressure, 1.1; number of islands, 5; niche size, 2; migrate, 10; mutate, The cell suspension was placed on the upper chamber of a two-compart- 95; crossover, 95. A 20-Å radius around the active site defined the binding ment microchemotaxis chamber (8-␮m pore size filters) in a 24-well for- pocket for automated docking. The ChemScore fitness function (32) mat. Cells were allowed to migrate into filters for 60 min in contact to the was employed for scoring the predicted receptor-ligand complexes. It lower chamber containing HEPES-buffered RPMI 1640 medium with 10% estimates the total free energy change that occurs on ligand binding and (v/v) FCS and DMSO (negative control) or fMLP (0.1 ␮M) as chemoat- was trained by regression against binding affinity data. As the fitness tractant. Cells in the filters were fixed with 4% formaldehyde and stained scores are dimensionless, they cannot be used explicitly as values for with Harris hematoxylin solution. The total numbers of cells in three high- binding energy or binding affinity. However, in each case, the scale of power microscopic fields (ϫ400) were counted. the score gives a guide as to how good the pose is: the higher the score, the better the docking result is likely to be. The ChemScore function Determination of catG activity in human plasma (ex vivo) and takes into account factors such as protein-ligand hydrogen bond energy, clinical study protocol lipophilic interactions, the entropic loss that occurs when single, acyclic In a randomized, double-blind, placebo-controlled, multicenter trial the bonds in the ligand become nonrotatable upon binding, and clashes tolerability and efficacy of orally administered B. serrata gum extract between protein and ligand atoms as well as ligand internal torsional PS0201Bo was investigated for maintaining the remission of Crohn’s dis- strain energy. Each docking run was repeated 10 times to obtain average ease. In the study addressing the tolerability (phase I), two soft gelatin score values with SE. The same method was used for redocking of the capsules of PS0201Bo each containing 400 mg (total of 800 mg of co-crystallized inhibitor. Root mean square deviation (RMSD) values PS0201Bo) were taken by healthy adult male volunteers in the morning between the x-ray co-crystal coordinates and the docking solutions were immediately after a defined meal. Venous blood samples were collected computed, and a mean value with SE was calculated. PyMOL was used from the volunteers after 0, 2, 4, 8, and 24 h, citrated, and immediately for visualization of docking poses (33). stimulated with 10 ␮M cytochalasin B and 2.5 ␮M fMLP for 5 min. Plasma was prepared and catG activity was measured (see above). Surface plasmon resonance (SPR) spectrometry In the study addressing the efficacy of PS0201Bo (phase II–III), the patients received two soft gelatin capsules of PS0201Bo each three times Experiments were conducted on a Biacore X device. Purified catG (100 daily (total dose of 2400 mg of PS0201Bo/d) or two capsules of placebo, ␮g/ml) was immobilized onto CM5 biosensor chip flow cells using the respectively, during or immediately after a meal over a period of 52 wk. standard amine coupling method according to the manufacturer’s instruc- Venous blood samples from the patients were collected before study med- tions. Flow cell 1 was not altered (reference), and catG (170 fmol/mm2) ication application and after 4 wk of treatment and catG activity in the was immobilized on flow cell 2 corresponding to 4500 resonance units. plasma was assessed as described above. In parallel, aliquots of the plasma Equilibration of the baseline was obtained by a constant flow of HBS-P were used for analysis of BAs (see below). buffer (10 mM HEPES, 150 mM NaCl, 0.01% P20, 1% DMSO (pH 7.4)) The study protocol and a sample patient information/consent form were for 4 h. Stock solutions of BAs, ␣-amyrin, and JNJ-10311795 were diluted reviewed by the competent Ethics Committee (IECs) of the University of into assay buffer. Measurements were performed at 25°C and a flow rate of Frankfurt, Germany, and a favorable opinion was issued on July 6, 2006 30 ␮l/min. After recording association, the liquid phase was replaced by (Gescha¨fts no. 158/06); regulatory authority approvals were obtained from assay buffer and the dissociation was monitored. The binding profiles were the competent Higher Federal Authority, the Bundesinstitut fu¨r Arzneimit- obtained after subtracting the response signal of the untreated reference cell tel und Medizinprodukte, Bonn, Germany (EudraCT no. 2006-002939-24, 1. Sensorgrams were processed by using automatic correction for nonspe- Vorlage no. 4031905 of July 18, 2006). This trial was conducted in com- cific bulk refractive index effects using Biacore evaluation version 3.1 pliance with the protocol and principles of the Declaration of Helsinki software. (1996) and the International Conference on Harmonisation/World Health Organization Good Clinical Practice guidelines. Intracellular Ca2ϩ measurements Determination of boswellic acids in human plasma by 8 Platelets (6 ϫ 10 /ml PBS plus 1 mg/ml glucose) were incubated with 2 LC-ESI-MS/MS ␮M Fura-2-AM for 30 min at 37°C. After washing, 108 platelets were resuspended in 1 ml of PBS plus 1 mg/ml glucose, preincubated with The content of BAs in human plasma was determined by liquid chroma- inhibitors for 10 min at room temperature, and incubated in a thermally tography-electrospray ionization-tandem mass spectrometry (LC-ESI-MS/ controlled (37°C) fluorometer cuvette in a spectrofluorometer (Aminco- MS). BAs were separated using a Gemini C18 column (150 mm ϫ 2mm ␮ Bowman series 2; Thermo Spectronic) with continuous stirring. CaCl2 internal diameter, 5 m particle size, and 110 Å pore size; Phenomenex) (1 mM) was added 30 s before the addition of catG (0.1 nmol/ml), and and determined with an API 4000 triple quadrupole mass spectrometer the fluorescence emission at 510 nm was measured after excitation at (Applied Biosystems) equipped with a Turbo-V-source operating in neg- 340 and 380 nm, respectively. Intracellular Ca2ϩ levels were calculated ative electrospray ionization mode. High-purity nitrogen for the mass spec- according to the method of Grynkiewicz et al. (34). In other experi- trometer was produced by a NGM 22-LC/MS nitrogen generator (cmc ments, 108 Fura-2-AM-loaded platelets were mixed with 107 unloaded Instruments). A linear gradient was employed at a flow rate of 0.6 ml/min neutrophils in 1 ml of PBS plus 1 mg/ml glucose and 0.1 mM EDTA, mobile phase with a total run time of 10 min (mobile phase A, water/ and preincubated with the inhibitors for 10 min at room temperature. ammonia (100/0.05, v/v); mobile phase B, acetonitrile/ammonia (100/0.05, Five min prior to stimulation with 100 nM fMLP, 10 ␮M cytochalasin v/v)). The gradient started from 90% phase A to 10% within 3.5 min. This

B was added, and CaCl2 (1.1 mM) was added 1 min prior to fMLP. The was held for 1.5 min at 10% phase A. Within 0.5 min the mobile phase ﬂuorescence was measured and intracellular Ca2ϩ concentration was shifted back to 90% phase A and was held for 4.5 min to equilibrate the calculated as described above. column for the next sample. The injection volume of samples was 20 ␮l. 3436 IDENTIFICATION OF CATHEPSIN G AS TARGET OF BOSWELLIC ACIDS

FIGURE 1. BAs selectively bind catG. A, Chemical structures of BAs and related pentacyclic triterpenes. B, Structure of the KBA-Sepharose affinity matrix (KBA- Seph), composed of KBA linked to glutaric acid via the C3-OH group to EAH Sepharose 4B (Seph). C, Super- natants (12,000 ϫ g) of neutrophil lysates were incubated with either KBA-Seph or Seph. Precipitates were separated by SDS-PAGE and proteins were visualized by silver staining. After in-gel digestion and analysis by MALDI-TOF-MS, the band at 26 kDa (arrow) was identified as human catG. D, Coverage of peptides matching the human catG sequence. Peptide fragments identified by MALDI-TOF-MS matched to the protein sequence of human catG (chain A), with a sequence coverage of 63% (bold). E, Proteins precipitated by Seph and KBA-Seph (see C) were analyzed by SDS- PAGE and Western blotting using specific Abs against catG. F, Isolated human catG (1 ␮g) was incubated with Seph or KBA-Seph. After extensive washing procedures, catG was analyzed in the precipitates by Western blotting. Similar results were obtained in at least three additional experiments, each.

Retention times of KBA, AKBA, ␤-BA, and A␤-BA, and internal standard tracted by liquid-liquid extraction. The prepared samples were extracted (2,5-dimethyl-celecoxib (DMC)) were 3.70, 4.13, 4.19, 4.34, and 5.24 min, twice with ethyl acetate. The organic phase was removed at 45°C under a respectively. The quantification of BAs was performed with Analyst soft- gentle stream of nitrogen. The residues were reconstituted with 50 ␮lof ware V. 1.4.2 (Applied Biosystems) employing the internal standard acetonitrile/water/ammonia (60/40/0.1, v/v), centrifuged for 2 min at (DMC). Ratios of analyte peak area and internal standard area (y-axis) were 10,000 ϫ g, and then transferred to glass vials (Macherey-Nagel) before plotted against concentrations (x-axis), and calibration curves were calcu- injection in the LC/MS-MS system. lated by least squares regression with 1/concentration2 weighting. BA stock solutions (500 ␮g/ml) were prepared in methanol and further Statistics diluted with methanol to obtain working standards with the following con- ␤ ␤ Statistical evaluation of the data was performed by one-way ANOVAs for centration ranges: -BA from 1.25 to 12,500 ng/ml and A -BA, KBA, and independent or correlated samples followed by Bonferroni post hoc tests. AKBA from 0.125 to 1250 ng/ml. The following working standards were Where appropriate, Student’s t test for paired observations was applied. A used for plasma to obtain calibration standards with the concentrations p value of Ͻ0.05 was considered significant. ranges: ␤-BA from 0.125 to 1250 ng/ml, and A␤-BA, KBA, and AKBA from 0.013 to 125 ng/ml. The lower limit of quantification belongs to the standard curve and is the first point that has (1) accuracy between 80 and Results 120%, and (2) five times the peak-to-noise ratio (according the Food and Protein fishing with immobilized boswellic acids reveals Drug Administration 2001 Guidance for Industry: Bioanalytical Method cathepsin G as boswellic acid-binding protein Validation) and was 0.125 ng/ml for ␤-BA and A␤-BA and 0.038 ng/ml for KBA and AKBA, respectively. KBA was linked at the C3-OH moiety using glutaric anhydride Samples for standard curve and quality controls were prepared with 200 yielding the half ester 3-O-glutaroyl-KBA (glut-KBA). The free ␮l of blank human plasma (Blutspendedienst Hessen, Deutsches Rotes OH moiety at C3 is unlikely to be part of a pharmacophore since Kreuz, Frankfurt, Germany), 50 ␮l of ammonia solution 25%, 20 ␮lof esterification (e.g., in AKBA) frequently improves the potency in working standards, and 20 ␮l of internal standard solution (25 ng/ml DMC). Plasma from patients was prepared similarly, but instead of 200 ␮l pharmacological assays (1). The de novo formed free carboxylic of blank human plasma and 20 ␮l of working standard, 200 ␮l of sample group of glut-KBA was amide-coupled with the primary amine of and 20 ␮l of methanol were added. BAs and internal standard were ex- EAH Sepharose 4B, yielding KBA-Seph (Fig. 1B). EAH Sepharose The Journal of Immunology 3437

Table I. Docking results of BAs and controlsa which excludes coprecipitation via a linker molecule and thus con- ﬁrms a direct interaction of BAs with catG. Ligand ChemScore H Bond Lipo RMSD (Å)

JNJ-10311795 52.97 Ϯ 0.08 3.46 Ϯ 0.01 312.15 Ϯ 1.07 0.22 Ϯ 0.03 Docking of boswellic acids to cathepsin G ␤-BA 26.51 Ϯ 0.21 0.99 Ϯ 0.01 230.62 Ϯ 2.43 0.51 Ϯ 0.08 A␤-BA 24.83 Ϯ 0.36 0.97 Ϯ 0.02 224.38 Ϯ 3.58 0.55 Ϯ 0.08 To further study the binding of BAs to catG, we used automated KBA 28.38 Ϯ 0.25 1.95 Ϯ 0.09 223.37 Ϯ 2.1 0.45 Ϯ 0.07 molecular docking. The co-crystallized synthetic catG inhibitor Ϯ Ϯ Ϯ Ϯ AKBA 26.12 0.32 1.71 0.08 221.08 3.16 0.44 0.06 K ␣-Amyrin 27.36 Ϯ 0.25 0.9 Ϯ 0.03 217.06 Ϯ 3.99 2.36 Ϯ 0.35 JNJ-10311795 ( i for catG of 38 nM) (28), designed and devel- oped based on structural information from the x-ray structure of a The docking scores are listed for each docked ligand together with the protein- catG, was successfully redocked into Protein Data Bank structure ligand hydrogen bonding (H Bond) and Lipophilic (Lipo) contributions to the score. RMSD values for all the docking solutions and, in the case of JNJ-10311795, between 1T32 (reference for catG). The acquired binding mode into the S1 the docking poses and the co-crystal structure are shown. Values are given as and S2 pocket was identical to the x-ray structure with an RMSD Ϯ means SE. Larger positive values indicate potential tight binding. of 0.22 Ϯ 0.03 Å and a ChemScore of 52.97 Ϯ 0.08. The protein- ligand interaction seems to be mainly driven by lipophilic/pi-pi interactions and a tight H-bonding network formed by the phos- 4B (Seph) without glut-KBA as ligand was used as negative con- phonate group, which is also reflected in the ChemScore (Table I). trol to discriminate unspecific protein binding. After incubation of The 2-naphthyl group occupies S1 and the 1-naphthyl group S2, neutrophil lysates with Sepharose beads, bound proteins were sep- with the latter being involved in an aromatic stacking interaction arated by gel electrophoresis and visualized by silver staining. with the imidazole of H57. One phosphonate oxygen is Several neutrophil proteins were precipitated by both KBA- H-bonded with N␧ of H57, another oxygen is H-bonded to N␧ Seph and Seph. However, a 26-kDa protein was significantly en- of K192, and the third is H-bonded to the backbone NH of G193 riched in KBA-Seph precipitates (Fig. 1C). This protein band was and O␥ of S195. The carboxamido-N-(naphthalene-2-carboxyl)pi- excised, in-gel digested by trypsin, and subjected to MALDI-TOF- peridine segment of JNJ-10311795 occupies the hydrophobic MS. Analysis of the obtained peptide fragments using a peptide S3/S4 cavity, which is defined by the side chains of Y215, I99, and sequence data base revealed that the peptides matched (sequence F172 (Fig. 2). Docking of BAs into the same box resulted in av- coverage of 63%) to the primary amino acid sequence of the hu- erage docking scores for AKBA of 26.12 Ϯ 0.32, KBA of 28.38 Ϯ man neutral serine protease catG (25.8 kDa, Fig. 1D). Immuno- 0.25, A␤-BA of 24.83 Ϯ 0.36, and ␤-BA of 26.51 Ϯ 0.21, with an logical analysis by Western blot confirmed the identity of catG, apparently identical binding mode occupying the same part (S1) of and also isolated catG bound to KBA-Seph (Fig. 1, E and F), the binding site as JNJ-10311795. For all BAs, we observed only

FIGURE 2. Rigid automated molecular docking of BAs into catG. Proposed binding modes of BAs and the controls to catG. A, JNJ-10311795, (B) ␤-BA, (C)A␤-BA, (D) KBA, (E) AKBA, and (F) ␣-amyrin. Indicated by dashed lines are the molecular interactions formed with the distinct amino acids of catG (Protein Data Bank: 1T32, shown in surface representation, light blue). 3438 IDENTIFICATION OF CATHEPSIN G AS TARGET OF BOSWELLIC ACIDS

ﬁned and inconsistent binding patterns were obtained (Fig. 3C), presumably originating from concentration-dependent aggregation or superstoichiometric binding of Bas, thus precluding determination of detailed binding kinetics.

Boswellic acids inhibit the proteolytic activity of catG To investigate if the interaction of BAs with catG may have a functional consequence, we first analyzed whether BAs modulate the proteolytic activity of catG. All BAs investigated potently inhibited isolated catG, whereas the related ␣-amyrin and ursolic acid showed no significant inhibition (Fig. 4A). Also glut-KBA was effective, indicating that esterification of the C3-OH by glutaric acid is not detrimental. Concentration-response studies (Fig. ␮ ␮ ␤ 4B) revealed IC50 values of 0.6 M for AKBA, 0.8 M for -BA, 1.2 ␮M for A␤-BA, and 3.7 ␮M for KBA. The potency of BAs was reduced by increasing the peptide substrate concentration, and Lineweaver-Burk blots imply a competitive inhibition of catG by A␤-BA (Fig. 4C). Moreover, wash-out experiments support a reversible inhibition of catG by BAs (Fig. 4D). FIGURE 3. Assessment of binding characteristics of BAs to catG. A, Isolated human catG (1 ␮gin500␮l final volume) was incubated with Effects of BAs on related serine proteases Seph or KBA-Seph in the presence or absence of 1 mM ␤-BA or 10 ␮M The effects of BAs on the activity of the closely related serine JNJ-10311795 as indicated. After extensive washing procedures, catG was proteases chymase, tryptase, HLE, proteinase-3, and chymotrypsin analyzed in the precipitates by Western blotting. Similar results were ob- were determined under individually optimized, but still compara- tained in at least three additional experiments, each. B and C, SPR spec- ble, experimental conditions. AKBA and A␤-BA (10 ␮M, each) trometry analysis of the binding of ␤-BA to catG. B, Responses for the catG inhibitor JNJ-10311795 (0.5 ␮M), ␤-BA (10 ␮M), and ␣-amyrin (10 failed to significantly inhibit chymase and tryptase, but slightly ␮M). C, Dilution series of ␤-BA (2, 4, 8, or 16 ␮M). Similar results were suppressed HLE and proteinase-3 activity (IC50 values of AKBA obtained in at least three additional experiments. and A␤-BA Ն30 ␮M) (Fig. 5A). A␤-BA inhibited chymotrypsin ␮ with an IC50 of 4.8 M (Fig. 5B). Also for JNJ-10311795 it was found before that among various serine proteases, chymotrypsin is one binding mode with RMSD values between 0.44 Ϯ 0.06 and rather sensitive against this compound (Ki of 490 nM) (28). To- ␤ 0.55 Ϯ 0.08 Å over all docking solutions (Table I). From the gether, apart from the inhibitory effect of A -BA on chymotrypsin, ligand poses, one can speculate that the contributions to the scores BAs seem to be rather selective for catG. are mainly based on lipophilic and H-bonding interactions (Table I). A␤-BA forms an H bond between the acetyl oxygen and the Boswellic acids inhibit neutrophil chemoinvasion and 2ϩ side chain hydroxyl of S40, and ␤-BA can be H-bonded with its catG-mediated Ca mobilization in human platelets hydroxyl group and the backbone carbonyl oxygen of H57, similar Once released into the plasma, catG may contribute to migration/ to KBA, with the latter forming also an H bond between the ketone invasion of neutrophils along chemoattractant gradients by degrad- oxygen and the backbone NH of G193 and the side chain hydroxyl ing extracellular matrix proteins (13). The chemoattractant fMLP of S195. In our model, AKBA forms H bonds between its acetyl increased (4.1-fold) chemoinvasion of human neutrophils through oxygen and the side chain hydroxyl of S40, and between the ke- Matrigel (assessed by modified Boyden chamber assay), which tone oxygen and the side chain hydroxyl of S195. Docking of was efficiently suppressed by JNJ-10311795 (0.1 ␮M) as well as ␣ ␤ -amyrin as a negative, nonbinding control yielded a ChemScore by AKBA or A -BA (Fig. 6A), with IC50 values of 2.7 and 4.5 of 27.36 Ϯ 0.25, with two different docking poses detected with ␮M, respectively (Fig. 6C). In contrast, the elastase inhibitor IV RMSD values of 2.36 Ϯ 0.35 Å over all solutions (Fig. 2). (0.2 ␮M) failed in this respect (data not shown). On the other hand, in the absence of Matrigel, neither JNJ-10311795 nor AKBA or Analysis of binding characteristics of BAs to catG A␤-BA reduced the chemotactic response of neutrophils toward To gain more insights into the specificity of the BA-binding to fMLP (Fig. 6A). Furthermore, preincubation of neutrophils with catG, we tested whether an excess of either soluble ␤-BA or BAs did not result in an increased chemokinetic activity in the JNJ-10311795 could prevent binding of the isolated protease to absence of fMLP (data not shown). immobilized KBA. As shown in Fig. 3A, JNJ-10311795 (10 ␮M) CatG, released from challenged neutrophils, also stimulates clearly antagonized catG-binding from KBA-Seph, supporting the platelets for aggregation and secretion by cleavage of the protease- hypothesis that BAs bind to the same pocket of catG as activated receptor-4, accompanied by mobilization of Ca2ϩ (19). JNJ-10311795. Also, soluble ␤-BA at 1 mM reduced catG-binding Addition of isolated catG (0.1 nmol/ml) to human washed platelets to immobilized BAs. Application of higher concentrations (Ͼ1 caused a marked but transient Ca2ϩ mobilization that was essen- mM) of BAs or analysis of ␣-amyrin was infeasible due to insuf- tially prevented by preincubation of platelets with AKBA (5 ␮M) ficient aqueous solubility of the compounds. or JNJ-10311795 (0.2 ␮M) to the same extent (Fig. 6D). In con- Next, we attempted to characterize the interaction of BAs with trast, AKBA (5 ␮M) failed to inhibit Ca2ϩ mobilization in plate- catG by SPR spectroscopy using a Biacore X device. The data lets activated by 0.5 U/ml thrombin (not shown), excluding general obtained indicate a reversible interaction between catG (immobi- suppression of Ca2ϩ mobilization by AKBA. In another type of lized) and BAs or JNJ-10311795 (used as analytes), whereas the experiment, fMLP caused no Ca2ϩ mobilization in Fura-2-loaded ␣-amyrin (negative control) does not interact with catG (Fig. 3B). platelets unless unloaded neutrophils were coincubated. This Ca2ϩ However, at higher BA concentrations (Ͼ8–10 ␮M), only unde- mobilization was strongly suppressed by 0.1 ␮M JNJ-10311795 or The Journal of Immunology 3439

FIGURE 4. BAs inhibit the proteolytic activity of catG. A, Isolated catG (0.2 ␮g) from neutrophils was preincubated with the test compounds (10 ␮M; except JNJ-10311795, 0.5 ␮M) or vehicle (veh., DMSO) for 20 min at 25°C. Then, Suc-AAPF-pNA (1 mM ﬁnal concentration in 200 ␮l of ﬁnal volume) was added to start the reaction, and the absorbance was measured at 410 nm. The enzyme activity was determined by the progress curve method. Results p Ͻ 0.001. B, Concentration-response curves ,ءءء ;are presented as percentage of the control (veh., DMSO), and data are given as means Ϯ SE, n ϭ 4–6 of BAs on catG activity. Inhibition of the activity of catG (0.2 ␮g) by BAs was determined as described above. Data are given as means Ϯ SE, n ϭ 4; p Ͻ 0.001. C, Kinetic analysis of the inhibition of catG (0.2 ␮g) by 10 ␮MA␤-BA and 0.1 ␮M JNJ-10311795. Data are given as ,ءءء p Ͻ 0.01 and ,ءء means (n ϭ 3–4) and results are presented as Lineweaver-Burke plots. The catG substrate concentrations were 0.1, 0.2, 0.3, 0.5, 1, 2, and 4 mM. D, Reversibility of catG inhibition. CatG (0.2 ␮g) was incubated with 0.5 or 10 ␮MA␤-BA or vehicle (DMSO) for 5 min at room temperature, each. Then, one aliquot of the sample containing 10 ␮MA␤-BA was diluted with assay buffer 20-fold, whereas the other aliquot was not altered, and catG substrate (1 mM) was added to each sample to start the reaction. The absorbance was measured at 410 nm, and catG activity was determined by the progress curve .p Ͻ 0.01 ,ءء ;method. Results are presented as percentage of the control (DMSO), and data are given as means Ϯ SE, n ϭ 4

3 ␮M AKBA (Fig. 6E). Together, BAs are able to block cellular ease), a single dose of 800 mg orally administered B. serrata gum functions by inhibiting catG. extract to healthy volunteers (phase I) transiently reduced catG activity ex vivo in cytochalasin B/fMLP-stimulated blood (Fig. Inhibition of catG by B. serrata gum extracts ex vivo 7A). In the phase II–III study, catG activity was ﬁrst measured in In a clinical trial (i.e., on the safety and effectiveness of frankin- plasma prepared from venous blood of Crohn’s disease patients cense extract in healthy volunteers and patients with Crohn’s dis- before medication. Four weeks after oral intake of 3 ϫ 800 mg/day

FIGURE 5. Effects of BAs on the activity of various serine proteases. A,A␤-BA and AKBA (10 ␮M, each) or vehicle (DMSO) were preincubated with the serine proteases (20 min, 25°C) and the activities of the respective proteases were assayed under standard conditions as described in Materials and ,p Ͻ 0.001. Prot-3 ,ءءء p Ͻ 0.01 and ,ءء ;Methods. Results are presented as percentage of the control (vehicle), and data are given as means Ϯ SE, n ϭ 4–5 proteinase-3; chymo., chymotrypsin. B, Concentration-response curves of A␤-BA and AKBA on the activities of HLE, proteinase-3, and chymotrypsin. .p Ͻ 0.001 ,ءءء ;p Ͻ 0.01 ,ءء ;p Ͻ 0.05 ,ء ;Data are given as means Ϯ SE, n ϭ 3–4 3440 IDENTIFICATION OF CATHEPSIN G AS TARGET OF BOSWELLIC ACIDS

FIGURE 6. BAs suppress catG-mediated functional cellular responses. A, Inhibition of chemoattractant-induced neutrophil invasion. Neutrophils, pretreated with BAs (10 ␮M, each), JNJ-10311795 (JNJ, 0.1 ␮M), or vehicle (veh., DMSO), were placed on the upper chamber of a two-compartment Boyden chamber. Cells that migrate through Matrigel-coated pore-size filters in the lower chamber containing buffer (negative control) or 0.1 ␮M fMLP within 40 min were fixed, stained with Gram’s Violet, and the absorbance was measured at 570 and 620 nm, respectively. Results are presented as fold p Ͻ 0.05. B, Effects of BAs and ,ء ;increase of the number of stimulated migrated cells vs vehicle-treated cells. Data are given as means Ϯ SE, n ϭ 5 JNJ-10311795 on neutrophil chemotaxis. The experimental settings were the same as above, except that pore-size filters were uncoated (no Matrigel). Data are given as means Ϯ SE, n ϭ 3. C, Concentration-response for A␤-BA and AKBA to inhibit neutrophil chemoinvasion. Migrated cells stimulated with fMLP were set to 100%. Data are given as means Ϯ SE, n ϭ 4. D and E, AKBA inhibits Ca2ϩ mobilization in platelets induced by catG or by fMLP-stimulated neutrophils. D, Fura-2-loaded platelets (108) were incubated with AKBA (5 ␮M), JNJ-10311795 (JNJ, 0.2 ␮M), or vehicle (veh., DMSO), and after 5 min the measurement of intracellular Ca2ϩ concentration was started. After 30 s, 0.1 nmol/ml isolated catG was added and the fluorescence was recorded for another 100 s. E, Fura-2-loaded platelets (108) were mixed with 107 unloaded neutrophils and incubated with AKBA (3 ␮M), JNJ-10311795 (JNJ, 0.1 ␮M), or vehicle (veh., DMSO), and after 5 min the measurement of intracellular Ca2ϩ concentration was started. After 30 s, 0.1 ␮M fMLP was added and the fluorescence was recorded for another 100 s. Curves are representative for at least four experiments.

extract (representing the steady-state plasma levels of BA, Table II) or of placebo, respectively, catG activity was analyzed in plasma again. CatG activity of patients receiving placebo was not significantly different during treatment, whereas a clear and significant reduction of catG activity in the plasma of patients that received frankincense extract was evident (Fig. 7B). No reduced activity of HLE or proteinase-3 in the plasma was immediately apparent (our unpublished data). Note that neutrophil numbers in the blood of verum-treated patients did not significantly change during treatment (6.9 Ϯ 1.1 cells/nl before treatment and 6.7 Ϯ 0.6 cells/nl after treatment).

Discussion Herein we provide evidence that catG is a functional target of BAs FIGURE 7. Treatment with frankincense extracts lowers catG activity in human subjects. A, Phase I trial. Healthy male volunteers received two with pharmacological relevance. First, catG was identified as a capsules containing B. serrata gum extract (PS 0201Bo) as single dose Table II. Steady-state concentrations of BAs in the plasma of patients (800 mg) application. Venous blood was taken at the indicated time points, a promptly stimulated with 10 ␮M cytochalasin B and 2.5 ␮M fMLP for 5 treated with B. serrata gum extract PS0201Bo min, and plasma was prepared and aliquots of the plasma were immediately analyzed for catG activity as described. Data are shown as means Ϯ SE, Boswellic acid Plasma Concentration (␮M) n ϭ 12 volunteers. B, Phase II–III trial. Blood was taken from Crohn’s ␤-BA 6.35 Ϯ 1.0 disease patients prior medication (100% value) as well as after 4 wk of A␤-BA 4.90 Ϯ 0.5 continuous administration of either 800 mg of PS 0201Bo or placebo three KBA 0.33 Ϯ 0.1 times each day (total 2400 mg/day). After venipuncture, the blood was AKBA 0.04 Ϯ 0.01 promptly stimulated with 10 ␮M cytochalasin B and 2.5 ␮M fMLP for 5 a min, plasma was prepared, and catG activity was assessed as described BAs were analyzed by LC-ESI-MS/MS in plasma prepared from venous blood Ϯ ϭ ϭ obtained from Crohn’s disease patients 4 wk after daily oral administration of 800 mg above. Data are given as means SE, n 5 (placebo) and n 3 (verum), of B. serrata extract (PS0201Bo) or placebo three times per day (total, 2400 mg/day). .p Ͻ 0.05. Values are given as means Ϯ SE, n ϭ 3 patients ,ء ;triplicates, each The Journal of Immunology 3441 selective BA-binding protein from neutrophils in a target-fishing We showed that BAs and JNJ-10311795 bound to catG with a assay using a BA-affinity Sepharose matrix. Because isolated catG similar binding mode, and in neutrophil/platelet coincubation ex- bound to this affinity matrix, BAs may bind directly to catG and periments, AKBA as well as JNJ-10311795 suppressed the in- not via a linker protein. Moreover, BAs docked into the active site crease in intracellular Ca2ϩ concentration of platelets evoked by of catG in a similar fashion as for the synthetic catG inhibitor catG or by catG released from fMLP-challenged neutrophils. Sim- JNJ-10311795 (28), and JNJ-10311795 antagonized catG-binding ilarly, BAs as well as JNJ-10311795 blocked chemoinvasion of to immobilized BAs. Also, SPR studies indicate a reversible in- neutrophils in response to fMLP (mediated by catG), but both teraction between BAs and catG. Second, BAs potently inhibit compounds failed to affect chemotaxis. Comparisons of pharma- catG activity with partially submicromolar IC50 values in a com- cological actions of BAs and of JNJ-10311795 in vivo reported in petitive and reversible manner. Third, BAs suppress catG-medi- the literature indicate that the substances exhibit related properties. ated cellular responses such as fMLP-induced chemoinvasion of Thus, JNJ-10311795 efficiently inhibited glycogen-induced rat neutrophils, as well as Ca2ϩ mobilization in platelets evoked by peritonitis and blocked LPS-induced rat airway neutrophilia (28). isolated catG or by fMLP-stimulated neutrophils. Fourth, studies Similarly, mixtures of BAs or frankincense extracts impaired car- with human subjects demonstrate a significant suppression of catG rageenan-induced pleurisy in rats accompanied by reduced neutro- activity ex vivo after oral administration of B. serrata gum extracts vs phil infiltration (40). Moreover, the increases in the pro-inflamma- placebo. Finally, catG may play roles in inflammatory diseases (i.e., tory cytokines TNF-␣ and IL-1␤ in inflamed rat tissues were rheumatoide arthritis, bronchial asthma) (11, 21, 36), where frankin- reversed by JNJ-10311795 (28) as well as by frankincense (1). cense extracts showed beneficial effects in several studies (2). CatG may play a role in the pathogenesis of various inflamma- Inhibition of a serine protease, that is, HLE by AKBA, was tory diseases (11, 28, 41), and catG inhibitors were suggested to shown before, but significantly higher concentrations (IC50 of 15 have potential for treating asthma, psoriasis, and rheumatoid ar- ␮M) were required (4) as compared with catG. We found that BAs thritis (11, 21), for which frankincense extracts were proposed to inhibited all three neutrophil serine proteases from azurophilic have therapeutic benefit (1, 2). Nevertheless, experiments with granules (catG, HLE, and proteinase-3) that have an overlapping catG-deficient mice led to controversial results regarding the im- range of extracellular substrates, with catG being the most affine portance of catG in chemoinvasion and in inflammatory processes. target. The successful identification of catG as BA target by the For example, MacIvor et al. showed that neutrophils from catGϪ/Ϫ affinity fishing approach is probably the result of the combination mice displayed normal phagocytosis, superoxide production, and of several fortunate circumstances. Thus, the relatively high affin- normal chemotactic responses to C5a, fMLP, and IL-8 (42). In ity of catG to KBA-Seph enabled binding under stringent condi- contrast, others found reduced neutrophil infiltration, myeloperox- tions and endured the thorough washing of the precipitates. More- idase, and chemoattractants (CXCL1 and CXCL2) in ischemic kid- over, the C3-OH moiety chosen to link KBA is no pharmacophore neys of catG-deficient mice (43). Also, mice lacking catG and since esterification with glutaric acid did not hamper the interfer- HLE failed to initiate cytoskeletal reorganization and cell spread- ence with catG or docking into the active center. In fact, the es- ing, and they exhibited severe defects in the release of MIP-2 and terified analogs of KBA (i.e., AKBA and Glut-KBA) were even reactive oxygen intermediates, and exogenously added, proteolyti- somewhat superior in inhibiting catG, and also for other targets, cally active catG largely restored these defects (18). However, no AKBA is frequently more potent than KBA (37–39). Finally, BAs catG inhibitor is available for therapeutic use, and data from clin- are highly rigid molecules, thus favoring tight binding to catG. ical trials with respective candidates are still missing. Note that besides catG, no other protein from neutrophils was ob- In conclusion, we provide evidence for catG as a functional viously enriched in KBA-Seph pull-downs vs Seph precipitates. target of BAs in vivo. The potent interference of ␤-BA and A␤-BA In the docking studies, BAs bound to the active center in catG with catG in relationship to their high achievable plasma levels as JNJ-10311795 with a similar binding mode blocking the S1 favors this interaction as a possible molecular basis underlying pocket. It is reasonable to assume that the interactions are mainly certain beneficial effects of frankincense observed in animal mod- hydrophobic together with H bonds formed between the ketone els of inflammation, as well as in human subjects suffering from oxygen and/or the hydroxyl group of the BAs and several amino inflammatory disorders. However, the importance and value of acids of catG. Notably, these are identical with the essential inter- catG as a therapeutic target in chronic inflammatory diseases has actions of JNJ-10311795. We showed experimentally that not been entirely assessed and is controversially discussed. On the ␣-amyrin, which differs from ␤-BA only in the carboxyl group and other hand, the knowledge about the functionality of BAs as catG the configuration of the C3-OH, does not bind to catG or inhibit its inhibitors may offer new possibilities for use of frankincense in activity. Even though this is not reflected in the docking scores, additional inflammatory diseases, for example, in chronic obstruc- one explanation might be that the carboxyl group of the BAs tive pulmonary disease, where catG may contribute, but also in shields the ligand from the solvent and this entropic contribution acute inflammatory states. Finally, investigations with defined BAs might be crucial for binding. Entropic contributions are not in- or with structurally optimized derivatives in animals and humans cluded in the scoring function of the docking program. This hy- may reveal the therapeutic potential and suitability of single sub- pothesis is supported by the fact that the carboxyl group of all BAs stances in vivo. always points toward the solvent in all docking solutions, never into the protein binding pocket. Acknowledgments Although AKBA and KBA have been considered as the phar- We thank Bianca Jazzar for expert technical assistance. macologically relevant BAs (1, 2), we find that BAs lacking the C11-oxo moiety are about equipotent to 11-keto-BAs to interfere Disclosures with catG. In view of the high plasma levels of ␤-BA/A␤-BA (e.g., The authors have no financial conflicts of interest. 2.4 - 10.1 ␮M) as compared with the fairly low levels of KBA and AKBA (Յ0.34 and 0.1 ␮M, respectively) achieved after oral ad- References ministration of frankincense preparations (Ref. 6 and this study), 1. Poeckel, D., and O. Werz. 2006. Boswellic acids: biological actions and molec- ␤ ␤ ular targets. Curr. Med. Chem. 28: 3359–3369. the interaction of -BA/A -BA with catG might be of pharmaco- 2. Ammon, H. P. 2006. Boswellic acids in chronic inflammatory diseases. Planta logical relevance. Med. 72: 1100–1116. 3442 IDENTIFICATION OF CATHEPSIN G AS TARGET OF BOSWELLIC ACIDS

3. Siemoneit, U., B. Hofmann, N. Kather, T. Lamkemeyer, J. Madlung, L. Franke, lets independent of Ca2ϩ and differentially interact with platelet-type 12-lipoxy- G. Schneider, J. Jauch, D. Poeckel, and O. Werz. 2008. Identification and func- genase. Mol. Pharmacol. 70: 1071–1078. tional analysis of cyclooxygenase-1 as a molecular target of boswellic acids. 24. Rosenfeld, J., J. Capdevielle, J. C. Guillemot, and P. Ferrara. 1992. In-gel diges- Biochem. Pharmacol. 75: 503–513. tion of proteins for internal sequence analysis after one- or two-dimensional gel 4. Safayhi, H., B. Rall, E. R. Sailer, and H. P. Ammon. 1997. Inhibition by bo- electrophoresis. Anal. Biochem. 203: 173–179. swellic acids of human leukocyte elastase. J. Pharmacol. Exp. Ther. 281: 25. Shevchenko, A., M. Wilm, O. Vorm, and M. Mann. 1996. Mass spectrometric 460–463. sequencing of proteins silver-stained polyacrylamide gels. Anal. Chem. 68: 5. Syrovets, T., B. Buchele, C. Krauss, Y. Laumonnier, and T. Simmet. 2005. 850–858. Acetyl-boswellic acids inhibit lipopolysaccharide-mediated TNF-␣ induction in 26. Corvey, C., P. Koetter, T. Beckhaus, J. Hack, S. Hofmann, M. Hampel, T. Stein, monocytes by direct interaction with I␬B kinases. J. Immunol. 174: 498–506. M. Karas, and K. D. Entian. 2005. Carbon source-dependent assembly of the 6. Buchele, B., and T. Simmet. 2003. Analysis of 12 different pentacyclic triterpenic Snf1p kinase complex in Candida albicans. J. Biol. Chem. 280: 25323–25330. acids from frankincense in human plasma by high-performance liquid chroma- 27. Jones, G., P. Willett, and R. C. Glen. 1995. Molecular recognition of receptor tography and photodiode array detection. J. Chromatogr. B Analyt. Technol. sites using a genetic algorithm with a description of desolvation. J. Mol. Biol. Biomed. Life Sci. 95: 355–362. 245: 43–53. 7. Siemoneit, U., C. Pergola, B. Jazzar, H. Northoff, C. Skarke, J. Jauch, and 28. de Garavilla, L., M. N. Greco, N. Sukumar, Z. W. Chen, A. O. Pineda, O. Werz. 2009. On the interference of boswellic acids with 5-lipoxygenase: F. S. Mathews, E. Di Cera, E. C. Giardino, G. I. Wells, B. J. Haertlein, et al. 2005. mechanistic studies in vitro and pharmacological relevance. Eur. J. Pharmacol. A novel, potent dual inhibitor of the leukocyte proteases cathepsin G and chy- 606: 246–254. mase: molecular mechanisms and anti-inflammatory activity in vivo. J. Biol. 8. Burdine, L., and T. Kodadek. 2004. Target identification in chemical genetics: the Chem. 280: 18001–18007. (often) missing link. Chem. Biol. 11: 593–597. 29. Berman, H. M., J. Westbrook, Z. Feng, G. Gilliland, T. N. Bhat, H. Weissig, 9. Taunton, J., C. A. Hassig, and S. L. Schreiber. 1996. A mammalian histone I. N. Shindyalov, and P. E. Bourne. 2000. The Protein Data Bank. Nucleic Acids deacetylase related to the yeast transcriptional regulator Rpd3p. Science 272: Res. 28: 235–242. 408–411. 30. Ponder, J. W., and D. A. Case. 2003. Force fields for protein simulations. Adv. 10. Harding, M. W., A. Galat, D. E. Uehling, and S. L. Schreiber. 1989. A receptor Protein Chem. 66: 27–85. for the immunosuppressant FK506 is a cis-trans peptidyl-prolyl isomerase. Na- 31. Halgren, T. 1996. Merck molecular force field 1: basis, form, scope, parameter- ture 341: 758–760. ization, and performance of MMFF94. J. Comput. Chem. 17: 490–519. 11. Pham, C. T. 2006. Neutrophil serine proteases: specific regulators of inflamma- 32. Eldridge, M. D., C. W. Murray, T. R. Auton, G. V. Paolini, and R. P. Mee. 1997. tion. Nat. Rev. Immunol. 6: 541–550. Empirical scoring functions, I: The development of a fast empirical scoring func- 12. Campbell, E. J. 1982. Human leukocyte elastase, cathepsin G, and lactoferrin: tion to estimate the binding affinity of ligands in receptor complexes. J. Comput. family of neutrophil granule glycoproteins that bind to an alveolar macrophage Aided Mol. Des. 11: 425–445. receptor. Proc. Natl. Acad. Sci. USA 79: 6941–6945. 33. DeLano, W. L. 2002. The PyMOL Molecular Graphics System (2002). DeLano 13. Capodici, C., and R. A. Berg. 1989. Cathepsin G degrades denatured collagen. Scientific, San Carlos, CA; www.pymol.org. Inflammation 13: 137–145. 34. Grynkiewicz, G., M. Poenie, and R. Y. Tsien. 1985. A new generation of Ca2ϩ 14. Heck, L. W., W. D. Blackburn, M. H. Irwin, and D. R. Abrahamson. 1990. indicators with greatly improved fluorescence properties. J. Biol. Chem. 260: Degradation of basement membrane laminin by human neutrophil elastase and 3440–3450. cathepsin G. Am. J. Pathol. 136: 1267–1274. 35. Steadman, R., P. L. St John, R. A. Evans, G. J. Thomas, M. Davies, L. W. Heck, 15. Berahovich, R. D., Z. Miao, Y. Wang, B. Premack, M. C. Howard, and and D. R. Abrahamson. 1997. Human neutrophils do not degrade major basement T. J. Schall. 2005. Proteolytic activation of alternative CCR1 ligands in inflam- membrane components during chemotactic migration. Int. J. Biochem. Cell Biol. mation. J. Immunol. 174: 7341–7351. 29: 993–1004. 16. Richter, R., R. Bistrian, S. Escher, W. G. Forssmann, J. Vakili, R. Henschler, 36. Miyata, J., K. Tani, K. Sato, S. Otsuka, T. Urata, B. Lkhagvaa, C. Furukawa, N. Spodsberg, A. Frimpong-Boateng, and U. Forssmann. 2005. Quantum pro- N. Sano, and S. Sone. 2007. Cathepsin G: the significance in rheumatoid arthritis teolytic activation of chemokine CCL15 by neutrophil granulocytes modulates as a monocyte chemoattractant. Rheumatol. Int. 27: 375–382. mononuclear cell adhesiveness. J. Immunol. 175: 1599–1608. 37. Altmann, A., L. Fischer, M. Schubert-Zsilavecz, D. Steinhilber, and O. Werz. 17. Chertov, O., H. Ueda, L. L. Xu, K. Tani, W. J. Murphy, J. M. Wang, 2002. Boswellic acids activate p42MAPK and p38 MAPK and stimulate Ca2ϩ O. M. Howard, T. J. Sayers, and J. J. Oppenheim. 1997. Identification of human mobilization. Biochem. Biophys. Res. Commun. 290: 185–190. neutrophil-derived cathepsin G and azurocidin/CAP37 as chemoattractants for 38. Poeckel, D., L. Tausch, S. George, J. Jauch, and O. Werz. 2006. 3-O-acetyl-11- mononuclear cells and neutrophils. J. Exp. Med. 186: 739–747. keto-boswellic acid decreases basal intracellular Ca2ϩ levels and inhibits agonist- 18. Raptis, S. Z., S. D. Shapiro, P. M. Simmons, A. M. Cheng, and C. T. Pham. 2005. induced Ca2ϩ mobilization and mitogen-activated protein kinase activation in Serine protease cathepsin G regulates adhesion-dependent neutrophil effector human monocytic cells. J. Pharmacol. Exp. Ther. 316: 224–232. functions by modulating integrin clustering. Immunity 22: 679–691. 39. Sailer, E. R., L. R. Subramanian, B. Rall, R. F. Hoernlein, H. P. Ammon, and 19. Sambrano, G. R., W. Huang, T. Faruqi, S. Mahrus, C. Craik, and S. R. Coughlin. H. Safayhi. 1996. Acetyl-11-keto-␤-boswellic acid (AKBA): structure require- 2000. Cathepsin G activates protease-activated receptor-4 in human platelets. ments for binding and 5-lipoxygenase inhibitory activity. Br. J. Pharmacol. 117: J. Biol. Chem. 275: 6819–6823. 615–618. 20. Sun, R., P. Iribarren, N. Zhang, Y. Zhou, W. Gong, E. H. Cho, S. Lockett, 40. Sharma, M. L., A. Khajuria, A. Kaul, S. Singh, G. B. Singh, and C. K. Atal. 1988. O. Chertov, F. Bednar, T. J. Rogers, et al. 2004. Identification of neutrophil Effect of salai guggal ex-Boswellia serrata on cellular and humoral immune granule protein cathepsin G as a novel chemotactic agonist for the G protein- responses and leucocyte migration. Agents Actions 24: 161–164. coupled formyl peptide receptor. J. Immunol. 173: 428–436. 41. Wiedow, O., and U. Meyer-Hoffert. 2005. Neutrophil serine proteases: potential 21. Maryanoff, B. E. 2004. Inhibitors of serine proteases as potential therapeutic key regulators of cell signalling during inflammation. J. Intern. Med. 257: agents: the road from thrombin to tryptase to cathepsin G. J. Med. Chem. 47: 319–328. 769–787. 42. MacIvor, D. M., S. D. Shapiro, C. T. Pham, A. Belaaouaj, S. N. Abraham, and 22. Jauch, J., and J. Bergmann. 2003. An efficient method for the large-scale prep- T. J. Ley. 1999. Normal neutrophil function in cathepsin G-deficient mice. Blood aration of 3-O-acetyl-11-oxo-␤-boswellic acid and other boswellic acids. Eur. 94: 4282–4293. J. Org. Chem. 4752–4756. 43. Shimoda, N., N. Fukazawa, K. Nonomura, and R. L. Fairchild. 2007. Cathepsin 23. Poeckel, D., L. Tausch, N. Kather, J. Jauch, and O. Werz. 2006. Boswellic acids G is required for sustained inflammation and tissue injury after reperfusion of stimulate arachidonic acid release and 12-lipoxygenase activity in human plate- ischemic kidneys. Am. J. Pathol. 170: 930–940.