Cathelicidin Anti-Microbial Peptide Expression in Sweat, an Innate Defense System for the Skin

The eccrine gland is one of the major cutaneous appendages and secretes sweat. Its principal function is thermoregulation during exposure to a hot environment or physical exercise. In addition to this function, we show that LL-37, a member of cathelicidin family of anti-microbial peptides, is expressed in sweat. LL-37 protein and mRNA was seen in the eccrine structures of normal human skin by immunohistochemistry and in situ hybridization. LL-37 was localized to both the eccrine gland and sweat ductal epithelial cells, whereas dermcidin, a previously described natural antibiotic in sweat, was expressed only in the gland itself. The anti-microbial activity of LL-37 and dermcidin against various bacteria in the sweat ionic environment was demonstrated by solution colony forming assay using synthetic peptides, and in sweat obtained from normal volunteers. These results indicate that cathelicidin is secreted in human sweat, has potent anti-microbial activity against both gram-positive and gram-negative bacteria, and can, after processing from the preproform, provide a barrier for protection against infection. Thus, sweat represents a unique mode of delivery for potent innate immune effector molecules in the absence of inflammation. The eccrine gland is one of the major cutaneous appendages and secretes sweat. Its principal function is thermoregulation during exposure to a hot environment or physical exercise. In addition to this function, we show that LL-37, a member of cathelicidin family of anti-microbial peptides, is expressed in sweat. LL-37 protein and mRNA was seen in the eccrine structures of normal human skin by immunohistochemistry and in situ hybridization. LL-37 was localized to both the eccrine gland and sweat ductal epithelial cells, whereas dermcidin, a previously described natural antibiotic in sweat, was expressed only in the gland itself. The anti-microbial activity of LL-37 and dermcidin against various bacteria in the sweat ionic environment was demonstrated by solution colony forming assay using synthetic peptides, and in sweat obtained from normal volunteers. These results indicate that cathelicidin is secreted in human sweat, has potent anti-microbial activity against both gram-positive and gram-negative bacteria, and can, after processing from the preproform, provide a barrier for protection against infection. Thus, sweat represents a unique mode of delivery for potent innate immune effector molecules in the absence of inflammation. In mammalian skin, two major classes of anti-microbial peptides have been identified: cathelicidins (Zanetti et al., 1995Zanetti M. Gennaro R. Romeo D. Cathelicidins a novel protein family with a common proregion and a variable C-terminal antimicrobial domain.FEBS Lett. 1995; 374: 1-5Abstract Full Text PDF PubMed Scopus (583) Google Scholar; Gallo et al., 1997Gallo R.L. Kim K.J. Bernfield M. Kozak C.A. Zanetti M. Merluzzi L. Gennaro R. Identification of CRAMP, a cathelin-related antimicrobial peptide expressed in the embryonic and adult mouse.J Biol Chem. 1997; 272: 13088-13093Crossref PubMed Scopus (305) Google Scholar; Nizet et al., 2001Nizet V. Ohtake T. Lauth X. et al.Innate antimicrobial peptide protects the skin from invasive bacterial infection.Nature. 2001; 414: 454-457Crossref PubMed Scopus (980) Google Scholar) and β-defensins (Ali et al., 2001Ali R.S. Falconer A. Ikram M. Bissett C.E. Cerio R. Quinn A.G. Expression of the peptide antibiotics human beta defensin-1 and human beta defensin-2 in normal human skin.J Invest Dermatol. 2001; 117: 106-111Crossref PubMed Scopus (201) Google Scholar; Harder et al., 1997Harder J. Bartels J. Christophers E. Schroder J.M. A peptide antibiotic from human skin.Nature. 1997; 387: 861Crossref PubMed Scopus (1162) Google Scholar; Stolzenberg et al., 1997Stolzenberg E.D. Anderson G.M. Ackermann M.R. Whitlock R.H. Zasloff M. Epithelial antibiotic induced in states of disease.Proc Natl Acad Sci USA. 1997; 94: 8686-8690Crossref PubMed Scopus (193) Google Scholar). The sole cathelicidin in humans is LL-37 (Agerberth et al., 1995Agerberth B. Gunne H. Odeberg J. Kogner P. Boman H.G. Gudmundsson G.H. FALL-39, a putative human peptide antibiotic, is cysteine-free and expressed in bone marrow and testis.Proc Natl Acad Sci USA. 1995; 92: 195-199Crossref PubMed Scopus (417) Google Scholar; Frohm et al., 1997Frohm M. Agerberth B. Ahangari G. Stahle-Backdahl M. Liden S. Wigzell H. Gudmundsson G.H. The expression of the gene coding for the antibacterial peptide LL-37 is induced in human keratinocytes during inflammatory disorders.J Biol Chem. 1997; 272: 15258-15263Crossref PubMed Scopus (650) Google Scholar), and it is expressed in leukocytes and in a variety of epithelial surfaces. Cathelicidins and β-defensins have distinct and overlapping anti-microbial activity (Bals, 2000Bals R. Epithelial antimicrobial peptides in host defense against infection.Respir Res. 2000; 1: 141-150Crossref PubMed Scopus (384) Google Scholar; Nagaoka et al., 2000Nagaoka I. Hirota S. Yomogida S. Ohwada A. Hirata M. Synergistic actions of antibacterial neutrophil defensins and cathelicidins.Inflamm Res. 2000; 49: 73-79Crossref PubMed Scopus (163) Google Scholar). The activity of these peptides has been proposed to provide innate defense against a variety of potential microbial pathogens, and in the case of cathelicidins has been shown to be essential for skin defense against group A Streptococcus (GAS) (Nizet et al., 2001Nizet V. Ohtake T. Lauth X. et al.Innate antimicrobial peptide protects the skin from invasive bacterial infection.Nature. 2001; 414: 454-457Crossref PubMed Scopus (980) Google Scholar). Cathelicidins are a widely expressed family of mammalian anti-microbial peptides and like many such molecules are synthesized as a preproprotein (Zanetti et al., 1995Zanetti M. Gennaro R. Romeo D. Cathelicidins a novel protein family with a common proregion and a variable C-terminal antimicrobial domain.FEBS Lett. 1995; 374: 1-5Abstract Full Text PDF PubMed Scopus (583) Google Scholar). This preproprotein consists of a highly conserved signal sequence and cathelin domain but have substantial heterogeneity between species in the C-terminal domain encoding the mature active peptide. The anti-microbial activity of cathelicidins is only activated following proteolytic processing of the active C-terminal peptide from the cathelin domain of the preproprotein. As originally isolated from neutrophils granules, the human cathelicidin C-terminal peptide is a 37 amino acid cationic peptide after cleavage (Gudmundsson and Agerberth, 1999Gudmundsson G.H. Agerberth B. Neutrophil antibacterial peptides, multifunctional effector molecules in the mammalian immune system.J Immunol Methods. 1999; 232: 45-54Crossref PubMed Scopus (149) Google Scholar;Agerberth et al., 2000Agerberth B. Charo J. Werr J. et al.The human antimicrobial and chemotactic peptides LL-37 and alpha-defensins are expressed by specific lymphocyte and monocyte populations.Blood. 2000; 96: 3086-3093Crossref PubMed Google Scholar). LL-37 was also detected directly in human skin keratinocytes, but only where inflammation was present, suggesting this class of anti-microbial peptides functions primarily in response to injury rather than in modulating the surface colonization of the skin (Frohm et al., 1997Frohm M. Agerberth B. Ahangari G. Stahle-Backdahl M. Liden S. Wigzell H. Gudmundsson G.H. The expression of the gene coding for the antibacterial peptide LL-37 is induced in human keratinocytes during inflammatory disorders.J Biol Chem. 1997; 272: 15258-15263Crossref PubMed Scopus (650) Google Scholar). The eccrine sweat gland is one of the major cutaneous appendages, and it is a secretory as well as an excretory organ (Sato et al., 1989Sato K. Kang W.H. Saga K. Sato K.T. Biology of sweat glands and their disorders I Normal sweat gland function.J Am Acad Dermatol. 1989; 20: 537-563Abstract Full Text PDF PubMed Scopus (477) Google Scholar). The sole function of sweat has been considered to be thermoregulation during exposure to a hot environment or during physical exercise. Recently, a novel anti-microbial peptide "dermcidin" was found in the human sweat gland, and showed sweat may have an additional important role against various bacteria (Schittek et al., 2001Schittek B. Hipfel R. Sauer B. et al.Dermcidin: a novel human antibiotic peptide secreted by sweat glands.Nat Immunol. 2001; 2: 1133-1137Crossref PubMed Scopus (475) Google Scholar). Though anti-microbial peptides that have been reported are cationic peptides, this novel peptide is an anionic peptide whose mechanism of action remains unknown. In this investigation we evaluated expression of cathelicidins in sweat and the sweat gland, and show sweat has abundant constitutive production of this broad-spectrum and potent natural antibiotic. These data provide additional support to the hypothesis that sweat can play an important part in establishing a skin defense barrier against microbes and should be considered as an integral part of the innate immune system. Normal axillary skins from three healthy adult volunteers were sampled using a 6 mm punch biopsy. Sample acquisition was approved by the Committee on Investigations Involving Human Subjects of the University of California, San Diego. Tissue was immediately embedded in OCT compound using liquid nitrogen after biopsy and kept it at –80°C. For immunohistochemistry, fresh frozen sections were cut at 8 μm, and stored at –80°C until use. For in situ hybridization, fresh frozen section were cut at 8 μm, then fixed with 4% paraformaldehyde for 10 min at room temperature, and immersed in 0.1% active diethyl pyrocarbonate in phosphate-buffered saline (PBS: 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4.H2O, 1.4 mM KH2PO4) at 4°C overnight before processing. For western blot analysis, human sweat was collected from three healthy volunteers. 2 ml of the crude sweat was centrifuged at 15,000 r.p.m. (50,000 g) for 10 min, then supernatant collected. After repeating once, 1 ml of sample was concentrated by lyophilization, then resuspended in 50 μl of twice glass distilled water, and the rest of sample was stocked as crude sweat. All samples were stored at –20°C until use. Total protein concentration of the collected sweat was evaluated by BCA assay (protein assay reagent, Pierce, Rockford, IL) according to the manufacturer's instruction. Rabbit anti-LL-37 polyclonal antibody was kindly provided from Dr Birgitta Agerberth (Microbiology and Tumor Biology Center, Stockholm, Sweden) (Gudmundsson et al., 1996Gudmundsson G.H. Agerberth B. Odeberg J. Bergman T. Olsson B. Salcedo R. The human gene FALL39 and processing of the cathelin precursor to the antibacterial peptide LL-37 in granulocytes.Eur J Biochem. 1996; 238: 325-332Crossref PubMed Scopus (435) Google Scholar; Frohm et al., 1997Frohm M. Agerberth B. Ahangari G. Stahle-Backdahl M. Liden S. Wigzell H. Gudmundsson G.H. The expression of the gene coding for the antibacterial peptide LL-37 is induced in human keratinocytes during inflammatory disorders.J Biol Chem. 1997; 272: 15258-15263Crossref PubMed Scopus (650) Google Scholar). Rabbit anti-dermcidin polyclonal antibody was prepared as previously described (Schittek et al., 2001Schittek B. Hipfel R. Sauer B. et al.Dermcidin: a novel human antibiotic peptide secreted by sweat glands.Nat Immunol. 2001; 2: 1133-1137Crossref PubMed Scopus (475) Google Scholar). Chicken polyclonal antibodies were derived against cathelin domain (CATH) peptides by Aves Labs (Tigard, OR). Peptide amino acid sequences used for immunization were CZNLYRLLDLDPRPTMD for CATH. Briefly, hens were injected with a KLH conjugated peptides, immune eggs collected from these hens, and the IgY fraction purified from yolks. IgY fractions were then passed over affinity columns prepared with the respective immunizing peptides, the columns washed, and antibody eluted. The stock concentration of the chicken anti-cathelin antibody was 1.62 mg per ml. Synthetic LL-37, dermcidin (DCD), and DCD-derived (DCD-1L) peptides were prepared by Synpep Corporation (Dublin, OR). Peptide amino acid sequences were LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES, SSL-LEKGLDGAKKAVGGLGKLGKDAVEDLESVGKGAVHDVKDVLDSV, and SSLLEKGLDGAKKAVGGLGKLGKDAVEDLESVGKGAVHDVK-DVLDSVL for LL-37, DCD, and DCD-1L, respectively. The synthetic peptides were purified through a HPLC column and identity confirmed by mass spectrometry. Lyophilized sweat samples (10 μl) were separated by 16.5% Tris–tricine/peptide gel (BIO-RAD, Hercules, CA), and then transferred on to a nitrocellulose membrane (OSMONICS, Westborough, MA). For positive control, 5 pmol LL-37 synthetic peptide was applied. The membrane was treated with blocking solution [0.1% TTBS: 5% nonfat milk in 0.1% Tween 20/Tris-buffered saline (TBS: 150 mM NaCl, 10 mM Tris Base, pH 7.4)] for 60 min at room temperature, and then rabbit anti-LL37 polyclonal antibody (1: 5000 in blocking solution) was incubated with the membrane overnight at 4°C. After washing the membrane three times with 0.1% TTBS, horseradish peroxidase labeled goat anti-rabbit polyclonal antibody (1: 5000 in the blocking solution) was incubated with the membrane for 60 min, room temperature. After washing the membrane with 0.1% TTBS, the membrane was immersed in ECL solution (Western Lightning Chemiluminescence Reagents Plus, New Lifescience Products, Boston, MA) for 60 s then exposed to X-ray film (Kodak). For further confirmation of the identity of the band detected by western blot, the filter was stripped of antibody (62.5 mM Tris–HCl (pH 6.8), 2% sodium dodecyl sulfate, 100 mMβ-mercaptoethanol) at 50°C for 30 min, then reacted with a chicken anti-cathelin antibody (1: 10000) and horseradish peroxidase anti-chicken IgY goat antibody (1: 10000, Aves Labs, OR). The filter was stripped of antibody again, then reacted with a rabbit anti-dermcidin antibody (1: 6000) and goat horse-radish peroxidase anti-rabbit IgG antibody as described above. To estimate the concentration of LL-37, quantitative dot-blot and western blot analysis was performed. Ten microliters of crude sweat sample were compared with a standard curve of various concentrations of synthetic peptide applied on to nitrocellulose filter. After blotting, LL-37 on the nitrocellulose filter was detected as described above. For detection of LL-37 and dermcidin peptides, rabbit anti-LL-37 and anti-dermcidin polyclonal antibodies were used for immunostaining. Tissue sections were immersed in PBS after 4% paraformaldehyde fixation for 5 min, and endogenous peroxidase activity blocked with a 30 min incubation in 0.3% H2O2 in methanol. After washing with PBS, sections were blocked with 2% goat serum in PBS for 30 min, then incubated with primary antibody, rabbit anti-LL-37 (1: 1000) or rabbit anti-dermcidin (1: 3000) in PBS, 3% bovine serum albumin for 60 min. Sections were washed in PBS and the signals detected with Vectorstain ABC Elite Rabbit kit (Vector Laboratories, Burlingame, CA) following the manufacturer's instructions. As a negative control, the polyclonal antibody was replaced by normal rabbit preimmune IgG diluted with PBS containing 3% BSA at the same protein concentration as that used for the primary antibody. Finally, sections were incubated in 0.02% diaminobenzidine with 0.05% H2O2 in PBS for 1 min and counterstained with Mayer's hematoxylin for 30 s. All procedures were carried out at room temperature. For preparing in situ probes, PCR products of the cathelin domain sequence (CDS, 236 bp) and LL-37 mature form sequence (LL-37mfs, 137 bp) were amplified, respectively. This product was sequenced to confirm identity, then inserted in the PCR II-TOPO vector (TOPO TA Cloning version M; Invitrogen, Carlsbad, CA). Plasmid DNA was subsequently purified using a DNA Extraction Maxi Kit (Qiagen, Valencia, CA) and linearized with restriction enzymes (HindIII and XbaI for CDS, HindIII and XhoI for LL-37mfs) for anti-sense and sense RNA probe, respectively. Digoxigenin-labeled riboprobes were prepared on linearized plasmid using the digoxigenin RNA labeling kit (SP6/T7, Roche Molecular Biochemicals, Mannheim, Germany) following the instructions provided. Two microliters of the plasmid containing cDNA for CDS and LL-37mfs were blotted on nitrocellulose filters (OSMONICS, Westborough, MA) and ultraviolet cross-linked to confirm the specificity of the labeled probes, respectively. After washing with PBS at room temperature for 10 min, sections were treated with a 1 M triethanolamine solution (pH 8.0) containing 0.25% acetic anhydride for 15 min at 37°C, washed with PBS at least three times, then treated with 100% ethanol for 5 min, and then dried. Prehybridization was performed with 50% formamide in ×2 sodium saline citrate (3 M sodium chloride, 0.03 M sodium citrate) for 30 min at 45°C. After removing excess solution, sections were hybridized for 16 h at 45°C with sense or anti-sense digoxigenin-labeled cRNA probes (1 μg per ml) in hybridization solution (1 mg per ml yeast tRNA, 20 mM Tris–HCl buffer (pH 8.0), 2.5 mM ethylenediamine tetraacetic acid, 1×Denhardt's solution, 0.3 M NaCl, 50% deionized formamide, 50% dextran sulfate). Stringent washing was performed for 60 min at 45°C with 50% formamide in ×2 sodium saline citrate, and for 10 min at 37°C with ×2 sodium saline citrate. RNase treatment was carried with 40 μg per ml of RNase-A (Roche Molecular Biochemicals) for 30 min at 37°C. After washing with the ×2 sodium saline citrate for 30 min at 37°C, sections were reacted with anti-digoxigenin alkaline phosphate Fab fragment antibody (1: 500, in PBS; Roche Molecular Biochemicals) for 5 h at room temperature. Alkaline phosphate was visualized by incubation with 5-bromo-4-chloro-3-indolyl phosphate (X-phosphate) and nitroblue tetrazolium with addition of levamisole solution (DAKO, Carpinteria, CA) overnight at room temperature. Methyl green stain was used for the nuclear counter stain. RNA preservation in tissues was confirmed by similar hybridizations using a keratin 14 anti-sense probe (insert 175 bp). To evaluate the anti-microbial activity of LL-37, dermcidin, and total sweat, colony-forming unit assay was performed with Staphylococcus aureus (isolated from clinical sample), GAS (NZ131), and enteroinvasive Escherichia coli O29 as described (Porter et al., 1997Porter E.M. van Dam E. Valore E.V. Ganz T. Broad-spectrum antimicrobial activity of human intestinal defensin 5.Infect Immun. 1997; 65: 2396-2401Crossref PubMed Google Scholar). Before analysis, we determined the concentration of the bacteria in culture by plating different bacterial dilutions (at A600, 1.0 corresponded to 3.5×109 per ml for S aureus, 2.5×108 and GAS, and 3.5×108 for E. coli). Cells were washed twice with 10 mM sodium phosphate buffer (20 mM NaH2PO4.H2O, 20 mM Na2HPO4.7H2O) and diluted to a concentration of 2×106 cells per ml (S. aureus, GAS) or 2×105 cells per ml (E. coli) in sweat buffer (40 mM NaCl, 10 mM KCl, 1 mM CaCl2, 1 mM MgCl2, 1 mM Na dihydrogen phosphate, pH 6.5) (Schittek et al., 2001Schittek B. Hipfel R. Sauer B. et al.Dermcidin: a novel human antibiotic peptide secreted by sweat glands.Nat Immunol. 2001; 2: 1133-1137Crossref PubMed Scopus (475) Google Scholar) or phosphate buffer. S. aureus and E. coli were incubated for 4 h at 37°C with various concentrations of LL-37 or dermcidin peptides in 50 μl of buffers using wells of a 96 well round bottom tissue culture plate (Costar 3799, Corning inc., NY). GAS was incubated for 1 h due to the poor ability of GAS to grow in these buffers. After incubation, the cells were diluted from 10× to 105×, and each of 20 μl of those solutions were plated in triplicate on tryptic soy broth (for S. aureus) and Todd Hewitt broth (for GAS and E. coli), then the mean number of colonies determined. The number of cfu per ml was calculated, and the bactericidal activities of the tested regents were calculated as follows: (cell survival after peptide incubation)/(cell survival after incubation without peptide)×100, which represented the percentage of cells that were alive (% live). Normal human sweat was evaluated by western blot with two antibodies independently derived against distinct domains of the cathelicidin proprotein: anti-cathelin and anti-LL-37. LL-37 was present in human sweat (Figure 1a ). Three bands were detected with anti-LL37 at 18, 14 and 5 kDa. A single band was seen with anti-cathelin, migrating identically to that seen at 18 kDa with anti-LL-37. This 18 kDa corresponds to the expected size of the full-length cathelicidin, whereas the 5 kDa band corresponds to the expected mature anti-microbial peptide. Parallel to exploring the LL-37 expression, dermcidin protein expression was also evaluated by western blotting. After detection of LL-37 protein, the filter was stripped of antibody and dermcidin peptide was evaluated (Figure 1a). The antibody detected three bands at 20, 14.5, and 10 kDa. To evaluate more quantitatively the concentration of cathelicidin and dermcidin in normal sweat samples, dot-blot and western blot analysis was performed by comparison with known amounts of synthetic LL-37 peptide. LL-37 content of the human sweat was found to vary considerably between patients, with an average concentration of ≈1 μM in the crude sweat collected from normal volunteers (Figure 1b: LL-37). Furthermore, western blot analysis of normal sweat demonstrated that cathelicidin was processed into active LL-37 and represented approximately 10% of the total immunoreactive material (Figure 1c). In contrast to the variability in LL-37 detected between individual sweat preparations, the relative amount of dermcidin was similar in identical patient samples (Figure 1b: DCD). The concentration of dermcidin could not be determined as the antigenic epitope of the dermcidin detecting antibody was directed against the pro-domain and is not present on the active dermcidin peptide. To explore further cathelicidin expression and localization, skin biopsies were examined for LL-37 protein expression by immunohistochemistry. Abundant LL-37 was detected in the eccrine gland (Figure 2a,c). Staining was diffusely located in the cytoplasm of both of the clear and dark cells in the secretory gland (Figure 2a). Ductal epithelium in dermis also showed protein expression, and strong staining could be found in the apical portion in the ductal cell (Figure 2c). Unlike results from western blots of individual collections of secreted sweat, no variability was seen between patient expression of LL-37 by immunohistochemistry. The localization of dermcidin contrasted with the localization of LL-37. Dermcidin was seen only in the secretory glands but not in the ducts (Figure 2b,d). Expression was the strongest on the surface of dark cells in the eccrine gland proper (Figure 2b). To explore expression of cathelicidin mRNA in eccrine appendages, in situ hybridization was performed. Nonradioactive in situ hybridization with digoxigenin-labeled cRNA anti-sense probes for the immature cathelin domain (CDS) (Figure 3a) and LL-37 mature peptide (mfs) (Figure 3c) showed that the LL-37 mRNA was expressed in tissues previously found positive by immunohistochemistry. Intense signal was found in the cytoplasm of the gland cells, around the nucleoli, and mRNA was also found in the ductal epithelium in dermis by using both CDS and LL-37mfs probes. No signal was detected with sense probes as control in every tissue (Figure 3b,d). Relative abundance of LL-37 mRNA was similar in all samples. To investigate the antibacterial activity of LL-37 in sweat, bactericidal assays were performed. S. aureus or E. coli was incubated for 4 h, and GAS was incubated for 1 h, at 37°C in sweat buffer (Figure 4a) or phosphate buffer (Figure 4b). LL-37 (8 μM), killed 97% of E. coli and 80% of S. aureus in sweat buffer, and 2 μM of LL-37 killed 79% of E. coli and 85% of S. aureus in phosphate buffer. In sweat buffer, 4 μM of LL-37 killed 68% of GAS and 4 μM of LL-37 killed 66% of GAS in phosphate buffers after 1 h. Synthetic dermcidin (DCD) or DCD-1L had no effect on E. coli, S. aureus, or GAS in either buffer at peptide concentrations under 16 μM, nor could we identify synergistic activity of LL-37 and DCD or DCD-1L. Synthetic DCD was effective in killing E. coli (100%) and S. aureus (40%) in sweat buffer at 80 μM, and 160 μM of DCD killed 66%S. aureus in sweat buffer (Figure 4c). To determine if crude human sweat was anti-microbial, individual sweat preparations previously evaluated for LL-37 and dermcidin composition in Figure 1 were inoculated with GAS and bacterial survival measured (Figure 4d). GAS survival varied with the observed concentration of LL-37. Patient samples 1 and 3, which had LL-37 in greatest abundance, killed over 95% of the initial inoculum, whereas in sample 2, which had little detectable LL-37 but similar amounts of dermcidin, the majority of GAS survived. Skin is constantly exposed to a variety of microbial pathogens. Thus, its function as a protective barrier to resist infection is important. Anti-microbial peptides are essential elements of epithelial defense participating in a variety of immune mechanisms both direct and indirect (Zasloff, 2002Zasloff M. Antimicrobial peptides of multicellular organism.Nature. 2002; 415: 389-395Crossref PubMed Scopus (6269) Google Scholar). Cathelicidin anti-microbial peptides are constitutively expressed in neutrophils, lung and gut epithelium, but have only been seen at significant levels in skin during inflamed conditions such as wound healing (Frohm et al., 1997Frohm M. Agerberth B. Ahangari G. Stahle-Backdahl M. Liden S. Wigzell H. Gudmundsson G.H. The expression of the gene coding for the antibacterial peptide LL-37 is induced in human keratinocytes during inflammatory disorders.J Biol Chem. 1997; 272: 15258-15263Crossref PubMed Scopus (650) Google Scholar; Gallo et al., 1997Gallo R.L. Kim K.J. Bernfield M. Kozak C.A. Zanetti M. Merluzzi L. Gennaro R. Identification of CRAMP, a cathelin-related antimicrobial peptide expressed in the embryonic and adult mouse.J Biol Chem. 1997; 272: 13088-13093Crossref PubMed Scopus (305) Google Scholar; Bals et al., 1998Bals R. Wang X. Zasloff M. Wilson J.M. The peptide antibiotic LL-37/hCAP-18 is expressed in epithelia of the human lung where it has broad antimicrobial activity at the airway surface.Proc Natl Acad Sci USA. 1998; 95: 9541-9546Crossref PubMed Scopus (584) Google Scholar). Simultaneously, anti-microbial peptides of the defensin family, in particular human β-defensin 1 (hBD1), 2 (hBD2), and 3 (hBD3), have recently been described in normal skin (Ali et al., 2001Ali R.S. Falconer A. Ikram M. Bissett C.E. Cerio R. Quinn A.G. Expression of the peptide antibiotics human beta defensin-1 and human beta defensin-2 in normal human skin.J Invest Dermatol. 2001; 117: 106-111Crossref PubMed Scopus (201) Google Scholar), but are most abundant in disease conditions such as psoriasis (Harder et al., 2001Harder J. Bartels J. Christophers E. Schroder J.M. Isolation and characterization of human beta-defensin-3, a novel human inducible peptide antibiotic.J Biol Chem. 2001; 276: 5707-5713Crossref PubMed Scopus (1126) Google Scholar). In this investigation we asked if cathelicidins are expressed in significant amounts in sweat, thus providing an additional innate anti-microbial defense system under noninflammatory conditions. Our results show that the human cathelicidin LL-37 is expressed in sweat, and is present in both mature and immature forms. A combination of distinct antibodies against different domains of human cathelicidin was used in this study. The rabbit anti-LL-37 antibody derived against the C-terminal mature peptide identified an 18 kDa immature form, a 14 kDa band and the mature peptide. A chicken antibody derived against the N-terminal precursor cathelin domain detected only the expected 18 kDa form. In all samples evaluated, the majority of soluble immunoreactive material was found as this propeptide (Agerberth et al., 1995Agerberth B. Gunne H. Odeberg J. Kogner P. Boman H.G. Gudmundsson G.H. FALL-39, a putative human peptide antibiotic, is cysteine-free and expressed in bone marrow and testis.Proc Natl Acad Sci USA. 1995; 92: 195-199Crossref PubMed Scopus (417) Google Scholar). This precursor protein blocks the anti-microbial activity of LL-37 until it is activated by proteinase cleavage to free the C-terminal active peptide fragment (Zanetti et al., 1990Zanetti M. Litteri L. Gennaro R. Horstmann H. Romeo D. Bactenecins, defense polypeptides of bovine neutrophils, are generated from precursor molecules stored in the large granules.J Cell Biol. 1990; 111: 1363-1371Crossref PubMed Scopus (125) Google Scholar; Shi and Ganz, 1998Shi J. Ganz T. The role of protegrins and other elastase-activated polypeptides in the bactericidal properties of porcine inflammatory fluids.Infect Immun. 1998; 66: 3611-3617PubMed Google Scholar). For example, proteinase 3 is responsible for cleavage of LL-37/hCAP-18 after exocytosis from activated human neutrophils (Sorensen et al., 2001Sorensen O.E. Follin P. Johnsen A.H. Calafat J. Tjabringa G.S. Hiemstra P.S. Borregaard N. Human cathelicidin, hCAP-18, is processed to the antimicrobial peptide LL-37 by extracellular cleavage with proteinase 3.Blood. 2001; 97: 3951-3959Crossref PubMed Scopus (663) Google Scholar). Finding LL-37 as both the inactive pro-form and active peptide in sweat suggests it can be processed after secretion, perhaps by one of the several proteases present at the skin surface (Fraki and Hopsu-Havu, 1976Fraki J.E. Hopsu-Havu V.K. Human skin proteases. Fractionation of psoriasis scale proteases and separation of a plasminogen activator and a histone hydrolysing protease.Arch Dermatol Res. 1976; 256: 113-126Crossref PubMed Scopus (34) Google Scholar). It is also possible that either host proteases released after barrier disruption, or microbial proteases expressed during infection, could serve to cleave sweat cathelicidin, thus further providing natural antibiotic protection when needed. Recently, a novel anti-microbial peptide "dermcidin" was reported as expressed in sweat glands (Schittek et al., 2001Schittek B. Hipfel R. Sauer B. et al.Dermcidin: a novel human antibiotic peptide secreted by sweat glands.Nat Immunol. 2001; 2: 1133-1137Crossref PubMed Scopus (475) Google Scholar). Dermcidin has a net negative charge of -2 in contrast to anti-microbial peptides such as cathelicidins and defensins that are highly cationic. This difference suggests the mechanism of bacterial killing by dermcidin is much different from cathelicidins. Our results show that expression of dermcidin and cathelicidin is different; dermcidin is secreted from granules in the dark cell, but LL-37 is present in the epithelium of both gland and duct. In this study, dermcidin synthetic peptides had poorer activity than synthetic LL-37 against E. coli, GAS, and S. aureus. We speculate that, unlike cathelicidins, dermcidin anti-microbial activity depends on structural conformations or the presence of additional cofactors not yet identified. This dependence on unknown cofactors or structure is suggested by the differences seen in activity between synthetic peptide and the more potent native peptide previously described (Schittek et al., 2001Schittek B. Hipfel R. Sauer B. et al.Dermcidin: a novel human antibiotic peptide secreted by sweat glands.Nat Immunol. 2001; 2: 1133-1137Crossref PubMed Scopus (475) Google Scholar). Further work is necessary to understand the mechanism of action of this unique anti-microbial protein. Cathelicidins have broad-spectrum activity against gram-positive and gram-negative bacteria such as S. aureus, Pseudomonas aeruginosa, and E. coli, as well as against fungi and envelope viruses (Bals et al., 1998Bals R. Wang X. Zasloff M. Wilson J.M. The peptide antibiotic LL-37/hCAP-18 is expressed in epithelia of the human lung where it has broad antimicrobial activity at the airway surface.Proc Natl Acad Sci USA. 1998; 95: 9541-9546Crossref PubMed Scopus (584) Google Scholar; Johansson et al., 1998Johansson J. Gudmundsson G.H. Rottenberg M.E. Berndt K.D. Agerberth B. Conformation-dependent antibacterial activity of the naturally occurring human peptide LL-37.J Biol Chem. 1998; 273: 3718-3724Crossref PubMed Scopus (482) Google Scholar; Turner et al., 1998Turner J. Cho Y. Dinh N.N. Waring A.J. Lehrer R.I. Activities of LL-37, a cathelin-associated antimicrobial peptide of human neutrophils.Antimicrob Agents Chemother. 1998; 42: 2206-2214Crossref PubMed Google Scholar). We have also shown that LL-37 or the murine cathelicidin CRAMP are potent inhibitors of GAS, one of the most prevalent and potentially invasive skin pathogens (Dorschner et al., 2001Dorschner R.A. Pestonjamasp V.K. Tamakuwala S. et al.Cutaneous injury induces the release of cathelicidin anti-microbial peptides active against group A Streptococcus.J Invest Dermatol. 2001; 117: 91-97Crossref PubMed Scopus (461) Google Scholar). In this prior work, as well as the present data on in vitro killing by LL-37 in sweat buffers, the concentration of LL-37 required for efficient anti-microbial action generally exceeds 2 μM; however, the concentration of mature LL-37 observed here was at least 20-fold less than what is apparently necessary for anti-microbial activity. A greatly increased local concentration of LL-37 would be expected after normal evaporation, and this may enhance the effectiveness of newly secreted anti-microbial peptides in solution. Crude sweat collected in this study, however, affected bacterial survival without concentration, and this activity correlated with the relative abundance of LL-37 detected in each sample. This observation may be explained by the presence of anti-microbial molecules other than cathelicidins or dermcidin. These other molecules may be primarily responsible for the anti-microbial action of the crude sweat samples, or may act synergistically with the cathelicidin and/or dermcidin that has been identified in sweat in this study. Precedence for synergistic activity has been observed with combinations of LL-37 and defensin (HNP-1) (Nagaoka et al., 2000Nagaoka I. Hirota S. Yomogida S. Ohwada A. Hirata M. Synergistic actions of antibacterial neutrophil defensins and cathelicidins.Inflamm Res. 2000; 49: 73-79Crossref PubMed Scopus (163) Google Scholar). One microgram of LL-37 per ml (about 0.1 μM) with 10 μg HNP-1 per ml kills 50% of E. coli and 60% of S. aureus in 10 mM phosphate buffer containing 150 mM NaCl. In our own experiments, 4 μM LL-37 increases the potency of hBD2 against S. aureus 4-fold (Ong et al., 2002Ong P.Y. Ohtake T. Brandt C. Strickland I. Boguniewicz Z. Ganz T. Gallo R.L. Leung D.Y. Endogenous antimicrobial peptides and skin infections in atopic dermatitis.N Engl J Med. 2002; 347: 1151-1160Crossref PubMed Scopus (1550) Google Scholar). Further study of the combined activities and regulation of skin and sweat anti-microbial peptides will greatly advance our understanding of this system of skin defense. An association between a lack of development of eccrine glands and decreased immune defense has been described (Doffinger et al., 2001Doffinger R. Smahi A. Bessia C. et al.X-linked anhidrotic ectodermal dysplasia with immunodeficiency is caused by impaired NF-kappaB signaling.Nat Genet. 2001; 27: 277-285Crossref PubMed Scopus (623) Google Scholar; Jain et al., 2001Jain A. Ma C.A. Liu S. Brown M. Cohen J. Strober W. Specific missense mutations in NEMO result in hyper-IgM syndrome with hypohydrotic ectodermal dysplasia.Nat Immunol. 2001; 2: 223-228Crossref PubMed Scopus (312) Google Scholar). In these patients the immune defect results not from a primary defect in sweat but from a common dependency on nuclear factor-κB signaling mechanisms in eccrine development and immune response. This intriguing association between development of the eccrine apparatus and the immune system supports a hypothesis that during evolution the immune defense role of the eccrine appendage was greater than at present. Now, human keratinocytes can rapidly respond to injury by production of anti-microbial peptides, and neutrophils are recruited as a means of delivery of natural antibiotics. These additional innate immune systems diminish dependence on sweat as a defense barrier. Furthermore, it has become clear that several unique systems of gene regulation and protein processing contribute to the function of anti-microbial peptides in different tissues. Thus, current information regarding anti-microbial effector molecules, combined with advancing understanding of the control of innate immunity, suggests these molecules are likely to play an important part in host defense and may provide important diagnostic and therapeutic advances towards the treatment of disease. This work was supported by VA Merit award and NIH grants AI48176 and AR45676 (R.G.), and the Japan-North America Medical Exchange Foundation (M.M.). We thank Dr K. Hase (Laboratory of Mucosal Immunology, Department of Medicine, University of California, San Diego) for his technical assistance, Prof. T. Koji (Department of Histology and Cell Biology, Nagasaki University School of Medicine, Japan) for histochemical advice, and Dr Birgitta Agerberth for gift of anti-LL-37 antibody.