Human Cathelicidin LL-37 Is a Chemoattractant for Eosinophils and Neutrophils That Acts Via Formyl-Peptide Receptors

Home , Cathelicidin, Major basic protein

Original Paper

Int Arch Allergy Immunol 2006;140:103–112 Received: January 3, 2005 Accepted after revision: December 19, 2005 DOI: 10.1159/000092305 Published online: March 24, 2006

Human Cathelicidin LL-37 Is a Chemoattractant for Eosinophils and Neutrophils That Acts via Formyl-Peptide Receptors

a a b G. Sandra Tjabringa Dennis K. Ninaber Jan Wouter Drijfhout a a Klaus F. Rabe Pieter S. Hiemstra

a b Departments of Pulmonology, and Immunohematology and Blood Transfusion, Leiden University Medical Center, Leiden , The Netherlands

Key Words ting using antibodies directed against phosphorylated Eosinophils Neutrophils Chemotaxis Antimicrobial ERK1/2. Results: Our results show that LL-37 chemoat- peptides Innate immunity Lung infl ammation tracts both eosinophils and neutrophils. The FPR antagonistic peptide tBoc-MLP inhibited LL-37-induced chemotaxis. Whereas the FPR agonist fMLP activated ERK1/2 Abstract in neutrophils, LL-37 did not, indicating that fMLP and Background: Infl ammatory lung diseases such as asth- LL-37 deliver different signals through FPRs. Conclu- ma and chronic obstructive pulmonary disease (COPD) sions: LL-37 displays chemotactic activity for eosinophils are characterized by the presence of eosinophils and and neutrophils, and this activity is mediated via an FPR. neutrophils. However, the mechanisms that mediate the These results suggest that LL-37 may play a role in in- infl ux of these cells are incompletely understood. Neu- fl ammatory lung diseases such as asthma and COPD. trophil products, including neutrophil elastase and anti- Copyright © 2006 S. Karger AG, Basel microbial peptides such as neutrophil defensins and LL- 37, have been demonstrated to display chemotactic activity towards cells from both innate and adaptive im- Introduction munity. However, chemotactic activity of LL-37 towards eosinophils has not been reported. Therefore, the aim of Eosinophils and neutrophils play an important role in the present study was to investigate the chemotactic ac- infection and infl ammation, and increased numbers have tivity of LL-37 for eosinophils and to explore the mecha- been demonstrated in various infl ammatory diseases, in- nisms involved in LL-37-mediated attraction of neutro- cluding asthma and chronic obstructive pulmonary dis- phils and eosinophils. Methods: Neutrophils and eo- ease (COPD). At the site of infl ammation, these granulo- sinophils were obtained from venous blood of healthy cytes contribute to the infl ammatory process by releasing donors. Chemotaxis was studied using a modifi ed Boy- a range of mediators including proteinases, antimicrobial den chamber technique. Involvement of formyl-peptide peptides, lipid mediators and reactive oxygen intermedi- receptors (FPRs) was studied using the antagonistic pep- ates. In addition to causing tissue injury, granulocyte tide tBoc-MLP. Activation of the mitogen-activated pro- products may contribute to the recruitment of infl amma- tein kinase (MAPK) ERK1/2 was studied by Western blot- tory cells. This is illustrated by recent studies demonstrat-

© 2006 S. Karger AG, Basel Correspondence to: Pieter S. Hiemstra 1018–2438/06/1402–0103$23.50/0 Department of Pulmonology, Leiden University Medical Center Fax +41 61 306 12 34 Building 1, C3-P, PO Box 9600 E-Mail [email protected] Accessible online at: NL–2300 RC Leiden (The Netherlands) www.karger.com www.karger.com/iaa Tel. +31 71 526 2950, Fax +31 71 526 6927, E-Mail [email protected] ing that neutrophil-derived antimicrobial peptides like mediate LL-37 induced processing and release of IL- , -defensins and LL-37 mediate chemotaxis of infl amma- but its involvement in chemotactic responses has not yet tory cells via activation of chemokine and formyl-peptide been reported. receptors (FPRs) [1, 2]. Therefore, the aim of the present study was to investi- LL-37 is the C-terminal active part of the only human gate whether LL-37 displays chemotactic activity towards cathelicidin that has been identifi ed to date, hCAP-18 [for human eosinophils, and to compare the mechanisms in- recent reviews, see 3, 4]. Cathelicidins are a family of an- volved in eosinophil and neutrophil chemotaxis by LL- timicrobial peptides that consist of a conserved N-termi- 37. Our results show that LL-37 chemoattracts both eo- nal cathelin domain and a variable C-terminal antimicro- sinophils and neutrophils via FPRs. In addition, we dem- bial domain, which can be cleaved off by proteinases [5, onstrated that formoterol, a 2 -adrenoceptor agonist 6]. Expression of hCAP-18/LL-37 has been demonstrated widely used in the treatment of infl ammatory lung dis- in neutrophils [7] , monocytes, T cells [8] , keratinocytes ease, inhibits LL-37-induced chemotaxis. Finally, activa- in infl amed skin [9] and in various squamous epithelia tion of the mitogen-activated protein kinase (MAPK) [10]. In the airways, hCAP-18/LL-37 expression was ERK1/2 by fMLP and LL-37 was studied. While both shown in surface epithelia of the conducting airways and fMLP and LL-37 were shown to chemoattract neutrophils in the submucosal glands [11], and hCAP-18/LL-37 was and eosinophils, only fMLP induced activation of the demonstrated in broncheoalveolar lavage fl uid [12] . LL- MAPK ERK1/2. The results from this study suggest that 37 was originally identifi ed as an antimicrobial peptide LL-37 may play a role in infl ammatory lung diseases such [13], and different studies have suggested that LL-37 as asthma and COPD. plays an important role in host defense [14, 15]. How- ever, additional functions have been described which suggest a regulatory role for LL-37 in infl ammatory respons- Materials and Methods es. These include LPS neutralization [16, 17], activation LL-37 Synthesis of epithelial cells [18, 19] and chemotactic activity to- LL-37 (amino acid sequence LLGDFFRKSKEKIGKEFKR- wards neutrophils, monocytes, T cells and mast cells [2, IVQRIKDFLRNLVPRTES) was synthesized by solid-phase pep- 20]. Collectively, these results show that LL-37 may in- tide synthesis on a TentagelS AC (Rapp, Tübingen, Germany) using hibit infl ammatory responses to microbial stimuli, but 9-fl uorenylmethoxycarbonyl (Fmoc)/tBu chemistry, benzotriazole- may also act as a proinfl ammatory mediator. In addition, 1-yl-oxy-tris -pyrrolidino-phosphonium hexafl uorophosphate/N- methylmorpholine for activation and 20% piperidine in N-methyl- its ability to affect differentiation of dendritic cells [21] pyrrolidone for Fmoc removal. The peptide was cleaved from the and chemotaxis of monocytes and T cells [2] , its angio- resin, deprotected with trifl uoroacetic acid/water and purifi ed by genic activity [22] and its involvement in epithelial wound reverse-phase high-performance liquid chromatography on a Vy- healing [23] suggest that LL-37 may also be involved in dac C18 column. The molecular mass was confi rmed by Maldi-Tof adaptive immunity and wound repair. mass spectrometry. While LL-37 displays chemotactic activity towards Chemotaxis Assay different cell types from both innate and adaptive immu- Granulocytes were isolated from heparinized blood obtained nity, chemotactic activity towards eosinophils has not from healthy volunteers by density gradient centrifugation using been demonstrated yet. In addition, limited data are Percoll (density: 1.082 g/ml) as previously described [25] . Eosino- available on mechanisms mediating LL-37-induced che- phils were isolated from the mixed granulocyte population by neg- ative selection using anti-CD16 beads [26] . The neutrophil and motaxis of infl ammatory cells. Two receptors have been eosinophil suspensions were at least 95% pure. The chemotactic described that are involved in LL-37-mediated signaling activity of eosinophils and neutrophils was determined using the in leukocytes, the FPR-like 1 (FPRL1) [2] and the puri- modifi ed Boyden chamber technique [27] . Briefl y, 26 l of chemo- nergic P2X receptor [24] . Yang et al. [2] suggested that tactic stimuli in assay buffer was added to the lower compartment 7 ! 6 LL-37-induced chemotaxis is mediated via the FPRL1 and 50 l of cell suspension (2.5 10 cells/ml in assay buffer) was added to the upper compartment of a 48-well chamber (Neuro- based on experiments using FPRL1-transfected HEK293 probe, Cabin John, Md., USA). The assay buffer was composed 2+ cells and cross-desensitization of LL-37-induced Ca of 20 m M N-2-hydroxyethylpiperazine-N -2-ethanesulfonic acid fl uxes in monocytes using an FPRL1-specifi c agonist. In (Hepes buffer; pH 7.4), 132 mM NaCl, 6 mM KCl, 1.2 mM KH2 PO4 , contrast, mast cell chemotaxis was suggested not to be 1 m M MgSO4 , 5.5 m M glucose, 0.1 m M CaCl 2 and 0.5% human mediated via the FPRL1 [20] . Direct evidence for a role serum albumin (Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam, The Netherlands) diluted of FPRL1 and related receptors in chemotaxis of granu- 1: 1 with serum-free RPMI supplemented with 2 mM L-glutamine, locytes is missing. The P2X7 receptor has been shown to 20 U/ml penicillin and 20 g/ml streptomycin. The compartments

104 Int Arch Allergy Immunol 2006;140:103–112 Tjabringa /Ninaber /Drijfhout /Rabe /

Hiemstra

Eosinophils Neutrophils 100 125 ✽

✽ 100 75

✽ 75 ✽ 50 ✽ 50 Cells per 6 HPF Cells per 6 HPF 25 25

0 0 0 0.1 1 10 100 0 0.1 1 10 100 µ LL-37 (µM) fMLP LL-37 ( M) fMLP

Fig. 1. Effect of LL-37 on migration of human eosinophils and neutrophils. Human eosinophils and neutrophils were isolated from peripheral blood, and chemotaxis induced by various concentrations of LL-37 and by fMLP (10–8 M ) was determined using a modiﬁ ed Boyden chamber technique. Cells in 6 high-power ﬁ elds (HPF) were counted and expressed as mean 8 SEM of three individual experiments using cells from different donors. Data were analyzed for statistical differences by Student’s t test for paired samples. * p ! 0.05 vs. control.

were separated by a lower fi lter with a pore width of 0.45 m (Mil- tibodies against phosphorylated ERK1/2 (New England Biolabs, lipore Products, Bedfort, Mass., USA) and an upper fi lter with a Beverly, Mass., USA), and secondary horseradish peroxidase- pore width of 8 m (Sartorius Filter, San Francisco, Calif., USA). conjugated anti-rabbit IgG antibodies. The enhanced chemolu- After incubation (90 min for neutrophils and 120 min for eosino- minescent Western blotting detection system (Amersham Phar- phils) at 37 ° C, the fi lters were removed, fi xed in ethanol-butanol macia Biotech, Upsala, Sweden) was used to reveal immunore-

(80: 20, v/v), and stained with Weigert solution. Migrated cells were activity. counted in six random high-power ﬁ elds (! 400). As a positive control, chemotactic activity of N-formyl-methionyl-leucyl-phenylalanine (fMLP; Sigma-Aldrich, St. Louis, Mo., USA) was used. The FPR antagonistic peptide tert-butoxy-carbonyl-methionyl-leucyl- Results phenylalanine (tBoc-MLP) [28] was obtained from Sigma (St. Lou- is, Mo., USA), the long-acting 2 -adrenoceptor agonist formoterol LL-37 Induces Chemotaxis of Eosinophils and was provided by Sigma, and the monoclonal anti-LL-37 antibody Neutrophils 1.10C8 was generated in our laboratory [18] . To study the effects of LL-37 on chemotaxis of eosin- Cellular Lysates ophils and neutrophils, chemotactic activity of various Neutrophils were suspended in RPMI medium and incubated LL-37 concentrations or the FPR ligand fMLP (as a pos- at 37 ° C. After 15 min cells were lysed by adding 1 vol lysis buffer itive control) towards human peripheral blood eosino- (5 mM Tris, 100 m M NaCl, 1 mM CaCl 2 , 1 mM MgCl2 , 1% (v/v) phils and neutrophils was evaluated using a modiﬁ ed Triton X-100, 2 mM Na3 VO 4 and protease inhibitors (mini com- plete protease inhibitor cocktail; Boehringer Mannheim, Basel, Boyden chamber technique. LL-37 was shown to induce Switzerland) in washing buffer), and incubated for 15 min on ice. a dose-dependent chemotaxis of eosinophils and neutro-

After centrifugation for 5 min at 13,000 rpm at 4 ° C, the superna- phils (fi g. 1). When LL-37 or fMLP was added to the up- tant was collected and stored at –20 ° C. Protein concentrations of per compartment or both compartments, no increased the lysates were measured by the bicinchoninic acid protein assay migration was observed (table 1). These results show that system (Pierce, Rockford, Ill., USA). LL-37 displays chemotactic rather than chemokinetic ac- Gel Electrophoresis and Western Blotting tivity for eosinophils and neutrophils. Gel electrophoresis and Western blotting were performed using Bio-Rad systems (Hercules, Calif., USA) according to the Monoclonal Anti-LL-37 Antibody Inhibits manufacturer’s instructions. Samples were subjected to sodium LL-37-Induced Chemotaxis dodecyl sulfate polyacrylamide gel electrophoresis on a 10% gly- cine-based gel, and dissolved proteins were transferred to a poly- To study the specifi city of LL-37-induced chemotaxis, vinylidene difl uoride membrane. Non-specifi c binding sites were LL-37 was preincubated for 2 h with a monoclonal anti- blocked, and the blots were incubated with rabbit polyclonal an- LL-37 antibody before neutrophil chemotaxis assay in the

Human Cathelicidin LL-37 Int Arch Allergy Immunol 2006;140:103–112 105 Table 1. Checkerboard analysis of LL-37-induced chemotaxis of caused by LL-37 and not by a contaminant in the LL-37 eosinophils and neutrophils preparation used.

Lower compartment Upper compartment LL-37-Induced Chemotaxis of Eosinophils and medium LL-37 fMLP Neutrophils Is Mediated via Formyl-Peptide Receptors Eosinophils Medium 0 0 0 LL-37-induced chemotaxis of neutrophils, monocytes LL-37 23 0 ND and T cells was suggested to be mediated via the FPR fMLP 23 ND 13 FPRL1, based on studies using FPRL1-transfected Neutrophils HEK293 cells [2] . However, involvement of FPRL1 in Medium 29 19 25 LL-37-induced chemotaxis of infl ammatory cells was not LL-37 54 28 ND fMLP 61 ND 20 formally demonstrated. Therefore, we studied the effect of pertussis toxin, which acts as a G-protein inhibitor by Chemotaxis was determined using the modifi ed Boyden cham- causing ADP-ribosylation of the -subunit of G-proteins, ber technique with cells present in the upper compartment. LL-37 and the FPR antagonistic peptide tBoc-MLP on LL-37- ! –5 –8 was used at 0.8 10 M, whereas fMLP was used at 10 M. induced chemotaxis of eosinophils and neutrophils (fi g. ND = Not determined. Results are from a single experiment (cell count in 6 high-power fi elds); similar results were obtained using 3). Inhibition of LL-37-induced chemotaxis by pertussis cells from another donor in a separate experiment. toxin did not reach statistical signifi cance (eosinophils: p = 0.0554; neutrophils: p = 0.1085). However, tBoc- MLP did signifi cantly inhibit LL-37-induced chemotaxis, suggesting that LL-37-induced chemotaxis of both eosinophils and neutrophils is mediated via a FPR. As expect- ed, fMLP-induced chemotaxis was fully inhibited by both 120 pertussis toxin and tBoc-MLP.

✽ 80 LL-37-Induced Chemotaxis Is Inhibited by the Long-Acting 2-Adrenoceptor Agonist Formoterol To study the effect of 2 -agonists on LL-37-induced 40

Cells per 6 HPF chemotaxis of human eosinophils and neutrophils, cells were preincubated with formoterol and LL-37-induced 0 chemotaxis was determined using a modiﬁ ed Boyden Blank LL-37 LL-37 + moAb chamber assay. Formoterol inhibited both LL-37- and fMLP-induced chemotaxis (ﬁ g. 4).

Fig. 2. Effect of a monoclonal anti-LL-37 antibody on LL-37-in- Combination of LL-37 and fMLP Inhibits Neutrophil duced chemotaxis of neutrophils. LL-37 was preincubated with the Chemotaxis monoclonal anti-LL-37 antibody 1.10C8 and neutrophil chemotaxis by LL-37 (0.8 ! 10–5 M ) was determined using a modifi ed To study the combined effect of LL-37 and fMLP on Boyden chamber technique. Chemotaxis was assessed by counting chemotaxis of neutrophils, chemotaxis assays were per- cells in 6 high-power fi elds (HPF) and expressed as mean 8 SEM formed using LL-37, fMLP and a combination of fMLP of three individual experiments. Data were analyzed for statistical and LL-37. While both LL-37 and fMLP separately in- differences by Student’s t test for paired samples. LL-37 signifi - creased chemotaxis of neutrophils, the combination of cantly increased chemotaxis as compared to the control. * Signifi - cant effect of anti-LL-37 monoclonal antibody on LL-37-induced LL-37 and fMLP decreased chemotaxis (fi g. 5). chemotaxis, p ! 0.05. LL-37 Does Not Activate Mitogen-Activated Protein Kinases in Human Neutrophils MAPKs have been demonstrated to be involved in a presence of the monoclonal antibody. While LL-37 in- variety of cellular processes, including chemotaxis. There- duced chemotaxis of neutrophils, the monoclonal anti- fore, the effect of LL-37 on MAPKs activation in human LL-37 antibody inhibited LL-37-induced chemotaxis neutrophils was studied. Cells were stimulated with LL- (fi g. 2). This confi rms that the observed chemotaxis is 37 or fMLP for 15 min, and activation of the MAPK

106 Int Arch Allergy Immunol 2006;140:103–112 Tjabringa /Ninaber /Drijfhout /Rabe /

Hiemstra

Eosinophils Neutrophils ✽ ✽

✽ ✽ 100 200

75 ✽

✽ 50 100 Cells per 6 HPF Cells per 6 HPF 25

0 0 _ _ _ _ PTX PTX tBoc tBoc tBoc tBoc PTX PTX Control Control

LL-37 fMLP LL-37 fMLP

Fig. 3. Effect of pertussis toxin and tBoc-MLP on LL-37-induced chemotaxis of human peripheral blood eosinophils and neutrophils. Cells were preincubated with pertussis toxin (160 ng/ml) or tBoc-MLP (50 M) and chemotaxis by LL-37 (0.8 ! 10 –5 M) and fMLP (10 –8 M) was determined using a modifi ed Boyden chamber technique. Cells in 6 high-power fi elds (HPF) were counted and expressed as mean 8 SEM of three individual experiments. Data were analyzed for statistical differences by Student’s t test for paired samples. fMLP and LL-37 increased chemotaxis of both eosinophils and neutrophils as compared to the control, while only LL-37-induced chemotaxis of eosinophils did not reach statistical signifi cance (p value 0.0821). * p ! 0.05.

Eosinophils Neutrophils ✽ 100 ✽ 150 ✽ 75 ✽ ✽ 100

50 Cells per 6 HPF Cells per 6 HPF 25

0 0 _ –4 –6 –4 –6 –4 –6 –4 –6 _ _ _ Control Control Form. 10 Form. 10 Form. 10 Form. 10 Form. 10 Form. 10 Form. 10 Form. 10

LL-37 fMLP LL-37 fMLP

Fig. 4. Effect of formoterol on LL-37-induced chemotaxis of human peripheral blood eosinophils and neutrophils. Cells were preincubated with formoterol (10–4 and 10–6 M) and chemotaxis by LL-37 (0.8 ! 10 –5 M ) and fMLP (10 –8 M) was determined using the modifi ed Boyden chamber technique. Cells in 6 high-power fi elds (HPF) were counted and expressed as mean 8 SEM of three individual experiments. Data were analyzed for statistical differences by Student’s t test for paired samples. fMLP and LL-37 increased chemotaxis of both eosinophils and neutrophils as compared to the control, while only fMLP-induced chemotaxis of neutrophils did not reach statistical signifi cance (p value 0.0687). * p ! 0.05.

Human Cathelicidin LL-37 Int Arch Allergy Immunol 2006;140:103–112 107 ERK1/2 was determined using Western blot analysis. LL- more, the long-acting 2 -adrenoceptor agonist formoterol 37 did not induce MAPKs activation in neutrophils, inhibited LL-37-induced chemotaxis of both eosinophils while fMLP activated ERK1/2 (fi g. 6). This suggests that and neutrophils, suggesting that 2-agonists may inhibit LL-37 and fMLP activate different cellular pathways in LL-37-mediated infl ammatory responses. Finally, LL-37 neutrophils. and fMLP appeared to activate different cellular pathways, since in contrast to fMLP, LL-37 did not activate ERK1/2 in human neutrophils. The results from this Discussion study suggest that LL-37 may contribute to infl ammation in infl ammatory lung diseases such as asthma and In this study, we demonstrate that LL-37 induces che- COPD. motaxis of human eosinophils and neutrophils. This ef- Chemotactic activity of antimicrobial peptides has fect was inhibited by the FPR antagonistic tBoc-MLP been suggested to be mediated via chemokine and FPRs peptide, indicating the involvement of an FPR. Further- [29]. Human -defensins, that are mainly produced by a variety of epithelial cells, chemoattract immature dendritic cells and memory T cells via the human CC-chemokine receptor 6 [1] , while the receptor involved in neutrophil -defensin-induced chemotaxis of T cells and immature dendritic cells has not yet been identifi ed [30] . ✽ Cathepsin G, a serine proteinase with antimicrobial ac- ✽ tivity which is abundantly present in the azurophilic gran- 120 ules of neutrophils, has been demonstrated to chemoat- 100 tract monocytes and neutrophils [31, 32]. Chemotactic responses to cathepsin G were shown to be mediated via 80 the high-affi nity FPR, which was supported by transfec- 60 tion studies using FPR-transfected cell lines [32] . In ad-

40 dition to the FPR, two homologues of this receptor have Cells per 6 HPF 6 per Cells been described, namely FPRL1 and FPRL2 [33] . In con- 20 trast to the low-affi nity FPRL1, the FPRL2 does not bind 0 Control LL-37 fMLP LL-37 + fMLP fMLP. LL-37-induced chemotaxis of neutrophils, monocytes and T cells was suggested to be mediated by the low- affi nity FPRL1 [2] . This conclusion was based on experiments showing that FPRL1-transfected HEK293 cells Fig. 5. Effect of LL-37, fMLP and the combination of LL-37 and fMLP on chemotaxis of neutrophils. Cells were incubated with LL- migrate in response to LL-37 whereas ETFR cells express- 37 (0.8 ! 10–5 M ), fMLP (10 –8 M ) or the combination of LL-37 ing FPR do not respond. Furthermore, the authors showed (0.8 ! 10 –5 M) and fMLP (10–8 M), and chemotaxis was deter- cross-desensitization of LL-37-induced Ca2+ mobiliza- mined using the modifi ed Boyden chamber technique. Cells in 6 tion in monocytes by the FPRL1-specifi c agonist Su-pep- high-power fi elds (HPF) were counted and expressed as mean 8 tide. In the present study the involvement of FPRs in SEM of three individual experiments. Data were analyzed for statistical differences by Student’s t test for paired samples. * p ! LL-37-induced chemotaxis of human eosinophils and 0.05. neutrophils was demonstrated using the FPR antagonis-

Fig. 6. Effect of fMLP and LL-37 on ERK1/2 activation in neutrophils isolated from human peripheral blood. Cells were pERK1/2 incubated for 15 min with various concentrations of fMLP or LL-37, and lysed in ly- –7 –8 –4 –5 –6 sis buffer. Activation of ERK1/2 was deter- – 10–6 10 10 10 10 10 mined by assessment of phosphorylated fMLP LL-37 ERK1/2 (pERK1/2) in cellular lysates using Western blot analysis.

108 Int Arch Allergy Immunol 2006;140:103–112 Tjabringa /Ninaber /Drijfhout /Rabe /

Hiemstra tic peptide tBoc-MLP. However, based on these results, we cannot discriminate to which FPR(s) LL-37 binds. In Neutrophils Eosinophils addition to chemotaxis, also LL-37-induced endothelial cell proliferation [22] , but not LL-37-induced activation of airway epithelial cells [18] and human monocytes [34] have been suggested to be mediated via an FPR. A recent Chemotaxis study shows that LL-37 increases processing and release of IL-1 in LPS-primed monocytes via the P2X7 receptor [24]. Furthermore, we demonstrate that while both LL-37 and fMLP induce chemotaxis of eosinophils and neutro- Release phils via an FPR, only fMLP activates the MAPK ERK1/2 Formoterol in human eosinophils and neutrophils. This suggests that LL-37 and fMLP activate different FPRs, or that they differentially signal through the same FPR. Differential hCAP-18/LL-37 signaling via the FPRL1 by multiple ligands was already demonstrated by Bae et al. [35, 36], suggesting that the proinfl ammatory activity of ligands of FPRs may differ Release Activation and that FPR-mediated chemotaxis is associated with additional cellular activities. In addition, other receptors than FPRs may be involved in MAPK activation. Final- ly, we observed that the combination of LL-37 and fMLP was less effective in inducing neutrophil chemotaxis than LL-37 or fMLP alone. Whether this is explained by com- petition for binding sites on the same receptor, activation of different signaling cascades by both components, a par- Fig. 7. Role of hCAP-18/LL-37 in chemotaxis of neutrophils and allel inhibitory effect of LL-37 on chemotaxis that is in- eosinophils. hCAP-18/LL-37 produced by neutrophils and epithe- dependent of FPRs, or other mechanisms, remains to be lial cells causes chemoattraction of neutrophils and eosinophils, via established. a mechanism that is inhibited by the 2 -adrenoceptor agonist for- The antimicrobial peptide LL-37 may affect the in- moterol. fl ammatory process by regulating recruitment of granulocytes to infl ammatory sites [2], by induction of release of the chemokine IL-8 from lung epithelial cells [18, 19] and by induction of cell death as demonstrated in airway enzae infection [39] . In contrast, Islam et al. [40] demon- epithelial cells and Jurkat T cells [Aarbiou, Tjabringa et strated that Shigella spp. decreased LL-37 expression in al., manuscript in preparation]. Recruitment of infl am- colon epithelial cells both in vitro and in vivo, a process matory cells to sites of infl ammation may result in a circle which was suggested to be mediated via cellular signaling of infl ammation via release of chemokines and antimi- through bacterial DNA. Bals et al. [14] showed that gene crobial peptides (fi g. 7). Various in vivo studies have dem- transfer of hCAP-18/LL-37 in a cystic fi brosis xenograft onstrated regulation of hCAP-18/LL-37 levels in infec- model restored bacterial killing to normal levels, suggest- tion and infl ammation. Both hCAP-18/LL-37 gene and ing that antimicrobial peptides protect against bacterial protein expression were shown to be increased in kerati- infection ex vivo. Therefore, LL-37 was suggested to be nocytes from infl amed skin from psoriasis patients, and involved in infection and infl ammatory diseases. psoriatic scales were shown to contain hCAP-18/LL-37 In the present study, we demonstrate that LL-37, [9]. In addition, tracheal aspirates derived from mechan- which is abundantly present in neutrophils, chemoat- ically ventilated newborns with pulmonary or systemic tracts both eosinophils and neutrophils. This suggests that infections were demonstrated to contain increased levels LL-37 may play a role in both neutrophil- and eosinophil- of hCAP-18/LL-37 as compared to non-infected new- dominated diseases including the infl ammatory lung dis- borns [37] . Furthermore, hCAP-18/LL-37 was found to eases COPD and asthma, which are characterized by an be increased in chronic nasal infl ammatory disease [38] , infl ux of neutrophils and eosinophils respectively. Vari- and in cultured pharyngeal cells after Haemophilus infl u- ous studies demonstrated mixed neutrophil-eosinophil

Human Cathelicidin LL-37 Int Arch Allergy Immunol 2006;140:103–112 109 infi ltration in these lung diseases. While in patients with ol inhibits chemotaxis via an increase in intracellular cy- chronic bronchitis neutrophil infl ux is accompanied by clic AMP (cAMP), since other cAMP-elevating agents eosinophilic infi ltration [41], in asthma not only eosino- such as phosphodiesterase inhibitors and dibutyryl cAMP phil but also neutrophil numbers in bronchoalveolar la- also block chemotaxis [57, 58]. Since LL-37 does not ac- vage or induced sputum increase following experimental tivate the MAPKs ERK1/2, p38 or JNK, this suggests that allergen exposure [42, 43] and during an exacerbation therapy targeting MAPKs does not affect LL-37-induced [44]. Furthermore, interactions between neutrophils and chemotaxis of eosinophils and neutrophils. eosinophils were suggested in infl ammatory lung diseas- In conclusion, we demonstrate that LL-37 chemoat- es, as in a study demonstrating a more extensive eosino- tracts both eosinophils and neutrophils via activation of phil degranulation in children with acute severe asthma an FPR. This suggests that LL-37 may regulate the in- with a combined neutrophil/eosinophil airway infl amma- fl ammatory response by selectively attracting infl ammation, compared to patients with isolated eosinophilic in- tory cells. Since both eosinophils and neutrophils are che- fl ammation [45] . In addition, eosinophil migration across moattracted by LL-37, LL-37 may be involved in both layers of cultured epithelial cells was enhanced by neu- eosinophil- and neutrophil-dominated diseases. Further- trophils, suggesting involvement of neutrophil-eosino- more, formoterol inhibits LL-37-induced chemotaxis of phil interactions [46] . A mechanistic basis for interac- granulocytes, suggesting that therapeutic intervention of tions between neutrophils and eosinophils was provided infl ammatory diseases with 2-agonists may affect LL-37- by studies suggesting that neutrophil elastase [47–49] and mediated infl ammatory processes. This study suggests an eosinophilic major basic protein [50, 51] may mediate interaction of eosinophils and neutrophils via the antimi- these interactions. Although extrapolation of in vitro crobial peptide LL-37. studies to the situation in vivo is diffi cult, our results demonstrating that neutrophil-derived LL-37 chemoattracts eosinophils suggest that LL-37 acts as a mediator in the Acknowledgements interaction between neutrophils and eosinophils. We used relatively high concentrations of LL-37 to induce This study was supported by a grant from The Netherlands Asthma Foundation (grant 98.46) and by a research grant from chemotaxis. However, these concentrations can be AstraZeneca (Lund, Sweden). reached in the close vicinity of degranulating neutrophils. In addition, seminal plasma has been found to contain high concentrations (up to 80 g/ml) of hCAP-18 [52] . Furthermore, as previously suggested by Yang et al. [2], the high concentrations required for LL-37 to induce chemotaxis indicate that the interaction between LL-37 and its FPR is a low-affi nity interaction. Current therapy for infl ammatory lung diseases in- cludes treatment with corticosteroids and 2 -agonists like formoterol. Also MAPKs, which play an important role in various cellular processes including chemotaxis, have been evaluated as target for drug therapy. 2 -Agonists act on a variety of cell types, resulting in e.g. smooth muscle cell relaxation, reduced epithelial cytokine expression and reduced infi ltration of infl ammatory cells [53] . The inhibitory activity of 2 -agonists is not restricted to responses mediated by FPRs. In the present study, we demonstrate that the 2 -adrenoceptor agonist formoterol inhibits LL-37-induced chemotaxis of both eosinophils and neutrophils, suggesting that therapeutic intervention with 2 -agonists in both COPD and asthma may reduce the proinfl ammatory effects mediated by LL-37. Previously, formoterol has been shown to inhibit fMLP-induced chemotaxis of eosinophils [54–56] . It is likely that formoter-

110 Int Arch Allergy Immunol 2006;140:103–112 Tjabringa /Ninaber /Drijfhout /Rabe /

Hiemstra

References

1 Yang D, Chertov O, Bykovskaia SN, Chen Q, 13 Gudmundsson GH, Agerberth B, Odeberg J, 25 Van Wetering S, Mannesse-Lazeroms SP, van Buffo MJ, Shogan J et al: Beta-defensins: link- Bergman T, Olsson B, Salcedo R: The human Sterkenburg MA, Daha MR, Dijkman JH, ing innate and adaptive immunity through gene FALL39 and processing of the cathelin Hiemstra PS: Effect of defensins on interleu-

dendritic and T cell CCR6. Science 1999; 286: precursor to the antibacterial peptide LL-37 in kin-8 synthesis in airway epithelial cells. Am J

525–528. granulocytes. Eur J Biochem 1996; 238: 325– Physiol 1997;272:L888–L896. 2 Yang D, Chen Q, Schmidt AP, Anderson GM, 332. 26 Dent G, Munoz NM, Ruhlmann E, Zhu X, Wang JM, Wooters J et al: LL-37, the neutro- 14 Bals R, Weiner DJ, Meegalla RL, Wilson JM: Leff AR, Magnussen H et al: Protein kinase C phil granule- and epithelial cell-derived cathe- Transfer of a cathelicidin peptide antibiotic inhibition enhances platelet-activating factor- licidin, utilizes formyl peptide receptor-like 1 gene restores bacterial killing in a cystic ﬁ brosis induced eicosanoid production in human eo-

(FPRL1) as a receptor to chemoattract human xenograft model. J Clin Invest 1999; 103: sinophils. Am J Respir Cell Mol Biol 1998; 18:

peripheral blood neutrophils, monocytes and 1113–1117. 136–144.

T cells. J Exp Med 2000; 192: 1069–1074. 15 Nizet V, Ohtake T, Lauth X, Trowbridge J, 27 Koenderman L, Kok PT, Hamelink ML, Ver- 3 Zanetti M: Cathelicidins, multifunctional pep- Rudisill J, Dorschner RA et al: Innate antimi- hoeven AJ, Bruijnzeel PL: An improved meth- tides of the innate immunity. J Leukoc Biol crobial peptide protects the skin from invasive od for the isolation of eosinophilic granulo-

2004; 75: 39–48. bacterial infection. Nature 2001; 414: 454– cytes from peripheral blood of normal

4 Bals R, Wilson JM: Cathelicidins – a family of 457. individuals. J Leukoc Biol 1988; 44: 79–86. multifunctional antimicrobial peptides. Cell 16 Larrick JW, Hirata M, Balint RF, Lee J, Zhong 28 McCormick BA, Colgan SP, Delp-Archer C,

Mol Life Sci 2003; 60: 711–720. J, Wright SC: Human CAP18: a novel antimi- Miller SI, Madara JL: Salmonella typhimuri- 5 Sorensen OE, Follin P, Johnsen AH, Calafat J, crobial lipopolysaccharide-binding protein. In- um attachment to human intestinal epithelial

Tjabringa GS, Hiemstra PS et al: Human cat- fect Immun 1995; 63: 1291–1297. monolayers: transcellular signalling to subepi-

helicidin, hCAP-18, is processed to the antimi- 17 Nagaoka I, Hirota S, Niyonsaba F, Hirata M, thelial neutrophils. J Cell Biol 1993; 123: 895– crobial peptide LL-37 by extracellular cleavage Adachi Y, Tamura H et al: Cathelicidin family 907.

with proteinase 3. Blood 2001; 97: 3951–3959. of antibacterial peptides CAP18 and CAP11 29 Yang D, Chertov O, Oppenheim JJ: Participa- 6 Sorensen OE, Gram L, Johnsen AH, Anders- inhibit the expression of TNF- by blocking tion of mammalian defensins and cathelicidins son E, Bangsboll S, Tjabringa GS et al: Process- the binding of LPS to CD14+ cells. J Immunol in antimicrobial immunity: receptors and ac-

ing of seminal plasma hCAP-18 to ALL-38 by 2001;167: 3329–3338. tivities of human defensins and cathelicidin

gastricsin: a novel mechanism of generating 18 Tjabringa GS, Aarbiou J, Ninaber DK, Drijf- (LL-37). J Leukoc Biol 2001; 69: 691–697. antimicrobial peptides in vagina. J Biol Chem hout JW, Sorensen OE, Borregaard N et al: The 30 Yang D, Chen Q, Chertov O, Oppenheim JJ:

2003; 278: 28540–28546. antimicrobial peptide LL-37 activates innate Human neutrophil defensins selectively che- 7 Sorensen O, Arnljots K, Cowland JB, Bainton immunity at the airway epithelial surface by moattract naive T and immature dendritic

DF, Borregaard N: The human antibacterial transactivation of the epidermal growth factor cells. J Leukoc Biol 2000; 68: 9–14.

cathelicidin, hCAP-18, is synthesized in my- receptor. J Immunol 2003; 171: 6690–6696. 31 Chertov O, Ueda H, Xu LL, Tani K, Murphy elocytes and metamyelocytes and localized to 19 Scott MG, Davidson DJ, Gold MR, Bowdish WJ, Wang JM et al: Identiﬁ cation of human

speciﬁ c granules in neutrophils. Blood 1997; D, Hancock RE: The human antimicrobial neutrophil-derived cathepsin G and azuroci-

90: 2796–2803. peptide LL-37 is a multifunctional modulator din/CAP37 as chemoattractants for mononu-

8 Agerberth B, Charo J, Werr J, Olsson B, Idali of innate immune responses. J Immunol 2002; clear cells and neutrophils. J Exp Med 1997;

F, Lindbom L et al: The human antimicrobial 169: 3883–3891. 186: 739–747. and chemotactic peptides LL-37 and -defen- 20 Niyonsaba F, Iwabuchi K, Someya A, Hirata 32 Sun R, Iribarren P, Zhang N, Zhou Y, Gong sins are expressed by speciﬁ c lymphocyte and M, Matsuda H, Ogawa H et al: A cathelicidin W, Cho EH et al: Identiﬁ cation of neutrophil

monocyte populations. Blood 2000; 96: 3086– family of human antibacterial peptide LL-37 granule protein cathepsin G as a novel chemo- 3093. induces mast cell chemotaxis. Immunology tactic agonist for the G protein-coupled formyl

9 Frohm M, Agerberth B, Ahangari G, Stahle- 2002;106: 20–26. peptide receptor. J Immunol 2004; 173: 428– Backdahl M, Liden S, Wigzell H et al: The ex- 21 Davidson DJ, Currie AJ, Reid GS, Bowdish 436. pression of the gene coding for the antibacte- DM, MacDonald KL, Ma RC et al: The cat- 33 Le Y, Murphy PM, Wang JM: Formyl-peptide

rial peptide LL-37 is induced in human ionic antimicrobial peptide LL-37 modulates receptors revisited. Trends Immunol 2002; 23:

keratinocytes during inﬂ ammatory disorders. dendritic cell differentiation and dendritic cell- 541–548.

J Biol Chem 1997; 272: 15258–15263. induced T cell polarization. J Immunol 2004; 34 Bowdish DM, Davidson DJ, Speert DP, Han-

10 Frohm NM, Sandstedt B, Sorensen O, Weber 172: 1146–1156. cock RE: The human cationic peptide LL-37 G, Borregaard N, Stahle-Backdahl M: The hu- 22 Koczulla R, von Degenfeld G, Kupatt C, Krotz induces activation of the extracellular signal- man cationic antimicrobial protein (hCAP18), F, Zahler S, Gloe T et al: An angiogenic role for regulated kinase and p38 Kinase pathways in

a peptide antibiotic, is widely expressed in hu- the human peptide antibiotic LL-37/hCAP-18. primary human monocytes. J Immunol 2004;

man squamous epithelia and colocalizes with J Clin Invest 2003; 111: 1665–1672. 172: 3758–3765.

interleukin-6. Infect Immun 1999; 67: 2561– 23 Heilborn JD, Nilsson MF, Kratz G, Weber G, 35 Bae YS, Park JC, He R, Ye RD, Kwak JY, Suh 2566. Sorensen O, Borregaard N et al: The cathelici- PG et al: Differential signaling of formyl pep- 11 Bals R, Wang X, Zasloff M, Wilson JM: The din anti-microbial peptide LL-37 is involved tide receptor-like 1 by Trp-Lys-Tyr-Met-Val- peptide antibiotic LL-37/hCAP-18 is expressed in re-epithelialization of human skin wounds Met-CONH2 or lipoxin A4 in human neutro-

in epithelia of the human lung where it has and is lacking in chronic ulcer epithelium. J phils. Mol Pharmacol 2003; 64: 721–730.

broad antimicrobial activity at the airway sur- Invest Dermatol 2003; 120: 379–389. 36 Bae YS, Yi HJ, Lee HY, Jo EJ, Kim JI, Lee TG

face. Proc Natl Acad Sci USA 1998; 95: 9541– 24 Elssner A, Duncan M, Gavrilin M, Wewers et al: Differential activation of formyl peptide 9546. MD: A novel P2X7 receptor activator, the hu- receptor-like 1 by peptide ligands. J Immunol

12 Agerberth B, Grunewald J, Castanos-Velez E, man cathelicidin-derived peptide LL37, in- 2003; 171:6807–6813. Olsson B, Jornvall H, Wigzell H et al: Antibac- duces IL-1 processing and release. J Immunol 37 Schaller-Bals S, Schulze A, Bals R: Increased

terial components in bronchoalveolar lavage 2004;172: 4987–4994. levels of antimicrobial peptides in tracheal as- ﬂ uid from healthy individuals and sarcoidosis pirates of newborn infants during infection.

patients. Am J Respir Crit Care Med 1999; 160: Am J Respir Crit Care Med 2002; 165: 992–

283–290. 995.

Human Cathelicidin LL-37 Int Arch Allergy Immunol 2006;140:103–112 111 38 Kim ST, Cha HE, Kim DY, Han GC, Chung 45 Gibson PG, Norzila MZ, Fakes K, Simpson J, 52 Malm J, Sorensen O, Persson T, Frohm-Nils- YS, Lee YJ et al: Antimicrobial peptide LL-37 Henry RL: Pattern of airway inﬂ ammation son M, Johansson B, Bjartell A et al: The hu- is upregulated in chronic nasal inﬂ ammatory and its determinants in children with acute se- man cationic antimicrobial protein (hCAP-18)

disease. Acta Otolaryngol 2003; 123: 81–85. vere asthma. Pediatr Pulmonol 1999; 28: 261– is expressed in the epithelium of human epi- 39 Lysenko ES, Gould J, Bals R, Wilson JM, 270. didymis, is present in seminal plasma at high Weiser JN: Bacterial phosphorylcholine de- 46 Zuurbier AE, Liu L, Mul FP, Verhoeven AJ, concentrations, and is attached to spermato-

creases susceptibility to the antimicrobial pep- Knol EF, Roos D: Neutrophils enhance eosin- zoa. Infect Immun 2000; 68: 4297–4302. tide LL-37/hCAP18 expressed in the upper re- ophil migration across monolayers of lung epi- 53 Johnson M, Rennard S: Alternative mecha- spiratory tract. Infect Immun 2000; 68: thelial cells. Clin Exp Allergy 2001; 31: 444– nisms for long-acting 2 -adrenergic agonists in

1664–1671. 452. COPD. Chest 120, 258–270. 2001 40 Islam D, Bandholtz L, Nilsson J, Wigzell H, 47 Lapperre TS, Hiltermann TJN, van Bree L, 54 Eda R, Sugiyama H, Hopp RJ, Okada C, Bew- Christensson B, Agerberth B et al: Downregu- Sterk PJ, Stolk J, Rabe KF et al: Relation be- tra AK, Townley RG: Inhibitory effects of for- lation of bactericidal peptides in enteric infec- tween neutrophil elastase and eosinophil cat- moterol on platelet-activating factor induced tions: a novel immune escape mechanism with ionic protein in induced sputum following eosinophil chemotaxis and degranulation. Int

bacterial DNA as a potential regulator. Nat ozone exposure of mild asthmatics. Eur Respir Arch Allergy Immunol 1993; 102: 391–398.

Med 2001; 7: 180–185. J 2000; 16: 219S. 55 Erjefalt JS, Korsgren M, Malm-Erjefalt M, 41 Tsoumakidou M, Tzanakis N, Siafakas NM: 48 Liu H, Lazarus SC, Caughey GH, Fahy JV: Conroy DM, Williams TJ, Persson CG: Acute Induced sputum in the investigation of airway Neutrophil elastase and elastase-rich cystic ﬁ - allergic responses induce a prompt luminal en-

inﬂ ammation of COPD. Respir Med 2003; 97: brosis sputum degranulate human eosinophils try of airway tissue eosinophils. Am J Respir

863–871. in vitro. Am J Physiol 1999; 276:L28–L34. Cell Mol Biol 2003; 29: 439–448. 42 Fahy JV, Liu J, Wong H, Boushey HA: Analy- 49 Van Wetering S, Mannesse-Lazeroms SP, 56 Wallin A, Sandstrom T, Soderberg M, How- sis of cellular and biochemical constituents of Ruhlmann E, Rabe KF, Hiemstra PS: Neutro- arth P, Lundback B, Della-Cioppa G et al: The induced sputum after allergen challenge: a phil serine proteinases induce eosinophilic cat- effects of regular inhaled formoterol, method for studying allergic airway inﬂ amma- ionic protein and LTC4 release by human eo- budesonide, and placebo on mucosal inﬂ am-

tion. J Allergy Clin Immunol 1994; 93: 1031– sinophils. Eur Resp J 2000; 16: 100S. mation and clinical indices in mild asthma.

1039. 50 Moy JN, Gleich GJ, Thomas LL: Noncytotox- Am J Respir Crit Care Med 1999; 159: 79–86. 43 De Monchy JG, Kauffman HF, Venge P, ic activation of neutrophils by eosinophil gran- 57 Yasui K, Agematsu K, Shinozaki K, Hokibara Koeter GH, Jansen HM, Sluiter HJ et al: Bron- ule major basic protein. Effect on superoxide S, Nagumo H, Yamada S et al: Effects of the- choalveolar eosinophilia during allergen-in- anion generation and lysosomal enzyme re- ophylline on human eosinophil functions:

duced late asthmatic reactions. Am Rev Respir lease. J Immunol 1990; 145: 2626–2632. comparative study with neutrophil functions.

Dis 1985; 131: 373–376. 51 Page SM, Gleich GJ, Roebuck KA, Thomas J Leukoc Biol 2000; 68: 194–200. 44 Fahy JV, Kim KW, Liu J, Boushey HA: Prom- LL: Stimulation of neutrophil interleukin-8 58 Kaneko T, Alvarez R, Ueki IF, Nadel JA: El- inent neutrophilic inﬂ ammation in sputum production by eosinophil granule major basic evated intracellular cyclic AMP inhibits che-

from subjects with asthma exacerbation. J Al- protein. Am J Respir Cell Mol Biol 1999; 21: motaxis in human eosinophils. Cell Signal

lergy Clin Immunol 1995; 95: 843–852. 230–237. 1995; 7: 527–534.

112 Int Arch Allergy Immunol 2006;140:103–112 Tjabringa /Ninaber /Drijfhout /Rabe /

Hiemstra