FEBS 24035 FEBS Letters 480 (2000) 147-150

LEAP-1, a novel highly disulfde-bonded human peptide, exhibits antimicrobial activity1

Alexander Krause, Susanne Neitz2, Hans-Juirgen Maigert, Axel Schulz, Wolf-Georg Forssmann, Peter Schulz-Knappe2, Knut Adermann*

Niedersaichsisches Institutfuir Peptid-Forschung (IPF), Feodor-Lynen-Strasse 31, D-30625 Hannover, Germany

Received 20 June 2000; accepted 31 July 2000 Edited by Pierre Jolles

Abstract We report the isolation and characterization of a novel human peptide with antimicrobial activity, termed LEAP-l (liver-expressed antimicrobial peptide). Using a mass spectrometric assay detecting cysteine-rich peptides, a 25-residue peptide containing four disulfide bonds was identified in human blood ultrafiltrate. LEAP-l expression was predominantly detected in the liver, and, to a much lower extent, in the heart. In radial diffusion assays, Gram-positive Bacillus megaterium, Bacillus subtilis, Micrococcus luteus, Staphylococcus carnosus, and Gram-negative Neisseria cinerea as well as the yeast Saccharomyces cerevisiae dose-dependently exhibited sensitivity upon treatment with synthetic LEAP-l. The discovery of LEAP-l extends the known families of mammalian peptides with antimicrobial activity by its novel disulfide motif and distinct expression pattern. © 2000 Federation of European Biochemical Societies. Published by Elsevier Science B.V. All rights reserved.

Key words: Antimicrobial peptide; Liver; Hemofltrate; Cysteine-rich peptide

l. Introduction

Antimicrobial peptides are an important component of innate immunity in many species, including plants, insects, fsh, amphibians and mammals [1-3]. Although especially the latter have a highly efcient system of adaptive immunity, there is growing evidence that in animals antimicrobial peptides such as defensins [2,3], protegrins [4] and cathelicidins [5] represent a considerable part of the immune system in the defense of cellular pathogens. Antimicrobial peptides are predominantly expressed in epithelial cells lining the respiratory, alimentary and genito-urinary tract as well as in granules of circulating phagocytes, and they exert their activity mainly by disrupting the cell membranes of cellular pathogens [3,6-8]. This rather unspecifc mode of action is suggested to be responsible for the broad-spectrum activity of many peptide antibiotics. Despite their diverging efects, most antimicrobial peptides share

*Corresponding author. Fax: (49)-511-5466132. E-mail: [email protected]

The nucleotide sequence data reported in this paper have been submitted to the GenBank/EBI Data Bank with accession number AJ277280. Scanning of this sequence against the data base resulted in the identifcation of related sequences with the accession numbers AD000684 and P81172. Present address: BioVisioN GmbH and Co. KG, Feodor-Lynen-Strasse 5, D-30625 Hannover, Germany.

common structural elements in being amphipathic and carrying net positive charges at physiological pH due to basic amino acid residues [1,7].

As many antimicrobial peptides contain multiple disulfde bonds, we intended to isolate small cysteine-rich peptides from human blood ultrafltrate, which was shown earlier to be a source of defensins [9]. Using a cysteine alkylation assay, we report the isolation and characterization of a novel extremely cysteine-rich 25-residue peptide that was termed LEAP-1 (liver-expressed antimicrobial peptide). By analysis of the primary structure and the corresponding cloned cDNA, we demonstrate that LEAP-1 is a novel peptide which resembles other peptide antibiotics in that it exhibits comparable structural elements and antimicrobial activity, but difers signifcantly in the predominant expression in liver.

2. Materials and methods

2.1. IsolationandstructuralcharacterizationofLEAP-1

LEAP-1 was isolated during a comprehensive search for cysteine-containing peptides in fractions from human blood ultrafltrate [10]. In brief, the cysteine content of a peptide is determined after reduction of disulfdes and subsequent carboxamidomethylation of the resulting free thiol groups inducing a defned mass shift for each single cysteine residue. 10 Jg of freeze-dried fractions were dissolved in 4-methylmorpholine (50 Jl, 200 mM, pH 8.3) and dithiothreitol (5 Jl, 400 mM) was added. After incubation for 60 min at 45°C, iodoacetamide (20 Jl, 500 mM) was added, and after a further 30 min at ambient temperature, the samples were desalted by solid-phase extraction. Peptides were eluted with 100 Jl 70% acetonitrile in 0.1% trifuoroacetic acid (TFA), and the fractions were subjected to MALDI-TOF mass spectrometry prior to and after carboxamidomethylation on a Voyager-DE STR (PerSeptive Biosystems). LEAP-1 was identifed as a peptide with a molecular mass of 2789.8 Da and a content of eight cysteine residues calculated from the resulting mass shift of +464.5 Da in the alkylated fraction (Fig. 1). For purifcation of LEAP-1, a 10-mg aliquot of the corresponding peptide fraction was dissolved in 0.06% TFA and separated in three reverse-phase HPLC steps. Final purifcation was performed on a C18 ODS-AQ column (4.6X250 mm, 5 Jm, 120 A , UV detection at 214/280 nm, YMC Europe, Schermbeck, Germany). NH2-terminal amino acid sequence analysis was performed by Edman degradation using a Procise 494 sequencer (PE Biosystems, Foster City, CA, USA).

2.2. Peptide synthesis

LEAP-1 was assembled on a preloaded Tentagel R Trt-Thr(tBu) resin (Rapp Polymere, Tuibingen, Germany) using standard Fmoc chemistry and trityl-protected cysteines. The peptide was deprotected and cleaved from the resin with a mixture of TFA, ethanedithiol and water(94:3:3). After fltrationandprecipitationwith tert-butylmethylether, the reduced peptide obtained was purifed by reverse-phase HPLC on a Vydac C18 column (The Separations Group, Hesperia, CA, USA). Fractions containing the product were identifed using electrospray mass spectrometry (ESIMS; Mr determined, 2797.5 Da; Mr calculated, 2797.3 Da) and analytical HPLC, and subjected

0014-5793/00/$20.00 © 2000 Federation of European Biochemical Societies. Published by Elsevier Science B.V. All rights reserved. PII: S0014-5793(00)01920-7

A. Krause et al./FEBS Letters 480 (2000) 147-150

to oxidation in phosphate bufer at pH 7.4. The fully disulfde-bonded peptide was then purifed by reverse-phase HPLC (Vydac C18). Purity was checked on a Nucleosil C18 PPN column (Macherey and Nagel, Duiren, Germany) and by capillary electrophoresis on a Beckman P/ACE System 2000. Synthetic LEAP-1 coeluted and comigrated with native LEAP-1 during reverse-phase HPLC and capillary electrophoresis, respectively. The molecular mass of the product was determined by ESIMS (Mr experimental 2789.5 Da; Mr calculated 2789.3 Da).

2.3. Antimicrobial activity

2.3.1. Radial diiusion assay [11]. Bacterial cells in mid-logarithmic phase (500 Jl) were added to 100 ml of previously boiled 10 mM sodium phosphate bufer (pH 7.2, 37°C) containing 0.03% (w/v) tryptic soy broth, 0.8% (w/v) NuSieve GTG agarose (FMC BioProducts, Rockland, ME, USA) and 0.02% (v/v) Tween 20. Bacteria growing at 37°C were plated with 0.2% type 1 agarose (w/v, low EEO, Sigma-Aldrich) and 0.6% NuSieve agarose. The mixture was pipetted into sterile Petri dishes and evenly spaced wells were obtained by punching out gel cylinders with a diameter of 3 mm. After incubation of the bacteria with synthetic LEAP-1 (1-11 Jg) for 18 h, growth inhibition appeared as plain circular zones, and their diameters were determined [12]. Synthetic casein-(16-57), a 42-residue peptide fragment of human casein a-s1, was used as a negative control in order to exclude any efect due to contaminants derived from peptide synthesis procedures. Experiments were carried out in duplicate. Water was used as additional negative control.

2.3.2. Determination of colony forming units (CFU). S. cerevisiae was cultured overnight at 30°C in YPD medium (1% yeast extract, 2% peptone, w/v, pH 6.5). To obtain mid-logarithmic phase germs, cells were subcultured at 1:100 for 3 h and harvested by centrifugation at 600Xg and 4°C for 10 min. After washing of the pellets a dilution containing approximately 2X10cells/ml in 10 mM sodium phosphate bufer was prepared. Dry LEAP-1 (1-12.5 Jg) was then mixed with 100 Jl of cell suspension and incubated for 2 h at 30°C. Subsequently, 10 Jl of this mixture was withdrawn and the reaction was stopped by a 500-fold dilution with ice-cold sodium phosphate bufer, and aliquots were manually spread on YPD agar plates. After 16 h at 30°C, colonies were counted and inhibition percentiles were calculated. Three single experiments were carried out at least in triplicate. An analogous procedure was performed with B. subtilis using LB medium (pH 7.4).

2.4. Molecular biological standard methods and cDNA cloning

RNA extraction, cDNA frst strand synthesis, polymerase chain reaction (PCR), reverse transcription/polymerase chain reaction (RT-PCR), and DNA fuorescence sequencing were performed as described earlier [13-15]. LEAP-1-specifc cDNA fragments were amplifed from human liver cDNA by means of standard RT-PCR methods using the primers SN-2S (CCGCTTGCCTCCTGCTCCTCCTCCTCCTCG) and SN-6AS (CGATGACAGCAGCCGCAGCAGAAAATGCAG) derived from a LEAP-1 gene-containing sequence (accession number AD000684, positions 31 400-34 200). Nested 3’and 5′-RACE PCR for amplifcation of LEAP-1 cDNA ends was performed using Marathon-Ready human liver cDNA (Clontech, Palo Alto, CA, USA). The gene-specifc primers SN-2S and SN-3S were used for 3′-RACE, and SN-5AS and SN-4AS for 5′-RACE according to the manufacturer’s instructions. The fragments obtained were cloned in pGEM T vector (Promega, Mannheim, Germany) and subsequently sequenced. The specifcity of the inserts for LEAP-1 was verifed by sequence comparisons using the MacMolly program package (Softgene).

2.5. Real-time RT-PCR analysis

To quantitatively evaluate the expression of the LEAP-1 gene in various human tissues we performed real-time RT-PCR [16] by means of a Prism 7700 Sequence Detection System (PE Biosystems). Primers and probes were designed with the assistance of the Primer Express software (PE Biosystems). The probe constructed fanked intron 2 of the gene (FAM-CCAGCTGGATGCCCATGTTCCAGA-TAMRA) while the PCR primers were specifc for exons 2 and 3. For relative quantifcation, the values calculated for LEAP-1 gene expression in each sample were divided by the corresponding values for the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) housekeeping gene. All quantitative real-time RT-PCR experiments were performed in duplicate using standard reagents and mixes from PE Biosystems and 96-well microplates containing cDNAs from 14 diferent human tissues, six standard curve samples (prepared from HUH-7 cells) and two non-template controls (purifed water, ACS reagent, Sigma-Aldrich).

3. Results and discussion

The peptide isolation procedure utilizing a mass spectrometric assay for cysteine-rich peptides (Fig. 1) resulted in the isolation of a peptide which was shown to be homogeneous by HPLC and capillary electrophoretic analysis. Sequence analysis revealed the amino acid sequence DTHFPIXIFXXGXXHRSKXGMXXKT (Fig. 2), where X represents residues which gave blank cycles during Edman sequencing. As a diference in molecular mass between the original sample and the reduced and subsequently carboxamidomethylated sample corresponded to eight cysteine residues, it was concluded that all Xs of the peptide sequence represent cysteine. The corresponding sequence of the peptide containing eight disulfde-bonded cysteine residues would have a molecular mass of 2789.3 Da. This value accords exactly with the molecular mass that was found experimentally for the isolated peptide which we termed LEAP-1. To our knowledge, a pep-tide containing four disulfde bonds within a stretch of 17 amino acid residues is unique in the human proteome. In the COOH-terminal moiety, LEAP-1 contains basic amino acid residues such as lysine and arginine. Therefore, LEAP1 exhibits structural features which are known from other peptide antibiotics such as the defensins [3,8]. To verify the structure determined and to investigate antimicrobial efects of LEAP-1, a corresponding synthetic peptide was produced. Oxidation of the reduced synthetic precursor readily produced a major product identical with the native counterpart.

To determine the amino acid sequence of the precursor protein and to investigate the expression pattern of the corresponding gene, the cDNA specifc for LEAP-1 was cloned and sequenced. For this purpose, sequence comparisons of the LEAP-1 amino acid sequence with entries of the EMBL data base were carried out. We could identify a putative gene for LEAP-1 being located within a sequenced contig of human chromosome 19 (accession number AD000684). From the given nucleotide sequence of the gene we then deduced the most probable exon regions by searching for typical exon/in-

Fig. 1. Comparison of MALDI-TOF mass spectra of peptide fractions 7-30 before (upper trace) and after carboxamidomethylation (lower trace). The mass shift of 464.5 Da indicates the content of eight cysteine residues in LEAP-1 which was confrmed by Edman degradation with the purifed peptide.

A. Krause et al./FEBS Letters 480 (2000) 147-150

Fig. 2. LEAP-1 precursor cDNA sequence and deduced amino acid sequence. Coding regions are printed in capital letters. The positions of the primers used for preparative RT-PCR and the positions of the putative introns 1 and 2 within the corresponding gene are indicated. The sequence representing the isolated antimicrobially active LEAP-1 and the potential polyadenylation signal are underlined. The typical secretory signal sequence 1-24 is printed in italics.

tron boundary sequences and open reading frames. Following this strategy, six PCR primers located within the putative exons 1 and 3 of the LEAP-1 gene were derived. In PCR reactions with cDNA from human lung, kidney, liver and tonsils, we obtained homogeneous bands with each of the possible nine primer combinations. The most homogeneous and intensive band was obtained from human liver cDNA and the primer combination SN-2S/SN-6AS. The PCR product was cloned, sequenced and verifed as LEAP-1-specifc. To amplify the 3′-as well as the 5′-end of the LEAP-1 cDNA, we performed nested 3′-and 5′-RACE PCR. The RACE PCR products obtained exhibited comparable sizes and could both be verifed as LEAP-1-specifc. By identifcation of overlapping regions and the largest fragments obtained, a 391 bp LEAP-1-specifc cDNA sequence exactly matching the three exon regions of the corresponding gene was fnally determined. Approximately 40 bp upstream of the 5′-terminal nucleotide of the cDNA sequence, a potential TATA box motif (ATAAAA) was identifed within the above-mentioned contig. Furthermore, a typical AATAAA polyadenylation signal is located approximately 28 bp upstream of the 3′-terminus of the cDNA. Both fndings indicate the cloning of an almost full-length cDNA specifc for LEAP-1.

The cDNA sequence exhibits an open reading frame encoding an 84-residue precursor protein which contains the isolated peptide at the COOH-terminus (Fig. 2). An NH2-terminal signal sequence (residues 1-24) characteristic for secretory proteins was identifed using prediction methods [17]. In addition, a basic penta-arginyl segment directly adjacent to the NH2-terminus of the isolated peptide is contained within the precursor protein. It appears that the LEAP-1 precursor is a secretory protein that is proteolytically processed at the poly-basic proteolysis site whereby the isolated peptide LEAP-1 is released. A further processing of LEAP-1 seems to be possible as an NH2-terminally truncated form (sequence positions 6584) was also identifed in human blood ultrafltrate by mass spectrometry. With its eight cysteine residues forming four disulfde bonds along a sequence of 17 amino acids, LEAP1 is an exceptionally structured molecule. Interestingly, data base comparisons using the entire amino acid sequence derived from cDNA translation resulted in an entry (SwissProt, P81172) describing the partial amino acid sequence 50-84 of the LEAP-1 precursor. However, this entry does not provide information regarding the occurrence of LEAP-1 in plasma and the cysteine connectivity, gene expression and the antimicrobial activity. Furthermore, tBLASTn searches in the Gen-Bank data base led to the identifcation of several mouse EST clone sequence entries (e.g. W41260, W36480, W34077) related to LEAP-1. The corresponding murine cDNA encodes a putative 84-residue protein which shows a sequence identity of approximately 51% with human LEAP-1, and exhibits an identical cysteine pattern at its COOH-terminus. Thus, it might represent the murine ortholog of LEAP-1.

Using the synthetic peptide, we examined the antimicrobial efects of LEAP-1 by performing an agar difusion assay [11]. Here, mainly Gram-positive bacteria exhibited susceptibility to LEAP-1 treatment (Fig. 3). Nevertheless, LEAP-1 also inhibited the growth of Gram-negative Neisseria cinerea and the yeast Saccharomyces cerevisiae, thus resembling the antimicrobial spectrum of human �-defensin-1 [18]. In contrast, no inhibition of the growth of the yeast Rhodotorula rubra and Gram-negative germs Escherichia coli BL21 and Pseudomonas fuorescens was detected. In order to determine the in vitro IC50 values of LEAP-1 against two sensitive strains, we performed CFU assays. These experiments revealed half-maximal inhibitory activity at approximately 40 Jg/ml (14.4 JM) against B. subtilis and 50 Jg/ml (18 JM) against S. cerevisiae. These efective peptide concentrations are in the same order of magnitude as known for other antimicrobial peptides [19-22].

Fig. 3. Dose-dependent antibacterial activity of LEAP-1 using the radial difusion assay. (A) Bacillus megaterium ATCC14581, (B) Bacillus subtilis ATCC6051, (C) Micrococcus luteus ATCC9341,

(D) Staphylococcus carnosus DSM20501, (E) Neisseria cinerea ATCC14685, (F) Saccharomyces cerevisiae ATCC9763. 1 inhibition unit (U) corresponds to 0.1 mm radius of growth inhibitory area. The results of two independent experiments with casocidin-I at 11 Jg as a positive control are shown. Casocidin-I has been shown to cause antimicrobial efects comparable to those of magainin II [18].

Fig. 4. Expression of the LEAP-1 gene by real-time quantitative RT-PCR. The indicated x-fold expression of LEAP-1 in diferent tissues was determined by standardization to the expression in the kidney, using a two-step calculation method. First, averaged LEAP1 expression values of each tissue were divided by the values for the GAPDH housekeeping gene. Then, the normalized target gene values obtained were divided by the normalized calibrator value. Tissue from kidney with the lowest LEAP-1 expression level served as calibrator (i.e. kidney exhibits a 1-fold expression). For all tissues three experiments were carried out in duplicate.

Increased peptide concentrations due to local stimulation of expression at infammatory sites or putative synergism between diferent antimicrobial components of suboptimal concentrations contribute to the fact that it is difcult to conclude in vivo efects from in vitro assays. Moreover, it might be possible for an antimicrobial peptide even in sublethal concentrations to be active against pathogens by receptor-mediated interactions or other mechanisms [20,23,24].

In order to analyze whether the expression pattern of LEAP-1 is similar to that of other human antibiotic peptides, we performed quantitative real-time RT-PCR assays (Fig. 4). The highest level of LEAP-1 gene expression was detected in the liver where it was found in a range comparable to that of the GAPDH housekeeping gene. Expression of LEAP-1 in heart and brain proved to be at least two times higher than in the other tissues tested. These organs have not been reported so far to harbor signifcant amounts of antimicrobial peptides. Lung, prostate gland, tonsils, salivary gland and trachea express LEAP-1 in a similar but low level. These results were confrmed using the superoxide dismutase gene as housekeeping gene for quantitative real-time RT-PCR and indicate a rather unusual expression pattern for an antimicrobially active peptide. In particular, the prevailing expression in liver gives rise to further investigations concerning the role of LEAP-1 in liver-specifc functions. Thus, it would be important to evaluate whether LEAP-1 is involved in the ac-

A. Krause et al./FEBS Letters 480 (2000) 147-150

tivation of liver-derived acute-phase proteins. However, several fndings suggest that antimicrobial peptides may have further functions other than the direct inhibition of microbes at mucosal barriers [23-25].

In conclusion, we identifed a novel peptide that circulates in human blood plasma and exhibits a unique motif of cysteine connectivity. LEAP-1 extends the known families of endogenous peptides with growth inhibitory efects on microbial organisms. Future investigations concerning the role of LEAP-1 should address the extraordinary level of expression in the liver, and possible synergistic efects with other components of the immune system.

Acknowledgements: This work was supported in part by the Deutsche Forschungsgemeinschaft (grant to A.K., IIIGK-GRK99/2-98).

References

[1] Andreu, D. and Rivas, L. (1998) Biopolymers (Pept. Sci.) 47, 415-433.

[2] Lehrer, R.I. and Ganz, T. (1999) Curr. Opin. Immunol. 11, 23

27.

[3] Schroider, J.M. (1999) Biochem. Pharmacol. 57, 121-134.

[4] Shi, J. and Ganz, T. (1998) Infect. Immun. 66, 3611-3617.

[5] Bals, R., Weiner, D.J., Moscioni, A.D., Meegalla, R.L. and Wilson, J.M. (1999) Infect. Immun. 67, 6084-6089.

[6] Shai, Y. (1995) Trends Biochem. Sci. 20, 460-464.

[7] Hancock, R.E.W. (1997) Lancet 349, 418-422.

[8] Ganz, T. and Lehrer, R.I. (1995) Pharmacol. Ther. 66, 191-205.

[9] Bensch, K.W., Raida, M., Maigert, H.-J., Schulz-Knappe, P. and Forssmann, W.-G. (1995) FEBS Lett. 368, 331-335.

[10] Schulz-Knappe, P., Schrader, M., Staindker, L., Richter, R., Hess, R., Juirgens, M. and Forssmann, W.-G. (1997) J. Chromatogr. A 776, 125-132.

[11] Lehrer, R.I., Rosenman, M., Harwig, S.S.S.L., Jackson, R. and Eisenhauer, P. (1991) J. Immunol. Methods 137, 167-173.

[12] Zucht, H.D., Raida, M., Adermann, K., Maigert, H.-J. and Forssmann, W.-G. (1995) FEBS Lett. 372, 185-188.

[13] Pardigol, A., Forssmann, U., Zucht, H.D., Loetscher, P., Schulz-Knappe, P., Baggiolini, M., Forssmann, W.-G. and Maigert, H.-J. (1998) Proc. Natl. Acad. Sci. USA 95, 6308-6313.

[14] Maigert, H.-J., Reinecke, M., David, I., Raab, H.R., Adermann, K., Zucht, H.D., Hill, O., Hess, R. and Forssmann, W.-G. (1998) Regul. Pept. 73, 165-176.

[15] Maigert, H.-J., Cieslak, A., Alkan, O., Luischer, B., Kaufels, W. and Forssmann, W.-G. (1999) J. Biol. Chem. 274, 444-450.

[16] Bieeche, I., Olivi, M., Champeeme, M.-H., Vidaud, D., Lidereau,

R. and Vidaud, M. (1998) Int. J. Cancer 78, 661-666.

[17] Nielsen, H., Engelbrecht, J., Brunak, S. and von Heijne, G. (1997) Protein Eng. 10, 1-6.

[18] Zucht, H.D., Grabowsky, J., Schrader, M., Liepke, C., Juirgens, M., Schulz-Knappe, P. and Forssmann, W.-G. (1998) Eur. J. Med. Res. 3, 315-323.

[19] Ganz, T. (1999) Science 286, 420-421.

[20] Chertov, O., Michiel, D.F., Xu, L., Wang, J.M., Tani, K., Murphy, W.J., Longo, D.L., Taub, D.D. and Oppenheim, J.J. (1996)

J. Biol. Chem. 271, 2935-2940.

[21] Mor, A., Nguyen, V.H., Delfour, A., Migliore-Samour, D. and Nicolas, P. (1991) Biochemistry 30, 8824-8830.

[22] Lowenberger, C., Charlet, M., Vizioli, J., Kamal, S., Richman, A., Christensen, B.M. and Bulet, P. (1999) J. Biol. Chem. 274, 20092-20097.

[23] Yang, D., Chertov, O., Bykovskaia, S.N., Chen, Q., Bufo, M.J., Shogan, J., Anderson, M., Schroider, J.M., Wang, J.M., Howard,

O.M.Z. and Oppenheim, J.J. (1999) Science 286, 525-528.

[24] Solomon, S., Hu, J., Zhu, Q., Belcourt, D., Bennett, H.P., Bate-man, A. and Antakly, T. (1991) J. Steroid Biochem. Mol. Biol. 40, 391-398.

[25] Shi, J., Ross, C.R., Leto, T.L. and Blecha, F. (1996) Proc. Natl. Acad. Sci. USA 93, 6014-6018.