Archives of Biochemistry and Biophysics 573 (2015) 14–22
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi
Antibacterial and membrane-damaging activities of mannosylated bovine serum albumin Chia-Yu Tsai a, Ying-Jung Chen a, Yaw-Syan Fu b, Long-Sen Chang a,c,⇑ a
Institute of Biomedical Sciences, National Sun Yat-Sen University, Kaohsiung 804, Taiwan Department of Biomedical Science and Environmental Biology, Kaohsiung Medical University, Kaohsiung 807, Taiwan c Department of Biotechnology, Kaohsiung Medical University, Kaohsiung 807, Taiwan b
a r t i c l e
i n f o
Article history: Received 31 January 2015 and in revised form 25 February 2015 Available online 6 March 2015 Keywords: Mannosylated BSA Bactericidal effect Membrane-damaging activity Fusogenicity
a b s t r a c t The aim of this study was to test whether mannosylated BSA (Man-BSA) exerts antibacterial activity on Escherichia coli (gram-negative bacteria) and Staphylococcus aureus (gram-positive bacteria) via its membrane-damaging effect. Man-BSA caused inhibition of growth of E. coli and S. aureus. Moreover, bactericidal action of Man-BSA on E. coli and S. aureus positively correlated with the increase in membrane permeability of the bacterial cells. Morphological examination showed that Man-BSA disrupted bacterial membrane integrity. Destabilization of the lipopolysaccharide (LPS) layer and inhibition of lipoteichoic acid (LTA) biosynthesis in the cell wall increased the bactericidal effect of Man-BSA on E. coli and S. aureus. Man-BSA also induced leakage and fusion of membrane-mimicking liposomes in E. coli and S. aureus. Man-BSA showed similar binding affinity for LPS and LTA. LPS and LTA strongly suppressed the membrane-damaging activity of Man-BSA, whereas an increase in the Man-BSA concentration attenuated the inhibitory action of LPS and LTA. Taken together, our data indicate that Man-BSA’s bactericidal activity depends strongly on its ability to induce membrane permeability. Moreover, the bactericidal action of Man-BSA proven in this study suggests that Man-BSA may serve as a prototype for the development of anti-infective agents targeting E. coli and S. aureus. Ó 2015 Elsevier Inc. All rights reserved.
Introduction The development of resistance of bacteria to antibiotics is a global problem that underscores the need for new therapeutic agents. Antimicrobial peptides and proteins are currently under consideration as a potential alternative to conventional antibiotics, on account of their widespread occurrence in nature [1,2]. Antimicrobial peptides and proteins display a broad spectrum of activities against a wide range of pathogens including bacteria, fungi and enveloped viruses. A common feature of most of these peptides and proteins is that they are cationic and have amphipathic properties. Because the bacterial membrane consists of abundant negatively charged phospholipids, it is believed that most of the antimicrobial peptides and proteins interact with
⇑ Corresponding author at: Institute of Biomedical Sciences, National Sun Yat-Sen University, Kaohsiung 804, Taiwan. Fax: +886 7 5250197. E-mail address: [email protected] (L.-S. Chang). http://dx.doi.org/10.1016/j.abb.2015.02.034 0003-9861/Ó 2015 Elsevier Inc. All rights reserved.
anionic phospholipids and kill microorganisms by permeabilizing the bacterial membrane, by thinning the membrane or by destabilizing the membrane structure [3]. Nevertheless, antimicrobial peptides and proteins may kill bacteria by inhibiting macromolecular biosynthesis and/or by interacting with specific vital components inside the bacterial cells [3]. Given that membrane composition of bacteria includes abundantly anionic phospholipids, an increase in the positive charge of the antibiotic proteins via blocking of negatively charged carboxylate groups may enhance the interaction of such proteins with the bacterial membrane and thus enhance the potency of the antibacterial effect. Several studies revealed that bovine a-lactoglobulin and bovine lactoferrin have antibacterial properties [4–6]. Pan et al. [7–9] found that amidation of bovine a-lactoglobulin and bovine lactoferrin increases the net positive charge of these proteins, sharply increasing their bactericidal activity. Nevertheless, there are no studies exploring the possibility that novel antimicrobial proteins could be prepared from nonbactericidal proteins after modification of carboxyl groups. Because antimicrobial proteins usually exert their activity by damaging the bacterial membrane, proteins that preferably interact with phospholipids but do not
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
show bactericidal activity can be a good candidate for testing of the above possibility. Various studies revealed that bovine serum albumin (BSA)1 can transfer a lipid amphiphile to lipid bilayer membranes [10], meaning that BSA can interact with lipid bilayers. Moreover, mannosylated BSA (Man-BSA), which is prepared by conjugation of BSA’s carboxyl groups with p-aminophenyl-a-mannopyranoside is reported to have membrane-damaging effects on zwitterionic lipid vesicles [11]. Thus, the aim of this study was to characterize the antibacterial activity and the mechanism of ManBSA’s action.
15
measuring the optical density (OD) at 550 nm (OD550nm 0.3 for E. coli and OD550nm 0.5 for S. aureus corresponding to approximately to 108 CFU/ml). The bacteria were pelleted, washed with PBS, and then concentrated to 5 107 CFU/ml in PBS. One hundred microliters of the bacteria suspension were incubated with Man-BSA at 37 °C for indicated time periods. Then the cells were cultured in 1 ml LB medium for E. coli or TSB for S. aureus for 6 h with gently shaking. Then antimicrobial activity of Man-BSA was measured using optical density (OD) at 550 nm. Antimicrobial activity of Man-BSA was calculated as (OD550 after Man-BSA treatment)/(OD550 without Man-BSA treatment) 100%.
Materials and methods Bacterial membrane permeability assessment BSA (Fraction V, fatty acid free, catalog number 775835, purity >99%) was obtained from Roche Applied Science. Calcein, rifampin, propidium iodide (PI), cardiolipin, 1-ethyl-3-(diethylaminopropyl) -carbodiimide hydrochloride (EDC), p-aminophenyl a-D-mannopyranoside, p-aminophenyl a-D-glucopyranoside, rhodamine isothiocyanate, sodium dithionite, egg yolk phosphatidylglycerol (EYPG), egg yolk phosphatidylethanolamine (EYPE), lipopolysaccharide (LPS) 0111:B4 from Escherichia coli and lipoteichoic acid (LTA) from Staphylococcus aureus were purchased from Sigma– Aldrich Inc., and N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)-1,2-dihexadecanoyl-sn-glycerol-3-phosphoethanolamine (NBD-PE) and lissamine rhodamine B 1,2-dihexadecanoly-sn-glycero-3-phosphoethanolamine (Rh-PE) were the products of Molecular Probes. Unless otherwise specified, all other reagents were analytical grade.
Bacterial suspensions (5 107 CFU/ml) were treated with ManBSA or PBS in a total volume of 100 ll under the same conditions used in antimicrobial assay. Subsequently, the bacterial cells (E. coli or S. aureus) were incubated with propidium iodide (PI) for 30 min in the dark at room temperature. Then the bacterial cells were washed with PBS, and resuspended in 1 ml PBS. PI fluorescence values were corrected for background (extracellular) dye fluorescence and expressed as a percentage relative to the value obtained after complete permeabilization of the cells by heating at 100 °C for 5 min. Uptake of PI was measured by Beckman Coulter Paradigm™ Detection Platform with excitation at 535 nm and emission at 595 nm. Competitive replacement of LPS-bound or LTA-bound rhodaminelabeled Man-BSA (Rh-Man-BSA) by unlabeled Man-BSA
Preparation of Man-BSA and glucosylated-BSA (Glu-BSA) Man-BSA
was
prepared
by
coupling
p-aminophenyl
a-D-mannopyranoside to BSA through water-soluble EDC according to the method described in [11]. BSA contains 41 Asp, 58 Glu and 1 C-terminal carboxyl groups. MALDI-TOF analyses showed that, as compared with BSA, Man-BSA showed an increase in the molecular weight by 24,050 Da [11]. Thus, the conjugation of approximately 95 p-aminophenyl a-D-mannopyranoside with the carboxyl groups in Man-BSA was calculated from the increment in the molecular weight. BSA was also modified with p-aminophenyl a-D-glucopyranoside according to the procedure described in [11]. MALDI-TOF analyses revealed that approximately 86 of the 100 carboxyl groups in BSA were conjugated with p-aminophenyl a-D-glucopyranoside. Bacterial strains E. coli (JM109) and S. aureus (ATCC 25923) were used in this study. E. coli were maintained on Luria–Bertani (LB) agar plate at 37 °C, and S. aureus were maintained on tryptic soy agar plate at 37 °C. Antimicrobial assay E. coli (JM109) were grown in LB medium and S. aureus were grown in tryptic soy broth (TSB) from a single colony with agitation at 37 °C. Bacterial number was then evaluated by 1 Abbreviations used: BSA, bovine serum albumin; Man-BSA, mannosylated BSA; LPS, lipopolysaccharide; LTA, lipoteichoic acid; PI, propidium iodide; EDC, 1-ethyl-3(diethylaminopropyl)-carbodiimide hydrochloride; EYPG, egg yolk phosphatidylglycerol; EYPE, egg yolk phosphatidylethanolamine; NBD-PE, N-(7-nitrobenz-2-oxa-1,3diazol-4-yl)-1,2-dihexadecanoyl-sn-glycerol-3-phosphoethanolamine; Rh-PE, lissamine rhodamine B 1,2-dihexadecanoly-sn-glycero-3-phosphoethanolamine; LB, Luria–Bertani; TSB, tryptic soy broth; OD, optical density; GAPDH, glyceraldehyde3-phosphate dehydrogenase; PE, phosphatidylethanolamine; PG, phosphatidylglycerol.
Modification of Man-BSA with rhodamine isothiocyanate was prepared according to the procedure described in [12]. Rh-ManBSA (1 lM) was titrated with increasing concentrations of LPS or LTA until maximal changes in fluorescence intensity of Rh-ManBSA was achieved. Then increasing concentrations of unlabeled Man-BSA were added to compete for binding of Rh-Man-BSA with LPS or LTA. Competitive binding was monitored at excitation wavelength and emission wavelength at 550 and 580 nm, respectively. Release of entrapped fluorescent marker from liposomes Membrane-damaging activity of Man-BSA was determined by measuring the release of the liposome-entrapped, self-quenching fluorescent dye calcein. EYPE/EYPG (75/25, mol/mol) or EYPG/cardiolipin (60/40, mol/mol) were dissolved in chloroform/ methanol (v/v, 2:1) and dried by evaporation. Buffer (10 mM Tris–HCl, 100 mM NaCl, pH 7.5) containing 50 mM calcein was added to the film of lipids, and after hydration the suspension was shaken vigorously. The multilamellar vesicles obtained in this way were extruded 10 times, above the transition temperature, through a 100-nm polycarbonate filter and applied to a Sepharose 6B column (2 15 cm) to separate the liposome from the free calcein. Leakage was induced by adding aliquots of Man-BSA to a vesicle suspension directly in the cuvette used for fluorescence determination at 30 °C. The kinetics of membrane damage were monitored by the increase in fluorescence with emission at 520 nm and excitation at 490 nm, and the signal was expressed as percentage of total calcein release after the addition of 0.2% Triton X-100. Fusion assay of phospholipid vesicles Liposome fusion induced by Man-BSA was measured by increase in fluorescence resonance energy transfer between two lipid probes (NBD-PE and Rh-PE). NBD-PE and Rh-PE were used as donor and acceptor fluorescent lipid, respectively. NBD-PE and
16
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
Rh-PE (1 mol% in phospholipid vesicles) were incorporated into phospholipid vesicles. The NBD emission was monitored at 530 nm with the excitation wavelength set at 465 nm. As the two liposome groups interacted, the fluorescence energy emitted from NBD-PE labeled liposomes was transferred to the Rh-PE labeled liposomes resulting in a decreased fluorescence signal of NBD. The percentage of fusion was defined by the following relationship: %Fusion = 100 (1 F/F0), where F and F0 are the fluorescence intensities in the presence and absence of the RhPE, respectively. For the inner monolayer phospholipid fusion assay, phospholipid vesicles were treated with sodium dithionite to completely reduce the NBD-labeled phospholipid located at the outer monolayer of the membrane. Final concentration of sodium dithionite was 100 mM and the inner leaflet mixing assay was carried out at the level in which 55–60% fluorescence intensity of NBD were reduced. Scanning electron microscopy (SEM) E. coli were grown at 37 °C to mid-exponential phase (OD550 = 0.5–0.6) and incubated with 1.1 lM Man-BSA under the same conditions used in antimicrobial assay. After incubation for 24 h, Man-BSA-treated cells were prepared for SEM analysis according to the procedure described in [13]. The micrographs were viewed at a 5.0 kV accelerating voltage on a Hitachi SU8000 SEM, and a secondary electron image of the cells for topography contrast was collected at several magnifications. Detection of TNF-a expression Human acute myeloid leukemia U937 cells were obtained from ATCC (Manassas, VA, USA). U937 cells were cultured in RPMI 1640 supplemented with 10% fetal calf serum, 1% sodium pyruvate, 2 mM glutamine and penicillin (100 units/ml)/streptomycin (100 lg/ml) in an incubator humidified with 95% air and 5% CO2. Cells were stimulated with indicated LPS or LTA concentrations in the absence or presence of Man-BSA for 24 h. Total RNA was isolated from cells using the RNeasy minikit (QIAGEN Inc., Valencia, CA) according to the instructions of the manufacturer. Reverse transcriptase reaction was performed with 2 lg of total RNA using M-MLV reverse transcriptase (Promega, Madison, WI) according to the manufacturer’s recommendations. A reaction without reverse transcriptase was performed in parallel to ensure the absence of genomic DNA contamination. After initial denaturation at 95 °C for 10 min, PCR amplification was performed using GoTaq Flexi DNA polymerase (Promega, Madison, WI) followed by 35 cycles at 94 °C for 50 s, 65 °C for 50 s, and 72 °C for 50 s. After a final extension at 72 °C for 5 min, PCR products were resolved on 2% agarose gels and visualized by ethidium bromide transillumination under UV light. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control to check the efficiency of cDNA synthesis and PCR amplification. Primer sequences were as follows: TNF-a-forward, 50 -ATGAGCACTGAAA GCATGATCCGG30 and TNF-a-reverse 50 -GCAATGATCCCAAAGTAGACCTGCCC-30 ; GAPDH-forward: 50 -GAGTCAACGGATTTGGTCGT-30 ; GAPDH-reverse: 50 -TGTGGTCATG AGTCCTTCCA-30 , and the reverse transcriptase-PCR products were subsequently confirmed by direct sequencing. Statistical analysis All data are presented as mean ± SD. Significant differences among the groups were determined using the unpaired Student’s t-test. A value of P < 0.05 was taken as an indication of statistical significance.
Results and discussion As shown in Fig. 1A and B, the antibacterial activity of Man-BSA toward E. coli and S. aureus was dose dependent. The half-maximal inhibitory concentration (IC50) of Man-BSA for E. coli and S. aureus was 0.85 ± 0.15 and 0.61 ± 0.11 lM, respectively. This finding indicates that Man-BSA has a stronger bactericidal effect on S. aureus than on E. coli. Moreover, Man-BSA killed E. coli and S. aureus most effectively after treatment for 24 h (Fig. 1C and D). On the other hand, intact BSA did not significantly suppress the growth of E. coli and S. aureus (Fig. 1A and B). The effect of Man-BSA on the permeability of the inner membrane of E. coli and S. aureus was monitored using a PI fluorescent probe. Inner permeability changes evoked by Man-BSA resulted in an increased intracellular concentration of PI, which is detected thanks to the fluorescence resulting from the binding of the probe to nucleic acids present in the cytoplasma. Fluorescence measurement showed that the percentage of permeabilized cells increased after incubation with Man-BSA (Fig. 2), whereas BSA did not increase membrane permeability of E. coli and S. aureus (inset in Fig. 2). The Man-BSA concentration that resulted in 50% of permeabilized E. coli and S. aureus cells was approximately 0.96 and 0.75 lM, respectively. This result also indicated that, compared to E. coli, S. aureus is more sensitive to the antibacterial action of Man-BSA. Furthermore, the antibacterial activity was authenticated by our ultrastructural studies; representative micrographs are shown in Fig. 3. Compared to BSA-treated bacterial cells, Man-BSA-treated E. coli and S. aureus showed pronounced morphological changes consistent with damage to the bacterial cells. The Man-BSA-treated bacteria displayed significant wrinkling, surface roughening and membrane blebbing. This suggests that the ManBSA-treated bacterial cells lost integrity of their membrane. LPS and LTA contain a saccharide moiety and are major constituents of the cell wall of E. coli and S. aureus, respectively [14]. Thus, the effect of LPS and LTA on Man-BSA’s antibacterial activity was examined too. Other studies suggest that LPS-associated cations tightly cross-link the LPS layer whereas EDTA treatment causes the loss of LPS [15,16]. In sharp contrast to Ca2+, EDTA enhanced the bactericidal effect of Man-BSA (Fig. 4A). Laganas et al. [17] showed that rifampin inhibits LTA biosynthesis. Fig. 4B shows that rifampin pretreatment enhanced Man-BSA’s bactericidal effect on S. aureus. Taken together, these findings indicate that LPS and LTA may inhibit Man-BSA’s antibacterial action. To characterize the interaction of Man-BSA with LPS and LTA, we performed replacement of LPS- or LTA-bound Rh-Man-BSA by unlabeled Man-BSA. As shown in Fig. 5, the fluorescence intensity of Rh-Man-BSA was reduced by titration with LPS or LTA, and the maximal reduction in fluorescence intensity of Rh-Man-BSA was noted with 200 lg of LPS or 230 lg of LTA. This result likely means that changes in the conformation of Rh-Man-BSA occurred after binding to LPS or LTA. Gradual restoration of the fluorescence intensity was observed after the addition of unlabeled Man-BSA; this phenomenon is indicative of competitive displacement of Rh-Man-BSA from LPS or LTA. Calculations that were based on the changes in fluorescence intensity showed that the dissociation constants of Man-BSA for LPS and LTA were 0.08 ± 0.03 and 0.09 ± 0.02 lM, respectively. On the other hand, unmodified BSA did not competitively displace Rh-Man-BSA from LPS and LTA (data not shown). Some studies indicated that the E. coli membrane contains phosphatidylethanolamine (PE) and phosphatidylglycerol (PG) at the molar ratio of approximately 75:25, whereas the S. aureus membrane contains PG and cardiolipin at the molar ratio of approximately 60:40 [18]. Accordingly, we tested whether ManBSA induces the leakage of calcein from EYPE/EYPG (75:25,
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
17
Fig. 1. Bactericidal activity of Man-BSA on E. coli and S. aureus. Bactericidal activity of Man-BSA against (A) E. coli and (B) S. aureus showed a concentration-dependent manner. E. coli or S. aureus were treated with indicated Man-BSA concentrations for 24 h. Bactericidal activity of Man-BSA against (C) E. coli and (D) S. aureus showed a time-dependent manner. E. coli and S. aureus were treated with 1.1 lM Man-BSA for indicated time periods. Data are presented as mean ± SD of three independent experiments with triplicate measurements (⁄P < 0.05 at all tested Man-BSA concentrations, bactericidal effect of Man-BSA vs. bactericidal effect of BSA).
Fig. 2. Membrane permeability of E. coli and S. aureus after Man-BSA treatment. Man-BSA induced increase in membrane permeability of E. coli and S. aureus in a concentration-dependent manner. E. coli and S. aureus were treated with indicated Man-BSA concentrations for 24 h. Man-BSA-treated bacteria were incubated with propidium iodide (PI) for 30 min before fluorescence measurement. Maximal membrane permeability signal was obtained by heating E. coli and S. aureus at 100 °C for 5 min. (Inset) Effect of Man-BSA and BSA on membrane permeability of E. coli and S. aureus. E. coli and S. aureus were treated with 2.2 lM Man-BSA or BSA for 24 h. Data are presented as mean ± SD of three independent experiments with triplicate measurements (⁄P < 0.05).
mol/mol) and EYPG/cardiolipin (60:40, mol/mol) vesicles. As shown in Fig. 6, Man-BSA caused the release of calcein from the EYPE/EYPG and EYPG/cardiolipin vesicles in a concentration-
dependent manner. The maximal calcein release from EYPE/EYPG and EYPG/cardiolipin vesicles was observed when the Man-BSA concentration was >0.6 lM. The Man-BSA-induced calcein leakage
18
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
Fig. 3. SEM of Man-BSA-treated E. coli and S. aureus. (A) BSA-treated E. coli; (B) Man-BSA-treated E. coli; (C) BSA-treated S. aureus; (D) Man-BSA-treated S. aureus. E. coli and S. aureus were grown at 37 °C to mid-exponential phase (OD550 = 0.5–0.6) and incubated with 1.1 lM BSA or Man-BSA. After incubation for 24 h, BSA- and Man-BSA-treated cells were prepared for SEM analysis according to the procedure described in Torrent et al. [13].
Fig. 4. Integrity of cell wall reduced antibacterial activity of Man-BSA. (A) Effect of EDTA and Ca2+ on bactericidal activity of Man-BSA against E. coli. E. coli were treated with 1.1 lM Man-BSA in the presence of 2 mM EDTA or CaCl2 for 24 h. (B) Inhibition of LTA biosynthesis enhanced bactericidal activity of Man-BSA. S. aureus were incubated with 1.56 lg/ml rifampin at 37 °C for 1 h before treatment with 1.1 lM Man-BSA at 37 °C for 24 h. Data are presented as mean ± SD of three independent experiments with triplicate measurements (⁄P < 0.05).
from EYPG/cardiolipin vesicles was noticeably stronger than that from EYPE/EYPG vesicles. In contrast, BSA did not cause the leakage of EYPE/EYPG and EYPG/cardiolipin vesicles (inset in Fig. 6). Because LPS and LTA with intact structure reduced the antibacterial activity of Man-BSA (Fig. 4), we assessed the effect of LPS and LTA on the membrane-damaging activity of Man-BSA. As shown in Fig. 7, LPS and LTA suppressed the membrane-damaging activity of Man-BSA in a concentration-dependent manner. On the other hand, an increase in Man-BSA concentration abrogated the inhibitory action of LPS and LTA. Given that Man-BSA is able to induce liposome fusion and membrane leakage [11], the fusogenicity of Man-BSA was then examined. Fig. 8 shows that a lipid mixing assay involving a fluorescence energy transfer between NBD-PE and Rh-PE was used to study the fusion event induced by Man-BSA. The fluorescence intensity of NBD-PE did not significantly decrease when the NBD-PE-labeled
vesicles were mixed with Rh-PE-labeled vesicles without the addition of Man-BSA. Fig. 8 shows that Man-BSA-induced fusion of EYPG/cardiolipin and EYPE/EYPG vesicles, and the fusogenicity of Man-BSA in relation to EYPG/cardiolipin vesicles was higher compared to that of EYPE/EYPG vesicles. On the other hand, BSA did not have a fusogenic effect on EYPE/EYPG and EYPG/cardiolipin vesicles (inset in Fig. 8). Under controlled conditions, sodium dithionite reduced the NBD attached to the lipid headgroup in the outer leaflet to a nonfluorescent derivative while leaving the NBD in the inner leaflet largely unaffected. When Man-BSA was added to the dithionite-treated EYPG/cardiolipin and EYPE/EYPG vesicles, a reduction in NBD fluorescence signal was observed, pointing to inner-leaflet admixing. Given that merging of the two inner leaflets should be concomitant with formation of the fusion pore [19], the inner leaflet admixing is expected to cause content mixing and/or leakage of vesicular content.
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
19
Fig. 5. Replacement of LPS- or LTA-bound Rh-Man-BSA by Man-BSA. Rh-Man-BSA (1 lM) dissolved in 10 mM Tris–HCl-100 mM NaCl (pH 7.5) were incubated with 200 lg LPS (A) or 240 lg LTA (B) for 1 h, and increasing concentrations of unlabeled Man-BSA were then added. Maximal reduction in fluorescence intensity of Rh-Man-BSA was noted when 200 lg LPS or 240 lg LTA was added. Replacement of LPS-bound or LTA-bound Rh-Man-BSA by unlabeled Man-BSA led to a restoration in fluorescence intensity of Rh-Man-BSA. The addition of unlabeled Man-BSA into the solution of Rh-Man-BSA did not change the fluorescence intensity of Rh-Man-BSA in the absence of LPS or LTA. Moreover, the fluorescence intensity of Rh-Man-BSA-LPS or Rh-Man-BSA-LTA complexes was not significantly restored by titration with buffer (10 mM Tris–HCl-100 mM NaCl, pH 7.5). Data are presented as mean ± SD of three independent experiments.
Fig. 6. Membrane-damaging activity of Man-BSA. Man-BSA induced calcein release from (A) EYPE/EYPG and (B) EYPG/cardiolipin vesicles in a concentration-dependent manner. The experiments were performed in 10 mM Tris–HCl–100 mM NaCl (pH 7.5). The final concentration of lipids was 2.5 lM. The signal was expressed as the percentage of total calcein release after the addition of 0.2% Triton X-100. (Inset) Effect of BSA on membrane permeability of EYPE/EYPG and EYPG/cardiolipin vesicles. The used BSA and Man-BSA concentrations were 0.66 lM. Data are presented as mean ± SD of three independent experiments with triplicate measurements (⁄P < 0.05).
Various studies revealed that an antibacterial peptide can abrogate LPS- and LTA-induced TNF-a expression [20–22]. Thus, we next analyzed the effect of Man-BSA on LPS- and LTA-induced TNF-a expression. As shown in Fig. 9A, LPS and LTA dose-dependently enhanced TNF-a expression in U937 cells: 10 lg/ml LPS and 0.1 lg/ml LTA notably increased TNF-a expression. Fig. 9B shows that cotreatment with Man-BSA actively inhibited the TNF-a expression in the LPS- and LTA-treated U937 cells, whereas BSA had no effect. Other studies showed that binding to LPS and LTA may enhance self-promoted uptake of antimicrobial peptides and proteins and thereby enhance their membrane-permeabilizing activity [23– 26]. Our data revealed that the binding of Man-BSA to LPS and LTA attenuates this protein’s bactericidal activity. Accordingly, LPS and LTA do not function as a polyanionic ladder for traversing Man-BSA from the outside through the cytoplasmic membrane. Together with the findings that Man-BSA induces permeabilization of the inner membrane (according to the PI fluorescence experiments, Fig. 2), the above result suggests that the bactericidal effect
of Man-BSA is directly linked to its membrane-permeabilizing activity. The finding that Man-BSA induced the leakage of bacterial-membrane-mimicking vesicles supports this notion. Moreover, Man-BSA induced the fusion of bacterial-membranemimicking vesicles in our experiments. Because the peptide-induced vesicle-vesicle fusion causes simultaneous leakage of vesicular contents [27,28], our data suggest that the membranedamaging effects of Man-BSA on EYPE/EYPG and EYPG/cardiolipin are mediated by fusogenic events. Furthermore, destabilization of LPS structure enhanced the bactericidal activity of Man-BSA (Fig. 4), whereas LPS suppressed Man-BSA’s membrane-damaging activity (Fig. 8). Likewise, LTA abrogated the membrane-damaging activity of Man-BSA, whereas inhibition of LTA synthesis enhanced Man-BSA’s bactericidal effect on S. aureus (Fig. 4). It is evident that binding to LPS and LTA reduces Man-BSA’s bactericidal activity. Nevertheless, increased Man-BSA concentration can attenuate the suppressive effect of LPS and LTA on the Man-BSA-induced membrane permeability. On the other hand, Man-BSA-induced calcein leakage from EYPG/cardiolipin vesicles
20
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
Fig. 7. Effect of LPS and LTA on membrane-damaging activity of Man-BSA. (A) Inhibitory effect of LPS on membrane-damaging activity of Man-BSA toward EYPE/EYPG vesicles. (B) Inhibitory effect of LTA on membrane-damaging activity of Man-BSA toward EYPG/cardiolipin. Man-BSA (0.66 lM) was incubated with indicated concentrations of LPS or LTA for 1 h, and then membrane-damaging activity of Man-BSA-LPS complexes or Man-BSA-LTA complexes was measured. The experiments were performed in 10 mM Tris–HCl–100 mM NaCl (pH 7.5). The signal was expressed as the percentage of total calcein release after addition of 0.2% Triton X-100. Data are presented as mean ± SD of six independent experiments.
Fig. 8. Man-BSA induced fusion of EYPE/EYPG and EYPG/cardiolipin vesicles. The experiments were performed in 10 mM Tris–HCl–100 mM NaCl (pH 7.5). Equal concentrations of NBD-PE-labeled and Rh-PE-labeled phospholipid vesicles were used. The final concentration of lipids was 200 lM. Man-BSA induced fusion of (A) EYPE/ EYPG and (B) EYPG/cardiolipin vesicles in a concentration-dependent manner. (Inset) Fusogenicity of Man-BSA and BSA on EYPE/EYPG and EYPG/cardiolipin vesicles. The used BSA and Man-BSA concentrations were 0.55 lM. Data are presented as mean ± SD of three independent experiments with triplicate measurements (⁄P < 0.05).
Fig. 9. Effect of Man-BSA on LPS- and LTA-induced TNF-a mRNA expression in U937 cells. (A) LPS (left panel) and LTA (right panel) dose-dependently induced up-regulation of TNF-a mRNA expression in U937 cells. (B) Man-BSA repressed LPS- and LTA-induced TNF-a mRNA expression.
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
21
Fig. 10. Bactericidal activity of Glu-BSA on E. coli and S. aureus. Bactericidal activity of Glu-BSA against (A) E. coli and (B) S. aureus showed a concentration-dependent manner. E. coli or S. aureus were treated with indicated Glu-BSA concentrations for 24 h. Data are presented as mean ± SD of three independent experiments with triplicate measurements.
was higher than that from EYPE/EYPG vesicles (Fig. 6). Man-BSAinduced fusion of EYPG/cardiolipin vesicles was also more pronounced than that of EYPE/EYPG vesicles (Fig. 8). EYPG and cardiolipin are anionic phospholipids, whereas EYPE is a zwitterionic phospholipid. Other studies showed that Naja naja atra CTX3 has a stronger membrane-damaging activity toward EYPG/cardiolipin vesicles than toward EYPE/EYPG vesicles [29]. Moreover, it was reported that CTX3 assumes a different membrane-bound state upon adsorption onto EYPE/EYPG and EYPG/cardiolipin vesicles [29]. Taken together, these observations suggest that the membrane-interacting mode of Man-BSA is related to the difference in Man-BSA’s damaging effects on EYPE/EYPG and EYPG/cardiolipin vesicles. This notion is supported by the finding that Man-BSA assumes different conformations after binding to lipid vesicles of different lipid composition [11]. LPS and LTA can be released from the cell wall when bacteria are killed by host immune cells or an antibiotic drug. LPS is an endotoxin and plays a key role in the pathophysiology of inflammation, sepsis and shock [30]. Like LPS, LTA also elicits an inflammatory reaction [31]. LPS and LTA induce the inflammatory response via binding to Toll-like receptor 4 (TLR4) and TLR2/6, respectively [30–32]. Several studies revealed that an antimicrobial peptide can suppress LPS- or LTA-induced TNF-a production in human monocytes or in macrophage cell lines [20–22]. The results of these studies suggest that the interaction of antimicrobial peptides with LPS and LTA impedes the binding of LPS and LTA to cell surface receptors, thereby attenuating the LPS- and LTA-induced TNF-a expression [20–22]. Because Man-BSA shows binding affinity for LPS and LTA (Fig. 5), a similar mechanism may explain the suppressive effect of Man-BSA on the LTA- and LPS-induced TNF-a expression. Evidently, in addition to suppression of bacterial growth, Man-BSA may be useful for development of the therapeutic strategies against LPS-and LTA-induced inflammation syndrome. Several studies showed that bovine a-lactoglobulin and bovine lactoferrin have antibacterial activity [4–6]. An increase in the net positive charge of bovine a-lactoglobulin and bovine lactoferrin via amidation further increases their bactericidal activity [7–9]. Native BSA, however, does not show any anti-bacterial activity (before the mannosylation of carboxyl groups). Thus, the results of the present study point to a possible new method for preparation of novel antibacterial proteins. In particular, Hoek et al. [33] and Haney et al. [34] found that antibacterial peptides can be isolated from a proteolytic digest of lactoferrin. Accordingly, further research is needs
to test whether the peptide fragment(s) of Man-BSA have bactericidal properties. Our previous results revealed that glucosylated BSA (Glu-BSA) prepared from conjugation of BSA with p-aminophenyl a-D-glucopyranoside also shows membrane-damaging activity [11]. To test whether the bactericidal activity of Man-BSA was due to the conjugated p-aminophenyl a-D-mannopyranoside or blocking of the negatively charged carboxylate groups, the bactericidal activity of Glu-BSA was also investigated. As shown in Fig. 10, Glu-BSA showed dose-dependently antibacterial activity toward E. coli and S. aureus. These observations reveal that blocking of negatively charged carboxylate groups with monosaccharide generates glycated BSA with bactericidal activity. In summary, the present study shows that Man-BSA damages membrane-mimicking liposomes in E. coli and S. aureus and has bactericidal effects on these bacteria. Our data suggest that the Man-BSA’s bactericidal effects on E. coli and S. aureus are closely associated with the membrane-damaging activity. Together with the finding that Glu-BSA shows anti-bacterial activity, our data suggest the notion that novel antimicrobial proteins could be prepared from nonbactericidal proteins after modification of carboxyl groups. Because of the incessant appearance of antibiotic-resistant microbes, new antibacterial peptides from natural sources attracted much attention in recent years [3,35]. The bactericidal action of Man-BSA and Glu-BSA proven in this study suggests that the glycated BSA may serve as a prototype for the development of anti-infective agents targeting E. coli and S. aureus. Acknowledgment This work was supported by Grant MOST 103-2320-B110-002MY2 from the Ministry of Science and Technology, Taiwan, ROC (to L.S. Chang). References [1] [2] [3] [4] [5] [6] [7]
M. Zasloff, Nature 415 (2002) 389–395. A. Izadpanah, R.L. Gallo, J. Am. Acad. Dermatol. 52 (2005) 381–390. K.A. Brogden, Nat. Rev. Microbiol. 3 (2005) 238–250. N. Orsi, Biometals 17 (2004) 189–196. H. Jenssen, R.E. Hancock, Biochimie 91 (2009) 19–29. L. Chaneton, J. Dairy Sci. 94 (2011) 138–145. Y. Pan, J. Wan, H. Roginski, A. Lee, B. Shiell, W.P. Michalski, M.J. Coventry, Lett. Appl. Microbiol. 44 (2007) 229–234. [8] Y. Pan, B. Shiell, J. Wan, M.J. Coventry, W.P. Michalski, A. Lee, H. Roginski, Int. Dairy J. 17 (2007) 1450–1459.
22
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
[9] Y. Pan, B. Shiell, J. Wan, M.J. Coventry, H. Roginski, A. Lee, W.P. Michalski, Int. Dairy J. 17 (2007) 606–616. [10] M.S. Abreu, L.M. Estronca, M.J. Moreno, W.L. Vaz, Biophys. J. 84 (2003) 386– 399. [11] S.Y. Yang, Y.J. Chen, P.S. Kao, L.S. Chang, Arch. Biochem. Biophys. 564 (2014) 43–51. [12] L.W. Chen, P.H. Kao, Y.S. Fu, S.R. Lin, L.S. Chang, Toxicon 58 (2011) 46–53. [13] M. Torrent, B.G. de la Torre, V.M. Nogues, D. Andreu, E. Boix, Biochem. J. 421 (2009) 425–434. [14] D.J. Scheffers, M.G. Pinho, Microbiol. Mol. Biol. Rev. 69 (2005) 585–607. [15] M. Vaara, Microbiol. Rev. 56 (1992) 395–411. [16] H.J. Marvin, M.B. ter Beest, B. Witholt, J. Bacteriol. 171 (1989) 5262–5267. [17] V. Laganas, J. Alder, J.A. Silverman, Antimicrob. Agents Chemother. 47 (2003) 2682–2684. [18] R.F. Epand, B.P. Mowery, S.E. Lee, S.S. Stahl, R.I. Lehrer, S.H. Gellman, R.M. Epand, J. Mol. Biol. 379 (2008) 38–50. [19] M.I. Gurr, T.L. Harwood, K.N. Frayn, Lipid Biochemistry, 5th ed., Blackwell Science, Oxford, 2002. [20] Q.P. Lin, L.F. Zhou, N.N. Li, Y.Q. Chen, B.C. Li, Y.F. Cai, S.Q. Zhang, Eur. J. Pharmacol. 596 (2008) 160–165. [21] K.L. Brown, G.F. Poon, D. Birkenhead, O.M. Pena, R. Falsafi, C. Dahlgren, A. Karlsson, J. Bylund, R.E. Hancock, P. Johnson, J. Immunol. 186 (2011) 5497– 5505. [22] Y. Ruan, T. Shen, Y. Wang, M. Hou, J. Li, T. Sun, Int. Immunopharmacol. 15 (2013) 575–580.
[23] K.L. Piers, M.H. Brown, R.E. Hancock, Antimicrob. Agents Chemother. 38 (1994) 2311–2316. [24] E. Isogai, M. Hirata, H. Isogai, K. Matuo, K. Kimura, K. Yokota, K. Oguma, M. Tojo, F. Kaneko, S. Kotake, S. Ohno, Adv. Exp. Med. Biol. 528 (2003) 195–200. [25] T. Koprivnjak, C. Weidenmaier, A. Peschel, J.P. Weiss, Infect. Immun. 76 (2008) 2169–2176. [26] S. Arcidiacono, J.W. Soares, A.M. Meehan, P. Marek, R. Kirby, J. Pept. Sci. 15 (2009) 398–403. [27] G. van den Bogaart, J. Mika, V. Krasnikov, B. Poolman, Biophys. J. 93 (2007) 154–163. [28] Y.L. Chiou, Y.J. Chen, S.R. Lin, L.S. Chang, Toxicon 58 (2011) 518–526. [29] P.S. Kao, S.R. Lin, W.P. Hu, L.S. Chang, Toxicon 60 (2012) 367–377. [30] J.A. Mitchell, M.J. Paul-Clark, G.W. Clarke, S.K. McMaster, N. Cartwright, J. Endocrinol. 193 (2007) 323–330. [31] M.C. Dessing, M. Schouten, C. Draing, M. Levi, S. von Aulock, T. van der Poll, J. Infect. Dis. 197 (2008) 245–252. [32] S. Knapp, S. von Aulock, M. Leendertse, I. Haslinger, C. Draing, D.T. Golenbock, T. van der Poll, J. Immunol. 180 (2008) 3478–3484. [33] K.S. Hoek, J.M. Milne, P.A. Grieve, D.A. Dionysius, R. Smith, Antimicrob. Agents Chemother. 41 (1997) 54–59. [34] E.F. Haney, K. Nazmi, F. Lau, J.G. Bolscher, H.J. Vogel, Biochimie 91 (2009) 141– 154. [35] Y.J. Gordon, E.G. Romanowski, A.M. McDermott, Curr. Eye Res. 30 (2005) 505– 515.
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi
Antibacterial and membrane-damaging activities of mannosylated bovine serum albumin Chia-Yu Tsai a, Ying-Jung Chen a, Yaw-Syan Fu b, Long-Sen Chang a,c,⇑ a
Institute of Biomedical Sciences, National Sun Yat-Sen University, Kaohsiung 804, Taiwan Department of Biomedical Science and Environmental Biology, Kaohsiung Medical University, Kaohsiung 807, Taiwan c Department of Biotechnology, Kaohsiung Medical University, Kaohsiung 807, Taiwan b
a r t i c l e
i n f o
Article history: Received 31 January 2015 and in revised form 25 February 2015 Available online 6 March 2015 Keywords: Mannosylated BSA Bactericidal effect Membrane-damaging activity Fusogenicity
a b s t r a c t The aim of this study was to test whether mannosylated BSA (Man-BSA) exerts antibacterial activity on Escherichia coli (gram-negative bacteria) and Staphylococcus aureus (gram-positive bacteria) via its membrane-damaging effect. Man-BSA caused inhibition of growth of E. coli and S. aureus. Moreover, bactericidal action of Man-BSA on E. coli and S. aureus positively correlated with the increase in membrane permeability of the bacterial cells. Morphological examination showed that Man-BSA disrupted bacterial membrane integrity. Destabilization of the lipopolysaccharide (LPS) layer and inhibition of lipoteichoic acid (LTA) biosynthesis in the cell wall increased the bactericidal effect of Man-BSA on E. coli and S. aureus. Man-BSA also induced leakage and fusion of membrane-mimicking liposomes in E. coli and S. aureus. Man-BSA showed similar binding affinity for LPS and LTA. LPS and LTA strongly suppressed the membrane-damaging activity of Man-BSA, whereas an increase in the Man-BSA concentration attenuated the inhibitory action of LPS and LTA. Taken together, our data indicate that Man-BSA’s bactericidal activity depends strongly on its ability to induce membrane permeability. Moreover, the bactericidal action of Man-BSA proven in this study suggests that Man-BSA may serve as a prototype for the development of anti-infective agents targeting E. coli and S. aureus. Ó 2015 Elsevier Inc. All rights reserved.
Introduction The development of resistance of bacteria to antibiotics is a global problem that underscores the need for new therapeutic agents. Antimicrobial peptides and proteins are currently under consideration as a potential alternative to conventional antibiotics, on account of their widespread occurrence in nature [1,2]. Antimicrobial peptides and proteins display a broad spectrum of activities against a wide range of pathogens including bacteria, fungi and enveloped viruses. A common feature of most of these peptides and proteins is that they are cationic and have amphipathic properties. Because the bacterial membrane consists of abundant negatively charged phospholipids, it is believed that most of the antimicrobial peptides and proteins interact with
⇑ Corresponding author at: Institute of Biomedical Sciences, National Sun Yat-Sen University, Kaohsiung 804, Taiwan. Fax: +886 7 5250197. E-mail address: [email protected] (L.-S. Chang). http://dx.doi.org/10.1016/j.abb.2015.02.034 0003-9861/Ó 2015 Elsevier Inc. All rights reserved.
anionic phospholipids and kill microorganisms by permeabilizing the bacterial membrane, by thinning the membrane or by destabilizing the membrane structure [3]. Nevertheless, antimicrobial peptides and proteins may kill bacteria by inhibiting macromolecular biosynthesis and/or by interacting with specific vital components inside the bacterial cells [3]. Given that membrane composition of bacteria includes abundantly anionic phospholipids, an increase in the positive charge of the antibiotic proteins via blocking of negatively charged carboxylate groups may enhance the interaction of such proteins with the bacterial membrane and thus enhance the potency of the antibacterial effect. Several studies revealed that bovine a-lactoglobulin and bovine lactoferrin have antibacterial properties [4–6]. Pan et al. [7–9] found that amidation of bovine a-lactoglobulin and bovine lactoferrin increases the net positive charge of these proteins, sharply increasing their bactericidal activity. Nevertheless, there are no studies exploring the possibility that novel antimicrobial proteins could be prepared from nonbactericidal proteins after modification of carboxyl groups. Because antimicrobial proteins usually exert their activity by damaging the bacterial membrane, proteins that preferably interact with phospholipids but do not
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
show bactericidal activity can be a good candidate for testing of the above possibility. Various studies revealed that bovine serum albumin (BSA)1 can transfer a lipid amphiphile to lipid bilayer membranes [10], meaning that BSA can interact with lipid bilayers. Moreover, mannosylated BSA (Man-BSA), which is prepared by conjugation of BSA’s carboxyl groups with p-aminophenyl-a-mannopyranoside is reported to have membrane-damaging effects on zwitterionic lipid vesicles [11]. Thus, the aim of this study was to characterize the antibacterial activity and the mechanism of ManBSA’s action.
15
measuring the optical density (OD) at 550 nm (OD550nm 0.3 for E. coli and OD550nm 0.5 for S. aureus corresponding to approximately to 108 CFU/ml). The bacteria were pelleted, washed with PBS, and then concentrated to 5 107 CFU/ml in PBS. One hundred microliters of the bacteria suspension were incubated with Man-BSA at 37 °C for indicated time periods. Then the cells were cultured in 1 ml LB medium for E. coli or TSB for S. aureus for 6 h with gently shaking. Then antimicrobial activity of Man-BSA was measured using optical density (OD) at 550 nm. Antimicrobial activity of Man-BSA was calculated as (OD550 after Man-BSA treatment)/(OD550 without Man-BSA treatment) 100%.
Materials and methods Bacterial membrane permeability assessment BSA (Fraction V, fatty acid free, catalog number 775835, purity >99%) was obtained from Roche Applied Science. Calcein, rifampin, propidium iodide (PI), cardiolipin, 1-ethyl-3-(diethylaminopropyl) -carbodiimide hydrochloride (EDC), p-aminophenyl a-D-mannopyranoside, p-aminophenyl a-D-glucopyranoside, rhodamine isothiocyanate, sodium dithionite, egg yolk phosphatidylglycerol (EYPG), egg yolk phosphatidylethanolamine (EYPE), lipopolysaccharide (LPS) 0111:B4 from Escherichia coli and lipoteichoic acid (LTA) from Staphylococcus aureus were purchased from Sigma– Aldrich Inc., and N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)-1,2-dihexadecanoyl-sn-glycerol-3-phosphoethanolamine (NBD-PE) and lissamine rhodamine B 1,2-dihexadecanoly-sn-glycero-3-phosphoethanolamine (Rh-PE) were the products of Molecular Probes. Unless otherwise specified, all other reagents were analytical grade.
Bacterial suspensions (5 107 CFU/ml) were treated with ManBSA or PBS in a total volume of 100 ll under the same conditions used in antimicrobial assay. Subsequently, the bacterial cells (E. coli or S. aureus) were incubated with propidium iodide (PI) for 30 min in the dark at room temperature. Then the bacterial cells were washed with PBS, and resuspended in 1 ml PBS. PI fluorescence values were corrected for background (extracellular) dye fluorescence and expressed as a percentage relative to the value obtained after complete permeabilization of the cells by heating at 100 °C for 5 min. Uptake of PI was measured by Beckman Coulter Paradigm™ Detection Platform with excitation at 535 nm and emission at 595 nm. Competitive replacement of LPS-bound or LTA-bound rhodaminelabeled Man-BSA (Rh-Man-BSA) by unlabeled Man-BSA
Preparation of Man-BSA and glucosylated-BSA (Glu-BSA) Man-BSA
was
prepared
by
coupling
p-aminophenyl
a-D-mannopyranoside to BSA through water-soluble EDC according to the method described in [11]. BSA contains 41 Asp, 58 Glu and 1 C-terminal carboxyl groups. MALDI-TOF analyses showed that, as compared with BSA, Man-BSA showed an increase in the molecular weight by 24,050 Da [11]. Thus, the conjugation of approximately 95 p-aminophenyl a-D-mannopyranoside with the carboxyl groups in Man-BSA was calculated from the increment in the molecular weight. BSA was also modified with p-aminophenyl a-D-glucopyranoside according to the procedure described in [11]. MALDI-TOF analyses revealed that approximately 86 of the 100 carboxyl groups in BSA were conjugated with p-aminophenyl a-D-glucopyranoside. Bacterial strains E. coli (JM109) and S. aureus (ATCC 25923) were used in this study. E. coli were maintained on Luria–Bertani (LB) agar plate at 37 °C, and S. aureus were maintained on tryptic soy agar plate at 37 °C. Antimicrobial assay E. coli (JM109) were grown in LB medium and S. aureus were grown in tryptic soy broth (TSB) from a single colony with agitation at 37 °C. Bacterial number was then evaluated by 1 Abbreviations used: BSA, bovine serum albumin; Man-BSA, mannosylated BSA; LPS, lipopolysaccharide; LTA, lipoteichoic acid; PI, propidium iodide; EDC, 1-ethyl-3(diethylaminopropyl)-carbodiimide hydrochloride; EYPG, egg yolk phosphatidylglycerol; EYPE, egg yolk phosphatidylethanolamine; NBD-PE, N-(7-nitrobenz-2-oxa-1,3diazol-4-yl)-1,2-dihexadecanoyl-sn-glycerol-3-phosphoethanolamine; Rh-PE, lissamine rhodamine B 1,2-dihexadecanoly-sn-glycero-3-phosphoethanolamine; LB, Luria–Bertani; TSB, tryptic soy broth; OD, optical density; GAPDH, glyceraldehyde3-phosphate dehydrogenase; PE, phosphatidylethanolamine; PG, phosphatidylglycerol.
Modification of Man-BSA with rhodamine isothiocyanate was prepared according to the procedure described in [12]. Rh-ManBSA (1 lM) was titrated with increasing concentrations of LPS or LTA until maximal changes in fluorescence intensity of Rh-ManBSA was achieved. Then increasing concentrations of unlabeled Man-BSA were added to compete for binding of Rh-Man-BSA with LPS or LTA. Competitive binding was monitored at excitation wavelength and emission wavelength at 550 and 580 nm, respectively. Release of entrapped fluorescent marker from liposomes Membrane-damaging activity of Man-BSA was determined by measuring the release of the liposome-entrapped, self-quenching fluorescent dye calcein. EYPE/EYPG (75/25, mol/mol) or EYPG/cardiolipin (60/40, mol/mol) were dissolved in chloroform/ methanol (v/v, 2:1) and dried by evaporation. Buffer (10 mM Tris–HCl, 100 mM NaCl, pH 7.5) containing 50 mM calcein was added to the film of lipids, and after hydration the suspension was shaken vigorously. The multilamellar vesicles obtained in this way were extruded 10 times, above the transition temperature, through a 100-nm polycarbonate filter and applied to a Sepharose 6B column (2 15 cm) to separate the liposome from the free calcein. Leakage was induced by adding aliquots of Man-BSA to a vesicle suspension directly in the cuvette used for fluorescence determination at 30 °C. The kinetics of membrane damage were monitored by the increase in fluorescence with emission at 520 nm and excitation at 490 nm, and the signal was expressed as percentage of total calcein release after the addition of 0.2% Triton X-100. Fusion assay of phospholipid vesicles Liposome fusion induced by Man-BSA was measured by increase in fluorescence resonance energy transfer between two lipid probes (NBD-PE and Rh-PE). NBD-PE and Rh-PE were used as donor and acceptor fluorescent lipid, respectively. NBD-PE and
16
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
Rh-PE (1 mol% in phospholipid vesicles) were incorporated into phospholipid vesicles. The NBD emission was monitored at 530 nm with the excitation wavelength set at 465 nm. As the two liposome groups interacted, the fluorescence energy emitted from NBD-PE labeled liposomes was transferred to the Rh-PE labeled liposomes resulting in a decreased fluorescence signal of NBD. The percentage of fusion was defined by the following relationship: %Fusion = 100 (1 F/F0), where F and F0 are the fluorescence intensities in the presence and absence of the RhPE, respectively. For the inner monolayer phospholipid fusion assay, phospholipid vesicles were treated with sodium dithionite to completely reduce the NBD-labeled phospholipid located at the outer monolayer of the membrane. Final concentration of sodium dithionite was 100 mM and the inner leaflet mixing assay was carried out at the level in which 55–60% fluorescence intensity of NBD were reduced. Scanning electron microscopy (SEM) E. coli were grown at 37 °C to mid-exponential phase (OD550 = 0.5–0.6) and incubated with 1.1 lM Man-BSA under the same conditions used in antimicrobial assay. After incubation for 24 h, Man-BSA-treated cells were prepared for SEM analysis according to the procedure described in [13]. The micrographs were viewed at a 5.0 kV accelerating voltage on a Hitachi SU8000 SEM, and a secondary electron image of the cells for topography contrast was collected at several magnifications. Detection of TNF-a expression Human acute myeloid leukemia U937 cells were obtained from ATCC (Manassas, VA, USA). U937 cells were cultured in RPMI 1640 supplemented with 10% fetal calf serum, 1% sodium pyruvate, 2 mM glutamine and penicillin (100 units/ml)/streptomycin (100 lg/ml) in an incubator humidified with 95% air and 5% CO2. Cells were stimulated with indicated LPS or LTA concentrations in the absence or presence of Man-BSA for 24 h. Total RNA was isolated from cells using the RNeasy minikit (QIAGEN Inc., Valencia, CA) according to the instructions of the manufacturer. Reverse transcriptase reaction was performed with 2 lg of total RNA using M-MLV reverse transcriptase (Promega, Madison, WI) according to the manufacturer’s recommendations. A reaction without reverse transcriptase was performed in parallel to ensure the absence of genomic DNA contamination. After initial denaturation at 95 °C for 10 min, PCR amplification was performed using GoTaq Flexi DNA polymerase (Promega, Madison, WI) followed by 35 cycles at 94 °C for 50 s, 65 °C for 50 s, and 72 °C for 50 s. After a final extension at 72 °C for 5 min, PCR products were resolved on 2% agarose gels and visualized by ethidium bromide transillumination under UV light. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control to check the efficiency of cDNA synthesis and PCR amplification. Primer sequences were as follows: TNF-a-forward, 50 -ATGAGCACTGAAA GCATGATCCGG30 and TNF-a-reverse 50 -GCAATGATCCCAAAGTAGACCTGCCC-30 ; GAPDH-forward: 50 -GAGTCAACGGATTTGGTCGT-30 ; GAPDH-reverse: 50 -TGTGGTCATG AGTCCTTCCA-30 , and the reverse transcriptase-PCR products were subsequently confirmed by direct sequencing. Statistical analysis All data are presented as mean ± SD. Significant differences among the groups were determined using the unpaired Student’s t-test. A value of P < 0.05 was taken as an indication of statistical significance.
Results and discussion As shown in Fig. 1A and B, the antibacterial activity of Man-BSA toward E. coli and S. aureus was dose dependent. The half-maximal inhibitory concentration (IC50) of Man-BSA for E. coli and S. aureus was 0.85 ± 0.15 and 0.61 ± 0.11 lM, respectively. This finding indicates that Man-BSA has a stronger bactericidal effect on S. aureus than on E. coli. Moreover, Man-BSA killed E. coli and S. aureus most effectively after treatment for 24 h (Fig. 1C and D). On the other hand, intact BSA did not significantly suppress the growth of E. coli and S. aureus (Fig. 1A and B). The effect of Man-BSA on the permeability of the inner membrane of E. coli and S. aureus was monitored using a PI fluorescent probe. Inner permeability changes evoked by Man-BSA resulted in an increased intracellular concentration of PI, which is detected thanks to the fluorescence resulting from the binding of the probe to nucleic acids present in the cytoplasma. Fluorescence measurement showed that the percentage of permeabilized cells increased after incubation with Man-BSA (Fig. 2), whereas BSA did not increase membrane permeability of E. coli and S. aureus (inset in Fig. 2). The Man-BSA concentration that resulted in 50% of permeabilized E. coli and S. aureus cells was approximately 0.96 and 0.75 lM, respectively. This result also indicated that, compared to E. coli, S. aureus is more sensitive to the antibacterial action of Man-BSA. Furthermore, the antibacterial activity was authenticated by our ultrastructural studies; representative micrographs are shown in Fig. 3. Compared to BSA-treated bacterial cells, Man-BSA-treated E. coli and S. aureus showed pronounced morphological changes consistent with damage to the bacterial cells. The Man-BSA-treated bacteria displayed significant wrinkling, surface roughening and membrane blebbing. This suggests that the ManBSA-treated bacterial cells lost integrity of their membrane. LPS and LTA contain a saccharide moiety and are major constituents of the cell wall of E. coli and S. aureus, respectively [14]. Thus, the effect of LPS and LTA on Man-BSA’s antibacterial activity was examined too. Other studies suggest that LPS-associated cations tightly cross-link the LPS layer whereas EDTA treatment causes the loss of LPS [15,16]. In sharp contrast to Ca2+, EDTA enhanced the bactericidal effect of Man-BSA (Fig. 4A). Laganas et al. [17] showed that rifampin inhibits LTA biosynthesis. Fig. 4B shows that rifampin pretreatment enhanced Man-BSA’s bactericidal effect on S. aureus. Taken together, these findings indicate that LPS and LTA may inhibit Man-BSA’s antibacterial action. To characterize the interaction of Man-BSA with LPS and LTA, we performed replacement of LPS- or LTA-bound Rh-Man-BSA by unlabeled Man-BSA. As shown in Fig. 5, the fluorescence intensity of Rh-Man-BSA was reduced by titration with LPS or LTA, and the maximal reduction in fluorescence intensity of Rh-Man-BSA was noted with 200 lg of LPS or 230 lg of LTA. This result likely means that changes in the conformation of Rh-Man-BSA occurred after binding to LPS or LTA. Gradual restoration of the fluorescence intensity was observed after the addition of unlabeled Man-BSA; this phenomenon is indicative of competitive displacement of Rh-Man-BSA from LPS or LTA. Calculations that were based on the changes in fluorescence intensity showed that the dissociation constants of Man-BSA for LPS and LTA were 0.08 ± 0.03 and 0.09 ± 0.02 lM, respectively. On the other hand, unmodified BSA did not competitively displace Rh-Man-BSA from LPS and LTA (data not shown). Some studies indicated that the E. coli membrane contains phosphatidylethanolamine (PE) and phosphatidylglycerol (PG) at the molar ratio of approximately 75:25, whereas the S. aureus membrane contains PG and cardiolipin at the molar ratio of approximately 60:40 [18]. Accordingly, we tested whether ManBSA induces the leakage of calcein from EYPE/EYPG (75:25,
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
17
Fig. 1. Bactericidal activity of Man-BSA on E. coli and S. aureus. Bactericidal activity of Man-BSA against (A) E. coli and (B) S. aureus showed a concentration-dependent manner. E. coli or S. aureus were treated with indicated Man-BSA concentrations for 24 h. Bactericidal activity of Man-BSA against (C) E. coli and (D) S. aureus showed a time-dependent manner. E. coli and S. aureus were treated with 1.1 lM Man-BSA for indicated time periods. Data are presented as mean ± SD of three independent experiments with triplicate measurements (⁄P < 0.05 at all tested Man-BSA concentrations, bactericidal effect of Man-BSA vs. bactericidal effect of BSA).
Fig. 2. Membrane permeability of E. coli and S. aureus after Man-BSA treatment. Man-BSA induced increase in membrane permeability of E. coli and S. aureus in a concentration-dependent manner. E. coli and S. aureus were treated with indicated Man-BSA concentrations for 24 h. Man-BSA-treated bacteria were incubated with propidium iodide (PI) for 30 min before fluorescence measurement. Maximal membrane permeability signal was obtained by heating E. coli and S. aureus at 100 °C for 5 min. (Inset) Effect of Man-BSA and BSA on membrane permeability of E. coli and S. aureus. E. coli and S. aureus were treated with 2.2 lM Man-BSA or BSA for 24 h. Data are presented as mean ± SD of three independent experiments with triplicate measurements (⁄P < 0.05).
mol/mol) and EYPG/cardiolipin (60:40, mol/mol) vesicles. As shown in Fig. 6, Man-BSA caused the release of calcein from the EYPE/EYPG and EYPG/cardiolipin vesicles in a concentration-
dependent manner. The maximal calcein release from EYPE/EYPG and EYPG/cardiolipin vesicles was observed when the Man-BSA concentration was >0.6 lM. The Man-BSA-induced calcein leakage
18
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
Fig. 3. SEM of Man-BSA-treated E. coli and S. aureus. (A) BSA-treated E. coli; (B) Man-BSA-treated E. coli; (C) BSA-treated S. aureus; (D) Man-BSA-treated S. aureus. E. coli and S. aureus were grown at 37 °C to mid-exponential phase (OD550 = 0.5–0.6) and incubated with 1.1 lM BSA or Man-BSA. After incubation for 24 h, BSA- and Man-BSA-treated cells were prepared for SEM analysis according to the procedure described in Torrent et al. [13].
Fig. 4. Integrity of cell wall reduced antibacterial activity of Man-BSA. (A) Effect of EDTA and Ca2+ on bactericidal activity of Man-BSA against E. coli. E. coli were treated with 1.1 lM Man-BSA in the presence of 2 mM EDTA or CaCl2 for 24 h. (B) Inhibition of LTA biosynthesis enhanced bactericidal activity of Man-BSA. S. aureus were incubated with 1.56 lg/ml rifampin at 37 °C for 1 h before treatment with 1.1 lM Man-BSA at 37 °C for 24 h. Data are presented as mean ± SD of three independent experiments with triplicate measurements (⁄P < 0.05).
from EYPG/cardiolipin vesicles was noticeably stronger than that from EYPE/EYPG vesicles. In contrast, BSA did not cause the leakage of EYPE/EYPG and EYPG/cardiolipin vesicles (inset in Fig. 6). Because LPS and LTA with intact structure reduced the antibacterial activity of Man-BSA (Fig. 4), we assessed the effect of LPS and LTA on the membrane-damaging activity of Man-BSA. As shown in Fig. 7, LPS and LTA suppressed the membrane-damaging activity of Man-BSA in a concentration-dependent manner. On the other hand, an increase in Man-BSA concentration abrogated the inhibitory action of LPS and LTA. Given that Man-BSA is able to induce liposome fusion and membrane leakage [11], the fusogenicity of Man-BSA was then examined. Fig. 8 shows that a lipid mixing assay involving a fluorescence energy transfer between NBD-PE and Rh-PE was used to study the fusion event induced by Man-BSA. The fluorescence intensity of NBD-PE did not significantly decrease when the NBD-PE-labeled
vesicles were mixed with Rh-PE-labeled vesicles without the addition of Man-BSA. Fig. 8 shows that Man-BSA-induced fusion of EYPG/cardiolipin and EYPE/EYPG vesicles, and the fusogenicity of Man-BSA in relation to EYPG/cardiolipin vesicles was higher compared to that of EYPE/EYPG vesicles. On the other hand, BSA did not have a fusogenic effect on EYPE/EYPG and EYPG/cardiolipin vesicles (inset in Fig. 8). Under controlled conditions, sodium dithionite reduced the NBD attached to the lipid headgroup in the outer leaflet to a nonfluorescent derivative while leaving the NBD in the inner leaflet largely unaffected. When Man-BSA was added to the dithionite-treated EYPG/cardiolipin and EYPE/EYPG vesicles, a reduction in NBD fluorescence signal was observed, pointing to inner-leaflet admixing. Given that merging of the two inner leaflets should be concomitant with formation of the fusion pore [19], the inner leaflet admixing is expected to cause content mixing and/or leakage of vesicular content.
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
19
Fig. 5. Replacement of LPS- or LTA-bound Rh-Man-BSA by Man-BSA. Rh-Man-BSA (1 lM) dissolved in 10 mM Tris–HCl-100 mM NaCl (pH 7.5) were incubated with 200 lg LPS (A) or 240 lg LTA (B) for 1 h, and increasing concentrations of unlabeled Man-BSA were then added. Maximal reduction in fluorescence intensity of Rh-Man-BSA was noted when 200 lg LPS or 240 lg LTA was added. Replacement of LPS-bound or LTA-bound Rh-Man-BSA by unlabeled Man-BSA led to a restoration in fluorescence intensity of Rh-Man-BSA. The addition of unlabeled Man-BSA into the solution of Rh-Man-BSA did not change the fluorescence intensity of Rh-Man-BSA in the absence of LPS or LTA. Moreover, the fluorescence intensity of Rh-Man-BSA-LPS or Rh-Man-BSA-LTA complexes was not significantly restored by titration with buffer (10 mM Tris–HCl-100 mM NaCl, pH 7.5). Data are presented as mean ± SD of three independent experiments.
Fig. 6. Membrane-damaging activity of Man-BSA. Man-BSA induced calcein release from (A) EYPE/EYPG and (B) EYPG/cardiolipin vesicles in a concentration-dependent manner. The experiments were performed in 10 mM Tris–HCl–100 mM NaCl (pH 7.5). The final concentration of lipids was 2.5 lM. The signal was expressed as the percentage of total calcein release after the addition of 0.2% Triton X-100. (Inset) Effect of BSA on membrane permeability of EYPE/EYPG and EYPG/cardiolipin vesicles. The used BSA and Man-BSA concentrations were 0.66 lM. Data are presented as mean ± SD of three independent experiments with triplicate measurements (⁄P < 0.05).
Various studies revealed that an antibacterial peptide can abrogate LPS- and LTA-induced TNF-a expression [20–22]. Thus, we next analyzed the effect of Man-BSA on LPS- and LTA-induced TNF-a expression. As shown in Fig. 9A, LPS and LTA dose-dependently enhanced TNF-a expression in U937 cells: 10 lg/ml LPS and 0.1 lg/ml LTA notably increased TNF-a expression. Fig. 9B shows that cotreatment with Man-BSA actively inhibited the TNF-a expression in the LPS- and LTA-treated U937 cells, whereas BSA had no effect. Other studies showed that binding to LPS and LTA may enhance self-promoted uptake of antimicrobial peptides and proteins and thereby enhance their membrane-permeabilizing activity [23– 26]. Our data revealed that the binding of Man-BSA to LPS and LTA attenuates this protein’s bactericidal activity. Accordingly, LPS and LTA do not function as a polyanionic ladder for traversing Man-BSA from the outside through the cytoplasmic membrane. Together with the findings that Man-BSA induces permeabilization of the inner membrane (according to the PI fluorescence experiments, Fig. 2), the above result suggests that the bactericidal effect
of Man-BSA is directly linked to its membrane-permeabilizing activity. The finding that Man-BSA induced the leakage of bacterial-membrane-mimicking vesicles supports this notion. Moreover, Man-BSA induced the fusion of bacterial-membranemimicking vesicles in our experiments. Because the peptide-induced vesicle-vesicle fusion causes simultaneous leakage of vesicular contents [27,28], our data suggest that the membranedamaging effects of Man-BSA on EYPE/EYPG and EYPG/cardiolipin are mediated by fusogenic events. Furthermore, destabilization of LPS structure enhanced the bactericidal activity of Man-BSA (Fig. 4), whereas LPS suppressed Man-BSA’s membrane-damaging activity (Fig. 8). Likewise, LTA abrogated the membrane-damaging activity of Man-BSA, whereas inhibition of LTA synthesis enhanced Man-BSA’s bactericidal effect on S. aureus (Fig. 4). It is evident that binding to LPS and LTA reduces Man-BSA’s bactericidal activity. Nevertheless, increased Man-BSA concentration can attenuate the suppressive effect of LPS and LTA on the Man-BSA-induced membrane permeability. On the other hand, Man-BSA-induced calcein leakage from EYPG/cardiolipin vesicles
20
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
Fig. 7. Effect of LPS and LTA on membrane-damaging activity of Man-BSA. (A) Inhibitory effect of LPS on membrane-damaging activity of Man-BSA toward EYPE/EYPG vesicles. (B) Inhibitory effect of LTA on membrane-damaging activity of Man-BSA toward EYPG/cardiolipin. Man-BSA (0.66 lM) was incubated with indicated concentrations of LPS or LTA for 1 h, and then membrane-damaging activity of Man-BSA-LPS complexes or Man-BSA-LTA complexes was measured. The experiments were performed in 10 mM Tris–HCl–100 mM NaCl (pH 7.5). The signal was expressed as the percentage of total calcein release after addition of 0.2% Triton X-100. Data are presented as mean ± SD of six independent experiments.
Fig. 8. Man-BSA induced fusion of EYPE/EYPG and EYPG/cardiolipin vesicles. The experiments were performed in 10 mM Tris–HCl–100 mM NaCl (pH 7.5). Equal concentrations of NBD-PE-labeled and Rh-PE-labeled phospholipid vesicles were used. The final concentration of lipids was 200 lM. Man-BSA induced fusion of (A) EYPE/ EYPG and (B) EYPG/cardiolipin vesicles in a concentration-dependent manner. (Inset) Fusogenicity of Man-BSA and BSA on EYPE/EYPG and EYPG/cardiolipin vesicles. The used BSA and Man-BSA concentrations were 0.55 lM. Data are presented as mean ± SD of three independent experiments with triplicate measurements (⁄P < 0.05).
Fig. 9. Effect of Man-BSA on LPS- and LTA-induced TNF-a mRNA expression in U937 cells. (A) LPS (left panel) and LTA (right panel) dose-dependently induced up-regulation of TNF-a mRNA expression in U937 cells. (B) Man-BSA repressed LPS- and LTA-induced TNF-a mRNA expression.
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
21
Fig. 10. Bactericidal activity of Glu-BSA on E. coli and S. aureus. Bactericidal activity of Glu-BSA against (A) E. coli and (B) S. aureus showed a concentration-dependent manner. E. coli or S. aureus were treated with indicated Glu-BSA concentrations for 24 h. Data are presented as mean ± SD of three independent experiments with triplicate measurements.
was higher than that from EYPE/EYPG vesicles (Fig. 6). Man-BSAinduced fusion of EYPG/cardiolipin vesicles was also more pronounced than that of EYPE/EYPG vesicles (Fig. 8). EYPG and cardiolipin are anionic phospholipids, whereas EYPE is a zwitterionic phospholipid. Other studies showed that Naja naja atra CTX3 has a stronger membrane-damaging activity toward EYPG/cardiolipin vesicles than toward EYPE/EYPG vesicles [29]. Moreover, it was reported that CTX3 assumes a different membrane-bound state upon adsorption onto EYPE/EYPG and EYPG/cardiolipin vesicles [29]. Taken together, these observations suggest that the membrane-interacting mode of Man-BSA is related to the difference in Man-BSA’s damaging effects on EYPE/EYPG and EYPG/cardiolipin vesicles. This notion is supported by the finding that Man-BSA assumes different conformations after binding to lipid vesicles of different lipid composition [11]. LPS and LTA can be released from the cell wall when bacteria are killed by host immune cells or an antibiotic drug. LPS is an endotoxin and plays a key role in the pathophysiology of inflammation, sepsis and shock [30]. Like LPS, LTA also elicits an inflammatory reaction [31]. LPS and LTA induce the inflammatory response via binding to Toll-like receptor 4 (TLR4) and TLR2/6, respectively [30–32]. Several studies revealed that an antimicrobial peptide can suppress LPS- or LTA-induced TNF-a production in human monocytes or in macrophage cell lines [20–22]. The results of these studies suggest that the interaction of antimicrobial peptides with LPS and LTA impedes the binding of LPS and LTA to cell surface receptors, thereby attenuating the LPS- and LTA-induced TNF-a expression [20–22]. Because Man-BSA shows binding affinity for LPS and LTA (Fig. 5), a similar mechanism may explain the suppressive effect of Man-BSA on the LTA- and LPS-induced TNF-a expression. Evidently, in addition to suppression of bacterial growth, Man-BSA may be useful for development of the therapeutic strategies against LPS-and LTA-induced inflammation syndrome. Several studies showed that bovine a-lactoglobulin and bovine lactoferrin have antibacterial activity [4–6]. An increase in the net positive charge of bovine a-lactoglobulin and bovine lactoferrin via amidation further increases their bactericidal activity [7–9]. Native BSA, however, does not show any anti-bacterial activity (before the mannosylation of carboxyl groups). Thus, the results of the present study point to a possible new method for preparation of novel antibacterial proteins. In particular, Hoek et al. [33] and Haney et al. [34] found that antibacterial peptides can be isolated from a proteolytic digest of lactoferrin. Accordingly, further research is needs
to test whether the peptide fragment(s) of Man-BSA have bactericidal properties. Our previous results revealed that glucosylated BSA (Glu-BSA) prepared from conjugation of BSA with p-aminophenyl a-D-glucopyranoside also shows membrane-damaging activity [11]. To test whether the bactericidal activity of Man-BSA was due to the conjugated p-aminophenyl a-D-mannopyranoside or blocking of the negatively charged carboxylate groups, the bactericidal activity of Glu-BSA was also investigated. As shown in Fig. 10, Glu-BSA showed dose-dependently antibacterial activity toward E. coli and S. aureus. These observations reveal that blocking of negatively charged carboxylate groups with monosaccharide generates glycated BSA with bactericidal activity. In summary, the present study shows that Man-BSA damages membrane-mimicking liposomes in E. coli and S. aureus and has bactericidal effects on these bacteria. Our data suggest that the Man-BSA’s bactericidal effects on E. coli and S. aureus are closely associated with the membrane-damaging activity. Together with the finding that Glu-BSA shows anti-bacterial activity, our data suggest the notion that novel antimicrobial proteins could be prepared from nonbactericidal proteins after modification of carboxyl groups. Because of the incessant appearance of antibiotic-resistant microbes, new antibacterial peptides from natural sources attracted much attention in recent years [3,35]. The bactericidal action of Man-BSA and Glu-BSA proven in this study suggests that the glycated BSA may serve as a prototype for the development of anti-infective agents targeting E. coli and S. aureus. Acknowledgment This work was supported by Grant MOST 103-2320-B110-002MY2 from the Ministry of Science and Technology, Taiwan, ROC (to L.S. Chang). References [1] [2] [3] [4] [5] [6] [7]
M. Zasloff, Nature 415 (2002) 389–395. A. Izadpanah, R.L. Gallo, J. Am. Acad. Dermatol. 52 (2005) 381–390. K.A. Brogden, Nat. Rev. Microbiol. 3 (2005) 238–250. N. Orsi, Biometals 17 (2004) 189–196. H. Jenssen, R.E. Hancock, Biochimie 91 (2009) 19–29. L. Chaneton, J. Dairy Sci. 94 (2011) 138–145. Y. Pan, J. Wan, H. Roginski, A. Lee, B. Shiell, W.P. Michalski, M.J. Coventry, Lett. Appl. Microbiol. 44 (2007) 229–234. [8] Y. Pan, B. Shiell, J. Wan, M.J. Coventry, W.P. Michalski, A. Lee, H. Roginski, Int. Dairy J. 17 (2007) 1450–1459.
22
C.-Y. Tsai et al. / Archives of Biochemistry and Biophysics 573 (2015) 14–22
[9] Y. Pan, B. Shiell, J. Wan, M.J. Coventry, H. Roginski, A. Lee, W.P. Michalski, Int. Dairy J. 17 (2007) 606–616. [10] M.S. Abreu, L.M. Estronca, M.J. Moreno, W.L. Vaz, Biophys. J. 84 (2003) 386– 399. [11] S.Y. Yang, Y.J. Chen, P.S. Kao, L.S. Chang, Arch. Biochem. Biophys. 564 (2014) 43–51. [12] L.W. Chen, P.H. Kao, Y.S. Fu, S.R. Lin, L.S. Chang, Toxicon 58 (2011) 46–53. [13] M. Torrent, B.G. de la Torre, V.M. Nogues, D. Andreu, E. Boix, Biochem. J. 421 (2009) 425–434. [14] D.J. Scheffers, M.G. Pinho, Microbiol. Mol. Biol. Rev. 69 (2005) 585–607. [15] M. Vaara, Microbiol. Rev. 56 (1992) 395–411. [16] H.J. Marvin, M.B. ter Beest, B. Witholt, J. Bacteriol. 171 (1989) 5262–5267. [17] V. Laganas, J. Alder, J.A. Silverman, Antimicrob. Agents Chemother. 47 (2003) 2682–2684. [18] R.F. Epand, B.P. Mowery, S.E. Lee, S.S. Stahl, R.I. Lehrer, S.H. Gellman, R.M. Epand, J. Mol. Biol. 379 (2008) 38–50. [19] M.I. Gurr, T.L. Harwood, K.N. Frayn, Lipid Biochemistry, 5th ed., Blackwell Science, Oxford, 2002. [20] Q.P. Lin, L.F. Zhou, N.N. Li, Y.Q. Chen, B.C. Li, Y.F. Cai, S.Q. Zhang, Eur. J. Pharmacol. 596 (2008) 160–165. [21] K.L. Brown, G.F. Poon, D. Birkenhead, O.M. Pena, R. Falsafi, C. Dahlgren, A. Karlsson, J. Bylund, R.E. Hancock, P. Johnson, J. Immunol. 186 (2011) 5497– 5505. [22] Y. Ruan, T. Shen, Y. Wang, M. Hou, J. Li, T. Sun, Int. Immunopharmacol. 15 (2013) 575–580.
[23] K.L. Piers, M.H. Brown, R.E. Hancock, Antimicrob. Agents Chemother. 38 (1994) 2311–2316. [24] E. Isogai, M. Hirata, H. Isogai, K. Matuo, K. Kimura, K. Yokota, K. Oguma, M. Tojo, F. Kaneko, S. Kotake, S. Ohno, Adv. Exp. Med. Biol. 528 (2003) 195–200. [25] T. Koprivnjak, C. Weidenmaier, A. Peschel, J.P. Weiss, Infect. Immun. 76 (2008) 2169–2176. [26] S. Arcidiacono, J.W. Soares, A.M. Meehan, P. Marek, R. Kirby, J. Pept. Sci. 15 (2009) 398–403. [27] G. van den Bogaart, J. Mika, V. Krasnikov, B. Poolman, Biophys. J. 93 (2007) 154–163. [28] Y.L. Chiou, Y.J. Chen, S.R. Lin, L.S. Chang, Toxicon 58 (2011) 518–526. [29] P.S. Kao, S.R. Lin, W.P. Hu, L.S. Chang, Toxicon 60 (2012) 367–377. [30] J.A. Mitchell, M.J. Paul-Clark, G.W. Clarke, S.K. McMaster, N. Cartwright, J. Endocrinol. 193 (2007) 323–330. [31] M.C. Dessing, M. Schouten, C. Draing, M. Levi, S. von Aulock, T. van der Poll, J. Infect. Dis. 197 (2008) 245–252. [32] S. Knapp, S. von Aulock, M. Leendertse, I. Haslinger, C. Draing, D.T. Golenbock, T. van der Poll, J. Immunol. 180 (2008) 3478–3484. [33] K.S. Hoek, J.M. Milne, P.A. Grieve, D.A. Dionysius, R. Smith, Antimicrob. Agents Chemother. 41 (1997) 54–59. [34] E.F. Haney, K. Nazmi, F. Lau, J.G. Bolscher, H.J. Vogel, Biochimie 91 (2009) 141– 154. [35] Y.J. Gordon, E.G. Romanowski, A.M. McDermott, Curr. Eye Res. 30 (2005) 505– 515.
