Antimicrobial activity of human β-defensin 4 analogs: Insights into the role of disulfide linkages in modulating activity

Peptides 38 (2012) 255–265

Contents lists available at SciVerse ScienceDirect

Peptides journal homepage: www.elsevier.com/locate/peptides

Antimicrobial activity of human ␤-defensin 4 analogs: Insights into the role of disulfide linkages in modulating activity Himanshu Sharma, Ramakrishnan Nagaraj ∗ CSIR – Centre for Cellular and Molecular Biology, Uppal Road, Hyderabad 500 007, India

a r t i c l e

i n f o

Article history: Received 19 June 2012 Received in revised form 21 August 2012 Accepted 21 August 2012 Available online 19 September 2012 Keywords: Antimicrobial activity Disulfide bonds Human ␤-defensins Membrane permeabilization Salt sensitivity

a b s t r a c t Human ␤-defensins (HBDs) are cationic antimicrobial peptides that are components of the innate immune system. They are characterized by three disulfide bridges. However, the number of cationic residues as well as the presence of lysine and arginine residues vary. In HBD4, the cationic residues occur predominantly in the N-terminal segment, unlike in HBD1–3. We have examined the antimicrobial activity of peptides spanning the N- and C-terminal segments of HBD4. We have introduced one, two and three disulfide bridges in the peptides corresponding to the N-terminal segments. Peptides corresponding to the N-terminal segment had identical sequences and variation was only in the number and spacing of cysteines and disulfide bridges. Antimicrobial activity to varying extents was observed for all the peptides. When two disulfide bridges were present, decrease in antimicrobial potency as well as sensitivity of activity to salt was observed. Enhanced antimicrobial activity was observed when three disulfide bridges were present. The antimicrobial potency was similar to HBD4 except against Escherichia coli and was attenuated in the presence of salt. While the presence of three disulfide bridges did not constrain the peptide to a rigid ␤-sheet, the activity was considerably more as compared to the peptides with one or two disulfide bridges. The peptides enter bacterial and fungal cells rapidly without membrane permeabilization and appear to exert their activity inside the cells rather than at the membrane. © 2012 Elsevier Inc. All rights reserved.

1. Introduction The cationic antimicrobial peptides defensins are important components of innate immunity [13,14,30,46]. Defensins have been the subject of extensive investigations with a view to understand how these peptides exert their antimicrobial activity [2,5,11,15,17,36]. They are classified into ␣, ␤ and ␪ defensins on the basis of disulfide connectivity. Human ␤-defensins (HBDs) are expressed predominantly in mucosal and epithelial surfaces [13,24,36]. Disulfide connectivities in ␤-defensins are between cysteines 1, 5; 2, 4 and 3, 6 [13,31,36,42]. Structural studies on HBD1–3 have indicated that the ␤-strands occurring at the middle and Cterminal regions are conserved [3,19,20,26,40,41]. Bactericidal and fungicidal activities of these peptides vary considerably and are attenuated at high ionic strength [36]. HBD1 and 2 kill Gram-negative bacteria and fungus more efficiently as compared to Gram-positive bacteria at low ionic strength. HBD3 is more potent against Gram-positive and Gram-negative bacteria as compared to HBD1, 2 and the activity is not attenuated at high salt concentrations [2,16,17,28,51]. HBD4 exhibits broad spectrum

∗ Corresponding author. Tel.: +91 40 27192589; fax: +91 40 27160591. E-mail address: [email protected] (R. Nagaraj). 0196-9781/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.peptides.2012.08.024

antimicrobial activity but only at low ionic strength [15]. The antimicrobial activities of smaller fragments of ␤-defensins, with or without disulfide constraints, have been investigated for a better understanding of structure–activity relationships [21,27–29,44]. Recently, it has been demonstrated that HBD analogs composed of different regions of HBD1 and 3 possess enhanced antibacterial activity at high salt concentrations and internal regions of HBD1 and C-terminal region of HBD3 are essential for tolerance to salt [44]. Chimeras of HBD2 and 3 have been recently shown to have potent activity that is not attenuated in the presence of high salt concentrations [23]. Linear analogs of HBD3 also show bactericidal activity comparable to the native peptide [8,58]. Reduced HBD1 shows potent antimicrobial activity against Candida albicans and anaerobic Gram-positive species Bifidobacterium and Lactobacillus [43]. We have previously shown that C-terminal analogs of HBD1–3 with a single disulfide bridge kill bacteria and fungi [28,29]. Although the cysteine connectivity in HBD4 is identical to other ␤-defensins, the distribution of cationic residues in the primary sequence is different. Cationic residues in HBD1–3 occur predominantly at the C-terminal end and after the third cysteine residue [13,36,49], while in HBD4, 6 of the 10 cationic residues are localized between residues 1 and 18 [15]. Also, 5 of the 6 cationic residues in this N-terminal segment occur sequentially, interspaced by cysteine residues as RCRKKCR. Therefore, it would be of interest to delineate the determinants of antimicrobial activity in HBD4.

256

H. Sharma, R. Nagaraj / Peptides 38 (2012) 255–265

Table 1 Primary structures of HBD1–4 and net charge at neutral pH. Peptide

Sequencea

HBD1 HBD2 HBD3 HBD4 N1 N2 N3 N4 N5 C1

DHYNC1 VSSGGQC2 LYSAC3 PIFTKIQGTC2 YRGKAKC1 C3 K GIGDPVTC1 LKSGAIC2 HPVFC3 PRRYKQIGTC2 GLPGTKC1 C3 KKP GIINTLQKYYC1 RVRGGRC2 AVLSC3 LPKEEQIGKC2 STRGRKC1 C3 RRKK ELDRIC1 GYGTARC2 RKKC3 RSQEYRIGRC2 PNTYAC1 C3 LRK ELDRIC1 GYGTARRKKC1 R ELDRIGYGTARC1 RKKC1 R ELDRIC1 GYGTARC1 RKKR ELC1 DRIC2 GYGTARC1 RKKC2 R ELC1 C2 DRIGC3 YTARC2 RKKRSC1 C3 L KRSQEYRIGRC1 PNTYAC1 LKR

a

Net charge +4 +6 +11 +7 +4 +4 +4 +4 +4 +5

Disulfide connectivities are shown by superscript numbers adjacent to cysteines. Cationic residues are underlined.

Synthetic peptides were designed spanning N and C-terminal regions of the sequence with disulfide bridges (Table 1). We have observed that peptides spanning the N- and C-terminal regions of HBD4 exhibited antimicrobial activity. The presence of three disulfide bridges in the N-terminal segment resulted in enhanced antimicrobial activity against E. coli, Pseudomonas aeruginosa, Staphylococcus aureus and C. albicans. A peptide corresponding to the C-terminal segment was active against only E. coli and C. albicans. The peptides appear to kill bacteria and fungi not by permeabilizing membranes but by rapidly crossing the membrane barrier and exerting their activity inside the cells. 2. Materials and methods 2.1. Reagents 9-Fluorenylmethoxy carbonyl (F-moc) amino acids were obtained from Novabiochem AG (Switzerland) and Advanced Chemtech (Louisville, KY). N-Fmoc-N -(4-methoxy-2,3,6trimethylbenzenesulfonyl)-l-arginine polyethyleneglycol-polystyrene (F-moc-l-Arg(Mtr)-PEG-PS) and polyethyleneglycol –polystyrene (PEG–PS) were purchased from Millipore (Bedford, MA). N-hydroxybenzotriazole hydrate (HOBt) and 2-(1Hbenzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU) were from Advanced Chemtech (Louisville, KY). Phospholipids PC (1-palmitoyl-2-oleoyl-sn-glycero-3phosphocholine), PE (1-palmitoyl-2-oleoyl-sn-glycero-3-phosand PG (1-palmitoyl-2-oleoyl-snphoethanolamine), glycero-3-phospho-(1 -rac-glycerol)) (sodium salt), were purchased from Avanti Polar Lipids (Alabaster, AL). CF (carboxyfluorescein) was obtained from Sigma–Aldrich, India. N-(3-triethylammoniumpropyl)-4-(6-(4(diethylamino) phenylhexatrienyl) pyridinium dibromide) (FM4-64), and propidium iodide (PI), were obtained from Molecular Probes. HBD4 (Code: 4406-s, Lot No. 590707) was purchased from Peptide Institute Inc. (Osaka). All other chemicals used were of the highest grade available. 2.2. Peptide synthesis HBD4 analogs N1–N5 and C1 (Table 1) were synthesized using Fmoc chemistry [1]. F-moc-l-Arg(Mtr)-PEG-PS (0.13 mmol/g) and PEG–PS resin (0.19 mmol/g) were used as solid support. Peptides were cleaved from the resin using a mixture containing 80% trifluoroacetic acid (TFA), 8% m-cresol, 8% thioanisole, and 4% ethanedithiol for 12–15 h at room temperature. Peptides were precipitated on ice-cold diethyl ether. Disulfide bond formation was accomplished in peptides having two cysteine residues by carrying out oxidation at a concentration of 0.4 mg/ml in 20% dimethylsulfoxide (DMSO) for 12–14 h [48]. Purification of peptides was carried out on a Hewlett Packard 1100 series HPLC instrument using a

reverse-phase Agilent 300SB-C18 Zorbax column with a water (solvent A) and acetonitrile (solvent B) containing 0.1% TFA. Peptides were eluted using a linear gradient from 5% to 100% of solvent B over 60 min at a flow rate of 0.5 ml/min. Two and three disulfide bridged peptides N4 and N5 were synthesized using orthogonal protecting group for side chain of cysteines. In the two disulfide containing peptide, N4, 1st and 3rd cysteines were protected with acetoamidomethyl (Acm) group and 2nd and 4th cysteines were protected with trityl group (Trt). In the three disulfide containing peptide, N5, 1st and 5th cysteines were protected with Acm, 2nd and 4th cysteines were protected with tertiary butyl (t-Bu) and 3rd and 6th cysteines were protected with Trt group which was removed during the acidic cleavage of peptide from resin. The first disulfide bond was formed in 20% DMSO [48]. Peptide was purified by reverse phase chromatography and was subjected for second disulfide formation by deprotecting Acm group. It was removed by iodine oxidation as mentioned elsewhere [27]. In brief, peptide was dissolved in acetic acid:water (4:1) at the concentration of 50–100 ␮g/ml and 20 equivalents of iodine were added into peptide. Solution was kept for 6–8 h at 37 ◦ C. Reaction was stopped by adding two times volume of deionized water. Iodine was removed from the mixture by washing with excess carbon tetrachloride until solution was colorless. The peptide was then purified by HPLC. The 3rd disulfide was formed by removing the t-Bu group by the method described by Kluver et al. [27] with few modifications. The peptide was dissolved in TFA (50 ␮g/ml). Reaction mix was formed by adding 500 equivalent of DMSO and 100 equivalent of anisole to the peptide solution. Mix was stirred overnight as it was found that incomplete oxidation occurs in 2 h of incubation. The peptide was then purified by reverse phase chromatography using linear gradient of 5–100% solvent B (acetonitrile containing 0.1% TFA, v/v) in 60 min on reverse phase C-18 column (Agilent 300 SB C-18 Zorbax). All the peptides were characterized by mass spectrometry on an AB4800 matrix assisted laser desorption ionization-time of flight/time of flight (MALDI TOF/TOF) mass spectrometer from Applied Biosystems (PerSeptive Biosystems, Foster City, CA). Peptides with two or three disulfide linkage were purified after each oxidation step and then were characterized by MALDI TOF/TOF before keeping for next oxidation step. For CF labeling, peptide attached to the resin was incubated with a mixture containing CF and activating agent (HOBt and HBTU) in DMF [56] and labeled peptides were cleaved from resin as described above. All the purified peptides were dried and dissolved in deionized water. The concentrations of peptides were determined using a molar absorption coefficient of 1280 M−1 cm−1 at 280 nm and for CF labeled peptide at 492 nm using coefficient of 65000 M−1 cm−1 . The concentration was also cross-checked by dissolving known weight of dried peptide in deionized water.

H. Sharma, R. Nagaraj / Peptides 38 (2012) 255–265

257

2.3. Antimicrobial assay

2.5. Circular dichroism (CD)

The bacterial strains used were E. coli (MG1655), S. aureus (NCTC 8530), P. aeruginosa (NCTC 6750) and antifungal activity was performed using C. albicans (ATCC 18804). Bacteria were grown overnight in nutrient broth (Bacto; Difco) at 37 ◦ C. 1% inoculum of overnight-grown bacteria was subcultured in nutrient broth for 2 h at 37 ◦ C. Bacterial cells were harvested at mid-log phase and washed with 10 mM sodium phosphate buffer (pH 7.4) and were resuspended in same buffer to obtain cell density of 106 CFUs/ml (colony forming units/ml). 100 ␮l of cell suspension was taken in 96-well plate and incubated with varying concentrations of peptides at 37 ◦ C for 2 h. Cells were then spread on nutrient agar and incubated at 37 ◦ C for 14–18 h. Colonies formed were counted and the minimum concentration of peptide at which 99.9% killing observed was taken as LC (lethal concentration). MFC (minimum fungicidal concentration) was determined exactly as described for bacteria. However, fungal cells were grown in yeast extractpeptone-dextrose (YEPD) (Difco) medium at 30 ◦ C for 24–30 h. Peptide activity was also checked in the presence of reducing agent DTT (dithiothreitol). Peptides were incubated with bacterial cells at varying concentrations in phosphate buffer containing 10 mM DTT. Percentage of killing was determined by counting the colonies with respect to colonies formed by untreated cells which were also kept in the same buffer containing DTT. Salt sensitivity was examined by pre-incubating the cells with varying concentration of NaCl (30–150 mM), CaCl2 and MgCl2 (250 ␮M–2 mM) followed by addition of peptides. Kinetics of killing was determined as follows: cells at a density of 106 CFUs/ml were incubated with lethal concentrations of peptides in 100 ␮l of 10 mM sodium phosphate buffer (pH 7.4) at 37 ◦ C. Cells were spread on nutrient agar at fixed time intervals during 2 h of incubation. Cells without treatment with peptide, was taken as a control. Plates were incubated for 14–18 h at 37 ◦ C and colonies formed were counted. Fungal cells were grown in yeast extractpeptone-dextrose (YEPD) (Difco) medium at 30 ◦ C for 24–30 h. All experiments were carried out thrice independently in duplicates. Values obtained in each experiment were within variation of 5%. Data shown here represents the average value of repeated experiments.

CD spectra were recorded on a Jasco J-715 spectropolarimeter. Peptides were dissolved in 10 mM phosphate buffer (pH 7.4), trifluoroethanol (TFE) or lipid vesicles (LUVs, large unilamellar vesicles) of PE–PG (7:3) (peptide–lipid molar ratio of 1:10) at a concentration of 25 ␮M just before measurement. All the spectra were recorded in 0.1 cm path length cell using a step size of 0.2 nm, band width of 1 nm and scan rate of 100 nm/min. The spectra were recorded by averaging 8 scans and corrected by subtracting the solvent/buffer spectra.

2.4. Preparation of lipid vesicles Desired amount of PC, PE and PG lipids was taken from their stock solutions made in chloroform/methanol to form PC and PE–PG (7:3) vesicles. Fluorescent labeled vesicles were made by adding 1 mol% rhodamine–PE (Rh–PE). Vesicles were prepared by drying lipids using nitrogen stream and then were kept for 5–6 h in vacuum to remove traces of organic solvents. For preparation of large unilamellar vesicles (LUVs), desiccated lipid film was hydrated in 10 mM phosphate buffer (pH 7.4) and kept at 4 ◦ C for overnight. Multilamellar vesicles were formed by vortexing lipid suspensions for several minutes until whole lipid film was removed from glass vial. Multilamellar vesicles were passed several times through polycarbonate membranes of 200 nm pore size using mini-extruder (Avanti Polar Lipid Inc.) [32]. Homogeneous size distribution around 200 nm was confirmed by dynamic light scattering analysis, using a Photocor Complex-dynamic light scattering (DLS) instrument (Photocor Instruments, MD). A laser of wavelength 632.8 nm was used for collecting the data. The data were processed using DynaLS software (V.2.8.3). Giant unilamellar vesicles (GUVs) labeled with Rh–PE were prepared using swelling method [9,22]. The desiccated lipid film was hydrated with 10 mM phosphate buffer, pH 7.4 at 45 ◦ C for 15 min followed by 10 h incubation at 60 ◦ C to get vesicles.

2.6. Microscopy Cellular localization of HBD4 analogs was studied using confocal microscopy. E. coli and P. aeruginosa were stained with lipophilic dye FM4-64 which preferentially stains the inner membrane [12]. Dye concentration was kept 3 ␮M in the bacterial suspension. To reduce the background due to the dye in solution, cells were washed with 10 mM phosphate buffer (pH 7.4) and resuspended in same buffer to get 107 CFUs/ml. Cells were then treated with CF labeled peptides CF-N2, -N4 and -N5 and then transferred to chamber slide and observed on a Leica TCS-SP5 ultraspectral confocal microscope (Leica Microsystems). Images with Z-sections of 0.25 ␮m thickness were taken after 10 min of incubation using 100× oil immersion objectives. Excitation wavelengths 488 nm and 543 nm were used for CF-labeled peptides and FM4-64 respectively. Emission bands in the range of 494–568 nm and 600–740 nm were used for detection of CF labeled peptide and FM4-64 respectively. In the case of C. albicans, propidium iodide (PI) was used which stains preferentially nucleus of nonviable cells. Excitation wavelength 543 nm and emission band in the range of 573–686 nm were used for PI. Images with Z-sections of 0.45 ␮m thickness were captured after 20 min of incubation. Bright-field images were also captured simultaneously using transmitted light detector. Binding of labeled peptides CF-N2, CF-N4, and CF-N5 with GUVs and MLVs was examined using confocal fluorescence scanning microscopy. GUVs and MLVs were treated with peptides (2.5–40 ␮M) and after treatment; vesicles were left for 30 min to settle on slide. Images were captured using 63× oil immersion objective on a Leica TCS-SP5 ultraspectral confocal microscope (Leica Microsystems). Excitation wavelength 543 nm and emission band in the range of 590–712 nm were used for Rh conjugated lipid vesicles. 2.7. Isothermal titration calorimetry Isothermal titration calorimetry was performed using a Microcal VP-ITC calorimeter (Microcal, Norhampton, MA). All solutions were prepared in 10 mM phosphate buffer of pH 7.4 and were degassed before the experiment. Peptides N1 and N2 at 50 ␮M concentration were titrated with 10 mM PC or PE–PG (7:3) LUVs at 25 ◦ C. Each time 4 ␮l of LUVs was injected in 8 s into the calorimeteric cell of volume 1.4 ml containing the peptide and stirred at 300 rpm for 300 s. Heat of dilutions was determined by injecting LUVs into buffer and was subtracted from heat determined due to lipid–peptide titration. Enthalpy change, H◦ , was calculated after ith injection at which heat change (corrected for heat of dilution) approaches zero as all peptide at this point almost bound to the lipid vesicles, using formula [45]: H ◦ =

 i

ıhi 0 V Cpep cell

(1)

where ıhi is heat change (corrected for heat of dilution) during each injection of lipid vesicles into the calorimeter cell of volume, Vcell

258

H. Sharma, R. Nagaraj / Peptides 38 (2012) 255–265

0 is the total peptide concentracontaining the peptide solution. Cpep tion in the cell.  Cumulative heat change of all injections, H, was

ıhi .

determined by

The hemolytic activity of peptides was also checked against rat erythrocytes. HBD4 as well short peptides did not lyse erythrocytes significantly even at 10 times lethal LC or MFC (Fig. S1).

i

3.3. Kinetics of killing 3. Results 3.1. Design of HBD4 analogs HBD4 has six cationic residues in the region of the first three cysteine residues, unlike HBD1–3. Peptides spanning the Cterminal cationic segments of HBD1–3, with a single disulfide bond exhibit antimicrobial activity [28,29]. Here, we have examined the antimicrobial activity of peptides spanning the N-terminal cationic segment comprising residues 1–18 and C-terminal residues 18–37 of HBD4. The sequences of HBD1–4 and peptides spanning the N- and C-terminal segment are shown in Table 1. The peptides have one, two and three disulfide bridges. Since the peptides correspond to the N-terminal and C-terminal segments of HBD4, it is not possible to introduce disulfide bonds as observed in full length HBD4. The spacing between the cysteines in the single disulfide bridged peptide N1–3 is such that the extent of constraint in the vicinity of the positively charged residues would vary. Peptide C1 corresponds to C-terminal of HBD4 sequence and has a single disulfide bond. In N3 and C1, the number of residues that occur between the cysteines were identical to the pattern between cysteines 1, 2 and 4, 5 respectively in HBD4. Spacing between cysteines in N4 and N5 were similar to conotoxins [50]. Specific disulfide connectivity in N4 and N5 was introduced using orthogonal protecting group on cysteine side chains [27]. Peptides were purified by HPLC and characterized by mass spectrometry after each step of disulfide formation (Table 2). Peptides N1–5, C1 and HBD4 eluted at 22.8, 22.7, 22.9, 24.3, 27.0, 23.0, and 26.4 min respectively, on reverse phase HPLC C-18 column. Retention times of the two and three disulfide bridged peptides N4 and N5 were closest to HBD4.

3.2. Antimicrobial activity The antimicrobial activity of the peptides is summarized in Table 3. All the disulfide bridged peptides show antimicrobial activity. While N1 and N2, having one disulfide bridge, show comparable activity, N3 where the cationic segment RKKR is outside the cyclic segment shows preferential activity against S. aureus and C. albicans. The potency toward E. coli and P. aeruginosa is considerably lower as compared to N1 and N2. Lethal concentration for single disulfide constrained peptide corresponding to C-terminus of HBD4 was comparable to the other single disulfide linked peptides in case of E. coli and C. albicans. However, the peptide did not exhibit activity against P. aeruginosa and S. aureus. When two disulfide bridges are present (N4), activity is observed against Gram-negative bacteria and C. albicans but not against S. aureus. When three disulfides are present, enhanced antimicrobial activity is observed. Antimicrobial activity of the disulfide bridged peptides was checked in presence of reducing agent DTT. The peptides were inactive at their respective LC in the presence of DTT. The antimicrobial spectrum of commercial HBD4 is also shown in Table 3. Activity is observed against E. coli, P. aeruginosa, S. aureus and C. albicans. Greater activity was observed against E. coli and C. albicans as compared to P. aeruginosa and S. aureus. HBD4 and N5 show comparable activity against P. aeruginosa and C. albicans. N5 exhibited greater against S. aureus but lower activity against E. coli as compared to HBD4.

The kinetics of microbial killing by peptides N1–N5 and HBD4 are shown in Fig. 1. The rate of killing in case of the analogs depends on the disulfide connectivity. Peptides N1 and N2 kill E. coli more rapidly as compared to N4 and N5. However, N3 kills S. aureus more rapidly as compared to the other peptides including HBD4. While N1 and N2 kill C. albicans relatively rapidly, the rate of killing is slower for N3, N4 and N5. It appears that when three disulfide bridges are present, the rate of killing is slower. Rate of killing for HBD4 is slow against S. aureus and C. albicans but comparable to N1 and N2 against Gram-negative bacteria. 3.4. Salt sensitivity The antimicrobial activity of ␣ and ␤-defensins are attenuated in the presence of high salt [2,17,28,33,36,42,51,53]. The effect of physiological extracellular concentrations of Na+ , Ca2+ , and Mg2+ on the antimicrobial activity of the analogs and HBD4 were examined and the data are shown in Fig. 2. Only marginal inhibition of activity is observed for N1 and N2 against E. coli and P. aeruginosa in the presence of Na+ , whereas Ca2+ and Mg2+ completely inhibit activity. The activities of N4 and N5 are attenuated considerably in the presence of Na+ , Ca2+ and Mg2+ . While the activity of HBD4 is diminished in the presence of Na+ and Mg2+ , the effect of Ca2+ on activity is marginal unlike for N1–5. Against S. aureus, Na+ completely inhibits the activity of N1 and N2 but not N3. The activities of N5 and HBD4 were also inhibited significantly. The divalent cations have no effect on the activity or show very marginal reduction in activity, except for HBD4 in the presence of Ca2+ . Attenuation or loss in activity of N1–5 is observed in the presence of Na+ against C. albicans. Divalent cations do not inhibit activity of the peptides against C. albicans except for N4. The activity of HBD4 against C. albicans is lost in the presence of salts. Although the sequences of the peptides with different disulfide bridges are identical except for the presence of cysteines, there are considerable differences in the inhibition of activity by Na+ and divalent cations. It appears that the disulfide connectivity patterns play an important role in determining salt sensitivity. 3.5. Circular dichroism Conformational changes with introduction of one, two and three disulfide bridges in shorter segment were examined by CD spectroscopy. Spectra of the peptides in buffer are shown in Fig. 3. All the peptides show a minimum in the region of 200 nm with crossover ∼190 nm. The spectra suggest that while a small fraction may populate ␤-turn conformation, the peptides exhibit considerable conformational flexibility [57]. Even, HBD4, which is presumed to have a small helical segment in the N-terminal region and three ␤strands shows spectra which suggests considerable conformational flexibility. The spectra in lipid vesicles (Fig. 4) are very similar indicating that they do not modulate the conformation as observed with linear cationic antimicrobial peptides [38,47]. The spectra in TFE (Fig. 5) suggest the tendency of the peptides to fold into ␤-turn conformation [57]. 3.6. Localization The interaction of peptides with bacteria and fungi was investigated by labeling the peptides at the N-termini with CF. Labeling

H. Sharma, R. Nagaraj / Peptides 38 (2012) 255–265

259

Fig. 1. Kinetics of cell death. Peptides were incubated with indicated bacteria for different time points. At each time point, peptide treated cells were plated and bacterial and fungal colonies were counted after 14–18 h and 24–30 h respectively. Log10 CFUs/ml was determined and plotted against time. () N1; () N2; () N3; (夽) N4; () N5; and (♦) HBD4.

Fig. 2. Effect of salts on antimicrobial activity. Bactericidal and fungicidal activity of HBD4 and analogs were determined in sodium phosphate buffer (10 mM, pH 7.4) supplemented with varying concentrations of monovalent (30–150 mM) and divalent ions (0.25–2 mM). Values here are indicated for 150 mM and 2 mM monovalent and divalent ions respectively. The symbol  represents complete loss in activity. () No salt; (

) 150 mM NaCl; (

) Ca2+ ; and (

) Mg2+ .

260

H. Sharma, R. Nagaraj / Peptides 38 (2012) 255–265

Table 2 Peptides with two disulfide bonds, N4 and three disulfide bonds, N5 were synthesized using orthogonal protection of cysteine side chains. At every step of deprotection and oxidation, peptides were characterized by HPLC and mass spectrometry. Peptide

N4

HPLC retention time (min)

25.7 24.3 29.2 29.6 27.0

N5

Trt removal and DMSO oxidation

Acm removal and I2 oxidation

t-Bu removal and anisole/DMSO/TFA oxidation

Theoretical mass

Observed mass

Theoretical mass

Observed mass

Theoretical mass

Observed mass

2369.8

2370.72 2227.6

2227.26

2831.5

2831.98 2689.3

2688.12 2575.1

2574.59

Table 3 Antimicrobial activity of HBD4 and its analogs. Values (in ␮M) are LC of peptides against bacteria and MFC against C. albicans.

a

E. coli P. aeruginosaa S. aureus C. albicans a b

HBD4

N1

N2

N3

5 12 15 7.5

20 20 20 40

20 20 20 40

100 (60%) 100 (60%) 20 40

N4b

N5

C1b

30 20 IA 30

10 10 7.5 7.5

20 IA IA 40

Values in parentheses denote percentage of killing. IA represents peptide is inactive against corresponding organism.

was carried out on-resin to ensure selective labeling at the Nterminus. Labeled peptides at a concentration of 5 ␮M were incubated with bacteria and yeast for 10 and 20 min respectively and after that images were captured. Fig. 6A shows images of E. coli in the presence of peptides. The FM4-64 labeled panels shows no distinctive breaks. The peptides appear to localize rapidly inside the bacteria (panels labeled CF). Similar observations were made with P. aeruginosa (data not shown). All the disulfide bridged peptides follow a similar mode of entry into the bacteria. Analysis of fluorescence intensities indicates CF-N5 is localized inside the cell as shown by the fluorescent intensity with respect to the line drawn across the bacterial morphology (Fig. 6B). Intensity due to peptide starts rising after the fluorescence due to FM4-64 and reaches its maximum where FM4-64 fluorescence is least. Rapid localization into the cytosol is also observed in the case of C. albicans (Fig. 7A). Membrane integrity appears to be intact in C. albicans too as with bacteria. The peptide is localized throughout the cytoplasm and uptake of propidium iodide (PI) indicates cell death. In case of untreated cells, PI staining was not observed. Fluorescent intensity

with respect to line drawn across the C. albicans morphology, indicates overlap between PI and CF-N5. It indicates that a population of peptide colocalizes with nucleic acids.

Fig. 3. CD spectra of peptides in buffer. Peptides were dissolved in sodium phosphate buffer (10 mM, pH 7.4) at a concentration of 25 ␮M.

Fig. 4. CD spectra of the peptides in presence of lipid vesicles. Peptides were mixed with lipid vesicles composed of PE:PG (7:3) at peptide–lipid ratio 1:10.

3.7. Interaction with model membranes In order to examine if the peptides associate with model membranes, their interaction with anionic and zwitterionic GUVs were monitored by fluorescence microscopy. The data shown in Fig. 8A shows faint membrane stain with anionic vesicles composed of PE:PG (7:3) while there was no fluorescence with zwitterionic vesicles composed of PC. Fluorescence intensity curve with respect to line drawn across the vesicle, shows peptides do not enter inside the vesicles and fluorescence due to peptide is just partially outside the fluorescence in the membrane (Fig. 8B). Other images of vesicles composed of multibilayers indicate that the peptides do not appear to penetrate the bilayer and localize in the other leaflets. Strength of peptide association with membranes was monitored by titrating peptides N1 and N2 with zwitterionic, PC and anionic,

H. Sharma, R. Nagaraj / Peptides 38 (2012) 255–265

261

Fig. 5. CD spectra of peptides in TFE. Peptides (25 ␮M) were dissolved in TFE just before measurement.

PE–PG (7:3) lipid vesicles using isothermal titration calorimetry. Titration curves show no significant change in heat during the titration of peptides with zwitterionic vesicles while with anionic vesicles heat release was observed as shown in Fig. 9. Molar enthalpy change, H◦ were calculated for N1 and N2 from summation of exothermic heat change (corrected for the heat of dilution) during each injection from Eq. (1). Values were −0.181 kcal/mol for N1 and −0.447 kcal/mol for N2 till 15th and 13th injection respectively as after 13th and 15th injection, decrease in exothermic heat with lipid injections was not very smooth as is evident from the titration curve. Calorimetric studies clearly indicate that the peptides do not interact with zwitterionic membranes while low values of molar enthalpy change for anionic vesicles suggest weak interactions. 4. Discussion Extensive structural investigations on ␤-defensins indicate a common structural motif of a short ␣-helix at the N-terminal region followed by three ␤-strands in a hairpin arrangement [3,19,20,40,41]. Yet, the antimicrobial activity as well as modulation of antimicrobial activity at high concentrations of NaCl varies considerably in human ␤-defensins [2,15–17,36,51]. Although bacterial killing by formation of peptide pores in the membranes was proposed based on the crystal structure of human ␣-defensin HNP3 [18], the structures of ␤-defensins in the crystalline state do not appear to support a common mechanism of membrane damage [7,19,36]. Structure–activity correlations based on the antibacterial activity of truncated analogs with non-native disulfide bridges is an attractive approach to delineate the antimicrobial activity and salt sensitivity of defensins [21,27–29,44]. We have investigated the activities of peptides spanning the N- and C-terminal regions of HBD4 with varying number of disulfide bridges. The connectivities in the two and three disulfide bridged peptides were different from those observed in HBD4. In the single disulfide bridged peptides N1–N3, the cationic segment RKKR was positioned in and outside of the disulfide bridged ring. In C1, all the cationic residues were outside the disulfide bridge segment. In the two and three disulfide bridged peptides, the spacing between adjacent cysteines were similar to conotoxins [50]. Antimicrobial activity was observed for the shorter peptides. Peptides with a single disulfide bridge have antimicrobial activity against Gram-negative and Gram-positive bacteria as well as C. albicans. However, when the segment RKKR

Fig. 6. Localization of HBD4 analogs in E. coli. (A) CF-labeled peptides, CF-N2, CFN4 and CF-N5 (all 5 ␮M) were incubated with 107 CFUs/ml E. coli stained with inner membrane dye FM4-64 (3 ␮M). Images were captured with Leica TCS-SP5 confocal microscope using 100× oil immersion objectives with excitation wavelengths 488 nm and 543 nm and emission bands in the range of 494–568 nm and 600–740 nm for CF-labeled peptides and FM4-64 respectively after 10 min of incubation. (B) Spectral region of interest (ROI)-1 and -2 shows fluorescence intensity with respect to lines drawn across the bacterial morphology in case of CF-N5 (A). Red line in the spectra represents inner membrane of the E. coli and green line represents CF-N5. ROI stands for region of interest shown by arrows.

is outside the disulfide loop as in N3, considerably diminished activity was observed against Gram-negative bacteria but not against Gram-positive bacteria and fungi. In peptide C1, where K and R residues are also outside the disulfide bridged segment, activity was observed only against E. coli and C. albicans. The results indicate that position of K and R in the single disulfide bridged peptides is important for exhibiting specific antibacterial activity but not antifungal activity. This suggests that the location of the cationic stretch in the disulfide loop could be an important determinant of activity. The selective activity of N4, which has two disulfide bridges, against Gram-negative bacteria, indicates that even in a short peptide, spatial arrangement and number of disulfide bridges can play a crucial role in determining selective

262

H. Sharma, R. Nagaraj / Peptides 38 (2012) 255–265

Fig. 7. Localization of HBD4 analogs in C. albicans. (A) Cells at the concentration of 107 CFUs/ml was treated with 5 ␮M of CF-N2, CF-N4 and CF-N5 and then transferred to chamber slide. After 20 min of incubation, images were captured using 100× oil immersion with Z-sections of 0.45 ␮m thickness under confocal microscope. Excitation wavelengths 488 nm and 543 nm and emission bands in the range of 494–568 nm and 573–686 nm for CF-labeled peptides and PI respectively were used and bright field (BF) images were also captured using transmitted light. (B) Spectra region of interest (ROI)-1 and -2 represents fluorescence intensity with respect to the line drawn across C. albicans morphology in case of CF-N5. Red and green lines represent fluorescence due to PI and CF-N5 respectively. ROI stands for region of interest shown by arrows.

activity. When three disulfide bridges are introduced, enhancement in activity is observed. The activity of short three disulfide bridged peptide is comparable to HBD4, particularly against P. aeruginosa and C. albicans. Also, in presence of a reducing agent all the disulfide bridged peptides lose their activity completely at their lethal concentration. The disulfide bridges appear to be essential for antimicrobial activity and potency. Our results highlight the importance of the segments rich in cationic amino acids for activity and specificity as also observed in HBD1–3 [21,37]. The shorter peptides with a single disulfide constraint do not lose activity in the presence of NaCl against Gram-negative bacteria unlike HBD4 and peptides constrained with two or three disulfide links. HBD1 and HBD2 lose their activity in presence of high salt [2,16]. We have also shown HBD4 also loses its activity in presence of physiological concentration of salts as also reported elsewhere [15]. Wu et al. also have shown that linear analogs of

HBD3 were more salt-resistant than folded HBD3 [58]. The difference in activity between the shorter peptides and HBD4 in the presence of NaCl is less pronounced against S. aureus and C. albicans. Divalent ions are known to stabilize the outer membrane of Gram-negative bacteria [52] which is also evident from the loss of activity of peptide in presence of divalent cations especially in presence of Mg2+ ion. It is probably due to inhibition of electrostatic interaction between cationic peptide and negatively charged membrane. Against C. albicans, the shorter peptides with a single disulfide bond and three disulfide bonds are active in the presence of divalent cations unlike HBD4. While the shorter peptides are less active as compared to HBD4, they are partially less sensitive to the effect of salts. One of the strategies that have been deployed to overcome the problem of salt sensitivity was to generate hybrids of full length HBD1–3 [23,44]. Chimeric defensins composed of HBD2 and 3 sequences did show improved antimicrobial activity that was

H. Sharma, R. Nagaraj / Peptides 38 (2012) 255–265

263

Fig. 8. Confocal microscopy with lipid vesicles. Labeled peptides CF-N2, CF-N4 and CF-N5 at concentration of 2.5–40 ␮M were incubated with anionic and zwitterionic GUVs and MLVs (PE:PG::7:3) containing 1% rhodamine (Rh) conjugated lipids. Excitation wavelengths 488 nm and 543 nm and emission bands in the range of 494–568 nm and 590–712 nm for CF-labeled peptides and Rh conjugated lipid vesicles respectively were used. Images here are shown for 40 ␮M labeled peptides. (A) Vesicles treated with peptides were left over chamber slide for 30 min and then images were captured. (B) Fluorescence measured with respect to the line drawn across the vesicle in case of CF-N5. Red line represents Rh fluorescence and green line represents due to CF-N5. ROI stands for region of interest shown by arrows.

not attenuated by high concentration of NaCl. The disulfide connectivities in these peptides were identical to the beta defensin motif [23]. Some of the hybrid peptides did show salt-resistant activity. We have observed that some of the shorter segments of HBD4 are considerably resistant to the effect of NaCl as compared to the full length peptide. Although the number of disulfide bridges in the N-terminal cationic segment appear to play an important role in determining the lethal concentration as well as sensitivity to mono and divalent cations, their role in modulating conformation of the peptides is not apparent by CD spectroscopy. The spectra may be interpreted as typical of a “loose” ␤-turn. Even the spectrum of full-length HBD4 is not reminiscent of a peptide in highly folded conformation. Despite the presence of three ␤-strands in hairpin conformation, the CD of HBD4 in the present study and also of other defensins suggest considerable conformational flexibility. Unlike in the case of linear cationic antibacterial peptides [38,47], lipids do not have the ability to induce HBD4 and its analogs to fold into ordered conformation. Investigations on the mechanism of killing indicate that the shorter peptides enter E. coli and C. albicans cells rapidly and are localized in the cytosol. Bac71–35 , polymyxin B, and buforin II [4,35] are also internalized inside the cell but these peptides cause membrane permeabilization and leakage of cellular content. There does not appear to be a strong evidence for membrane permeabilization and pore formation as evident from calcein release assay (Fig. S2). Also staining of FM4-64 labeled cells do not show

membrane damage. Lipid binding is not avid with LUVs. Also, fluorescence microscopy study with GUVs and MLVs clearly indicate that the peptides do not have the ability to cross the lipid bilayer in model membranes. However, in bacterial cells, the peptides cross the membrane barrier very rapidly and exert their action inside the cells. A recent study on HBD3 shows that it acts on the cell wall biosynthetic machinery and it does not act by forming lesions in membrane [39]. Vylkova et al. has been shown HBD2 and 3 do not cause gross membrane permeabilization or membrane disruption in case of C. albicans [54,55]. Targeting of lipid II by HNP1 and HBD3 in S. aureus has lead to the proposal that the initial target on the S. aureus surface is likely to be lipid II and penetration of the lipid bilayer may not be an important step in bacterial killing [10,39]. In the case of Gram-negative bacteria membrane, activity against LPS was observed for HBD3 [6]. The variants of HBD4 and also possibly full length HBD4 appear to kill microorganisms in different manner rather than proposed mechanisms for linear cationic antimicrobial peptides [25,34,38,47]. In conclusion, our study suggests that even in a short peptide, spatial arrangement and number of disulfide bridges can play a crucial role in determining selective activity. The location of the cationic stretch in the disulfide loop appears to be an important determinant of activity. We propose that the cationic peptides exert their activity by interacting transiently with microbial membranes. This causes membrane destabilization resulting in the formation of transient

264

H. Sharma, R. Nagaraj / Peptides 38 (2012) 255–265

Fig. 9. Isothermal titration calorimetry of N1 and N2 peptides with LUVs. Peptides (50 ␮M) in calorimetric cell of the volume 1.40 ml in sodium phosphate buffer (10 mM, pH 7.4) were titrated with anionic, PE–PG (7:3), LUVs and zwitterionic, PC, LUVs. Concentration of lipid was 10 mM and 4 ␮l of vesicles was injected each time. Heat of dilution was corrected from heat generated during peptide lipid titration.

defects which provide a pathway for rapid accumulation of peptides in the cytoplasm. Excess of cationic peptides in cytoplasm could then have an inhibitory effect on several metabolic processes leading to cell death. Acknowledgements We thank Ch.V.B. Swamy, CCMB, for technical help with mass spectrometry of peptides. We thank Ms. Nandini Rangaraj, CCMB, for assistance in the use of confocal microscope. The project was funded by Network Project (NWP05), Council of Scientific and Industrial Research, India. Ramakrishnan Nagaraj is the recipient of the JC Bose fellowship from the Department of Science and Technology, India. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.peptides. 2012.08.024. References [1] Atherton E, Sheppard RC, editors. Solid phase peptide synthesis: a practical approach. Oxford: IRL Press; 1989. [2] Bals R, Wang X, Wu Z, Freeman T, Bafna V, Zasloff M, et al. Human beta-defensin 2 is a salt-sensitive peptide antibiotic expressed in human lung. J Clin Invest 1998;102:874–80. [3] Bauer F, Schweimer K, Kluver E, Conejo-Garcia JR, Forssmann WG, Rosch P, et al. Structure determination of human and murine beta-defensins reveals structural conservation in the absence of significant sequence similarity. Protein Sci 2001;10:2470–9. [4] Benincasa M, Pacor S, Gennaro R, Scocchi M. Rapid and reliable detection of antimicrobial peptide penetration into gram-negative bacteria based on fluorescence quenching. Antimicrob Agents Chemother 2009;53:3501–4. [5] Bensch KW, Raida M, Magert HJ, Schulz-Knappe P, Forssmann WG. hBD-1: a novel beta-defensin from human plasma. FEBS Lett 1995;368:331–5. [6] Bohling A, Hagge SO, Roes S, Podschun R, Sahly H, Harder J, et al. Lipid-specific membrane activity of human beta-defensin-3. Biochemistry 2006;45:5663–70. [7] Brogden KA. Antimicrobial peptides: pore formers or metabolic inhibitors in bacteria? Nat Rev Microbiol 2005;3:238–50.

[8] Chandrababu KB, Ho B, Yang D. Structure, dynamics, and activity of an allcysteine mutated human beta defensin-3 peptide analogue. Biochemistry 2009;48:6052–61. [9] Chaudhary N, Nagaraj R. Hen lysozyme amyloid fibrils induce aggregation of erythrocytes and lipid vesicles. Mol Cell Biochem 2009;328:209–15. [10] de Leeuw E, Li C, Zeng P, Diepeveen-de Buin M, Lu WY, Breukink E, et al. Functional interaction of human neutrophil peptide-1 with the cell wall precursor lipid II. FEBS Lett 2010;584:1543–8. [11] Ericksen B, Wu Z, Lu W, Lehrer RI. Antibacterial activity and specificity of the six human {alpha}-defensins. Antimicrob Agents Chemother 2005;49:269–75. [12] Fishov I, Woldringh CL. Visualization of membrane domains in Escherichia coli. Mol Microbiol 1999;32:1166–72. [13] Ganz T. Defensins: antimicrobial peptides of innate immunity. Nat Rev Immunol 2003;3:710–20. [14] Ganz T, Selsted ME, Szklarek D, Harwig SS, Daher K, Bainton DF, et al. Defensins. Natural peptide antibiotics of human neutrophils. J Clin Invest 1985;76:1427–35. [15] Garcia JR, Krause A, Schulz S, Rodriguez-Jimenez FJ, Kluver E, Adermann K, et al. Human beta-defensin 4: a novel inducible peptide with a specific salt-sensitive spectrum of antimicrobial activity. FASEB J 2001;15: 1819–21. [16] Goldman MJ, Anderson GM, Stolzenberg ED, Kari UP, Zasloff M, Wilson JM. Human beta-defensin-1 is a salt-sensitive antibiotic in lung that is inactivated in cystic fibrosis. Cell 1997;88:553–60. [17] Harder J, Bartels J, Christophers E, Schroder JM. Isolation and characterization of human beta-defensin-3, a novel human inducible peptide antibiotic. J Biol Chem 2001;276:5707–13. [18] Hill CP, Yee J, Selsted ME, Eisenberg D. Crystal structure of defensin HNP-3, an amphiphilic dimer: mechanisms of membrane permeabilization. Science 1991;251:1481–5. [19] Hoover DM, Chertov O, Lubkowski J. The structure of human beta-defensin1: new insights into structural properties of beta-defensins. J Biol Chem 2001;276:39021–6. [20] Hoover DM, Rajashankar KR, Blumenthal R, Puri A, Oppenheim JJ, Chertov O, et al. The structure of human beta-defensin-2 shows evidence of higher order oligomerization. J Biol Chem 2000;275:32911–8. [21] Hoover DM, Wu Z, Tucker K, Lu W, Lubkowski J. Antimicrobial characterization of human beta-defensin 3 derivatives. Antimicrob Agents Chemother 2003;47:2804–9. [22] Ishitsuka R, Yamaji-Hasegawa A, Makino A, Hirabayashi Y, Kobayashi T. A lipidspecific toxin reveals heterogeneity of sphingomyelin-containing membranes. Biophys J 2004;86:296–307. [23] Jung S, Mysliwy J, Spudy B, Lorenzen I, Reiss K, Gelhaus C, et al. Human beta-defensin 2 and beta-defensin 3 chimeric peptides reveal the structural basis of the pathogen specificity of their parent molecules. Antimicrob Agents Chemother 2011;55:954–60. [24] Kaiser V, Diamond G. Expression of mammalian defensin genes. J Leukoc Biol 2000;68:779–84.

H. Sharma, R. Nagaraj / Peptides 38 (2012) 255–265 [25] Klocek G, Schulthess T, Shai Y, Seelig J. Thermodynamics of melittin binding to lipid bilayers. Aggregation and pore formation. Biochemistry 2009;48:2586–96. [26] Kluver E, Adermann K, Schulz A. Synthesis and structure–activity relationship of beta-defensins, multi-functional peptides of the immune system. J Pept Sci 2006;12:243–57. [27] Kluver E, Schulz-Maronde S, Scheid S, Meyer B, Forssmann WG, Adermann K. Structure–activity relation of human beta-defensin 3: influence of disulfide bonds and cysteine substitution on antimicrobial activity and cytotoxicity. Biochemistry 2005;44:9804–16. [28] Krishnakumari V, Rangaraj N, Nagaraj R. Antifungal activities of human beta-defensins HBD-1 to HBD-3 and their C-terminal analogs Phd1 to Phd3. Antimicrob Agents Chemother 2009;53:256–60. [29] Krishnakumari V, Singh S, Nagaraj R. Antibacterial activities of synthetic peptides corresponding to the carboxy-terminal region of human beta-defensins 1–3. Peptides 2006;27:2607–13. [30] Lehrer RI, Ganz T. Defensins of vertebrate animals. Curr Opin Immunol 2002;14:96–102. [31] Lehrer RI, Lichtenstein AK, Ganz T. Defensins: antimicrobial and cytotoxic peptides of mammalian cells. Annu Rev Immunol 1993;11:105–28. [32] MacDonald RC, MacDonald RI, Menco BP, Takeshita K, Subbarao NK, Hu LR. Small-volume extrusion apparatus for preparation of large, unilamellar vesicles. Biochim Biophys Acta 1991;1061:297–303. [33] Mandal M, Nagaraj R. Antibacterial activities and conformations of synthetic alpha-defensin HNP-1 and analogs with one, two and three disulfide bridges. J Pept Res 2002;59:95–104. [34] Matsuzaki K. Why and how are peptide–lipid interactions utilized for selfdefense? Magainins and tachyplesins as archetypes. Biochim Biophys Acta 1999;1462:1–10. [35] Park CB, Yi KS, Matsuzaki K, Kim MS, Kim SC. Structure–activity analysis of buforin II, a histone H2A-derived antimicrobial peptide: the proline hinge is responsible for the cell-penetrating ability of buforin II. Proc Natl Acad Sci U S A 2000;97:8245–50. [36] Pazgier M, Hoover DM, Yang D, Lu W, Lubkowski J. Human beta-defensins. Cell Mol Life Sci 2006;63:1294–313. [37] Pazgier M, Prahl A, Hoover DM, Lubkowski J. Studies of the biological properties of human beta-defensin 1. J Biol Chem 2007;282:1819–29. [38] Saberwal G, Nagaraj R. A synthetic peptide corresponding to the hydrophobic amino terminal region of pardaxin can perturb model membranes of phosphatidyl choline and serine. Biochim Biophys Acta 1989;984: 360–4. [39] Sass V, Schneider T, Wilmes M, Korner C, Tossi A, Novikova N, et al. Human beta-defensin 3 inhibits cell wall biosynthesis in Staphylococci. Infect Immun 2010;78:2793–800. [40] Sawai MV, Jia HP, Liu L, Aseyev V, Wiencek JM, McCray Jr PB, et al. The NMR structure of human beta-defensin-2 reveals a novel alpha-helical segment. Biochemistry 2001;40:3810–6.

265

[41] Schibli DJ, Hunter HN, Aseyev V, Starner TD, Wiencek JM, McCray Jr PB, et al. The solution structures of the human beta-defensins lead to a better understanding of the potent bactericidal activity of HBD3 against Staphylococcus aureus. J Biol Chem 2002;277:8279–89. [42] Schneider JJ, Unholzer A, Schaller M, Schafer-Korting M, Korting HC. Human defensins. J Mol Med 2005;83:587–95. [43] Schroeder BO, Wu Z, Nuding S, Groscurth S, Marcinowski M, Beisner J, et al. Reduction of disulphide bonds unmasks potent antimicrobial activity of human beta-defensin 1. Nature 2011;469:419–23. [44] Scudiero O, Galdiero S, Cantisani M, Di Noto R, Vitiello M, Galdiero M, et al. Novel synthetic, salt-resistant analogs of human beta-defensins 1 and 3 endowed with enhanced antimicrobial activity. Antimicrob Agents Chemother 2010;54:2312–22. [45] Seelig J. Titration calorimetry of lipid–peptide interactions. Biochim Biophys Acta 1997;1331:103–16. [46] Selsted ME, Ouellette AJ. Mammalian defensins in the antimicrobial immune response. Nat Immunol 2005;6:551–7. [47] Shai Y. Mode of action of membrane active antimicrobial peptides. Biopolymers 2002;66:236–48. [48] Tam JP, Wu CR, Liu W, Zhang JW. Disulfide bond formation in peptides by dimethyl sulfoxide. Scope and applications. J Am Chem Soc 1991;113:6657–62. [49] Taylor K, Barran PE, Dorin JR. Structure–activity relationships in beta-defensin peptides. Biopolymers 2008;90:1–7. [50] Terlau H, Olivera BM. Conus venoms: a rich source of novel ion channeltargeted peptides. Physiol Rev 2004;84:41–68. [51] Tomita T, Hitomi S, Nagase T, Matsui H, Matsuse T, Kimura S, et al. Effect of ions on antibacterial activity of human beta defensin 2. Microbiol Immunol 2000;44:749–54. [52] Vaara M. Agents that increase the permeability of the outer membrane. Microbiol Rev 1992;56:395–411. [53] Varkey J, Nagaraj R. Antibacterial activity of human neutrophil defensin HNP-1 analogs without cysteines. Antimicrob Agents Chemother 2005;49:4561–6. [54] Vylkova S, Li XS, Berner JC, Edgerton M. Distinct antifungal mechanisms: beta-defensins require Candida albicans Ssa1 protein, while Trk1p mediates activity of cysteine-free cationic peptides. Antimicrob Agents Chemother 2006;50:324–31. [55] Vylkova S, Nayyar N, Li W, Edgerton M. Human beta-defensins kill Candida albicans in an energy-dependent and salt-sensitive manner without causing membrane disruption. Antimicrob Agents Chemother 2007;51: 154–61. [56] Weber PJ, Bader JE, Folkers G, Beck-Sickinger AG. A fast and inexpensive method for N-terminal fluorescein-labeling of peptides. Bioorg Med Chem Lett 1998;8:597–600. [57] Woody RW. Circular dichroism. Methods Enzymol 1995;246:34–71. [58] Wu Z, Hoover DM, Yang D, Boulegue C, Santamaria F, Oppenheim JJ, et al. Engineering disulfide bridges to dissect antimicrobial and chemotactic activities of human beta-defensin 3. Proc Natl Acad Sci U S A 2003;100:8880–5.